UNIVERSIDAD NACIONAL 
AUTÓNOMA DE MÉXICO 
 
 
FACULTAD DE QUÍMICA 
 
PROGRAMA DE MAESTRÍA Y DOCTORADO 
EN CIENCIAS BIOQUÍMICAS 
 
 
 
CONTROL DE LA GLUCÓLISIS EN CÉLULAS 
TUMORALES DE RÁPIDO CRECIMIENTO 
 
 
 
T  E  S  I  S 
 
QUE PARA OBTENER EL GRADO DE: 
 
DOCTOR EN CIENCIAS (BIOQUÍMICAS) 
 
P      R      E      S      E      N      T      A: 
 
M. EN C. ALVARO MARÍN HERNÁNDEZ 
 
 
 
 
Tutor: DR. RAFAEL MORENO SÁNCHEZ 
 
 
 
 
MÉXICO, D. F.     Febrero 2011 
 
 
 
UNAM – Dirección General de Bibliotecas 
Tesis Digitales 
Restricciones de uso 
  
DERECHOS RESERVADOS © 
PROHIBIDA SU REPRODUCCIÓN TOTAL O PARCIAL 
  
Todo el material contenido en esta tesis esta protegido por la Ley Federal 
del Derecho de Autor (LFDA) de los Estados Unidos Mexicanos (México). 
El uso de imágenes, fragmentos de videos, y demás material que sea 
objeto de protección de los derechos de autor, será exclusivamente para 
fines educativos e informativos y deberá citar la fuente donde la obtuvo 
mencionando el autor o autores. Cualquier uso distinto como el lucro, 
reproducción, edición o modificación, será perseguido y sancionado por el 
respectivo titular de los Derechos de Autor. 
 
  
 
 
 
CONTROL DE LA GLUCÓLISIS EN CÉLULAS TUMORALES DE 
RÁPIDO CRECIMIENTO  
 
 
 
 
RECONOCIMIENTOS 
 
 
Esta tesis doctoral se realizó bajo la dirección del Dr. Rafael Moreno Sánchez en el Departamento 
de Bioquímica del Instituto Nacional de Cardiología “Ignacio Chávez”. 
 
El Comité Tutoral que asesoró el desarrollo de esta tesis estuvo formado por: 
Dra. Irma Ofelia Bernal Lugo                             Facultad de Química, UNAM 
Dr. Jesús Adolfo García Sáinz                         Instituto de Fisiología Celular, UNAM 
Dr. Rafael Moreno Sánchez                                             Instituto Nacional de Cardiología 
 
Se reconoce la colaboración de la Dra. Emma Cecilia Saavedra  Lira, del departamento de 
Bioquímica del Instituto Nacional de Cardiología, en el desarrollo de los modelos cinéticos. 
 
Se reconoce la colaboración de la Dra. Marina Macías Silva, del Instituto de Fisiología Celular, 
quien nos proporcionó hepatocitos de rata. 
 
Se reconoce la colaboración del Dr. Alejandro Zentella Dehesa, del Instituto de Investigaciones 
Biomédicas, quien nos proporcionó células endoteliales de cordón umbilical. 
 
 
El proyecto fue apoyado parcialmente por CONACYT (43811-Q, 80534, 123636). Durante los 
estudios de doctorado gocé de una beca otorgada por CONACYT para la realización de la presente 
tesis. 
 
  
El Jurado de Examen Doctoral estuvo constituido por: 
Presidente Dr. Armando Gómez Puyou             Instituto de Fisiología Celular, UNAM 
Vocal  Dr. Jesús Adolfo García Sáinz                     Instituto de Fisiología Celular, UNAM 
Secretario Dr. Alejandro Zentella Dehesa                    Instituto de Inv. Biomédicas, UNAM 
Suplente  Dr. Juan Pablo Pardo Vázquez  Facultad de Medicina, UNAM 
Suplente Dr. Marco Antonio Cerbón Cervantes Facultad de Química, UNAM 
 
 
 
DEDICATORIAS 
 
 
 
 
 
 
 
A mi esposa, una mujer incansable que me apoya incondicionalmente y de la cual 
estoy profundamente enamorado. 
 
A mis hijos, esas personitas a las que amo. 
 
A mis padres, por darme la vida y por inculcarme el sentido de la responsabilidad y de 
la honradez. 
 
A mi hermano por su apoyo. 
 
A las familias: Marín y Robles   
 
A todos mis amigos por su amistad. 
 
 
 
 
 
 
 
 
 
 
AGRADECIMIENTOS 
 
 
 
 
 
 
Quiero expresar mi agradecimiento: 
 
 
 
A mi tutor el Dr. Rafael Moreno Sánchez por apoyar  mi formación, por su confianza 
y por  haberme permitido ser parte de su grupo. 
 
A cada una de las personas que forman parte del Departamento de Bioquímica. 
 
A mis alumnas, por la confianza que depositaron en mí. 
 
Al Instituto Nacional de Cardiología “  Ignacio Chávez”.  
 
Al  CONACYT por el financiamiento del proyecto y por la beca que me otorgo.    
 
    
 
 
 
 
 
 
 
“La alegría de ver y entender es el 
 más perfecto don de la naturaleza”. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
“La mayoría de las ideas fundamentales de la ciencia 
 son esencialmente sencillas y, por regla general pueden 
 ser expresadas en un lenguaje comprensible para todos”. 
                                                                         Albert Einstein  
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
“El principio de la investigación  
necesita más cabezas que medios”.  
Severo Ochoa    
 
 
 
 
INDICE página 
 
Capítulo 1    La glucólisis en el cáncer 
 
 
1.1 Introducción       1 
1.2 La glucólisis             3 
1.3  La glucólisis tumoral 3 
1.4 ¿Qué ventajas ofrece a las células tumorales  
el incremento de la glucólisis?               
 
6 
 
Capítulo 2     Mecanismos que inducen el incremento de la glucólisis            
 
 
2.1 El factor inducido por la hipoxia (HIF-1α)  
(Manuscrito No. 1: Mini Rev. Med. Chem. 2009, 9:1084-1101).    
 
8 
2.1.1 Comentarios sobre el manuscrito No. 1 31 
2.2 Genes supresores de tumores y oncogenes 32 
2.2.1 PTEN                                            32 
2.2.2 p53                                                                                       33 
2.2.3 c-Myc                                                                                                  33 
2.3 Cambios en los mecanismos de regulación de la hexocinasa 
 y de la fosfofructocinasa tipo I. 
34 
2.3.1 Hexocinasa (HK)                                                                                            34 
 2.3.2 Fosfofructocinasa tipo 1 (PFK-1) 35 
 
Capítulo 3    La glucólisis como un blanco terapéutico 
 
3.1 La glucólisis como un blanco terapéutico en el tratamiento del cáncer           38 
3.2 Análisis de control metabólico                                                                       39 
3.2.1 Análisis de elasticidades                                                                                40 
3.2.2 Modelos matemáticos de vías metabólicas (In silico biology)                        43 
  
 
 
 
 
Capítulo 4 
4.1 Planteamiento del problema                                                       45 
4.2 Hipótesis                                                                                     45 
4.3 Objetivo General                                                                                               46 
4.4 Objetivos particulares                                                                                        46 
 
Capítulo 5     Resultados 
 
5.1 Distribución de control en la glucólisis de las células AS-30D  
(Manuscrito No. 2: FEBS J. 2006, 273:1975-1988).                               
 
 
47 
5.1.1 Datos no mostrados                                   66 
5.1.2 Comentarios sobre el manuscrito No 2          68 
5.2 Construcción de los modelos cinéticos de la glucólisis tumoral 
 (Manuscrito No. 3: BBA-Bioenergetics, 2010).          
 
 
71 
5.2.1 Comentarios manuscrito No. 3              99 
5.3 Inhibición de la hexocinasa (Manuscrito No. 4)    100 
5.3.1 Comentarios  manuscrito No. 4               126 
 
Capítulo 6  
 
6.1 Discusión general              127 
6.2 Conclusiones generales       132 
6.3 Perspectivas     133 
6.4 Referencias                          137 
6.5 Apéndice                  143 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
“CONTROL DE LA GLUCÓLISIS EN CÉLULAS  
TUMORALES DE RÁPIDO CRECIMIENTO”  
 
 
RESUMEN  
Un incremento significativo en la velocidad de la glucólisis se observa en la mayoría 
de las células tumorales con respecto a sus tejidos de origen. Dada la constante 
duplicación de las células tumorales, se considera que la glucólisis es una vía metabólica 
de gran importancia, ya que proporciona los esqueletos de carbono necesarios para la 
síntesis de proteínas, DNA y triglicéridos. Además puede contribuir en la síntesis de ATP 
cuando las células se ven sometidas a bajas concentraciones de oxígeno o cuando la 
mitocondria es disfuncional. 
Este incremento en la glucólisis se atribuye a un aumento en la expresión de todas 
las enzimas y transportadores de esta vía, promovido por algunos oncogenes (C-MYC) o 
por el factor inducido por hipoxia (HIF-1α). Además, la hexocinasa (HK) y  la 
fosfofructocinasa tipo 1 (PFK-1) presentan cambios en sus mecanismos de regulación. 
Estos antecedentes nos llevaron a proponer que la distribución de control en la glucólisis 
de las células tumorales era diferente a lo que se observa en las células normales. 
Para evaluar nuestra propuesta inicialmente se caracterizó la glucólisis en células 
normales (hepatocitos de rata) y en células tumorales (AS-30D). En ambos modelos se 
midió el flujo, la concentración de algunos metabolitos y la actividad de cada una de las 
enzimas de la glucólisis. Los resultados indicaron que en las células tumorales (en 
comparación a las células normales) había un incremento en la actividad de la mayoría de 
las enzimas de la glucólisis de 1 a 4 veces, aunque el cambio más significativo que se 
observó fue el incremento en la actividad de la HK  y de la PFK-1 de 306 y 56 veces, 
respectivamente. En consecuencia, también hubo un aumento en las concentraciones 
intracelulares de sus productos: la glucosa 6-fosfato (G6P) y  la fructosa 1,6-bifosfato 
(FBP) entre 5 y 250 veces; y un incremento en el flujo glucolítico de 9 veces. 
 
 
Posteriormente, para determinar la distribución de control se empleo el análisis de 
elasticidades y el modelado cinético, y se encontró que el control de la glucólisis en las 
células AS-30D, recae en el transportador de glucosa (GLUT) (CJ
Ei= 0.2), en la HK (CJ
Ei= 
0.44) y en la hexosa fosfato isomerasa (HPI) (CJ
Ei= 0.4). Asimismo, el modelado cinético 
de la glucólisis en las células tumorales humanas HeLa determinó que los principales 
sitios de control son el GLUT (CJ
Ei= 0.4) y la degradación de glucógeno (CJ
Ei= 0.57). Estas 
diferencias en la distribución de control entre ambos tipos de células tumorales se 
atribuyeron a la expresión de diferentes isoformas del GLUT y al mayor contenido de 
glucógeno en las células HeLa.  
Esta distribución de control apoyó parcialmente nuestra propuesta, pues la PFK-1 
no ejerció un control significativo en las células tumorales mientras que en células 
normales esta enzima es uno de los sitios principales de control.  La razón mecanística 
residió en la presencia de una alta concentración de los principales activadores de la 
enzima en el citosol de las células tumorales como son la fructosa 2-6 bisfosfato, el AMP y 
el fosfato (Pi), los cuales, por un lado, llevan a la enzima a su máxima actividad y, por otro, 
bloquean la inhibición que ejercen el ATP y el citrato. Por otra parte, la HK mantiene un 
control significativo de la glucólisis tumoral debido a que independiente de su localización 
(citosol o membrana externa mitocondrial) continua siendo fuertemente inhibida por la 
G6P. 
Una vez que determinamos que la HK era uno de los principales sitios de control de 
la glucólisis en AS-30D, se planteó disminuir el flujo de esta vía al inhibir a esta enzima. 
Sin embargo, no se observó una baja en el flujo glucolítico, por lo que concluimos que era 
necesario inhibir a más de una enzima para poder observar una disminución significativa 
en la glucólisis.     
 
 
ABSTRACT  
A significant increase in the rate of glycolysis is observed in most tumor cells with 
respect to their tissue of origin. Due to the continuous duplication of tumor cells, it is 
thought that glycolysis is an essential metabolic pathway, providing the carbon skeletons 
necessary for the synthesis of macromolecules (proteins, DNA and triglycerides). 
Additionally, it can help in the synthesis of ATP when cells are subjected to low 
concentrations of oxygen or when the mitochondrion is dysfunctional. 
This increase in glycolysis is attributed to an increase in the expression of all 
enzymes and transporters of this pathway, promoted by some oncogenes (C-MYC) or the 
hypoxia-induced factor (HIF-1α). In addition, hexokinase (HK) and phosphofructokinase 
type 1 (PFK-I) undergo changes in their regulatory mechanisms. Our proposal was that the 
distribution of control in glycolysis of tumor cells was different from that observed in normal 
cells.  
  To evaluate our proposal we began characterizing glycolysis in both non-tumor cells 
(rat hepatocytes) and tumor cells (AS-30D). In both models the fluxes, concentration of 
some metabolites and activity of each of the enzymes of glycolysis were determined. The 
results indicated that in tumor cells (with respect to normal cells) all glycolytic enzymes 
increase their activity by 1 to 4 times, although the most significant change was observed 
in HK and PFK-1 activities with increases of 306 and 56 times, respectively. As a 
consequence, the intracellular concentrations of glucose 6-phosphate (G6P) and fructose 
1,6-bisphosphate (F1,6BP), which are products of HK and PFK-1, respectively, increased 
by 5- and 250-fold as well as the glycolytic flux  (9 fold). 
To determine the distribution of control two approaches were used: elasticity 
analysis and kinetic modeling. Thus, it was established that the main control of the 
 
 
glycolytic flux in AS-30D cells resided in the glucose transporter (GLUT) (CJ
Ei= 0.2), HK 
(CJ
Ei= 0.44) and hexose phosphate isomerase (HPI) (CJ
Ei= 0.4) whereas in HeLa cells 
were the glycogen degradation pathway (CJ
Ei= 0.57) and GLUT (CJ
Ei= 0.4). The differences 
in the distribution of control between tumor cells were attributed to the expression of 
different GLUT isoforms, the over-expression levels of HK, and the glycogen content. 
This distribution of control partially supported our original proposal because the 
PFK-1 did not exert significant control in tumor cells, in marked contrast to non-tumor cells 
in which this enzyme is one of the main controlling steps. The mechanistic reason was that 
the high cytosolic concentration of the PFK-1 activators in tumors cells fructose 2-6, 
bisphosphate, AMP and Pi brought about both an increased PFK-1 activity and protection 
against ATP and citrate inhibition. Moreover, HK maintained significant control because 
regardless of their location (cytosol or mitochondrial outer membrane), it was strongly 
inhibited by G6P. 
Finally, considering that HK was one of the main controlling steps of glycolysis in 
AS-30D, it was examined the possibility of inhibiting tumor HK with the anti-cancer drug 
casiopeina II-gly to thus blocking the glycolytic flux. However, the glycolytic flux was not 
significantly diminished leading to the conclusion that it is necessary to inhibit more than 
one enzyme to observe a significant decrease in tumor glycolysis. 
 
 
CONTENIDO  
Esta tesis está dividida en 6 capítulos. En el capítulo 1 se describen generalidades sobre 
la glucólisis tumoral. El capítulo 2 comprende los mecanismos que inducen el incremento 
de la glucólisis en las células tumorales y aquí se incluye un artículo de revisión sobre el 
factor inducido por la hipoxia (HIF-1α) publicado en la revista Mini Reviews in Medicinal 
Chemistry (Manuscrito No. 1). En el capítulo 3 se relatan las generalidades de la teoría de 
control metabólico, mientras que el plantemiento del problema, la hipótesis y los objetivos 
se incluyen en el capítulo 4. El capítulo 5 comprende los resultados que se dividen en tres 
partes. En la primera y en la segunda parte se establecen las enzimas que controlan el 
flujo de la glucólisis tumoral, mediante dos herramientas experimentales del análisis de 
control métabolico: el análisis de elasticidades (Manuscrito No. 2: FEBS J. 2006, 
273:1975-1988)  y el modelado cinético (Manuscrito No. 3: BBA-Bioenergetics, 2010). La 
tercera parte trata de la inhibición de la hexocinasa para reducir  la glucólisis; este trabajo 
se encuentra en preparación y se muestra la primera versión del artículo (Manuscrito No. 
4). Finalmente, el capítulo 6 incluye la discusión general, las conclusiones, las 
perspectivas, las referencias y un apéndice. 
 
 
 
ABREVIATURAS 
1,3BPG: 1,3 bisfosfoglicerato 
2PG: 2-fosfoglicerato 
3PG: 3-fosfoglicerato 
ALD: Aldolasa 
AMPK: AMP cinasa                                                                                                             
DHAP: Dihidroxiacetona fosfato 
ENO: Enolasa 
F1,6BP: Fructosa 1,6 bisfosfato 
F2,6BP: Fructosa 2,6 bisfosfato 
F6P: Fructosa 6-fosfato  
G3P: Gliceraldehído 3-fosfato 
G6P: Glucosa 6-fosfato 
GAPDH: Gliceraldehído 3-fosfato deshidrogenasa 
GLUT: Transportador de glucosa 
HIF-1α: El factor inducido por la hipoxia 
HK: Hexocinasa 
HPI: Hexosa fosfato isomerasa 
LDH: Lactato deshidrogenasa 
MTC: Transportador de monocarboxilatos 
PEP: Fosfoenolpiruvato 
PFK-1: Fosfofructocinasa tipo 1 
PFKFB3: Fosfofructocinasa tipo 2 
PGAM: Fosfoglicerato mutasa 
PGK: Fosfoglicerato cinasa 
Pi: Fosfato 
PYK: Piruvato cinasa 
TPI: Triosa fosfato isomerasa 
 
 
 
 
1 
 
Capítulo 1. La glucólisis en el cáncer  
 
1.1 Introducción 
El cáncer es una enfermedad que surge de una serie de alteraciones genéticas 
(mutaciones, amplificaciones y pérdidas de genes) producidas por defectos genéticos o 
agentes ambientales (radiación, carcinógenos y virus oncogénicos), que dejan sin 
regulación la proliferación y la diferenciación celular, propiciando que una célula pueda 
llegar a generar un tumor (Vogelstein y Kinzler, 2004; Tysnes y Bjerkvig, 2007). Este 
proceso involucra la activación de oncogenes (HER2, Src, H-Ras, c-Myc) y la 
inactivación de genes supresores de tumores (PTEN, p53, pRb) los cuales intervienen 
en la regulación del ciclo celular y la proliferación (Vogelstein y Kinzler, 2004; Gillies et 
al., 2008). 
Durante el 2007, en México, el cáncer ocupó el tercer lugar como causa de 
muerte con 514, 420 defunciones. Entre las tres principales causas de muerte por 
tumores en hombres se encuentran el cáncer de próstata, el de pulmón y el de 
estómago. En las mujeres, las muertes correspondieron principalmente a el cáncer de 
mama, al de cuello del útero (cérvico-uterino) y al de hígado (INEGI, 2009). 
El tratamiento del cáncer se basa en la mayor susceptibilidad que tienen las 
células tumorales (con respecto a las células normales) al daño inducido por la 
radiación y la quimioterapia, que dejan severos efectos secundarios y en muchos casos 
los tumores son refractarios. Esto hace indispensable la búsqueda de nuevas 
estrategias terapéuticas que sean más selectivas y eficientes contra el cáncer 
(Hornberg et al., 2007). Por ello, encontrar diferencias a nivel molecular entre las 
células no tumorales y las células tumorales es esencial (Pelicano et al., 2006). 
 
 
2 
 
A este respecto, una de las alteraciones que se mantiene en la mayoría de las 
células tumorales, es un alto consumo de glucosa en comparación a las células 
normales, que las lleva a excretar una gran cantidad de lactato, fenómeno conocido 
como efecto “Warburg”, el cual resulta del aumento en la expresión de todas las 
enzimas y transportadores de la glucólisis inducido por algunos oncogenes y por el 
factor inducible por hipoxia-1 (HIF-1) (Walenta y Mueller-Klieser, 2004; Moreno-
Sánchez et al., 2007).  
El incremento en la velocidad de la glucólisis tumoral puede favorecer la síntesis 
de macromoléculas (DNA, trigliceridos, proteínas) necesarias para una constante 
duplicación. Además puede contribuir en la generación de ATP cuando la mitocondria 
tumoral se encuentra dañada (Carew y Huang, 2002) o cuando las células tumorales se 
encuentran bajo hipoxia (Xu et al., 2005).  
Por lo anterior, diversos grupos de investigación consideran que inhibir a la 
glucólisis tumoral podría ser una adecuada estrategia en el tratamiento del cáncer 
(Pedersen et al., 2002; Pelicano et al., 2006; Bartrons y Caro, 2007; Clem et al., 2008; 
Evans et al., 2008; Kumagai et al., 2008). Sin embargo, para que dicha inhibición sea 
más efectiva se debe considerar inhibir solo a las enzimas que la controlan. El análisis 
de control metabólico permite identificar a este tipo de enzimas aplicando diferentes 
estrategias experimentales (Moreno-Sánchez et al., 2008). 
Mediante el empleo del análisis de control metabólico, en este trabajo de tesis se 
determinaron los principales sitios de control de la glucólisis en las células tumorales. 
 
 
 
 
 
3 
 
1.2  La glucólisis 
La vía de Embden-Meyorhof-Parnas, mejor conocida como glucólisis, fue 
descubierta por Eduard Buchner en 1897, aunque fue hasta 1930 cuando se describió 
la vía completa gracias a los estudios de Otto Warburg, Hans von Euler-Chelpin, Gustav 
Embden, Otto Meyerhof y Jacob Parnas (Nelson y Cox, 2005). Esta vía metabólica 
genera, a partir de una molécula de glucosa, dos moléculas de piruvato, y la energía 
liberada en esta conversión se conserva en forma de ATP y NADH. En aerobiosis, el 
piruvato formado es oxidado casi por completo por la mitocondria a CO2 y H2O. En 
cambio, cuando no hay suficiente oxígeno para soportar la oxidación del piruvato y del 
NADH, el piruvato se reduce a lactato, aceptando electrones del NADH y formándose el 
NAD+ necesario para mantener a la glucólisis funcionando (Figura 1.1).  
 Es interesante señalar que aunque en los libros de texto, en general, no se 
consideran como parte de la glucólisis a los transportadores de glucosa y de 
monocarboxilatos, ambos transportadores son esenciales, ya que de ellos depende la 
entrada de la glucosa y la salida del lactato (que bajo anaerobiosis es el producto final) 
(Figura 1.1). 
 
1.3  La glucólisis tumoral 
Las células tumorales presentan como característica común un incremento 
significativo en la velocidad de glucólisis con respecto a sus tejidos de origen, que se ve 
reflejado en un aumento en el consumo de glucosa y en la producción de una gran 
cantidad de lactato (aún en presencia de oxígeno). Este fenómeno fue descrito por 
Warburg de ahí que sea conocido como efecto “Warburg” (Pedersen, 1979; Eigenbrodt, 
et al., 1985; Walenta y Mueller-Klieser, 2004; Moreno-Sánchez et al., 2007).  
 
 
4 
 
G6P
F6P
F1,6BP
G3PDHAP
3PG
2PG
PEP PIR
1,3BPG
+
G3P
F2,6BP
GLU
LAC
HK
PFK-1
ALD
GAPDH
PGK
ENO
PYK
LDH
PFK-2
TPI
HPI
LAC
LAC
H+
H+
PGAM
GLU
ATP
ADP
NAD+ + Pi
NADH
ADP
ATP
NAD+NADH
ATP ADP
ATPADP
GLUT
MCT
Pentosas fosfato
Ribosa + NADPH
Triglicéridos
Serina
Cisteína
Glicina
Alanina
ATP
ADP
PIR
PIR Mitocondria
Síntesis-Degradación
de glucógeno
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Figura 1.1 La glucólisis y su contribución en la síntesis de algunos intermediarios. GLUT, 
transportador de glucosa; GLU, glucosa; HK, hexocinasa; G6P, glucosa 6P; HPI, hexosa fosfato 
isomerasa; F6P, fructosa 6P; PFK-1, fosfofructocinasa tipo 1; F1,6BP, fructosa 1,6 bifosfato; 
PFK-2, fosfofructocinasa tipo 2; F2,6BP, fructosa 2,6 bifosfato; ALD, aldolasa; DHAP, 
dihidroxiacetona fosfato;G3P, gliceraldehído 3P; TPI, triosa fosfato isomerasa; GAPDH, 
gliceraldehído 3P deshidrogenasa; 1,3BPG, 1,3 bifosfoglicerato, PGK, fosfoglicerato cinasa; 
3PG, 3-fosfoglicerato; PGAM, fosfoglicerato mutasa; 2PG, 2-fosfoglicerato; ENO, enolasa; PEP, 
fosfoenolpiruvato; PYK, piruvato cinasa; PIR, piruvato; LDH, lactato deshidrogenasa; LAC, 
lactato; MTC, transportador de monocarboxilatos.   Modificado de Mini. Rev. Med. Chem. 2009, 
9:1084-1101.  
 
 
5 
 
En células tumorales de rata (Novikoff, H-91, Jensen, Rous y Erlich) la 
producción de lactato llega a ser de 5 a 28 veces mayor que en los tejidos normales 
(riñón, hígado, músculo, cerebro y placenta) (Pedersen, 1979). En diversos tipos de 
carcinomas humanos (cervico-uterino, cabeza y cuello, colon-rectal) la concentración de 
lactato alcanzan los 40 μmol por gramo de tejido (peso húmedo), la cual es muy 
elevada considerando que los tejidos sanos mantienen una concentración de lactato de 
alrededor de 2.5 μmol/g de tejido (peso húmedo) (Walenta y Mueller-Klieser, 2004). En 
el suero de pacientes el lactato alcanza una concentración de hasta 9 mM, 
concentración 4 veces mayor a la concentración que se encuentra en individuos sanos 
(Nishijima, et al., 1997; Fisher, et al., 2007).  
El incremento de la glucólisis se ha relacionado con el grado de malignidad de 
los tumores. En hepatomas de rata (tumores experimentales) se ha encontrado que en 
aquellos con una baja malignidad (reducida velocidad de crecimiento) mantienen una 
producción de láctico similar al tejido de origen (hígado) (0.16 nmol/min * mg de peso 
húmedo), mientras que en tumores malignos (elevada velocidad de crecimiento) 
mantienen una producción de láctico entre 7 a 27 veces mayor  (1.6-4.3 nmol/min*mg 
de peso húmedo) con respecto al tejido normal (Pedersen et al., 1978; Stubbs et al., 
2003).  
En tumores humanos de cerebro, mama, pulmón, hígado, cérvix, cabeza y cuello 
niveles altos de lactato tienen pobre prognosis (Gillies et al., 2008), además de que este 
incremento se asocia con una alta incidencia de metástasis (Walenta et al., 2004). 
Asimismo, la expresión de la LDH-A está relacionada con la agresividad, metástasis y 
pobre prognosis en los tumores de colon, pulmón, endometrio y estómago 
(Giatromanolaki et al., 2006).   
 
 
6 
 
Por último, es importante señalar que el incremento en la glucólisis no es 
exclusivo de las células tumorales ni ocurre en todos los tipos de cáncer (próstata; Liu 
et al., 2010). Este aumento en la glucólisis también se observa en células proliferativas 
normales (linfocitos y otras células hematopoyéticas) (DeBerardinis et al., 2008).  
 
1.4 ¿Qué ventajas ofrece a las células tumorales el incremento de la glucólisis? 
La mayor parte de las macromoléculas requeridas para la proliferación se 
generan a partir de intermediarios de la glucólisis (Bauer et al. 2005). Por ejemplo, la 
glucosa 6-P (G6P) es el precursor de la ribosa 5P (R5P), metabolito necesario para la 
síntesis de ácidos nucleicos; en paralelo se genera NADPH, útil en la destoxificación de 
radicales libres, en la síntesis de lípidos y de la desoxirribosa; también hay síntesis de 
UDP-glucoronato usado en la producción de proteo-glicanos, glicoproteínas y en la 
detoxificación de xenobióticos. La síntesis de triglicéridos y fosfolípidos comienza con la 
dihidroxiacetona fosfato (DHAP).  A partir del 3-fosfoglicerato (3PG) y del piruvato (PIR) 
se sintetizan varios aminoácidos (serina, cisteína, glicina y alanina) para la generación 
de proteínas (Figura 1.1). 
Por otra parte, la glucólisis puede ser una fuente muy valiosa de ATP cuando las 
células se ven sometidas a bajas tensiones de oxígeno como ocurre cuando el tumor 
aumenta de tamaño y el crecimiento desorganizado de los nuevos vasos (angiogénesis) 
deja grandes zonas hipóxicas (Gatenby  y Filies, 2004) o cuando la mitocondria deja de 
ser funcional debido a la alta velocidad de mutaciones que se presentan en el ADN-
mitocondrial (Carew and Huang, 2002).  
Las grandes cantidades de láctico (producto de la glucólisis) expulsado por las 
células tumorales acidifica los alrededores del tumor, lo que induce la muerte de las 
 
 
7 
 
poblaciones vecinas de células normales que son incapaces de adaptarse a la acidosis, 
esto permite que las células tumorales invadan otros tejidos y además incrementa la 
radio-quimio resistencia y las metástasis (Gatenby y Gillies, 2004). Asimismo, se ha 
descrito que el láctico afecta el funcionamiento del sistema inmune, lo cual protege a las 
células tumorales. La acidificación puede inhibir la actividad de las células NK y la 
migración de los polimorfonucleares (Lardner, 2001) permitiendo que las células 
tumorales no sean eliminadas. El lactato también reduce la proliferación, la producción 
de citocinas y la citotoxicidad en los linfocitos T (Fisher et al., 2007). Al mismo tiempo, el 
lactato puede permitir la estabilización del HIF-1α, factor muy importante en el 
desarrollo tumoral. 
   
 
  
 
 
 
8 
 
Capítulo 2. Mecanismos que inducen el incremento de la glucólisis 
 
2.1 El factor inducido por la hipoxia (HIF-1α) 
HIF-1α tiene una función muy importante en el incremento de la glucólisis ya que 
induce la transcripción de la mayoría de los genes de las enzimas y transportadores de 
la glucólisis. Además como se indicará posteriormente, los genes supresores de 
tumores y los oncogenes puede indirectamente incrementar la glucólisis, al  permitir la 
estabilización de este factor de transcripción (aun bajo condiciones no hipóxicas). En la 
siguiente revisión, se discute la importancia de este factor en el metabolismo energético 
de las células tumorales.    
 
 
Mini Reviews in Medicinal Chemistry 2009, 9: 1084-1101
 
 
 
9 
 
Factor de impacto 2009: 2.97 
Citas: 15 lllN4 
HiF-la Moduiates Energy Metabolism in Cancer CeUs by inducing 
Over-Expression of Specific Glycolytic lsoforms 
Ah",irad: To tle'rdop ncw antl more efficicnl antl-canccr ii wiii he importanl lo characteriLc the üf 
[Ta"""'I""'" fac1m ~scntiall{.r One such fae!nr i" factor-1ft (HfF-l.u), a tran-
low oxygen cOJiditlons útld fouud ill high levels m rnúl!gnanl so lid flJntOts, hUí not ¡Ji nonnal 
ti5sues or slow-growiug tumors" [n fast·growitlg tUti101'S, HJF-Ja 16 iuvolved ln {he adlvutiün ofullmeroos cel1ular proc-
esses Jnduding resistance agaiU5t apoptosis:. over,expr,;:ssiol1 01' drug eftlu'X membniue pumps. 'VascUlar remooe1i1'ig al1d 
áJiglogenésis as wd] as Ine:mswis, In táliéer te11$" HIF· lÚlnduées úver"expressiml and increase:d activiry of several 
cniytic pmtein isnfhrms lhaí ulffcr from ihnse fmmtÍ in non-malignaní -cclb, incluuing inmspuncTh {(¡LLT I GLCT3) antÍ 
<'-41Lym<.-"S (HKL HKIL PFK~L ALD-A, ALD-C. PGK 1, El\'O-o., PYK-Ml, LDH-A, PFKFB-3" The enhaneed iumor gly-
Hux by HIF-l ú alw imolves changes in tbe killt'tt;.' púnems úf l!;úÍ(:¡nns ú1' key glytotyuc en-
ZylUéS. 1"1Ie HIF-1 (J iudul'c..i iwfonns pfovide C¡mCét rells wHIt reduoed sensltÍvJty to physJúlogkal jnhihitors, lower atTin-
it)' für pmthzcts ami high>:% catal:vtic c3padty (Vmaxf) in fOT\\"uru reactions becausc 01' marked oVeT*Cxpress.ion compareJ 
tú rnúsc iwfúl1l1S ir} núnYlaJ tiS3t\éS. Some of the HIFi (J-indúeed búfúrms ruw particlpaté irt SUf\ 1vál 
pathway~, mdudlug tr3ns.:npl]onal aCiivatitm of H2B ni~ttme lby LDI-j-AJ. inhihillon nI' armrrto~is (hy f-jKIi) ano pmmo-
!itm uf ec1! migmlinn (by El\'O-at fHF-la ~{ciiHU m~{y ahm modulate milHchunJrial functi(m and oxyg:en ctm:.umptiun by 
"lacte, lIlHlg tbt· pynl\ are iri Sorne tllmor tYPC$, or e úxidtlh'! subullí! 4 
expte$$Jon fO incfease oxJdativé phosphot)ltatioH in otllef CfU¡,¡;cr cell lines. In ll1Js rev íevv, the foles of HlF· 1 ü and H1F t (f-
induca.! glycolytlc enL)'méh are examina.! JUlil it is conchJJed tÍlJl targeting the HIF ln-indueed glucose tranhpOl1ér anJ 
hcxúkitltl..&é, impol1tl1it tú tlux eúntmL providt" bcttcr targds for ttlrnúr ana 
progre~,,¡ml than targeting HIfItLltseU: 
Ke~ \" ords: G Iw.::o~c lf,¡:nlSporb,:;FS. ÍI":;_\OK ¡¡¡aSés. ] I j [" ~ I fL ~h,éll'¡V1;", tnitocl!(flltlfia. """ ,,,vme illli ibitOf:',. lni{oc!¡ündrt;:¡ I ¡nh ihi-
ton;. 
I'iTRODlICllOl\: 1111'-1 
Th~ oC [wooue ;0 suHd tumors ls a 
n:curn:nt f~atun: \vhích 1::: linkeu lo the pnJC'"'''' uf 
lhUlt lnl.rl:.Il:)ffllatiorl. !i1dOstll:-'1S ond ti¡) d)CnlU-, 
immurlo- ~Ind ¡. 2]. Thc 
fhelor "~;~=~:,:,~thut a 
e "l' nree[cil1S 
In cr:iU1[roPulcVClS, ,(.:;;;Uu lar nneJlll él'a-
and :::arvtvaL v:t3,;,'ular v:I:tümtHor ecmlroL 
tHéhlbolk 
¡\Jmv"It'" "c't eral I:'OfUr1H:' of U] F ex ij<¡f f Uf/- L 
-2 and -3 i see hcltm ror ¡nore d.:taH), the fucus of re\, ll:\V 
wil] be on IHf~l tx:cau3,e ü:s failCtlOil::: aré the mo"t weH ue-
fhiétl In fdlllk:HI tu IIKldllylng T11e lHoh:culllf hi\)-
IH,.:;,,·hnniS1!l:', ~"CU!uy ílíF-l dé[:,- a~ él u¡¡ns"ríp[iemI111 
activatiüo factor re¡:Cllallog 
~wdkd4 sirle..: 
0.\10""¡,00 ha\\! bl.-'t'Y1 Gxli::n-
pUI"[''''IIO'' ,,1' S":-!i1":;ll/;:¡ el 
aud r;;v ícw,,>d prCircm cC'vicw Ilx:us:.'s 
on Ol!.jS~ {ha1 TU!':S in tl1t: 
eh;,",!'" in nux inshJé CJmC":;f cdl~ arl,J WltOSé 1,,0-
f;'lnns ar..: iHI'-i imd 
IIlF-l 
4. 
D'\IA 
tmil 
fáétor ¡:; M ti Mthunils. 
\"hich hotb >comain (lnc bt:1~1 i 1,,1h.-
) ,md }IVO (P[R-ARNT 
sub-
lit.: 
tlw 
250 O;:: diss(J!\;cd irl w~it..:r ami :nÚc). 
it:, haH',.Iif\; oC 5 min, j IIF-la (:<)i!h:H1 i~ 
 
 
 
10 
 
1-II F _ l a .. e l h ·., 
·r ....... r .. .... pr ........ n. 
1· ~3 1··,· .. .: ,..., 
_ .,-rL· 
II ~ H 2 .. " " 
I - H .. , S 
e P V II L ::'> 
" 
• hltllt • '''ASI 
111 1"_ 1 0 1 .... .., 1 1 ... .., 
Fi g. ( 1). Rrg ulat ion of HII '~ - Ia sta llility and ar thrity. 
Milli-Rt.'lliell'.'· i" M f!lliám¡/ (1U' lIIi.wry. 2U09. V"I. 9. NIi. 9 108 5 
0. 
(-) 
C O . 
, ~ .. c 
FU .H 
' ·· y r 
s .. ..,.., 
o. 
(-) 
C O . 
Under nonnoxia, prolyl hydroxylase (p1·ID) hydroxy lates prolin e (p ro) res id ues (402 and 564) of HrF-la in a region call ed the oxygen-
dependent degrada tion (O-DD) dorna in, which faci litates its interaction with the von Hippel-Lindau protein (PVHL) and hence with an ubiq-
uitin-pro tein ligase complex that marks HfF-la for destruc lion by Ihe proteasome. Asparaginyl-aspartyl hydroxy lases (A Hs) by hydroxy lat-
ing an Asn residue (803) in the carboxy-tenninal transcr iptiona l activation dornain (C-TAD) of HTF-Ia, inhibits the binding of cofactors, 
such as p300 and CBP that are requ ired fo r the transcription oftargel genes. HrF-la is a heterodirner thal binds to hypox ic responsive ele-
ments (HRE) contained in Ihe promoler region of the glycolytic genes. Abbreviations: 2-oxo, 2-oxoglul.:1rale; Succ, succinate; N-TAD, 
amine-tenninal transcriptional activarion domain; Lac, lactate; Fum, fumarme; Pyr, pyru vate; Asc, ascorbate, (-), in hibition. 
thal mim ics hypox ic conditions, thc hal r- lifc bccomes in-
ereased 10 3 O 111 in [8). 
In particular, enhanced I IlF-l express ion has bcen de-
tected in the l1lajori ly 01' brain, pancreas, l1lal1lmary gland, 
colon, ovary, lung and prostate pril1l3ry tumors, and in thei r 
l1le l3Slasis, b ut not in the l1lajority 01' bclli gll !ul1l ors or nor-
mal ti ssues [9-15]. lI igher expression 01' I II F-Ia correlates 
with poor surviva l in breas!, head and neck, esophagus, 
stomach and IUllg cancers, although ror cervical cancer this 
assoc iation is no! so c lear [ 15]. Both the biological complex-
ity 01" the I [l F systcm and methodological difficu lties such as 
the criteria used 10 ident iry 111 F positi ve ce ll s, immunohisto-
chemical protocols al1(l source or tumor tissue for its experi-
mental eval uation most probably account for any conflic ling 
data [15]. 
Regulalo ry M echanis l11s Co ntrollin g Hl f- I Activity 
The von lI ippel-Lindau prolein (pVl IL), a component o r 
the ubiqui tin ligase E3 complex, reg ulates IUF-Ia degrada-
tion (Fig. 1). For pVIlL-IlI F interactioll , IIl F-la l11 uSI fi rs t 
be hydroxy lated al prol ines 402 and 564 in the oxygen-
dependenl deb'fadalion (O-DO) dOl11 ain by prolyl-4-hydroxy-
lases (P il Os). III F-I a transcriptional aClivi ty can also bc 
directly inhibi ted by I\.sn 803 hydroxy lat ion ca lalyzed by 
asparaginyl-aspartyl hyelroxylases (1\. I Is; also known as fac -
tors inhibiti ng IITF- I, FllI s). lIydroxyla tion prevcllls re -
crui l1l1enl 01' p300/CB P co-act iva to rs that would otherwisc 
combine together with ¡ IIF-Ia, rorming the ac tivc transc rip -
lional eompl ex [7]lhal binds largel genes (Fig. 1). 
PilO and 1\. 11 enzymalic activi ties both requi re Fe2
+, 2-
oxoglutarate, ascorbate and oxygcn (Fig. 1). Il encc, one way 
to reduce I!l F hydroxylation would be by decreasing the 
oxygcll level below to Ihm req uired fo r Pil Os. In Ihis l11an -
ner, it was proposed that these enzymes sense intracellu lar 
oxygen lcve ls [ 16, 17] ane! that under hypoxic cond itions, 
PI ro s and A l is are inactivatcd wilh the result that In F-I a. 
becolllcs s ta bilized alld activa ted [ 16]. 
The ro le of PI¡ Os as <ln ill lracell ular oxygen sens illg sys-
tcm remains 1I1lce rtain beca llse the hydrox ylase Km vallles 
ror O, are reported lo be mueh higher (> 90 ¡1M) [ 18, 19 ] 
than the actual O2 concelllration thal ex ists in the cytosol 
(12.5-25 ~ M) and in lhe eapi llaries and anerioles (20-50 
 
 
 
11 
 
1086 M ;I/i-Hel';ell'$ il/ J /e{li('jm¡/ 0 /1'111;.\"11')' , 20 09, V{II. 9, NIJ. 9 
¡1 M) [ 17, 20-22]. Conseq uent ly, III F- Ia should nOI be inae-
tivated by hydroxylat ion at an [02] 01' 10 Jl M. Under such 
condi tions, hydroxylase aC livilY, wi th a Km valuc or 100 ~lM 
should only be 9% ol"maximal vcloci ty (Vm), which is mOSI 
likely nOI suffi eienl lo inaeliva le III F- Ia. Inlh is regard, I IIF-
la slabili1 ..a lion has becn rcpo rted 10 occur in inlac l cells al 
[O,] below 50 ¡1M [17]. 
Severa I possibili lies have bcen suggeslcd to explain Ihe 
apparent discrepancy betwccn the high Km va lues or Pll Ds 
dctcnnin ed ror O2 and Ihe raet that IIIF-la hydroxyla lion 
and associa led degradat ion oceurs under nonnox ia. For in-
stancc, kinelic sludi cs havc not takcn inlo account thc contri-
bu tion o r Ihe Icng th 01" the pcptide subs tra tes used in assay-
ing PIID activi ty nor the role that III F-lo subs trate binding 
plays in rac ilitating oxygen bind ing, which may lower the 
ac tual Km (02) 10 morc physiologically relevan l va lues (re-
viewed in [23]). In adcl ition, Ihe expression leve ls 01' PIID2 
and PIID 3 are thcl11 selves íncreased under hypoxic condí-
lions by 111 F -1 a [24, 25]. Therefo re, inereased PII D ex pres-
sion wou ld be expccwd 10 also help reduce II IF- lo levcls. ln 
this regard , Pl ID2 was shown to be the most prominently 
ex pressed iso lom1 in a large range 01' cancer cell lines wi lh 
po ten t ac tivity tow~ l r d s III F- I a [24]. 
M itochondria ll nvolvemen t in the Regula tion of I·II F-I o 
A n additional lllechanisl11 explaining how 1 UF- lo levels 
are increased du ring hypox ia is thal Ihe decrease in [02] 
causes an increase in the generat ion or radical oxygen spe-
cies (ROS) in mitochond ria by respiralOry complexes 1 and 
111 p6, 27]. The increased ROS induces oxidation of Fe2
+ to 
Fe3 
, which would function 10 dilllinish lhe hydroxylase ac-
tiv ity or PIID and A l! ( Fig. 1). In agreement wi th th is hy-
pothesis, 111 F-l o is not stabilized in anti-oxidant treated cells 
(hepatoma, smooth muscle, cardiomyocy tes, gastr ic cpi the-
lium , renal tubulc epitheliulll, macrophages) undcr hypox ia. 
In contrast, where ROS product ioll is low, such as in cell s 
laeking (a) mi lochondria l DNA (rho zero, pO ee li s), (b) eylo-
chrome e, or (e) complex 11 1 Rieske iron-sulfur prOlcin , and 
(d) in cells trcated w ilh stigmatellin , an inhibi tor 01' complex 
11 1 [2 8], IIIF- Ia hydroxylation proeeeds emeiently under 
hypox ia [22 , 29, 30]. 
Regarding the role 0 1" lhe rcspiratory ehain , it has becn 
proposed that Ihc reducl ion in [02 ] during hypoxia leads 10 a 
dccrease in cy tochrol11e e oxidase (COX ; complex IV) act iv-
ity [31 ], resu lt ing in thc accumu lalion and ovcrloading of the 
reduced intermediates, ubiquinol and semiquinone, particu-
larly the laller, which then promote superoxide gcncra tion 
(see Fig. 3 fo r chemical structll res). Specific inhibi tion ofthe 
respira lory complexes, by either cyanide (COX), ant imycin 
(complex 111 ; cylochro l11e b-c] complex ), Ihcnoyltrinuoro-
accIOne (TTFA ) or o-tocophery l succinate (comp lcx 11 ; suc-
cinatc dchydrogenase) or ro lenone (eomplcx 1) (Fig. 3), can 
also promote Ihe generation of ROS under normoxia, be-
cause these respirato ry inhibito rs alTecl lhe electron transport 
by respira tory chain complex es to induce increased levcJs or 
sCllliquinone (and other free radica l Illolccul es) [32 -34]. In 
contrast, block ing entry or encrgy substrates to inhibi t thc 
respiralory chain at the eleelTOn cntranec level, by using 
malonate to inhibi t cOlllplex 11 , phenylsucci nate or n-butyl-
Illalonatc to block transpon of succinate and other diear-
Mari,,-IIC/'míllll e: el al. 
boxylate Krebs cycle intermedia tes, or a-cyano-hydroxy-
cinnama tes to prevent pyruvate uptake into mitochondria 
(Figs. 2 and 3), is not expcctcd to induee gencrat ion or ROS, 
as thc.'.:;e inh ibitors do not dircct ly modiry the respirato ry 
chain al the levcl or clcetron fl ow. 
Changes in the Krebs cycle are also likely to contribute to 
the regulation o r IIIF-I o act ivity. The enzylll e2 o r lhc 2-
oxoglutara te dchydrogcnase cO l11plcx (2-0GD II) can be tar-
gctcd for ubiquit ination-dependcnt dcgrada lion by Siah2, thc 
RING fingcr ubiquitin-prolcin isopcpt ide ligase [35]. As 
Siah2 is induccd by hypoxia, d isrup tion 0 1" Illi tochondrial 
mctabolislll by affccting 2-0G DII would lead to loss 01' 
mi tochondrial stability and cell death. 
J-I ow is I·II F- Ia Maintain ed S t:lb lc and Act ivc in C:l ncc r 
Cclls'! 
It is lhought tha l III F-Ia in tumor cc ll s is s tabil izcd duc 
10 the hypoxic environment developcd in ccrtain reg ions, 
part icularly in solid tUlllors 1 mm diameter or larger [36, 37]. 
Although tumors may have an aetive angiogenesis, unorgan-
ized, Ihin and frag ile ne\V vesse ls are formed that afrect the 
normal dynamies 01' the blood flux. Conscqucntly, some tu-
mor scctions \ViII become excluded, leading to hypoxie rc-
gions [ 15, 38]. IIIF-I o stabili za tion is also promotcd by ac ti-
va lion 0 1' cert ain oncogenes such as v-sre, IIER 2""" and J I-
RAS, 01" by inactivation of some tumor suppressors such as 
p53 ami PTEN [3, 6]. 1I 0\Vever, the molecular mechanisms 
operat ing in thesc processes have not been elucida tcd. A 
high incidence o r pV11 L mutatiolls is associated wi th kidncy 
and central llcrvous sys tem tUl11ors. TIlese pVIIL mU lations 
1110dify or dele te ei ther lhe a-dol11ain in thc C-terminal rc-
gion which binds to elongin-C in Ihe pro teaso l11e, or the p-
dOl11ain that in teracL<; with the II IF- Ia O-DD dOl11ain and is 
required for lluclearlcy tosolic trafficking, prevent ing 111 F- I ex. 
degradation (Fig. 1) [3, 39]. 
Under nonlloxia, III F- lo can be stabilized by Ihe high 
laclate and pyruvate levels generated by active tumor glyco-
Iys is. It has been shown that these monocarboxyla les, and 
oxaloaccta tc, inhibit PIID ac tivity by cOlllpcling with 2-
oxogllltara tc for binding [29, 40]. Sillli larly, mll talions or 
down-rcgulation of suceina tc dchydrogcnase (SD II) and fu-
Illa ratc hydrata sc (F II ) induce a statc 0 1' pscudo-hypox ia that 
Illakes caneer eclls behave as if they were hypoxic, which 
leads 10 111 F -1 a stabi li zat ion and enhancel11cn t [41-43]. T hese 
Illutations inhibi t SDII alld FII ac ti vities,leading to succillalc 
and fuma rate accumulat ion, \Vit hout associated ROS prodllc-
tion, and 10 prodllct-inhibition of hydroxylases [41-43] (Fig. 
1). Moreover, SDII a11(1 FI I mUl'ations, o r their down-
regulalion, are associatcd \Vith devclopmcl1t 01' phaeochro-
mocylomas, paragangliomas, leiomyomas, leiomyosarcomas, 
renal cell, gastric and colon carcinomas, and papillary lhy-
roid eancer [22, 29, 41 -43]. 
Add itiona ll·II F Isofo rms 
Three isofo nlls of IIIF-a ha ve been deseribed (111 F-
la, II IF-2aIEPAS I and IIIF-3u11 PAS) and Ihree IIIF-Ip 
iso forms (III F- I O/¡\RNTI, IIIF-20/A RNT2, and 111 F-30/ 
ARNTI), although their exact rc lationships in fonning hel-
erodimcrs are not known. HIF-I a and IUF-2atEPAS 1 share 
similar struc ture, hypoxic stabili zation and excl usi ve dimeri-
 
 
 
12 
 
TI/II/or Glyco!p 'i\', 1-11 F-Ia tl11tlllypox;" M i"i-Rt'I';l'II'S ;" Ml'llidtwl Chemi.\'l ry . 2009, Vol. 9, NII. 9 1087 
GtU+A', " ATI' + GLU O 11 1(1 r'K' A P ADI' .... ~ ~, 
(2) 
GLU.o" / 
'lt GLUTJ 
ADP+ G6I' 
~ T ilofhondrla 
O Monomer 
6 0hner 
m Tt' .... mtr 
o I 'sororms 
C)losol 
, , 
e l. in 
G6 P +ADI' 
IIKI 
, ' , ' 
\(.) H) 
_--__. ',~ _ G61' _ .... ----+- G I )'fO¡;~ n 
~ G61' P('nloses p. th" -.,, 
.... .. _ .... ---...... ~ ¿ 11 1'1 
.. r..¡ rnU' Il-J 
~ (-) ."61' 7<; Fl.6BP 
AT P~ \ ......... l'rK-L ' 
...... !'DP ft':. ATI' AOI' : 
ALIl~:ffi ~~ AU;¿'" ~ .!,/ 
TCs <E---- DIIAI' + GJ I' PYK-M2 
T I" tlru>.< ffi 
ATP + GLU GlP 
A D ' ~ CA I'n l! 
~~ !1 F1.61l1> 
II I'V· 16 E7 
A DI 
1.J1l1'G 
ADP -y'¡ PGK! 
ATP ~ 
~ 3PG 
éPGA M
-
S 
21'C 
t:NO-G é 
PU' 
- g.g 
1\la l . 
r ,\teJ' .. 
"$;jI! N Ir I..AC r--...J-- ti . 
PYR I.AC I..AC 
LD II-A 
.~ . 
AI)P ATP " YR 
"YR 
f ig. (2). C lyco lyl ic isofo rm s up rrgulatcd by Hl f- I in (':III('('r cclls. 
GLUT, glucose transporte r; HK, hexokinase; HPI , hexosephosphate isomerase; PFK 1, phosphofructokinase type 1; ALD, aldolase; PFKFB3, 
phosphofructokinase type 11; TPI, triosephosphate iso merase; GAPD H, glyceraldehyde-3-phosphate dehydrogenase; PGK, phosphoglycerate 
kinase; PGAM , phosphoglycerate mutase; ENO, enolase; PYK, pyruvare kinase; LDH , lactate dehydrogenase; MCT, monocarboxylare 
transporter; PDH, pyruvare dehydrogenase compl ex; PDK, pyruvate dehydrogenase kinase; GLU, gl ucose; G6P, glucose 6-phosphate, F6P, 
[ructose 6-phosphate; F2,6BP, fructose-2 ,6-bisphosphare; Fl ,6SP, [ructose 1,6 bisphosphate; DHAP, dih ydroxyacetone phosphate; G3P, 
glyceraldehyde-3-phosphate; 1,3 BPG, 1,3 bisphosphoglycerate; 3PG, 3-phosphoglycerate; 2PG, 2-phosphoglycerate; PEP, phosphoenolpy-
ruvate; PVR, pyru vate; LAC, lactate; TGs, oiacy lglycetides; Ser, setine; Cys, cysteine; Gly, glycine; Ala, alanine; (+) ac tiva(ion; (-) inhibi-
tion. 
zation with IIIF-lp [44]. IIIF-Ip may also dimerize with aryl 
hydrocarbon receptors, allowing cross-talk wi th xenobiolic 
metabolism. I lowever, complexcs eontaining IIIF-2a act i-
vate a distinet s llbsct of genes, comparcd to III F- I a , that are 
nOI in vol ved in regulating glycolytie genes [45]. 1I1F-2a fi s-
sue express ion occurs in a limited number of non-parenchy-
mal cells (in kidney, pancreas and brain) and parcnchymal 
cc lls (in ¡i vc r, intcstinc and hcart ) [45]. 1I0wcvcr, IIIF-2o. is 
also in vol ved in tumor progrcssion and incrcascd cx prcssion 
has bccn obscrvcd in divcrsc solid lumors, including bladdcr, 
brain , breasl, colon, ovar y, prostatc and renal carcinomas 
[13, 45]. 
The role 0 1' 111 F-3o. is not clear. Three splice variants can 
be produccd from the IIIF-3o. gcnc. Isororm-2, also ca llcd 
IPAS (inh ibitory PAS domain, a natura l lllF-1 a anlagonis t), 
lacks an Asn-containing transactiva tion domain (C-TAD), 
such that it aCls in a dom inant negat ive manner f'onn ing tran-
scriptiona lI y inacti vc hetcro-dimers \Vi th I I I F- I ~ , thereby 
prevcnting I II F- Ia dimerizing with IIIF- lp [44]. In the cor-
neal epitheli llm , where the IPAS conccntration is high, cor-
ncal neo- vascularization is inhibi lcd [46]. On Ihc olhcr hand, 
II IF- l p is constitll lively cxpresscd undcr nonnoxic condi-
tions and is slightly increased by Ihe same effcctors 111at up-
regulatc I II F-In cx pression (hypoxia, EGF, CoCI,) [47]. 
GLYCO LYS IS AN D I'II F-MEDIATED REGU LATlON 
MOS l cancer cells show cnhanccd g lycolytic capacity 
compared to Iheir tissucs of origin [48,49]. This occurs bc-
cause many 01' the glycoly tic enzymes can be cxpressed as 
severa I difTcren t isofonns (Ta ble 1) and Ihe isofonns ex-
prcssed in cancercells are di fTcrcn 1. Th is process is regulatcd 
by II IF-la which ac ts as a transcriptional fac tor for most 
0 (' the glyco lytic enzymes and transportcrs (Figs. 1 and 2) 
 
 
 
13 
 
1088 M il/i-Rel'it'll'l' il/ !l1nliá l/lI / Che/l/i,wfJ' , 2009, 11,,/. 9, N". 9 ll/uríl/-lIl'rl/(Ím/e:. el ti l. 
GLyeOL YTle INHIB ITORS 
1·10 
HO 
Gossypol Koningic ac id 
y; O 
HO 
B'YOH 
OH OH O 
2-dcoxyg lucosc 3-bromopyruvate 
MITOCHONDRIAL INHIBITORS 
~
o O-
HO O O : 1 o /'O~~ 
Phenylsuccinatc n-buthylmalonatc Malonatc 
0-
-O 
H O~ 1 '" Ne 
ó ó o 
O O 
H 
o-
a-cyan,,4-hydroxyc ilmamalc Dicloroacclalc Rotcnonc 
O 
1ñcnoyltrifluoroacctonc 
o OH 
}---{ 
1) O 
Oxalatc 
o NH, 
} --{ -
-O O 
lodoacctatc Oxamatc 
'" 
.& 
N"'\ 
lj ~ h 
OH 
:? el 
'" 
Clotrimazolc 
Stigmatcllin 
oeH'X;=0
0 
R 11 
~ e..... o 
Q: ~ OCOCH,c¡-I(CH, ), 
",..-::. o eH) 
OH 
NJ-iCJ-iO 
R= Hcxyl : Antimycin A l 
R- Butyl: Ant imycin Al 
a-tocophcryl succinalc 
Hypoxia 
HIF-10 INHIBITORS 
%, 
ag 
HO, 
O 
OCH;, 
O : 
Manassantin B
ol 0H 
O 
N(CH;)¿*HC1 
HO. 
o    
HO 
HO 
OH 
Vitexin 
  
Mini-Reviews in Medicinal Chemistry, 2009, Vol. 9, No. 9 1089 
(Fig. 3. Contd....) 
  
A 0H o ÓN 
OCH, 
V-ATPase INHIBITORS 
OCHz CH¿3 | CH; 
OH 
  
OCH, CHz CH, 
H,C 
CH; CH¡ CHz  OCH; CH; 
CHj3 CHz3  OCH; CHz 
Bafilomycin A, 
Concanamycin A
OTHER COMPOUDS 
o ps o* 2 
H,CO (CH3-CH=C-CHa)10-H o H,CO (CH,-CH=CÉCH))yo-H 
ox 
H,¿CO CH 
HO CEs Methylglyoxal > 3 
Oo OH 
Sa ical 
Ubiquinol-10 
Semiquinone-10 radica! 
Fig. (3). Chemical structures of some anticancer drugs that block energy metabolism, 
[4, 44, 50-54]. Interestingly, HIF-1a activation only in- 
creases the transcription of one particular isoform for each of 
the HIF-1a regulated glycolytic proteins. The following sec- 
tion discusses the specific HIF-1a mediated regulation of 
glycolysis and why some of the glycolytic isoforms may 
prove to be suitable drug targets for cancer therapy. 
Glucose Transporters (GLUTs) 
The glucose transporter family consists of three different 
classes. Class 1 contains four members, GLUTI-GLUT4 
(Table 1) whose preferential substrate is glucose. Class 2 and 
3 transporters are selective for other carbohydrates [55]. 
GLUTI and GLUT3 expression is up-regulated by HIF-1a 
(Fig. 2). GLUTI is expressed in all tissue types, whercas 
GLUT3 is preferentially expressed in the brain. Apparently, 
GLUT! can form dimers and tetramers [56]. It has been ar- 
gued that the HIF-1a-mediated GLUT1 and GLUT3 over- 
expression in cancer cells is related to their high glucose 
affíinity (low Km) [57]. However, it is somewhat surprising 
that the kinetic parameters of glucose transporters have been 
determined only for glucose analogues such as 2-deoxy- 
glucose(2-DOG) or 3-O-methyl glucose, but not for glucose 
itself: hence, substantial differences in the kinetic parameters 
have been reported: GLUT1, Km = 6.9-50 mM, Vm = 6.5- 
  
 
 
 
14 
 
Tumor (il)'colp:is. N I F-Ia ((mi /- ~rfJoxit l i/li ('1'it!ll'S ll eflid/Ud /lemislry , 09, 1101. , o.  H9 
(Fi~ . J. Onld ...) 
o 
I - Ia I I RS 
O 
H 
O  
anass lm B :::::-... OCH
J B ~ O H 1 / 
H,C 
 
 
 O 
itcxin 
OCH, 
Di lJlxin 
OH 
- TPasc N I I RS 
CH] H, , 
OH O "'- "'- H 
H}C 
H,C 
}C :::,... 
'<:: 
H3 e 3 O H3 e 3 
ilom ycin Al 
n in A 
HER POUDS 
O CH3 
H ' CO ~ (C ' . ' '-C H ~ ¿ CH ')IO -H O 
D ~ 
O· 
H ' CD~ (C H ' - C H 
H,C VCH, Mcthy glyoxal 
,CDY CH, 
  
Ubiquinol- IO 
Scmiqu inonc- l O radical 
ig. ). he mi cal rures f e t cer r gs at l ck ergy etabo lism. 
, 4, - 54]. t r s li g ly, I IF-Ia e li li n nly -
ses I e Ira ri ti n f e arti ular nn r ch f 
I e I IF-Ia ulated l eo lytie r leins. he lo ing c-
li n i u ses I e c ific I IF-Ia I11 cdialcd ulali n f 
l co lys is d hy Oll1e f I e l colytic so nns ay 
r ve 10 e il ble r g l ets r nccr rapy. 
lucose ran sport ers LUTs) 
he l cose lra sporter i ly nsis l'S f rce i T renl 
J ses. la s I ntains ur embcrs, UT I-  LUT4 
able ) hose r ferential bs lrale  l cosc. la s  d 
 ra orters re cl cti e r t e!' r hydrates 51. 
LUTI d LUD pre sion  -r ul aled y II IF- la 
ig. ). LUTI  re scd  l i e I es, hereas 
Un  re ferenlia ll y pre scd  I e rain. parenlly, 
LUT I n rl11 l11ers d l lr J11 erS 6]. t as ecn r-
ued I al I e II IF- la-l11ed ialed UTI d LUT3 er-
pre sion  cer l s  l t d  eir i h l cose 
ff ity I  ) 7]. Ilowcvcr, t  cwhat r risi g 
at e i clic r rnetcrs 01' lleose I porters ave ecn 
Clcnnill cd l ly r r l cosc l lles ch s eoxy-
l 2- G) r 0- ethyl l cose, ut OI r l cose 
lC; clf; nce, bstall lial i c ccs  I e i etic ara clcrs 
ave een oned : LUTI , ili ~ . - 0 111M, III ~ . -
 
 
 
15 
 
1090 J f il1;-R(!'Ilü!lI's ill J'l e(liál/(/1 Chemi,wfJ', 2009, Vt}l. 9, No. 9 MtlrI;/-lIermí lUle:. el l/l. 
T:tblc l. Isoform sof Glyco lytic I'rotcins 
Tnl nSllO rCl.' r or Enzyllll.' Cl.'nl.'s l soforms Oligolll l.' ric Sla ll.' AnlicllllCl.' r J)rugs 
GLUT 4 GLUTI , GLUT2, GLUT3 . GLUT4 M?, O, T 
HK 4 HKI, HKII , HKIII , HKJ V M.T 3-BrPyr, clolrÍlnazolc 
HPI I No iso forms O 2-DOG 
PFK-I 3 PFK -L, PFK-M. PFK-P T c1olrimazolc 
A LD 3 ALD-A, ALD-B, ALD-C T clOlrimazolc 
Tri I No iso fonns o 
GA PDH 2 GAPD I. GAPD2 T Arscnilc,Goss, IAA, 3-BrPyr, Koningic acid 
PGK 2 PGK I. PGK2 M 3-BrPyr 
PGAM 2 PGAM-A, PCiAM- B o Oxamalc, Oxalalc 
ENO 3 ENO.a., ENO-~ , ENO-., o 
PYK 2 PYK-R, PYK -L, PYK-M 1, PY K-M2 T Oxamalc, Oxatalc 
LDH 3 LDH-A, LDH-B, LDH -C T Goss, Oxamalc, Oxa lalc 
PFK-2 4 PFKFBI, PFKFB2, PFKFB3, PFKFB4 o 
MCf 4 MCT I, Men, MCT3 , MCf4 M? 
M. monomer: D, dlmer: T. tetramer. ¡AA, IOMacetaIC: 2-00G, 2-deoxyglucose: 3-BrPyr, 3-bmmopyruvale: Goss, gossypoL Data takcn from (55,56,63,65,73,77.78,81,100, 
107,109, 11 1,116. 121,124. 128,130, 1321· 
700 pmo Vmin/ooeyle; GLUT2 . Km ~ 17-42 mM. Vm ~ 3. 1-
900 pmo l/min/oocyle; GLUD, Km ~ 1.8- 10 mM. Vm ~ 2.2 -
850 pmoVmin/oocytc; G LUT4, Km = 4.6- 100 mM, Vm = 
ISO pmol/m in/oocyle) [5 8-60]. Based on Ihe kineIic para me-
lCrs dClCrmined for 2-DOG, and thc assumption that Ihe Km 
valucs fo r 2-DOG arc close 10 those for glucose, it can be 
conc1uded that G LUT3 is the transporter with the highcst 
afTini ty and ca laly tic efficiency (Vm/Km ; G LUD>G LUTI > 
GLUT2>GLUT4), while GLUT2 ovcr-expression would bc 
physio logically irrcJevan t at nonnal blood g lucose levels of 
arollnd 5 mM. 
GLU Tl is the transporler mOSI widcJy ovcr-expressed in 
canccr cclls (T able 2), part icu larly in highly proliferative and 
malignant tllmors [55, 61 ]. GLUT3 is a lso over-expresscd in 
lung, co lon, ovary, larynx and mal11l11ary gland IUl110rS (Ta-
ble 2); high levels of G LUTI or GLUD have been lIsed as 
indicators o f bad prognosis [55]. Inleresling ly, G LUT I and 
GLUTI are one of Ih e l11 ain con tro lling steps of glyco lysis in 
sOl11e fas t-growth lumor ce ll s [62; Rodriguez-Enriquez S., 
Marín- I lernández A, Gallardo-Pérez J.c. , Moreno· Sánchez 
R. , unpublis hed dala] , and henee il providc...:; a sui lable Ihera-
pcutic targc l ror glycolytic and hypoxic tllmors . I lowever, 
in hibilors o r G LUT Ihal specifically targel cancer cells ha ve 
11 0t ye l been developed. 
l'lexokinase (1'1 K) 
Monomeric I IK has fo ur isoforms (Table 1) with molecu-
lar mas ses of lOO KDa for IIKI, IIKII alld IIKIII or 50 KDa 
ror I IKI V, or glucokinasc (G K). Their Km va lues ror glucose 
rangc fr0111 0.003 10 8 mM in Ihe o rder of re lati vc aninity 
( l/Km ): II KIII> IIK I> IIKII> IIK IV. The aC livi ly o f isoforms 
1-111 is slrongly inh ibilCd by the prod ucl, G6P, whereas GK 
is fully inscnsiti ve to Ihis l11etabolile [63 ]. IIKI and IIK II 
gcnes are 111 F-I a !arget.:; (Fig. 2) [44]. 11 KII over-expression 
occllrs in th e l11ajority of tumors, although in brain, tes ti s, 
and hcad and neck lumors II KI is prcfcrclltially over-
ex prcsscd (Table 2) [64] and may fonn lelramers [65]. These 
IwO iso forms can bind 10 Ihe exlernal milochondrial mel11-
branc by mcalls or a 15 hydrophobic amino ac id segmcnl , 
MIAS II LLAYFFTELN . in Ihe amino- Ienninal region [66]. 
In somc tumor cclls, Ihc mi tochondria-bound I IK accou nls 
for 50-70% oftotal ce llular IIK [62] . Ilowever, in Ihe major-
ily of kinclic sludies in cancer cell s, Ihe ana lysis 0 1' IIK ac-
livity has bccn derived from Ihc free or cy toso lic iso fonll, 
whilc the contr ibu lion of l11embranc-bollnd IIK has oftcn nOI 
been eva lualcd , Ihereby underestimating Ihe tota l I IK aCliv-
ily. 
Apparently, IrK prefcrcntially interac ts wi lh the I11CI11-
branc pcnneabi lity lrans ilion (MPT) pore through thc volt-
agc-depcndent an ion chan ncl (VDAC), which leads to Ihe 
blocking of cytochromc c release induced by lhc pro-apop-
10lic proteins Bax alld Bid and proleclion or cancer cell s 
from apoptosis [66, 67]. In turn , inactivat ion 01' cyclophilin 
D, a mal rix componcnl o f Ihc MPT pore, induces Ihe release 
of 1I KII from milocho ndria and cnhances Bax-media led 
apo ptosis in cancer ce lls [68]. 
Mitochondrial IIKI and IIKII have prcfercn lial acccss to 
ATP produccd by oxidalive phosphorylation bccause of thci r 
 
 
 
16 
 
TI/lll llr GlyC(l lp 'L\', NI F·l a IIml Nyp(lxitl Mitli -Hl'I'il'II'S il1 M {'(lil'Ílla l Chelll i.wrJ', 200 9. VIII. 9, No. 9 109 1 
Tllbk 2. Isoforms ofGl ucosr T rllll spor l(' rs lln d G lyco ly1k Enzy lll (,s Ex p r('ssrd in Huma n T umo rs 
Isofor m s Typ('S o ftulllor 
~ ro >. = o " 
o 
3 .::! ::; " " " 
~ = .: ~ " = ¡; d ·2 = " E t .. = o " t t· 1': z el> o ..l :;: > ;. .,; ; :i -= . 
~ ~ . -:;: 
~ 
o; .. " o == ..l ¡; ~ Z '" 'E " :J ~ ..l '" U .;¡ > u .:.¡ .... :::': .:;, .::: " ¡:: l:. 
'" 
o ¡;; ~ U 
GLUT I X X X X X X X X X X X X 
GLUn x x x x X 
HKI X X X 
HK II X X X X X X X X X X X X 
HPI X X X X X X X X X X X X X X 
PFK-L X X X X X X X X 
A L D-A X X X X X X X X X X X X X X X 
TPI X X X X X X X X X X X X X X 
GA PDH X X X X X X X X X X X X X X X X X X X X 
PG KI X X X X X X X X X X X X X X 
PGAM-B X X X X 
ENO-a X X X X X X X X X X X X X X X 
PYK·M2 X X X X X X X X X X X X X X X X X X X X 
LDH A X X X X X X X X X X X X X 
PFKFBP3 X X X X X 
Mfe. thcre are not rcpons 
. . .. Data taken fmm (55, 6 1, 11 3. 114. 165, 166] , Mg. Manmlary gllUld; EOOo, enoomClmulI , l llN . head aOO neck, LN, Iymphauc ood uJes, NS, nervous systCm; RL, retlcular lymphOllla , 
BM, bone marrow, 
proxima l locat ion to mi lochondria (Fig. 2) [64] and, as a 
result , are rcpo rted ly less sens ilive to inhib ilion by G6P [69] . 
I lowever, rcsults from our laboratory have revealed strong 
G6P in hibition of bOlh mitochondria l and cytosolic I JK [62] 
when enzyme aClivily was assayed under near-physiological 
condil ions (37°C, 1'" 7 and concenlTations of glucose and 
G6 P ~ 1 mM). The G6 P concenlralion has been reported al 
0 .6-5 mM in lumors [621 and Ihe inhib ilion conslant (Ki) or 
le,. va llles fo r I I K vary bel\Veen 20 and 2 10 !!M [63, 70]. 
Consequently, Ihe II K activity would be predicled 10 be 
strongly G6 P-inhibited under sucll concl itions (Fig. 2). Fur-
therlllore, G6P ( 1 111 M) induces the re lease o f mi tochondri -
ally bou nd IIK in both malignant and non-ma lignant cells 
[7 1, 72], lIence, IIK wOll ld be predominantly free in the cy-
tosol in cancer cell s with high [G6P] such as AS-30D hepa-
tocarcinoma (G6P ~ 5 mM), whereas in 11Imors with lo w 
G6P slleh as lIe La ce lls (G6P=Q.6 111 M), IIK l11ay be pre-
dominantly boulld 10 mitochondria l extemal membrane. 
It should also be po inlcd OUI Ihat, in sOllle slud ics, Ihe 
relative Icvels o f II KI and II KII activity in cy toso lic frac-
tions have very Iikely been under-estilll atc<l , because the 
ATP concell tration used (3-5 111 M) \Vas not saturat ing g iven 
Ihe Km values of 0 .4- 1 mM. In order 10 corrcctly eSlimate 
IIK aClivity ( Vmax), at leasl lO times the Km value (~ 10 
mM ATP) should have been lI sed f'or these kinelic assays. 
This provides an addit ional uncen ai nty in interpre ting Ihe 
dala from Ihe stud ies 01" olhers when detenllining Ihe overall 
ra tio 0 1' IIK and relal ive contributions from cytosolic versus 
rni lochondrial aClivity. 
I1 KI and IIKII b illding lo mitochondria inhibi l apoptos is 
and cnsure Ihal milochondria l ATP is prefe renlially ll sed for 
hexose phosphorylation, Ihereby contribut ing 10 lhe survival 
advantagc o f tumor cells. Th is regulalory Illechanism o f tu-
mor 11 K supporls an essenti al ro le for Ihe ell zyme in Ihe con-
lro l of Ihe g lyeo lylic nllx [62] . Morcover, II KII over-
express ion promotes enhanccd glycolytic flu x because 1I KII , 
together with GLUT, cxerts Ihe main control on the glyeo-
Iylic Ta te in tumor cells [62], Therefore, mitochondrial I rK I 
and 11 make att ractive targelc; for therapeulic interven lion 10 
suppress lumor growlh. 
Apparenl speci fic inhibilion 01" II K by 3-bromopyruvate 
(Table 1; Fig. 3) has beell reporIed [73]. lIo lVever, Ihe eylo-
loxic aClivity againsl cancer ccl ls \Vas of low potency (lCso-
uM) [74, 75] and other glycolytic (150 uM induces 70% 
inhibition of GAPDH and PGK) (Table 1) and mitochondrial 
(PDH, SDH, glutamate dehydrogenase, pyruvate transporter) 
proteins [76, 77], as well as the mitochondrial proton leak, 
are also sensitive to similar low concentrations of this com- 
pound [77]. Clotrimazole (Fig. 3) induced HK detachment 
from mitochondria in B16 melanoma cells, but also detached 
PFK-1 and ALD from the cytoskeleton in mouse LL/2 Lewis 
lung cancer cells, leading to diminished G6P, F1,6BP, and 
ATP levels, and glycolytic flux [78]. Clotrimazole reduces 
cellular proliferation and viability of human CT-26 colon, 
Lewis lung and breast MCF-7 carcinomas (1Cso= 50- 80 uM) 
[78, 79]: and the size and development of intracranial glio- 
mas (C6 and 9L), prolonging survival in rodents [80]. 
Clcarly, although more specific HK inhibitors are required, 
some of the already known HK inhibitors might assist treat- 
ment by sensitizing cancer cells to other anti-cancer drugs. 
Hexosephosphate Iso merase (HPI) 
HPI is a homodimer with 63 KDa subunits and with no 
isoforms (Table 1) [81]. HPI is over-expressed under hy- 
poxia through a mechanism not directly dependent on HIF- 
la [54]: HIF-1 specific consensus sequence for binding in 
the HPI gene has not been described (Fig. 2). Moreover, the 
enzyme does not seem to exert significant flux control over 
glycolysis [62], and hence HPI over-expression in cancer 
cells (Table 2) may be related to its other les 
functions. In addition to participating in glycolysis, HPI also 
promotes cell migration, proliferation and metastasis [82]. In 
tumor cells, HPI is expelled to the extracellular space where 
acts as a cytokine (i.e., autocrine motility factor, AMF) thus 
altering several functions including tumor progression and 
metastasis. 
HPI is inhibited by erythrose 4-phosphate (ERI4P) and 
F1,6 BP (Ki values of 0.7 and 170 uM, respectively) [83: 
Marín-Hernández A, Moreno-Sánchez R, Saavedra E, manu- 
script in preparation], which may exist at relatively high 
concentrations inside tumor cells (16 ÚM and 10-25 mM, 
respectively) [62, 84]. Hence, HPI modulation by ERI4P and 
F1,6BP can be proposed as one mechanism for limiting ex- 
cessive flux through the glycolytic pathway, regulating the 
supply of G6P for the pentose phosphate and glycogen syn- 
thesis pathways. 
 
  
2-DOG is a glucose analog recognized by glucose trans- 
porters, phosphorylated by HK and dehydrogenated by glu- 
cose 6-phosphate dehydrogenase (G6PDH). However 2-DOG 
(as 2-DOGG6P after HK phosphorylation) is not isomerized 
by HPI, which it inhibits, thereby diminishing glycolytic flux 
(Table 1). In addition, 2-DOG effectiveness is drastic: 
reduced in the presence of glucose, due to the comp 
for GLUT, HK and G6PDH. The 2-DOG primary effect is 
on HPI because HPI inhibition by 2-DOG6P (Ki=5 mM) 
[85] induces, in turn, the accumulation of G6P, which is a 
more potent HK inhibitor (Xi=20-200 uM [63, 70]) than 2- 
DOG (Ki20.4 mM) and 2DOG6P (Ki4.3 mM) [86]. In con- 
sequence, glycolytic flux in the presence of 2-DOG dimin- 
ishes because HK is inhibited by G6P and HPI is inhibited 
by 2-DOG6P. 
  
Interestingly, 2-DOG is more toxic for osteosarcoma p” 
cells than for the parental osteosarcoma cells (1C5p values of 
Marín-Hernández e al. 
32-100 uM and 0.6-6 mM, respectively), presumably be- 
cause the p? cells are only dependent on elycolysis for ATP 
generation [87]. 2-DOG (500 mg/kg weight) does not exhibit 
anticancer activity in osteosarcoma nude mouse xenografts 
and non-small cell lung cancer [88], which may be because 
they rely more on OxPhos for their energy supply. However, 
2-DOG significantly enhances the anticancer activity of 
etoposide, camptothecin, and Hocchst-33342 in a range of 
other cancers, including cerebral glioma BMG-1, squamous 
carcinomas 4451 and 4197, and malignant glioma U-87 cells 
[89]. It is possible that cells treated with the other agents 
become heavily reliant on glycolysis and therefore become 
more sensitive to 2-DOG. Combining 2-DOG with adriamy- 
cin, paclitaxel or ctoposide diminishes the size and prolifera- 
tion of human osteosarcoma, xeno-transplanted MV 522 lung 
carcinoma and Ehrlich hepatoma-bearing mice in compari- 
son with tumors treated with 2-DOG or anticancer drugs, 
separately [88-90]. This increased sensitivity towards anti- 
cancer drugs induced by 2-DOG is attributed to the high gly- 
colysis-dependence of the tumor for ATP supply and may 
result from increased demands for ATP made by the cell 
damaging agents. 
    
2-DOG also affects protein glycosylation, induces accu- 
mulation of mis-folded proteins in the endoplasmic reticu- 
lum, leads to a decrease in the amount of HK associated with 
mitochondria and induces the expression of P-glycoprotein 
[91, 92]. Therefore, the drug is not a specific glycolysis in- 
hibitor and its anticancer activity may as a result be limited. 
Phosphofructokinase Type 1 (PFK-1) 
PFK-1 is a homo- or hetero-tetramer of 380 KDa, with 
three isoforms (Table 1). The main isoforms expressed in 
liver and platelets are PFK-L and PFK-P (or C), respectively, 
whereas skeletal muscle only has PFK-M. A mixture of the 
three isoforms is found in all other tissues [81], but HIF-1a 
only increases expression of the PFK-L isoform [4, 44]. 
PFK-M shows the highest affinity for F6P (K/50.6-2 
mM) and it is the least sensitive to ATP inhibition. PFK-L is 
the least sensitive isoform to inhibition by the Krebs cycle 
intermediate, citrate (ICsy=0.18 mM). PFK-P has the lowest 
affinity for F6P (Kpy5=1.4-4 mM) and is more sensitive to 
citrate inhibition (1Csy=0.08 mM) [93]. Therefore, to in- 
crease flux through this enzyme (and hence increase glycoly- 
sis and ATP synthesis) tumors preferentially over-express M 
and L, over the P isoform, exploiting their reduced sensitiv- 
ity to feed-back inhibition by ATP and citrate. A lower pH 
also inhibits PFK-1 activity, decreasing both the affinity for 
F6P and Vin [Moreno-Sánchez R, Marín-Hernández A, En- 
calada R, Saavedra E, unpublished data]. Due to their higher 
glycolytic flux resulting in lactic acid production, cancer 
cells have a more acidic cytosol and extracellular pH [94], 
which would decrease PFK-1 activity. Hence, it is not sur- 
prising that tumors express greater levels of PFK-1 induced 
by HIF-1a to compensate for the lower pH. 
The role for activators such as F2,6BP and AMP, which 
would also be expected to promote an increased flux vía the 
L and M isoforms, is currently unclear because unfortu- 
nately, detailed kinetic studies on PFK-1 are scarce. The 
systematic kinetic analysis of AS-30D and HeLa PFK-1 are
 
 
 
17 
 
1092 Mi"i-RI'I,j{' ll's i" M l'lJit'Ímli C/u'mislr)' . 200 9, IIvl . 9, No. 9 
50 ~M ) 4, ] d l er l eo lylie ISO ~ M ces  
ibili n 01'  PDII d K) able ) d il chondria l 
I I, II , l l alc ydrogenasc, ruvatc I sporler) 
r leins 6, 7], s e ll s I e Íl chondrial r l n Ic k, 
re l o ns iti c 10 i ilar o  centra lions f I is -
und 7]. lolr azo le ig. ) eed II K l enl 
 il chond ria   16 e l a e ll s, ul l  el hed 
- I d  LD  I e l skeJclon  ouse LI2 e is 
g e cer ls, e i g 10 i inished 6P, I , BP, d 
 TP cls, d l co lylie nllx ( ]. l t azo le uces 
l lar r liferati n d i bili ty f an T-26 l n, 
ewis g mi reas t CF-7 as I so  0-  Jl ) 
8, ]; d e i e mI el en t f t ranial lio-
as 6 d L), r l ging r iva l  enls 0] . 
lcarly, l gh o re eci fí e IIK ibitors re uired, 
e f e l y n 1I K ibilOrs ight ss ist at-
ent y siti i g ncer ls 10 t er t i cer rugs. 
l'lcxoscphospha lc o cnl SC P I) 
11 PI   odimer ith  Da bunÍls d Ílh o 
s ab lc ) 1]. JIPI  cr-c pre scd dcr y-
xia gh  echanis  o t irec tly endent n I IF-
. 4] ; I I F-I cci fí c nscnsus cnce r i i g  
I e 1I PI ne as OI ecn escri ed ig. ). oreovcr, e 
e ocs OI   ert i nifi ant n x nt ro l e r 
l co lys is 2], d nce JIP I er- pre s ion  cer 
ls able ) ay e l t d  Íl  ther ss well-k nown 
ll ctions.  dition  arti i a ti ng  l co lysis, IIPI lso 
r o tes ll 1l1 igrat ion, r lif ra ti n d 1l1etastasis ].  
o r ll s, P I  pe lled  e trace llular ace here 
CIs s  t kine ., lOcrine o t lity tor, MF) s 
l i g era I cti ns n l ing l or r gre sion d 
etast4:ts is. 
11 PI  i iled y l rose osphate I4 P) d 
I,6 P i l es 01" .  d 0 ~ M, cclive ly) 3; 
arín- I lemández , oreno-Sánchez , aavee!ra , allll -
ript  r paration], hic h ay ist t l t e ly i h 
ncenlrat ions ie!e l or e ls 6 11  d -25 M , 
celively) 2, ]. I len ce, IIPI od ulalion y RI4P " d 
 1 , P n e osee! s ne ll1echanis  r i Íl g -
ss ive x gh e l co lylic l ay, ulat i g I e 
pply f 6P r e ntose osphate d l gen n-
esis t ays. 
G   l cose al g gnized y l cose l s-
orlers, osphorylaled y I IK e! e! ydrogenated y l -
se - osphate ydrogcnase PD II). Ilo\Vever G 
s O 6 P fl er II K osphorylalion)  ot s ll1erized 
y I [PI, hich l ibi ts, h r y lll in ishing l co lytie n  
able ).  dilion, G c ffc t ne s  rasti ca ll y 
ced  e r s nce f l cose, e 10 e pet ition 
r LUT, JlK d 6PDJI . IlC G r ary rfeCl  
n lIP I ccause JIP I i i li n y G6P ( Ki ~ 5 ) 
] uces,  m , e ulation f 6P, hich   
ore Olenl lI K ibilor ( Ki ~ 2 0 - 2 00 ~M 3 , 0]) l n -
G ( K i ~ O .4 M) d G6P i=l.3 M) 6].  n-
uence, l colytic x  e r sence f G in -
cs cause IfK  i ilCd y 6P d IfP I  i ited 
y G6P. 
t res ti ngly, G  ore ic r e a O 
ls n r e arenla l c a ls l so lues f 
ad" -lIL,,,,,iI1lJe: N ul. 
-1 0 J.l  d . -6 M, p clivcly), r ably e-
se e O ls rc nly ependcnt n gl co lys is r TP 
neration 7] . 2-D  0 Illglkg \Veighl) es Ol hibil 
l cer C livily  t C01l1a dc ouse nografts 
d n-small ell g cer 8], hich ay e cause 
y ly ore n x Phos r I eir ergy ply. Ilowever, 
G ifi nlly ances e l cer ti ity f 
l oside, plo lhecin , d lIoechst- 3342   ge f 
t er e cers, n l i llg rebral l ll1a G -I , ous 
as 5 1 d 197, e! alignanl l o a -87 e ls 
9]. t  sib le al e ls I t d ith e l er ents 
e eavi ly li nt n l colysis e! h r fore  e 
ore siti e 10 OG. ombining G ith e! ri Ill Y-
i , acJitaxe l r e os ide i in ishes e i e d rolifera-
n f an r a, no-Iransplamed V5 2 g 
a d hrlich al a-bearing ice  p ari-
n i th l ors tcd i th G r t i ncer rugs, 
a rately -90]. 11lis r ased ns itivity IO ards li-
cer rugs n uced y G  t led 10 e i h ly-
l i d cnce f e or r TP ply d ay 
sult r J11 sed ancls r  TP ade y I e ell 
aging ents. 
G l  ff cts r le in l cosy lat ion, ces cu-
ulatioll f Illis-f lded r teins  I e oplasmic ti ll-
, s   crease  e ount f I IK ciated ith 
it chondria d ces e pre sion f - l coprolei n 
1, 2]. herefore, e e!r g  Ol  ecifíc l co lys is -
i il r e! L'.i li cer ti Íly ay   ult e i ited. 
osphofruclokin ase l' pe I K- I) 
FK-I   o- r el r -t l er f 0 Da, ith 
ree s ab le ). 111e ain s r s pre sed  
er d l lel els re K-L d K- P r ), eclive ly, 
hereas e lela l use le nly as -M.  ix ture f I e 
rec s nlls  d  l l cr ss cs 1], ul 1 1 F- I a 
ly cases pre s ion f I e - L r  , 4]. 
-  s l e i hesl lTi ily r P o.rO .6-2 
) d   I c s l s iti c  1\ TP ibition. -L  
c st sitive s r   ibiti n y e rebs c le 
!c cd ia te, it r tc so 0.18 M). - P as I e cs t 
ff ity I' r 6P 1<0.  1.4-4 ) d  ore siti c  
t tc i iti n I o 0.08 ) 3] . hcrcfore,  -
ase l x l gh I is e d cnce r asc l eo ly-
is d TP nthesis) u ors r l'erenlia ll y er-ex pre s  
d , cr c  f r , c plo iti g I cir c uced siti v-
l  10 ack ibition y TP d i rate. 1\ o er l l 
l  ibits K- I tivity, e! creasing th e m ity r 
6P d m oreno-Sánchez , arín-Il crnández , n-
l da , avedra , lI published ata]. ue 10 eir i her 
l co ly ti c n  ulti g  lie ie! r d uction, nce l' 
ls ave  ore i ic lOsol d X lra e llu lar ll 4], 
hich ou le! ccrease FK-I Clivity. lIence, t  o t r -
risi g at u ors pre s r ater Ic cls f K-I uced 
y 1  I F ~ 1a. 10 pensale r I e er i l. 
he le r Cli a lo rs ch s , P d MP, hich 
ou ld l o e p ctcd 10 r o te n ased n x i  e 
 d  s r s,  tTenlly clear ca use fo rl U-
ately, etailcd i elic t i cs n - I re rce. he 
t atic i etic a lys is f S-30D e! lIeLa - I  
 
 
 
18 
 
TI/lllor (il)'w l)'si, .. 1-11 F·I o. (lml I-I)'poxi(l 
clI ITcnt ly lInder investigation in ollr labora tory [Moreno· 
Sánehez R, Marín-I lem ández A, Encalada R, Saa vedra E, 
lInpublished dala]. The resu lts have shown Ihal the Ki values 
for A TP and citrate are 1.7 and 4- 17 mM , whereas the Ka 
vallles fo r F2,6B P and AMP are 0.1-33 ~ M and 0.4-3 mM , 
respcctivcly, in a K+·basecl l11ecliulll ; K+ is also a PFK·I acti · 
vator wi lh a Ka 01' 11.5· 13.5 111M . Usuall y, Ihese affíni ly 
consta ntli ha ve been cletcrmined in reaclion medi ul11 with no 
K- . 
The physiological concentra lions of ATP ( 1.4-9.2 111M), 
AM P (0.15-3.3 mM), ci trate (0.4- 1.7) and F2 ,6 BP (5-50 ~ M) 
prcsent in tumor cells [62, 95-97] would indicalc Ihat ATP 
inhibition is l110re likely 10 be relevant than ei lTate inhibition , 
and Ihat F2 ,6BP would prevail over AM P aC li valion. Fur-
Ihcnnorc, il has bccn cs tab li shcd thal PFK-I aC livalion by 
F2,6BP ovcrcolllcs ATP and ci lTatc inhibition [62]. IIcnce, 
the high F2 ,6B P levels present in caneer eells [96, 97] does 
nOI supporl a major role ror activa ted PFK in the nux-conlrol 
o f glyco lysis [62] . On Ihe o lher hand, PFK-I in tumors w ilh 
low expression or low F2,6BP cont enl Illigh l still exert sig-
nifican l nux con lro l ol' glycolys is [Marín-Ilernández A, Mo-
reno-Sánchez R, Saaveclra E, manuscript in preparat ion]. 
Ald olase (A LD) 
A LD is a homo-Ielramer of 40 KDa subllnits, with three 
isoforms (Table 1). A LD-A, B, and C preclomina le in skcle-
lal l11usclc, liver and brain, respcclivcly . Combinat ions of Ihe 
Ihree iso l'orms are fOllnd th rollghout all ti ssucs [8 1]. 
ALD-A ancl C are more effí cient ( 10-20 limes) Ihan B in 
Ihe fo rward (g lycolylic) reaclion [98J. ALD-B shows higher 
affin ity for G3 P and DI IAP, which racili lates the reverse 
reaclion. TIlUS, ALO-A and e are preferen lially localized in 
lisslles wi th h igh glycolysis such as skelelal l11uscle, ery Lhro-
cyles and bra in, whereas A LD-B is in gluconeogen ic ti ssues 
such as liver and k idney. Therefore, II IF-I a up-regulates the 
express ion of A LD-A and e (Fig. 2) [4, 44] with ALD-A 
predominan Il y expresscd in Lumors (Table 2) [6 1]. 
T riose ph osp ha te ISO ll1enlSe (T PI ) 
TPI is a homodimeric enzyme of 27 KDa subunits with-
out isofonn s (T abl e 1), a lthough posl-translational regula Lion 
has been deser ibcd [8 1] . TPI gene is a 111 F-I n target [5 1] 
(Fig. 2). Enzyma lic ac tivily of T PI is onc 01' the highcs t 
fOllnd in natllre and in tumors its ac tivity is 6-61 U/mg pro-
lein , whereas other g lyeo lytic enzymes have much lower 
aClivities in Ihe 0.003-0.8 U/mg prolein range [62]. There-
fore, TPI does not cxert nux control o f glyco lys is [62], ancl 
itli eleva lCd content in cancer most Iikely have a still un-
know n fun ction. 
T PI deficiency in patien ts induces an incrcase in DIIAP 
cO l1 centration. TI1C OIIAP acculllu lation fa vors itl) non-
enzymatic decomposilion to methylglyoxa l (Fig. 3), which is 
a highly react ive aldehydc that modifies protcins and ONA. 
Interestin gly, some studies have sllggested Lhat I11c thyl-
glyoxal has anticancer propcrties. In part icular, il inh ibits 
glycolys is and mi loehondria l respimtion in human Icukacmia 
ce lls, bu l not in nonna l ce ll s [99] . Therel'ore, the physiologi-
cal mcan ing ol'an increased expression ofTPI in tumor cc\ ls, 
induced by I II F-Ia, may be related 10 avoiding the acc llmu-
lation of DIIAP, and Ihc gcneration ol' melhylglyoxa l. 
.Hi//i-H('l' i('lI'.'I i// MCllil'illltl C"('misl r)'. 2009. 1'01. 9. NIJ. 9 1093 
C lycera ldehyde-3-P hosp ha le Oehyd rogen:lse (CA POI'I) 
GAPD II is a hOlllo- Lctramcr of' 37 KDa subunits [ lOO] 
wi lh no isofonns, cxcept for a sccond gcne (gapd2) exclll-
sively expressed in spermatozoa [10 1]; posl-translational 
regula tion has also bccn describcd. lis Km va lues for G3 P 
and NAD+ are 240 and 80 J.lM , rcspective ly [Marín-lIemán-
dez A, Moreno-Sánch ez R, Saavedra E, manuscript in prepa-
rat ion]. IIIF-Ia up-reglllates its ex pression. In adclition 10 
glyco lys is and gluconeogeneis, GA PO II participa tes in tran-
scrip tional regulation as a nuclear tRNA ex port pro tein, and 
in replica lion and repair of DNA (acling as umcyl-ONA gly-
cosylase, by removing uracyl res idues). GAPDI I can also 
lllediaLe endocytos is by its intcraclion with lubulin and it can 
bc requircd for progmmmed neuronal cell dealh [82, lOO]. 
I lowcvcr, the role of its nuclear lTans location in canccr de-
veloplllent and grow lh has not yct been cS lablished. 
GA PD II is a house-kceping pro tein lhat is lIsed in nu-
Illerous stlldics as a cy tosolic markcr o r control for protein 
loading in SDS- PAG E. Ilowevcr, th is prolein is over-
expressed to variab le dcgrees in cancer ce ll s, and its sub-
cellu lar loca lization varies wilh Ihe cellular g rowth s ta te. In 
quiescelll ce lls, GAPDII is loca lizcd in the cylOso l bUI, in 
proliferating cells, GA PD II is delected also in Ihe nuclcus 
[ lOO]. Therc l'ore, it is no t id cal 10 use the GAPOII protcin or 
mR A as a load ing control for westem or northcm blo tting 
studics or as a cytosolic marker in studies wi th tumor cell s. 
Gossypol, a po lyphcnolic aldehydc derived from conon 
secds (Table 1; Fig . 3), is an in hib ilOr of GA POI I, but il al so 
inhibi ts other NA O+-dependent enzymes such as LO! I (Ta-
blc 1) and somc mitochondrial dehydrogcnases (e.g., iso -
cit ratc deh ydrogenase). Gossypol can also affcct a number of 
eellu lar functions associaLed wilh ecllula r prolif'era Lion, in-
cluding ion transport , melllbrane propcrt ics, g lyco lysis, res-
pinHion, glucose u pt, ~ k e, and calciulll hOmCOS{¡lSis by inhibit-
ing ca lcineurin [102]. Al 1-9 J.lM, gossypol ind uces growlh 
inhibition of several human cancer cell lines (breast, cervix, 
mclanoma, ovary, and colon) [ 102, 103]. SITuctural data ami 
molecular Illodcling s tud ies have show n the direc t interaction 
of gossypo l with Bcl-2 and Bcl-XL and support its ability 10 
inhibi t Ihe pro-survival ac tiv ity of thesc pro tcins in cancer 
ceJls, prol11oting apoptosis [104]. At low doses (30 I11 gikg), 
the drug reduced Lumor size by 65%, and Illortality was re-
dueed to 8% in nude mice with the human SW-13 adrenocor-
ti cal carcinom a [ 105]. In a phase I clin ical tria l, gossypol 
decrcased glial and adrenal tumor size by 10-50% [ 106]. 
Koningic ac id (heptc lidic acid, avoee tt in) is a sesquitcr-
pene anlibiolic (Fig. 3) that inhibils GA PD II activity by re-
acting wilh the essenlial th iol group (cys leinc 149). The Iypc 
of inhibition is cOlllpetili ve against G3 P (Ki= I .1 J.l M) [107]. 
Al I mglKg, it suppresses tumor growth 0 1' Erlich asci les in 
vivo, whereas at 36 J.lM (10 J.lg/mL), it inhi bits Ihe growth of 
several cancer cell hnes with high glycolysis but not the 
growth o f' nonn al cells lines with low gl yco lysis [ 108] . 
I lowcvcr, koningic acid also shows acute loxici ty in mice (at 
3 1.5 mg/Kg o r abou t 600 ~l g pe,. mousc) and delcterious el"-
f'ects on nonnal cells with high glycolysis (crythrocy tcs) 
[ 108] . 
For o ther GAPD II inh ibilOrs, such as arsenite (AsO/) 
<\nd iodoacetatc (Table 1; Fig. 3), information aboll t Iheir 
 
 
 
19 
 
1094 Mi"i-H('l,it' I\'.\· i" Mn/id" ,,' Cltcmislr)' , 2009, 1'01.9, No. 9 
effects on tumor ce lis are scarce. Arsenate (J 12As04
1
- +-+ 
IIAsO/- at neutral p ll ) can al so block glyco ly ti c flu x be-
cause GA PDII lllay use il, ins tead of phosphate, 10 fo rm 1-
arseno-3-phosphoglycerate, which is spon la neous ly and rap-
idly hyd rolyzed in water back to G3 P with no assoc iated 
synlhes is 01' ATP (Table 1) [fo r fllfl her delails see 109]. 
Phosp hoglycc rate Kinase (PGK) 
PGK is a monomer 01' 48 KDa with two iso forms (Table 
1). PGKI is expressed in all sOlllatic and cancer ce ll s (Table 
2), whereas PG K2 only appears in spermatozo ids [8 1]. I!l F-
la II p-regllla tes express ion 0 1' PG KI ( Fig. 2) [4, 44J. As 
PGK does not have s ignificant fl ux contro l 0 1' g lycolys is, its 
over-expression in tumor ce ll s lllay have o ther runctions. 
Tumor cells secrete PGK, and extrace llular disulfide bond 
reduclion 01' plasminogen by PG K acting as a disulfide re-
ducrose leads 10 production of angioslat in by promoting 
autoprotcolylic clcavage or plasminogcn [1 10]. Angiostalin 
inhibi ts the plasma melllbrane FoF¡-ATPase, normall y pre-
sent in mitochondria but found expressed in cancer cell s. As 
a result cytoso lic acidiricat ion occurs and both angiogenesis 
and Ille tastasis are inhibited. 
Phosp hoglyccnltc M li tase (PGAM) 
PGA M is a di l11 eri c enzyme consisling 0 1' A or B iso-
ronlls (AA, A B "nd BB) (T"ble 1) e"e h reqllir ing 2,3 -
bisphosphoglyeerate (2,3 BPG) as a eo raetor [ 11 1]. PGA M-B 
(hol11odilller of B subunits) shows a highcr affinity ror 3PG 
and 2,3 BPG (KJII ~ 0.5 mM and 25 ~ M) lhan PGA M-A 
(Km=O.S 111 M and 60 J.l M, respectivcly) whereas the aflinity 
fo r 2PG is similar in bo th isoforms (Km=0 .2S mM) [ 112]. 
PGAM-B over-expression has becn reported in liver, lung, 
colon, and l11 amm ary gland tUlllors (Table 1) [ 11 3, 114]; in 
adclition, 1I 1 F- I (l regulation of PGAM -B expression has bccn 
ind icated [50], suggesling that PGA M-B may play an impor-
tan l ro le in Illalignancy. Furlh ennore, it has been reported 
that inereased expression of the IwO PGA M iso ronns favors 
the proliferation and imlllortali za tion or fib roblasls, whereas 
decreased express ion induces premature senescence [115]. 
This observation suggests that PGAM promotes the immor-
lali za tion of cancer cells ra lher Ihan arrccting increased g ly-
colys is, as this enzyme is no t a flu x-controlli ng slep [62]. 
¡ Ience, PGA M is not highly rclevant to the theme 0 1' this 
review. 
Enolase ( ENO) 
ENO is a dimcric cnzymc ro n11 ed from three di rfcrcnt 
subunits o f S2- 1 00 KDa (Table 1). n, e main iSOron11s are 
aa, a~ , ~ ~ , ay and n [ 11 6]. ENO-a (homodimer or a sub-
units) is dist ributed in most lissues whereas ENO-p and 
EN O-y arc expressed prcfcrclll ially in skelclal musclc and 
brain, respcctivcly [11 6]. The three iso fo rms show simi lar 
arrini ly ror 2PG ( K III ~ 30 ~M) [ 117]. E O has an essenlial 
reqllirement ror diva len t metal ion s in the f o ll ow in ~ order 0 1' 
po tency: Mg2+> 2n 2
+ > Mn2+ > Fc2+> Cd 2+> Co-+> i2+> 
Sm3+> Tb3+ [ 116]. 
III F-Ia up-regula les E O-a expressioll [4, 44] 10 signiri-
cant leve ls in several tumor types (Table 1). ENO-a acts as a 
receptor for plasminogen which favors tumor growth and 
metas tasis [ 116]. COIl sidered logclher wilh Ihe role of PG K 
!1Iarlí,- l/ermím(e:. ('/ al. 
discusscd aboye as a fac ilitator ofplasminogen ac tivat ion by 
all toproteo lysis to plasmin, which is involved in catalyz ing 
Ihe degradation o f fi brin aggregates, the cvidence sugges ls 
Ihat Ihe combin ation or lhese IwO g lyco ly tic enzymes is 
like ly to fac ili ta te cancer mctas las is. Again, however, Ihese 
enzymes are nOl high ly rclevant lO lhe pll rpose 01' Ihis re-
v¡cw. 
Pyruv .. tc Kinnsc ( PYK) 
PYK is a homo-tclramcr with four iso fol111S (Table 1). 
PYK- L is localized mainly in liver and kidney (g lll-
coneogenic tissues) and PYK-R is expressed in eryth rocytes. 
These IwO isoforms are encoded by Ih e same gcne (which 
has 12 exons), bul are expressed rrom altemalive promoters 
such thal specifi c promoters ror L and R isororms are in 
exon l and 2, respeclivcly. The two OI her isofo l111S are PYK-
MI , loca lized in brain , heafl and skelelal mllscle, <tnd PYK-
M2, which is exprcssed in cmbryonic and stem cclls, ICllko-
eytes, pla lelelS, and caneer ee lls (Fig. 2). The MI and M2 
iso forms are also encoded by the same gcne through alterna-
tive splicing [S I J. 
PYK-MI is the only iso fonn with no cooperative kineti cs 
in relation to ils subslrate, PE P, but has the highest affinity 
(Km = 0.08 mM) whereas PYK-R ex hibits the lowest affinity 
(Kili ~ 1.4 mM) [ 118]. TIle lhree other iso lo nlls (R, L and 
M2) that ex hi bil cooperalive kinelics are also pOlenl ly acti-
vated by F 1 ,6BP (KlF 0.06-0.4 '1M), an upslream glyeolylie 
intermediary that es tablishes a fecd-rorward regulatory 
mechanism. A T P strong ly inhibits the activity 0 1' the L and R 
iso forms (K i= 0. 1 and 0 .04 mM, respcclivcly) but only 
mildly inhibile; MI and M2 aClivity (Ki= 3 and 2.5 mM, re-
spec tivcly). Phosphorylat ion 01' lhe L ancl R iso fonlls by pro-
tcin kinases rll ll y abo lishes aClivÍly, whereas the MI and M2 
iso forms arc nOI suscepl ible to phosphoryla lion and hence 
are nOI d ircc tly rcgll latcd by the aClion or hormone binding 
[li S]. The kinelie prope rties of lhe M2 iso fo n11 sugges ls thal 
il is highly active in lumor ce lis at phys iologica l concenlra-
lions o r PEP (0.1-0.3 mM), FI ,6BP (0.6-25 mM) and ATP 
( IA- 9.2 mM) [64, 96] and, IhcreJo re, Ihis is nol a conlro l-
ling st ep or glyeo lys is [62]. 
II IF-Ia only up-regulates PYK-M2 expression, which is 
Ihe main iso form found in tumors (Table 2). As this isoform 
is rcla lively inscnsilive to ATP in hibi tion and it is nOl regu-
laled by phosphory la tion, it secms clear thal ile; ovcr-
express ion is favored in tUlllors to allain an enhaneed g lyco-
Iylic n ux. Furthermore, PYK-M2 undergocs a dimer (illac-
tive)- lelramer (aclive) Iransilion (Fig. 2) which is modula ted 
by FI ,6BP and the oncoproleins pp60 -v-src and II PV-1 6 E7 
[119] . PYK-M 2 binds lO lyrosine-phosphory-Iated peptides 
which are induced in growth I'aclor-s timulated cells [ 120] . 
The in leraclion o r thesc peptides and oncoproteins with 
PYK -M2 induces Ihe rclease 01' the allosleri c activator 
F 1,6BP, prollloling PYK -M 2 dilllcrizalion ,~ n d inaclivalion 
[ 119, 120] (Fig. 2); Ihe less ac tive d illler sccms 10 predomi-
nate in cancer cells [ 119]. 
It has bccn proposed that the increase in the inactive cn-
zyme lead 10 accumula tion or ups lTcam g lycolytic interm edi-
aries which are in turn channeled 10 synthesis o f ribose and 
nllcJeic acids, pro teins and lipids esscnt ia l ror ce llll lar prolir-
¿lycolysis, HIF-1a and Hypoxia 
eration [119, 120]. However, no experimentation has been 
performed to support this conclusion. Kinetic analysis of 
PYK, carried out in different rodent and human cancer cell 
lines, reports higher activity than in their normal counter- 
parts, which suggests that PYK in cancer cells is not inactive 
and it is not in a dimeric form. Furthermore, glycolytic flux 
is unaffected in cancer cells by PYK inhibition because this 
enzyme does not exert significant flux-control [62] due to a 
marked over-expression, null ATP inhibition and potent 
F1,6BP activation; in other words, the PYK-M2 tetramer 
might be less abundant than the dimer, but it would still ex- 
ceed the pathway demands. 
Lactate Dehydrogenase (LDH) 
LDH is a homo- or hetero-tetrameric enzyme of 33.5 
KDa subunits with two main isoforms. LDH-A (also LDH-5 
or LDH-M) is abundant in skeletal muscle, LDH-B in heart, 
and five other subunit combinations have been found in other 
tissues (Table 1): LDHI(Ba), LDH-2 (B3A), LDH-3 (B2A5), 
LDH-4 (BA3) and LDH-5 (Ay). LDH-CA is expressed exclu- 
sively in the testis and spermatozoids [121]. 
In glycolytic tissues such as liver and skeletal muscle, 
LDH-4 and $5 (isoforms with high subunit A content) are 
predominant. In contrast, in tissues that consume lactate 
(heart, kidney, erythrocytes), LDH-1 and 2 (isoforms with 
high subunit B content) predominate, because of their higher 
lactate affinity (Km = 4 mM) compared to the LDH-4 and $ 
isoforms (Km = 7 mM) [122]. Not surprisingly, HIF-1a up- 
regulates LDH-A (LDH-5) expression [4, 44] (Fig. 2) favor- 
ing an enhanced glycolytic flux. A high LDH-A level corre- 
lates with aggressive forms of several different tumor types 
[123]. 
In addition, LDH-A over-expression stimulates other 
non-glycolytic functions. GAPDH and LDH-A bind to sin- 
gle-stranded DNA. NADH addition diminishes the formation 
of GAPDH- or LDH-DNA complexes indicating that the 
NADH/NAD” ratio may regulate DNA binding of these gly- 
colytic enzymes [82]. GAPDH and LDH-A constitute the 
transcription factor complex OCA-S, which increases histone 
transcription (H2Bgene) to maintain the replication process 
and function of eukaryotic chromosomes [82]. 
Oxamate and oxalate (Table 1; Fig. 3) are classical LDH 
inhibitors. Oxamate is a competitive inhibitor of LDH that 
inhibits glycolysis (albcit at very high concentrations with an 
ICgo= 80 mM) [124], but which is much more potent as an 
inhibitor of tumor cell growth (ICsy=10-47 uM) [87]. In 
monolayer leukemia cultures, tumor micro-spheroids, and in 
vivo tumor models (mouse melanoma), oxalate and oxamate 
induce apoptosis and cellular death at sub-millimolar doses 
[125, 126]. Unfortunately, in these studies, the inhibitory 
effect on LDH and glycolysis was not determined. Oxamate 
and oxalate are not very specific for LDH, as they also affect 
other glycolytic (PGAM, PYK), and non-glycolytic enzymes 
including transaminases, PDH, pyruvate carboxylase and the 
mitochondrial pyruvate transporter [127, 128]. This makes it 
difficult to define the importance of LDH inhibition in the 
control of the glycolytic flux in cancer cells. 
However, LDH-A knock down in breast cancer cells in- 
creases mitochondrial respiration and decreases mitochon- 
Mini-Reviews in Medicinal Chemistry, 2009, Vol. 9, No. + 1095 
drial membrane potential, compromising the ability of these 
tumor cells to proliferate under hypoxia [129]. The tumori- 
genicity of the LDH-A-deficient cells was severely dimin- 
ished, and this phenotype was reversed by complementation 
with the human ortholog LDH-A protein. These results dem- 
onstrated that LDH-A plays a role in tumor maintenance 
[129], although it remains to be determined whether similar 
knock-down of other glycolytic steps also induces the same 
described phenotype. 
Phosphofructokinase Type 2 (PFK-2) 
PFK-2 is a bi-functional homodimeric enzyme with 52- 
58 KDa subunits. The enzyme, through its kinase and phos- 
phatase activities regulates the concentration of F2,6BP, the 
most potent activator of PFK-1. Therefore, the PFK-2 kinase/ 
phosphatase activity ratio determines the actual F2,6BP cel- 
lular level and the degree of PFK-1 activation. The PFK-2 
activities are oppositely regulated by PEP, a-glycero-phos- 
phate and citrate, and by protein kinase C phosphorylation, 
all of which inhibit the kinase activity and stimulate the 
phosphatase activity [130]. 
There are four genes (pfkfb-1, 2, 3, and 4) in the rat and 
human genomes that encode four different PFK-2 isoforms 
(liver, heart, placenta, and testis, respectively). It seems that 
HIF-1a regulates the expression of the four genes, although 
the specific consensus sequence for HIF-1a binding has only 
been described for pfkfb-3 [53], consistent with the over- 
expression of the placenta-type PFK-2 in a great variety of 
tumors (Table 1). A high content of the placenta-type PFK-2 
promotes an increased level of F2,6BP, because the phospha- 
tase activity of this isoform is very low (0.2 mU/mg recom- 
binant protein), but its kinase activity is relatively high (140 
mU/mg recombinant protein). The placenta-type PFK-2 
kinase/phosphatase activity ratio is therefore about 710, 
which is the highest compared to the other isoforms (0.4- 
4.1). Moreover, the placenta-type PFK-2 kinase activity can- 
not be inhibited by phosphorylation because it lacks the re- 
quired Ser residue [130]. 
The high F2,6BP level in cancer cells [96, 97], brought 
about by over-expression of the placenta-type PFK-2 over- 
comes ATP and citrate inhibition and induces full activation 
of PFK-1 [62] which favors an increased glycolytic fux. 
Monocarboxylate Transporter (MCT) and Plasma Mem- 
brane H'"-ATPase 
Enhanced glycolysis elevates levels of lactate and HP, 
which must be actively expelled from cancer cells to keep 
the cytosolic pH and osmotic balance under control [131]. 
The MCT family consists of 9-14 members from which 
MCT1-MCTA catalyze the reversible co-transport of lactate, 
pyruvate or ketone bodies and H” (Table 1; Fig. 2). Lactate 
extrusion is favored by an acidic cytosolic pH, or an alkaline 
extracellular pH [131, 132]. 
MCT!1 is in all types of tissues, whercas MCT2 is mainly 
expressed in the liver, stomach, skin, kidney and brain, 
MCT3 is exclusive to the retina and MCT4 is abundantly 
expressed in tissues with high glycolysis such as skeletal 
muscle, lcukocytes, testis, lung, placenta and heart. The af- 
finity for lactate and pyruvate (Km= 0.7-28 mM: and 0.1-150 
   
 
 
 
 
20 
 
Tumor (;~rm /J' si. , ' , I- I/;".. / a uml liJJHI.d(I 
t i n 19, 0] . 1I0wever,  peri en tat ion as een 
cr m1cd  ppo rt is cJusion. inct ic alysis f 
K , rricd ut  ifTcrcnt ent d an cer ell 
h s, orts i her Clivity I n  eir nnal u nter-
a rt s, hich ggeslCi at K  ccr c ls  ot cli c 
d   OI   i cric nn, r! cnn ore, l co lyti c n  
 a fcc tcd  ncer c ls y K ibiti n ecause i s 
e ocs ot crt i fi eant nu - on tro l 2] ue 10  
arkcd cr-cxpre sioll, u l TP i iti n d len! 
 I , P eti at ion;  l er ords, e - 2 ra er 
ighl e ss ndant n I e i er, ut t ould ti l -
ed e l ay ands. 
L ~ l c l ~ l t e ehyd r o ge n ~ l se H) 
OI I   Ol11o- r t r -t eri c e f .5 
Oa bunits ith \Vo ai n r s. OII-A lso OI I-5 
r O II-M)  ndant  eletal uscJe, OII-B  ean, 
d e t er bun it b inalions ve cen nd  l er 
l es ab1c ): II (B ), II -2 3 ), II-3 ,A,), 
II-4 ,) d II-5 .¡) . II-C4  pressed clu-
cly  l e sti s d nn atozoids 2 1]. 
 colytic is es eh s Ii cr d eleta l l11usc1e, 
OII-4 d  s nns ith i h buni t  ntcnt) re 
r ominan!.  ntrast,  s es at e etate 
ean, i ney, t rocyles), OII- I d  so nn s ith 
i h bllnit  ntenl ) l11 inate, ccause f thei r i hcr 
l te ffi i ty    M) parcd  e O II-4 d 5 
s s  ~  M) 2 ]. 0 1 s U~3fi s ingl y, I IF- Ia -
ulalCs II-A II -5) pre sion , ] ig. ) or-
  anced l colytic n .  i h II-A el rre-
l s i th gre sive nns 0 1' eral i Terent lU or es 
 23] . 
 <l di tion , OI I-A er-e pre s ion ulates t er 
n-glycolytic ctions. POII d OI I-A i d  i -
l ed ONA. OII d iti n ini shes l e nna tion 
0 1" POII - r OI I-ON A plexes i at i g at e 
OIIINAO+ ti  ay ulate O A i ing fthcse ly-
e lylie es ]. PDII d II-A e ns litu le I e 
scri tion tor plex CA-S, hich r ases i t ne 
scri t ion 112Bgene) 10 ainlain I e l i ti n r ce s 
d ct i n fe aryotic osomes 2] . 
xamate d alate able ; ig. ) re 1a sical OI I 
ib ilOrs. xamate   petiti ve ibitor f OII at 
ibits l colysis l ei t t ery i h cent rations ith  
so   ) 4], ut hic h  uch o re otent s  
ibitor 01" or eH b,.Tfow th so 10-4 7 ~lM) 7].  
onolaycr eu ia lt ures, or icro-spheroids, d  
i o or ode ls ouse elanoma), ala te d lllate 
ce optosis d l ul ar ea th t - i l i o lar ses 
25, 6]. nfo rtunalely,  se d ies, e i it ry 
ffec t n OI I J1(1 l co lysis as ot t ined. xamate 
d alate re t ery ccifi c r O I 1, s ey l o f c! 
t er l colytic M, K), d n-g lycolytic es 
ing ra inases, O l i, r vate xylase d e 
it chond rial ruva te orter 7, 8]. his akes t 
i i ult  fi e e port ance 01" OII ibit i n  e 
ntrol l" e l co lytic n x  cer e ll s. 
I lowever, O I I-A ock n  rea sI ncer  is -
ses itochondrial irati n d crcases it chon-
¡lfi" ;- I'I 'il'lI ',' ;,, !Hl'lliámtl lle",;sl r)', U09, "I. , fI. 9 95 
rial embra ne telll ial, promis ing e ility f se 
or e ls 10 rolif rate der ypoxia 29). he ori -
cnicity f e OII A-def cicnl e l s as c crely i in-
ed, d is enotype as ersed y plemen ation 
ith e an rt log OI I-A rotein. hese sults -
llslra ted at OII-A l ys  le  l or aintcnance 
29], th gh t ains  e et incd hethcr i lar 
k- owll 01' t er l co lytic t s l o ces e e 
escr i ed cnotype. 
hosphofrll ctokinase ypc  K-2) 
-2   i ctional odillleric llle ith -
8 Oa buni ls. hc ll lllc, h  i ase d os-
atase ctivit ies ulates e ncentration f 2, P, I e 
ost tent ti ator f K- I . herefore, I e -2 i asel 
osphatase ! i ity t i  t l ines e t al , P cl-
l r el d e egr e 01' K- I ti ation. 11le -2 
ti it ies re ppositely lated y EP , l er -phos-
hale d i t te, d y rotein i ase  osphorylation, 
l f hich ibil l e i ase Cli vily d lll ulate e 
osphalase C livily 130]. 
h crc re ur enes Pj j - / , , , Qud )  l e l d 
an CIlOIllCS at c ode ur iffcren l K-2 nn s 
cr, eart , l centa , d tis, ectivcly). t s at 
I IF-I a c u latcs e pre sion f I e m enes, ll llgh 
e ecific sensus ence r I IF-I a i cl ing as nly 
een escribed r flj -3 ], ns istenl ith l e er-
re s ion f I e l ta-t pe K-2   reat ariety f 
ors able ).  i h ntent f e l nla-t pc -2 
r otes  sed e l f 2, BP, cause e hospha-
e ti ity 01" I is nn  ry  .  /mg -
i anl ro tein), t  i ase tivity  l t i cJy i h 40 
IllU/mg binan t rotcin). l e l ccn ta-l pe K-2 
osphatase ti ity ti   h r fore out 10, 
hich  e i hest Olllpared 10 I e l cr s nlls . -
.1). oreovcr, l e l enla- t pe K-2 i ase tivilY n-
OI e i itcd y osphorylation ecause  ks e -
uired er e 30]. 
he i h 2, P cl  ncer ls 6, 7], r llghl 
ou t y er-c pre sion 01' e t  type K-2 er-
es  IP d itratc ibi ti n d ces I! t atí n 
f - 1 ] hich ors n sed l colytic nux. 
MO ll o c~ l rb o xy hll e nl ns porte r l\t. C l) ~ l nd Pl: l s m ~ l e m-
ra ne )-I +-A Pase 
nhanced l colysis atcs cls f l te d I r+, 
h ich us t e t i ely pc lcd  cer e ls  eep 
e t solic ll d o lÍc l nce der n lro l 31 ] . 
he CT ily ns ists 01" - 14 embers  h ich 
CTI- Cf4 t l ze e ersible -t spon f tate, 
yruvate r t ne dies d Ir able ; ig. ). actate 
t lls i n  recl y  i ic t so lic ll , r n l ali e 
Xlr ce l ular l 1 31 , 2]. 
CT 1   l cs f es, hereas CT2  Illa in ly 
pressed  I e er, ach, in, iclney lld rain , 
Cn i  c lus ive 10 e i a d CT4  undan tly 
ressed  l s es itil i h l colysis cll s cleta l 
uscle, e ocytes, sti s, g, l centa d ear!. he f-
ity r tate d yruva te = . -28 M ; d . - 150 
 
 
 
21 
 
1096 M ill i-Rl'l'it' lI'S ill Ml,(lic;,U11 Chl' lII islr)' , 2009, Vol. 9, No. 9 
mM, respectively) differs among the four isofonlls with 
MCT4 sho\Ving the 10\Ver amnity [ 13 1, 132]. MCT4 is the 
predominant isoform expressed in so me breast caneer cell 
lines [ 133 ] bu t in oxidati vc tumor cclls such as O ICo-2, 
\ViOr, FaOu, Sil la and PC-3 ccJls and primary human ca n-
eers it is MCTI [1 34, 135]. II IF-laonly up-regulates MCT4 
expression [52], although inhibi tion of MCT 1 dccrcases tu-
mor growth and rcndcrs it scns it ive to irradia tion [135], sug-
gc..o;;ling thal bOlh isoform transporters playa rc1evant role in 
the operation of glycolys is in lumors. 
A sccond sys tcm uscd 10 regulate th c cy lOsolic pll is lhe 
p lasma membrane V -type II+-A TPa se [ 132, 136]. This el1 -
zyme is over-expressed in tumors and is invol ved in the tu-
mor in lerslilium acid ification from pll 6.8 to p ll 6.5 [137], 
which in tum proJ11otc Jlletastasis [1 38]. TIlUS, inh ibition of 
the V- typc A TPasc and cy tosolic acidification can induce 
ccll death and could constitutc a promis ing and novel thera-
peulic approach. For inslancc, inh ibiting V -ATPasc wlth 
macrolide an libiolics, bafilomycins 0 1' concanamyci ns (Fig. 
3 ) [139], or down-regu laling its cxpression [140], induce 
ca neer cell death . B10cking othcr cylOsol ic pI! reglll ators 
such as lhe Na+/lr an tiportcr, lhe MCT fami ly dcscribcd 
abo ve, or Ihc Na+-depend ent Cr /I ICOJ ' exchanger mighl a lso 
be suimble anti-canccr largets and some spccific inhibilOrs 
have been found [1 32 ]. 
HI I'- Ia AND R EGULAT IO N 01' M ITOC IION DR IAL 
ENZYMES, C YTOC I·IROM E C OX IDASE AND 
PYRUVAT E D EI·IYDROGENASE KINASE 
Th is seclion discll sses the two mitochondrial enzymcs, 
cy tochromc c oxidase (COX; complcx IV) (111(1 POK, known 
to be regula led by II1 F- Ia.. COX is the rcspiratory complcx 
tha! consumes O2 and is inhibited by cyanide (Fig. 3), I hS , 
CO, CO2 or NO. COX is a dimer in which each monomer 
comprises 13 subunits. Subunits 1-3 arc cncoded by lhe mi-
tochondrial ONA, are highly conserved, and constitu le Ihe 
ea talytie coreo Subuni l 4 participa tes in the inilial s teps of 
COX assembly and binds A TP, whieh induces COX inhibi-
tion. III F-1a regula les COX s llbunit 4- 1 cx pression, ca using 
an isofonll switch from the usual subunit 4-2 lO 4-1. The nel 
effect is an increase in COX activity, but on ly a s light in-
crease in O2 consumption and A TP lcvels [ 141 , 142], in 
agree lllen t with lhe negligi ble ro le 01' COX in the cont rol 01' 
the rcspiratory flux and oxidat ive phosphorylat ion rates 
[143]. 
The pyruvate dehydrogenase complex (PO l i) is inhibi ted 
by phosphorylat ion in a reac tion catalyzed by POl i kinase 
(POK ). Four POK iso fo llllS have been ident i fi ed in mammal-
ian ce lls (POK 1-4). TIlese enzymes are dimers wlth subllnits 
01' 46 KOa [144J. POKI is expressed ahnost exel usivc!y in 
heart. POK2 is found in heart , skelerul muscle, placenta , 
lung, brain, kidney, pan creas and liver. ¡ Ieart and skelcta l 
muse le also express PDK3 and PDK4 [ 145, 146] . 
POKs phosphorylate three serine residues (s itc 1, Ser-
264; s ite 2, Ser-27 1; site 3, Ser-203) of the PD II-EI a 
subuniL POKI can phosphorylate all three siles whereas Ihe 
other iso follllS only phosphoryla lc sites I and 2 [ 147, 148J. 
IIIF- laup-regulates POKI and has bcen proposed lO playa 
major rolc in Ihe i n ~ l ctivat i on of the POli cnzymc compl ex, 
J\luril1- l-Ií'rmímle:, l' l (11. 
thcreby decreasing pyruvate oxida lion th rough lhe Krebs 
cyelc, oxygcll consumpt ion, and milochondrial fllllction in 
eaneer eells [ 149,150]. Ilo\Vever, this proposal [5,149- 152 ] 
assulllcs tha t pyrllvalc is thc mai n oxidizable substrate in 
hypoxic cancer cclls and that PO l i is thc ratc-limi ling s lep of 
Krcbs cyc!c. Morcovcr, completc PO l i phosphorylation and 
inhibition has not becn demonslrated. Neither has a POK 
induccd significant diminu tion in Ihc ratc of mitochondrial 
respiration or Ox.Phos nor an associaled cn hanccment in gly-
colys is been shown to occur. Oichloroacclate (OCA; Fig. 3 ) 
has been uscd as an inhibitor 0 1' POK as a Illeans to highlight 
thc importance of POK in tUlllor cell metabolislll [152]. In 
this s tlldy, DCA \Vas sho\V n lO induce cancer cc ll dcath by 
increa sing ROS produclion and apoplOsis. 1I0\Vcver, DCA 
probably also alr ecls othcr cellu lar func tions anc! hence, thc 
role 01' POK is less than cen ai n based on these resul to;;. Fur-
Ihcnllore, tUlllor milochondria are able 10 oxidizc scveral 
alternative energy subs tra tes, such as glulami nc, glutamate, 
fa lly acids and ketone boches [49J, at high ra tes and indc-
pcndcntly of POII complcx ac ti vi ty. 
CAN SMA LL DR UG INI'IIBI TO RS 01' ~ II I'- I a ACTI-
VA TlON BE DES IGNE D AN D DEVE LOPED AS 
NOVE L CANCE R T HE RA PI ES? 
Given the obvious importance of III F- Ia activ ity to the 
cn hanced proliferation, promotion and survival of cancer 
ce lls, it fo llows that inhibi tors of IIIF- Ia would likc!y be 
impon ant canccr Iherapies (revic\Ved in [153]). Unfortll-
natcly, many of the substanccs fou nd to in hibit IIIF-la have 
provcn too cylOtox ic lO be uscful as drug candidatos. A con-
siderable effort has been madc 10 idenlify therapeutically 
lI scful 111 F- I a small dru g inh ibitors, many of which are natu-
ral produ cto;; or synlhctic compounds based on natura l prod-
lIets. Among the mOSI reeent inlercsting devc!opmcnts are 
thc manassan tins [ 154] such as manassanlin B (Fig. 3 ), a 
complcx dineo lignan cx tracted from Sallnmls chinel1sis and 
cemllllS, herbs used in Chincse and Korcan fol k mcdicinc. It 
inhibits both the gro\Vth of hypoxic cancer cc lls and IIIF- I a 
ac tiva tion wilh nM ICso values. Unlikc other cOlllpounds tha t 
atl ack hypoxic cancer cells mana ssantin B has very 10\V tox-
¡city, and as such, is alead compound in thc devclopmcnt of 
nc\V non-toxic anti-cancer thcrapcutic agents . The second 
interestlng devc!opmcnt is lhe discovcry lhat cardiac g ly-
cosides, such as digoxin (Fig . 3), are po tent inhibi tors of 
III F- Ia synthcsis (in thc submicromolar range) [155]. At 
low concenlrations, Ihesc drugs have also been shown to 
inhibit lumor growth in vivo. These rcsults suggest that pre-
viously difficult lO trea t hypoxic tUIllOrs wi th high In F- I a 
aC livi ly may now be largetcd. 
CO NCL USIO NS 
111 F-I a is a major transcri ption factor rcgulating the 
gcncs encoding g lyco ly tic enzymcs and transportcrs. li s ac-
tivity is mainly targcted to those g lyco lytie iso forllls that 
increa se path\Vay flux bu t also on other fllnct ions sllch as 
rcgulation 01' gcne transcription, O A repair, ceH ular migra-
lion, invas ion and Illetastas is, and inhibilion of apoptosis 10 
favor tumor dcvclopmcIl t and gro \V lh. For Ibcsc rcasons, 
III F- Ia is a logieal therapcutic larget fo r lhc treatment o f 
eaneer[1 56] . 
Glycolysis, HIF-1a and Hypoxia 
HIF-1a inhibition by cither RNA interference or by the 
drugs vitexin or topotecan (Fig. 3) induces a reduction in 
tumor growth and metastasis [157-159]. Hypoxia depresses 
proliferation of tumor from HIF-1a positive embryonic stem 
cells but, in marked contrast, it does not affect proliferation 
of HIF-1a-deficient (HIF-1 0”) tumors from embryonic stem 
cells [160]. In contrast, HIF-1aó astrocytes can generate 
tumors in the vascular-rich brain parenchyma but not in the 
poorly vascularized subcutancous environment [161]. Thus, 
HIF-1a may have different roles in tumor growth and devel- 
opment. HIF-1a also participates in the developing heart and 
vascular system, in the working skeletal muscle, in the adap- 
tation of ischemic cardiovascular disease, in the female re- 
productive tract and in osteoblast development in addition to 
bcing one of the key transcriptional factors for embryonic 
development and maintenance of the immune system [3, 
162, 163]. Therefore, given the wide-ranging activities and 
potential for HIF-1a targeted drug induced toxicity, it will be 
essential that the multiple functions of this transcription fac- 
tor should be fully elucidated before embarking on clinical 
trials targeting HIF-1a for the treatment of cancer. General 
inhibition of HIF-1a activity certainly promotes pronounced 
side effects [10]. 
  
The complete characterization of the HIF-1a regulated 
mitochondrial proteins and their functions should first be 
undertaken to better understand why certain isoforms are 
preferentially synthesized in cancers and to facilitate the 
identification of the best therapeutic targets. 
From the perspective of flux control analysis, it appears 
that GLUT and HK, but not PFK-1 and PYK, provide the 
best targets for therapeutic intervention at the level of energy 
metabolism in hypoxic and glycolytic tumors. It follows that 
specific, potent and cell permeable inhibitors of these two 
controlling steps of glycolysis may prove to be preferred 
targets rather than HIF-1 a. For specificity, it is also desirable 
that putative drugs should only interact with the tumor pro- 
teins and not with the non-tumor proteins. It may also be 
possible to exploit the more acidic extracellular pH in tumors 
because some compounds such as a-tocopheryl-succinate 
become more potent anticancer drugs at lower pH than at 
neutral pH [164]. For potency, preferred compounds will be 
those with low nanomolar range Ki values and drug design 
should consider that the compound has to penetrate into the 
cancer cells, for which a hydrophobic chemical segment may 
prove beneficial. 
ACKNOWLEDGEMENTS 
The present work was partially supported by CONACyT- 
México grant No. 80534. The authors wish to thank Prof. 
P.K. Ralph for his stimulating and helpful observations. 
AMH was the recipient of CONACyT-México fellowship 
159991. 
ABBREVIATIONS 
HIF =  Hypoxia-inducible factor 
HRE =  Hoypoxic responsive elements 
pVHL = Von Hippel-Lindau protein 
PHDs =  Prolyl-4-hydroxylases 
Mini-Reviews in Medicinal Chemistry, 2009, Vol. 9, No. 9 1097 
ALD 
EPI 
GAPDH 
PGK 
PGAM 
ENO 
PYK 
LDH 
MCT 
PDH 
PDK 
G6PDH 
G6P 
F6P 
F2,6BP 
F1.6BP 
DHAP 
G3P 
1,3BPG 
2,3BPG 
3PG 
2PG 
PEP 
PYR 
Asparaginyl-aspartyl hydroxylases 
Radical oxygen species 
Oxygen-dependent degradation 
Asparagine-containing transactivation do- 
main 
Thenoyltrifluoroacetone 
Succinate dehydrogenase 
Fumarate hydratase 
Membrane permeability transition 
Voltage-dependent anion channel 
Oxidative phosphorylation 
Glucose transporter 
Hexokinase 
Hexosephosphate isomerase 
Phosphofructokinase type 1 
Phosphofructokinase type 2 
Aldolase 
Triosephosphate isomerase 
Glyceraldehyde-3-phosphate dehydrogenase 
Phosphoglycerate kinase 
Phosphoglycerate mutase 
Enolase 
Pyruvate kinase 
Lactate dehydrogenase 
Monocarboxylate transporter 
Pyruvate dehydrogenase complex 
Pyruvate dehydrogenase kinase 
Glucose 6-phosphate dehydrogenase 
Glucose 6-phosphate 
Fructose 6-phosphate 
Fructose 2, 6 bisphosphate 
Fructose 1.6 bisphosphate 
Dihydroxyacetone phosphate 
Glyceraldehyde-3-phosphate 
1,3 bisphosphoglycerate 
2,3-bisphosphoglycerate 
3-phosphoglycerate 
2-phosphoglycerate 
Phosphoenolpyruvate 
Pyruvate
 
 
 
22 
 
TI/lllor G/J l·II~ I '. \ ' i s . IF- Ia (1IllI IIYJHIXitl 
I!l F- la ibiti n y eil er A ce r y I e 
gs it in r t ean ig. ) uccs  ction  
or th d e tastasis -1 59], 11 ypoxia epre ses 
roliferati n f u or lll 1 1 F-I a ositive bryonie Sl  
ls ut,  arked ntrast, t oes ot rreet r li f ratí n 
f IUF -1 a-deficíen t IUF - 1 a./-) ors  cll1bryoníc c  
 is 0].  nt rast, I I - Ia-l - t y tes n cneratc 
ors  e scular-rich rain rna UI o!  I e 
orly scularized uta eous i enl 6 1]. 1l111S, 
I I F-Ia ay ve fI rent l es  or r th d vel-
ent. I IF-Ia o art i i a tes  I e cloping eart d 
scular t , í  I e o rking clelal1l1usc lc, í  I c ap-
í n f sc íc i vascular i scase,  e c alc c-
r duclive ct d  st oblast cl ent  di tíon 10 
ei g ne f e ey ri li na l t rs r c bryonic 
cl elll d aíntcnancc f e ll1u nc t l11 , 
2, 3]. hercfore, i en e ide-ranging C liv ilies d 
otenlial r 111 F-la l eted r g ced icity, t ill e 
en tial at e Illultiple c ti ns f I is r cri ti n c-
r ould c ll y c l i a tcd efo re c bark ing n e li ica l 
ls a ti g 111 F-l a r e r enl f ceL eneral 
i iti n f 111 F-I a ti i ty rt i l y 01110les unced 
e rr elS 0] . 
he plete araclerízat ion 0 1' I e II IF- I a laled 
it chondrial r teins d i r cli ns uld t e 
dert ken 10 e tcr ders lalld hy rt i n so mls re 
refcremia lly l cs ized  cers d  ilit tc I e 
ntifi ati n fl e est I cr culic eK 
r  l e erspect ivc f  Ol1lrol alysis, l pcars 
at LUT d IIK , u t o t K-l d K, r vide I e 
est l ets r I peut ie ent ion t I e cl 0 1' ergy 
l1le tabolisl1l  poxie d l eo lytic 11I11l 0rs, t r l \Vs al 
ec ifie , tent d eell cnneable i i lors f se I O 
ntro l i g l s f l colysis ay r ve  e rcfc red 
els t er I an I1I F-I . or ccifi cily, l  lso s irab le 
l al ta tive r gs uld ly lcracl ilh l c or ro-
l i s d OI ilh I c n-t or rOleins. t l11 ay o e 
o sib le  ploit I e ore i ic X lr ce lu lar 1'11  Ul110rs 
cause e e pollnds ch s t pheryl-s ll c ina te 
e ore Olent ti nccr r gs l cr 1'11 n l 
eutral l l 4]. r t ncy, r fe red pollnds i ll e 
se ith  n01110 lar ge i l es d rug esi n 
lld Ol1sider I m l e pound as  cnetrate l  I e 
cer ells, r hich  drophobic e ical cnt ay 
r ve nc fi cia!. 
LEDGEM ENTS 
he r sent ork as arlially ported y  ACyT-
éx ico r nt o. 534. 1l1e l ors ish 10 k roL 
. . a lph r is l11u lat ing d elpful servalions. 
II as I e c ipienl 0 1'  ACyT-México c Jlo ship 
91. 
BREVIATIONS 
111 F 
II E 
lIL 
IIDs 
11 xia-i ducible t r 
Il y xic c nsive c en ls 
on I lippc l- indau r lcill 
Proly 1-4-h xylases 
illi- el'h'lI's ll Jf t!{licill ,1 OW/Ili.wq, 09. tl/' . o.  97 
AlI s 
ROS 
O-DD 
C-TA D 
TTFA 
SDII 
FII 
Ml'T 
VDAC 
Ox Phos 
GLUT 
IIK 
IIPI 
PFK-I 
PFK-2 
LD 
TPI 
PDII 
K 
M 
 O 
K 
II 
CT 
II 
K 
6P I I 
6P 
6P 
, P 
I, P 
II AP 
3P 
3  
3 P  
 
 
P 
R 
sparaginyl-asparty l droxylases 
adical ygcn ccics 
xygcn -depc ndcnl g radatíon 
sparaginc-containing ra cti at i n cl -
l11 all1 
henoyl Lr in r acelonc 
u cinalc c ydroge llase 
u arate ydra tasc 
cmbranc cnneab l ity r siti n 
ol l e- cpcndcll t i n anncl 
xidative osphorylation 
 lueose sporter 
Ilcxokinasc 
Ilexoscphosphatc s l11crasc 
osphol'ructokinasc c  
ospho l'ruclok inasc I c  
 Idolase 
ri hosphale erasc 
 Iyceralclc ydc-3-phosphate c ydrogenase 
osphoglyceratc i asc 
osphoglycerate utase 
nolase 
yruvale i ase 
ac ta te c drogcnasc 
onocarboxylatc lra ortcr 
yruvatc c drogcnasc plex 
yruvatc c drogenasc i ase 
lucose osphate ydrogcnasc 
l cose osphatc 
r ctosc - osphale 
r ctosc ,  osphate 
ruclOsc ,  i osphalc 
i ydroxyacctone osphalc 
 Iyeer. I   yde-3 -phosphate 
,  i sphoglyccrate 
, sphogl yccra tc 
osphoglyccrate 
- hosphoglyccrate 
osphoenolpyru vatc 
r vat  
 
 
 
23 
 
109M Mi"i-Rt'ú ews ¡" Mn liámtl Chell/ü lt)'. lOO\}. IItll . 9. ¡V tl. 9 
LAC 
ERI4P 
2-DOG 
Lactatc 
Eryth rosc 4-phospha lc 
2-dcoxyglucosc 
REFERENCES 
11 1 
[2] 
[J I 
[41 
[5 1 
[61 
171 
181 
191 
[101 
[1 11 
1121 
1131 
[14 1 
[1 51 
[1 61 
[171 
1181 
[191 
[201 
[2 11 
Knowlcs, HJ .; HaITis, AL. Hypoxia and oxidative stress in breast 
cancer. Hypoxia and rumourigenesis. Breasl Callcer Res., 100 1, 3, 
318-22. 
Lu, C.W.; Lin, S.C.; O leo, K.F.; Lai, Y.Y .; Tsai , SJ. Induclkm o r 
pyru vatc dchydrogenasc kinasc-3 by hypoxia-inducible ractor· 1 
promotes mctabo lic switch and drug resistancc. 1. Biol. Chem. , 
l UOM, 183, 28 1 06- 14. 
Weidemann, A.; Joh n. . oo, R.S. Biology or HIF- I alpha. Cell Deaflt 
DitJér., 200N, /5, 621 ·7. . 
Scmen7.a, G.L; Roth, P.H.; Fang, H.M.; Wang, G.L. Transcrlp-
üonal rcgulaüon or genes encoding glycolytic cnzymes by hypoxia 
- induc ible ractor 1. J. Bio/. Chem., 199~ , 169, 23757-63, 
Dcnko, N. C. Hypoxia, HIF I and glucosc mctabolism in Ihe so lid 
tumour. Nm. Re\'. C(UlCer, 1008, 8, 705- 13. 
Kim, J.W.; Gao, P.; Dang, C. V. Effccts or hypoxia on twnor me-
tabolism. Cancer Melrutasis Rev .. 2007, 26, 29 1-8 . 
Lisy, K; Pect, DJ . Turn me 00 : rcgulating HIF transeriptional ac-
tivity. CeJ/ Deafh Differ. , 2008, 15, 642-9. 
Huang L.E.; Gu, 1 ; Schau, M .; Buno, H.F. Regulation orhypoxia· 
inducible ractor I alpha is mediatcd by an 02-dependent degrada· 
tioo domain 'Ta the ubiquilin-proteasomc pathway. Proc. NO//. 
Acad Sci USA , 1998, 95, 7987-92. 
Zhong, H.; De Marzo, A.M.; Laughncr, E; Lim, M.; Hilton, O.A.; 
Zagzag, D.; Bucchlcr, P.; lsaaes, W.; Scmenza, G.L.; SimOlls, J.W. 
Ovcrcxpression o r hypoxia· inducible ractor l a lpha in oommon 
human cancers and thcir mClaslascs. C(lIIcer Res., 1 9~ , 59, 5830-
5. 
Luodgren, K.; Holm, c.; Landberg, G. Hypox ia 31ld breast eallecr: 
progooSlic aod therapeutic imp lieations. Cell Mol. Life Sci., 2007, 
64, 3233-47. 
Sun, H.L. ; Liu, Y.N.; Huang, Y.T.; Pao, S.L.; Hunng, D.Y.; Guh, 
J. H.; Lec, F.Y.; Kuo, S.C.; Teng, C.M. YC- I inh ibilS HI F- l ex· 
prcssioo io prostalC eanccr cc l1s: eomributioll or AktlNF·kappaB 
signaliog to HIF- Ialpha accumulation during hypoxia. O"eogene, 
2007, 26, 3941 ·51. 
Shyu, K.G.; Hsu, F.L.; Wang, M.J .; Wang, B.W.; Lin, S. Hypoxia-
inducib1c ractor ) alpha regulatcs lung ooenocareinoma cell inva-
sion. ETp. Cell. Res., 2007, 3/3, 11 81-91. 
Talks, K.L. ; Turtey, H.; Gancr, K.C.: Maxwc ll , P.H.; Pugh, C.W.; 
RatclifTe, P.J.; Harris, A. L. Thc exprcssion and distributioo or the 
hypoxia· induc ible raclors HIF- Ialpha and HI F-2alpha in normal 
human lissues, eanccrs, and tumor-associated macrophagcs. Am. 1. 
PlIfhol., 2000, 157,411-21. 
Monsc r, N.; H e1czyosk~ L.; Lundwa ll , A ; PaII llOan, S.; Anders-
Bjartcll . Loca lization of imm unoreactive H1F- l al pha and HIF· 
2alpha in ncurocndocrine cclls or both bcnign and malignant pros· 
tate glands. PrOSf(lle, 2007,67, 12 19-29. 
Vaupcl, P.; Maycr, A. Hypoxia in canccr: signifieance and impacl 
on clin ica l outoomc. Callcer Metastasis Rell., 2\1 07,16,225-39. 
Fong, G.H.; Takeda, K. Role and regulatioo or prolyl hydroxylase 
domain prolCins. Cell Deafh Differ. , 200M , 15,635-4 1. 
Waro, lP. Oxygen scnsors in eonlexl. Biochim. Bioph)'s. Acta , 
200M, 1777, 1- 14. 
Koivunen , P.; Hirsita, M. ; Kivirikko, K.I.; Myllyhalju, J. The 
Icngth or pcptide substratcs has a marited effcct 011 hydroxylation 
by the hypoxia·inducible ractor prolyl 4-hydroxylascs. 1. Biol. 
Chem., 2006 , 28/ , 287 12-20. 
Ehrismann, D.; Flashman, E.; Gcnn , D.N.: Mathioudakis, 
Hcwitsoo, K.S., Ratcliffc, P J.; Schoneld, CJ . Studies 00 thc activ-
ity or the hypoxia-inducible·raclor hydroxylases usiog an oxygcn 
eonsu"1>lion assay. Biochem. J.. 2007 , 401, 227-34. 
Tsa i, A.G.; Johnson , P.c.; Intaglieta, M. ls the distribution ortissuc 
p0(2) homogenoous? Amioxid Redo:c SignaJ., lUU 7, 9, 979-84. 
Cabra lcs, P.; Tsa~ AG.; Frangos, JA ; Intaglicta, M. Role ofendo· 
thelial ni trie oxide in microvaseular oxygen dcl ivcry and eonsump· 
lioo. Free Radie. Biol. Med , 2005,39, 1229-37. 
[2 21 
[231 
[24J 
[251 
126[ 
1271 
[28 1 
[29 1 
[301 
[311 
1321 
[33J 
[34[ 
[35 1 
1361 
1371 
[381 
[391 
1401 
141 1 
1421 
Marí,,·J/ermi mle:. el al. 
Lec, K.; Roth, R.A.; LaPrcs, JJ . Hypoxia, drug therapy and toxic-
ity. Plwm/Oco/. nJer., 20tH , J 13, 22946. 
Faodrey, J.; GOIT, T.A.; Gassmann , M . Rcgul31ing cellular oxygeo 
sensing by hydroxylalion. Cardiwasc. Res., 2006, 71,642-51. 
Appc\hoIT, R.J. ; Tian, Y.M.; Raval, R.R.; Turley, H.; HaITis, A.L.; 
Pugh, C.W.; RatcJirre, PJ .; Gleadle, J .M. Di fferem iaJ runction or 
lhe prolyl hydroxy lases PH 0 1, PH D2, and PHD3 in thc regulat ion 
o r hypoxia-indueible ractor. J. Bio/. Chem. , 200~ , 179, 38458-65. 
Sl i ch~ O.P.; Wirlhncr, R.; KOditz, J .; Spie hnann , P.; Camcnisch, 
G. ; Wenger, R.H. lncrcascd prolyl 4-hydroxy lase domain proleins 
eompcnsalC ror dccrcased oxygen lcvcls. Evidcncc ror an aU loregu · 
Ialory oxygen·scnsing systelll. J. Bio/. Chem., 2006 , 181 , 23482-9 1. 
Klimova, T.; Chande l, N.S . Milochondria l complex [J] regulalCs 
hypoxic aelivation or H1 F. Cell De(Jfh Difler., 20UM, 1.5, 6.60-6. 
Taylor, C.T. Mitochondria and ccllu lar oxygcn scnsmg m the HI F 
pathway. Bioche", J. , 2ooH, 409, 19-26. 
Cov iao, R.; Pardo, J.P.; Moreno-Sánchcz, R. Tight binding or in· 
hibitors to bov ine be l oomplex is indepcndenl or the Rkskc prolcin 
redox state . Conscqucnccs for scmiquinone stabilizat ion in the qui · 
nol oxidation site . J. Bio/. Chem., 2002,277, 48449-55 . 
Pouysségur, 1 ; Mcchta-Grigoriou, F. Rcdox regulation of lhe hy-
poxia·inducib lc ractor. Biol. Chem., 2006, 187, 1337-46. 
Bcl~ EL ; Klimova, T.A.; Eiscnbart, J .; Moracs, C.T.; Murphy, 
M. P.; Budinger, G.R.S.; Chandcl, .S. Thc Qo site of thc mito-
chondrial complcx IU is rcquircd ror the transduction or hypoxie 
s lgnahng \ '10 rcacllve oxygcn spccics prodUctloo J Cell BJO/ 
2007, 177, 1029-36. 
Horvat, S.; Bcyer, C.; Amold, S. Errect orhypox ia on the transcrip-
lion panem o r subunit isoronns and lhe kinetics of eylOChrome e 
ox idase in cortical astrocytcs and ccrebcllar ncuron. .. . J. Nellro-
chem., 2U06, 99, 937-5 J. 
euzi l, J.; Dyason, J.C.; Frccman, R.; Dong. L. F.; Prochazka, L.; 
Wang, X. F.; Sche 01 cr, l.; Ralph. SJ. Mitoeans as anti-caneer 
agcnlS targcling mitoehondrin : lessons rrom sludies wilh vitamin E 
aoalogucs, inh ibitors of complcx 11 .1. Bioenerg. Biomembr., 2007, 
39,65-72. 
Pitkanen, S.; Robinson, n .H. Mitochondrial eomplcx 1 dcfieicney 
Icoos to increascd production orsuperoxidc mdicals and induction 
of superoxide dismutasc. 1. Clin. h n>e5I .• 1996, 98,345-5 1. 
Dong, L.F .; FreClnan, R.; Liu, J .; Zobalova, R.; Marín.Hcm.ándcz, 
A.; Stantic, M.; Roh lena, J.; Rodrlgue'L.Enriquez, S.; Vahs, K.; 
Butchcr, 8.; Goodwio, J.; Brunk, U.T.; Witting, P.K.; Moreno-
Sánehez, R..; Schemcr, I. E.; Ralph, SJ.; Ncuzil, J. Suppre~sioo or 
tumor growth in vim by the mitocan alpha-Iooophcryl SUCCUlalC re-
quircs respiratory com plex 11 . Clú,. Caneer. Res., 2U09, 15, 1593-
600. 
Habclhah, H.; Lainc, A.; Erdjumcnt·Bromage, H.; Temps!, P.; 
Gershwio, M.E.; Bowtc l ~ 0 .0 .; Ronai , Z. Regulatioo of 2-
OXoglular3lC (alpha-kelOg lutarale) dehydrogcnasc Slabil ily by me 
RI G finger ubiqu itin ligase Siah. J. Bio/. Cltem., 2004, 179, 
53782-8. 
Li, X.F.; Cari in, S.; Urano, M.; Russcll, J.; Ling, C.C.; 
O'Donoghuc, J.A . Visualizat iOll or hypoxia in m;croscop ic Iwnors 
by immunofluorcsccnt microscopy. Calleer Res., 2007, 67, 7646-
53. 
Leo, C.; Horn , L.c.; Eincnke l, J .; HcnlSchel, 9.; HOc kel, M. Twnor 
hypoxia and exprcss ion or c- met in cervical cancer. Gynecol. On· 
ca!., 2007, 104, 18 1-5 . . . . 
Li ao, D.; Johnson, R.S. Hypox ia: a key regulatororanglogenesls 10 
cancer. COllcer Me/rulasis Rev .. 2007, 16, 28 1-90. 
Bonicalz~ M.E.; Groulx, l. ; Paulsen , N.; Lee, S. Role of exon 2· 
cncodcd beta ·domain orthe von Hippel·Liodau lwnor suppressor 
protcin . J. Biol. Chem., 200J , 276, 1407-16. . 
Lu, H.; Dalgard, CL; Mohyeklio, A.; rvk Fatc, T .; Talt, A.S.; 
Varma, A. Reversible inactivation or HI F- l prolyl hydroxylascs al-
lows ecll mClabo lism to control basal HI F-1. J. Biol. Cltem., 2005, 
280, 4 1928-39. 
Sclak, M.A.; Annour S.M.; Mac Kenzie, E.D.; Boulahbel, H.: Wal· 
son, D.G.; Mansncld, K.O.; Pan, Y.; Simon, M.C.; Thompson , 
C. B.; Gottlicb, E. Suecinate links TCA eyele dysfunctioo 10 onco-
gcncsis by inhibiting H1F-a lpha prolyl hydroxylasc. Cancer Cell, 
2005 , 7,77-85 . 
Pollard, PJ .; Bricre, JJ.; Alam, N.A.; Barwell , 1 ; Bare lay, E.; 
Wonham, N.C. ; Hunt, T.; Milchell , M.; Olpio, S.; Moa!, S.1.; Har-
grcaves, I.P.; Hca lcs, SJ .; Chung, Y.L.; Griffi ths, J. R.; Dalgleish, 
 
 
 
24 
 
Tlllllor (ilJwly.,·is, 11/ F- I a (lf/(I IIJpoxüt 
(43 ) 
(44) 
(45) 
(46) 
(47) 
(48) 
(49) 
[50) 
(5 1) 
[52 J 
[53 ) 
[54) 
[55 ) 
[56) 
[57) 
[58) 
[59J 
(60) 
[6 1] 
(62 ) 
[63 J 
A.; McGrath, J.A.; Glccson , M.l ; Hodgson, S.V.; Poulsom, R.; 
Ruslin, P.; Tomlinson , LP. Accumulal ion of Krebs cycle inlerme-
diales and ovcr-exprcssion of H IFlalpha in nunours which resuh 
from gcrml ine FH and SDH mUlalions. IIlIm, Mol. Gener" 2005 , 
/4 , 223 1-9. 
Lec , S.; Nakamura, E. ; Yang, H.; Wei, W.; Linggi, M.S. ; Sajan, 
M.P.; Farese, R.V.; Frecman , R.S. ; Carler, B.O.: Kacl in , W.G. Jr.; 
Sehlisio, S. Neuronal apoptosis linked lo EgIN3 prol y[ hydroxylase 
and familial pheochromocytoma genes: developmenla[ cull ing and 
cancer. Callcer Cell, 1005 , 8, [55-67. 
Rankin , E. B.; Giaccia, AJ . The rolc ofhypox ia-inducib le fac lors in 
lumorigenesis. Cell Vea/h Differ. , 2008, /5, 678-85. 
Palel, S.A .; Simon, M.C. Bio1ogy of hypoxia-inducib le factor-2alpha 
in dcvelopment and discase. Cell Deal/¡ Dilfer. , 2008, 15, 628-34. 
Jang, M.s. ; Park, J .E.: Lec, J.A.; Park, S.G.: Myung, P.K. ; Lec, 
O.H. ; Park, B.C. ; Cho, S. Bind ing and regulalÍon of hypox ia-
inducible factor-I by lhe inhibilol)' PAS proleins . Biochem . Bio-
phys. Res. Comm/lll., 2005, 337, 209-15 . 
Zhong, H.; Hanrahan, c.; van der Poel, H.; Simona, J.W. Hypoxia-
ind ucib le factor l alpha and Ibeta proleins share common signaling 
pathways in hwn an prostale canccr cells. Biochem, Biophys, Res. 
CommuII .. 200 1, 284, 352-6. 
Pedersen , P.L. Mitochondrial evenlS in the Jife and dcath of animal 
cells: a brief overview. 1. Bioellerg. Biomembr., 1999, 31 , 291-304. 
Moreno-Sánchez, R.; Rodríguez-Enríquez, S.; Marín-Hemández, 
A .; Saavedra, E. Energy mctabotism in lumorcclls. FEBSJ., 2007, 
274, 13934 18. 
Takahashi, Y.; Takahashi, S.; Yoshimi , T.; Miura, T. Hypoxia-
induccd expression of phosphoglyccrale mu tase B in fibroblaslS. 
Eur.J Biochem., t 998, 254 , 497-504. 
Gess, B.; Hofbau cr, K.H.; DcUlzman n, R.; Kurtz, A. Hypoxia up-
regulales trioscphosphate isomerase expression I lk¡ a HIF-
dependen l palhway. PJIlIger. .. Arc"-, 2004, 448, 175-80. 
UlIah , M.S. ; Oavies, AJ .; Haleslmp, A. P. TIle plasma mcmbrane 
lactale tmnsporter MCf4, but not MCf I, is up-regulated by hy-
poxia through a H[F-I alpha-dcpendent meehanism. J Biol, Chem. 
20 06,281,9030-7. 
Obach, M.; avarro-Sabate, A. ; Caro, 1 ; Kong, X.; Duran, J .; 
Gómez, M.; Perales, J .c.; Ventura, F.; Rosa, J .L. ; Sanrons, R. 6-
Phosphofruclo-2 -kinasc (Pfldb3) gene promoler contains hypoxia-
inducible factor- I binding siles neceSS31)' for Imnsactivation in re-
sponsc to hypoxia . J. Biol. Chem., 200 4, 279, 53562-70. 
Fun asaka, T.; Yanagawa, T.; Hagall, V.; Raz, A. Regu lation of 
phosphoglucosc isomerase/auloerine motility factor cxpression by 
hypoxia. FASEBJ., 2005, /9, 1422-30. 
Macheda, M.L.; Rogers, S.; Best, J.O. Molecular and cellular rcgu· 
lalÍon of g lucosc transporter (GLUT) proteins in cancer. 1. Cel/. 
Phy .\·iol. , 2005, 202, 654-62. 
Pessino, A .; Hebcrt, D.N.; Woon, C. W.; Harrison, S.A.; Clallcy, 
S.M.; Suxton, J.M.; Carruthers, A.; Czech, M. P. Evidence thal 
fUl1C lionaJ crythrocyte-rype glucose tran sporters are oligomers. J. 
Biol. Chem., 199 1, 166, 2021 3-7. 
Zhao, F.Q.; Kealing, A.F. Functional propcrt ies and genonlics of 
g lucose transponers. ClIrr. Genomics, 2007, 8, 11 3-28. 
Suranl, C.F.; Sell, G .1. Mammalian faci li tat ive glucose transport-
ers: evK1el1Ce for similar substrale reeogn ition sites in fu nctionally 
monomcric proleins. Biochemistry, 1992, 31 , 10414-20. 
Gould, G.W.; TIlomas, H.M.; Jess, T J. ; Bell , G .I. Expression of 
human glueose transporters in XellOpll.f oocytes: kinetic charactcri-
zation and substrate specific itics ofthe crythrocyte, liver, and bra in 
isoforms . Biochemiflry , 199 1, 30,5 [394 5. 
Liu, Q .; Vera, J.C. ; Peng, H. ; Golde, H. TIlc prcdicled ATP-binding 
domains in the hcxose transporter GLUTI e rilicall y affeet trans-
porter aCl ivily. Biochemil'lry, 200 1, 40, 7874 -8 1. 
Alterberg, S.; Greu lich, K.O. Genes of glycolysis are ubiquitously 
ovcrcx presscd in 24 cancer c1asscs. Genomics, 2004, 84, 1014-20. 
Marín-Hemándcz, A.; Rodrlguez-Enríqucz, S.; Vital-González, 
P.A.; Flores-Rodríguez, F.L.; Mac ias-Si lva, M.; Sosa-Garrocho, 
M.; Moreno-Sánchcz, R. Dctenn ining aOO uOOerstanding the con· 
trol o f glyco[)'sis in faSl-growth rumor ce ll s. Flux control by an 
over-cxpresscd but strong ly product.inhibited hexokinase. FEBS 1. 
20 06, 273, 1975-88. 
Wilson, lE. Isozymes ofmammalian hexokinase: struclure, subccJ-
luJar JocaJization and metabohc function. J. Exp. Biol., 2003 , 206 , 
2049-57. 
[64J 
[65) 
[66) 
[67) 
(68) 
[691 
[70) 
[7 1 ) 
(72) 
[73 ) 
[74) 
[75J 
[76) 
[7 7) 
(78) 
[79J 
(80) 
[8 1) 
(82) 
[83) 
M illi-Rt!I'it'l I?>' i" M t!(l icim¡/ Oll'lII islrJ'. 200 9, V"I. 9. No. 9 ) 099 
Pedersen, P.L.; Malhupala, S.; Rempel, A.; Geschwind, J.F.; 
Ko, Y. H. MilochoOOrial bound type JI hexokinasc: a key player in 
the growth and survival of many canccrs and an ideal prospect for 
therapeulic intervention. Biochim. Biophy.f. Acta , 20 02, 1555, 14-
20. 
Xie, G.; Wilson, J.E. Tetrameric strucrure of mitochondrially 
bound rat brain hexok ina se: a c rosslink ing srudy. Arch. Biochem. 
Biophys., 1990 ,276,285-93 . 
Pastorino, J.G .; Shulga, N.; Hoek, J .B. MitochondriaJ binding of 
hexokinase JI inhib its Bax-induccd cylochrome c rcJease and apop-
tosi s.J Biol. Chem. , 2002, 277, 7610-8. 
Majewski , N. ; ogueira, V.; Robcy, R. B.; Hay, N. Akt inhibits 
apoptos is downSlrcam of BID c lcavage vio a glucosc·dcpendent 
mcchanism involving mÍlochoOOrial hexokinascs. Mol. Cel/. Biol., 
200-t, 24, 730-4O. 
Mach ida , K.; Oh la, Y.; Osada, H. SuppressKlO of apoplosis by 
cycJophilin O \ 'k¡ stabilizalion ofhcxokinase 11 mitochondrial b ind-
ing in cancercells.J Biol. Chem., 2006, 281, 14314-20. 
Nakashima, R.; Paggi, M.; Seon, LJ .; Pederscn, P.L. Purificalion 
and eharacterization of a bindablc fonn of mitochondrial bound 
hexokinase from the highly gl)'co ly tic AS-300 ral hepatoma cell 
line. Concer Re .~., 191'18, 48, 9 t 3-9. 
Ardehali , H.; Prin tz, R.L. ; Koch, S.; WhitescJI, R.R.; May, J.M .; 
Granner, O.K. FWlcl ional inleraction bctwecn Ihe N.o and C-
term inal halvesof human hexokinasc 11. 1. Biol. Chem., 1999, 274, 
15986-9. 
White, T.K .; Wilson , J .E. Rat brain hexokinase: location of the 
allosteric regu latory s ile in a struclural domain at the N-tcrminus of 
the enzyme. Arch. Biochem. Bi op hy .~ .. 1987, 259, 402 -[ l . 
Radojkovic , J .; Ureta, T Hexokinase isocnzymcs from !he No-
vikoff hcpatOflla. Purificalion, kinetic and structurnl characteriza-
tion, with emphasis on hexokinase C. Biochem. 1. , 1987, 242, 895 -
903. 
Ko, Y.H.; Pedcrsetl , P.L.; Geschwind, l F. Glucosc catabolism in 
the rabbit VX2 tumor model for liver canccr: characteri zmion and 
ta rget ing hexokinasc . Callcer Lell" 200 1, 173,83-91. 
Ko, Y. H.; Smith, B.L.; Wang, Y.; Pomper, M.G.; Rin i, O.A.; Tor-
benson, M.s.; Hullihen, 1 ; Pcdersen, P.L Advanccd cancers: 
eradicalÍon in all cases us ing 3-brom opyruvale Iherapy to deplete 
AT P. Biochem. Bioph)'!I. R e~ · . CDrll/mm., 2004, 324, 269-75. 
Xu , R.; Pelicano, H.; Zhou, Y. ; Carew, J .S.; Feng, L.; Shalla, K.N.; 
Keating, MJ.; Huang, P. Inh ibition of glycolysis in canccr cells: a 
novel stratcgy lO overcOfne drug res istancc assoc ialcd with mito-
chondrial respiratory defect and hypoxia. Concer Res., 2005, 65, 
613-2 1 
Jones, A.R. ; GilIan , L.; M ilmlow, O. TIle ant i-glycolYlic aCliviry of 
3-bromopyruvate on marure boar spennatoz03 ill vilro. COll tracep-
lion, 1995, 52, 3[7-20. 
Pe reim da Sil va, A.P.; El-Sacha, T.; Kyaw, N.; dos Santos, R.S .; 
da-Silva, W.S.; Almeida, F.C.; Da Poian A.T.; Galina, A. Inhibi-
tion of energy-produc ing pathways of HepG2 ce ll s by 3-
bromopyru vate. Biochem. J., 2009, 4/ 7, 7 17-2 6. 
Penso, J .; Bd tner, R. Dctachment of glycolytic cnzymcs from cy-
toskcleton of Lewis [Wlg carc inoma and co lon adenocarc inoma 
cells induccd by c1otrimazole and ilS corrclatK:lO lo ccll viability 
and morphology. Mol. Geller. Melab., 2001, 76, 18 1-8. 
Meira, 0 .0.; Marinho-Carvalho, M.M. ; Teixe ira, C.A. ; Veiga, 
V.F.; Da Poian , A.T.; Holandino, C.; de Freitas, M.S.; Sola-PeMa, 
M. Clotrimazo le dccrcascs human breast cancer cc[[s viabiliry 
through alterations in cytoskeleton-associated g lycolyt ic cnzymes. 
Mol. Gene/, Me/ab., 2005, 84, 354-62. 
Khalid, M.H.; Tokunaga, Y.; Capu ry, A. J.; Walters, E. Inhib il ion 
of rumor growth and pro longcd surviva[ of ralS with inlracranial 
glklmas following administration of clolrimazo le. 1. Neuroswg. , 
21)05,103, 79-86. 
van Wijk, R.; van Solinge, W.W. Thc energy-lcss red blood cell is 
losl: cl)'throcyte enzyme abnonnalities of glycol)'sis. Blood, 2005, 
106, 403442. 
Kim, J.W.; Oan g, C.V. Multifacetcd roles of glycolytic cnzymes. 
Trends Biochem. Sci. , 2005, 30, 142-50. 
Chirgwin, J.M.; Parsons, T F.; Nolunann, E.A. Mechanislic imp li-
cations of Ihe pH indcpcndence of inhibit ion o f phosphoglucose 
isomerase by neutral suga r phosphales. 1. Biol. Chem., 1975, 250, 
7277-9. 
 
 
 
25 
 
1100 Jlilli-R('l· iell'.\' ill M ellidllltl Chelllislr'}'. 2009, VIII. 9. NI}. 9 
[84 J Reitzer. L J.; Wiee, B.M.; Kennell, O. The pentose cyele. Control 
and essential fune tion in HeLa eell nllClcie acid synthcs is. 1. Bio/. 
Chem. , 1980 , 255. 56 16-26. 
[85] Horlon , R.W.; Mcldrum, B.S.; Bachc lard, H.S. Enzymic and cerc-
bral metabolic effccts of 2-dcoxy-0 -glucosc. J . Nellrochem .. 1973. 
21 , 507-20. 
[86J Manuel y Kccnoy , B.; Zlihner, O.; Malaissc, W.J. Dissociated ef-
fccts of 2-dcoxy-D-glueosc and D-[2-1HJglucosc ami D{5-
lHJglueosc convcrsion inlo lHOH in ral crythrocytcs. Biochem.J 
1992, 188. 433-38. 
[87] Liu, H.; Hu. Y.P.; Savaraj. N.; Pricbe, W.; Lampidis, T,J. Hypcr-
sensitization of twnor cells to glycolytic inhibitors. Biochemistry. 
2001. 40, 5542-7. 
[88] Maschek. O.; Savaraj, N. ; Pruebe. W.; Braunschweigcr. P.; Hamil -
ton. K.; Tudmarsh, O.F. ; De Young. L R.; Lampidis. T,J. 2-dcoxy-
O-g lucose increases the efficacy of adriamycin and paclilaxel in 
human osteosareoma and non-smalJ cell lung cancers in "ivo. Can-
cer Re:.· .• 2004, 64, 31-4 . 
[89] Dwarakanalh, S .s.; Khailan. D.; Ravindra nath. T 2-dcoxy-D-
glucose cnhanccs the eytOloxicily of lopoisomerase inhibi lors in 
human lumor eclllincs. Callcer Bio/. Ther .. 2004 , 3, 864-70. 
[90J Oupta, S.; Mathur, R.; Dwarakanath , B.S. Thc glycolyt ic inhibitor 
2-deoxy-0 -glucose enhances the efficaey of etoposide in Ehrl ich 
ascites rumor-bcaring mice . Callc er Bio/. Ther ., 2005, 4, 87-94. 
[9 1] Kang, H.T.; Hwang, E.S. 2-Deoxyglucosc: an anticancer and ami-
viral therapcutic, bUI nol any more a low glucose mimctic. LiJe Sci.. 
2006, 78, \392-9. 
[92J Lynch, R.M.; Fogarty, K.E.; Fay, F.s. Modulation of hexokinasc 
association wilh miloehondria ana lyzed wilh quanlilalive three-
dimensional eonfoca l microscopy.J Cel/Biol., 199 1. 112,385-95. 
[93J Dunaway. O.A.; Kaslcn, TP.; Sebo, T ; Trapo, R. Ana lysis of the 
phosphofructokinasc subunits and isocnzymes in hwnan lissucs. 
Biochem. 1. , 1988 , 251 , 677 -83. 
[94] Rodriguez-Enriquez, S.; Juárez, O.; RodríguC"/.-Zavala, lS .; Mo-
reno-Sánchcz. R. Multisite con trol of the Crabtree effce l in aseites 
hcpatoma ceUs. Ellr. 1. Biochem .• 2001 , 268, 25 12-9. 
[95] Oonzá lcz-Malcos, F.; 06mcz, M.E.; Gueía-Salguero, L.; Sánchcz, V.; 
Arng6n, lJ. Inhibilion of glycolysis by amino adds in ascitcs tumor 
ceUs. Spccificity and mcchanism 1. Biol. Chem .• 1993, 168, 7809-17. 
[96] Denis , c.; Paris, H.; Mural, J .c. Hormonal control of fructose 2,6-
bisphosphate concentration in the HT29 human colon adcnocarci-
noma cell line. Alpha 2-adrenergic agonl<Jls countcracl cffect of 
vasoactive intestinal pcplidc. Biochem. 1., 1986 , 139. 53 1-6. 
[97] Loiscau, A.M. ; Rousseau , 0.0.; Hue, L. Fructose 2,6-bisphosphate 
and the control of glycolysis by glucocorticoili and by olhcr agenls 
in rat hepatoma cc ll s. Cancer Re.f .. 1985, 45 , 4263-9. 
[98J Pezza , J.A.; Choi, K.H.; Serardin i, T.Z.; Seemink. P.T.; Allen, 
K.N.; Tolan, D.R. Spalial cluslering of l<Jozyrne-specific residues 
revcals unlikcly delenninants o r isozyme spccific ity in rructose-
1 ,6-bisphospha le aldo lasc. 1. Biol. Chem., 2003. 278. 17307- 13. 
[99] Biswas, S.; Ray, M.; Misra, S.; [).¡tta, O.P.; Ray, S. Sclcctive inhibi-
tion of mitochondrial respiratioo aod glycolysis in human lcukaemic 
leucocytes by rre!hy lglyoxal. Biochem. J .• 1997, 313, 343-8. 
[100] Sirover, M.A. Role o f the glyeolytic prolein. glyecrakiehydc-3-
phosphale dehydrogenasc , in nonnal ccll function and in ccll pa-
Ihology. J Cel/ Biochem., 1997. 66, 133-40. 
[101] Welch, J.E.; Srown, P.L.; O ' Srien , D.A.; Magyar, P.L.; Sunch, 
0 .0.; Mori, e. ; Eddy, E. M. Human glyceraldchydes 3-phosphate 
dehydrogcnasc-2 gene is cxpresscd spccifically in spcrmalogenic 
cells.J Andro/. . 2000, 21 . 328-38 
[102] Jaros:zewski, J.\V. ; Kaptan, O.: Cohen, J.s. Action of gossypol and 
rhodamine 123 on wild typc and multldrug-resistant MCF-7 human 
breast cancer cells: 31 P nuclear magnetic resonance and toxicity 
srudies. Ctmcer Res., 1990, 50,6936-43. 
[ 103] Tuszynski, O.P.; Cossu, O. Differenlial cytotoxic cffecl of gossypol 
on hwnan melanoma. colon carcinoma, and othcr tissue culture ccll 
lines. CtmcerRes .• 1984, 44, 768-7 1. 
[104] Oliver, C.L.; Miranda, M.B.; Shangary, S.; Land, S.; Wang, S.; 
Johnson, D.E. (-)-Oossypol ael~ directly on !he milochondria 10 
overcomc Scl-2- and BcI-X(L}medialed apoplos is rcsistancc . Mol. 
Cancer Ther. , 2005, 4, 23 -31. 
[105] Wu , Y.W.; Chik, CL ; Knazek, R.A. An in vi/ro ami in l'im srudy 
of anlitumor effects of gossypol on human SW-13 adrcnocortical 
carcinoma. Callcer Res., 1989 ,49.3754-8. 
!\Itl r li,-Jlerlllí llllc ; el,i1. 
[106) Sushunow. P.; Reidenberg. M.M.; Wascnko, J ; Lorenzo, S .; Lcmke. 
S.; Himplcr, S .; Corona, R.; Coyle, T. Gossypol trcalment ofrccurrcnt 
adult malignant gliomas.J Nellrrxmcol. . 1999, 43, 79-86. 
[107J Sakai , K.; Haswni , K.; Endo, A. lnaclÍvation or rabbil muscle glye-
craldehyde-3-phosphatc dehydrogcnase by koningic neld. BBA-
Pr % ~~ Stmct. Mol. Emy mol., 1988, 951, 297-03. 
[1081 Kumagai. S.; Narasaki, R.; Haswni. K. Olucosa-dependent active 
AT P dcplction by koningic aeid ki lls high-glycolytic cells. Bio-
chem. Biophy:.·. Res. Commlln. , 20 08, 365, 362-68. 
l109J Ralph . S.l Arscnie-based anlincop lastie drugs and thcir rrechanisms 
or action. MelOl &¡:.;ed Drllgs., 200S, 1008, Article ID 260 146. 
[1 10] Lay. AJ.; Jiang, X.M.; Kiskcr, O.; Flynn, E.; Underwood, A.; 
Condron, R.; Hogg, PJ. Phosphoglyccrale kina.<Je acts in rumour 
angiogcnesis as a disulphlde reductasc. Nalllre , 2000, 408. 869-73. 
lJI IJ Repiso. A.; Ramirez-Bajo, MJ.; Vives-ColTons, J.L; Carrerns, J.; 
Climenl, F. Phosphoglyecrate mutase BS isoenzyme deficiency in 
a patient with non-spherocytic anemia: familial and metabolic stud-
ies. Haelll(l/ologica . 2005, 90, 257-9. 
[1 12] Fundele. R.; KrielSeh , W. K. Purifieation and propcrtics of the 
phosphoglyccrale mutase isozymes from the mouse. Comp. Bio-
chem. Physiol. B. , 1985,81,965-8. 
[113] Durany, .; Joscph, 1 ; Campo. E. ; Molina, R.; Carreras, 1 
Phosphoglycerate mutasc. 2,3 bisphosphoglycerate phosphatase 
and cnolase activilY and isocnzymes in lung, colon and ¡ivcr carci -
nomas. Br. 1. Cancer, 1997,75. 969- 77. 
lJ I4J Durany, N.; Joscph , 1 ; Jimcncz. O.M.; Cli ment. F.; Femández. P.L; 
Rivera, F.; Carreras, 1 Phosphoglyccratc mutase, 2,3-
bisphosphoglyccrate phosphatasc, crcatine kinasc and cnotase activily 
and isocozymcs in breast carcinoma. Br. J. Callur, 2000 , 82, 20-7 . 
[115] Kondoh, H.; Llconart, M.E.; Oil, 1; Wang, 1 ; Dcgan, P.; Petcrs, 
O.; M:uti ncz, D.; Carnero. A.; Seach, D. Olycolytic enzymcs can 
modulatc cell ular life span. Callcer Re.\·., 2005. 65, 177-85. 
[116] Pancholi, V. Multifunctional alpha-eno lasc: its role in discases. 
CeJl. Mol. Lije Sei. , 2001, 58, 902-20. 
[117] Shimizu , A.; Suzuki, F.; Kalo, K. Characlerization of alpha alpha, 
beta beta , gamma gamma and alpha gamma human enolase 
isozymes. and preparalion o f hybrld cno lases (alpha gamma, beta 
gamma and alpha beta) from homodimeric fonns. BBA -Pror. 
S/rllcr. Mol. Enzymol., 1983, 748. 278-84. 
l118] Imamur.l, K.; Tanaka, T. Pyruvate kinase isozymes from ral Me//¡-
O(lf EII=ymol., 1982, 90, 150-65. 
rr 191 Mazurek , S.; Orimm, H.; Bosehek, e. B.; Vaupel, P.; Eigenbrodl, E. 
Pyruvatc kinasc Iypc M2: a crossrood in the tumor melabo lome. 
Brh J. Nlltr. , 2002 , 87, S23-9. 
l120] Christofk , H.R.; Vander-Heiden, M.O.; Wu, .; Asara, 1M .; 
Cant1cy, L.e. Pyruvate kinase M2 is a phosphotyros ine-binding 
prolein. Natllre, 2008, 452. 18 1-6. 
[12 11 Drcnl, M.; Cobben , NA; Henderson, R.F.; WOUICrs, E.F.M. ; van 
Die ijen-Visser. M. Uscfulncss o r taetale dchydrogenasc and its 
isoenzymes as indicators of lung damage or inOammalion. Ellr. 
Re;:.;pir. 1. , 1996, 9,1736-4 2. 
[122J Suh l, S.N.; Jackson , K.Y.; Vandcrlinde, R.E. Thc effccl of tcm-
perarure on Ihe kinctic constants of human laCtalC dchydrogenasc 1 
and 5. Clill. Chim. Acta, 1977. 80, 265 -70. 
l l 23J KoukoW'3kis, M.I.; Oialromanolaki. A.; Simopoulos, C.; Poly-
ehronidis, A.; Sivridis, E. Lactatc dehydrogenase 5 (LOH5) relates 
to up-rcgulated hypox ia inducible factor pathway and rnetastasis in 
co lorcctal cancer. Clin. Efp. Metastasis, 2005, 11, 25-30. 
[i 24] Elwood , lC. Effcct of oxamatc on glycolysis ami respiralion in 
sarcoma 37 ascites eclls. Cancer Res. , 1968, 28, 2056-60. 
1I25J Jow, O.M. ; Wu, Y.C. ;Ouh. 1 H.; Teng , C.M. Armcpavine oxatate 
induces ccll dcath on CCRF-CEM Jcukemia cell linc Ihrough an 
apoptotic pathway. Life Sei .• 2004 , 75, 549-57. 
[1261 OOrlach, A.; Ackcr, H. p02- and pH-gradicnts in mulliccllular 
sphcroids and their rclationship 10 cellu lar metabolism and radia-
lion scnsitivity of malignant human lWnor cells. BBA-Mo/. B(l s i . ~ 
Dis., 199 4, 1217, 105- 12. 
[127J Martin- Rcqucro, A.; Ayuso, M.s.; Parrilla, R. Rate-limiling stcps for 
hcpatic gluconcogcnesis. Mcchanism of oxamate inhibition of milO-
chondrial pyruvale melabolism. J. Biol. Chem., 1986.261. 13973-8. 
[128] Seutler, E. ; Forman. L ; Wesl , e. EffeCI of oxa lale and malonale on 
red ecl l metabolism. BloOO, 1987. 70, 1389-93. 
[129] Fantin, V. R.; SI-Picrre, J.; Lcdcr, P. Attenuat ion of LDH-A expres-
sion unoovcrs a link bctween glyoolysis, mitochondrial physiology. 
and tumor mainlcnance. COllcer Cell, 2006 , 9, 425-34. 
1131) 
[132] 
[133] 
[134] 
[135] 
[136] 
1137] 
[138] 
[139] 
[140] 
[141] 
[142] 
[143] 
[144] 
[145] 
[146] 
[147] 
[148] 
[149] 
1 a Trends Biochem. Sci, 2001, 26, 
30-5. 
Halestrap, A.P.; Meredith, D. The SLCI6 gene family-from mono- 
carboxylate transporters (MCT's) to aromatic amino acid transport- 
ers and beyond. Pflugers Arch., 2004, 447, 619-28. 
Izumi, H.; Torigoe, T.; Ishiguchi, H.: Yoshida, Y.; Tanabe, M.; Ise, 
T.; Murakami, T.; Yoshida, T.; Nomoto, M.: Kohno, K. Cellular 
pH regulators: potentially promising molecular targets for cancer 
chemotherapy. Cancer Treat. Rev., 2003, 29, 541-9. 
Gallagher, S.M.; Castorino, J.J.; Wang, D.; Philp, N.J. Monocar- 
boxylate transporter 4 regulates maturation and trafficking of 
CD147 to the plasma membrane in the metastatic breast cancer cell 
line MDA-MB-231. Cancer Res., 2007, 67, 4182-9. 
Martín-Venegas, R.; Rodríguez-Lagunas, M.J.; Geraert, P.A.; Fer- 
rer, R. Monocarboxylate transporter 1 mediates DL-2-Hydroxy-14- 
methylthio)butanoic acid transport across the apical membrane of 
Caco-2 cell monolayers. ./. Nutr., 2007, 137, 49-54. 
Sonveaux, P.; Végran, F.; Schroeder, T.¿Wergin, M.C.; Verrax, J.; 
Rabbani, Z.N.; De Saedeleer, C.J.; Kennedy, K.M; Diepart, C.; 
Jordan, B.F.; Kelley, M.J.; Gallez, B.: Wahl, M.L.; Feron, O.; De- 
whirst, M.W. Targeting lactate-fueled respiration selectively kills 
hypoxic tumor cells in mice. J. Clin, Invest., 2008,118, 3930-42. 
Bowman, E.J.; Bowman, B.J. V-ATPases as drug targets. ./. Bio- 
energ. Biomembr., 2005, 37, 431-5. 
Gerweck, L.E.; Vijayappa, S.; Kozin, S. Tumor pH controls the in 
vivo efficacy of weak acid and base chemotherapeutics. Mol. Can- 
cer Ther., 2006, 5, 1275-9. 
Sennoune, S.R.: Luo, D.; Martinez-Zaguilan, R. Plasmalemmal 
vacuolar-type H+-ATPase in cancer biology. Cell Biochem. Bio- 
phys., 2004, 40, 185-206. 
Bowman, E.J.; Gustafson, K.R.; Bowman, B.J.; Boyd, M.R. Identi- 
fication of a new chondropsin class of antitumor compound that se- 
lectively inhibits V-ATPases. J. Biol. Chem., 2003, 278, 44147-52. 
Lu, X.; Qin, W.; Li, J.; Tan, N.; Pan, D.; Zhang, H.; Xie, L.; Yao, 
G.; Shu, H.; Yao, M.; Wan, D.; Gu, J.; Yang, S. The growth and 
metastasis of human hepatocellular carcinoma xenografts are inhib- 
ited by small interfering RNA targeting to the subunit ATP6L of 
proton pump. Cancer Res., 2005, 65, 6843-9. 
Fukuda, R.; Zhang, H.; Kim, J.W.; Shimoda, L.: Dang, C.V.; Se- 
menza, G.L. HIF-1 regulates cytochrome oxidase subunits to optimize 
efficiency of respiration in hypoxic cells. Cell, 2007, 129, 11-22. 
Campian, J.L.; Gao, X.; Qian, M.; Eaton, J.W. Cytochrome c oxidase 
activity and oxygen tolerance. J. Biol. Chem., 2007, 282, 12430-8. 
Moreno-Sánchez, R.; Saavedra, E.: Rodríguez-Enríquez, S.; Olín- 
Sandoval, V. Metabolic control analysis: a tool for designing 
strategies to manipulate metabolic pathways. J. Biomed. Biotech- 
nol., 2008, 2008, Article ID 597913. 
Roche, T.E.; Hiromasa, Y. Pyruvate dehydrogenase kinase regula- 
tory mechanisms and inhibition in treating diabetes, heart ischemia, 
and cancer. Cell Mol. Life Sci., 2007, 64, 830-49. 
Gudi, R.; Bowker-Kinley, M.M.; Kedishvili, N.Y.; Zhao, Y.; 
Popov, K. Diversity of the pyruvate dehydrogenase kinase gene 
family in humans. J. Biol. Chem., 1995, 270, 28989-94. 
Bowker-Kinley, M.M.; Davis, W.L; Wu, P.; Harris, R.A.; Popov, 
K.M. Evidence for existence of tissue-specific regulation of the mam- 
malian pyruvate dehydrogenase complex. Biochem, J., 1998, 329, 
1916. 
Korotchkina, L.G.; Patel, M.S. Site specificity of four pyruvate dehy- 
drogenase kinase isoenzymes toward the three phosphorylation sites of 
human pyruvate dehydrogenase. J. Biol. Chem,, 2001, 276, 37223-9. 
Kolobova, E.; Tuganova, A.; Boulatnikov, L.; Popov, K.M. Regula- 
tion of pyruvate dehydrogenase activity through phosphorylation at 
multiple sites. Biochem. J., 2001, 358, 69-77. 
Papandreou, L.; Caims, R.A.; Fontana, L.; Lim, A.L.: Denko, N.C. 
HIF-1 mediates adaptation to hypoxia by actively downregulating 
mitochondrial oxy gen consumption. Cell Metab., 2006, 3, 187-97. 
      
Received: 16 December, 2008 Revised: 21 April, 2009 Accepted: 23 April, 2009 
[150] 
[151] 
[152] 
[153] 
[154] 
[155] 
[156] 
[157] 
[158] 
[159] 
[160] 
[161] 
[162] 
[163] 
[164] 
[165] 
[166] 
Mini-Reviews in Medicinal Chemistry, 2009, Vol. 9, No. 9 1101 
Kim, J.W.; Tchernyshyov, 1.; Semenza, G.L.; Dang, C.V. HIF-1- 
mediated expression of pyruvate dehydrogenase kinase: a meta- 
bolic switch required for cellular adaptation to hypoxia. Cell Me- 
tab., 2006, 3, 177-85. 
McFate, T.; Mohyeldin, A.; Lu, H.; Thakar, J.: Henriques, J.: 
Halim, N.D.; Wu, H.; Schell, M.J.; Tsang, T.M.; Teahan, O.; Zhou, 
S.; Califano, J.A.; Jeoung, N.H.; Harris, R.A.:Verma, A. Pyruvate 
dehydrogenase complex activity controls metabolic and malignant 
phenotype in cancer cells. J/. Biol. Chem., 2008, 283, 22700-8. 
Bonnet, S.; Archer, S.L.; Allalunis-Turner, J.; Haromy, A.; Beau- 
lieu, C.; Thompson, R.; Lee, C.T.; Lopaschuk, G.D.; Puttagunta, 
L.: Bonnet, S.; Harry, G.: Hashimoto, K.; Porter, C.J.; Andrade, 
M.A.; Thebaud, B.;: Michelakis E.D. A mitochondria-K+ channel 
axis is suppressed in cancer and its normalization promotes apopto- 
sis and inhibits cancer growth. Cancer Cell, 2007, 11, 37-51. 
Martínez-Sánchez, G.: Giuliani, A. Cellular redox status regulates 
hypoxia inducible factor-1 activity. Role in tumour development. J. 
Exp. Clin. Cancer Res., 2007, 26, 39-50. 
Hossain, C.F.; Kim, Y.P.; Baerson, S.R.; Zhang, L; Bruick, R.K.; 
Mohammed, K.A.; Agarwal, A.K.; Nagle, D.G.; Zhou, Y .D. Sawurus 
cernuus lignans-potent small molecule inhibitors of hypoxia-inducible 
factor-1. Biochem. Biophys. Res. Commun., 2005, 333, 1026-33. 
Zhang, H.; Qian, D.Z.; Tan, Y.S.; Lee, K.; Gao, P.; Ren, Y.R.; Rey, 
S.; Hammers, H.; Chang, D.; Pili, R.; Dang, C.V.; Liu, J.O.; Se- 
menza, G.L. Digoxin and other cardiac glycosides inhibit HIF- 
lalpha synthesis and block tumor growth. Proc. Natl. Acad. Sci, 
USA, 2008, 105, 19579-86. 
Kong, D.; Park, E.J.; Stephen, A.G.; Calvani, M.; Cardellina, J.H.; 
Monks, A.: Fisher, R.J.; Shoemaker, R.H.; Melillo, G. Echinomy- 
cin, a small-molecule inhibitor of hypoxia-inducible factor-1 DNA- 
binding activity. Cancer Res., 2005, 65, 9047-55. 
Takahashi, Y.; Nishikawa, M.; Takakura, Y. Inhibition of tumor 
cell growth in the liver by RNA interference-mediated suppression 
of HIF-lalpha expression in tumor cells and hepatocytes. Gene 
Ther., 2008, 15,572-82. 
Choi, H.J.; Eun, J.S.; Kim, B.G.; Kim, S.Y.; Jeon, H.; Soh, Y. 
Vitexin, an HIF-lalpha inhibitor, has anti-metastatic potential in 
PC12 cells. Mol. Cells, 2006, 22, 291-9. 
Puppo, M.; Battaglia, F., Ottaviano, C.; Delfino, S.; Ribatti, D.; 
Varesio, L.; Bosco, M.C. Topotecan inhibits vascular endothelial 
growth factor production and angiogenic activity induced by hy- 
poxia in human neuroblastoma by targeting hypoxia-inducible fac- 
tor-lalpha and -2alpha. Mol. Cancer Ther., 2008, 7, 1974-84, 
Carmeliet, P.; Dor, Y.; Herbert, J.M.; Fukumura, D.; Brusselmans, K.; 
Dewerchin, M.; Neeman, M.; Bono, F.; Abramovitch, R.; Maxwell, P.; 
Koch, C.J.; Ratcliffe, P.; Moons, L.; Jain, R.K.; Collen, D.; Keshert, E. 
Role of HIF-lalpha in hypoxia-mediated apoptosis, cell proliferation 
and tumour angiogenesis. Nature, 1998, 394, 485-90. 
Blouw, B,; Song, H.; Tihan, T.; Bosze, J.; Ferrara, N.; Gerber, H.P.; 
Johnson, R.S.: Bergers G. The hypoxic response of tumors is depend- 
ent on their microenvironment. Cancer Cell, 2003, 4, 13346. 
Ameln, H.; Gustafsson, T.; Sundberg, CJ; Okamoto, K.; Jansson, E; 
Poellinger, L.; Makino, Y. Physiological activation of hypoxia induc- 
ible factor-1 in human skeletal muscle. FASEB J., 2005, 19, 1009-11. 
Critchley, H.O.; Osei, J.; Henderson, T.A.; Boswell, L.; Sales, K.J.; 
Jabbour, H.N.; Hiranio, N. Hypoxia-inducible factor-lalpha expres- 
sion in human endometrium and its regulation by prostaglandin E- 
series prostanoid receptor 2 (EP2). Endocrinology, 2006, 147, 744-53. 
Neuzil, J., Zhao, M., Ostermann, G., Sticha, M., Gellert, N., Weber, 
C., Eaton, J.W., Brunk, U.T. Alpha-tocopheryl succinate, an agent 
with in vivo antitumour activity, induces apoptosis by causing 
lysosomal instability. Biochem. J., 2002, 362, 709-15. 
Balinsky, D.; Platz, C.E; Lewis, J.W. Isozyme pattems of normal, 
benign, and malignant human breast tissues. Cancer Res., 1983, 43, 
5895-901. 
Atsumi, T.; Chesney, J.; Metz, A.; Leng, L.; Donnelly, S.; Makita, 
Z.; Mitchell, R.; Bucala, R. High expression of inducible 6-phos- 
phofructo-2-kinase/fructose-2,6-bisphosphatase (iPFK-2; PFKFB3) 
in human cancers. Cancer Res., 2002, 62, 5881-7. 
    
    
     
 
 
 
26 
 
T,/lllor G!J(·d!Js is . 111 F .. ¡ a (1m! 1I),I'I.I.\'i" 
(UO] Okar, O.A.; Man7..aJ1O, A.; Navarro-Sabate, A.; Riera, L.!.; Banrons, 
R.; Lange, AJ . PFK-21FBPasc-2 : maker and brcaker of lhe esscntial 
bklfactor fruclosc-2,6-bisphosphalc. Trefld ~ i chem. ei., 01, 6, 
-5. 
(l I] aleslrap, . P.; eredilh , O. he C I6 ne i l - f  ono-
r oxylatc sportcrs Cfs)  alic ino id sport-
rs d yond . fll/gers rch., 04, 47, 19-28. 
( 21 l i, .; origoc, . ; I shig u ch~ .; oshida, .; anabe, .; e, 
. ; urnkam i, .; osh ida, .; omOlo, .; ohno, . e lular 
 c ulalors: otenlia ly r ising o l cula r l ets r cer 
c othcrnpy. Ollcer r ol. eo". , 1 03, 9, 1-9. 
(I 3J a lagher, . .; astorino, .; ang, O.; hilp, J. onocar-
xylale ortcr  u!ales aturat ion d te i g f 
O 147 10 l e l a embrane  l e etastatic r ast ccr e ll 
e DA-MB-23 l . tmcer Res., 00 7, 7, 182-9. 
( 4) aní n- enegas, .; odri uez-Lagunas, J.; eraert,  . .; er-
r, . onocamoxylalc orter I ediatcs OL-2 - ydroxy{4-
cthyllh io)butanoie id sport ero s ! e ical embranc f 
aco-2 e ll onolayers. J. I/Ir., 07, / 7, 9-5 4. 
( 5J nveaux, .; égran, .; ehrocdcr, .; crgin , .C. ; e rax, .; 
a ban i, . .; e acdelccr, J .; c ncdy, . ; icpan, .; 
OI' an , 8.F.; e ley, J.; Oallcz, 8.; ahl, .L.; eron, .; e-
h irst, . . argeting l c\ed irali n lccli e ly i ls 
poxic or e ls  ice. . lin III'es t., 1 8, /8, 30-4 2. 
( 6) owman, .; owman, . - T Pases s r g ets. J. io-
erg. i embr. , 05, 7, 1 -5 . 
( 71 Ocr eek , .E.; ij a pa , . ; ozin, . mor  ntrols e  
;\'O l cy f eru: id d ase e otherapcutics. ol. an-
r er., 06, , 75-9. 
( 8J $ noune, . .; uo, D.; artinez-Zaguilan , . l al mal 
cuola r-typc +- TPase  ccr i l gy. e/J i chem . io-
hys., 04 , 0, -206. 
( 9J o man, J .; Oustafson, .R.; o man, J .; oyd, . R. l nt'i-
ti n f  c  e ndropsin l s f l or pound at -
l ely i ilS - T Pa es. io/. hem., U 03 , 78, 147-52. 
( 0J u, .; in , .; i, 1 ; an , . ; an, .; hang, .; ic, .; ao, 
.; u, .; 30, .; \Van, .; Ou , .; ang, . he r th d 
ctastasis f an cpatocellular i a nografts re ib-
lc  y all lc c g A l eti g  e bunit TP6 L f 
r t n nl'. ancer es ., 005, 5, 43-9. 
( 1] ukuda, . ; hang, .; i , l .; i oda, .; ang, . .; $e-
cnza, O  IF- I ulates e i ase unils to opti izc 
i cy ofrc i ll n  poxic ecJIs. ell, 07, 9, 1- 2. 
( 2] a pian, ; ao, .; ian, . ; aIOIl, . yt chr c e i ase 
e li iry d gen n e. . iol, hem" 07, 2, 30-8. 
( 3] oreno-Sánchez , .; aavedrn, E.; odrigucz-Enriqucz, .; lín-
ndoval, . Clabolic nll'ol a lysis:  l r si ning 
ll' t gies 10 anipulale etabolic t ays. i med. iolech-
o l., 08, 08, rticJe  79 13. 
4J oche, .E.; ir asa, . yruvate drogenase i ase ula -
ry cchani s s d i iti n  c t g i betes, eart ia, 
d nccr. e" ol. if  c;., 07, 4, 30-49. 
( 51 Oudi, .; owkcr-Kinlcy, .M.; edishv li, . .; hao, .; 
pov, . iversil)' f e ruvalc c ydrogenase i ase cne 
il y  ans. iol. hem., 95, 0, 989-94 . 
[146] owker-Kinlcy, .M.; avis, . J. ; u, .; arris, . .; pov, 
. . vi cncc r iS{encc of ti suc-sp ciflC c ulati n f c am-
al ian ruva te drogenase p lex . i chem, ., 98, 9, 
1-6. 
( 7J orotchkina, . ; alcl, .S. ilc ccificil)' f r ruvatc c y-
cnasc i ase soc cs ard I e I rcc osphorylatkln itcs f 
hunEn r va te ydrogcnase. ioJ. hem., 01, 6, 23-9. 
( 8] olobova, .; uganova, .; ou latnikov, l.; pov, . . egu la-
 f r vate c drogenase tivity l gh osphorylal ion t 
ultip lc ilcs. i hem. . , 001, 58, - 7. 
( 9) andreou, !.; ai s, . .; ntana, .; i , .L.; c nko, .C. 
IF-I ediates aptat ion  pox ia y ti e ly nregu lating 
it chondrial xy cn sumption. ell e tab., 06 , , 7-97. 
n 'M d: 1 "priL 09 rup lfd : J " rU. ro09 
illi-R('l, it!II'S ll ¡\ /(!(lici/UIi ht!mislr)', 09. ol. , o.  I ItI I 
0J i , . .; ehernyshyov, l ; $ enza, O.L; ang, . . IF-I-
ediatcd c pre s iol1 f ruvate c drogcnase i ase:  ela-
olic ilch c ired r cll lar aplati n  poxia. ell e-
l b. , U 06, , -85. 
I 1J c Fate, .; ohycldin, .; u, .; hakar, .; cnrlques, .; 
a li , .D.; u , .; ehell , J .; sang, . . ; eahan, .; hou, 
.; alifano, . .; wlg, .H.; a ri s. . .; enna, . yruvatc 
c ydrogcnasc plcx Cl ivil)' nll'ols etabolic d rro li nan t 
cnol)'pe  cer e lls. iol. hem., 0H, 3, 7 0-8. 
1 2} o ncl, .; rcher, . L.; l l n is- umcr, .; aromy, .; eau-
, .; ompson, .; c , e. .; opaschuk, 0.0 .; n unta , 
.; o net, .; a ry, O. ; ashimolo, .; orter, l; ndrade , 
.A.; hebaud, .; ichelaki s . .  it ndri - + role! 
is  pre scd  ccr d  rnla lization otes opto-
is d i nhib i l~ eer r th. allcel" ell, 07, 1, -51. 
( 3) ani ez-Sánchcz, O.; Oiuliani , . e lu lar x S{atus ulales 
pox ia uc ible tor-I ti ity. olc  our op t. J 
ep. /in . ancel' R e~., 07, 6, -50. 
4] o sain, .F.; i , .P.; acrson, .R.; hang, ; ruiek , .K.; 
oha med , .A.; garwal, .K.; aglc, 0 .0 .; l u, . . Salfnll7l ~ 
ml/llS ig -potent al l ol cule i it rs f poxia-indueible 
t r-!' i chem. iophys. & .~. OIIIIIIlIII., 05, 3, 26- 3. 
51 hang, H.  ian, O.Z.; an, .S.; c, .; Oao, .; en, . R.; ey, 
.; a mers, .; hang, .; il i, .; Oang, . V. ; iu, . .; $e-
enza, O. L. igoxin d t cr r iac l osklcs hibit  IF-
ha thesis d l ck or r lh . roc. at!. cad. ei. 
SA, 008, 05, 579-86. 
61 ong, .; ark, J .; tc hen , .O.; alvani, .; arde llina, . .; 
onks, . ; isher, .l ; ocmaker, . .; el il lo, O . chi y-
ein,  a l - o lecule ibilor f pox ia-induciblc t r- lO A-
i ing li iry. allcer es., 05, 5, 47- 5. 
7} akahashi , .; ishikawa, .; akakura, . i i li n f l or 
c U r lh  e i r y A erfcrel1 c edia ted prcssion 
f IF- I alpha prc s ion  l or c ls d epatoeytcs. elle 
nler. , 0H, , 72 -82. 
8] hoi, J .; Wl, . .; i , 8.0. ; i , . .; con, .; oh, . 
ilcx in,  I -Ialpha hibitor, as t i etastatie otcntial  
 12 e ls. ol. ells, 06, , 91-9. 
9J u po, . ; anaglia, ., n viano, C.  c lfi no, .; ibaui, .; 
arcsio, .; osco, .e. opot can ibits scular c dolhe lial 
th Clor r duction d i genic ti ilY ced y y-
xia  an r l l a y eling ypoxia-inducible -
I l ha d alpha. ol. tlllcer her. , 0H, , 4-84. 
[160] armcliel, .; or, .; erbcrt, . .; lrumu ra, .; r ssc l an s, .; 
cwcrehin , .; ce an, .; ono, .; bramoviteh , .; axwcll, .; 
och, J .; alclilTe , .; ons , .; in, .K.; oUen, .; cshcrt, . 
ole f IF- Ialpha  xia - edialed plOSis, c ll r li cr l i n 
d r our ang iogene is. ahlre, 98, 4, 5-90. 
1] l uw, 8,; ng, .; ihan, .; os7.c, 1 ; rara, .; Oerbcr, .P.; 
nson, .S. ; ergers . he poxic nse of ors  end-
t  l cir icrocnvironment. ancel' ell, 03, , -46. 
2J mclu, .; ustaf son , .; ndberg, J .; kamolo, .; sson, ,; 
oc li ger, . ; akino, . hys iological Cli alKm f poxia c-
le t r- I  an cJetal usc le. SEBJ. , 05, , 09- 1 . 
1I 3] ril hley, .O.; sei, 1; enderson,  ; Bos\\o~ll , .; ales, .l ; 
bour, .N.; iranio, . ypoxia-induciblc t r-l a pres-
cn  an ctri  d lS lat i n  r st landin -
rics r sl noid  2 P2). ndocrinology, t1t16, 7, 4-53 . 
I 4J euzil, 1 , hao, ., Slcnna n, O., li ha, ., Oe llcrl, . , eber, 
., aton, . , runk, .T. l ha-tocophcryl ccinate,  en! 
ilh  ""'  i -tmnour etivity, uces oplosis y sing 
I al S{ability. i chem. ., 02, 62, 9 - 15. 
5] ali sky, .; lalZ, . f. ; ewis, . . l c n s f onnal , 
nign, d ali nant ntan reast cs. ancer es" 3, , 
95-90 1. 
( 6) tsumi, .; hesncy, .; elz, .; eng, .; o nclly, .; akita, 
.; itchell, .; ocala, . igh pre sion f c ible - hos-
ofr ct -2-kinase/ fr ct sc- , - i osphalase K-2; FB3) 
 an nccrs. allcer es., 02, , 8 1-7. 
 
 
 
27 
 
From: aneelafarooqi@benthamscience.org 
To: morenosanchez@hotmail.com; rafael.moreno@cardiologia.org.mx 
Subject: MRMC-192807: Dr. Gennari: Acceptance of Manuscript # 840 after Minor 
Changes 
Date: Thu, 26 Mar 2009 09:21:40 +0500 
March 25, 2009 
Correspondence reference no. MRMC-192807 
 
Dear Dr. Sanchez, 
You will be pleased to know that your article entitled "HIF-1α modulates energy metabolism in cancer cells by modifying the 
status of glycolytic enzymes " will be accepted for publication in MRMC only after incorporation of the following minor changes 
suggested by the referees along with the confirmation letter for color figures, if any. 
Best Regards 
Ms. Aneela Farooqi 
Manager Publications 
Bentham Science Publications 
E-mail: aneelafarooqi@benthamscience.org 
 
Referees' comments: (1) 
The review article entitled "HIF-1α modulates energy metabolism in cancer cells by modifying the status of glycolytic 
enzymes" is aimed at discussing the role of HIF-1a in the regulation of tumor glycolysis. 
Growing body of evidences substantiate the metabolic switch of cancer cells to aerobic glycolysis. This metabolic 
modulation involves generation and utilization of cellular ATP under adverse conditions like diminished oxygen 
supply (hypoxia), cellular stress etc. Hypoxia inducible factor -1α (HIF-1α) has been known to be involved in the up 
regulation of several genes that favor tumor growth. This review article discusses about the role of HIF-1α in the 
up regulation of glycolytic enzymes. 
However, the article does not enlighten how HIF-1α modulates the glycolytic up-regulation, except stating that HIF-1a 
induces the over-expression of key glycolytic enzymes. 
Significantly, key references, which are relevant to the topic of this review, are not discussed. For example, the recent 
review by Denko (2008) in Nat Rev Cancer, and the report by Fulda and Debatin (2007) in Cell Cycle all discuss about 
Hypoxia, HIF-1 and glucose metabolism. Similarly, De Palma et al., (2007) have recently reported the metabolic 
modulation by hypoxia that would have been ideal for discussing the metabolic switch in hypoxia in this review. The 
pioneering work published by Semenza et al., (1994) which is entitled "Transcriptional regulation of genes encoding 
glycolytic enzymes by hypoxia-inducible factor 1" has not been referred. This reference is more appropriate and 
relevant for this review article, for a thorough discussion. 
Table-1 does not cite any references from where those data were taken from. 
It is unclear why the role of HIF-1 α has not been discussed for the glycolytic enzymes such as Hexosephosphate 
Isomerase (HPI), Phosphofructokinase type 1 (PFK-1), Triosephosphate Isomerase (TPI), Phosphofructokinase type 2 
(PFK-2). 
 
Regarding 3-bromopyruvate, the sentence on page-14, line-15 reads as follows citing a reference; " ..... the cytotoxic 
activity against cancer cells was of low potency (IC50~50uM) [76]". This is incorrect, as that reference does not relate to 
cancer cells, in fact that report is related to the activity of "boar spermatozoa". 
 
Page 19, line 15 states " GAPDH is a homo-tetramer of 37 kDa subunits [102] with no isoforms". GAPDH has been 
reported to have different isoforms either at mRNA level or at post-translational levels, (e.g, Choudhary et al., (2000), 
Saunders et al., (1999), Glaser and Gross (1995)). It is erroneous to define that GAPDH has no isoforms. 
 
 
 
28 
 
 
Unfortunately, the review article does not mention about the inhibitors of GAPDH in detail. For example, Koningic 
acid has been known as an inhibitor of GAPDH and has been under in-depth investigation for its potential application as 
an anti-cancer agent. 
Finally, on page-28, the first paragraph emphasizes that targeting cytosolic pH regulators, and v-type ATPase as very " 
……. promising and novel therapeutic approach". On the contrary, in the abstract, the concluding sentence states as 
"……it is concluded that targeting the HIF1α-induced glucose transporter and hexokinase important to glycolytic flux 
control might provide better therapeutic targets for inhibiting tumor growth and progression than targeting HIF-1α itself ". 
There is no explanation or convincing justification why targeting hexokinase or glucose transporters is better or 
advantageous than other targets. 
Referees' comments: (2) 
I would suggest a few minor changes that might further clarify the work: 
In the title, the phrase "by modifying the status of glycolytic enzymes" is confusing. This could be changed to "through 
precise effects on glycolysis" or "by facilitating a glycolytic phenotype" or something similar. 
There are a few typographic errors: On page 6, the authors use PDH or PDH2 when they 
should use PHD to abbreviate prolyl hydroxylase. Also, the end of the sentence that starts page 30 is scrambled. 
The authors point out that the drug 2-deoxyglucose (2-DOG) inhibits HPI, although it is 
usually referred to as a competitive inhibitor of hexokinase. The authors point out that in the 
presence of glucose, 2-DOG does compete with glucose at HK and elsewhere in the glycolytic 
pathway. Under standard conditions of culture or tumor growth in vivo (when glucose is present), is 2-DOG's primary effect 
on cell growth/survival exerted at the level of HPI or at these other sites? 
PYK-M2's involvement in tumor cell metabolism is well-publicized but still difficult to explain and somewhat confusing as 
presented here. Do the oncoproteins pp60-v-src and HPV-16 E7 promote formation of the inactive dimer or the active 
tetramer? And can the authors resolve the issue of why the less active dimer seems to predominate in tumor cells, even when 
the Warburg effect (i.e. lactate production) is so prevalent in these same cells? Finally, many discussions of this topic 
hypothesize that PYK-M2 dimerization facilitates tumor growth by allowing upstream glycolytic intermediates to 
accumulate and be used for biosynthesis of nucleic acid and other macromolecules. But is there any good evidence that this 
occurs? The two papers cited do not provide any compelling evidence to this point. Reference 117, for example, shows an 
enhancement of conversion of glucose carbon into lipids in the presence of PYK-M2, but this activity presumably requires 
the forward activity of that enzyme, yielding pyruvate whose carbon can ultimately enter fatty acid synthesis. Can the 
authors resolve this question? 
Although it is not known to be a HIF-1α  target, MCT1 has been shown to support tumor 
growth in animal models (Sonveaux et al, J Clin Invest 118:3930-3942). This reference should be included. 
In Figure 2, why is MCT4 shown as a filled symbol when it is the only MCT isoform shown? 
Referees' comments: (3) 
The first part describing glycolytic enzymes is very comprehensive and detailed. The second part, when the authors attempt 
to discuss HIF-1 is a target is poorly elaborated (pg 30). Either they remove this paragraph, which in n ay case seems to be 
at odd with the rest of the manuscript, or they expand it which might make the manuscript too long. 
1) Pg 4: please check half-life of HIF under hypoxia. 
2) Typos: pf 9 In 13: liomyomas; pg 10 In 13, sentence is unclear (down-regulate?); pg 30 In 1 
revise sentence. 
3) Reference for Topotecan as HIF inhibitor is missing.  
 
Typographical errors 
References incorrectly presented
 
 
 
29 
 
21 April 2009 
 
Ms. Aneela Farooqi 
Manager Publications 
Bentham Science Publications 
 
Thank you for your kind letter of 25 March 2009, in which we were informed that our manuscript will be 
accepted for publication in Mini-Reviews in Medicinal Chemistry after incorporation of the changes 
suggested by the three reviewers.  We would like to thank the reviewers for their favorable comments and 
minor criticisms.  We have modified the text taking into account their suggestions.  We have outlined our 
responses to each of the points they have made, in turn.   
Due to the changes made and the addition (and deletion) of references, the reference numbering was 
modified.  The reference format was also changed.  A running header was included on the Title page.  As 
required, we are re-submitting the revised manuscript in both a Word file, with all the changes made in 
bold letter, and a pdf file, with all the contents included and inserted as they should appear in the printed 
version.   
 
Sincerely yours, 
 
Rafael Moreno-Sanchez, Ph. D. 
 
 
Response to reviewer 1 
We agree with the reviewer comment that the paper does not enlighten on how HIF-1 modulates the 
glycolytic up-regulation.  However, the genetic (molecular biological) mechanism by which HIF-1 
modulates over-expression of specific glycolytic genes was outside the scope of our mini-review, and of 
the journal (focusing on Medicinal Chemistry).  The central point of our review was to address the up-
regulation of glycolytic gene isoforms and the possible biochemical (kinetic) reasons for this selection in 
tumor cells.  Despite almost all the glycolytic pathway are up-regulated by HIF-1, the important point we 
make is in regard to those steps that play the most critical roles in controlling the glycolytic flux in cancer 
cells (a paragraph clearly stating the main focus of the paper was included on p. 4).  It is these proteins 
that should therefore be considered the best targets for inhibitors as anti-cancer agents because blocking 
at such points would be predicted to show a stronger effect at inhibiting glycolysis (this was originally 
stated on the Abstract and p. 32). 
Some of the references suggested by the reviewer [4,5] were included in the text.  On the other hand, the 
De Palma paper is on skeletal muscle subjected to hypoxia, which is not a highly proliferative tissue like 
tumors; the Cell Cycle paper is a very short (two and a half pages) review on the subject.  References 
were added to Table 1. 
We have now modified the manuscript to mention that TPI gene is also up-regulated by HIF-1α (p. 19), 
and that HPI is over-expressed under hypoxia through a HIF-1α independent mechanism (p. 15).  
Discussion on HIF-1α-mediated PFK-1 and PFK-2 over-expression was included in the original 
manuscript on p. 17 and 26, respectively. 
The incorrectly cited reference on 3BrPyr was changed (previous [76], now [74, 75 ]). 
Isoform is defined as a protein encoded by a specific gene and hence, with amino acid sequence and 
kinetic and structural properties different to other proteins catalyzing the same physical (transporters) or 
chemical (enzymes) reaction.  Therefore, the original statement, “GAPDH has no isoforms” was correct.  
The publications mentioned by the reviewer which use the word isoform are not isoforms, but concern 
post-translational modifications of GAPDH.  However, another GAPDH isoform has been certainly found 
in spermatozoa.  Then, the text was slightly modified (p. 20).  
A detailed description of the properties of Koningic acid as a GAPDH inhibitor was incorporated on p. 21 
and 22, as suggested. 
The reviewer has taken out of context the conclusion that targeting HIF-1 versus GLUT or HK as a better 
option.  This was one of the key proposals developed throughout the review and is quite distinct from the 
possibility of targeting cytosolic pH regulators and ATPases (p. 29 and 30) as the reviewer pointed out.  
Drugs targeting HIF have high toxicity in general to many cell types and this was outlined on p. 32, line 3 
and again in the conclusion on p. 33, line 14.  Consequently, targeting of the downstream targets of HIF, 
 
 
 
30 
 
particularly GLUT and HK could provide more specific targets with less side effects, as outlined and 
described on p. 12 for GLUT and p. 14 and 15 for HK, respectively.  This proposal arises from the 
observations of the authors that GLUT and HK constitute potent regulatory points in the control of 
glycolytic flux in cancer cells.  Hence, targeting these two sites would be predicted to have strong effects 
inhibiting glycolysis in cancer cells.  This was discussed on p. 14 and 16? with two examples of drugs, 3-
BrPyr  and 2DOG that have been shown as potent glycolytic inhibitors. 
 
 
Response to reviewer 2 
1, 2, 6.  These observations were corrected.  
3.   2-DOG.  The question on 2-DOG’s primary effect was further elaborated on p. 16. 
4.   PYK-M.  Interaction of oncogenes with this isoform induces the transition of tetramers to dimers.  This 
was stated in the original version on p. 24, line 10.  An explanation on why the dimer prevalence in cancer 
cells does not affect glycolysis was also added (p. 25, 3
rd
 paragraph).  Hard data on intermediary 
metabolite accumulation is still missing. 
5.   The Sonveaux et al reference was added (p. 29, 2
nd
 paragraph) as suggested, and the associated text 
was slightly modified. 
 
 
Response to reviewer 3 
We consider to be inappropriate the point made by this reviewer about taking exception to the section on 
HIF-1 as a target being poorly elaborated.  This section follows as a logical conclusion to the review and 
possible lead candidate for novel anti-cancer therapies aimed at targeting the subject of the review.  Thus, 
we emphasize that it is NOT “at odds with the rest of the manuscript” but provides a purposeful flow to the 
logic addressed here.  Unfortunately, there has not been significant development in the subject of 
targeting HIF yet, although some interesting initial results have been obtained with the manassantins as 
addressed on p. 32.  Hence, this section is relatively short, but still remains highly pertinent to the themes 
developed in the review.  
Again minor comments and typographical errors corrected. 
Half life of HIF-1α under hypoxia.  Values were checked and confirmed (p. 5, line 4).  
Topotecan reference was not missing – Ref. 157. 
 
 
 
31 
 
2.1.1 Comentarios sobre el manuscrito No. 1 
Este primer manuscrito nació de una idea que surgió al leer sobre los efectos del 
HIF-1α en el metabolismo de la glucosa en células tumorales. Una primera versión de 
este trabajo la publiqué como autor único en la Revista de Educación Bioquímica de la 
Facultad de Medicina de la UNAM (REB 2009, 28: 41-51) algunos meses antes.  
En este  manuscrito se analiza el papel que juega HIF-1α en la regulación de la 
glucólisis en las células tumorales, así como las ventajas que ofrece en estas células la 
expresión de isoformas diferentes de las enzimas (HKI, HKII, PFK-L, ALD-A, ALD-C, 
PGK1, ENO-α, PYK-M2, LDH-A y PFKFB-3) y transportadores (GLUT1 y GLUT3) 
glucolíticos inducida por este factor de transcripción. De este análisis concluimos que la 
expresión de HIF-1α facilita el desarrollo tumoral porque incrementa la glucólisis tumoral 
al promover la expresión de isoformas de las enzimas  que tienen una menor afinidad 
por sus productos (ALD-A, ALD-C, MTC4) y una menor sensibilidad a sus inhibidores 
fisiológicos (PFK-L y PYK-M2), lo cual promueve el flujo hacia lactato y ATP. Además, 
algunas de estas enzimas al mantener funciones distintas contribuyen en la regulación 
de la transcripción (LDH-A), en la inhibición de la apoptosis (HKII) y en mediar la 
migración celular favoreciendo la metástasis (ENO-α).  
Posteriormente, para publicar este manuscrito en la revista Mini Reviews in 
Medicinal Chemistry se incorporó información sobre inhibidores de HIF-1α y de las 
enzimas de la glucólisis, así como información sobre la regulación que ejerce HIF-1α 
sobre enzimas mitocondriales (piruvato deshidrogenasa cinasa y citocromo c oxidasa) 
que pueden modificar el funcionamiento de la mitocondria. Con base en la discusión de 
esta información concluimos en este trabajo que el HIF-1α, al tener la capacidad de 
 
 
 
32 
 
regular el metabolismo energético en las células tumorales, podría ser un blanco idóneo 
para el tratamiento del cáncer, aunque en paralelo sería necesario reducir 
específicamente al GLUT y a la HK, proteínas que controlan la glucólisis en células 
tumorales.  
2.2 Genes supresores de tumores y oncogenes 
Es interesante remarcar que aquellos genes (genes supresores de tumores y 
oncogenes) que conducen al desarrollo tumoral también están involucrados en el 
incremento de la glucólisis. 
2.2.1 PTEN (Phosphatase and tensin homolog deleted on chromosome ten)   
PTEN es una fosfatasa que específicamente remueve el fosfato en posición 3’ 
del anillo de inositol de los fosfoinositidos. Como consecuencia reduce su actividad 
Akt/PKB y las PDK´s (1 y 2) disminuyendo la fosforilación de sustratos involucrados en 
el crecimiento y la supervivencia celular. En las células tumorales, la activación de esta 
vía se produce por mutaciones que inhiben la actividad de PTEN, las cuales son muy 
comunes en algunos tumores (Jones y Thompson, 2009; Robey y Hay, 2009).  
El incremento en la glucólisis inducido por esta cascada de señalización se debe 
a que la activación de Akt promueve la expresión de varios transportadores de glucosa 
(GLUT1, GLUT2 y GLUT4) así como su translocación a la membrana plasmática e 
induce la asociación de la HKII a la membrana externa mitocondrial. Además, 
indirectamente activa la fosfofructocinasa tipo 1 (PFK-1) al activar por fosforilación 
directa a la fosfofrutocinasa tipo 2 (PFK2), cuyo producto es la fructosa -2,6- bisfosfato 
(F26BP), potente activador de la PFK-1.  Al mismo tiempo, Akt evita la supresión de la 
transcripción de los genes glucolíticos al inhibir al factor de transcripción FoxO. 
 
 
 
33 
 
Finalmente con la activación de mTORC1, Akt incrementa la presencia de HIF-1α el 
cual, como ya se documentó en la sección anterior (y en el primer manuscrito 
publicado), es el principal factor de transcripción que regula la expresión de la mayoría 
de las enzimas y transportadores de la glucólisis (Yeung  et al., 2008; Robey y Hay, 
2009; Marin 2009). 
2.2.2 p53 
 La proteína p53 es un factor transcripcional inducido por estrés y encargado de 
mantener la estabilidad genética en la célula al modular la proliferación celular.  Es 
importante señalar que en un gran número de tumores humanos esta proteína se 
encuentra inactiva por presentar mutaciones (Jones y Thompson, 2009; Olovnikov et 
al., 2009).   
p53 inhibe la glucólisis a diversos niveles, reprime la transcripción de los genes 
de algunos transportadores de glucosa (GLUT1 y GLUT4) al unirse directamente al 
DNA inhibiendo a sus promotores; induce la ubiquitinación y la degradación de la 
fosfoglicerato mutasa (PGAM); y reduce la concentración de F2,6BP al inducir a la 
enzima TIGAR (TP53-induced glycolysis and apoptosis regulator) que es una isoforma 
especial de PFK-2 con actividad predominante de bisfosfatasa que convierte a la 
F2,6BP (potente activador de la PFK-1) en F6P, con lo cual se reduce la actividad de la 
PFK-1. Por lo tanto, la perdida de p53 puede inducir un incremento en la glucólisis 
(Yeung et al., 2008; Jones y Thompson, 2009; Olovnikov et al., 2009). 
2.2.3 c-Myc 
c-Myc se transcribe activamente en diversos tipos de cáncer (linfomas de 
Burkitt), su producto, la proteína c-Myc, participa en la regulación de la apoptosis y de la 
 
 
 
34 
 
proliferación y diferenciación celular. c-Myc es un factor de transcripción que forma 
heterodímeros con la proteína MAX; estos heterodímeros se unen a sus genes blanco 
en la secuencia consenso 5´-CACGTG-3´ (Dang y Semenza, 1999). Diversos reportes 
indican que la activación de c-Myc altera el metabolismo de la glucosa, promoviendo el 
incremento de la glucólisis al aumentar la transcripción de varias enzimas (GLUT1, HPI, 
PFK, GAPDH, PGAM, ENO y LDH-A) (Shim et al. 1997; Osthus et al. 2000). 
Otros oncogenes como H-ras y src pueden estimular la glucólisis indirectamente 
al estabilizar al HIF-1α (Kim y Dang, 2006). 
 
2.3 Cambios en los mecanismos de regulación de la hexocinasa y de la 
fosfofructocinasa tipo 1. 
Diversos grupos han descrito que en células no tumorales de mamífero (eritrocito 
humano, músculo de ratón y corazón de rata) el control del flujo de la glucólisis lo 
mantienen principalmente la hexocinasa (HK) y la fosfofructocinasa tipo 1 (PFK-1) 
(Rapoport et al., 1974; Kashiwaya et al., 1994; Jannaschk et al., 1999). El control que 
ejercen estas enzimas oscila alrededor del 70 y 30%, respectivamente (Rapoport et al., 
1974; Kashiwaya et al., 1994; Jannaschk et al., 1999).  En un estudio en corazón se 
señala que también el transportador de glucosa puede contribuir significativamente al 
control de la glucólisis (Kashiwaya et al., 1994). 
 
2.3.1 Hexocinasa (HK) 
La HK cataliza la reacción de fosforilación de la glucosa (Figura 1);  existen 
cuatro isoenzimas (HKI, HKII, HKIII y HKIV), siendo la glucocinasa (HKIV) la única que 
 
 
 
35 
 
no es inhibida por su producto la glucosa 6-fosfato (G6P), mientras que las restantes 
mantienen constantes de inhibición (Ki) que oscilan entre 20 y 100 μM (Wilson, 2003).  
En las células tumorales, la actividad de esta enzima, y predominantemente de la 
isoforma HKII, se llega a incrementar entre 5 y 500 veces en diversas líneas tumorales 
de rata (hepatomas de Morris, AS-30D, Novikoff, Erhlich-Lettre,  glioma C6, células 
ovales de hígado, carcinoma de tiroides), de humano (MIA, adenocarcinoma de 
páncreas; K562, leucemia; LoVoD, carcinoma de colon), y en diversos tipos de tumores 
humanos (pulmón, mama, colon, recto) (Shonk et al., 1964; Kosow y Rose, 1968; Parry 
y Pedersen, 1983; Nakashima et al., 1988; Smith, 2000; Pedersen et al., 2002; Stubbs 
et al., 2003). La HKII se llega a encontrar unida a la membrana externa mitocondrial, lo 
que de acuerdo con varios autores permite que sea menos sensible a la inhibición por 
su producto la G6P (Bustamante et al., 1977; Nakashima et al., 1986; Widjojoatmadjo et 
al., 1990; Marín-Hernández et al., 2006). 
 
2.3.2 Fosfofructocinasa tipo 1 (PFK-1) 
La reacción que cataliza la PFK-1 produce fructosa 1, 6 bifosfato (F1,6BP) por 
medio de la fosforilación de la fructosa 6-fosfato (F6P) (Figura 1). Esta enzima se inhibe 
por el ATP, citrato y H+, y se activa fuertemente por la fructosa 2,6 bisfosfato (F2,6BP), 
el AMP y Pi. De esta enzima se conocen tres isoformas (músculo M, plaqueta C o P e 
hígado L), las cuales difieren en su sensibilidad a los activadores e inhibidores y tienen 
pesos moleculares ligeramente diferentes.   
En las células tumorales la PFK-1 incrementa su actividad hasta en 5 veces en 
algunos tumores y leucemias humanas (Vora et al., 1985; El-Bacha et al., 2003). En 
diversos modelos tumorales (carcinoma de tiroides, leucemia y gliomas) se ha 
 
 
 
36 
 
observado que la PFK-1 muestra entre 5 y 300 veces mayor sensibilidad por su 
activador la F2,6BP (Oskam et al., 1985; Colomer  et al., 1987; Staal et al., 1987) y una 
menor sensibilidad al efecto inhibitorio inducido por el ATP y el citrato entre 1 y7 veces, 
respectivamente (Meldolesi et al., 1976; Oskam et al., 1985; Staal et al., 1987). Las 
diferencias observadas se han atribuido a la expresión preferencial de las isoformas C y 
L sobre la M (Oskam et al., 1985; Vora et al., 1985; Staal et al., 1987). Sin embargo, es 
sorprendente descubrir que no hay datos que apoyen esta última explicación, en 
particular porque la caracterización cinética de las isoenzimas de la PFK-1 en tejidos 
sanos indican que la isoforma M es la que  tiene una mayor sensibilidad por la F2,6BP y 
se inhibe en menor proporción por el ATP y el citrato  (Vora et al., 1985b; Dunaway et 
al., 1988).  
Además, en células cancerosas existe un aumento significativo en la 
concentración de la F2,6BP debido a la expresión de una isoforma de la enzima 
bifuncional fosfofructocinasa tipo 2 (PFKB-3), que tiene la característica de mantener 
una elevada  actividad de cinasa sobre la actividad de fosfatasa (es decir predomina la 
síntesis de F2,6BP sobre su degradación) (Yalsin et al., 2009). El que esta enzima 
mantenga esta elevada actividad de cinasa, de acuerdo a datos cristalográficos, se 
debe a que algunos residuos de su extremo N-terminal bloquean el sitio catalítico de 
fosfatasa. Además con respecto al resto de las isoformas, ésta presenta diferencias en 
las secuencias de unión del ATP y de la F6P que al parecer facilitan su unión. 
Asimismo, el dominio C-terminal que en el resto de las isoformas bloquea el sitio de 
cinasa, en esta enzima es flexible y no lo bloquea (Yalsin et al., 2009).  
 
 
 
37 
 
Como se sabe, la PFK-1 participa en el conocido efecto Pasteur, que consiste en 
la inhibición de esta enzima por algunos intermediarios mitocondriales como el citrato y 
el ATP que se producen activamente durante aerobiosis y por lo tanto la PFK-1 se 
inhibe y la glucólisis disminuye. Sin embargo, como en las células tumorales la 
concentración de F2,6BP se mantiene elevada, este activador bloquea el efecto 
inhibitorio del ATP, el citrato y el H+ provocando que disminuya el efecto Pasteur 
(Eigenbrodt et al., 1985). 
 
 
 
 
38 
 
Capítulo 3. La glucólisis como un blanco terapéutico 
 
3.1 La glucólisis como un blanco terapéutico en el tratamiento del cáncer  
Como el incremento en la glucólisis es una de las alteraciones metabólicas que 
se mantiene en la mayoría de las células tumorales y está relacionada con resistencia a 
la quimioterapia y a la radioterapia (Fanciulli et al., 2000; Maschek et al., 2004; Xu et al., 
2005; Lee et al., 2007), diversos grupos de investigación consideran  que la inhibición 
de la glucólisis es una opción terapéutica (Pelicano, 2006), dado que podría disminuir el 
crecimiento de los tumores e incrementar la eficacia de la quimioterapia y la 
radioterapia. Aunque se propone inhibir, individualmente, a diversas enzimas (HKII, 
PFKFB3, GAPDH, LDH-A) o transportadores (GLUT1 and MCT) (Pedersen et al., 2002, 
Pelicano et al.,  2006; Kumagai et al., 2008; Bartrons et al., 2007; Evans et al., 2008; 
Clem et al.,  2008) de esta vía metabólica, no se ha determinado previamente el control 
que ejercen sobre la glucólisis. Para que la inhibición de la glucólisis sea efectiva, se 
debe considerar inhibir a las enzimas que controlan a esta vía ya que sólo reduciendo 
su actividad en una baja proporción, el flujo se reducirá drásticamente y no se 
necesitarán altas concentraciones del inhibidor.  
Por lo tanto en esta tesis se consideró que como la HK  y la PFK-1 presentan 
cambios en sus mecanismos de regulación en las células tumorales, era difícil suponer 
que la distribución del control de la glucólisis en las células tumorales fuera similar a 
aquella encontrada en las células normales. Por lo tanto, en este trabajo se determinó 
la distribución de control en células tumorales aplicando el análisis de control 
metabólico, para poder establecer cuáles serían las enzimas o transportadores que 
podrían ser considerados blancos terapéuticos. 
39 
 
3.2 Análisis de control metabólico 
El análisis de control metabólico permite identificar a las enzimas que controlan a 
una vía metabólica, al determinar cuantitativamente el grado de control que ejercen 
cada una de ellas sobre el flujo de una vía metabólica o sobre la concentración de 
metabolitos. Para ello es necesario determinar la estructura de control, que está 
constituída por el coeficiente de control de flujo, por el coeficiente de control de 
concentración y por los coeficientes de elasticidad. El coeficiente de control de flujo 
indica el grado de control que ejerce una enzima sobre el flujo; en tanto que el 
coeficiente de control de concentración es el grado de control que ejerce una enzima 
sobre la concentración de un metabolito. Ambos coeficientes son propiedades de la vía 
metabólica y están determinados por los coeficientes de elasticidad, que se definen 
como la habilidad que tiene una enzima de variar su actividad al cambio en la 
concentración de alguno de sus ligandos (sustrato o producto).   
El coeficiente de control de flujo se define como: 
                                             
0
0
J
vi
dvi
dJ
C J
vi   
la variación en el flujo (J) cuando se hace un cambio infinitesimal en la concentración o 
en la velocidad de una enzima (vi). Así que cuando un pequeño cambio en la velocidad 
de una enzima o transportador induce una variación proporcional en el flujo, la enzima 
ejerce un control elevado; pero si la variación en la actividad de la enzima no modifica el 
flujo, esta enzima no ejerce un control significativo (Fell, 1997; Moreno-Sánchez et al., 
2008). Para obtener los valores de CJ
vi normalizados y adimencionales es necesario 
multiplicar por el factor escalar (vi0 /J0) que es la relación entre los valores iniciales de 
velocidad y flujo. 
40 
 
Para poder determinar el coeficiente de control de flujo de una enzima son 
necesarias pequeñas variaciones en su actividad, sin alterar el resto de la vía 
metabólica, midiendo los cambios en el flujo. Varias estrategias experimentales se han 
desarrollado para poder determinar el coeficiente de control de una enzima, de las 
cuales en este trabajo se emplearon el análisis de elasticidades y modelado matemático 
(in silico biology).  
 
3.2.1 Análisis de elasticidades 
El coeficiente de elasticidad (ε) se define como el cambio en la velocidad de una 
enzima o transportador cuando la concentración de uno de sus ligándos (sustrato, 
producto o activador) varía en una proporción infinitesimal. Los coeficientes de 
elasticidad son positivos para los metabolitos (sustratos y activadores) que incrementan 
la velocidad de una enzima o transportador y son negativos para los metabolitos que la 
disminuyen (productos e inhibidores) (Fell , 1997; Moreno-Sánchez et al., 2008). 
Una enzima que tiene un coeficiente de elasticidad bajo no modifica su velocidad 
con el cambio en la concentración de sus ligandos y, por tanto, no modifica el flujo 
metabólico, lo que permite que mantenga un control alto sobre el flujo de la vía. En 
cambio una enzima con un coeficiente de elasticidad alto será muy sensible a los 
cambios en la concentración de sus ligandos, su actividad se ajustará a dichos cambios 
y, entonces, ejercerá un control bajo sobre el flujo. Esta relación inversa entre los 
coeficientes de elasticidad y los coeficientes de control se conoce como el teorema de 
conectividad (Fell, 1997; Moreno-Sánchez et al., 2008). Para simplificar el análisis 
teórico y experimental de una vía metabólica ésta se puede dividir en dos bloques, en 
41 
 
un bloque productor (EP) y en un bloque consumidor (EC) de un metabolito (m) (Fell, 
1997; Moreno-Sánchez et al., 2008), quedando el teorema de conectividad como sigue:   
EP  m  EC           0 C
C
P
P
E
m
J
E
E
m
J
E CC   
Para calcular los coeficientes de control es necesario determinar 
experimentalmente los coeficientes de elasticidad. Esto se puede lograr al variar la 
concentración del sustrato inicial (glucosa en el caso de la glucólisis) y de esta manera 
incrementar la concentración del metabolito de interés y el flujo (generación de láctico). 
Las concentraciones y los flujos son normalizados, considerando el punto de interés 
(que es la condición en la cual se quiere conocer la distribución de control) como el 
100%. En nuestro caso el punto de interés fue 5 mM de glucosa (concentración 
fisiológica de glucosa en sangre).  A partir de estos valores se obtiene la gráfica % 
metabolito vs % flujo (Figura 3.2.1) y se calcula la pendiente de la curva al punto de 
interés, que es el coeficiente de elasticidad del bloque consumidor ( CE
m ). Finalmente, 
con el empleo de un inhibidor de alguna de las enzimas del bloque consumidor se 
determinan la concentración del metabolito y el flujo, y del gráfico se obtiene el 
coeficiente de elasticidad del bloque productor ( PE
m ). Cabe señalar que los 
experimentos se tienen que realizar bajo condiciones de estado estacionario. 
Los coeficientes de control son calculados a partir del teorema de conectividad, 
considerando que la suma de los coeficientes de control de una vía son igual a 1 (CP + 
Cc = 1) (Fell, 1997; Moreno-Sánchez et al., 2008). Entonces tenemos que: 
                     P
m
C
m
C
mJ
EP
C



               P
m
C
m
P
mJ
EcC



     
42 
 
Pendiente= Ec
m
Glucosa
Pendiente= - Ep
m
Inhibidor
40 60 80 100 120 140
40
60
80
100
120
%
 F
lu
jo
% Metabolito
m
EP EC
m
EP EC
 
 
Figura 3.2.1 Gráfico para la obtención de los coeficientes de elasticidad. Una vía metabólica se 
puede dividir en dos bloques, en un bloque productor (Ep) y en un bloque consumidor (Ec) de 
un metabolito (m). Para determinar el coeficiente de elasticidad del Ec se tiene que variar la 
concentración del sustrato inicial de la vía (glucosa en el caso de la glucólisis) de esta manera 
se incrementa la concentración del metabolito de interés y el flujo. Mientras que para determinar 
el coeficiente de elasticidad del Ep se emplea un inhibidor de alguna de las enzimas del bloque 
consumidor con lo que se reduce el flujo y se incrementa la concentración del metabolito. Las 
concentraciones y los flujos son normalizados, considerando el punto de interés (que es la 
condición en la cual se quiere conocer la distribución de control) como el 100%. A partir de 
estos valores se obtiene la gráfica % metabolito vs % flujo y se calculan las pendientes de cada 
curva en el punto de interés, que corresponden al coeficiente de elasticidad del bloque 
consumidor (
CE
m ) y al coeficiente de elasticidad del bloque productor (
PE
m ). 
 
A partir de aplicar el análisis a varios metabolitos de la vía es posible determinar 
los sitios que ejercen la mayor parte del control. El método tiene la ventaja de que no se 
necesitan inhibidores específicos de cada una de las enzimas o transportadores de la 
vía, ya que únicamente es necesario producir una variación en la concentración del 
metabolito de interés (Moreno-Sánchez et al., 2008). Sin embargo, para poder aplicar el 
análisis de elasticidades es necesario que los intermediarios de interés se encuentren 
43 
 
en concentraciones en las cuales se puedan detectar por los métodos analíticos 
convencionales.  
El análisis de elasticidades se ha aplicado en diversos estudios entre los que se 
destacan los siguientes: 
a) En hepatocitos se determinó que la piruvato carboxilasa (CJ
Ei = 0.83) controla la 
gluconeogénesis (a partir de lactato) estimulada por glucagon (Groen et al., 
1986). 
b) Posteriormente también en hepatocitos el grupo de Martin Brand estimó que la 
respiración dependiente de la oxidación de glucosa es controlada por el bloque 
productor de NADH entre el 15 y el 30% mientras que por el bloque consumidor 
de NADH entre un 70 y un 85% (Brown et al., 1990). 
c) En el hepatoma AS-30D se determinó la distribución de control de la fosforilación 
oxidativa mostrándose que el complejo I y los sistemas consumidores de ATP 
mantienen un control significativo (CJ
Ei= 0.7) (Rodríguez-Enríquez et al., 2000). 
Además de los estudios anteriores en células  de mamíferos, este análisis se ha 
aplicado en diferentes  vías metabólicas (fotosíntesis, la síntesis de serina y treonina  la 
glucólisis y la síntesis de galactosa) en microorganismos como E. coli y levadura. 
(Moreno-Sánchez et al., 2008).  
 
3.2.2 Modelos matemáticos de vías metabólicas (In silico biology) 
Con la idea de hacer un análisis más profundo de una vía metabólica, se 
consideró el diseño de modelos matemáticos con los cuales se pudiera lograr 
reproducir su comportamiento en una gran variedad de condiciones experimentales 
(Poolman et al., 2004).  
44 
 
Existen dos tipos de modelado: modelado estructural y modelado cinético. El 
modelado estructural considera la estequiometría del conjunto de reacciones químicas 
que se llevan a cabo en la vía metabólica sin incluir datos cinéticos. En cambio, para el 
modelado cinético, además de la estequiometría es necesario conocer:  
-  la ecuación de velocidad con todos sus parámetros cinéticos (KmS, KmP, Vmf, Vmr, Ki, 
Ka), 
-  las constantes de equilibrio (Keq) (o en su caso las velocidades máximas en  
“forward” y “reverse” usando la ecuación de Haldane) de cada reacción, 
-   las concentraciones de intermediarios de la vía bajo una determinada condición de 
estado estacionario, 
-   un apropiado programa de computación en donde es construido el modelo. Entre los 
mas usados se encuentran “Copasi”, “Gepasi”, “Metamodel”, “WinScamp”, “Jarnac” and 
“PySCeS” (Poolman et al., 2004; Moreno-Sánchez et al., 2008). 
El modelo cinético permite examinar y entender los mecanismos de regulación 
de una vía que no se logra con sólo atender a las propiedades cinéticas de cada 
enzima por separado (Moreno- Sánchez  et al., 2008). Como el  modelo se valida con 
datos experimentales, se asegura que las predicciones que arroja sobre la estructura de 
control de una vía metabólica sean confiables.  
 
45 
 
Capítulo 4   
4.1 Planteamiento del problema 
 A nivel mundial y en nuestro país, el cáncer se encuentra entre las 
principales causas de muerte. Por tal motivo, diferentes grupos de trabajo han 
considerado que la glucólisis, por las ventajas que brinda a las células tumorales, puede 
ser un adecuado blanco terapéutico para reducir su crecimiento. Sin embargo, para 
lograr una reducción importante de la glucólisis  es necesario inhibir a las enzimas que 
mantiene un control alto sobre el flujo de esta vía.  El control de la vía en células  
normales recae principalmente en la HK y la PFK-1.  Sin embargo, no se puede 
suponer que estas mismas enzimas sean los principales sitios de control de la glucólisis 
en células cancerosas, pues existe marcada sobre-expresión de ellas y presentan 
cambios en sus mecanismos de regulación. Por lo tanto, en este trabajo se consideró 
determinar, en una primera etapa, qué enzimas mantienen el control de la glucólisis 
tumoral y al mismo tiempo entender los mecanismos bioquímicos involucrados. Con la 
información generada se planteó una segunda atapa en la que se proponen los sitios 
más adecuados para intervención terapéutica y así estar en capacidad de diseñar 
nuevas estrategias antitumorales.    
 
4.2 Hipótesis 
El control de la glucólisis en las células tumorales de rápido crecimiento no recae 
totalmente en la hexocinasa y en la fosfofructocinasa tipo I por los cambios que 
presentan en su expresión y en sus mecanismos de regulación. 
 
 
46 
 
 
4.3 Objetivo general 
Determinar la distribución de control en la glucólisis de células tumorales de rápido 
crecimiento. 
 
4.4 Objetivos particulares 
 Caracterización de la glucólisis en células tumorales (AS-30D y HeLa) y en 
células normales (hepatocitos de hígado de rata) a partir de la determinación de 
las concentraciones de intermediarios y de las actividades de cada una de las 
enzimas que forman parte de esta vía.  
 Determinar la distribución de control en la glucólisis de las células AS-30D, al 
emplear el análisis de elasticidades. 
 Construir modelos cinéticos de la glucólisis de las células AS-30D y HeLa.  
 
 47 
Capítulo 5.  Resultados 
  
5.1 Distribución de control de la glucólisis en las células AS-30D  
Como parte de nuestro primer objetivo, se realizó una caracterización de la 
glucólisis en modelos tumorales (hepatoma AS-30D de roedor y células HeLa de cáncer 
humano cérvico-uterino) y en un modelo de célula normal (hepatocitos de rata). Para 
ello determinamos las actividades de las enzimas, las concentraciones de algunos 
metabolitos y el flujo glucolítico en los tres modelos.  Con base en los resultados de 
esta caracterización se consideró determinar la distribución de control en la glucólisis de 
las células AS-30D mediante el análisis de elasticidades, debido a la ausencia de 
inhibidores específicos para la glucólisis. Los resultados obtenidos fueron publicados en 
la revista FEBS Journal 2006, 273: 1975-1988.  
 48 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Factor de impacto 2009: 3.04 
Citas:30 
Determining and understanding the control of glycolysis 
in fast-growth tumor ce lis 
Flux control by an over-expressed but strongly product-inhibited 
hexokinase 
Alvaro . HE,miánrJez , Sara 
Rodríguez!, Marina M"c,aS·"'ilvi'· 
2 
Keyworm 
.'a,)t,,,,, cc;¡?tt¡d€;:1:; ',ui,<cnt,olcoe:":¡ciétlt; 
hOiiok:mliO tYP€l 2; rnetaboLc 
p:1OSj)hof; uctox;nase ~yp>2 1 
Correspondence 
ft Morerv-S¿rchez !rsTituto 
Caro'OiogLa. Departar:len!C ce B;oquÍlTl;ca, 
Juan Badiano ¡" Col, SeCC'Ófl 
Méxicc '14080, 'vieluco 
FAX 52 55 55730920 
52 55732911 oxe 1422,1298 
&n8E' r;¡f,lol.mo'ono@ca"jiolo;¡:a.o;rg.rnx, 
11 1742·4658,2OCHi05214<x 
, Paola A, ,L Flores, 
M "me 1" Sosa·Gairr<lcn,oL and Rafael Moreno.Sánchez 1 
Secoon XVI, Mé);)(:o, Mex'Co 
Control of 1he flux vvas carrie<i oul in lwo ta;¡HlrOWlll 
tumor c'ell types 01' numan und roden! and AS,30D. respoct, 
Detennintitlon of the maximal tV'BUlA) of the lO gb,e~,ly! 
enzyrnt::;i from h.:xokinasc 10 lacuue reveak:d lhal ht::xokin-
as<: (l53-30¡) aud hao 
ín ral AS,3llD <:eHs thnn in nornnd 
isolated rat ",,¡,atoe:!le,. 'f\.{oreover, the ste'"dv·"',,le concentrations-
01' lhe glycol:llic melab,oliles, nalrlícularl, lhoS{' 01' Ihe of hexo· 
kíua,e ancl PFK-J, were íUéreased comparecl wílh hellal,(lC, 
cel1s. VlllÜX v<:tlues anu metaboHle concentrations for lhe 10 
In HeLa 
cn.zyrnc wcrc aJso i l1t:rc:ascd, hut to a much ksscr extt::nt 
limes ror hoth hexokinase and PFK-l). of lhe 
"""o.IVll;C Ilu, ín AS,3llD cdl, shmveu lhal the block "r enzymc.,; 
transponer. hexokíuase. 
ísomerase, PFK,j. ¡¡na lhe Gle6P hranehesl e,crtea mosl 01' Ihe flux con, 
when:a:s the block aldolase: lo hH:taie 
cxhibiled lhe conlroL Tha 
bloél, and heXükill:ase) alst, sÍlowed flux control 
whien índícated low tlux PFK,L Kíuetíe of 
PFK-l showe:d low towards íts allosteríc lnnlhitors cürate and 
A TP, il:t ool1centralions 01' lhe ae1ívator Ou lhe 
otner ha ud, hCXükíuase was íuhíbíled bul 
eoncentralíons 01' GleóP. Therefore. the cuhanead glyéolylÍé 
flux 111 tumor ceBs was stiH controUed un 
but Glc6P-inhihitcd hcxokinaSt;;, 
It ís wel! document<:d lhal fa'¡l-!!rowrh tumor cells have 
rutes of iactate formati0fi even under uerobic 
condüiol1s than cd}s. For instance. dif· 
fer¡,:Hl lyp.:S uf (Reub:r, Múrri:s, Dunings 
aud Ilhrosarcoma 1929 exhibil "dcs of 
0.1-2.7 ~lmo[ proteinf!, wherc:as normal 
Abbr~\llati(ín$ 
The increase in tumor glyeolys,i, has hc:en fL;;;s.ociated 
with lhe activaliL)n uf se"\ieral onéügenes ras anó 
sre 1 or wÍlh Ihe of the hy:po:,ia·írll:l "(,íble 
faclor IHIF,J:z) in !mnsformed human lyrnphohla:s«)íd 
dihydro:xyacBtone p:lospha:e; G6PDH, gJucose-6-prcsoLsle dehydroger¡ase, GI'i.PDH, g:vcerald€':lvoo-3-ohosohatB cej"rclroger¡ase, 
del'ycl:cegB"dSe: PF<-l. ohosoh(,lmolck 
f-Ei3S 1915 
 49 
Contro l of g lyco lys is in tumo r cells 
and hu man U87 glioma (3,4]. As a result of oncogene 
activa tion and expression, the over-expression of severa l 
genes encoding eighl glycolytic proleins, including lhe 
glucose l ransporter (G luT), lakes place [5J. The over-
expression of a plasma membrane H + -ATPase in rat 
fi broblas ts, to alter lhe cytosolic pH regula tion, and pre-
sumably enhance A TP eonsum ption, a lso promotes a 
sevenfold slimulalion of glycolys is, in addilion to indu-
cing malignanl tra nsformation (6]. 
In comparison with hepatocytes, in several fast-
growth tumor cells (AS-30D, NovikofT) lhere is over-
expression of hexokinase-lI [7,SJ, due to the activation 
of its own promoler, through a demethylation process 
[9] or Lhrough prot ein p53 activation (an abundant 
protein in fast-growlh lumor cells) [IOJ. 
Binding of tumoral hexokinase-II lo lhe mitochond-
rial outer membrane apparently changes its kinetic 
properLies, compared with the cytosolic isoenzyme, i.e. 
mitochond rial hexokinase-I1 shows lower sensitivily 
('" 30%) lo inhibilion by ilS prod ucl Glc6P [7J. The 
close vici ni ly of hexokinase-I1 to lhe adenine nucleo-
tide translocase in tumor milochond ria ensures lhat 
mitochondrial A TP is preferentiaUy used for hexose 
phosphoryla lion [S]. Jt has also been reported lhal 
hexokinase-I I plays an important role in preventing 
apoptolic evenls, such as cytochrome e release in HeLa 
cells, by interfering wi lh lhe bind ing of lhe pro-apop-
totie prolein Bax to lhe outer mitochondrial membrane 
[1 1 J. 
In normal ti ssues, citrate and ATP a re potenl allo-
sLeric inhibiLors of phosphofrucLokinase Lype 1 (PFK-
1) [12J, where iL is mainly consliluled by M subunils, 
but this does not oceur when the predominant subunit 
is L or e (13]. The tumoral isoenzyme is less sensitive 
to the inhibilo ry efTect of Lhese two a ll oslerie efTectors 
[13, 14]. In lhis regard, il has been observed lhal the 
subunit L or e eontenl of tumor PFK- I inereases, 
whereas lhat of subunit M decreases, which explai ns 
the sma ller effeeL of ilS negali ve modulaLors [1 5-17J. Jt 
has also been reporled lhal lhe conlenl of Fru(2,6) P) 
(a poLenl PFK- I activalor [12]) in HeLa cells, Ehrlich 
ascites cells and HT29 human colon adenocarcinoma 
is much higher than in norma l hepatocytes [20-80 ver-
sus 6 pmol·(mg prolei n¡- ' , respeclivelyJ [1 8-2 1]. These 
observations suggesl lhal PKF-I is highly aCLive 111 
tumors cells [13,20]. 
In mammalian nontumorigenic syslems, such as 
human eryth rocytes and rat perfused hea rt, glycolytic 
flux is mainly conlrolled by hexokinase (60-80% ) and 
PFK-I (20-30% ) [22,23J. In lumor cells, an expecled 
consequence of lhe over-expression of several glycolyt-
ie enzymes and glucose transpo rters, and kinelie ehan-
ges in hexokinase and PFK- I, is a large modifiealion 
A. Marin-Hernández el al. 
of lhe regulatory mechanisms and fuctioning of the 
palhway. Henee, Lhe ass umption that conlrol of lhe 
glyeolytie flux in tumor ce lls is simi lar Lo lhat of nor-
mal cells is apparenl ly not well supporled. Therefore, 
lo identify the fl ux-cont rolling siles of tumoral glycol-
ysis, we firslly delermined the Vmax of each glycolytic 
enzyme from hexokinase lo lacta le dehydrogenase 
(LDH) in AS- 30D and HeLa cells. Measuremenl of 
enzyme activity under Vmax conditions ensures the 
determination of the contenl of ac tive enzyme and 
a llows lhe degree of over-expression eompared with 
normal cells lo be eslab lished. Secondly, we deler-
mined the steady-state concentrations of several inter-
mediate metabolites to idenLify enzymes that may 
impose limila tions on the glyeolytic flux , a lthough such 
inferences do not always hold , pa rt ieularly for intenne-
diates involved in more than two reactions. 
To evaluate quantita t.i ve ly fl ux conlrol in tumoral 
glycolysis, we used lhe lheory of lhe melabolic co nlrol 
analysis [24J by applying an elaslicily analysis. This 
consists of lhe experimenta l determination of the sen si-
tivily of segments of lhe pa lhway t ow~ lrd s a common 
inLermediaLe. Once we had identified the main siles of 
flux control, we performed experiments lo determine 
which biochemical mechanisms a re involved in estab-
lishing why some enzymes exert significant control and 
others do no1. 
Results 
Maximal activities of glycolytic en zymes in 
hepatocytes and fast-growth tumor cells 
In normal freshly isolated hepatocytes, the enzymes 
with lower aClivi ty (and hence less content) were hexo-
kinase < PFK- I < a ldolase, enolase (Table 1). This 
activity pattern is in agreement with thal found in 
hepatocytes by olher authors [7,15J. In whole liver, lhe 
activities of all glycolytic enzymes were very simila r, 
excepl for pyruvate kinase, which was 3 times lower 
lhan lhaL oblained in isolaled hepalocyles (dala noL 
shown). In an attempt to establish a proliferating, non-
tumorigenic cell system, to make a more rigorous com-
parison wilh the lum or cell lines used in lhis work, we 
also isolated hepat ocytes from regenerating rat liver; 
organ regeneration was induced by prior treatment 
wilh CCl4 [0.39 g·( kg body weighl)- 'J for 12 or 24 h. 
In two difTerenl cell preparaLions, Lhe Vmax va lues of 
the glycolytic enzymes were essentially identical with 
those found for normal isolated hepatocyles (data not 
shown). 
In ral AS-30D hepaloma cells, Lhe enzymes with 
lower aeLi vity were hexokinase, PFK- I and aldolase, a 
1976 FEBS Journal 273 (20061 1975-1988 CI 2006 The Authors Journal compilation Cl 2006 FEBS 
 50 
A. Marin-H ernández el al. Control of glyeolysis in tumor eells 
Table 1. Maximal activity of glyeolytie enzymes in hepatocytes and tumor eells. AS-30D, HeLa and hepatocytes (65 mg protein'mL- 1
) were 
incubated in Iysis buffer as deseribed in Experimental procedures. Activities of all enzymes were determined in the cytosolic-enriehed frae-
tion at 37 oc. Specifie activities are expressed in U·(mg protein)- l. The values shown represent the mean ± SO with the number of different 
preparations assayed in parentheses. HK, hexolcinase; HPI, hexosephosphate isomerase; TPI, triosephosphate isomerase; GAPDH, glyceral-
dehyde-3-phosphate dehydrogenase; PGK, phosphoglycerate lcinase; PGAM , phosphoglycerate mutase; PYK, pyrwate lcinase; LDH, lactate 
dehydrogenase: G6PDH, glucose-6-phosphate dehydrogenase; a-GPDH, Cl-glycerophosphate dehydrogenase; PGM, phosphogtucomutase; 
NO, not detected. 
AS.JOD/ 
Enzymes Hepatocytes A5-30D hepatocytes HeLa 
HK 0.003 ± 0.002 131 0.46 ± 0.1" 171 153 0.02 ± O.006tt 141 
3.0 ± 1.7 141 HPI !I 0.4 ± 0.05 131 1.6 ± 0.7' 141 4 
PF K- l b 0.01 ± 0.002 131 0.2 1 ± 0.1' 141 22 0.09 ± 0.02 tl51 
0.2 ± 0.05151 
42±13131 
A1dolase 0.09 ± 0.02 131 0.23 ± 0.01' 141 2.7 
TP I !I 15.6 ± 5.6 131 56± 15 ' 141 3.6 
GAPDH 0.32 ± 0.07 131 1 ± 0.28' 131 2.7 2 ± 0.74 151 
2.5 ± 0.8 151 
13±6t 151 
1.4 ±1141 
0.36 ± 0.15151 
GAPOH !I 0.66 ± 0.23 131 0.9 121 1.4 
PGK 8.2 ± 5.8131 27 ± 10 l> (4) 3.3 
PGAM 11 ± 2131 20 ± 5l> (4) 2.3 
Enolase 0.11 ± 0.03 131 0.51 ± 0. 13' 131 4.3 
PYK 0.8 ± 0.36 131 6.6 ± 1.5" 141 8. 1 3 ± 1.3 t 141 
1.7 ± 0.6 131 
0.22 ± 0.08tt 151 
0.42 ± 0.13141 
ND 
LDH 4.4 ± 1.9131 6.4 ± 3.7 141 1.5 
G6PDH 0.03 ± 0.003131 0.05 ± 0.02 131 1.4 
PGM e 0.37121 0.2 1 ± 0.06 151 0.6 
Cl-GPDH d 0.11 121 0.002 ± 0.001 131 0.05 
l> P < 0.05 versus hepatocytes, H P < 0.005 versus hepatocytes, t P < 0.05 versus As..30D, t t P < 0.005 versus AS-30D . Student's Hest fO( 
nonpaired samples. 11 Activity in the reverse reaction . b Activity determined in the presence of 16-20 mM NH4 +. e The PGM activity was 
determined in the absence of glucose-l ,6- bisphosphate. d The reaction was started by adding OHAP. 
pa tte rn lha l also agrees \Vith that repo rted fo r the 
Slllne cells [7] and fo r o lher tumo r cell types [25]. The 
AS-30Dlhepatocyte activity ratio revealed lhm hexo-
kinase and , to a lesser extenl, PFK-I were lhe enzymes 
that were most over-expressed in tumor cell s; all o ther 
glycolylic enzymes, including glucose-6- phosphate 
dehydrogenase (G6PDH), were also over-expressed in 
AS-30D tumo r cells (Table 1). 
In H eLa cells, all glycolytic enzymes, except phospho-
glycerate mutase, also exhibited a higher activity than 
that shown by hepatocytes . However, in these hum an 
tumor ce lls neither hexokinase nor PFK- I were highly 
over-expressed as lhey \Vere in rodenl AS-30D cells. 
In H eLa cells, hexosephosphate isomerase, PFK-I , 
tr iosephosphate isomerase and pyruvate kinase, 
Logether with G6PDH, showed greater over-expression 
eompa red \Vith hepaloeytes (Ta ble 1). Vma> fo r phos-
phoglycerale mutase in HeLa eells \Vas 14 and 8 times 
lo\Ver than lha t found in AS-30D ce lls a nd hepaloeytes , 
respectively; such low phosphoglycerate mutase activi ty 
has al so been observed by olher a utho rs [25]. egligible 
ex-glyeerophosphate dehydrogenase aetivity \Vas fo und 
in both tumor ce ll types. A similar observation has been 
deseribed fo r lhe Mo rris hepalomas 3924A, 5123 D, 
7793 a nd 44 [26], \Vhieh are fasl or moderate-gro\Vth 
tumo r lines [27]. 
Glycolytic flux and intermediary concentrations 
As ex pected from lhe general increase in glycolyt ic 
enzymes, steady-state generalion of lac ta te in lhe pres-
enee o f 5 mM glucose \Vas marked ly higher in AS-30 D 
and He La cells (9- 13 times) than in hepal oeyles 
(T able 2) . In Lhe absenee of added gJ ucose, lhe glyeo-
Iytic flux diminished drastically in both tumor cell 
types, be ing negligible in AS-30D cells. T he differenee 
between the rates o f lactate formation with and with-
Table 2. Glycolysis in hepatocytes and tumO( cetls . AS-30D and 
HeLa ce lis (15 mg protein'mL - 1) and hepatocytes (30 mg pre>-
tein'mL - 1) were incubated in Krebs- Ainger medium as described in 
Experimental procedures. Under these conditions, the rate of lac-
tate formation in AS-30D cells was constant after 2 min and up to 
10 min from glucose addition (i.e. steady-state gtycolysis) . The 
intraceltular concentratian of Fru( 1,6)P2 was also invariant between 
the 2- and 10-min points, alter the addition 01 glucose (data not 
shown). Gtycotytic fluxes are expressed in nmot·min- 1·(mg cett pro-
tein)- l. The values shown represent the mean ± SO with the num-
ber of different preparations assayed in parentheses. The negative 
flux value indicates lactate consumption. 
Condition 
+ Glucose 
- Glucose 
Hepatocytes 
2.4 ± 1.7161 
-0.4±1161 
A5-30D 
21 ± 9 1401 
- 2.2 ± 2.6 1171' 
HeLa 
32± 10181 
7 ± 9161 
FEBS Journal 273 (2006) 1975-1988 CI 2006 The Authors Journal compilation CI 2006 FEBS 1977 
 51 
Control of glycolysis in tumor cells 
out added glucose indicates that neL glycolylic flux 
depends on external glucose, which was 8-9 times 
higher in AS-30D a nd HeLa eells than in hepa toeytes. 
The elevated glyeolytie flu x in HeLa eells in the 
absenee of added glueose \Vas probably sustained by 
endogenous sources, i.e. glycogen degradaLion. The 
content o f glycogen was appa ren lly not depleted in 
HeLa cells by the 10 min preineubation at 37 oC. In 
eontrast, the total dependenee of the glyeolytie flux on 
ex ternal glueose in AS-30D eells suggests depletion of 
glycogen induced by the 10 min preincubation al 
37 oc. The glyeolytie flux values reported in this \Vork 
are in the same range as reported for other tumor cell 
types [2]. 
The steady-state concenLraLions o f all glycolylic 
metabolites in AS-30D tumor eells also signifiea ntly 
increased , except for phosphoenolpyruvate and pyru-
vate (Table 3). In particular, Fru( I,6)P, inereased 250 
times and dihydroxyaeetone phosphate (DHAP) 16.6 
times. The cytosolic pyridine nucleotide redox slate 
(NAD H/NAD +), es timated from the laetate!pyruva te 
ratio. was more reduced in AS-30D cells, a situation 
that fa vors Hux through biosynthetie path\Vays. The 
eoncentration of A TP \Vas a lso higher in AS-30D cells 
than in hepatocytes; however, the ATP/ AD P ratio was 
similar (2.3 and 2.4). The lalter values are similar lo 
those previously reported [28] for normal organs su eh 
as ra t hea rt (5.7) and Ii ver (4.9), as \Vell as mouse Erhl-
ieh ascites cells (2.3) and 3924A hepatoma cells ( 1. 2). 
Table 3. Steady-state concentrations (mM) of glycolytic intermedi-
ates in nO(mal rat hepatocytes and hepatoma cells. See legend to 
Table 2 and Experimental procedures fO( experimental details. Val-
ues shown are the mean ± SO. The number of experiments is 
shown in parentheses. NO. not detected; NM. not measured; 
OHAP. dihydroxyacetone phosphate; G3P. glyceraldehyde 3-phos-
phate; PEPo phosphoenolpyruvate; Lac. lactate; Pyr. pyruvate . 
Metabolite Hepatocytes AS-300 HeLa 
Glucose NM 6.2. 1 (31 NM 
Glc6P 0 .96.0.16 (31 5.3.2.6" (231 0.6 • 0.16tt(41 
Fru6P 0.4 • 0.03 (31 1.5 ± 0.7"· (22) 0.22 • 0.09tt (41 
Fru(1.6)P2 0.1 • 0.05 (31 25. 7.6" (191 0.29 • 0.06tt (41 
DHAP 0.6.0.1 (31 10.2.3"(141 0.93 • 0.07tt (31 
G3P 0 .09.0.01 (31 0.9 ± 0.4*(7) ND 
PEP 0 .1 (21 0.1 .0.02 (31 0.32 (21 
Pyr 1.6.0.7 (31 2.1 • 1 (71 8.5 • 3.6tt (51 
Lac • 9.6. 1.3 (31 27. 11' (31 33 (21 
ATP 3.6 • 0.24 (31 5.6. 12' (91 9.2 • 1.9t (41 
ADP 1.6.0.6 (41 2.4 • 0 .7 (71 2.7.1 .3 (31 
Lac/ Pyr ratio 6 .3 12.9 3 .9 
11 L-Lactate was intracellularly Iocated . • P < 0.05 versus Hepato-
cytes . .. P < 0.005 versus Hepatocytes. tP < 0 .05 versus AS-300. 
ttP < 0.005 versus AS-300 . 
A. Marin-l-l ernandez er al. 
In contrast with AS-30D cell s, the steady-state con-
centrations of G\c6P, Fru6P, Fru( I,6)P2 and DH AP 
in HeLa cells were similar to those observed in hepato-
cytes, whereas the A TP and pyruvate concentrations 
\Vere 1.6 and 4 times higher than in AS-30D cells 
(Table 3). 
Determination of flux control coefficients for 
glycolysis in hepatoma cells 
Me La bolic conLrol ana lysis establishes how to deter-
mine quantitali vely the degree of cont rol (named flux 
control coefficient. CEi) that each enzyme Ei exerts 
over the metabo lic flux J [24]. In the oxidative phos-
phorylation paLhway, c'Ei va lues can be determ ined 
by titrating the flux \Vi th specifie inhi bit ors [24,29,30]. 
However, there are no specifi c, permeable inhibitors 
fo r eaeh glyeolytie enzyme. 
An alternative approach called elasticity analysis 
[3 1-34] consists of experimental determination of the 
sensitivity o f enzyme blocks Lowa rds a common inter-
mediate metabolite m. By applying the surnma tion and 
connectivity theorems of metabolic control analysis 
(see Eqns I and 2 in Experimenta l procedures), the 
CJ
Ei values can be caJculated. Varia tions in the steady-
state activi ty o f the enzyme blocks ca n be attained by 
adding difIerent concentrations o f the initial subst rate 
or inhibitors o f either block, which do not have to be 
specific fo r only one enzyme but they do have to inhi-
bit only one block. The block of enzymes that gener-
ates the comlTIon intermedia te is named the prod ucer 
block, \Vhereas the block of enzymes eonsuming that 
metabo lite is named the consumer block . 
For glycolys is, and olher pathways, any metabo lite 
may be used as the common intermediate that con-
nects prod ucer and consumer branches. However. to 
reach consistent results. it is more convenient to use, 
as common intermedia tes, metaboli tes that are present 
at relatively high concenlrations and that are only con-
nected to the specifie pathway, sueh as F rll (I,6)P, . 
Altho llgh o ther metabolites sueh as Glc6P, Fru6P a nd 
DH AP may be present at high concentrations. they are 
eonnected \Vith other path\Vays (glyeogen synthesis and 
degradation, pentose phosphate cycJe , glycerol and lri-
acyglycerol synthesis). Ho wever, lhis last gro up o f 
metabo lites may still be used in e lasticity-based analy-
sis as long as the flu x througll the o ther pathways is 
low (or it is assumed to be negligible) [33,35,36] o r by 
actually determining the e lTect of the branching path-
ways on the main flux and on the concentralion o f the 
eonneeting metabolite [23]. 
To determine the elastic ity coefficients of the con-
sumer block for the common metabolite (EEim ) , we 
1978 FEBS Journal 273 (2006) 1975-1988 e 2006 The Authors Journal compilation e 2006 FEBS 
 52 
A. Marín-Hernández er al. 
incubated hepatoma cells with different glucose con-
centrat ions (4-6 mM) or with the hexosephosphate 
isomerase inhibitor 2-deoxyglucose (0.5-10 mM), which 
induced variat ions in flux and in the sleady-state con-
centrat ions of the melaboli te. The elasl icity of the pro-
ducer block was de term ined by tit ra ting flux with the 
LDH inhibi tor, oxa late (0.5-2 mM), or lhe glycera lde-
hyde-3-phosphate dehydrogenase inhibi tor, arsenile 
(5- 100 ~ M ) . Thus, the glycolytic flu x (measured as the 
ra te of lacla te formation) a nd lhe co ncenlration of 
severa l intermed iates [G lc6P, Fru6P, Fru( I,6) P, and 
DHAP] were detennined under bot h conditions. The 
ta ngenl s to the curves, o r the slra ight Hnes, taken at 
lhe reference, conlrol points (100 %) in lhe norma lized 
plots of flux versus [metabolite] obtai ned with glucose 
and oxalate, or 2-deoxyglucose a nd a rsenite, represent 
the elasticities towa rds lhe inLermedia Le metaboli te of 
the consumer and producer blocks, respecLi vely. 
We a re awa re tha t lhe experimental points in lhe 
flux vers us [metaboli te] plot should be fi tted to a 
hyperbolic curve rather than lo a straight line, as most 
of lhe glycolylic enzymes and tra nsporters fo lJow a 
Michaelis-Menlen kinetic pa Ltern; a nea r-linea r rela-
tion between rate and substrate concentration might 
be atta ined when the product concentral ion va ries con-
comita ntly. However, lhe lack of suffici enl experimen-
ta l point s nea r lhe reference, una ltered stale may 
generate high, un rea listic slope values ( ~ 2) fo r the 
es timation of elasticity coefficients (F ig. 1) either by fi t-
ting to hyperbolic or linear equal ions. The definiti ons 
of the elaslici ty a nd flux cont rol coefficients as well as 
the lheorems of metabo lic control a nalysis a re based 
on d ilTerentia ls. However, it was not easy to proo uce 
small changes (and much less infinitesimal cha nges) of 
flux a nd meta boHLe concenlration by using lhe experi-
menta l protocols described . In consequence, slope 
values were calculated with both a pproximations, non-
linea r hyperbolic fit ting a nd linea r regression. In gen-
eral, similar elasticity coefficients resulted from eilher 
approximation, a lLhough less dispersion was a ttained 
with the linear regression (see legend lo Table 4 for 
values). 
Titration of the glycolytic fl ux with exogenous gl u-
cose and oxalate (Fig. lA), or with 2-deoxyglucose 
a nd arseni te (Fig. lB), induced changes in flu x a nd the 
Fru( I,6) P2 concenlration. Analysis of both segments 
showed tha t the Fru( I,6) P, consumer block (formed 
by enzymes from aldolase to LDH) showed a higher 
elasticity than lhe producer block (comprising GluT to 
PFK- I). In lhe fi rst case (with glucose or 2-deoxy-
glucose), the slope had a positive va lue beca use 
Fru(l ,6) P, is a subst ra te fo r the consumer block. On 
lhe other hand, with oxalate or arsenite t itra tion, the 
A 
160 
120 
VJ .¡¡; 
>-
(5 
80 u 
~ 
l? 
-:.R. o 
40 
O 
B 
60 
Control of glycolysis in tumor cetts 
+ Glucose 
o 
, . 
, ... 
o 
'. 
J 
o 
m = 3.56 
m = -1.7 
... 
+O xa ~l a~l~e __ ......... 
80 100 120 140 
% F-1 ,6-BP 
+ 2DOG +Arsenile 
o 
50 
, , m = -O. 17 
, 
m = 1.14 ... 
100 150 200 250 300 
% F-1 ,6-BP 
350 
Fig. 1. Experimental determination of elasticity coefficients for glyc-
oIyt ic intermediates in tumor cells . AS-30D hepatoma cetts (15 mg 
protein·mL - 1) were incubated in Krebs-Ringer medium at 37 oc. 
After 10 min, different concentrations of glucose (O, 4-6 mM) and 
oxalate 1.0\, 0.5- 2 mM) in (A), or 2-deoxyglucose (O, 0.5-1 0 mM) 
and arsenite (0\ , 5- 100 ~ M) in (B), were added to the cell suspens-
ion. When a glycolytic inhibitor was added, glucose was kept 
constant at 5 mM. The hexokinase and PFK-1 activities, which are 
part of the Fru(1.6)Prproducing block, were not affected by 10 mM 
oxalate or 1 mM arsenite (data not shown), Thus, the effect of 
these two inhibitors on flux was due to their interaction with 
enzymes of the Frull ,6)P2-consuming block, most likely LDH [661 
(data not shown) and glyceraldehyde-3-phosphate dehydrogenase 
1671. The va lues of the straight lines, or the tangents to the curves, 
at 100% Fru( l ,6)P2 (m), which are the elasticity coefficients in 
these normatized plots, are shown on the traces. 
slope had a nega tive va lue beca use Fru( I,6) P2 is a 
prod uct of the producer block, i.e. Fru( I,6)P, accumu-
lation inhibits lhe producer activily. 
FEBS Journal 273 (2006) 197&-1988 e 2006 The Authors Joumal compilat ioo e 2006 FEBS 1979 
 53 
Cont ro l o f g lyco lys is in tumor cell s A. Marín-Hernández er al. 
Table 4. Control (q and elast icity (&) coefficients values of AS-30D hepatoma cells. Control coefficients were calculated from elasticity co-
efficients, derived Irom data such as those shown in Fig. 1, and applying summation and connectivity theOfems (Eqns 1 and 2; see Experi-
mental procedures) . ¿'m, elasticity 01 producer block; cC
m. elast icity of consumer block; cJP' control coefficient 01 producer block; cJc, control 
coeHicient of consumer block. A11 the elasticity coefficients shown in the table were calculated by using slope vaJues derived Irom linear 
regression, although similar values were attained by nonlinear regression . For ins tance, the nonlinear regression for the titrations with 2-de-
oxyglucose and arsenile gave CCFBP and &PFBP of 1.99 ± 0.79 and - 1.5 ± 1.6, which yielded the Ce and Cp 01 0.39 ± 0.24 and 0.61 ± 0 .24. 
respectively. The number of experiments is shown in parentheses, and values are mean ± SO. OHAP, Oihydroxyacetone phosphate . 
Metabol ite ,cm c'c 
p 
e m c'p 
+ Glucose + Oxalate 
Glc6P 2 1 ± 17 161 0.29 ± 0.17 131 - 0.86 ± 0.41 131 0.71 ± 0.17131 
Fru6P 1.2 ± 0.45 141 0.31 ± 0.14131 - 0.67 ± 0.43 131 0.69 ± 0.14131 
Fru(1,6lP2 2.2 ± 0.95 151 0.44 ± 0.08 151 - 1.35 ± 0.66 161 0.62 ± 0.08 151 
DHAP 1.27 ± 0.47 151 0.49 ± 0.18151 - 1.5 ± 1 151 0.51 ± 0.18151 
+ 2-0eoxyglucose + Arsenite 
Fru(1,6lP2 0.93 ± 0.2 131 0.24 ± 0.06 131 - 0.25 ± 0.1131 0.75 ± 0.05 131 
+ Glucose + Arsenite 
DHAP 1.18 ± 0.2 131 0 .37 ± 0 .15 131 - 0.5 ± 0 .1 131 0.63 ± 0.14131 
h is worth emphasizi ng lhat, owing to the multitude 
of variables involved in deLermining flu x and in temle-
diary concenlra lions, which have lo be kepl consla nl 
during lhe experimenla l determinal ion of elastici ties 
towa rds one m e t ~ l bo l ite , lhe d ispersion of lhe experi-
menul l points ca n be considerable in some cell prepa-
ralions. This can be a pprecialed in Fig. l and in lhe 
values shown in Table 4. Nonelheless, it is possible to 
reach relevant concl usions a bout which steps exert sig-
nifica nL flux control of glycolysis a nd which steps have 
low or negligible control (Ta ble 5). 
The elasticily coefficients, estim ated from experi-
ments such as lhose shown in Fig. 1, are summarized 
in Table 4. The flux cont ro l coefficients derived from 
the elast ic ity coefficients a re also shown. These data 
cJearly estab lished that the ma in cont rol of the glyco-
lylic flux in AS-30D cells resides in the upstream parl 
of the pa thway. The experiments with glucose a nd oxa-
la te show a high value fo r the flu x control coeffici ent 
of lhe GJc6 P producer block [CJelGk6PJ]' which ind i-
cales lha l GluT, hexokinase a nd perha ps lhe degrada-
Tab le 5. Oistribution 01 control 01 glycolysis in AS-300 cells. GluT, 
Glucose transporter; HK, hexokinase; HPI, hexosephosphate isCF-
merase; TPI, triosephosphate isomerase; GAPOH, glyceraldehyde-
3-phosphate dehydrogenase; PGAM, phosphoglycerate mutase; 
PYK, pyruvate kinase; LOH, lactate dehydrogenase. 
Enzymes or branches QlEi 
GluT + HK 0.71 
Pentose phosphate cycle + HPI + glycogen synthesis -0.02 
PFK-1 0.06 
Aldolase, TP1, GAPDH, PGAM, enolase, PYK, LOH, 0.25 
Pyr branches, ATP demand 
1.00 
lion of glycogen, steps that lead to the fo rma tion of 
G Ic6P, were the sites thm exerted most of the flux con-
trol. In turn, lhe experiments wit h 2-deoxyglucose and 
a rseni le revealed lhat lhe producer block of Fru(J ,6) P, 
exerted most of the cont ro l, which indicales that flux 
cont ro l was mai nly located in GluT, hexokinase a nd 
glucogenolysis together wilh hexosephosphate ¡so-
merase and PFK-1. 
The same concl usion may be drawn from the high 
CJ
P(Fru6 P) value (Table 4). Howeve r, the glycolylic 
flu x was negligible in the absence of added gl ucose 
(Table 2), indicatíng lhal G lc6P a nd Fru6P forma-
lion fr om glycogen degradation was no t significant 
in AS-30D cells. Hence, from lhe d ifference belween 
lhe values of C p(Glc6P) and C P(Fru6P), which were 
determined under lhe same experimental conditi ons 
with glucose and oxalate, it was possible lO calculate 
a specific flu x conl ro l va lue of -0.02 fo r lhe GJc6P 
branches, pentose phosphale cycle a nd glycogen syn-
lhesis (Table 4). 
Secause of lhe less lha n perfecl malch of e e a nd C e 
values es timaLed fro m lhree d ifTerent experimental pro-
locols (Table 4), il \Vas difficult lo oblai n a reliable flu x 
cont ro l coefficient fo r lhe DHAP consumer bra nch. 
Wilh 2·deoxyglucose and arseni te, the C p (Fr u( 1.6) p2) 
value of 0.75 suggesls thal lhe res t of lhe pa lhway (from 
a ldolase lo LDH) exerls a flu x conl ro l of O.25 (Ta ble 5). 
However, the C C(OHAP) values with g)ucose and oxalate 
o r with glucose and arsenite were 0.49 a nd 0.37, respect-
ively, which revealed some discrepancy with Lhe 2-
deoxyglucose and arseni te protocol. If lhe tota l 
summalion of e Ei values was higher than 1.0, lhen 
bra nching in t.he middle and lower segments of glycol-
ysis might be significa nt, bringing about negative flu x 
cont ro l coefficienls. 
1980 FEBS Journal 273 (2006) 1975-1988 e 2006 The Authors Journal compilation CI 2006 FEBS 
 54 
A. Marin-fl ernandez et al. 
The flux contro l coefficient of PFK-I (Table 5) was 
estimated from lhe CP(Fru(1.6) P2) va lue attai ned with 
2-deoxyglucose and a rsenile minus lhe c'P(Glc6P) value 
altained with external glucose and oxalate (Table 4). 
Only posit ive d ifTerences between Cp va lues are to be 
ta ken into account fo r elucidating flux control fo r spe-
cHic enzymes. With negative d ifTerences [for insta nce, 
wi th C P(F ru( 1,6)P2) minus c'P(Fru6P) both a tta ined with 
glucose and oxa la te], the explanation is tha L there is a 
pathway bra nch at the measured metabolite concentra-
tion, o r that the experi mental dispersion masks small 
d ifTerences, o r that indeed there is no difTerence 
between the enzyme block s ana lyzed . 
Kinetic analysis 01 tumoral hexokinase and PFK-' 
To understand why hexokinase retained a significa nt 
degree of contro l on glycolytic flux, despite its high 
over-expression, and why PFK- I cont ro l became negli -
gible, the kinetic properties of the two enzymes were 
ana lyzed in cell extracts . T he affi nity of hexokinase fo r 
glucose a nd A TP in both the cytosolic and mitochond-
ria l fractions (Table 6) was in the sa me range as 
reported by Wilson [37] for hexokinase from nontu-
morigenic mamm alia n tissues. Hexokinase was equally 
Control of glycolysis in tumor cell s 
distributed bet ween the cyt osol a nd mitochond ria in 
AS- 30D hepatoma cells. 8 0 th hexokinase isoenzymes, 
cytosolic a nd mitochondrial, were 81-93% inhibi ted by 
I mM Glc6 P (Fig. 2A). 
The P FK-l in the cytosolic-enriched fraction from 
AS- 30D cells exhibited Km a nd Ka.5 va lues for ATP 
and Fru6P (Table 6) similar to those reported fo r 
PFK- I from other tumor cell lines [13,15]. In the 
absence of added eITectors, the PF K-I kineLic pat lern 
was sigmoidal with respect to Fru6P [a ltho ugh 
", 8-12 mM (NH.¡),S04 coming from the coupling 
enzymes was present] , and hyperbolic with respect to 
low concentrat ions of AT P (0.0 1- 1 mM). At high con-
centrations ( > I mM), A TP was inhibito ry fo r PFK- I 
ac tivity. Ci lrale \Vas also inhibitory al relatively low 
« 1 mM) Fru6P concentrations. However, when 
1. 5 mM Fru6P (physiological concentration) o r higher 
concentrations were used, citra te was innocuous, even 
at concentral ions as high as 10 mM (dala not shown), 
in the presence of 8- 10 mM (N H4),S04. Fru(2 ,6) P, 
was lhe most potent acLi vator of tumoral PFK- l , fo l-
lowed by AM P and H4 + (Table 6). 
The mea n ± SD intracellula r concentrations of 
AM P and citrate de termined under glycolyt ic steady-
sta te conditions in AS-30D cells were 3.3 ± 1.4 
Table 6. Tumora l hexokinase and PFK-l kinetic parameters. The activities of hexokinase and PFK-1 were determined at 37 oC as described 
in Experimental procedures . For hexokinase, the Km value fer ATP was determined in the presence of 5 mM glucose, whereas that fer gluc-
ose was determined with 10 mM ATP. For PFK-l, the Km value fer ATP was determined in the presence of 10 mM Fru6P. whereas the Ka.s 
value fer Fru6P was determ ined with 0.25 mM ATP. The ammonium concentration in the assay mixture, proceed ing from the coupling 
enzymes, was 16-20 mM. The Ka.s values for NH4 +, AMP and Fru(2,6)P2 were determined in the presence of 2 mM Fru 6P and 0 .8 mM ATP, 
and with Iyophilized coupl ing enzymes (I.e . in the absence of contaminating ammonium). The number of independent experiments is shown 
in parentheses . Units of Km and Ko_5 are ¡..lM; Vmax, U·(mg protein)- l. 
Hexokinase 
M itochondrial 
Cytosolic 
Type 111 
Type 11 11 
Type 11 1 a 
PFK-l 
a Values taken from 1371. 
AlP 
Km 
696 ± 180 (3) 
990 ± 50 (3) 
500 
700 
1000 
AlP 
Km 
14 ± 2 (3) 
NH4 + 
Ka. 
1400 ± 800 (3) 
Fru(2,6)P2 
Ka .• 
0.96 ± 0.3 (3) 
1.96 ± 0.4 (3) 
0.52 ± 0.1 (3) 
0.2 ± 0.09 (3) 
V~, 
0. 51 ± 0.2 (3) 
V~, 
0.52 ± 0.16 (3) 
FEBS Journal 273 (2006) 1975-1988 e 2006 The Authors Journal compilatlon te 2006 FEBS 
Glucose 
146 ± 12 (3) 
180 ± 40 (3) 
30 
300 
3 
Fru6P 
Ka.5 
200 (2) 
AMP 
Ka .• 
100 ± 50 (4) 
1.65 ± 0.18 (3) 
0.44 ± 0.1 (3) 
0.2 (2) 
V~, 
0.58 ± 0.1 4(4) 
1981 
 55 
Contro l of glycolysis in tumor eells 
A 
e 
: ~ 
tí 
<{ 
,g o 
B 
100 
80 
60 
40 
20 
o 
0.0 0.2 0.4 0.6 0.8 1.0 
G6P (mM) 
¡-----1-----11) 
e) 
}----------' 100 / ' ?-- / --" 
d) 
e) 
I ~ b) 
~ a) 
80 . / 1/" 
,¡;- l ' .• 60 I T 
Ji 
~ :~~ 
00.00 0.05 5 10 1'5 20 
(mM) 
0.0 0.2 0.4 0.6 0.8 
Aetivi ty (U/mg protein) 
Fig. 2. Effect of modulators on tumoral hexokinase and PFK-l. (A) 
Inhibition of mitoehondrial bound (O) and eytosotie hexokinase CA) 
by GIc6P. Values shown represent the mean ± SD from three dif-
ferent preparations assayed, exeept for the experiments with cyto-
solie hexokinase at O and 1 mM Gle6P, in whieh nine different 
preparations were analyzed. (B) Effect of modulators of PFK-1. 
PFK-l act ivity was determined in the presence of 1.5 mM Fru6P 
and (a) 0.8 mM ATP; (b) 3.9 mM ATP; (e) 3.9 mM ATP +1.7 mM eit-
rate; (d) 3.9 mM ATP + 1.7 mM citrate + 3.2 mM AMP; (e) 3.9 mM 
ATP + 1.7 mM citrate + 3.2 mM AMP + 51-1M Fru(2,6)P2 ; and (f) 
3.9 mM ATP + 1.7 mM citrate + 3.2 mM AMP + 50 ~I M Fru(2,6IP2 . 
Values shown represent the mean ± SD from three different prepar-
ations assayed. Inset : Activation 01 PFK-l by different eoneentrat-
ions of Fru(2,SIP2 (e l. AM P (. ) and NH 4+ (A l, in the presenee of 
2 mM Fru6P and 0.8 mM ATP. 
(11 = lO) and 1.7 ± 0.7 mM (11 = 6) , respectively. 
Thereafter, Lhe PFK-I activiL y was determjned in lhe 
presence of the intracellular concentmli ons of its sub-
slra les (A TP, Fru6P), inhibilors (ATP, cilrale) and 
aClivalors [AM P, Fru(2,6)P,]. PFK- I aClivily was fully 
inh ibited in lhe presence of AT P and citrate; this activ-
ily was only partially reslored by AM P (Fig. 28). The 
A. Marin-Hernández er al. 
potent ATP + citrale inhibilion was tolally overcome, 
or even surpassed, by Fru(2,6)P2 at concenlrations 
found in lumor cells (1 8-2 1]. 
Discussion 
Distributio n of glycolytic fl ux control 
Metabolic control analysis has been applied lo deter-
mine the cont rol structure of glycolysis in several nor-
mal mammalian systems, such as human erythrocytes, 
rat heart and mouse skeleLal muscle extracts [22,23,38]. 
Wi lh this quantitati ve fram ework, hexokinase and 
PFK- I have been idenlified as lhe main co nlrolling 
steps. Fast-growth tumor cells develop a nontypical 
melabolism [39,40], which ineludes an acceleraled gly-
colylic nux. As glycolysis in different tumor lines has 
been considered to be an extremely faSl pa thway [39], 
lhe idenlification of which enzyme(s) conl rols glycolyl-
ic nux becomes c1inically relevanl. 
The lO enzymes of lhe AS-30D hepaloma glycolytic 
palhway showed higher ac tivity Lhan in norma l mt 
hepa tocytes. Despite showing lhe greatest over-expres-
sion, hexokinase and PFK- I, together with aldolase, 
had lhe lowesl V"'a> values (Table 1). Olher groups 
ha ve described a sioti lar pallern for AS-30D [7] and 
o ther tumor cell lypes [25]. However, in all previous 
papers [7,15,25,41,42] il was difficull lo eslablish a 
strict activity sequence arder, as not all glycolytic 
activilies were determined; moreover, the activity 
assays were performed aL nonphysiologica l pH ( > 7) 
a nd temperalure « 37 OC) . Indeed, detennining Vmax 
under near-physiologica l conditions establishes the true 
content of active enzyme, which is not possible when 
mRNA , or protein, is measured. 
The tumoral hexokinase and PFK- I, enzyl11es that 
in normal tissues control the nux (100% in hum an 
erylhrocy1es, 59% in isolaled ral hean, and 100% in 
ral skelelal musele reconsl iluled palhway) [22,23,38], 
exhibiled lhe hi ghesl activily enhancemenl (306-fold 
a nd 22-56-fold increase versus hepalocyles; Table 1). 
In add ilion, lhe cylosolic concenlralions of Glc6P and 
Fru( I,6) P2, which a re products of hexokinase and 
PFK-I , respeclively, increased by fivefold and 250-fold 
(Table 3). The upper limils of ac tivity incremenl \Vere 
eSlablished by laking inlo accounl bolh lhe cylosolic 
a nd mitochondrial hexokinase activi ties (Table 6), and 
the PFK-I maxi mal aCLi vity attained in the presence of 
Lhe activalor Fru(2,6) P, (Fig. 28 insel). 
The nux control coefficients caJculated from elaslici-
ties are, to a great extent , determined by the definition 
of producing and consuming blocks [24,43,44]. Elasli -
city-based analysis requ ires that (a) lhe metabolic 
1982 FEBS Journal 273 (2006) 1975-1988 CI 2006 The Authors Joornal compilation CI 2006 FEBS 
 56 
A. Marín-Hernández et al. 
pathway reaches a quasi-sLeady-sLate (see legend to 
Table 2 fo r experimenlal de lails on how lhe glyeolylie 
sleady-slale flux was eSlablished), a nd (b) lhe inlenne-
diales lhal link lhe bloeks are no l signifiea ntly affecled 
by olher palhways. However, sorne of lhe glyeolytie 
intermedia tes used in this work for the estimation o f 
elasl ieity eoeffieienls sueh as GIc6P, Fru6P and DH AP 
are indeed connected to other pathways, and hence 
changes in the flux of glycogen synthesis and degrada-
tion, pentose phosphale cycle, and glycerol and triacyl-
glycero l synlhesis might aITect the concentrations o f 
lhese melaboliles . 
Another impo rtanl ass umption (c) in th is e lasticity-
based analysis is lha l producing a nd consuming blocks 
affeel eaeh o lher only lhrough lhe eommon melabo lile. 
However, the glycolytic segments analyzed in this wo rk 
may also interacL through the mo iety-conserved pools 
of ATP- ADP- AM P and NAD(P)H- NA D(P) + . The 
discrepancy in the calculated flu x co ntrol coeffic ients 
for producing and consuming blocks o f DHAP and 
Fru( I,6) P2, by using diITerent ex perimental pro locols 
(T able 4), mighl pa rtly be due lo va rialion in Lhe nieo-
tinamide nucleotide redox state and adenine nucleotide 
energy state, respective ly. 
Furthermore, the elasl icity-based analys is might be 
fla wed, and the contro l distri bution repo rted mighL be 
erroneous, if the concentrations of adenine nucleot ides 
also varied in lhe tilration experiments. It is worlh 
mentioning that several other sludies involving elasti-
city-based anal ys is have nol la ken in lO account the 
potential interactions between prodllcing and consu-
ming bloeks mediated by the pool of adenine nueJeo-
lides [23,33-36,45]. Therefo re, the varialion in lhe 
co ncentrations o f ATP, AD P and AM P was deter-
mined under the conditions o f lhe experiments shown 
in Fig. l . The results indica te that none o f the adenine 
nueJeotides ehanged signifieantly (11 = 4) on varying 
either glueose (from 4.5 lo 5.5 mM) o r oxalate (fr om 
O to l mM); with 2 nHvl oxa late, a 10-30% increase in 
ATP, ADP and AM P was observed . T hus, these find-
ings support the control distribution o f glycolysis 
(T able 5) der ived from the elas tieity-based ana lysis 
shown in Table 4 . 
DH AP is certa inly connecled wilh nonglycolytic 
reactions that involve N AD + such as cx-glycerophos-
phate dehydrogenase, which was however, negligible in 
AS-30D ceJls (Table 1). Li kewise, Fru( I,6) P2 format ion 
may affeel lhe ATP/ ADP ratio, which in l uro establi-
shes communication with glycolytic downstream reac-
tions (phosphoglycerate ki nase, pyruvate kinase) and 
wilh o ther energy-dependent reacLions (adenylate 
kinase, A TPases, biosynthetic pathways). Moreover, 
PFK- l is activa ted by AM P (Table 6 and Fig. 28), 
Control of glycolysis in tumor cel ls 
which connects with lhe adenylate kinase reaction. 
Therefore, the connectivity theorem lIsed here for lhe 
ca lculation of lhe flux cont rol coefficients from elast ici-
Ues towards some intermediates [Eqn 2 in Experimen-
ta l procedures] was apparently incomplete and too 
sim plistic lo describe all interactions, and iL may be 
necessary to consider more complicated relationships 
[24,43,44]. 
NotwithsLanding the aboye argumenLs, lhe elastic ily-
based analysis, as used here, revealed tha t the Glc6P-
produeing block (GluT and hexokinase) exerled lhe 
main control o f flux. The finding that the intracellu la r 
concentration o f free glucose was high and saturati ng 
fo r hexokinase (Table 3) suggests that most o f lhe con-
lrol exerted by the Glc6P-producing block might reside 
in hexokinase. Moreover, high over-express ion o f 
GluT has been docllmented for HeLa cells and other 
human tumor cell types [46,47]. However, before we 
ca n conclude that hexokinase is the main controlling 
step, we should further examine lhe contenl and activ-
ity of the GluT under phys iological conditions, as 
product inhibi tion of Lhe GluT aC li viLy has no t been 
explo red [48]. 
In He La ceJls, aJl glycolyt ic enzymes exce pt LDH 
were also over-expressed compared with hepatocytes. 
However, it should be nOLed tha t lo achieve a more 
rigorous compari son, a normal pro liferating endot hel-
ia l ceJl line should be used instead of hepatocytes. The 
over-expression in He La cells was much less lhan that 
in AS-30D ceJls (Table 1); however, the glyeolytic ra tes 
were similar (Table 2). This was probably d ue lo a 
high ra te o f degradation of glycogen (high lactate fo r-
malion in lhe absence of added glucose; Table 2) a nd 
am ino acids (e levated concentration of pyruvate; 
Table 3) , the prod ucls o f which bypass hexokinase, lhe 
presumed main co nt rolling step, lo enter the glycolytic 
pathway. 
In HeLa ceJls, the conlrol o f glycolylic flux may a lso 
reside in hexokinase, as in no rmal hepal ocyles and 
AS- 30D cells, as weJl as in PFK- I, because lhe two 
enzymes have the lowesl V m a.'( values, and they are no t 
highly over-expressed . The eighLfo ld lower concenlra-
lion of GJc6P in HeLa ceJl s, compa red with AS- 30D 
ceUs, is expected lo exert a low or negligible inhibition of 
hexokinase. The low Glc6P concentration in HeLa cells 
may be relaled to a higher activily of lhe Glc6P bran-
ches, glycogen synLhesis a nd penlose phosphale cyeJe. 
Indeed, G 6PDH ac ti vily was 4-7 times higher in HeLa 
cells lhan in hepalocytes a nd AS-30D ceJls (Table 1). 
A higher A TP concenlration in HeLa cells than in 
AS- 30D ceJls suggesls a lower ae tivit y of ATPases a nd 
o ther ATP-dependent cell processes or, alternatively, 
lhat A TP production by oxidat ive phosphoryla tion 
FEBS Journal 273 12006) 1975-1988 C 2006 The Authors Journal compilation e 2006 FEBS 1983 
 57 
Contro l of glycolysis in tumor cells 
was faster. Indeed, lhe ra te of o ligomycin-sensitive 
respiration, which reflects the rate of oxidat ive 
phosphorylation [40,49], in lhe presence of 5 mM 
glucose + 0.6 mM gl utamine was 43 ± 4 ng-aloms 
oxygen·min- I. IO- 7 cells (11 = 4) in AS-30D cells and 
92 ± 16 ng-aloms oxygen·min- I. IO- 7 cells ( 11 = 6) in 
HeLa cells. As the enzymalic assay of ADP delermines 
lolal bul nol free ADP, lhe A TPI ADP ratio was nol 
highly reliable as an indicalor of the cellular energy 
sta lus. 
There are lwo Cl-glycerophosphale dehydrogenase 
isoenzymes in mammalian cells, one bound to the mit-
ochondria l inner membrane and another in the cytosol, 
which regllla te lhe cytosolic ADH/NAO + ratio and 
are involved in lhe synthesis of lri acylglycerols [50] . The 
activi ty of the cytosolic isoenzyme decreases or becomes 
negligible in fasl-growlh hepalomas [26,27] and AS-30D 
and HeLa cells (Table 1), which may induce DH A P 
aCCllmulalion (Table 3). An alternalive roule for tr iacyl-
glycerol synthesis has been described that does not 
require ll-glycerophosphate. This palhway slarl s with 
the acylation of OHAP, in a reaction catalyzed by 
OH AP acyltransferase, which is present in rat ti ver, kid-
ney, spleen and ad ipose lissue; al the subce llu lar level, 
this enzyme is localized in lhe mitochondrial and micro-
soma! fractions [51]. Such a roule mighl be operati ng in 
t umor cells. 
Biochemical mechanisms underlying the 
evaluated distribution 01 Ilux control 
1 t should be emphasized that lhe ana lysis of Vmax v~ li ­
lIes only (i.e. cellula r contenl of active enzyme) lo 
reach conclusions on lhe melabolic pat hway control is 
certa inly incomplete, as the enzyma tic activities are 
determined in the absence of their physiological activa-
tors and inhibito rs, at sat urati ng substrate co ncentra-
tions, and in lhe absence of products; these factors 
discard the role of lhe reversibi lity of react ions under 
physiological conditions on conlrol of flu x [52]. 
Accllmlllation of prodllcts may decrease the forwa rd 
reaction. In th is rega rd, it is well documented that 
Glc6P is a polenl inhibilor of hexokinase-I, hexo-
kinase- II and hexokinase-II I [37]. In consequence, lhe 
increased hexokinase activit y might be cOllnLer-
balanced by stronger Glc6P inh ibition. One way lo 
circumvenl this blockade is to over-ex press hexokinase-
111, an isoenzyme wi lh a higher K; for Glc6P (O. I mM). 
Alternat.ively, hexoki nase binding to mitochondria may 
prolecl il aga insl Glc6 P inhibilion [7,53]. However, 
the Glc6P inhibition was strong and similar for bOlh 
mitochondria l alld cytosolic hexokinase isoenzymes 
(Figure 2A) , when lhe activi ly was assayed under 
A. Marin-Hernández et al. 
Ilea r-physiologica l conditi ons of pH (7.0) a nd lempera-
lure (37 oC) and high concenlrations of glucose 
( > I mM) and Glc6P (;;' I mM). On the o ther ha nd, 
Nakashima el al. [7] and Buslamanle el al. [53] deler-
mined Glc6P inhibition of mitochondrial hexokinase at 
22-30 oC, pH 7.9, and in a hypotonic medium wÍlh 
nonphysiological co ncenlralions of glucose « I mM) 
and Glc6P « I mM). The presence of lhis Glc6P regu-
latory mechanism in tumoral hexokinase sllpport s an 
essential role for this enzyme in lhe control of flux. 
Fou r hexokinase isoenzymes have been identified in 
mammalia n cells: Type-l , 11, 111 a nd IV (glucokinase), 
from which lhe fi rsl lhree are Glc6P-sensitive [37]. 
Hexokinase-I and hexokinase-II may bind lo lhe o ll ter 
mitochondrial membra ne, as they have a specific 
hydrophobic N- lermina l segmenl [54]. Hexokinase- l is 
predominantly located in brain, kidney, retina and 
breast, whereas hexokinase- II is abundanl in skelelal 
musc le and adipose tissue [8]. In tumor cells, hexokin-
ase-II is apparently Lhe main over-expressed isoe nzyme 
[7,8,37], excepl fo r brain lumors, in whieh hexoki nase-l 
is over-expressed [8]. 
Analysis of lhe hexokinase kinelic properlies and 
subcellula r redistr ibution lowards mitochondria sug-
gesled lhal lhe isoenzyme over-expressed in AS-30D 
ceJls was type JI , as previously suggested using a sim-
ilar analysis [7]. The amount of mitochondrial hexo-
kinase in AS-30D (50%) and HeLa cells (70%; dala 
not shown) was also similar to that observed for 
Novikoff and AS-30D aseiles tumor cells of 50-80% 
[55,56]. This observation explains, al leasl in par!, lhe 
enha nced glycolylic flux (Table 2, [8]) and lhe resisl-
ance to apoptosis [11] in these tumor cells. 
The kinetic a nalysis of PFK- I revealed lhal the iso-
enzyme present in AS-300 cells was completely insen-
sitive to lhe usual allosteric inhibitors, ATP and 
cit rate, in the presence of a low, physiological concen-
trat ion of F ru(2,6) P2, and that il was highly sensilive 
lo lhe aelivalors NH4 + , AM P a nd Fru(2,6) P,. The 
high, physiologieal concenlration of AMP in AS-30D 
did not suffice lo potenlly aClivate PFK-I in lhe pres-
ence of ATP a nd citrate . The expression of a PFK- 2 
isoenzyme with a low fructose-2,6-bisphosphatase 
activity in several human tumor lines has been des-
cribed [57,58], which ensures a high concentration of 
Fru(2,6) P,. I ndeed, Fru(2,6) P, was lhe mosl pO lenl 
aeLivalor of PF K- I and blocked lhe inhibition by ATP 
and cilrale (Fig. 2B and Table 6, and also [12]). 
Therefore, il s kinelic properlies prediel lhal PFK- I 
activity ca nnot impose a flu x limitation on glycolysis 
in AS-30D cells. Under nea r-physiologica l cond il ions, 
lhe esLimaled elasLieily coeffieienl of PFK- I fo r Fru6P 
was high (~tFK-1 Fru6P = 1.2), which pro vides the bio-
1984 FEBS Journal 273 (2006) 1975-1988 CI 2006 The Authors Journal compllat ion e 2006 FEBS 
 58 
A. Marin-Hernández el al. 
chemica l mechanism fo r its low fl ux control coefficient 
(Table 5). The PFK- I elastieity a nd kinelie properties 
also provide a biochem ical basis for understanding lhe 
diminished Pasteur elTeet observed in AS-30D [49] a nd 
other tumor cell types [39]. 
Experimental procedures 
Chemicals 
Hexoki nasc, G6PDH, hexoscphosphate isomerasc, aldolasc, 
a.-glycerophospha te dehydrogenase , l ri oscphosphale isom-
erJse, glycc raldehyde-3-phosphale dehydrogenase, pyruvate 
kinase and LDH were purchascd from Roche Co. (Mann-
he im, Gennany). Recombin ant enzymes enolase, pyruva te 
phosphate dikinase and phosphoglyccratc kinasc from 
Emamoeba hisrolyrica were kindly providcd by E Saavcdra , 
Instituto acional de Cardiología, México. Glucosc, G \c6P, 
Fru(1 ,6) P" glyceraldehyde 3-phosphate, 2-phosphoglyccrate, 
phosphocnolpyruva te, Fm(2,6) P2' pyruva te, citrate, A TP, 
AMP, ADP, GTP, dithiothre itol, cystei ne, NA DH, NA D+, 
NA DP and oxala tc wcrc from Sigma Chcmical (St Louis, 
MO, USA). Fm6P and 3-phosphoglyccra lc wcrc from Roche 
(Indianapolis, IN, USA). Methoxy[' H]inulin was from Per-
kio- Elmcr Lifc Sci (Boston, MA, USA). AIl ot her reagents 
were of analytical grade from commercial sources. 
Isolation 01 tumor and liver ce lis 
AS-30D hepatoma cc lls [(2-4) x lO' ccl ls·mL-' ) were propa-
gatcd in female Wistar ra ts (200 g) by iotrapcritoncal trans-
plantalion. Hcpatoma ce ll s werc isolated as describcd 
c1sewherc [59]. 
HeLa ce lls (1.5 x 104 cell s' mL- 1
) were grown in Dul -
bccco's minimal csscntial mcdium (Gibco Lifc Tcchnol-
ogies, Rock viJle, M D, USA), supplemented with 10% fetal 
bovine sc rum (G ibco), lO 000 U'm L - 1 slrep tomycin/ penicil-
li o, a nd fu ngizone (ampholericin B; Gibco) in 175-cm2 
fla sks (Coming, ew York, NY, USA) at 37 oC in 5% 
COy-95%Ol. I-Iepatocytes were isolated by perfusion of 
isola ted liver with collagenasc IV (Worthington, La kewood, 
NJ, USA) from fed Wistar rals [60]. The ccll ular viabilily 
assayed by lrypan blue exc! usion was 80-85%. Experimen-
tal manipulation of human and rodent cclls and animal spc-
cimens were carried out following the Institu to aciona l de 
Card io logía de Mexico gll idcli nes in accordance with the 
DecJaralíon of I-Ielsinki and Ihe US JH guidelines for ca re 
and use of experi mental animals. 
Determination 01 steady-state concentrations 
01 metabolites 
Cells (15 mg protein'm L- 1
) werc incubared in K.rebs- Ringer 
medium with orbital sha king (1 50 r.p .m.) at 37 oC in a 
Control of glyco lysis in tu mor cel ls 
plaslic fl ask. Afte r lO min, 5 mM glucosc was addcd to thc 
cc ll sll spension. The reac tion was stoppcd 3 min la ter with 
icc-cold perchlo ric ac id (3%, v/ v, fi nal concent ration). For 
the detennination of Ihe intracellular L-Iactale, an aliq llo t 
(1 mL) of thc ccll suspension was rapidly wilhdrawn and 
ccntri fugcd a L 20800 g fo r 10-15 s al room temperalure. 
The supernatan t was discarded, and the pellet rinscd with 
fresh K rc bs- Ringer medium a nd rhen resuspended in 3% 
perchloric acid. The samples \Vere neutralizcd with 3 M 
KOI-l / O.I M Tris. The concentrations of GIc6P , F m6P, 
Fru(I, 6)P" DHAP, glyccraldehyde 3-phosphate, phospho-
enol pyruvale, pyruvale, ATP, ADP, AMP, citrate and 
L-lactate were determincd by standa rd enzYlllutic assays [61J. 
Protein was dctermined by lhe biurct method using BSA 
as standard [62]. 
Determination 01 intracellular glucose 
AS-30D cc lls (15 mg'm L- 1
) \Vere preincubatcd for 10 min in 
the absence of exogenous substratcs at 37 oc. Exogenous glu-
cose was added after lO min, and the cclls were incllbatcd for 
3 min morc. After\Va rds, an aliq uot (0.5 mL) was carcfully 
pOllrcd inlo microcenl rifuge t ll beS containing, from boltom 
to top, 30% (v/ v) pcrchloric ae id (0.3 mL), I-bromodode-
cane (0.3 m L) and fresh K rebs- Ringer medium (0.3 mL). 
The salllp le \Vas centrifuged for 2- 3 min in a refrigcrated 
Microfuge (Eppcndorf centrifuge 5804 R, Eppendorf, 
Hamburg, Germany) a t 20 800 g. The botlom ¡ayer was col-
lecled and ncutra lized with 10% NaO H. Glucosc was dcter-
mined by enzymatic ana lysis [61J. 
To correcl fo r lhe presence of exogenolls glucose from lhe 
incubation mcdium in Ihe ccllu la r pe llel, the content of ex ter-
nal wa ter \Vas cval uatcd by lIsing mcthoxye H]in ulin. An 
aliquot of cells (5- 30 mg protein'IllL - J) was incubated 
in 0.5 mL K.rebs- Ringer medi llm with e H]inulin 
(0.15 mg·m L- J, specific rad ioac ti vity 5200 c.p.m: j.lg-J) for 
15 s. Thcrcaftcr, thc ccU suspcnsion \Vas carcfully laycrcd 
into microcentrifuge lubes prcpared as described above. The 
mdioactivily of the bottom layc r \Vas mcasurcd in a ¡iq uid 
sc intillation analyzer (Packard l nstrumenls/ Canberra, 
Meriden, er, USA). The concenlrarion of glucose carried 
from the exLrace ll ular mil ieu was 1.28 111M . This va lue was 
subtraclcd from thc original intraccll ular conccntralion. 
Determination 01 glycolytic Ilux in liver and 
tumor ce lis 
Cells (15 mg protein'mL - 1) wcrc incll ba tcd in 3 mL Krebs-
Ringer mediulll \V ith o rbital stirring (150 r.p. m.) al 37 oC in 
a plastic nask. Aftcr 10 min, 5 mM glucose was added to 
thc cells; lhe rcaction \Vas stoppcd with cold 3% perchlor ic 
acid O and 3 min later. The samples were neulralized with 
3 M KO H/ O.I M Tris. L- Lactate generJ tcd was detennined 
by enzymatic analysis [61]. 
FEBS Journa l 273 (2006) 1975-1988 ~ 2006 The Authors Journal compilation ~ 2006 FEBS 1985 
 59 
Contro l o f g ly co lysis in tumo r cel ls 
Cell extraets of tumor and liver eells 
CeUs (65 mg protein 'mL - 1) were resuspcndcd in 25 mM 
T ri sIH CI bulTer, pH 7.6, wilh I mM EDTA, 5 mM dilhio-
threitol a nd 1 mM phenylmethancsulfonyl nllo ride. The ceU 
suspcnsion was frozen in liquid nilrogen and th awed in a 
wa ter ba lh a l 37 oC; this procedure was repeatcd lhree 
times. Cell Iysa tes were centrifuged at 39000 g fo r 20 min 
and 4 oc. Afterwards, the supcrnata nl was coUccted for 
delenn ination of enzyme aClivily. Activities of hexokinase, 
hexosephosphate isomerase, PFK-I, triosephospha le iso-
merase, aldolase, glycc raldehyde-3-phosphale dehydroge-
nase, phosphoglycerate kinase, phosphoglycera te mutase, 
enoJase, pym va te kinase, LDH , o:-glyccrophosphate dehy-
drogenase, phosphoglucomulase and G6PDH were deter-
mined spcclropholOmelríca ll y by standard assays [63,64]. 
The ¡ncubalion buffer was 50 mM Mops, p H 7, at 37 oc. 
Phosphoglyce rale kinase aClivíty was delcrmí ncd in 50 mM 
pOlassium phosphate, pH 7, al 37 oc. To assay milochond-
ría l hexo kinase, isola ted mitochondria were preparcd as 
desc ri bcd elsewhere [59]. The assays for bOlh cyloso lic and 
milochondriaJ hexokinase isoenzymcs were carried oul a l 
37 oC in I mL KME burrer ( 100 mM KCI, 50 mM Mops, 
0. 5 mM EGTA , pH 7.0) plus 2 U G6PDH, I mM NADP +, 
5 mM MgCI" glueosc (rrom 0.05 10 10 mM) and I ~I M 
oJigomyci n. The reaction was sta rted by lhe addition of 
ATP (0.05- 5 mM) arter 2 min ineubation. 
Assay for hexokinase inhibition by G lc6 P was carried 
out in 3 mL of KME buffer a l 37 oC, in lhe presence of 
3 mM ATP, 5 mM MgCh, I ~l M o ligomycin and 0.2- 1 mM 
G Ic6 P. Owing 10 the hi gh unspccific oxidalion of added 
NAD H by lhe cytoso lic-enriched and milochondr ia l frac-
lions (despite lhe addition of rotenone), the ac tivity was 
calcula led from lhe ADP genera ted, which was delermined 
by a sta nda rd assay [61). The hexokinase acti vity was cor-
rected for lhe A DP formcd in lhe absenee of addcd glucose. 
BOllnd and free isoenzymes were prcincllbaled for 2 min, 
and then lhe hexokinase reaclíon was slarted by add ing 
3 mM glucose. Afler 30 s, lhe reaction was slOppcd wit h 
ice-cold pcrchlo ric acid (3 %, v/ v, fi nal concentration). The 
sam ples were neutra li zcd wit h 3 M KOH/ O.I M Tris. The 
conlro l activities for both hexokinase isoenzymes were com-
para ble 10 those determined by the G6P DI-I coupli ng assay. 
For PFK-I ac ti vi ly, freshly prepared exlraclS were incu-
bated al 37 oC in 50 mM Mops, pH 7. The reaction assay 
conlained 0. 15 mM NADH , 5 mM MgCh, 1 mM EDTA, 
Fru6P (from 0.0 1 to lO mM) and amm onium sulfa te suspcn-
sions of lhe following coupl ing enzyrnes: 0.36 U aldolase, 
9 U triosephosph ate isomerase, and 3.1 U (l- glyceralde-
hyde·3-phosphate dehydrogenase. The reaclion was started 
by lhe addi tion of exogenous A TP (from 0.01 lO 2 mM). For 
lhe deterrn inalion of Ko.s for NH 4 +, AMP and Fru(2,6) P2, 
as well as fo r lhe effecl of activalOrs a nd inhibilors, Iyophi-
lized (ammonium-free) coupling enzyrnes we rc used in KM E 
buffer. 
A Marín-Hernández et al. 
Determination of flux control eoeffieients 
The nux cont rol coefficients (CJEi) were detenn ined by 
using lhe elas ticity-based analysis [31,34,45,65,). This 
approach quantifies the sensitivity of a given enzyrne or 
block of enzymes 10 va ri ations in its subslra le or product 
when lhe sleady-s tate flux is modified. Glycolysis was stí-
mulated by exogenous glucose (4-6 mM) o r ínhibited by 
oxalale (0.5- 2 mM), 2-deoxyglueosc (0.5- 10 mM) or arsenile 
(5- 100 ~I M). Under lhese conditions, the va rialÍon in lhe 
glycolytic fl ux and in several in lermediales was detennined. 
Flux control coeff¡cien ts were eSlima lcd using lhe connec-
livily (Eqn 1) and summalion lheorems (Eq n 2) [24J, 
(1 ) 
cb +q , = 1 (2) 
in which J is Ihe glycolytic fl ux (mte of lactate formalion), 
El is lhe enzyme block lhal produces lhe inlermediate (m) 
and E2 is lhe enzyme block lhat consumes m. 
Acknowledgements 
This \Vo rk \Vas partia lly suppo rted by gra nl no. 438 11 -Q 
from CONACyT -México. 
References 
Warburg O (1 956) On rcspiratory impairment in cancer 
eells. Sciellce 124, 269- 270. 
2 Zu XL & Guppy M (2004) Caneer melabo lism: raelS, 
fa n tasy, a nd fiction. 8ioc/¡em Biop/¡ys Res CommllJl 313, 
459-465. 
3 Dang C & Semenza GL (1999) Oncogenic altera lions of 
metabolismo Trends 8ioc/¡em Sci 24, 68- 72. 
4 Lu H, Forbes RA & Verma A (2002) Hypoxia-indueible 
facto r 1 ac tiva tion by aerobic glycolysis implica tes lhe 
Warburg effecl in carcinogenesis. J Biol C/¡em 277, 
23 111 - 23 11 5. 
5 OSlhus RC, Shim H, Kim S, Li Q, Reddy R, Mukhctjee 
M, Xu Y, Wonsey D, Lec LA & Dang C (2000) Dercg-
ulalÍon of glucose lra nspon er I and glycolytic gene 
expression by e-Mye. J Bio! C/¡em 275, 2 1797- 2 1800. 
6 G ill ies RJ , Martinez-Zaguillan R, Martinez GM, Serr-
ano R & Perona R (1990) Tum ori genie 3T3 ee ll s main-
ta in an alka line intracc llular pH under phys iological 
eondilions. Proe Nat! Acad Sci USA 87, 741 741 8. 
7 akashima R, Paggi M, Seoll U & Pedersen PL (1988) 
Purifica ti on and characleriza tion of a binda ble form of 
mitochond riaJ bound hexoki nase from the highly glyco-
Iylic AS-30 D ra l hepaloma ce ll line. CaJlcer Res 48, 
91 3- 919. 
8 Pedersen PL, Malhupala S, Rcm pel A, G csehwind J F & 
Hee Ko Y (2002) Milochondrial bo und lypc [J 
1986 FEBS Joornal 273 (2006) 1975-1988 e 2006 The Authors Journal compllatlon e 2006 FEBS 
 60 
A. Marín-Hernandez et al. 
hexokinase: a key playe r in the growth and survival of 
many ca nccrs and a n ideal prospect for thcmpeutic 
intervent ion. Biocltim Biopltys A cta 1555, 14-20. 
9 Goel A, Ma thupala SP & Pedersen PL (2003) Glucose 
mctabolism in cancer. Evidencc that dcmethylation 
events playa role in activa ting type 11 hexokinase gene 
exprcssion. J Bio/ Chem 278, 15333- 15340. 
lO Ma thupala SP, Hcese C & Pedersen PL ( 1997) Glucose 
ca tabolism in cancer ce ll s. Thc type n hcxokinase prom 0 -
ter contains fllnctionall y active response cJements for the 
tumor suppressor p53. J Bio/ Chem 272, 22776- 22780. 
II Pas tor ino JG, Shulga N & Hoek JB (2002) Mit ochon-
dria l binding of hcxokinase n in hibits Bax-induced 
cytochrome e rclease a nd apoptosis. J Biol Cltem 277, 
76 !()- 7618. 
12 Depré C, Rider MH & Hue L ( 1998) Mechanisms or 
control of heart glycolysis. Eur J Biocltem 258, 277- 290. 
13 Staa l GEJ , Kalrr A, Hecsbeen EC, Van Veelen CWM & 
Rijksen G (1987) Subunit composilio n, rcgulatory prop-
ert ies, and phosphorylation of phosphofrllctokinase, 
from human gliomas. Cancer Res 47, 5047- 505 1. 
14 Mcldolesi MF, Macx:h ia V & Lacx:eti P (1976) Dirrer-
ences in phosphofruclOkinase rcgula tion in normal and 
tumor ra t thyroid cells. J Biol Cltem 251, 6244-6251. 
15 Oskam R, Rij ksen G, Staal GEJ & Vora S (1985) Isozy-
mic composition and regulatory properties of phospho-
fructokinasc from we lJ -dirTcrcntiated and anaplas tic 
medullary thyroid carcinomas of Ihe mI. Callcer Res 45, 
135- 142. 
16 Vora S, Halpcr IP & Knowles DM (1985) Alterations 
in the activity and isozymic profilc of human phospho-
fructokinase during malignant transformatio n in ViltO 
and in vitro: transformation-and progrcss ion-linkcd dis-
criminants of malignancy. Cancer Res 45, 2993- 3001. 
17 Sánchcz- Martínez C & Aragón 11 ( 1997) Analysis or 
phosphofruclOkinasc subunÍls and isozymcs in asciles 
tumor ce ll s and its original ti ssue, murine mammary 
gland. FEBS Leu 409, 8 90. 
18 Denís C, París H & Murat JC (1986) Hormonal contro l 
of fructose-2,6-bisphosphate concentration in the HT29 
human colon adcnocarcinoma ccll line. Biochem J 239, 
53 1- 536. 
19 Mojeda M, Bosca L & Hue L (1985) Errect or gl uta-
mi ne on fmclOsc-2,6-biphosphatc and on glucosc Illeta-
bolism in HeLa cells and in chick-embryo fibroblasts. 
Biochem J 232, 521 - 527. 
20 Loisca u AM, RoussC'du GG & Hue L (1985) Fructose-
2,6-bisphosphate and the contro l of glycolysis by gluco-
cort icoid s and by othcr agcnts in ra t hepalOma cell s. 
C(llIcer Res 45, 4263-4269. 
21 Nissler K, Petermann H, Wenz I & Brox D (1995) Fruc-
lOsc-2,6-bisphosphate mctabolisl11 in Ehrl ich asci tes 
tumour cell s. J Callcer Res Clill Ollcol 121 , 739- 745. 
22 Rapoport T, Heinr ich R, Jacobasch G & Rapoport S 
(1974) A li near steady-state trcatment of enzymatic 
Control of glycolysis in tumor cells 
chains. A malhematica1 modcl of glycolysis of human 
erytrocytes. EI/r J Biochem 42, 107- 120. 
23 Kashiwaya Y, Sato K, Tsuchiya , Thomas S, Fe ll 
DA, Vecch RL & Passonncau JV (1994) Control or glu-
cose utiliza tion in working perfused rat hearl. J Biol 
Chem 269, 25502- 255 14. 
24 Fell D (1997) Umlerswl1dil/g rhe COllfrol o/ M etabolism. 
Plenum Press, London. 
25 Wu R (1959) Regulatory mechanisms in carbohyd ratc 
metabolism. V. Limiting factors of glycolysis in HeLa 
cc ll s. J Bio/ Chem 234, 2806- 28 10. 
26 Hardíng JW, Pye rit z EA, Monis HP & Whíte HB 
(1975) Proportional ac tivities of glycerol ki nase and 
glycerol 3-phosphate dehydrogcnase in ral hepatomas. 
Biochem J 148, 545- 550. 
27 Pcderscn PL (1978) Tumor mitocho ndria and the bioener-
ge tics of cancer cells. Prog Exp Tumor Res 22, 190- 274. 
28 Traut TW (1994) Physiological concentrations of 
punnes and pyrimidines. M ol Cell Biochem 140, 1- 22. 
29 Moreno-Sánchez R & Torres-Márquez ME (199 1) Con-
tro l of oxidative phosphoryla tion in mi tochondria , cell s 
a nd tissues. 1m J Biochem 23, 11 63-1174. 
30 Groen AK , Wa nders RJ , Westerhofr )-1 V, van der Meer 
R & Tager J M (1982) Quan tifica tion of the contribu tion 
of vanous stcps 10 Ihe control of mitochond ria l respira-
tion. J Biol Chem 257, 2754-2757. 
31 Kacser H (1983) The control of enzymc sys tcms iJl vivo: 
elasticity analysis of lhe slcady sta te. 8 iochem Soc Trans 
11, 35-40. 
32 Moreno-Sánchez R, Bravo C & Westerhorr HV (1999) 
Determining and understa nding the control of nux. An 
ill ustration in submitochondrial pa rticles of how lo va li-
date schemes or metabolic con trol. EI/r J 8iochem 264 , 
427-433. 
33 G rocn AK , van Rocrm und CW, Vervoom Re & Tager 
JM (1986) Control of glllconeogencsis in rat liver ccl ls. 
F lux control coefficients of the enzymes in the gluconco-
gcnic palhway in the abscnce and prescnce of glucagon. 
Biochem J 237, 379- 389. 
34 Harncr RP, BrolVn GC & Brand MD (1990) Analysis or 
lhe control of respira tion ra te, phosphoryla tion ra te, 
pro ton leak rate and pro tonm otive force in isola ted 
mi tochond ria lI sing the ' top-down' approach of meta-
bolic contro l theory. E/lr J Biochem 188, 313- 319. 
35 Thomas S, Mooney PJ, Burrell MM & Fell DA (1997) 
Metabolic contr ol analysis of glycolysis in luber lissue 
of pota to (S olal1/1111 fIIberosl/m): explana tion fo r the 10\V 
control cocfficien t of phosphofructokinase over respira-
to ry flu x. Biochem J 322, 119- 127. 
36 Aínscow EK & Brand M D (1999) Top-dolVn control ana-
Iysis of ATP tu rnover, glycolysis and oxidative phospho-
ry lation in ra t hepatocytes. E/lr J Biochem 263, 671 - 685. 
37 Wilson JE (2003) Isozymcs of mammalian hexokinasc: 
struclure, subcellular localiza tion and mcta bol ic func-
tion. J Exp Bio/ 206, 2049- 2057. 
FEBS Journal 273 (2006) 1975-1988 e 2006 lhe Authors Journal compilatioo CI 2006 FE BS 1987 
 61 
Control of glycolysis in tumor cel ls 
38 Jannaschk K D, Burgos M, Centerlles JJ , Ovadi J & 
Cascante M (1999) Applica tio n o f metabolic contro l 
ana lysis 10 the sludy of toxic e ffects of coppcr in muscle 
glycolys is. FEBS Lett 445, 144-148. 
39 Eigenbrodl E, Fisle r P & Reinaeher M (1985) ew per-
spcctives in carbo hydratc metabolism in tumor cclls. 
Jn. Regulaúoll o/ Carboll)'drwe Metahoh<;m (Reitner, R, 
ed.), vo l. 2, pp 141 - 179. C RC Press, Boca Ralon, FL. 
40 Rodríguez-Enriq uez S & Mo reno-Sánchez R ( 1998) 
Intermediary metabo lism of fa st-growth tumor cell s. 
Arch Med Res 29, 1- 12. 
41 Mazurel S, Eigenbrodl E, Failing K & Sleinberg P 
( 1999) A ltcrat io ns in the glycolytic and glutamino lytic 
pathways aner mal ignant transformat ion of ra l li ver 
oval ee ll s. J Cell Physiol 18 1, 136-146. 
42 Mazurel S, Miehel A & Eigenbrod E (1997) Elfeel of 
eX lracc ll ular AM P o n cell prolifera tio n and mctabolism 
o f breast cancer ce ll lines with high and low glycolytic 
rales. J Biol Chem 272, 4941-4952. 
43 Shusler S, Kahn D & Westerholf HV (1993) Modular 
ana lysis of control of complex mctabolic pathways. 
Biophys Chem 48, 1- 17. 
44 Ainseow EK & Brand MD (1998) Control analysis of 
systcms wi lh rcact ion block s that 'cross-talk'. Biocllim 
Biophys Acta 1366, 284-290. 
45 Brown GC, Lakin-Tomas L & Brand MD (1990) Con-
trol of respiralion and ox idati ve phosphorylation in 
iso lated nll liver cclls. Ellr J Biocllem 192, 355-362. 
46 Kitagawa T, Tsurllhara Y, Hayashi M, Endo T & 
Stanbridge EJ ( 1995) A tumor-associated glycosylatio n 
change in the glllcosc lransporter G LUTI cont rolled by 
tumor supprcssor funct io n in human ce ll hybrids. J Cel! 
Sci 108, 3735- 3743. 
47 Maeheda M L, Rogers S & Besl JD (2005) Molecula r 
and ccllular regulation o f glucosc transpo rter (G LUT) 
proteins in cancer. J Cel! PII)'sioI 202, 654-662. 
48 Colville C, Seauer MJ , Jess TJ, Gould GW & Thomus 
HM ( 1993) Kinetie analysis of lhe live r-lype (G LUT2) 
and brain-Iype (G LUT3) glucosc Iransporters in 
X enoplls oocyles: sllbslrale spccifici lies and c fTec ls of 
transport inhibitors. Biocltem J 290, 70 1- 706. 
49 Rodrígucz-Enriq ucz S, Jm1rez O, Rodrigucz-Zava la JS 
& Moreno-Sánehez R (2001) Mult isile control of lhe 
Crablrcc efTccl in ascites hcpatoma ccl ls. Eur J Biocltem 
268,2512- 2519. 
50 HoUSlek J, Cannon B & Lindberg O (1975) Glyeerol-3-
phosphatc shuule and its funlion in inlcmledi ary mcta-
bolism of hamslcr brow-adiposc tissue. El/r J Biocltem 
54, 11 - 18. 
51 Hajra AK (1997) Di hydroxyaeelone phosphale aeyl-
transferasc. Biochim Biopltys Acta 1348, 27- 34. 
52 Mendoza-Cozall DG & Moreno-Sanehez R (2006) Con-
trol of glutat hione and phylochelatin synthesis llnder 
cadmi um stress. Pathway modeling for plants. J Tlteor 
Biol 238, 919- 936. 
A. Marín-Hernández et al. 
53 Bustamanle E & Pedersen PL (1977) High ae robie gly-
colysis of ral hepatoma ce ll s in cult ure: role o f mito-
chondrial hexo kinase. Proc Natl Aca(/ Sci USA 74, 
3735-3739. 
54 Smith TAD (2000) Mammal ian hexokinases and Ihe ir 
abnormal cxpression in canceL Br J Biomell Sci 57, 
17()-178 . 
55 Arora KK & Pedersen PL (1988) Funeli onal signifi-
cance o f mitochondri al bo und hexokinasc in tumo r ce ll 
melabolism. J Biol Chem 263, 17422- 17428. 
56 Parry DM & Pedersen PL (1983) Intraeellu lar localiza-
tion and propcnies of particulate hexokinase in the 
oviko fT asci les tumor. J Biol Chem 258, 10904-1091 2. 
57 Alsumi T, Chesney J, MelZ C, Lcng L, Donnell y S, 
Makila Z, Milehell R & Bueala R (2002) High 
exprcssion of indllc ible 6-phospho fmcto-2-k inase/ fruc-
losc-2,6-bisphosphalase (iPFK-2; PFKFB3) in human 
cancers. Callcer Res 62, 588 1- 5887. 
58 Chesney J, Milehell R, Benigni F, Baeher M, Spiegel 
L, AI-Abed Y, Han JH , Melz C & Bueala R (1999) An 
ind ucible gene producl for 6-phosphofruclo-2-kinasc 
wi lh an AU-rich instabili ly element: role in tumor cell 
glyco lysis and the Warbllrg effccl. Proc Nall Acml Sci 
USA 96, 3047- 3052. 
59 Lópcz-G ómez F & To rrcs-Márqllez & Morcno-Sánchez 
R (1993) Cont rol of ox idative phosphorylation in 
AS-30 D hepaloma mitochondria. /111 J Biochem 25, 
373- 377. 
60 Bcrry M & Friend DS (1969) High-yic\d prcparalion 
of iso laled ral liver parcnchymal cells. A biochemical 
and fine slruelu",1 sludy. J Cell Bio143, 506- 520. 
6 1 Bergmcyer HU (ed .) ( 1974) Mef//Od<; o/ Enz)'matic AlIa-
I)'sis. Verlag Chemie, Weinheim. 
62 Gorna ll AG, Bardwill CJ & David MM (1949) Delermi-
natio n o f scmm pro leins by mcans o f biurct reaclion. 
J Biol Chem 177, 751 - 766. 
63 Bergmeyer HU (ed .) (1983) Methods oJ EI/z)'matic AI/a-
I)'sis . Verlag Chemie, Weinheim. 
64 Saavedra E, Encalada R, Pineda E, Jasso-Chávcz R & 
Morcno-Sánchcz R (2005) Gl ycolysis in Entamoeba '1/:<;-
1Olytica. Biochemica l characlcrizat ion of recombinant 
glyco lytic cnzymes and flux contro l analysis. FEBS J 
272, 1767- 1783. 
65 Tager JM , Wandcrs JA, Grocn AK, Kunz W, Bohnen-
saek R, Küsler U, Lclko G, B6hme G, Duszynski J & 
Wojlczak L (1983) Control of mi lochondri al rcspirat ion. 
FEBS Lett 151, 1- 9. 
66 Simpson RJ, Brindle KM, Brown FF, Campbell ID & 
Foxall DL ( 1982) SI lid ies of lae la te dehydrogenasc in 
lhe purified slate and in inlact ery lhrocyles. Biochem J 
202,58 1- 587. 
67 Meunier JC & Dalziel K (1978) Kinelies sludies of gly-
ceraldehyde-3-phosphate dchydrogenasc from rabbit 
muscle. El/r J Biochem 82 , 483-492. 
1988 FEBS Journal 273 (2006) 1975-1988 e 2006 The Authors Journal comptlation e 2006 FEBS 
 62 
 
FEBS Journal 
27 Jan 06 
Reference no.: FJ-05-1056.R1 
 
Title: Determining and understanding the control of glycolysis in fast-growth tumor cells. Flux control by an over-
expressed but strongly product-inhibited hexokinase. Authors: 1) Rafael Moreno-Sánchez 2) Alvaro Marín-
Hernández 3) Sara Rodríguez-Enríquez 4) Paola Vital-González 5) Fanny Flores-Rodríguez 6) Marina Macías-Silva 
7) Marcela Sosa-Garrocho. 
Editor: Hans Westerhoff 
 
Dear Dr. Moreno-Sánchez 
Thank you for submitting your paper for publication in FEBS Journal. 
 
I am pleased to inform you that your paper has received favourable comments from the referees and is likely to be 
acceptable for publication in the Journal after incorporation of their comments, copies of which are appended below. 
Please note in particular the comments from the Editor. 
 
Editor's comments: 
I think the following objection by one of the reviewers: "I am concerned upon the likely possibility that the 
consuming and producing blocks interact via metabolites other than those measured. The existence of such 
interactions would invalidate the approach used in this manuscript. Particularly the concentrations of ATP, ADP and 
AMP are likely to change during the titration experiments and are likely to affect the rate of enzymes on both, 
consuming and producing blocks. "is fully justified. I can also see that it may be hard to deal with this. However, you 
should clearly mention this as a limitation to your approach and come with suggested future work. You should also 
remove some of the speculations. 
 
Please note that no guarantee of acceptance of the revised version can be given in advance. If accepted, the dates of 
receipt of both the original and revised papers will be printed. 
 
We look forward to receiving the revised version of your manuscript. 
 
Yours sincerely 
Dr Vanessa Wilkinson 
FEBS Journal 
98 Regent St 
Cambridge CB2 1DP, UK 
Email: febsj@camfebs.co.uk 
 
REFEREES' COMMENTS 
 
Referee 1 Comments: None 
 
Referee 2 Comments: 
 General comments: 
 
This manuscript describes a series of experiments aimed at describing the flux control distribution in healthy and 
tumoral hepatocytes. This is achieved through elasticity-based analysis, in which the control of producing and 
consuming blocks may be calculated from the (measurable) elasticities of consuming and producing blocks to their 
common intermediate. 
 
Although the manuscript has clearly improved, there are a number of speculations for which the authors offer no 
explanation and that hinder the comprehension of the manuscript. Examples of 
such speculations are noted in the "particular comments". 
 
An issue which concerns this reviewer is the likely possibility that the consuming and producing blocks interact via 
 63 
metabolites other than those measured. Particularly: ATP, ADP and AMP. If 
the concentrations of these metabolites change during the titration experiments, the elasticity based analysis will be 
flawed and the control distribution reported, erroneous. 
 
Particular comments: 
Page 8: The authors wrote: ". F1,6-BP increased 250 times and DHAP 16.6 times, which suggested a strong 
activation of PFK-1 and an attenuated utilization of DHAP for glycerol and tracylglyceride synthesis." The rationale 
behind this expectation is not explained. DHAP concentration depends also upon the glycolytic pathway downstream 
of glyceraldehydes-3-phosphate. 
 
Page 10: The authors wrote: "However, this last type of metabolites may be still used in elasticity-based analysis as 
long as the flux through other pathways is low or negligible". 
The question that follows immediately is whether these pathways indeed have a negligible effect on those 
metabolites. References of experiments should be given to support applicability of the 
elasticity-based analysis. 
 
Page 10: The authors, rightly, point out that the rate dependence upon a substrate (or product) is likely to be 
hyperbolic for most glycolytic enzymes. However, in these experiments both substrate(s) and product(s) may change 
simultaneously and the rate relation of the flux and, say, the substrate needs not to be hyperbolic if the concentration 
of the product is also varying during the titration. 
 
Page 10: The authors wrote: ". only one point determines the value of the tangent.", when referring hyperbolic fits to 
the data. This is not true. The tangent's slope at a given operational point will depend upon the derivative of the 
function evaluated in that point, and would therefore depend upon the points on which the fit was based (i.e. all). 
This reviewer doubts that tangent slopes are an appropriate method for elasticity calculations when so few data 
points are available. 
 
Page 15, second paragraph: The authors wrote that increased concentrations of G6P and F1-6BP suggest that HK and 
PFK-1 may exert a lower flux control in AS-30D cells than in hepatocytes. 
The rationale for this suggestion is not explained. 
 64 
         20 February, 2006 
 
 
Subject: Manuscript FJ-05-1056.R1 
 
Dr. Vanessa Wilkinson 
FEBS Journal 
(European Journal of Biochemistry) 
 
Dear Dr. Wilkinson: 
 
 Thank you for your kind letter of January 27, 2006 concerning our manuscript entitled 
“Determining and understanding the control of glycolysis in fast-growth tumor cells. Flux control by an 
over-expressed but strongly product-inhibited hexokinase”.  We were informed that the manuscript may 
become acceptable for publication after incorporation of the comments of the editor and referees.  
Therefore, the text was modified accordingly, in regard to the potential drawback in the elasticity analysis 
and to some speculations made in the original manuscript.  New experimental data were also included in 
the revised manuscript.  Corrections in the manuscript were highlighted in bold type, as requested. 
 
The changes made to the manuscript are described below.  We hope that the manuscript may be 
now found acceptable in its present form for its publication in FEBS Journal (European Journal of 
Biochemistry). 
 
     Sincerely, 
 
     Rafael Moreno-Sánchez, Ph. D. 
 
Response to Referee No. 1 
 
Referee 1 did not indicate any comment. 
 
Response to Referee No. 2 
 
We thank this reviewer for the many helpful and constructive criticisms of our manuscript.  
 
General comments 
Certainly the glycolytic consuming and producing blocks interact via metabolites other than those 
measured, particularly the adenine nucleotides. So, we fully agree with the reviewer that our elasticity 
analysis may be flawed if the concentrations of ATP, ADP and AMP also vary in the titration experiments.  
Therefore, to avoid further speculation in justifying our results, it was decided to determine the variation in 
the concentrations of the adenine nucleotides under the same conditions used in the titration experiments.  
This new experimentation was described, and the text modified, on p. 16, last paragraph, in which the 
reviewer’s argument was incorporated.  The concentrations of the adenine nucleotides (Table 3 for ATP 
and ADP; p. 14, line 8 for AMP) now include the values of the new results.  A figure with the new results is 
enclosed in this letter. 
 
Particular comments 
 
The statements regarding the original speculations on p. 8 (“…which suggested a strong activation of 
PFK-1 and an attenuated utilization of DHAP for glycerol and tracylglyceride synthesis”) and p. 15 (“These 
results suggested that HK and PFK-1 may exert a lower flux control in AS-30D cells than in hepatocytes”) 
were removed, as recommended. 
 
The statement on p. 9, last two lines, p. 10, first two lines (“…this last type of metabolites may be still used 
in elasticity-based analysis as long as the flux through…”) was slightly modified and some references 
added. 
 65 
 
The argument about the hyperbolic rate dependence upon substrate or product (p. 10, lines 17-19) was 
modified to include the observation made by the reviewer. 
 
The statement “…only one point determines the value of the tangent” was erased and the text modified 
accordingly on p. 10, lines 19-22.  
 
The statement on p. 8, lines 16-17 was modified because ATP/ADP rate change with the news values of 
ATP.   
0
25
50
75
100
%
 A
T
P
0
25
50
75
100
125
%
 A
D
P
0
25
50
75
100
%
 A
M
P
Glucosa (mM)  4. 5      5.0       5.5      5.0      5.0
Oxalato  (mM)     _          _          _         1.0     2.0
 
Figure 1. Concentrations of the adenine nucleotides under the same conditions used in the titration 
experiments. ATP 100 % was 7 ± 1  mM,  n= 4; ADP 100 % was 2.1 ± 0.7 mM, n= 4; AMP 100 % was 3.3 
± 1.4 mM, n= 4. 
 
 
 
 66 
5.1.1 Datos no mostrados  
Para poder aplicar el análisis de elasticidades fue necesario establecer la 
condición de estado estacionario, la cual se define como aquella en la que el flujo de la 
vía y las  concentraciones de intermediarios se mantienen constantes con respecto al 
tiempo. En la figura 5.2.1, se observa que el flujo (medido como la cantidad de lactato 
producido/ min) y la concentración de F1,6BP fueron constantes entre 2 a 10 minutos, 
después de haber agregado la glucosa. A partir de este resultado consideramos realizar 
nuestros experimentos incubando con glucosa por 3 minutos. 
 
 
 
 
  
 
 
 
 
Figura 5.2.1. Establecimiento de la condición de estado estacionario en la glucólisis de las 
células AS-30D. Flujo (nmoles de lactato/min x mg proteína) y F1,6BP (nmoles/mg proteína). 
(n=2) 
 El siguiente paso fue determinar que los inhibidores empleados para reducir el 
flujo glucolítico (oxalato y arsenito) al inhibir enzimas de la parte baja de la glucólisis 
(GAPDH y LDH) no afectaran la actividad de alguna de las enzimas de la parte alta (HK 
y PFK-1). Se analizaron sobre la actividad de la HK y de la PFK-1 concentraciones de 
oxalato y de arsenito de 10 mM y 1 mM (5 veces las concentraciones empleadas en las 
0 2 4 6 8 10
0
5
10
15
20
F
lu
jo
 (
n
m
o
l/
m
in
* 
m
g
) 
Tiempo (min)
0 2 4 6 8 10
0
5
10
15
20
F
1
,6
B
P
 (
n
m
o
l/
 m
g
)
 67 
titulaciones 2 y 0.2 mM, respectivamente), las cuales no modificaron significativamente 
su actividad. En cambio, la LDH disminuyó su actividad en el intervalo de 
concentraciones de oxalato empleadas en los experimentos (0.5-2 mM) (Figura 5.2.2). 
El efecto del arsenito sobre la actividad de la GAPDH no se determinó. 
 
 
 
 
 
 
 
 
 
Figura 5.2.2. Efecto del arsenito y del oxalato sobre la actividad de la HK, PFK-1. (n=1) 
 
Finalmente, debido a que la HK ejercía un control importante sobre la glucólisis 
tumoral, se consideró caracterizar a la enzima. Para ello se determinó su distribución 
entre el citosol y la mitocondria en células AS-30D (Tabla 5.2.1). Los resultados 
mostraron que el 50% de la HK estaba en el citosol y el resto estaba unido a la 
mitocondria, de manera similar a lo observado en otras líneas tumorales (Parry y 
Pedersen, 1983; Arora and Pedersen, 1988). En HeLa la HK se encontró en mayor 
proporción en la mitocondria (70%).  Debido a que fue difícil obtener mitocondrias de 
estas células, se estimó primero la cantidad total de enzima en las células completas y 
luego en extractos celulares se determinó nuevamente la actividad en las fracciones 
HK PFK-I LDH
0
20
40
60
80
100
 
O
x
a
la
to
 2
 m
M
O
x
a
la
to
 0
.5
 m
M
O
x
a
la
to
 1
0
 m
M
O
x
a
la
to
 1
0
 m
M
A
rs
e
n
it
o
 1
 m
M
%
 d
e
 A
c
ti
v
id
a
d
A
rs
e
n
it
o
 1
 m
M
 68 
citosólica (citosol en la tabla 5.2.1) y mitocondrial (mitocondria en la tabla 5.2.1). Pero 
nos faltó medir la actividad de una enzima marcadora de la fracción mitocondrial como 
lo habíamos hecho para AS-30D.  
 
Tabla 5.2.1 Distribución de la HK en las células AS-30D y HeLa.  
 % de actividad 
Enzima Citosol Mitocondria 
 AS-30D  
HK 52 ± 9 (3) 48 ± 9 (3) 
HPI 88 (2) 12 (2) 
Citocromo C oxidasa 15.5 (2) 84.5 (2) 
 HeLa  
HK 31.7 (2) 67.8 (2) 
HPI 92.8 (2) 7.3 (3) 
 
Con base en nuestros resultados sobre el análisis de elasticidades se diseñó un 
problema que fue publicado en la Revista de Educación Bioquímica en la sección 
Problema Bioquímico (ver apéndice). 
 
5.1.2 Comentarios sobre el manuscrito No. 2 
En este manuscrito, por medio del análisis de elasticidades, logramos establecer 
que el bloque productor de G6P (GLUT + HK) es el que mantenía la mayor parte del 
control sobre la glucólisis. Además, también determinamos que la PFK-1 no ejercía un 
control significativo. Lo anterior nos llevó a caracterizar cinéticamente a la HK y a la 
PFK-1 para encontrar una explicación mecanística de estos resultados. La 
 69 
caracterización reveló que la HK mantenía un control importante porque 
independientemente de su localización (citosol y membrana externa mitocondrial) no se 
modificaba la afinidad por sus sustratos (glucosa y ATP) ni tampoco su sensibilidad al 
efecto inhibitorio de la G6P: aunque había una sobre-expresión de la HK en células 
tumorales, su actividad incrementada producía más producto y por tanto más inhibición. 
Por otra parte, se demostró que la PFK-1 no controlaba porque la presencia de la 
F2,6BP y el AMP impedían que la enzima fuera inhibida por el citrato y el ATP. Estos 
resultados nos permitieron concluir que el GLUT y la HK son los principales puntos de 
control de la glucólisis tumoral y que al no ejercer un control significativo la PFK-1 se 
explicaba la ausencia del efecto Pasteur en las células tumorales. 
 Cabe señalar que no realizamos una caracterización cinética del GLUT en este 
estudio, sin embargo posteriormente en nuestro laboratorio se logró determinar que la 
isoforma que predomina en las células AS-30D del transportador de glucosa es el 
GLUT3 que tiene una alta afinidad por glucosa (Km = 0.5 mM) (Rodríguez-Enríquez et 
al., 2009), pero que con respecto al resto de las enzimas, es el transportador el que 
tiene  la menor velocidad. Este resultado apoyaba nuestra propuesta de que el  
transportador ejercía un control significativo sobre la glucólisis.  
A partir de la fecha de la publicación de nuestros resultados hasta estos 
momentos no se encuentran documentados estudios sobre el control de la glucólisis en 
células tumorales o estudios similares al nuestro aplicando el análisis de elasticidades u 
otra estrategia experimental derivada del análisis de control metabólico en otra vía 
metabólica. Esto puede deberse a lo complejo que puede resultar el aplicar el análisis 
de elasticidades, dado que los metabolitos que deben emplear son aquellos que se 
pueden considerar como metabolitos “esclavos”, es decir aquellos que únicamente son 
 70 
producidos o consumidos en la vía de estudio o que su consumo por otras vías es 
limitado. Con ello se facilita el agrupar a las enzimas que forman la vía en dos bloques, 
el bloque productor y el bloque consumidor del metabolito de interés.  Por ende, 
metabolitos como los adenin-nucleótidos (ATP, ADP y AMP) que son productos o 
sustratos de diversas vías metabólicas (metabolitos promiscuos) no pueden emplearse 
para este tipo de estudios.  
Por otra parte, aunque podría considerarse ventajoso el empleo de inhibidores 
poco específicos para modular la actividad de los bloques, debe considerarse que 
únicamente deben observarse cambios en las concentraciones de los metabolitos 
esclavos, sin afectarse drásticamente las concentraciones de los metabolitos 
promiscuos. 
 
 
 71 
5.2 Construcción de modelos cinéticos de la glucólisis tumoral 
Debido a que por medio del análisis de elasticidades no era posible determinar el 
control que ejercía individualmente la HK y el GLUT en AS-30D por las dificultades 
técnicas para medir la glucosa libre intracelular y sobre todo porque para extender 
nuestro estudio a células tumorales humanas era necesario disponer de una gran 
cantidad de material biológico, se decidió construir  en el programa “Gepasi” un modelo 
cinético de la glucólisis de las células AS-30D, considerando las ecuaciones cinéticas 
correspondientes para cada enzima. Para la validación experimental del modelo 
cinético, se determinaron en paralelo las concentraciones de los metabolitos y los flujos 
de la vía, además de comparar los valores de la distribución de control que el modelo 
predecía con aquellos obtenidos previamente por medio del análisis de elasticidades 
(Manuscrito No. 2 de esta tesis). Para construir el modelo fue necesario determinar los 
parámetros cinéticos (velocidades máximas “forward”, Vmf; velocidades máximas 
“reverse”, Vmr; afinidad por sustratos, Kms; afinidad por productos, Kmp; constantes de 
activación, Ka; constantes de inhibición Ki) de cada una de las enzimas, la velocidad de 
la glucólisis, los flujos de algunas ramificaciones (síntesis-degradación de glucógeno y 
la vía de las pentosas fosfato) y las concentraciones en estado estacionario de todos 
los metabolitos de la vía . Una vez validado este modelo, se uso como base para la 
construcción de un modelo de la glucólisis en células tumorales humanas (HeLa, cáncer 
cervico-uterino). El manuscrito con estos resultados fue aceptado para su publicación 
en la Revista Biochimica et Biophysica Acta-Bioenergetics. 
 
 72 
  
 
Factor de impacto 2009: 3.68 
!JI cancer 
1~ 
1. Introduction 
\2010; 
AB TRAer 
!V1ost cancer ceEs exhibit an acce}erated glymlysis rati' compared fhxmal celbL This met2bolic change 1:; 2~ 
assod¿ted wíth the uvcH:xpycsaio!1 of al! the and transporte::-s (as mdw.l:d by Hit· la and 
:~~r;b~~~~~~~~~~:D;~:, ;¡~sl 
underJymg contro~ mecndnisms, kinetlc 
Heta (eH" were c011strudéd with experimental lleré for e¿ch path'Ndy step :éJlZyItle kiBeth: 
Sü:adjl<H¿fé ffiétábo;ifé Cl}lléént;ii!:ir;(¡s &id flU;<l~$), TlR: mód0>; p",di<:lc,d 
flUXC5 2nd living cancer ce!]::; Ph'fSl01ogicaJ and 
I ~'fc~I~::;;:~::C()ndit¡OIH pre':\lmng du!",ng \lImos progression, 
A degradation:: GLUT> HK in He~2 wcre 
the rn¡¡in fllJ.x~ and A11' coticentr:ttiOCl:~m:Jtro1 steps, ModclJng 31so rcv51!ed th2t, i:J diminis.'1: the 
glyc 01 yuc flux ot thé ,lI¡Th"óOCl:c;1t (¿rlon by SU>;, i{ W J$ féqulíed {ú déaédSc G!lJ1 ú r HK or H 1'1 by 7b~t ~ AS, 3IJD ~, 
and dcgrildaüon by 37"Sg:;; (HeLa;. decrNsing mélltionéd SlC¡:tS by 
4;7%, TIlus, procein<; ¿fe propoS0d tú be fOff'mtlst therapeutk: targeu the1r :;¡mu]taneOtH 
Inhl0¡Holl wili llave antdgO¡)l$l!C effecl$ {JO tumor eúe-rgy mr::t;¡boh$J11 than inh¡b¡tion oí .lit othe-:' 
gl¡'coll}'llc, Don,cc>n!lRJillnll' ern:y:m::L This: article js part oi a Spccia! lssue cnYitied Biflcner¡;c'lrs nf ,'n',rrr 
:¿(nO LJsevler JlV, 
------------------------------------------------------------'* 
tú :normal noo-prolirerating ce[ls. However, lreir severe side effet1:S in 
[ ""l"enl I.c k o~:,~~~O\~~'':~:; I;(~d'~' E~,t::O: ;i,n here fl t () r 
té$istance) make neces,sary ihe ór novel 
therapeutic strategies tú fight mis Hence, tf1é ldéntifi-
catiOIl of the main differences at the molecular 2Ild f'unctioIl21 leve!s ,¡ 
between normal and lumor ('e115 15 essentia! in th:e search túr ¡¡~V 
,"efa peu!''' Ia;gets, 
tile mÓ$t fflétaholk álleratión iIl tné 
majo;,ity tumor ('el!s, 1s a gh¡cosc consumptiol1 (with a 
lTI thi" 
c~':o~n~':'¡O~~O~1\;¡:lraJn;t~J:~~r:~l;;a~C'I~.~I~e~i:;~~~f~:~~~ir;~~'r7:it'r to norm:tl cells ~; in 'Thé t;¡e ",7 
'Y'llh'eslS n<;"cessary tu tne ,;H:tlve po,liley¡,¡km 
MOf{'OVér, glyeóly51$ tan thé elíérgy~générat¡ng 
palt"Yfay llTlder (of1(:htlons where mitoc;,ondrial functton absent 
{)[ dimlIllshed sucl1 as in {he inltia! non-vascular stages Qr tumor 
prugresslolt whére hypoxk t:úndilions pre\taB. 
Toe cenlltar mecnamsms inVO¡VVle~d~=iol:;~tfc:e~~':l; 
tne simultarteous increased tr 
g~~~~n:~~~~~~¡ and ttansporters In prucesses 
h t;'1(,1,fJf In (HIFla) ano other OI1"og""'" 
rile higher toupled w lranslátimi aré 
respo:nsllole fUf the jncreased (ontent, and actw:ities 
 73 
ARTICLE IN PRE55 
2 A Marin-Hernánckz {1 al ¡ Biochimica el Biophysica Acta 10(}( (2010) XJCJC-IOO( 
70 enzyrnes and transporters which result in an increased pathway flux. 
71 HIF-lo: induces theexpression ofsperific glycolytic protein isofol111s (Le .• 
72 GWT1. GWTI. HKI and 11. PfK-l. ALOO-A and C. PGK1. EN<J:a, PYK-M2. 
73 LDH-A and PFKFB-3) which are nor usually found together in normal 
74 cells. Sorne of these ¡soforms show lower sensitivity to physiological 
75 inhibitors and lower affinity for products, thus favoring a higher forward 
76 (glycolytic) Tate compared to normal cells [reviewed in S). 
n Because of its relevance on tumor metabolism, inhibirían of 
78 glycolytic flux has been considered an alternative strategy for cancer 
79 treatment [reviewed in 6 ). However, since both tumor and non-tumor 
80 cells cantain a similar ser of glycolytic enzymes, it is mandatory to find 
81 striking differences in pathway control so that the tumor enzyme(s) 
82 with the highest control be different to those in normal ceUs. Perhaps 
&1 sorne of the side-t'ffects induced by anticancer glycolytic drugs have 
84 derived from theír lack of specificity, targeting both cancer and non-
85 cancer cells. 
86 Metabolic Control Analysis (MCA) has shown that the control of a 
87 pathway flux or a metabolite concentration is dis tributed in different 
88 degrees among aU the pathway enzymes. thus drcumventing the 
89 mísleadíng and qua lítative concept of the U rate-Iimiting step" and 
90 moving into a quantitative analysis of the control of metabolíc 
91 pathways (7 - 9). MCA allows for the quantitatíve determinatíon ofthe 
92 degree of control thar each enzyme has over the pathway flux and rhe 
93 intermediary concentrations, namely, the flux control coefficient (Gil 
9.t and the concentration control coefficient ( C~n, respectively, wherej is 
95 flux, M is a metabolite concentration and Ei is a pathway enzyme. A 
96 practical definirion ofthese coefficients is the percentage of change in 
97 flux or merabolite concentrarion when a 1% change in a pathway 
98 enzyme acrivity is attained. 
99 By applying different experimental approaches such as elasticity 
100 analysís and metabolic modeling. the control distribution of a 
101 metabolic pathway can be determined P,8) . The advantages of the 
102 elasticity analysi s strategy are thatit allows 1..etelTIlining of the control 
103 distribution in living ceUs, and that it is not required to know the 
104 complete set of the kinetic parameters of the pathway enzymes. In a 
105 previous work. elasticity analysi s was applíed to glycolysis in rcxlent 
106 AS-30D hepatoma 110~ The results indicated that the pathway was 
107 mainly controlled by the glucose transponer (Gurr ) plus hexokinase 
108 ( HK ) segment (71 %; Le., Cb.ur+ HK = 0.71 ), with lesser control 
109 attained by phosphofructokínase 1 (PFK-l ) (clPFK_1 = 0.06). The rest 
110 of the control ( C ~= 0.25 ) resided within the aldola.se (ALDO ) to 
111 lactatedehydrogenase ( LDH ) segment [l O). 
112 As elastícity analysis relies on quantifiable smaU changes in 
113 pathway intermediary concentrations, thus dividing the analyzed 
114 pathway in segments. thi s approach determines control coefficíents 
115 only by group of enzy mes. Hence. in our previous study it was not 
116 possible to define in the glucose-6-phosphate prcxlucing segment 
117 whích individual step, GLLIT, HK or even glycogen degradation, was 
118 exe rting the main flux control. 
119 Kinetic mcxleling of a metabolic pathway integrates all available 
120 information (detailed kinetic propenies of the individual enzymes and 
121 transJX>ners, metabolite concentrations and pathway fluxes) into a 
122 functional, cross-talking network that anempts to simulate whatoccurs 
123 in the intracellular milieu (reviewed in 8,11 ). Thus, the aim of kinetic 
124 modeling is to construct a model able to predict the fluxes and 
125 concentrations found in a cell under different specific metabo lic steady-
12(i states. Validated mcxlels allow for the identification ofthe mechanisms 
127 by which a metabolic pathway is controlled and also provide the 
128 individual di and C~ foreach pathway step. Byextending this approach 
129 to normal (host, nOlTIlal cells ) and pathologic systems ( tumor cells, 
130 parasites ) the enzymes and transporters with the highest control in the 
131 latter and low control in the former can be identified and hence 
132 proposed as the best and most adequate therapeutic targets [8,12). 
133 In the present work, we report on the kinetic mcxleling of 
13-1 glycolysis in rat AS-30D hepatoma and in human HeLa tumor cells 
135 under normoxic and physiological glucose concentration condirions 
cha racterí stics of in vitro culture experimentatíon and under hypoxic 136 
and glucose deprivation conditions prevailing during tumor forma- 137 
tion. The res ults indicated that in the evaluated conditions, the flux 138 
control was shared mainly among HK ~ HP I > GurrinAS-30Dcells and 139 
glycogen degradation ;;::; GLLIT> HK in HeLa cells. Funher modeling 140 
analysis showed that s imultaneous inhibition of these controlling 141 
steps has greater negative effects on tumor glyco lysis than inhibirion 142 
of non-controlling reactions. 143 
2. Experimental procedures 1 ... 
2. 1. Enzymes and chemicals 145 
HK. G6PDH. HPI. ALOO. GAPDH. aGPDH. aGPDH(fPI. PYK/LDH. 146 
LDH, G1 !lt G6P and F6P were purchased from Roche (Manheim, 147 
Germany ). Recombinant enzymes ENO, PPi-PFK and PGK from 148 
Entamoeba hiswlytica were those prevíously described (13). Glucose 149 
(Glu ). 6PG. FBP. G3P. 2PG. 3PG. PEPo Pyr. Rib5P. Ery4P. Xy5P. A11'. ADP. 150 
GTP. DlT. cysteine. NADH, NAD+, NADP+, 6PGDH. amyloglucosidase, 151 
11<, TA. MgCh and Iyophilized ALDO, aGPDH andTPI were from Sigma 152 
(St Louís, MO. USA). DHAPwas from Fluka (Buchs, Switzerland). Mops 153 
and Hepes were from Research Organics (Cleveland, Ohio, USA ). 15-1 
Potato PPi-PFK was purified as previously described [14). 155 
2.2. Enzyme activities 156 
Oarified cytosolicextracts from AS-30D and HeLa cells were prepared 157 
as previously described [lO). Glycolytic enzyme activities and affinity 158 
constants were determined for the forward (glycolytic) and reverse 159 
(gluconeogenic) reactions in 50 mM Mops buffer pH 7.0 (assay buffer) at 160 
37 oC by following the NAD(P)+ reduction or NAD(P)H oxídatían at 16l 
340 nm in a spectrophotometer (Agilent; Santa Clara, CA. USA). Standard 162 
kinetic assays were initia lly used [15,16). but the substrate concentra- 163 
tions were always adjusted to l nsure they were saturating for the tumor 164 
cell enzymes. The activities were detelTIlined under initial-rate condi- 165 
rions: the reactions were started by adding the specific subsrrates and the 166 
absorbance baseline in the absence of one substrate was always 167 
subtracted. PGM, AlaTA and 3PCDH activities were detelTIlined by 168 
standard assays [15,16). TI< and TAwere measured according to (17). 169 
The Ery4P, FBP and 6PG ínhibitíon on HPI in the forward direction 1 iO 
was determined in assay buffer containing 5 mM MgCI2 , 1 mM pp¡, 171 
0.15 mM NAOH. 1 U E. h~tolytica PPi-PfK. 0.36 U ALOO. 9 U 11'1. 3.1 U 172 
aGPDH and 0.OO3~0.014 mg of cyrosolic extract plus 13.8~28 ¡1M Ery4P. 173 
30-60 J.IM 6PG or 3-1 O mM FBP. The reaction was started by addingG6P 174 
(0.1-7 mM ). HPI inhibition was also evaluated ín the reverse reaction in 175 
assay buffercontaining 1 mM NADP+, 2 U G6PDH, and 13-55 J.IM Ery4P, 1 i(i 
40-90 ¡1M 6PG or 3~ 10 mM FllP. and 0.001-0.01 mg of extracto The 177 
reaction was started with the addition of 0.05-1 O mM F6P. 178 
PFK- l activity was determined according to [10) with Iyop hilized 179 
(ammonium-free ) couplíng enzymes. 180 
2.3. 5teady-state metabolite concentrations 181 
AS-30D hepatoma cells were collected from rat ascites fluid. HeLa 182 
cells were cultured under normoxic (95% air-5% CO2 ) conditions in 183 
Dulbecco-MEM medium at 37 oC as previously d escribed 1181. For 184 
hypoxic conditíons. HeLa cells were inirially grown under normoxia: 185 
then at 75-90% confluency, cells were subjected for 24 h to 0.1-0.2% 186 
oxygen in a humidified hypoxia incubator chamber ( BiIlups-Rothen- 187 
berg, California, USA) as described before [19). 188 
The experimental protocol to atrain glyco lytic flux under steady- 189 
state conditíans was described elsewhere [10). Briefly, the cells were 190 
harvested, washed and re-suspended in Krebs-Ringer medium 191 
( 125 mM NaCl, 5 mM KCI. 1 mM MgCI2• 1,4 mM CaCl2, 1 mM KH2P04 , 192 
and 25 mM Hepes. pH 7,4). The cells were incubated at 37 oC under 193 
orbital shaking: after 10 min, 5 mMglucose (or 1 mM for AS-30Dunder 19-1 
P1ease cite this article as: A Marín-Hernández, et al., Mcxleling caocer glycolysis, Biochim Bíophys. Acta (2010), doi:1O.1016jj. 
bbabio.2010.11.006 
 74 
 
ARTICLE IN PRESS 
A. Marín-Heroondez el al. I Biochimi:a et Biophysica Acta xxx (1010) XXX-IUC( J 
195 low glucose concentrarian ) wasadded and incubated for another 3 min 
196 and the reaetían was tenninated by perchloric acid (3% vlv. final 
197 concentration) extrartion. 
198 Glycolytic metabolite concentrations were detennined as previ-
199 ously described (l O]. The concentrations ofNAO+, G1P, alanine and 
200 6PG were determined in neutralized extracts as described befare by 
"'" using LDH; PGM plus G6PDH; AlaTA plus LDH and 6PGDH. 
202 respectively [20) . The Ery4P concentratian was determined in 
""3 50 mM Hepes. 1 mM EGTA. pH 7.4. plus 0.7 mM Xy5P. 5 mM MgCI,. 
""4 0.5 mM NADP+. 0.04 mM thiamine pyrophosphate (TPP). 1 U/mi of 
"'" eaeh HPI and G6PDH. and 0.5 U/mi ofTl(. 
206 For F2.6BP determinatíon, after the 3 min incubatíon with glucose. 
207 the cells were quickly spun down; the pellet was re-suspended in 
208 25 mM Tris-HCI pH 7.6 and the cells were disrupted by freezing in 
209 liqu id N2 and thawing at 37 oC three times. The homogenate was 
210 centrifuged at 20,800 x g for 3 min at 4 oc. The supernatant was 
211 alkalinized with NaOH to a final concentration ofO.1 mM and heated 
212 at 80 °C for 5 min. The sample was centrifuged and the supernatant 
213 was neutralized with acetic acid. F2,6BP was determined by using 
214 potato PPi-PFK as described elsewhere (14] . 
215 For determination of the total intracellular Pi, cel ls were incubated 
216 for 3 min with glucose, and then a sample of 1 mi was withdrawn and 
217 centrifuged at 800 xg for 30 S. The cell pellet was washed three times 
218 with saline (0.9% NaCl) and re-suspended in 1 mlofthe same so lution. 
219 The cellular suspension was divided l nto two: the first sample was 
220 centrifuged at 20,S00.fg for 1 min and the supernatant was recovered 
221 whereas the second sample was extracted with perchloric acid (3% v/ 
222 v final volume), centtifuged for 5 min as aboye and the supernatant 
223 was recovered. Pi was determined in the supernatants as described 
224 elsewhere (21]. The intraceJlular total Pi content was calculated from 
225 the difference between the Pi present in the first supernatant 
226 (extracellular Pi ) from that in the second su pematant (extracellular 
227 plus intraceJlular Pi). The concentration of cytosolic free Pi was 
228 calculated assuming that only 53% of the total content was in its free 
229 form, as previously determined in hepatocytes (221. 
230 2.4. Metabolic fluxes 
231 For the determination of the glycogen synthesis and degradation 
232 fluxes, AS-30D cells (15 mg protein/ml ) were pre-inrubated at 37 oC for 
233 10 min in Krebs-Ringer medium. Then, an aliquot was withdrawn 
234 (t = O) and 5 mM glucose wasadded to the ceHular suspension and kept 
235 under the same conditions. Aliquots were taken at 10 and 30 min, 
236 immediately centrifuged, and the cellular peHet was frozen in liquid N2 
237 and kept at - 70 oC until use. The pellet was further re-suspended in 
238 0.3 mi of 30% KOH and heated for 30 min at 90 oC; afterwards, the 
239 sample was mixed with 0.1 m16% Na2504 and 09 mi 1 00% ethanol and 
2-10 centrifuged at 20,SOOg for 5 mino The pelletwas washed once with 1 mi 
241 of80% ethanot and let dI)' at room temperature. Then, the pellet was re-
242 suspended in 0.5 mi d:2 M acetate buffer pH 4.8 plus 5 U amylogluco-
24.3 sidase and inrubated for 1 h at 37 oc. Glucose units derived from 
244 glycogen breakdown were determined by enzymatic analysis by using 
215 HK and G6PDH (20] ancl reported as nmol of glucose equivalents. After 
246 the addition of glucose, the glycoge'it degradation rate was calculated 
247 from the difference of glucose equivalents in glycogen from the 10 min 
2-18 minus O min inrubation, whereas the glycogen synthesis rate was 
249 calcuJated fro m the difference of glucose equivalen ts in glycogen from 
250 the 30 min minus 10 min incubation. For HeLa cells, glycogen synthesis 
251 and degradation rates were detennined in cells (5-7 mg protein ) 
252 inrubated with and without glucose, respectively. The samples were 
253 taken at O ancl 3 min and treated as described aboye. 
254 For the pentose phosphate pathway (PPP) flux, AS-30D cells were 
255 incubated as described aboye for glycogen metabolism analysis. After 
256 glucose addition, samples were withdrawn at 2, 4 and 6 min and 
257 immediately precipitated with perchloric acid, centrifuged and the 
258 supe rnatant neutralized. The 6PG content was determined in the 
neutralized extracts by using a 6PGDH assay (20] . The rate was 259 
calculated from the differences in 6PG concentrations at2 ifferenttime 260 
points. 261 
The pyruvate consumption rate by mitochondria (mitochondrial 262 
pyruvate metabolism; MPM) was estimated from the total cellular 263 
oxygen consumption rate minus that in the presence of 5 J..1M 264 
oligomycin, and assuming that the main substrate consumed by 265 
mitochondria was the endogenous pool of pyruvate generated by the 266 
cell s (equivalent to 4 mg cellular protein ) incubated with 5 mM 267 
glucose for 5 min. For the calculations it was assumed that ~ mol ofÚ2 268 
(12 atoms oxygen) are consumed per mol of completely oxid ized 269 
glucose. lt is noted that the MPM flux is the resu lt ofthe coordinated 270 
activities of the pyruvate transporte r, pyruvate dehydrogenase 271 
cornplex. Krebs cycle enzyrnes, respiratory chain complexes, adenine 272 
nucleotide translocator, Pi carrier and ATP synthase. 273 
For the majority of the parameters described aboye (enzyrne 274 
activities, metabolite concentrations and fluxes) at least two diffe rent 275 
batches of cells were assayed. The high dispersion in the experimental 276 
data is very common when using non-chemostat cultures and when 277 
using different batches of cellular cultures. On the other hand. lower 278 
dispersions can be certainly obtained by measuring triplicates of one 279 
homogeneous ceH culture. but such results only represent one 280 
independent sarnpling. 281 
2.5. Consrruroon 01 the kinetic models ,.2 
The models were constructed by using the software Gepasi v 3.3 283 
(23] available at the web site http://wW\Y.gepasi.org/. Fig. 1 dia- 284 
grarnmed the pathway reactions considered in both models. The 285 
reactions were wrinen in the program as described in Table 51 and a 286 
surnrnary of all the used kinetic parameters is provided in Tables 52, 287 
53, and 54 of supplemental)' material. 288 
AH glycolytic reactions (including those ofHK PFK-l and PYK) were 289 
considered reversible. For most of them, their Vm values were 290 
determined in cytosolic extracts in the forward and reverse ..reactions 291 
(Table 1 ) thus avoiding the use of Keq values taken from the literature in 292 
the rate equations (see below ). 5ince these Vm values actually 293 
corresponded to those of the enriched cyrosolic fractions. these were 294 
also experimentally determined in whole cells which were solubil ized 295 
with 0.02% Triton X-lOO to obtain the corresponding Vm in total cellular 296 
protein units (Table 53 ); t Ji. Ianer Vm values were used in the mooels. 297 
The kinetic parameters Vmf, Vmr and Km for substrates. products 298 
and modulators were all determined under the same conditions of pH 299 
(7.0 ) and temperature (37 oC) (Table 1). whereas those ofthe glucose 300 
transporte rs were previously determined by our research group [24]. 301 
The oxidative and non-oxidative PPP sections. the glycogen 302 
synthesis and degradation pathways and the MPM braoches were 303 
simplified and considered as non-reversible reactions with the constant 304 
flux values shown in Table 2. The glycogen metabolism reactio ns only 305 
induded the initial substrates and the final products without consid- 306 
ering the intermediate reactions catalyzed by PCM. The oxidative PPP 307 
section did not indude the NADPH production whereas the non- 308 
oxidative seaion only induded the reaction catalyzed by TI<. 309 
The celluJar ATP consuming processes (ATPases ) were represented 310 
as a mass-action irrevers ible reaction with an adjusted k value orO.0042. 311 
To maintain the pyridine and adenine nucleotide balances. the DHases 312 
and adenylate kinase (AK ) reactions were used, respectively. with mass- 313 
action reversible kinetics whose values were adjusted at k, = 250 and 314 
k2 = 1 and k, = 1 and k2 = 2.26, respectively. Conserved moieties were 315 
IATPI + IADPI + IAMP] = 113 mM and INADH]+ INAD+]=1.35 mM. 316 
In the pathway mooeling, metabolite concentrations were initial - 317 
ized at the values found in the cells (Tables 3 and 4 ). Fixed rnetabolite 318 
concentrations were used for lactate. glycogen, 6PG, Ery4P. F2,6BP. Pi, 319 
citrate and Xy5P at the values shown in Tables 2~4 (summarized in 320 
Table 54 of supplementary material ) except for Pi, for which only 53% 321 
of the tota l Pi content was assumed to be free Pi (22J . 322 
Please cite this article as: A Marín-Hernández. e t al.. Mooeling cancer glycolysis. Biochim. Biophys. Acta (2010). doi: 1O.1016/j. 
bbabio.2010.1 1.006 
323 
329 
330 
331 
r 
     
. ¿sica 
Pi 
in 
: Glycogen degradation 
A “6-6 ¿55— G6P,,. 
  
G6P Glycogen       
Glycogen synthesis 
  
7 / FBP (4) 
4 / BPE permea > | HPI 
Los Ear FOP 
1 Y ATP ADP +2Pi 
2 ENAP PFK-1 
1 FBP 
s 
“ DXysP ALD 
G3P + DHAP 
TPI 
G3P 
NAD+*+ Pi 
NADH <——> NAD+ 
DHases ne] GALAH ATP + AMP >? ADP 
1,3BPG 
ATP 
ADP PGK 
3PG 
PGAM 
 
  
ADP +Pi € ATP E (2 
ATPases ade PYK ao - FBP 
r + 12.5 ADP + 12.5 Pi mem? 12.5 ATP 
NADH 
LDH 
NAD+ 
Lactate ip 
Lactate out 
Fig 1. Pathway reactions induded in the kinetic models of glycolysis in AS-30D and HeLa tumor cells. Solid arrows indicate whether reactions are considered non-reversible or reversible in 
Gepasi; dotted arrows indicate allosteric modulators; dashed arrows indicate PPP existing reactions not included in the model. DHases; NADH consuming enzymes; ATPases, ATP-consuming 
processes; PPP, pentose phosphate pathway; MPM, mitochon drial pyruvate metabolism. ATP stoichiometry was calculated by using P/Oratios of 2.5 and 1.5 for NADH and FADH,, respectively. 
2.6, Rate-equations 
The rate-equations used for each reaction are described below. 
The kinetics of GLUT was introduced as a monosubstrate reversible 
Michaelis-Menten equation (Haldane's equation) [25]: 
vin (Gl - 
Ko (1 + GQ) + Gita 
Gluin 
yv= Keg 
in which Glujur and Gluin and Kauour and Kain are the extra- and 
intra-cellular glucose concentrations and the enzyme's affinity 
constants (Km values), respectively; Keq is the equilibriunt constant; 
and Vmfis the maximal velocitv in the forward reaction. 
Please 
bbabio.2010.11.006 
 
The HK rate equation was described as random bi-substrate 
Michaelis-Menten [25]: 
vinf _ PIO) 
$ KaKb (a Keg ) 
SA EA, A, 0, A, A, AE 
Ka — Kb KaKb ' Kp  Kq  KpKq KaKq — KpKb 
in which [A] and (B] represent the substrate concentrations (Glu and 
ATP, respectively), whereas [P] and [Q] represent the product 
concentrations (G6P and ADP, respectively). Ka, Kb, Kp and Kq 
represent the enzymes Km values for their respective ligands. 
The HPI rate equation was introduced as a monoreactant reversible 
equation with competitive inhibition by Ery4P, 6PG and FBP: 
vn 82 _ yyy LES 
v= Kosp Keop 
1 + 166? , [F6P] , TERYAP] , [6PG , [FER] 14 1687, TF6A Yap] , J6PG] , [FBP         
Keso — Króp — Kervao Ko Ko 
cite this article as: A. Marín-Hernández, et al., Modeling cancer glycolysis, Biochim. Biophys. Acta (2010), a] 
332 
333 
333 
336 
337 
338 
339 
340
 75 
 
4 
ARTICLE IN PRESS 
A Marfn-Hernández et al ¡ Biochimica et Biophy i  Aaa xxx (2010) XJUt-}()()( 
________ ~ G ~ IU ~ OO ~ ;~r G ~L~U~T-----------
AT P~ t · H K ~ ~ 
A DP ~ (_)! ~/ ycoge n deg r ada ~ 
,. ... - - - .... -.... -,- - - - SPG ( ppp 6P .. ....... 6P l en 
,,' " , : ~~ ··-¡. ... .. ..... ..\:1 ....... > t HPI Glycogen synthesis , . ~ 
" Ery4P : F6P 
~ •• •• A T P ~ P OP  2Pi , , , 
\ , 
Ery4P ADP +-1- PF K- ~ ,,··· ...... \:) .... .. i"CATP 
~"":>- ..... ___ FBP •••••• ~ •• Itrate 
- ? Xy5P TK t ALD · · · ·· · !·~! · · F2 . 6BP 
3P  O AP 
t 1 
3P 
AD- + ¡::t 
NAOH -- NAO- NADH GAPDH 
ases P  P ~ 2 OP 
A P + Pi ( ATPases 
TP 
~
, 3 PG 
ADP PGK 
 t AM 
2P¡ G 
ENO 
PEP .. ~:) .. .. .. .. ATP 
A TP~ ",'" +) 
ADP +-1- PYK <t. .. ........ · FBP 
~
+ 12 . 5 AOP + 1 25 P I 
H 
H 
D-
ct te In 
ct te t 
MPM ) .5 P 
i . . t ay ti os ded  e kinetic odels of l colysis  S-300 d eLa or Us. lid s m icare hether cti ns re <! rIln-reversible r ersible  
Cepasi; t d r s i te l a t ric odulators; sr <l s ¡ a te P i t  ti ns ot dud <l  e rnodeL Hases; DH rning es; lFases, 1P<Onsu ing 
r ce ses; P, rltose osphate t ay; PM, i tochondrial ruvate etabolis . TP etJy as l t d  i g ratiosof 2.5 d .  rOl' H ro H20 ectively. 
323 . . t - quati ns he  te uati n as scri ed s  bstrate 2 
324 
325 
326 
3" 
9 
0 
1 
he uations s d r ch cti n re scri ed l . 
he i etics f T as ced s  onosubs trate ersible 
ichaeli enten uat ion aldane's uation) J : 
Vmf [Glu,,,I [ ~~; I ) 
v = --,-'--¡;=-,T.:::!...L..-
v _ (1 [GIU;,I) GI 
" Gluoor  KCluin + uour 
 h ich llloul d 1uin d auoul d GJuin re e tra- d 
t - e l lar cose cen trations d e zyme's ffi it  
stants  lues ), ectively; eq  e il ri t" ns tant; 
nd f is e aximal l city  e OTWard cti n. 
ichaeli enten 1 : 3 
Vmf (1A1IBJ _ [PIIQI) 
KaKb eq 
v = -, -+' [All'j' +---'¡B"'j-+" [Ail11I!BiITj"'+' ¡P;;-1 -+' ¡"ñQjF+, ¡pmjrn¡QjO-+' ¡Il'ji17¡QIT¡-+"""",¡Pm1I1B"'1 
a b b p K  K q q b 
 hich J d [ J resent e strate centrations lu d ) § 
P, ectively ), hereas J d J resent e r duct 6 
ncentra tions 6P d DP, ectively). a, b, p d q 7 
r sent e es  l es r eir ective g ds. 8 
he Pl te uation as ced s  onoreactant ersible 9 
uation ith peti tive i iti n y r 4P, G nd P: ..1  
Vmf jG6p] _ Vmr [F6P] 
Ka;p F6P 
v = -, - """'¡C"'6P"'j-+-n¡F"6"'PI"'+""'¡"'ER"'yr.4P'"j"'+"""¡ "'PGI""--+' ¡"'f ""Pj 
c6P F6P ERy4P ~pc FBP 
M ase it  is rtide s: . arí - ernández, t l.. odeling cer l colysis. i him i phys. cta 10). doi: 1O.1016/j. 
bi . 0. 1. 06 
Marín-Hernández et al. / Biochimica et Biophysica Acta xxx (2010) xxx-—xxx 5 
 
tul Ta 21 
Kinetic parameters of AS-30D and HeLa glycolytic enzymes and transporter. 
3 Enzyme AS-30D Hela Enzyme AS-30D Hela 
t1L4  GLUT Vmp 0.055* 0.017* GAPDH Vmy P > 
tL5 Km tu 0.52% 93" Vm, 09 250 
16 HK Vmy 046 0.06* Kima3p 0.29 0.19 
t17 Ka Gtu 0.18> 0.1 Kms 38m 0.02 0.022 
tL.8 Km are 099 11 KmnaD+ 0.08 0.09 
+19 Ki c5p 0.02 0.02- Kmnaoe 0.004 0.0 
110 HPI Vmy 49+19 (3) 12402 (3) Km ps 11+1 (3) 29 
Ln Vm; 34+1.1 (6) 2.8 (2) PGK Vmp 27 13 
1.12 Km sp 09+02 (4) 0.4+ 0.03 (3) Vm, 43 38 
t1.13 Km sp 0.07 +0.03 (6) 0.05 (2) Kmi3orc 0.035 0.079 
t114 — PFK-1 Vmy 0.273 0.0784 Kmsr 0.12 0.13 
t1.15 Km pop 4.6% 10% KMaop 0.67 0.04 
11.16 Km ar 0.048* 0.021* Kmarr 0.15 027 
LAT Kin 1.75% 20 PGAM Vmy 20 14 
+1.18 Kierr 3.90 68% Vm, 13 053 
t1.19 Karsar 1810 8.4x 10% KmarG 0.18 (2) 0.19 
t1.20 ALDO Vmy 023 0.25 Kmarc 0.04 0.12 
t1.21 Vm, 0.18 NM ENO Vmy 0,51 036> 
t1.22 Kmesp 0.01 0.009 Vm, 0.74 04 
113 Kimcap 016 NM Kmaro 0.16 0.038 
11.24 KMomar 0.08 NM Kipgp 0.04 0.06 
4125 TPL Vmy 5.6 5 PYK Vny 66 > 
11.26 Vm; 56 42 Kmpep 0.4 0.014 
1.27 KMomar 19 16 Kmapp 03 04 
128 Kimcap 0.41 051 LOH Vmy 134 114 
1.29 Vm, 18 NM 
11.30 KMpyr 0.13 03 
+13 Kimtac 47 NM 
t1.32 Kmnap 0.07 NM 
1.33 Km nana 0.002 NM 
Vm in the forward (Vmy) and reverse (Vm,) reactions in Ux (mg cytosolic protein)” *; Km in mM. Values taken from *[24]; *[10]; 766], [Moreno-Sánchez R, Marín-Hernández A, 
Encalada R and Saavedra E, unpublished results]. “Vm in U x (mg total cellular protein). The number of independent batches of cells assayed is shown in parentheses; the absence of 
t1.34  parenthesis indicates one assayed preparation. NM, not measured. 
  
  
  
  
12.1 Table2 
Properties of some glycolytic branches in tumor cells. 
AS-30D Hela Table 3 13.1 
- Glycolytic flux and intermediary concentrations obtained in vivo (cells) and by in silico 
124 Glycogen metabolism modeling for AS-30D cells, 
125 PGMactivity' 032+0.1 (3) 075+02 (3) 132 
2.6 Kméip NM 0.07 Metabolite 5 mM glucose 1 mM glucose* t3.3 
2 zo a 
glycogen content E E de ) E Em y ) In vivo? Model In vivo? Model t3.4 
G1P content? 0.08 + 0.04 (3) NM Gltin 62+1 3.4 NM 0.8 
glycogen synthesis flux* 22403 (3) 24 (2) G6P 5,3426 6.5 2405 (4) 3.0 
glycogen degradation flux! 1.2 (2) 1242 (3) ro 1,5 +0.7 0.03 0.7402 (4) 0.016 
FP 25+7.6 5.2 0.6403 (3) 0.36 
Pentose phosphate pathway DHAP 10423 14 1403 (3) 40 
G6PDH activity! 0.05* 022* G3P 0.9+0.4 0.3 0.38 (2) 
6PG content? 0.35 +0.13 (5) 039 1,3BPG ND 0.01 == NM 
PPP flux* 0.096+0.03 (3) NM 3PG ND 0.01 NM 
TA activity! 0.043+-0.006(3) 0.033 (2) 2PG ND 0.04 NM 
TK activity" 0.010 40.001 (3) 0.037 PEP 0,140.02 0.003 NM 
Ery4P content? 1403 (3) 0016> Pyr 2141 0.84 0.72 (2) 
XyI5P content NM 0.016 lactate 27411 Fixed NM 
F2,6BP (x 107?) 6+1(3)> Fixed NM 
Triglyceride synthesis citrate 1.740.7 Fixed NM 
ALGPDH activity* ND* ND* ATP 5.6412 7.9 6(2) 
ADP 2.4407 2.1 1.5 (2) 
Amino acid metabolism AMP 3.3414 1,3 NM 
3PGDH activity" ND ND Pi 4.8+1.9 (3)* Fixed! 5 (132 
AlaTA activity' 0.046+-0.022 (3) 0.012 (2) NADH NM 0.005 NM 
Alanine content” ND ND NAD? 1.3+05 (4)> 1.34 NM 
Glycolytic flux 2149 29 10.5 (2) 
 
Mitochondrial pyruvate metabolism 
Metabolite concentrations in mM; flux in nmol lactate min” * (mg cellular protein)”*. E a 
Elec? pyraralecconsumed by mituciondia"! ESTA nu Values taken from *[10]. This study. In the model, when this condition was simulated, 
TU (mg eytosolic protein)” *; ?nmol glucose equivalents (mg total cellular protein)*; Gltzur concentration and glycogen synthesis flux were fixed at values of 1 mM and 
“in mM; *nmol min” * (mg total cellular protein)” *. Values were taken from *[10] and 1nmol min”? (mg of cellular protein)”*, respectively. “53% of the total Pi 
*[39]. “The glycogen concentration was calculated by assuming that 1,8 mg total concentration shown was assumed to be free Pi [22] which was used for pathway 
cellular protein has a volume of 2.28 tl [38]. The values are mean + SD and the number modeling. Figures in parentheses indicate number of independent cellular extracts 
of independent batches of cells astayed is shown in parentheses: the absence of assayed. NM, not meastired; ND, not detected. Fixed values in the model were at those 
12.32 parenthesis indicates one preparation assayed. NM, not measured; ND, not detected, experimentally determined in cells, 13.26 
16/). 
bbabio.2010.11.006     
 76 
 
ARTICLE IN PRESS 
A. Marin ~ HtTrtdrr:1ez l't l. I i chi iro rt i hysica cta 100( 10) x- x 
l.1 ablt 1 
inetic r eters r 5- OD !'ld eLa l colytic l es .lOO lr<lnsporter. 
tl.2 
tl.3 
l.4 
l.5 
e 
G f 
t l .6  
l.7 
AS-300 
¡ . 5 .... 
Km ~ . 2" 
Vm¡ D.46b 
mCI  D. 1Sb 
,,1P U 9' 
H," e 
O.O 1 7 " ~ DH ¡ 
.3" Vrn, 
. 6' "'<;3P 
0.1 Km I.3B~ 
.1 "w'!'D+ 
 
A5-30D ,,. 
l ' i' 
O.gb .5' 
 U 9 
0.02 . 2 
0.08 . 9 l.8 
tl.9 
t .1O PI 
i CliP 
Vm¡ 
o. c 0.02" 
. ± 1.9 ) .2±02 ) 
NIIDH . 4 
~ ~ K m ~ % 1 ) 
Vm, .  ± .1 ) .  ) tU I CK Vm¡ 7' ' 
t 1.12 a,p D.9 ± 0.2 ) . ±o' 3 ) Vrn, .3 .8 
l. 13 Km ~ o.07±O.03 ) . 5 ) '.,lDfG . 5 . 9 
l .I-I - I ¡ o.273d O.07Bd m3A:: 0.12 un 
l .15 Km ~ ~6 ' .0' Km ~ 0.67 . 4 
tl 6 ,,1P D. 8d O. 1d 
K~ TP . 5 Q27 
Ki.t1P . Sd 0' tl.17 PGAIVI ¡ 0' . b 
tl.18 Kicrr ,,,., 
.Sd rn, .3 U53 
l 9 KOF2liBP 1.8 X 10- 4<1 8.4 x 10- 4<1 3A:: . 8 ) UI  
l al OO ¡ Q23' ' 2A:: . 4 . 2 
, U 8  l 1 O rn¡ O.S l b Q.36b 
. '2 
tl.Zl 
t .2-I 
tl. 5 
tl 6 
tl:n 
tl.28 
t . 9 
tl.:I) 
tI.JI 
l 2 
tl 3 
l 3-I 
t .  
a.2 
t2.3 
t2.4 
t .5 
t2.6 
t2. 7 
a.S 
t2_9 
t2. 10 
t2.1I 
t2. 12 
t2. 13 
t2. 1"\ 
12. 15 
t2_16 
t2. 17 
t2.18 
12. 19 
t2.3) 
t2_21 
t2.22 
t2_23 
t2. 2-l 
t2.25 
".26 
t2.27 
t2.28 
t2.29 
a.'" 
t2.31 
Km ~ UOI . 0 rn, 0.74 U4 
fTlGlP UI 6  lA:: . 6 . 38 
mOHAP U 8  Km EP . 4 Uo. 
¡ 6  
m, 56' " 
I PYK m¡ .6b i' 
",,", .  . 4 
mDlw> .9 .6 Km ~ .3 .4 
rTI(;lP U 1 .51 H ¡ I A A 
Vrn, .8  
m". . 3 U3 
",,< .7  
_o . 7  
NNJH 0.002  
  l e ard ¡) d r  ,. ) cti ns  x g tosolic r tei ) 1:   M. alues l en  · [ 4); 1 0); "1 6), d[Moreno-Sáochez . Marín~ H emánde z , 
ncalada J.nd vedra , pu blished r e s ults~ ~Vm  g tal Uular rotein).1 e ber fin endent lches f Us .ayed  n  rentheses; l e sence f 
pa thesls l i ates e <L yed r paralion. , ot easured. 
il bw 2 
r perties f rne l colytic t..-aoches  or Us. 
A5~30 D H,,. il ble  
lycolyt ic l x d ediary ncenlralions t i ed  i  Us) ilnd y  i/  
Glycogm metabolism odeling r A5~30D Us. 
 aClivilyl .32±0.1 ) . S ±02 ) 
!,p  U 7 Met.tbolile   l cose   l cosec 
g lycogen contenrl 33±30 (S) 171 ± 55 (4 ) 
 i o" Model  i t! od,' ( 26 mM)" ( 135 mM )C 
1 P conlenlJ .08± . 4 ) NM Glu;n .2 ± 1 .4  .  
l gen lhesis l · ± 0 .3 ) 4 ) G6P 5.3 ± 2.6 6.5 ± 05 ) .  
l gen gradation 1 4 .  ) ± 2 ) ffiP .  ±0.7 0 .03 7 ± 02 ) . 16 
fRP 25± 7.6 .2 ± .3 ) . 6 
mt se sPOOlt ol ay DHAP ± 2.3   ± 03 ) .0 
6 OH Clivityl UOS' Q 2' 3P . ±OA 0.3 ~ 038 ( 2 ) 0.09 
 nlenlJ . S ± 0.13 ) U39 3  O 0 .01 -  0 .002 
P 1 4 . ±0.03 )  O O . 1  0.005 
 ti ityl  ± 0. 06 ( 3) . 3 ) O O 0.04  0 .016 
 ti ityl . 10 ± 0 . 01 3) . 7 PEP .1 ± 0.02 0 . 03  0 .001 
r 4P ntenrl ± .3 ) .016b Py' ± 1 0.84 . 2 ) 0.78 
ylSP ntent  0 b lactate ± 11 Fixed  Fixed 
n.68   10- 3) ± 1(3 )b Fixed  Fixed 
ri l ri  fhl'sil cilrale 7± 0.7 Fixed  Fixed 
aG  ilcti i ty 1 O' ND' ATP 6 ± 12 .'  ( 2) 4.9 
ADP 2 ± .7 . 1 .5 ) 2.9 
mino oro 'toboli  AMP 3 ± 1.4 .3  3.9 
OH Clivityl O ND Pi 8 ± 1.9 )' Fixl."d" )' Fixed d 
l  ti ityl  ± 0.0 2 ) . 12 ) NAOH  0.005  0.005 
lanine ntenrl O ND NAO' 3 ± 0.S )b . 4  1.34 
lycolytic   ± 9 9 0.5 ) 14 
ilOchondrial ruvale / boli  
elabolite ncenlralions  ;   ol t le i  1 g ll lar r t i ) 
Flux ofpyruvate consumed by m itocnondria4 1.8 ( 2) NM 
alues .1ken · IOJ. b-¡"his tudy. eln e odel, hen is ndition as u lated, 
l  g c t solic ein ) ' ; 2 ol l cose uivalents g tal ll lilr t  .. lUoul ncenlralion d l gen nthesis  ere  l i l es f   d 
< 1 
<3.2 
< .3 
<3 ' 
<3.5 
<3.6 
13.7 
<3.S 
13.9 
13.10 
13. 11 
13_1 2 
13. 13 
13. 14 
13. 15 
13. 16 
13_17 
13. 18 
13_19 
<3.2IJ 
tJ.2 1 
<3.22 
13.23 
13.24 
13.25 
J¡  M: 4 ol in- 1 g l .11 ll lar r tei )- I. alues ere .t en  · I OJ d I nmol in - 1 g f Uular r le in ) - I, spective ly. dS  f e Ola l Pi ~ 
1 9). "r e l gen ncenlration as l lated y ing at .  g l tal r ncen l rdlion n WdS d ed 10 '* r e i 1 2) hich WilS sed r dl dy ==-
ll lar r tein as  VOlumeo r 2~ 8J. 1he i l e s are ean± O d l e ber odeli ng. i res  pare l ses i a te ber f endenl ll lar tracts 
"" 
f endent atches f ls ed  n  renthese s: l e sence f sayed. , ot easb red: O, old t cted. i ed alues  e odel e re t l se 
renthesis i a tes e r paralion yed. , ol easured; O, ot etected. eri entally i <!  ls. 
Please cite this artiele as: A Marín-Hernández, et al., Modeling cancer glycolysis, Biochim. Biophys. Acta (2010), doi :l0.10 fj. 
abiO.2010.11.006 
<3." 
 77 
 
".1 
tA.2 
".3 
"A 
"., 
".0 
w.i 
".s 
" .. 
1A.IO 
tA.] 1 
w. 12 
1A. 13 
1.4.14 
l.-t i :) 
14.16 
tA.J i 
IA. IB 
tA.19 
".20 
14.21 
tA. Z2 
1.4.23 
tA. 24 
tA.25 
14.26 
341 
343 
344 
345 
3 .. 
348 
349 
350 
351 
352 
353 
35' 
355 
356 
357 
358 
359 
350 
361 
• A Marin-Hernández t>f al l Biochimica ti: Biophysica Acta xxx. (2010) 100(-100( 
T.1ble 4 
Glyoolytic nux and intermediary concentrations obtained in vivo «(e lls) and by in si/ro 
modeling for HeLa cells. 
Metabolite Nonnoxia Hypoxiac 
In \.1VOo Model 'n vivd' Model 
GIUin NM O.61 ~ NM l.. 
C6' l.3 ± OA (5)b 0.66 - l.4 ± OA (5 ) 1.0 
ffi' O.5±02 (5 )1> 0.01 05± 02 (5) 002 
FB' 0.38(2)" 0.14 0.23 (2) 052 
OHA' O.93± OJIl 2.0 0.54 16 
G3P NO 0.08 NM 01. 
1.3BPG NO 0.0009 NM 0001 
3PG NO 0.006 NM 0000 
2PG NO 0.003 NM 0004 
PE, 0.32 0.0002 NM 00003 
"Y' 8.5 ± 3.6 2.5 42 (2) ,. 
lactare 33 Fixed NM ",," F2,6BP (x 10- 3 ) 42±O.8 (3)" Fixed NM ",ed 
a lrate NM Fixed NM ",ed 
ATP 8.7 ± 3 (5)" 8.' 7.9 ± 4 (5) 7.7 
ADP 2.7 ± 1.3 22 1.8 (2) " AM P 0.4 (2)" 13 NM 1.2 
p; 75 (2)" Fixed 7.8 (2)" Fixed 
NAOH NM 0.005 NM 0005 
NAO+ NM U. NM U. 
Glycolytic nux 16± 12 ( 5)" 20 21 ± 9 (5)b 29 
Melabolile roncentrations in mM; nux in nmol lactate min 1 (mg cellular prOlein ) 
Values taken frorn ~ 1101 . bPresenL study. cFor lhis condition. HK and HPI activities were 
determine<! in cells expose<! to hypoxia by 24 h; GWT activity was estimate<! frorn lhe 
protein content oblained by Weslem-blOl analysis (Fig. 2); lhe ATPase rale was .lIso 
adjuste<! to a value of 4.5 x lO- l . Figures in ..r:a rentheses indicate number of 
independent cellular lloltches assayed. NM, not measuf'ed ; NO. nOl detecte<!.. The fixe<!. 
values use<!. for modeling were those experimen ~ r m i ne<!. in cells, except for Pi 
which Wd5 adjusted lO lhe free (OIKfmlr.ttion. -
with Vm,- as the maximal rate in the reverse (g luconeogen ic) 
direction. 
The kinetics for TPI, PGAM and ENQ were depicted by mono-
substrate s imple reversible Michaelis-Menten equa tion: 
v = 
in which [SJ and [PJ are the respective concen trat ions of subs trates 
and products with their respective affin ity constants . 
PFK-1 is an allosteric enzyme showing cooperative behavior with 
respect to F6P and hyperoolic kinetics with respect to ATP. The rate 
equation for th is tetrameric enzyme was the concerted transition 
model of Monod, Wyman and Changeux for exclusive ligand binding 
(F6P, activators, and inhibitors ) [25J together with mixed-type 
activation (F2,6BP or AMP or Pi) and s imple Michaelis-Menten 
terms for ATP and reve rse reaction. ATP (at high concentrations ) and 
citrate are the allosteric inhibitors. L is the allosteric transition 
constant ; KaF26BP is the activation constant fo r F26BP; Kicrr and Ki.~ TP 
are the inhibition constants forcitrate and ATP; a and ~ are the factors 
by which Kf6P and Vm change when an activator is bound to the active 
enzyme form (R conformatian in the Monad model ). 
1( ,K",Ktq ( ~ 1) - ¡ADI1 i rn ~ ¡AlF!M 
Y,;;' + "K,;' + ~ + I 
The ALDO rate equation was the revers ible Uni-Bi Michaelis- 363 
Menten equation. 
V ,( [FBP[ V [DHAP[[G3P[ 
v = _ ---,rm"--m _ ' -i K'""O'p"" - .----m _ ',,;.K<i D 1f!l!,,,.K=,~/fu_ """"''>ffi 
1 + _[FB_P] + _[DH_A_PJ + _[G_3'_1 + "i ILD",H,-,APJ'-4i'[G",3P-,-] 
KFBP Kow,p Kc3I' KDHAPKc3P 
364 
366 
The GAPDH kinetics was described by a simplified ord ered Ter-Bi 367 
reversible Michaelis-Menten equation: 368 
VmJ INADIIG3PIIPI] Vm, ~NADH] ~ 
_ KHAD l<cJpKp¡ I<tm:;KHJ.OO -
v - [NAD] [NAD[[G3Pj [NADIIG3pIlPí j[llPqjNA [NAbAl · 
1+ -- + + + + --
KNAD KHJ.oKGJ p KHAd<GlP KFl ~ K HADH KHADH 
300 
Rate equations far PGK and LDH were represented by the random 371 
Bi-Bi reversible Michaelis-Menten equation for non-interacting 372 
substrates (fl and ~ = 1). 373 
Vmf [AIIB] - Vmr [PIIQ] 
aKaKb ¡3KpKq v= ---",,-----,;;;-....::'TIi;ii;r- --ifiiT.!ii"'- ¡¡;;-- "" + [A] + [B] + [AIIB] + [PIIQ] + [PI + [Qr 
Ka Kb aKaKb ¡>KpKq Kp Kq 
370 
PVK kinetics was represented by the concerted transition model of 376 
Monod, Wyman and Changeux rate equation for exclusive Iigand 3n 
binding includ ing PE P as well as FBP activation and ATP inhibition, 378 
with simple Michaelis-Menten terms for AOP and reverse reaction. 379 
:![\+;: ] =- ( ~;i::~ ) 
( 
[A11'1) - ¡AIPI + [PY1I1 + IA"W'YRj + I 
1 1 +¡¡:¡;; [1 ""'J' ~ -;r,;;:; ~ 
( FBP)' + + ¡¡-
I + - ](Q ~- .. 
380 
The ATPases rate defined as the ATP-consuming cellu la r processes 382 
was depicted as mass-action irreversible reaction, in which k was the 383 
rate constant and Si was the substrate concentration. 384 
386 
The DHases and AK were included as reversible mass -action 387 
reactians, in which k, and k2 are the rate constant s, Si is the 388 
concentration of substrate and Pi is the concentration of producto 389 
390 
The rest of glycolytic branches were adjusted to irreve rsible 392 
constant flux. 393 
3. Resul ts and d iscussion 39-1 
3. 1. Glycolytic enzyme activities, intermediary concentrotions and fluxes 395 
in tumor cells 396 
A recurrent di fficu lty encountered in building kinetic models of 397 
metabolic pathways is the lack of all the required kinetic data of the 398 
pathway under study, evaluated in the same cellular mooe l, and 399 
determined under the same experimental conditions. Oespite these 400 
inconveniences, sorne models have been reported for glycolyt s in 401 
erythrocytes [26,27J, yeast [28) and the human-infecting parasites 402 
Trypanosoma brucei [29) and Entamoeba histolyóca (16). 403 
Although there is sufficient kinetic infarmation on glycoly tic 4().t 
enzymes from a wide variety of tumor cells, most stud ies are not 405 
Please cite this article as: A Marín-Hernández, et al., MooeJing cancer glyco lysis, Biochim Biophys. Acta (2010), doi: 10.1016jj. 
bbabio.2010.11.006 
 78 
 
ARTICLE IN PRESS 
A. Marin-H!'T'nández ef al. I Biochimira et Biophysica Acta 100{ (2010) 1001.- 1CXX 7 
406 usually accompanied by thorough cellular merabalic characterizations 
407 reporting metabolite concentrations and metabolic f1uxes. This is 
408 partirularly evident in studies based only on determinations of 
409 transcriptional changes. from which hypotheses and explanations are 
410 formulared about changes in metabolic fluxes and cellu lar function 
411 (reviewed and disrussed in 9].Hence. in arder tohave a complete kinetic 
412 characterization ofthe glycolytic pathway and its branches in the tumor 
413 cells studied heTeo the maximal velocities in the farward (Vm¡) and 
414 reverse ( Vmr) reactions and Km values for subs trates (Kms ), products 
415 (Kmp ) and effectors for each glycolytic enzyrne were determined in 
416 cytosolic extracts from AS-30D and HeLa cells at 37 oC and pH 7.0 
417 (Table 1). A pH value of 7.0 was chosen because the intracellular pH 
418 varies from 7.2 to 6.8 when ce lis are actively consuming glucose (30] . 
419 Due to thermcxlynamic constraints, HK. PA<-l and PYK activities were 
420 only deteITIlined for the forward reartion. 
421 In general. most of the enzyme activity (i.e., Vm ) va lues (Table 1) 
422 were in agreement with those previo usly reported for the same tumor 
423 celllines (31 ,32). In tumor ceUs, approximately 50-70% of HK activity is 
424 bound to mitochondria [10,33,34) : thus, to have an acrurate determi-
425 nation ofthe total HKartivity, detergent-peITIleabilized cells were used 
426 (Table 53). Remarkably, HK. PA<-l and PGAM activities found in 
427 cytosolic extracts we re 7.6, 3.5 and 14 times higher in AS-30D than in 
428 He La cells: which contrasted with the similar values obtained for the 
429 rest of the pathway enzymes in both cell s (Table 1 ). HK and PFK-l are 
430 153 and 27 times lower in rat normal hepatocytes (3 and 10 mU/mg 
431 protein, respective ly) than in the AS-30D hepatocarcinoma (1 O~ On the 
432 other hand, HKactivity in human muscle celis is 10 times lower whereas 
433 PA<- l artivity in human brain is 3 times higher than in HeLa cells (6 and 
43·1 258 mU/ mg protein. respectively ) (35,36) ; however. for a more rigorous 
435 compari son w ith HeLa cells, normal proliferating cervix epithelial ce ll s 
436 from which these tumor cells derived, should be used. 
437 The Km values were simi lar in the two cell types (Table 1) and 
438 were within the same interval found in the BRENDA database ( http: // 
439 www.brenda-enzymes.infol) determined for normal and tumor cell s. 
4.10 The previously reported value for the glycolytic flux in AS-300 ce lis 
4.11 under normoxia and S mM glucose was 21 nmol min- 1 (rng ceUuJar 
4·12 protein ) - l [lO) which was used as reference in the present 
443 study (Table 3 ). On the other hand, in HeLa cells a value of 16 ± 
4·14 12 nmol min- 1 ( mg cellular protein )- l ( n = 5 ) under normoxic con-
4.15 ditions was obtained in the present study (Table 4), which was lower 
446 than the previously reported value of32 ± 1 O nmol min - I (mg celluJar 
447 protei n)-l 110), indicating variability within the same ceU line from 
4.18 culture passage to culture passage. For comparison, nOITIlal rat 
449 hepatocytes and normal Chinese hamster ovary celis (CHO-Kl ) show 
450 glycolytic flux values of 2.4 and 8 nmol min- 1 ( mg cell protein )- l, 
451 respectively 110,37). Thus, the tumor cells used here indeed exhibit 
452 increased glyco lysis rates campa red to normal ceUs. Glycolytic 
453 metabolite concentrations were also previously reported for AS-300 
454 cells inrubated with S mM glucose (lO) and used as reference (Table 3), 
455 whereas for HeLa celis under nOITIloxia. values for sorne metabolites 
456 were re-determined (Table 4 ), finding no stati sticaJ difference with 
457 those previously reported [lO). 
458 The enzyme artivities, fluxes and inteITIlediary concentrations ofthe 
459 main glycolytic branclleS we re aiso evaluated (Table 2): glycogen 
460 synthesis and degradation, ppp, triglyce ride and serine synthesis, and 
461 MPM. 
462 Regarding g lycogen. HeLa cells showed 5.2 fold highe r content than 
463 A5-30D cells and 2.3 rold increased PGM activity (Table 2 ). The glucose-
46.1 l -phos phate (Gl P) intracellula r concentration (0.08 mM; Table 2 ) was 
465 Iower than its glycolytic precursor G6P in AS-300 celis (5.3 mM: 
466 Table 3 ). Under the steady-state conditions used, in which extracelluJar 
467 glucose concentration is S mM (see Experimental procedures seaion ), 
468 the glycogen syn th esis and deg radan on rates were 10-18 times lower 
469 (Table 2 ) than the glycolytic nuxes (Tables 3 and 4 ) in both twnor cells; 
470 with the exception ofthe glycogen degradation flux in HeLa cells which 
471 was 75% the glycolytic flux. The most l robable reason for the higher 
activity in the glycoge n degradation branch in HeLa cells was that these 472 
are cu lrured in Dulbecco-MEM medium containing 25 mM glucose, 473 
whereas the gl ucose concentration found in AS-300 ascites fluid is 474 
0.026 mM (38~ Cultivation of HeLa ce lis under high (25 mM ) glucose 475 
induces the ex pression ofthe 18 .. gltlt:8se am A i ~ LUT1, whereas AS- 476 
300 hepatoma developed intra-peritoneally in rats predominantly 477 
express the l'igh gll:lE8se affiRity GLUf3 (24) . In consequence, the 478 
inrubation of HeLa cell s in fluxes assay medium containing lower 479 
glucose (compared to culture medium) mostprobably promoted strong 480 
activation of glycogen degradation to compensa te for the diminished 481 
glucose uptake: this compensatory mechanism was apparently not 482 
required in AS-300 hepatoma. 483 
In the oxidative section of PPP (Table 2 ), higher G6POH activity 48.1 
was found in HeLa than in AS-30D ceUs, but similar 6PC concentra- 485 
tions. On the other hand. the activities of the PPP non-oxidat ive 486 
section enzymes TA and TI( were similar to reported values for t he 487 
same and other tumor cells (0.019-0.208 and 0.022-0.108 U 488 
(mg prote in)-I, respectively) (17,39). A lowe r Ery4P concentration 489 
in comparison to that of glyco lytic intermediaries (G6P, FBP, and 490 
OHAP) was determined, which was slightly higher than the previously 491 
reported value for Krebs ascites hepatoma cells (O.s nmol/mg ceUular 492 
prot ei n ) 140). The lower PPP flux and enzyme activities (Table 2) 493 
compared to glycolytic flux (see text aboye for values ) and enzyme 494 
activities (Table 1 ) suggested low nuc1eotide synthesis and hence 495 
negligib le cellular proli fera tio n, as expected from incubations made in 496 
a saline buffer for sho rt times (3-10 min). For A549 lung carcinoma 497 
and C6 rat glioma, PPP fluxes of 1 and 6 nmol/min':'mg protein, 498 
respectively, were determined by using radio-Iabelled glucose and 499 
lengthy incubations of 4-6 h in culture medium with glucose and fetal 500 
bovine se rum, conditions under which cellular growth is favored 501 
(41,42) . Then , it seems possible that the PPP fluxes, determined here 502 
from 6PG changes (Tab le 2 ), were underestimated as the variation in 503 
the cellular 6PG content was small and clase to the lower detection 50.1 
limit ofthe methcxl used. Further evaluation ofthe PPP flux in AS-300 505 
and HeLa cells either with radio-Iabelled glucose or under ce ll growth 506 
conditions for longer times should l esolve this apparent discrepancy. 507 
Low or negligible nGPOH activ it ies, which are in volved in 508 
triglyce ride synthesi s, have been reported for several tumor cells 509 
(43) ; accordingly, nGPOH activity was not detected in AS-30D and 510 
HeLa cell s (10). Triglyceride synthesis can al so be initiated by OHAP 511 
acylation in a reaction catalyzed by OHAP-acyltransferase 144) : 512 
however, thi s enzy me was not determined in the present study and 513 
this branch was not included in the model . 514 
A previous report indicated that serine metabolism is acce lerated 515 
in tumor cells (45) : however, the activity o f 3-phosphoglycerate 516 
dehyd rogenase was not detected when measured in the 3PG 517 
oxidation direction (Table 2 ) and thus, this branch was not inc1uded 518 
in the mode!. Alanine transaminase activity (AlaTA) measured in the 519 
pyruvate synthesis direction was low in both AS-300 and HeLa cells 520 
(Table 2 ). These low activities correJated with undetectable alanine 521 
levels in cellular extracts. 522 
The MPM rate was 11.7 fold lowe r than the lacta te production rate, 523 
suggesting that anaerobic glycolysis was favored. It might be possible 524 
that the mitochondrial pyruvate transporter expressed in AS-300 cells 525 
also has low affinity and low rate as described for Morri s 44 and 3924A 526 
hepatomas ( Vm = 5-12 mU/ mg mitochondrial protein and 527 
KmPyr= 0.74-1.1 mM) (46~ The ex pression of a low artivityflow affinity 528 
pyruvate transporter would favor a higher pyruvate flux towards lactate 529 
because of LDH higher activity and affinity (Vm = 2.0 U/ mg celluJar 530 
protein, Table 53; KmPyr= O.13 mM, Table 1 ). Contributing to the 531 
unusually high endoge nous pyruvate pool (Tables 3 and 4 ) may be the 532 
active oxidation of alternative substrates such as glutamine, glutamate 533 
and proline, which generate maJate and hence pyruvate by the aaion of 534 
over-expressed malic enzyme in cancer cell s 138,47). In addition, the 535 
onsel o f the Crabtree effect (OxPhos inhibition by external glucose) 536 
should also promote an elevation in the intrace llular pyruvate poo l 130). 537 
Please cite thi s artic1e as: A Ma rín-Hernández, et al., Mcxleling cancer glycolysis, Biochim. Biophys. Acta (2010), doi: 1O.1016/j. 
bbabio.2010.11.006 
8 
Marín-Hernández et al / Biochimica et Biophysica Acta xxx (2010) xxx-xxx 
3.2. -uilding and refinement of the kinetic model of AS-30D glycolysis 
The kinetic model of glycolysis in AS-30D cells was constructed by 
using the Gepasi software. The kinetic parameters Vm, Km for substrates, 
products and modulators for the pathway enzymes and glucose 
transporter from external glucose to lactate production as well as the 
enzyme activities and fluxes of the pathway branches experimentally 
determined were used to fulfill the parameters of the rate-equations 
describing each reaction as detailed in the Experimental procedures 
section. The metabolite concentrations and the pathway fluxes of 
glycolysis determined in vivo under steady-state conditions (Table 3) 
and the control distribution obtained by elasticity analysis previously 
reported by our research group [10] (Table 5) were used as reference 
parameters for model validation. 
In consequence, for model building and validation, a refinement 
process was recurrently necessary, which in turn prompted an 
experimental re-evaluation of the kinetic properties of some enzymes 
as well as the determination of neglected metabolite concentrations and 
the inclusion of some branches to ¡mprove model reproduction ofin vivo 
pathway behavior. Thus, this dynamic interplay between modeling and 
experimentation led to several model and rate equation modifications, 
from which the three most important are described below. 
3.2.1. Kinetics of PFK-1 
The first draft of the model included the reactions from external 
glucose to lactate production and the branch of glycogen metabolism 
(Fig. 1), in which a simplified version of the PFK-1 rate-equation 
(hyperbolic kinetics) was used. However, G6P level was too high 
whereas those of DHAP and G3P were too low. Thus, an improvement 
of the PFK-1 reaction was necessary. 
There is a lack of detailed kinetic analyses of PFK-1 in normal and 
tumor cells, in which a simple Hill equation for only one modulator is 
used, Hence, a kinetic characterization of PFK-1 was carried outin AS- 
30D and HeLa cells (see Experimental procedures section), resultingin a 
rate-equation that followed the concerted transition model of Monod, 
Wyman and Changeux together with non-essential activation by F2,6BP 
(or any other allosteric activator) (Moreno-Sánchez R,, Marín-Hernán- 
dez A., Encalada R. ¿nd Saavedra E., unpublished results). The 
intracellular concentrations of F2,6BP, AMP, ATP and citrate were also 
determined (Tables 3 and 4). As theactivating effect of F2,6BP overcame 
the inhibitory effect of citrate and ATP, at the tumor physiological 
concentrations of these modulators (see a simulation in Fig S1 in 
supplementary material), the PFK-1 ratein the presence of physiological 
F2,6BP was 2 fold higher with the concomitant increased in FBP and 
DHAP to near-physiological levels. However, a marked decrease in the 
Glin, G6P and F6P concentrations was attained. 
3.2.2. Kinetics of HPI 
Through modeling it was noted that a decrease in the HP activity 
resulted in a better prediction of the G6P concentration and the 
glycolytic flux. Hence, the effect of a wide variety of metabolites (ATP, 
AMP, Pi, PPi, citrate, Ery4P, DHAP, G1P, lactate, 6PG, FBP and F2,6BP) 
on HPI activity was examined. Most of them showed no effect (Table 
S5 in supplementary material); in contrast low Ki values for Ery4P 
(0.8-2.5 HM), 6PG (6.8-18 hM) and FBP (60-170 uM) were deter- 
mined (Table $6). Thus, multiple competitive-type inhibition by Ery4P 
(Fig. S2), 6PG and FBP was incorporated in the HPI rate-equation. The 
potent HPI inhibition by these metabolites was not particular of the 
tumor enzyme, since it was previously reported for the normal rat 
brain, and rabbit liver and muscle HPls [48-50] and they also 
modulated the enzymes from rat hepatocytes, yeast and E. histolytica 
(Table S6). Therefore, the intracellular concentrations of 6PG and 
Ery4P were also determined in the tumor cells, and in consequence 
the oxidative section of PPP and TK reaction with their corresponding 
fluxes (Table 2) and metabolite interactions with HPI were also 
included in the model (Fig. 1). 
bbabio.2010.11.006 
3.2.3. Kinetics of the FBP consuming block of enzymes and GAPDH 601 
By including the two modifications described above, the modeled 602 
FBP concentration was still several times lower than the physiological 603 
concentration, indicating an unrealistic high rate by the gonsumption 604 
block of this metabolite. Thus, the effects of G6P, F6P, Pi, ATP, AMP, 605 
Ery4P, F2,6BP and PPi on the enzymes of the ALDO-ENO segment 606 
were tested (see Table S5). 607 
ALDO activity was not modulated by any of these metabolites. 608 
Although GAPDH was inhibited by ATP (with non-competitive 609 
inhibition, Ki=4.3 mM) and ENO was inhibited by DHAP and ATP 610 
(non-competitive inhibition, Ki=31 mM, and uncompetitive inhibi- 611 
tion, Ki=14 mM, respectively), these Ki values were far from the 612 
physiological concentrations, and their effects were not included in 613 
the rate-equations. 614 
GAPDH from AS-30D cells showed low affinity for Pi (Table 1; 615 
Fig. S3) which suggested that Pi availability might regulate the 616 
enzyme activity. High Kmp values for GAPDH from bovine liver, 617 
human muscle and sarcoma [51,52] have also been documented. 618 
Thus, contrary to other glycolysis kinetic models in which Pi is 619 
considered saturating and hence it is not included as pathway 620 
metabolite, in the present model Pi was included in the GAPDH 621 
rate-equation. On the other hand, from the total intracellular Pi 622 
concentration determined (Table 3) only 53% [22] was assumed to be 623 
free (ie., 2.5 mM); this free Pi concentration was used for modeling 624 
and was in agreement with values previously determined by nuclear 625 
magnetic resonance in rat liver and heart (1.3-2.8 mM) [53,54]. 626 
3.2.4. Properties of the kinetic model of glycolysis in AS-30D hepatoma 627 
cells 628 
By incorporating (i) the PFK-1 complex rate equation, (ii) the 629 
inhibition of FBP, 6PG and Ery4P on HPI, and (iii) Piin the GAPDH rate- 630 
equation, the kinetic model was able to predict with high accuracy the 631 
concentration of most of the pathway metabolites and flux rate 632 
determined in living cells (Table 3). The most significant exceptions 633 
were the predicted FBP (4.8 times lower) and F6P (50 times lower) 634 
concentrations. It was found that the FBP steady state concentration 635 
directly depended on the Pi concentration within the reported 636 
physiological range (Table $7). On the other hand, only by decreasing 637 
to 20% the PFK-1 activity, the F6P reached the physiological level; 635 
however, this decrease was not experimentally supported because of 639 
the strong activation by F2,6BP, AMP and/or Pi. Moreover, under this 640 
last condition of20% activity, PFK-1 became the main flux control step, 641 
which was in disagreement with the results obtained by elasticity 642 
analysis [10]. Hence, still another unknown kinetic mechanism seems 643 
to also modulate the in vivo PFK-1 rate. 644 
3.2.5. Control distribution of glycolysis in AS-30D 645 
The analysis performed by Gepasi provided the dh of all the 646 
pathway steps (Table 5). Modeling indicated that the first three 647 
reactions of glycolysis exerted most of the flux-control with 648 
HK>HPI>GLUT (2C;, =1.04; Table 5). This control distribution was 649 
in agreement with the C/, obtained by elasticity analysis, which 650 
indicated that the G6P producer block (GLUT and HK) exerted the 651 
main flux-control (XC!, =0.65-0.71) [10], whereas the Cp value was 652 
not possible to be discerned in the previous study, because the high 653 
experimental dispersion masked flux-control differences among the 654 
different blocks of enzymes analyzed. 655 
HK and HPI showed high flux-control because they were strongly 656 
inhibited by glycolytic and PPP intermediaries. lt was previously 657 
demonstrated [10] that HK exerted significant control on flux in 658 
cancer cells because it was strongly inhibited by its product G6P, 659 
despite its higher levels of expression compared to normal cells. In 660 
particular, AS-30D cells over-expressed the HK-II isoform [24] which 661 
partitioned between the cytosol and mitochondria, the latter showing 662 
the same affinity constants for glucose, ATP and G6P than the former, 663 
cytosolic HK II [10]. Thus, although a hallmark of tumor cells is the 664 
Please cite this articie as: A. Marin-Hernandez, et al, Modeling cancer glycolysis, Biochim. Biophys. Acta (2U1U), d01:10.1016/j. 
 79 
53 
539 
010 
011 
542 
013 
' 14 
015 
016 
&17 
018 
01. 
550 
55 1 
552 
553 
554 
555 
556 
557 
558 
55. 
560 
561 
562 
563 
564 
565 
566 
567 
56 
13 569 
570 
571 
572 
573 
574 
575 
576 
577 
578 
57. 
580 
581 
582 
583 
584 
585 
586 
587 
588 
589 
590 
591 
592 
593 
594 
595 
596 
597 
598 
599 
600 
ARTICLE IN PRESS 
8 A. M(Jrfn-H~nándt l t l l i i ica t i hysico cto x OJO) KXJ(-XJO( 
. . Buil i g d n ent 01 me i etic odel 01 S-300 i olysis 
he i etic odel fg lysis  S-  ls as nstrurted y 
i g  epasi l't are. he i etic r eters .  r bstrates, 
nxiucts d odulators r e t ay rnes d l cose 
sporter r  terna l l ose e t re r duction s ell s e 
rne ti ities d es f e t ay r ches eri entally 
t ined ere sed  lfi l e r eters f e uations 
scri i g ch cti n s etailed  e eri ental cedu5,..es 
ti n. he etabolite centrations d e t ay es f 
l colysis t ined  i o der y-state nditions able ) 
d e ntrol i t ti n tai ed y l sti ity alysis r i usly 
oned y ur arch r up lO) able ) ere sed s r ce 
r eters r odel l i ation. 
 sequence, r odel il i g d li ation,  ent 
r ce s as u rently ce sary, hich  rn r pted  
eri enta l al ati n f e i etic r perties f rne es 
s e ll s e t ination f glected etabolite centrations d 
e rw:lusion f e r ches l prove odel r duction fin i  
t ay ehavior. hus, is a ic lay t en odeling d 
eri entation   eral odel d te uation odifications, 
 hich e ree ost portant re scri ed l . 
. . J. i eti s / K- J 
he st raft f e odel l ed e cti ns  ternal 
l cose  tate r duction d e r nch f gen etabo lis  
ig. ),  hich  plified ersion f e · l uation 
perbolic i etics) as sed. owever. 6P el as o i h 
hereas se f O AP d 3P ere o . hus.  pr vement 
f e f1(·l cti n as ece sary. 
here    f eta iled i etic alyses f f1(·l  r al d 
or ls,  hich  ple iII uation r ly e odulator  
sed ence.  i etic aracterization f f1(·l as rri d t in S-
0 d eLa e ls e peri ental r cedures cti n ). lt  in  
uation at o ed e ncerted slti n odel f onod, 
yman d hangeux ether ith n-e sential ri ati n y , 8P 
r y t er l steric ti ator) orencrSánchez .. a rín· ernán· 
ez ., nca lada . and avedra .. published sults ). he 
e lular ncentrat ons f 2, P. P, TP d it te ere l o 
t ined bies  d ). s t eacti ati geffect F . P er e 
e i it ry ff ct f it te d TP. t e or ysiological 
centrations f se odulators e  ulation  i . 5   
l entary aterial ), e f1(-l t  in e r sence f siological 
, 8P as  l  i her ith e comitant r ased  P d 
O AP  ar- hysiologicalle els. owever,  arked crease  e 
IUin, 6P d P centrations as t i ed. 
. .2. i eti s / PI 
hrough odeling t as ted at  ecrease  e I cti ity 
lt d   e ter r diction f e 6P centration d e 
l colytic . ence, e ffect f  ide ariety f etabolites TP, 
P. i. P¡. it te. r 4P. HAP. l P. I etate. . 8P d . 8P) 
n PI ti ity as ined. ost f  ed o ffect able 
S  pl entary aterial ) :  ntrast  i lues r r 4P 
. -2.5 ¡1 ).  . - 8 ¡1 ) d P 170 ¡1 ) ere eter-
ined able 56). hus, ultiple petiti e-type i iti n y r 4P 
ig. 52),  d P as rporated  e PI t · uation. he 
tent PI i iti n y se etabolites as ot arti ular f e 
or e. ce t as r iously orted r e r al t 
rain, d bit er d uscle Pls -50] d y l o 
odulated e es  t patocytes, east d . i t l ti  
ble 56), herefore. e eJlular ncentrations f  d 
r 4P ere lso t ined  e or ls, d  sequence 
e idative cti n f P d I< cti n ith eir r nding 
es able ) d etabolite cti ns ith PI ere l o 
el ed  e odel ig. ). 
. . i eti s al m  P ing l k al es d DH 1 
y l ding e o odifications scri ed oye. e odeled 2 
P ncentration as ti l eral es er n e ysiological 3 
centration, i ati g  realistic i h te y e.1:9ASl:lffif)h9R .:1 
l ck f is etabolite. hus, e cts f 6P, 6P. i, TP, P. 5 
r 4P, , P d Pi n e es f e l - O ent 0 
ere t d e able S ). 7 
l O ti ity as ot odulated y y f ese etabolites. 8 
lt ugh PDH as ibited y TP ith n -compe itive 9 
ibition, i 43 M) d O as ibited y AP d TP 0 
n· ompe itive ibition, i  31 M, d compe itive hibi· 1 
n, i  14 M, ectively ), se i l es ere r  e 2 
ysiologica l ncentrations, d eir f cts ere ot l ed  3 
e t · uations. 4 
PDH  - 0D ee ls ed J  ffi ity r i ab le ; 5 
ig. 5 ) hich gested at i ailabi li ty ight ulate e 6 
e ti ity. igh fi l es r P H  vine er, 7 
an uscle d a , 2) ave l o en cumented. 8 
hus, e ntrary e t er l colysis i etic odels  hich i  9 
si ered t rati g d nce t  ot l ed s t ay 0 
etabolite.  e r sent ode l i as l ed  e P H 1 
uation. n e t er nd,  e tal t e lular i 2 
centration t ined able ) ly  1 2J as ed  e 3 
e i e.. .  M ); is e i ncentration as sed r odeling 4 
d as  r ent ith l es r i usly t ined y clear 5 
agnetic ance  t er d eart 3 .8 ) ,54). 6 
. .4. r perti s f  i etic odel DI l colysis  ·  t a 7 
cells 8 
y r r orating i) e -l r plex te uation, ii ) e 9 
ibiti n fF P.  d ry4P n Pl. d ii ) i in e PDH te· 0 
uation, e i etic odel as le  redict ith i h racy e 1 
ncentration f ost f e t ay etabolites d  te 2 
t r ined  i g ls able ). he ost ificant ceptions 3 
ere e r dicted 8P .8 es er) d 6P 0 es er) 4 
ncentrations. t as d at e P y te ncentration 5 
ir ctly ended n e i ncentration ithin e orted 6 
ysiologica l ge able 57). n e t er nd. nly y creasing 7 
 0% e f1(·1 tiv ity, e P hed e ysiological el; $ 
ever, is crease as ot peri entally ported ecause f 9 
e g ti ati n y , P. P d/or i. oreove r, der is 0 
st ndition f 0% ti ity, K-l e e e ain  ntrol t p, ·U 
hich as  ent ith e sults t i ed y l sticity ,12 
alysis 1 0). ence, ti l other o n i etie echanis  s 3 
 lso odulate e  i  f1(- l te. 4 
. . ( IToI i t ti  / glyroly s ~  -  5 
Ite alys is r nned y epasi r i ed e CJI f 1l e ,16 
t ay s able ). odeling i ated at e t ree 7 
cti ns f glycollsis erted ost f e · ntrol ith -18 
HK ~ HP I > GLlJf ( ~C f¡= 1.04; able ). his ntrol i t ti n as ·19 
 r ent ith e ci¡ t i ed y l sticity alysis. hich 0 
i ated at e 6P r ducer l ck lJf d ) erted e 1 
ain · ntrol "iCf¡  0.65-0.71 ) lO] . hereas e C ~pt alue as 2 
ot sible  e i r ed  e r vious y, cause e i h 3 
eri ental i ersi n asked · ontro l i ces ong e 4 
i rent l cks f es alyzed. 55 
K d PI ed i h · ntrol ecause ey ere gly 6 
i ited y l colytic d P ediaries. It as r iously 7 
onstrated ] at  e rted ifi ant ntrol n   58 
cer ls cause t as gly i ited y t  r duct 6P, 9 
espite  i her els f ression e pared  r al e Us.  0 
articular. S-300 ls er· pre sed e K-II  ) hich 1 
arti ed t en e t sol d itochondria. e tt r ing 2 
e e ffi ity nstants r l eose, TP d 6P n e r er, 3 
t solic  11 lO). hus, gh  l a rk f or ls  e 4 
l ase ite is rtiele s:  a rín· ernández, t l., odeling ncer l colysis, i chim i phys. cta 010), oi: 10.1016/j. 
abi . 10.11.006 
 80 
 
ARTICLE IN PRESS 
A. Marín--Hernáooez el al. I Biochimiro rf Biophysica Acta}(}{}( (2010) xxx-xxx • 
"'. 1 Tabie S 
Aux contro l coefficients obtained in vivo (cells) <lOO by kinetic modeling. 
"'.2 eli "'.3 Enzyme 
"' .. AS- AS-300 model Hela model 
300 
"'.5 In viva' 5 mM lmM Normoxia Hypoxia 
Clurose Glucose 
"'." j:b' Jf I IlK ~ O.2 ~ 0.12 03. 032 
"" O.71 a 0.44 ~ 0.50 0 .08 02 
"'.S HPI+:::: g l)( ~ - 0.02' O." 0.46 0.05 0.1 
"'." -- - 0.004 - 0.1)9 - 0.009 - 0.005 
1.6.10 -0.1 - 0.1 - 0.1 - 0.06 
1.5. 11 Glycogen degradation NM n05 0.11 057 0,]3 
1..5. 12 PFK-l O.06a 0.02 0.03 0.03 0.12 
t.6.13 ALOO + lPl + GAPDH + 02 4a n08 0.009 0.01 on. 
PGK + PGAM + ENO + (GAPDH O.OS ) 
PVK + LDH 
t5. 14 MPM NM 0.003 0.02 - 0.004 - 0.001 
1.5. 15 TK NM 0.01 0.023 0 .014 0 .01 
t.6.16 ATPases NM -0.1 - 0.16 - 0.02 - 0.05 
a Values taken fmm [101. PI'P, pentose phosphate p.1thway; MPM, mitochooorial 
pyruvate metabolism NM, not measured. For AS-JOD, both condition s were modeled 
under normoxia. Fo r HeLa, both conditions were mode led in lhe presence of 5 mM 
t.6.17 glurose. 
665 increased activity ofthe glycolytic enzymes. the same enzyme activity 
666 regulatory mechanisms are apparently preserved to avoiding unre-
667 stricted increase in flux. which in turn can lead to ceH death because of 
668 the turbo des ign of glycolysis (55). On the other hand. HK also exerts 
669 signifi cant flux control in normal ce ll s 126,56) making this step less 
670 su itable for specific drug targeting in cancer cells. 
671 In turno HPI displayed a high d. despite being one of the most 
672 active enzymes in the pathway when determined under Vm 
673 conditions in the absence of modulators (Table 1). However, the 
674 potent combined HPI inhibition by FBP and the PPP intermediaries 
675 6PG and Ery4P led to severe activity decrease, becoming now limiting 
676 for pathway flux. The modulation of HPI activity by these metabolites 
677 might be a control mechanism ofthe G6P concentrat ion which in turn 
678 may exertcontro l on (a ) the synthesis of glycogen; (b) the PPP flux for 
679 NAOPH and ribose-SP syn theses; (c) the production of UO P-
680 glucuronate for proteo-glycan and glyco proteins synthe sis and 
681 xenobiotic detoxification; and (d ) the ceHula r ATP homeostasis by 
682 avoiding its depletion through an enhanced HK reaction. Remarkably, 
683 this HPI inhibition was also present in the yeast and ameba enzymes. 
684 As the potent inhibition on HPI here described was not considered in 
685 previous published models of glycolysís, which might change the 
686 control di stribution. the putative HPI controlling role remains to be 
687 experimentally eva luated in other biological systems. On the other 
688 hand. the sign ificant HPI flux control in AS-300 cancer cell s (cf 
689 Table 5 ). and its negligible controlling role in normal cells 126,56). 
690 indicates that thi s step is a suitable therapeutic target. 
691 The resu ltson HPI and GAPOH kinetiG emphasize theimportance. for 
692 metabolic modeling. of an extensive knót..ledge of the enzyme kinetic 
693 properties, which shou ld not be restricted to the sale characterization 
(l9.1 with their substrates, but also to its interaction with its products and with 
695 other pathway intermediaries. which are not always assayed when 
69(i studying the purified enzymes. This has been previously demonstrated 
697 by reconstitution of pathway segments of amebal glycolysis (57). in 
698 which interactions with non-substrate metabolites but pathway gly-
699 colytic intermediaries were elocidated for several enzymes. 
700 GLlJf displayed significant flux control (Cb.ur= o.2; Table 5) as 
701 previous ly suggested by elasticity analysis [10J and its kinetic 
702 properties [24J. From the main four isoforms (GLlJf1-4) , AS-30D 
703 cells mainly express GLlJf3 which has a high affinity for glucose 
704 (Km=0.52 mM) 124J. Moreover, by over-expressing all GLlff iso-
705 forms in Xenopus oDCytes and determining the ki netics parameters 
706 with 2-deoxyglucose, it was previously concluded that GLlJf3 is the 
707 transporter with the highest catalytic efficiency (reviewed in S). In AS-
300 cells. GLUT displayed lower activity and catalytic efficiency (Vrn/ 708 
Km ) than HK and HPI; however, the inhibitory effect on the lanertwo 709 
by glycolytic and PPP metabolites decreased their activities to levels 710 
that conferred flux-controlling roles. In contrast, in the kinetic models 711 
of glycolysis in yeas t 128~ T. brucei 1291. E. histolytica 1161 and heart 712 
[56), GWT contributed by >50% to flux-control. This difference in 713 
GLUT flux control between tumor cell s and heart indicates that this 714 
step might not be an adequate target for cancer treatment. beca use its 715 
inhibition would preferentially affect normal cell s over tumo r cells. 716 
The rest of the flux-control (0.08) was found in the AlDO-LDH 717 
segment (Table 5 ), which was also in agreement with that obtained by 718 
elasticity analysis (I Ctl = 0.24 ) [10]. Within this segment, most ofthe 719 
control was accounted by GAPDH (C& PDH = O.OS ). due to its Pi low 720 
affinity and hence its dependence on the variation of the Pi 721 
concentratíon in AS-300 cells (Table 57). The ATP demand (ATPases) 722 
showed a low flux control but ofthe negat ive sign. indicating that this 723 
reaction competes with the glycolytic ATP-consuming s teps for ATP 724 
thus inhibiting the glycolytic flux. 725 
3.2.6. VVhy a step contToIs flux? 726 
The HK elasticity coefficient (Table 58 ) and the catalytic effidency 71.7 
for ATP (Vm jKm) were l mong the lowest in both AS-30D and HeLa 728 
glycolysis, indicating near-satu ratio n with concomitant activity 729 
insensitivity to varíations in ATP. relatively lower transfonnation 730 
efficiency and hence flux limitation at this leve!. In turno the reasons 731 
for the HPI high flux control coefficient resided in its step-wise 732 
increasing elasticities (sensitivities ) towards the potent inhibitors 733 
FBP, 6PG and Ery4P. and low elasticity for its product F6P, ind icating 734 
satu ration, and hence strong inh ibit ion and prevalence of the reverse 735 
over the forward reaction. Simila rly. GLlff catalytic efficiency was the 736 
Jowest among the pathway steps, ind icating strong flux restriction; in 737 
addition, in HeLa cell s, GLlff eJasticity for its producto GIUin, was one of 738 
the lowest. ind icating near-saturation and perhaps product-inhibition 739 
and significant reaction reversibility and correlating with a h igher 740 
GLUT flu x control in these cells. On the other hand, the low flux- 741 
control exerted by PFK- l is explained by the low elas ticity towards 742 
activators, indicating saturation and hence full activation. If PFK-l 743 
were not fully activated, it would very like ly the main controlling step 74.1 
as its elasticity towards F6P is the lowest. On the other hand, the PFK-l 745 
elasticities for citrate, and for their products FBP and ADP. were 746 
extremely low because these metabolites were not sensed due to the 747 
high Ki and Km. and low intrace llular levels. 748 
3.2.7. Control of metabolite concentrations 749 
The control ofthe concentration of G6P. PEP and ATP depended on 750 
Gurr, HK. HPI and ATPase activities in AS-30D glycolysis and on GLUT, 751 
glycogen deg radarían and ATPase in HeLa glycolysis (Table 59). For 752 
the F6P concentration. PFK-l in both AS-30D and HeLa glycolysis was 753 
the predominant controlling step, as expected. 75·1 
3.2.8. Model robustness 755 
The modeled pathway behavior ( ~ ., flux-control and concentra- 756 
tion control distribution, flux rates and metabolite concentrations) 757 
was not signifi cantly altered by changing (decreasing by half o r 758 
increasing by 2 times) the ki netic parameters of most steps and PPP 759 
flux (data not shown ). Exceptions were (I) control d istribution andjor 760 
metaboJite concentrations, which varied by more than 50% when 761 
changing the main controll ing steps GLlff, HK. HPI, PFK- l and GAPDH 762 
Vm values (AS-30D model ) or GLlJf. glycogen degradarían. PFK-l and 763 
GAPDH Vm values (HeLa mode l, see below); (ii ) FBP concentration, 764 
which varied by 21-100% when changing ALDO Vm and Km vaJ ues in 765 
both models; and ( iii ) DHAP concentration , which varied by 9--100% 766 
when changingTPI. GAPDH. PGK. PGAM. ENO and ?YKKm va lues (AS- 767 
300) or only TPI Km values (HeLa ). 80th models were also highly 768 
sensitive to changes in the ATPase kinetics. revealing that this step is 769 
the weakest component in conferring stability to the pathway. 770 
Please cite this artide as: A Marín-Hernández. et al., Modeling cancer glycolysis, Biochim. Biophys. Acta (2010), doi: 1O.1 016jj. 
bbabio.2010.1 1.006 
 81 
 
ARTICLE IN PRESS 
10 A Marin-Hernández t>f al l Biochimica l1 Biophysica Acta KXX (2010) 1W(.-1OOf. 
77 1 Notwithstanding the enlisted exceptions. the predicted pathway 
772 behavior showed satisfactory robustness. further vaJ idating the 
773 models constructed for both cancer celllines. 
774 3.3. Kinetic model 01 glycolysis in HeLa tumor ce l/s 
775 The kinetic madel for HeLa cell s glycolysis was constructed based 
776 .liso on the experimental parameters determined in these cell s 
777 regarding enzyrne activities (Table 1). metabolite concentrations 
778 and fluxes of the glycolytic pathway (Table 4 ) and ¡ts branches 
779 (Table 2), and the Tate equations used in the AS-30D model. 
780 The majority ofthe metabolite concentrations and fluxes predicted 
781 by the mode l correlated we ll with the in vivo concentrations except 
782 for F6P and PEP, which were 50- and 1600-times lower, respectively 
783 (Table 4). These discrepancies indicated that some kinetic experi-
784 mentation-based refinement was still needed, most probably at the 
785 level of PFK- l and ALOO, and PVK. 
786 According to the modeL the main flux controlling steps in HeLa 
787 ce lis, grown under normoxic and high glucose (25 mM ) conditions, 
788 and assayed in the presence of phys iological glucose (5 mM), were 
789 the glycogen degradation pathway and GLlIT ( C ~ymg~ n dq-radaion + 
790 C ~ wr = 0.96), whereas the control exerted by the block of HK + HPI + 
791 PFK-l was 0.16 (Table 5). The higher control attained by GLlIT in HeLa 
792 cells compared to AS-30D was caused by the over-expression of the 
793 GLUTl isoform, which displays low rate in these cells (0.017 nmol! 
79-1 min ~ mg cellular protein) and low affinity for glucose (9 mM), resulting 
795 in a lower ratalytic efficiency than that ofGLurJ in AS-30D cells (24). At 
796 physiological glucose concentration (5 mM ), the rate equation for GLUTl 
797 predicts an activity of 25% of its maximal acrivity (4 nmol min- I mg 
798 cellular protein- I
), which cannot completely account for the obselVed 
799 glycolytic flux (16±12nmol lactate min- I mg cellular protein-I).ln 
800 consequence, the glycogen degradation pathway activates to provide the 
801 bulk of glucose equivalents to l each the observed high glycolytic flux. 
802 The high flux through the glycogen degradation pathway and the 
803 higher PGM acrivity determined in HeLa cells, compared toAS-300 cells, 
804 supported a relevant role fo r this pathway in delivering glucose 
805 equivalents for glycolytic flux (Table 2).l n fact, HeLa ce ll s incubated in 
806 the absence of glucose were able to produce lacrate (7 nmol · min- I ·mg 
807 cellular protein- 1), which corresponded to 22% of the lacrate produced 
80 in its presence. In contrast, lactate production in AS-300 cells was 
809 completely dependent on available external glucose (10). On the other 
810 hand at 25 mM glucose, the pred icted GLUTl rate in HeLa ce lls 
811 (11.7 nmol min- I mg cellular protein- I
) can easily cope with the 
812 required glucose supply for theglycolytic pathway. which brings about a 
813 decreased flux control exerted by both glycogen degradation and GLUT 
814 (data not shown). Moreover, it is well documented that glycogen 
815 phosphorylase is inhibited by high glucose and G6P levels. with the 
816 consequent decrease in glycogenolytic flux (58] . It might be possible 
817 that when HeLa ceJls are incubated with 5 mM glucose, the intraceJlular 
818 concentrations of these metabolites concomitantly a1 so diminish as a 
819 consequence of the low GLlIT activity, thus favoring glycogen 
820 phosphorylase acrivation, increasing glycogenolysis and hence acquir-
821 ing higher flux-control 
822 Thus, the diffe rent control distribution between AS-30D and HeLa 
823 cells was probably related to the environmental prevailing conditions 
824 in which ceJls we re grow n. The low glucose concentration encoun-
825 tered by AS-30D cells in the rat intraperitoneal cavity (0.026 mM) 
826 favored the expression of a high affinity GLlIT and led to a relatively 
827 low glycogen contento whereas the HeLa ceU high-glucose containing 
828 culture medium (25 mM) favored the expTession of a low affinity 
829 GLlIT and led to high glycogen content. 
830 3.4. Kinetic modeling under different steady-states 
831 It is worth emphasizing that the control distribution as obtained by 
832 modeling can only be applied to the specific steady-state cond ition 
analyzed, which depended on a particular set of enzyme activities 833 
expressed in the ceJl s. Therefore. ir was interesting to test the validity 834 
of the model predictions in cells subjected to other environmental 835 
conditions. 836 
During tumor growth, ceJls localized far from blood vessels are 837 
exposed to hypoxia and nutrient starvation. It is well known that 838 
under hypoxia the transcription facto r HI F-lcx induces an increase in 839 
the transcription of most glycolytic genes resulting in .In over- 840 
express ion of the enzymes and transporters of this pathway 841 
(reviewed in S]. On the other hand, under low glucose concentration, 842 
an increase transcription of HK-ll, PFKB3, GLlITl and GLlIT3 induced 843 
by AMP kinase has been observed (59,60] . 844 
With the valida ted kinetic models, low availability of external 845 
gl ucose (for AS-300) and hypoxia (for HeLa) conditions were 846 
examined. To this end, it seemed that only a small se t of data was 847 
needed to be experimentally re-evaluated by focusing on the steps 848 
with the highest control Although a generali zed increased activity 849 
under hypox ia in the glycolytic enzymes has been documented for 850 
most tumor cell s (reviewed in 11, a higher increase activity in non- 851 
controlling steps does not affect pathway flux rate and flux-control 852 
distribution (Section 3.2.8 ..... above). 853 
3.4.1. Modeling AS-30D glycolysis under low g/ucase concentrorion 854 
Steady-state intermediary concentrations and glycolytic flux 855 
determined in AS-30D cells incubated with 1 mM glucose were 856 
-50% lower for G6P, F6P, G3P and Pyr, and flux rate, and one order of 857 
magnitude lower for FBP and OHAP (Table 3). The AS-300 kinetic 858 
model predicted well the values for flux and G6P, FBP, G3P and Pyr 859 
concentrations found in the cell s, but for OHAP and G3P the values 800 
were 4 times higher and 42 times lower, respectively (Table 3). 861 
Essentially the same flux-control distribution was anained under low 862 
glucose ( HK ~ HPI>GLlIT), with the glycogen degradation branch 863 
gained more control, as expected (Table 5). 864 
3.42. Modeling HeLa glycolysis under hypoxic condirions 865 
HeLa cell s exposed to hypoxia did not show signifi cant changes in 866 
G6P, F6P and ATP steady-state concentrations, whereas a 31% 867 
increased glycolytic flux was indeed found compared to cells grown 868 
under normoxia (Table 4). Increased enzyme activities under hypoxia 869 
for HK and HPI were also determined (37% and 30%, respectively ) 870 
whereas a 252% increased protein content for GLlIT was determined 871 
by Western-blot analysis (Fig. 2) and then a corresponding increase in 872 
transport activity was assumed fo r mode ling. By introducing these 873 
new parameters. the kinetic mode l satisfactorily predicted the flux 874 
rate and the metabolite concentrations (Table 4 ). In addition, hypoxia 875 
did not bring about a change in the control distribution (glycogen 876 
degradation >GLlIT> HK), although the difference in the control 877 
exerted by glycogen degradation and GLlIT compared to normoxia 878 
(31%) was mainly transferred to HK, PFK-1 and the final pathway 879 
segment (Table S). 880 
The experiment of subjecting HeLa cells to hypoxia and nutrient 881 
stalVation and determining fluxes, activities and metabolite concentra- 882 
tions is difficult to achieve because oflow cell numberyields. Therefore. 883 
the HeLa model was used to predict the pathway behavior under both 884 
conditions (hypoxia and 1 mM glucose: data not shown). The results 885 
showed a simila r control distribution to that obtained under normoxia 886 
and 5 mM glucose (co lumn 5 ofTable 5) although with decreases in flux 887 
and metabolite concentrations by 36-86%. compared to hYJX)xia alone 888 
(calumn 6 of Table 5). These accurate predictions under a variety of 889 
physiological steady-state conditions revealed aga in high robustness of 890 
the kinetic models adding further validation. 891 
3.5. Enzyme ritrarion Ior the identification 01 the best drug torgets 892 
The ki netic mode ls allow the relationship between glycolytic flux 893 
and a given enzyme aa ivity to be analyzed (Fig. 3). As long as the 89-1 
Please cite this artide as: A Marín-Hernández, et al., Modeling cancer glyco lysis, Biochim Biophys. Acta (2010), doi: 10.1016jj. 
bbabio.201 0.11.006 
nica 
  
ó 
(kDa) _N_H_ 5,00 | 
Ss 
E E 150 
. o 
a-tubulin 55 — A 2 
100     
 
     
GLUT-1 
Fig. 2. Glucose transporter protein content after hypoxia treatment in HeLa cells. Bar data represent the mean + SD of at least three different preparations. N, normoxia; H, hypoxia. 
Western-blot analysis was carried out according to reference [19]. 
models may reproduce the in vivo pathway behavior, they are a valid 
tool for identifying the drug targets with the highest therapeutic 
potential in a metabolic pathway. Thus, GLUT, HK, HPI, PFK-1, TPI and 
PYK activities were varied in the AS-30D hepatoma glycolysis model, 
or GLUT, HK, HPI, TPI and the glycogen degradation branch inthe HeLa 
model, to establish how much these potential drug targets should be 
inhibited for attaining significant decrement in the glycolytic flux and 
ATP concentration. To decrease the pathway flux (or the ATP 
concentration; data not shown) by 50%, it was required to inhibit, 
individually, the main controlling steps HK, HPI or GLUT by 76-78% in 
AS-30D glycolysis (Fig. 3A), whereas those of HeLa glycolysis GLUT, 
HK, HP or glycogen degradation required 87-99% inhibition (Fig. 3B). 
In contrast, to achieve a similar flux diminution, the non-controlling 
steps PFK-1, PYK or TPI in AS-30D glycolysis needed to be inhibited by 
71%, 99,2% and 99.5%, respectively (Fig. 3A). Interestingly, 50% 
  
0 20 40 60 80 100 
Fl
ux
 %
    
0 20 40 60 80 
Enzyme activity % 
100 
Fig. 3. Dependence of tumor glycolytic flux on enzyme activity. The reference 100% enzyme 
activity values were those corresponding to the respective Vm values for the forward 
reaction (Tables 1 and $3) whereas 100% fluxes were those predicted by each model (A for 
AS-30D; B for HeLa cells), which were the fluxes through LDH (29 and 20 nmol min” ' (mg 
cell protein)”, respectively; Tables 3 and 4). In A, enzymes, transporters and braches were 
(a) GLUT + HK + HPI; (b) GLUT + HK; (c) HK or HPI; (d)GLUT, (e) PFK-1; (£) PYK and (g)TPL. 
In B, (a) GLUT+ glycogen degradation; (b) glycogen degradation; (c) GLUT; (d) HK; (e) HPI 
and (f) TPI. When two or more steps were titrated, identical variation in activity was applied. 
A decrease of the Vmy value was accompanied by a proportional decrease in the Vm, value. 
[E cite this article as: A. Marín-Hernández, et al. Modeling cancer glycolysis, Biochim. Biophys. Acta (2010), doi:10.1016/j. 
bbabio.2010.11.006 
inhibition of flux (or ATP concentration) was attained, in AS-30D 
glycolysis, when GLUT+-HK or GLUT+HK+HPI were decreased 911 
simultaneously by the same magnitude, 63% and 47% inhibition, 912 
respectively, whereas in HeLa glycolysis, when GLUT+glycogen 91: 
910 
degradation were decreased simultaneously by 48% (Fig. 3B). 914 
4. Concluding remarks 915 
The kinetic models for both tumor cells predicted reasonably well 916 
fluxes and metabolite concentrations found in living cells, To achieve 
high robustness, the models were built up with the kinetic properties : 
of each individual enzyme determined experimentally including 
(a) the most common ligands (substrates, products and modulators), 92 
and (b) the interactions of several other pathway intermediaries with 9 
the enzymes (Fig. 1). Also, the great majority of the changes made to 922 
improve and refine the models were based on “wet” experimentation. $ 
This allowed us to identify some mechanisms by which a traditionally 
considered non-controlling enzyme such as HP! may exert significant 
control on tumor glycolysis. However, further rounds of model 
refinement are still necessary to reach more prediction accuracy, 
especially for F6P and PEP concentrations. Nevertheless, the present 
models may serve to predict the pathway behavior under other 
physiological conditions and in other cancer cell lines as long as the 
most controlling steps are experimentally evaluated, 
Thus, kinetic modeling is a helpful theoretical-experimental 
approach to identifying the most promising therapeutic targets. It 93: 
has been proposed that inhibition of glycolysis in tumor cells can be o: 
an alternative therapeutic strategy [6,61] and several therapeutic 935 
targets have been proposed (GLUT1, HKII, PFKFB3, GAPDH, LDH, and 936 
MCT) [37, reviewed in 62-65]. It then results of clinical relevance to 937 
experimentally evaluate whether such targets are indeed flux control 938 
steps of the tumor cells' pathway. Moreover, it would be desirable that 
such targets have high flux control in tumor cells, but low control in 
host, normal cells. 
In normal cells, HK and PFK-1, together with GLUT, exert the main 
control on the glycolytic flux [reviewed in 8]. However, due to the 
different enzyme activity levels, different control distribution was 
expected, and demonstrated here, between HeLa (glycogen degrada- 945 
tion>GLUT>HK) and AS-30D (HK> HPI>GLUT) cells; and between 946 
these and normal cells: PFK-1 played a minor role on flux- and 947 
metabolite-control in cancer cells, except for the hypoxia condition. 948 
The main flux-controlling steps are the targets with the higher 949 
potential to diminish the glycolytic flux in the two cancer cells 950 
examined. Although the glycogen degradation pathway was a high 951 
controlling step in HeLa glycolysis, it could not be an appropriate drug 9: 
target because glycogen degradation cannot be sustained by pro- 953 
longed periods of time. 954 
In summary, drug-design studies targeting the most controlling 955 
enzymes or transporters in tumor glycolysis might have more 956 
promissory results than targeting non-controlling enzymes which 957 
usually require more potent and specific inhibitors to accomplish the 958 
required full inhibition to achieving effect on function. Another 959 
940 
941 
2 
93 
944 
    
  
 82 
 
895 
896 
897 
898 
89. 
900 
901 
902 
903 
90' 
905 
906 
907 
908 
OO. 
ARTICLE IN PRESS 
A. Marin.-Ht"f'oondI'Z el aI. I Biochimi a l'I Biophysica Acta 100( (2010) 100(- 100( 11 
300 
MW ~ 
N H > 
K a) o 250 
0= 
-¡¡ '¡;C 
GLUT-1 55 - ~ O e E 200 
C. " • o 
~ o 
'O 0 
-t bu lin 5 - '" 0 
LUT-l 
ig. . lucose sporter r tein ntent fter poxia l ent  eLa ls. ar ala resent l e ean ±  oral sl r e if rent r paration s. , r oxia; , poxia. 
eslem-blol a[ysis as rri d ut r rding lO ce 9). 
odels ay uce e  i  t ay ehavior, ey re  ali  
l r tif i g e r g ets ith e i hest r eutic 
tential   etabolic t ay. hus. WT, K, PI, K-1. PI d 
VK ti ities ere ri    -  t a l co lysis odel, 
ar Llff, K. PI. PI d e gen gradation r ch  t e ...:l 
odel,  t blish  uch se otential r g ets uld e 
i ited r n i g ifi ant cr ent  e l colytic  d 
1l' ncentration. o crease e t ay  r e TP 
centration; ata ot n ) y 0%. t as uired  ibit, 
i i ually, e ain ntro li g s K, PI r llIT y - 8%  
S-  l colysis ig. ), hereas se f eLa l colysis LUT, 
K, I r l gen radation uired - 9% ibiti n ig. B). 
 ntrast, r  ieve  ilar  i inution, e n-contro li g 
ps K-l, K r PI  S-30  l colysis eded  e i ited  
. .  d . %. ectively ig. ). t r stingly.  
A 
100 
80 
'" 60 • ~ 
¡¡: 
40 
20 
O 
B 
100 
80 
"' 60 • 2 
u. 
40 
20 
g~ . 
,. (1) 
,." 
O 
, 
,1e 
. - . - . - . ::: ~ -:;; ;.. .. 
. ~ . 
0 0 0 0 0 
, - - . - . ::....-~-.-.-.. .. 
(d) ~ . ~ · ::--(bl 
~ ' ~&I - - (.) .-. .. 
nzy e ti it   
f"1g. 1 epeooen<r of or glycoIytic flux n e tivity.1 e r nce !()()  rre 
ti ity l es ere se e r nding 10 e ective v  l es r r e r ard 
cti n ables  dS ) hereas  es ere se <licted y ch odel  r r 
- OO:  for eLl Js) . hieh ere eflu es gh H 9 d 0 ol in- 1 g 
l t - 1. s ectively: bles  d ).  . es, sporters  rames ere 
(a) LUf + K + HPI : ) Wf  K ; e) K r 1 ; ) LUf; ) -I ; 1) PYJ(a¡yJ )l I. 
 , ) UJ  + l r gen radabon: ) l r gen egradation: e) LUf: ) K; ) PI 
d 1) 11'1. hen o r ore t ps ere tit <!. enbcal ri ti n inactivitywasapp~ed . 
 crease of the ¡ l e as ro dJ'lied  a r portXlI\a1 crease  e ,. alue. 
i it n f  r TP ncentration ) as t i ed,  S- 0  0 
l colysis. hen W  + HK r   HK + HPI ere cr ased 1 
ult neously  e e agnitude, 3  d  ibiti n, 2 
ectively, hereas  eLa l colysis, hen llff + glycogen 13 
radation ere creased ult neously y  Hg. B). 4 
. oncluding arks 15 
he i etic odels r th orc lls r dicted nably eU 6 
es d etabolite ncentrations d   Us, o i ve 917 
i h us tness. e odels ere uilt p ith e i etic r perties 918 
f ch i i ual e t ined peri enta lly J i g 919 
) e ost on ds bstrates, r ducts d odulators). 20 
d ) e t cti ns f eral t er t ay ediaries ith 21 
e es ig. ). lso, e reat ajority f e nges ade  2 
prove d fi e e odels ere ased  " et" eri entatíon. 923 
his o ed s  n tify e echanis s  hich  it a ly 924 
si ered on-controlling e ch s PI ay ert ifi ant 925 
trol n or l colysis. owever, rt er ds f odel 926 
in ent re ti l cessary  ch ore r dicti n uracy, f127 
ecially r P d P ncentrations, evertheless. e resent 928 
odels ay r e  redict e t ay havior der t er 929 
ysiological nditi ns d  t er cer l s s g s e 930 
ost ntro ling s re eri entally aluated. 931 
hus. i etic odeling   elpful reti l eri ental 932 
r ach  ti f i g e ost r ising r peutic ets. t 3 
as en osed at ibiti n f l colysis  o r ls n e 93-1 
 ati e r peutic y , 11 d eral r peutic 5 
ets ave en osed lUn, KII , FB3, PDH, H. d 6 
Cf) 7, ed  -65 1,  n sults f ical ce r  7 
eri entally a luate hether ch ets re eed  ntrol 8 
ps f e orc ls' t ay. oreover, t ould e sirable at 939 
ch ets ve i h  ntro l  or ls, ut  ntrol  !)..lO 
ost, or al l s. - 1 
 r al ls. K d K-l, ether ith llIT, e rt e ain 9-12 
ntrol  e l colytic  I ed  1. owever, ue  e -13 
i rent e ti ity els, if r nt ntrol i t ti n as -14 
pected, d onstrated ere, t en eLa en grada- -15 
ion " LlIT> HK) d - 0D " I GLlIT) lls; d t een 6 
se d r al ls: K-l l ed  inor le n l - d -17 
etabolite-control  cer l s, cept r e poxia ndition, -18 
he ain - ntro li g s re e ets ith e i her !)..I9 
tential  i inish e l colytic   e o ce r ls 0 
ined lt ugh e gen radation t ay as  i h 1 
ntrolli g p  ela l colysis, t uld ot e  propriate r g 52
et ca use en radation not e t i ed y ro- 3 
ed ri ds f e. 4 
 mary , g-design ies ti g e ost ntro li g 5 
es r orters  or l colysis ight ave ore 6 
i sory sults n eti g n-controlling es hich 7 
aUy uire ore tent d ecific i itors  plísh e 8 
uired l i it n  i i g ct n cti n. nother 9 
Please it  is rti le s:  arí - ernández, t l., odeling cer l colysis, i chi . iophys. cta 10), i: 1O.1016/j. 
abi . 10. 1.006 
 83 
 
960 
961 
962 
963 
964 
965 
966 
967 
968 
960 
970 
971 
972 
973 
974 
Q 4 975 
976 
977 
978 
979 
980 
981 
982 
983 
984 
985 
986 
987 
988 
989 
990 
991 
992 
993 
994 
995 
996 
997 
998 
990 
1000 
1001 
1002 
1003 
10(>l 
1005 
1000 
1007 
1008 
1009 
1010 
1011 
1012 
1013 
1014 
1015 
1016 
1017 
1018 
1019 
1020 
1021 
1022 
1023 
102.1 
1025 
1026 
1027 
1028 
1029 
ARTICLE IN PRESS 
12 A Marfn-Hernóndez et al ! Biochimica et Biophysica Acta xxx ( 2010) xxx-xxx 
strategy would be tha t of a multi- targeted therapy against the mast 
can trolling steps as previously suggested lef. Fig. 3 ; see also 9 ). 
From a therapeutic point ofview, the kinetic models are useful for 
predicting how pathway fluxes and ATP concent ration mayvary when 
one or more steps are inhibited. The results o f the simu lations 
ind icated that only the simultaneous inhibition of the controlling 
steps may have signi fican t impact on glycolyt ic flux and ATP 
concentrat ion. Qther interve ntio ns such as inhibition of solely one 
controlli ng step, or wo~e, of non-con trolli ng steps will bring about 
negligible effects on pathway behavior. For the latte r, their negligible 
flux-con trol coefficien ts demand the design of highly potent and very 
specific inhibitors for the tumor steps or their full gene tra nscription 
or t ranslation blockade. 
Supplementary mate rials related to this article can be found on line 
at doi: 1O.1016/j.bbabio.2010.11.006. 
5. Uncited reference ~ 
Acknowledgements 
This research was supported by CONACyT-México grants No. 
80534 and 123636 to RMS and 83084 to ES. and Ins tituto de Ciencia y 
Tecnologia del Distrito Federal grant No. PICS08-S. AMH was 
supported by a CONACyT feIlowshi p (No. 59991). 
References 
111 R. Moreno·S.íncllez, S. Rodríguez-Enríquez, A M,;¡ñn -Hernández. E. S,),}vedr,;¡. 
Energy met.lbolism in tumor cells, FEBSJ. 274 (2007) 1393-1418. 
121 R. Moreno-Sáncllez. S. Rodñguez-Enñquez. E. Sa.';¡VedlOl. A M,;¡ñn-Hernández.J.C. 
Gall,;¡rdo-l'érez. TIle bioenergetics of cancer: is glycolysis the m,;¡in ATP supplierin 
all tumor cells7 Biofactors 35 (2009) 209-225. 
131 e.V. [)ang. j.W. Kim. P. Gdo, j. Yustein. The interpl,;¡y between MYC,;¡nd HJF in 
cancer. NaL Rev. C,;¡ncer 8 (2008) 51-56. 
14J S.J. Yeung.J. Pan, M.H. Lee, Roles of p53. Myc and HJF-l in regulating glycolysis-
tlle seventh hólJtm,;¡rk of Cdl'lcer. CelL Mol. üfe Sci. 65 (2008 ) 3981- 3999. 
15] A. Mañn- Hernández, j .e. Gallardo-Pérez. SJ. Ralph, S. Rodríguez-Enriquez, R. 
Moreno-Sáncllez. HIF-lo. modul,;¡tes energy metabolism in c,;¡ncer cells by 
induong over-expression of spe6flc glycolytic isoforms, Mini Rev. Med. Chem. 
9 (2009) 1084-1 10 1. 
161 S. Rodriguez-Enríquez. A. M,;¡rfn-Hernández. l e. G,;¡IJardo-Pérez. L C,;¡rrei\o-
Fuentes. R. Moreno-Sánchez, Targeting of cancer energy metaboJism, Mol. Nutr. 
Food Res. 53 (2009) 29-48. 
P] D. Fell, Underst.lnding the Control ofMetabolism. Portl,;¡nd Press. lOndon, 1997. 
18] R. Moreno-Sáncllez. E. $,a,;¡vedra. S. Rodríguez-Enríquez. V. OIín-Sa.ndov,;¡l. 
Met.lbolic control ,;¡na.lysis: ,;¡ tool for designing str,;¡tegies to m,;¡nipul,;¡U' 
pathways, J. Biomed. Biotech. 2008 (2008) 597913. 
191 R. Mort'no-Sáochez. E. Sa.,;¡vedr,;¡, S. Rodñguez-Enñquez. j.e. Gdll,;¡rdo-Pérez. H. 
Quez,;¡d,;¡. H.V. Westerholf. Met,;¡bolic control ,;¡n,;¡Jysis indic,;¡tes ,;¡ ch,;¡nge of 
StlOltegy in the treatment ofcancer. Mitochondrion ~ . 
1101 A. M,;¡ñn -Hernández. S. Rodríguez-Enríquez. PA Vit,;¡l-Conzález. F.L Flores-
Rodñguez, M. Macias-Silva. M. SOSd-Garrocho, R. Moreno-Sáochez. Detennining 
,;¡nd underst,;¡nding tlle cootrol of glycolysis in f,;¡st-growth tumor rells. Flux 
control by,;¡n over-expressed but strongly product-inhibited Ilexokina.se. FEBSJ. 
273 (2006) 1975- 1988. 
111 ] j.L Snoep, The silicon cell initi,;¡tive: working tow,;¡rds ,;¡ det,;¡iled kinetic 
description,;¡ t the cellul,;¡r 1evel. Curro Opin. BiotechnoL 16 (2005 ) 336-343. 
1121 jJ. Hornberg. F.j. Bruggema.n, BM. B.akker. H.Y. WesU'rholf. Melilbolic control 
';¡rldlysis to identify optima.l drug t,;¡rgets, Prog. Drug Res. 64 (2007) 173-189. 
113J E. Sa.avedlOl, R. Encalada. E. Pineda. R.jasso-Chávez. R. Moreno-Sánchez. Glycolysis 
in Entamoeba hiSlolylica. Biocllemic,;¡1 chólracteriz.ation of rerombirldnt glyrolytic 
enzymes and Oux control arldlysis, FEBS J. 272 (2005) 1767-1783. 
1141 E. v,;¡n Schaftingen. B. Lederer. R. &lItrons, H.G. Hers. A kinetic sludy of 
pyrophosphate: rructose-6-phosphólte phosphotransferase from pot.lto tubers. 
Eur. J. Biochem. 129 ( 1982) 191-195. 
1151 H.U. Bergmeyer. Methods of Enzym,;¡tic Ana.lysis. Verlag Chemie. Winlleim, 1983. 
116] E. Saavedra A. Marín-Hemández, R. Encal,;¡d,;¡. A. OJivos. G. Mendoz,;¡-Hemández. 
R. Moreno-Sánchez. Kinetic modeling can describe in vivoglycolysis in Enmmoeba 
his ro/yrica. FEBS J. 274 (2007 ) 4922-4940. 
117] P.e. Heinrich. H.P. Morris, G. Weber, Behavior oftransaldolase and tlOlnsketolase 
,;¡ctivities in norma.l. neoplastic. dilferenti,;¡ting ,;¡nd regenerdting liver. C,;¡ncer Res. 
36 ( 1976) 3189-3197. 
118J S. Rodñguez-Enñquez. P. Vit.ll-Gónzalez. F.L Flores-Rodñguez. A Mañn-Hernán-
del, L Ruiz~R,;¡mírez. R. Moreno-S.ínchez. Control of cellul,;¡r prolifer,;¡ tion by 
modulation of oxid,;¡tive phosphoryl,;¡tion in huma.n ,;¡nd rodent f,;¡st-growing 1030 
tumor ceUs. Toxico!. Appl. F'h,;¡nn,;¡col. 215 (2006 ) 2(11-217. 1031 
119] S. Rodriguez-Enriquez. L C.vreño-Fuentes.1.e. G,;¡II,;¡rdo-l'érez, E. $,a,;¡vedr,;¡. H. 1032 
Quez.ada, A. Vega, A Marín-Hemández, V. OIín-Sa.ndoval, M.E. Torres-Márquez, R. 1033 
Moreno-Sánchez. Oxid,;¡tive phosphoryl,;¡tion is impaired by prolonged hypoxi,;¡ in 1Q3.t 
breast but not incervix carcinoma.. lnt.J. Biochem.Cell Biol. 42 (2010) 1744-1751. 1005 
120J H.U. Bergmeyer, Methods ofEnzymaticAnalysis. Weinheim, VerlagChemie, 1974. 1036 
121] ES. Ból.ginski. P.P. Fol, B. Zólk, Microdetermin,;¡tion of inorg,;¡nic phosph,;¡te. 1037 
phospholipids, ,;¡nd total phosphate in biologic ma.teri,;¡!s. Clin. Chem 12 (1967) 1008 
326-332. 1039 
122] T.P.M. Akerboom, H. Bookelma.n. P.F. Zuurendonk. R. Meer. J.M. T,;¡ger, Intr,;¡ffiÍ- 1().t(1 
tochondrial and extlOlmitochondrial concentrations of adenine nucleotides and 1041 
inorg,;¡nic phosph,;¡U' in isal,;¡U'd Ilepatocytes from fasted ralS. Eur.j. Biocllem. 84 1M2 
(1978) 4 13-420. 1043 
123J P. Mendes. CEPAS I: ,;¡ softw,;¡re package for modeling tlle dyna.mics, ste,;¡dy states 1()4.t 
,;¡nd control of biochemical ,;¡nd other systems, CompuL Appl. Biosci. 9 (1993) 1().l5 
563-57 1. 1046 
124] S. Rodñguez-Enríquez. A Marín-Hemández. j.e G,;¡II,;¡rdo~pfrez. R. Moreno- l().t7 
Sánchez, Kinetics of tlOlnsport and phosphorylation of glucose in cancer cells, 1().t8 
j. Cell. F'hysiol. 221 (2009) 552-559. 1049 
125] I.H. Segel, Emyme Kinetics. Wiley, New York. 1975. 1050 
126J T. Rapoport. R. Heinrich. G. Jacobasch. S. Rapoport, A mathema.tical model of 1051 
glyrolysis ofhuma.n erythrocytes, Eur. j . Biocllem. 42 (1974) 107- 120. 1052 
127] F.B. du Preez. R. Con radie, G.P. Penkler, K. Holm, F.Lj. v,;¡n Dooren, J.L Snoep, A 1053 
compar,;¡tive ,;¡n,;¡lysis of kinetic models of erythrocyte glycolysis. j . Theor. Biol. 252 1054 
(20m) 488-496. 1055 
128J B. Teusink. J. Passarge, CA Reijenga, E. Esgalhóldo, e.e. van der Weijden, M. 1056 
Schepper, M.e. W,;¡lsh. B.M B.akker. K. v,;¡n [)amo H.V. Westerhoff. j.L Snoep. can 1057 
yedst glycolysis be understood in terms of in vitro kinetics of the constituent 1058 
enzymes7 Testing biochemistry. Eur.J. Biochem 267 (2000) 5313-5329. 1059 
129] B.M. Ból.kker. M.e. W,;¡lsh. B.H. Ter Kuile. He. Mensonides. PAM Michels. F.R. 1060 
Opperdoes, H.V. Westerhoff, Contribution of glucose tlOlnsport to the control of 1061 
the glyrolytic nux in Trypanosoma bruce~ Proc. N,;¡tl Acad. sa. USA 96 ( 1999) 1062 
10098-10103. 1063 
130J S. Rodríguez-Enñquez, O. Juárez. JS. Rodríguez-Zólvala, R. Moreno-Sánchez, 1064 
MuJtisite control of the Cr,;¡btree effect in asOtes Ilep,;¡tom,;¡ rells. Eur. j. Biocllem. 1005 
268 (2001) 2512-2519. l()(i(i 
131] R. Wu, Regul,;¡tory mecha.nisms in carbohydr,;¡te metabolism V. Limiting f,;¡ctorsof 1067 
glycolysis in Heló!. cells.J. Biol. Chem. 234 (1959) 2806-2810. 1068 
132J R. Nakashima., M. p,;¡ggl Lj. Scott, P.L Pedersen, PurifKdtion ,;¡nd crnracteriz,;¡tion 1069 
of,;¡ bind,;¡ble ronn ofmitochondri,;¡l bound hexokina.se from tlle highly glycolytic 1070 
A$-30D IOlt hepatoma ceO line, Cancer Res. 48 ( 1988) 913-919. 1071 
133] O.M P,;¡rry. P.L Pedersen, Intracellul,;¡r localizdtion ,;¡nd properties of particul,;¡te 1072 
hexokin,;¡se in the Novikoff,;¡scites tumor.J. BioL Chem. 258 (1983) 10904-10912. 1073 
134J K.K. Arora. P.L Pedersen. Functiooal significance of mitochondrial bound 1074 
Ilexokinase in tumor cell met,;¡bolism,J, Biol. Chem. 263 ( 1988 ) 17422- 17428. 1075 
135] G.E.j. St.l,;¡I, A. Kallf, E.e. Hessbeen, e.W.M. van Veelen, G. Rijksen, Subunit 1076 
composition. regul,;¡tory properties ,;¡nd phosphoryl,;¡tion of phosphofructokina.se 1017 
from hum,;¡n gliom,;¡s. C,;¡ncer Res. 47 (1987) 5047-5051. 1078 
136J Lj. Mandarino, R.L Printz. KA Cusi, P. Kinchington, R.M. O'Ooherty, H. Osawa, e. 1079 
Sewell, A. Consall O.K. Cranner, RA Oefronzo. Regul,;¡tion of Ilexokinase lI,;¡nd lOSO 
glycogen synthase mRNA, protein, ,;¡nd ,;¡ctivity in huma.n muscle, Am. J. Physiol. 1¡J31 
2m (1995) E701-E708. 1082 
137] S. Kuma.g.ai. R. Nar,;¡saki. K. Hasumi, Glucose~dependent ,;¡ctive AlP depletion by 1083 
koningic aod kills high-glycolytic cells, Biochem Biophys. Res. Commun. 365 1084 
(20m) 362-368. 1085 
138] S. Rodñguez-Enñquez. M.E. Torres-Márquez, R. Mort'no-Sánchez, Substrate 1086 
oxidation and AlP supply in AS-300 Ilepatoma. cells, Arch. Biocllem Biophys. 1¡J37 
375 (2000) 21-30. 1()38 
139J Lj. Reitzer, B.M. Wice, O. Kennell, TIle pentose cycle.j. Biol. Chem. 255 (1980) 1089 
5616- 5626. 1090 
140] KA Cuma.,;¡, P. McLe,;¡n. The pentose phosph,;¡te pdthw,;¡y ofglucose met,;¡boIism, 1091 
Biochem.j. 115 (19m) 1009-1029. 1092 
141] j.e Port,;¡is. R. Schuster, M. Merle. P. C,;¡nioni. Met,;¡bolic nux determina.tion in C6 1093 
glioma. cells using carbon-13 distribution upon 11-Be] glucose inrubation, Eur. j. 1 ()9.t 
Biochem. 217 ( 1993) 457-468. 1095 
142] e.M. Met.lllo.J.L Walter. G. Steph,;¡nopoulos. Ev,;¡lu,;¡tion of IlC isotopic tr,;¡cers rOl' 1096 
met.lbolíc nux analysis in mammalian cells, j. Biotechnol. 144 (2009) 167- 174. 1097 
143] j.W. H,;¡ rding. EA PyerilZ. H.P. Morris, H.B. White, Proportiorldl ,;¡ctivities of 1098 
glycerol kina.se and glycerol 3-phosphólte dehydrogenase in IOlt hepatomas, 1099 
Biochem.j. 148 ( 1975) 545-550. 1100 
144] B.W. Agr,;¡norf. AK. H,;¡jr,;¡. The ,;¡cyl dihydroacetone phospha.te pathw,;¡y for 1101 
glycerolipid biosynthesis in mouse liver and Ehrlich ascites tumorceJls. Proc. Natl 1102 
Acad. So. USA 68 (1971 ) 411 -415. 1103 
145] K. Snell. G. Weber. Enzymic imbal,;¡nce in serine met.lbolism in r,;¡t hepatom,;¡s, 1l(J.t 
Biochem.j. 233 (1986) 617-620. 1105 
146] G. P,;¡r,;¡dies, F. C,;¡pu,;¡no. C. P,;¡lombini, T. G,;¡leotti. S. P,;¡pa. Tr,;¡nsport or pyruv,;¡te in 1106 
mitochond ria from different tumor cells, Cancer Res. 43 ( 1983) 5068-5071. 1107 
147] SJ Ralph, S. Rodriguez-Enriquez. 1. Neuzil. R. Moreno-S.íochez. Bioenergetic IIOB 
pathways in rumor mitochondrla as targets for cancer thelOlPy and the importance 1109 
of thl- ROS-induced ,;¡poptDtic trigger, Mol. Aspects Med. 31 (2010) 29-59. 1110 
148] j. ZólJitis, I.T. Oliver. Inhibition of glucose phosphólte isamer,;¡se by met,;¡bolic Illl 
intermediates of fructose, Biochem. j. 102 ( 1967) 753-759. 1112 
149] j.M. Chirgwin, T.F. Pusons. AN. Ernst. Mechólnistic implic.ations of tlle pH 1113 
independence of inhibition of phosphoglucose isamer,;¡se by neutral sug,;¡r 1114 
phosprntes.j . Biol. Chem. 250 ( 1975) 7277-7279. 1115 
Please cite this artide as: A Marín-Hernández. et al., Modeling cance r glyco lysis, Biochim Biophys. Acta (2010), doi: 1O.1016fj. 
bbabio.2010.11.006 
 84 
1116 
111 7 
BIS 
1119 
11 20 
11 21 
11 22 
11 23 
Il :u. 
11 25 
11 26 
11 27 
11 28 
11 29 
11 30 
11 31 
11 32 
1133 
11 34 
11 35 
11 36 
11 37 
113$ 
11 39 
11 ·10 
1141 
1168 
ARTICLE IN PRESS 
A. Marin-Hernámez el al. I Biochimica et Biophysica Arra}()Oi (2010) UX-XX)i 13 
1501 M.K. Gaitonde, E. Murray, V.J. (unningham, ECreet of 6-phosphogluconate on 1591 H. Yun, M. l ee, S.-S. Kim. J. Ha, Glucose deprivalion ¡ncreases mRNA stability of 
phosphoglurose isomerase in rat brain in vitro and in vivo, J. Neurochem. 52 vascular endothelial groWlh factor through activation of AMP-activated protein 
( 1989) 1348-1352. kinase in OU14S prostate carcinoma,J. BioL Chem. 280 (2005) 9963-9972. 
151( F. Heinz, K.O. Kulbe, Glyceraldehyde phosphate dehydrogenase from liver: l. 1"'1 M Natsuizaka, M. Ozasa, S. Dannanin, M. Miyamoto. S. Kondo, S. )(amada, M. 
Isolation and characterization of lhe bovine liver enzyme, Hoppe Sey[ers. 1.. Shindoh, F. Higashino. W. Suhara, H. Koide, K. Aita, K. Nakagawa. T. Kondo, M. 
Physiol. Chem 351 (1970) 249-261. Asaka, F. Okada, M. Kobayashi, Synergistic up-regulation of hexokinase-2, glucose 
1521 S. Palra. S. Ghosh. S. Bera. A RUj, S. Ray, M. Rdy. Molecular characterization of transporters aOO angiogenic factors in pancreatic caneer cells by glucose 
tumor associated glyceraldehyde-3-phosphate dehydrogenase, Biochemistry deprivation 0100 hypoxia, Exp. Cell Res. 313 (2007) 3337-3348. 
(Mose) 74 (2009) 717-727. 161( H. Pelicano, 0.5. Martin, R.H. Xu, P. Huang. Glycolysis inhibition for anticancer 
1531 A. Tanaka, B. Chanee, B. QJistorff, A possible role of inorganic phosphate as a treatment, Oncogene 25 (2000) 4633-4646. 
regulator of oxidative phosphorylation in combined urea synthesis and 1621 P.L Pedersen , S. Mathupala, A Rempel, J.F. Geschwind, Y.H. Ko, MitochoOOrial 
gluconeogenesis in perfused rat liver, J. Biol. Chem. 264 ( 1989) 10034-10040. bound type 11 hexokinase; a key playerin the growth 0100 survival of many caneers 
1541 L Nascimben, J.5. lngwall, B.H. Lorell, 1. Pinz, V. Schultz, K Tomheim, R. Tían, a~ ideal prospect for therapeutic intelVention, Biochim. Biophys. Acta 1555 
Mechanism for increased glycolysis in the hypertrophied rat heart, Hypertension (2 ) 14-20. 
44 (2004) 662-667. 1631 ML Macheda, S. Rogen, j.0. Best, Molecular and cellular regulation of glucose 
1551 B. Teusink, M.e. Walsh, K. van Dam, H.V. Westerhotr, The danger of metabolic transporter (GnIT) proteins in cancer,J. Cell. PhysioL 202 (2005) 654-662. 
pathways with turbo design, TreOOs Biochem. Sci. 23 ( 1998) 162-169. 1641 S.P. Mathupala, CB. Colen, P. Parajuli, A.E. Sloan, Lactate and malignant tumon: a 
1561 Y.K. Kashiwaya. K. Sdlo, N. Tsuchiya, S. Tllomas, DA Fell, R.L Veech, j.V. therapeutic target at the stage of glycolysis, J. Bioenerg. Biomembr. 39 (2007 ) 
Passonneau, Control of g1ucose utilization in working perfused rat heart, J. Biol. 73-77. 
Chem 269 (1994) 25502-25514. 1651 A Yalcin. S. Telang, B. dem, j. Chesney. Regulation of glucose metabolism by 6-
1571 R. Moreno-Sánchez. R. Encalada, A. Marin-HernáOOez, E. Saavedra, Experimental phosphofructo-2-kinase/fructose-2, 6-bisphosphatases, Exp. Mol. Pathol. 86 
validation of metabolic pathway modeling. An iIIustration with glycolytic (2009) 174-179. 
segments from fnramoroa histolyrica, FEBS J. 275 (2008 ) 3454-3469. [661 J.E. Wilson, Isozymes of mammalian hexokinase: structure, subcellular localiza-
(58J S. Aiston, B. Andersen, L Agius, Glucose 6-phosphate regulates hepatic Hon and metabolic function, j. Exp. Biol. 206 (2003) 2049-2057. 
glycogenolysi s through inactivation of phosphorylase. Diabetes 52 (2003 ) 1671 " ImamuFa. +. +aRaloa, JI!, Al lale hiRase i s eeR ~ ymes ff:em FOIl. UelRBds ~ MymB I. 
1333- 1339. 99 (1 98:11 ) 1§9 lb§. 
Please cite this artide as: A. Marín-Hernández. et al., Modeling cancer glycolysis, Biochim. Biophys. Acta (2010), doi:10.1016/j. 
bbabio.201O.11.006 
11 42 
11 43 
11 44 
11 45 
11 ,1(; 
11 47 
11 48 
1149 
11 50 
11 51 
11 52 
11 53 
11 5-1 
11 55 
11 50 
11 57 
11 58 
11 59 
11 60 
11 61 
11 62 
11 63 
11 64 
11 65 
""'" 11 67 
 85 
Supplementary Material 
 
Table S1. Model reactions as written in Gepasi 
  
Enzyme or branch  Reaction 
GLUT Glu out = Glu in 
HK Glu in + ATP = G6P + ADP 
HPI G6P = F6P 
PFK-1 F6P + ATP = FBP + ADP 
ALDO FBP = DHAP + G3P 
TPI DHAP = G3P 
GAPDH NAD + G3P + Pi = 13BPG + NADH 
PGK 1,3BPG + ADP = 3PG + ATP 
PGAM 3PG = 2PG 
ENO 2PG = PEP 
PYK PEP + ADP = Pyr + ATP 
LDH NADH + Pyr = Lac + NAD 
Glycogen synthesis G6P + ATP -> glycogen + ADP + Pi + Pi 
Glycogen degradation  Glycogen + Pi -> G6P 
ATPases ATP -> ADP + Pi 
AK ATP + AMP = ADP + ADP 
DHases NADH = NAD 
PPP G6P -> 6PG 
TK Xy5P + Ery4P -> G3P + F6P 
MPM Pyr + 13ADP + 13Pi -> 13ATP 
PPP, pentose phosphate pathway; MPM, mitochondrial pyruvate metabolism; TK, transketolase.  The 
computer program GEPASI used for building the pathway model only accepts entire figures and therefore 
the oxidative phosphorylation stoichiometry for ATP synthesis from pyruvate oxidation was included as 13 
instead of 12.5. 
 
 
 Table S2. Summary of the kinetic parameter values used in Gepasi for kinetic models of glycolysis 
in AS-30D and HeLa tumor cells*  
Enzyme  AS-30D HeLa  Enzyme  AS-30D HeLa 
GLUT Km GLU 
Keq 
Kmp 
0.52 
a
 
1 
b
 
10 
b
 
9.3 
a 
1 
b
 
10 
b 
 
 PGK Km1,3BPG 
Km 3PG 
Km ADP 
Km ATP 
α 
β 
0.035 
e
 
0.12 
e
 
0.67 
e
 
0.15 
e
 
1 
b
 
1 
b
 
0.079 
e
 
0.13 
e
 
0.04 
e
 
0.27 
e
 
1 
b
 
1 
b
 
HK Km GLU 
Km ATP 
Ki G6P 
KADP 
Keq 
0.18 
c
 
0.99 
c
 
0.02 
d
 
3.5 
b
 
651 
f
 
0.1 
e
 
1.1 
e
 
0.02 
d
 
3.5 
b
 
651 
f 
 
PGAM Km 3PG 
Km 2PG 
0.18 
e
 
0.04 
e
 
0.19 
e
 
0.12 
e
 
HPI Km G6P 
Km F6P 
KiEry4P 
KiFBP 
Ki6PG 
0.9 
e
 
0.07 
e
 
0.0017 
g
 
0.17 
g
 
0.0094 
g
 
0.4 
e
 
0.05 
e
  
0.001
g
 
0.06 
g
 
0.015 
g 
 
ENO Km 2PG 
Km PEP 
0.16 
e
  
0.04 
e
 
0.038 
e
 
0.06 
e 
 
PYK Km PEP 
Km ADP 
Km Pyr 
Km ATP 
Keq 
KaF16BP 
KiATP 
0.4 
e
 
0.3 
e
 
10 
j
 
0.86 
j
 
195172 
k
 
4 x 10
-4 s
 
2.5 
s
 
0.014 
e
 
0.4 
e
 
10 
j
 
0.86 
j
 
195172 
k
 
4 x 10
-4
 
s
 
2.5 
s
 
PFK-I Km F6P 
Km ATP 
Ki ATP 
Ki CIT 
4.6 
h
 
0.048 
h
 
1.75 
h
 
3.9 
h
 
1.0 
h
 
0.021 
h
 
20
 h
 
6.8 
h
 
 86 
*The Vm values were those from Table S3.  Km, Ka and Ki in mM.  
 
a
 taken from [24]. 
b
 arbitrary value 
c
 taken from [10] 
d 
taken from [66] 
e 
Present study 
f
 recalculated from the ΔGº´ = - 16.7 KJ mol
-1
. 
g
 adjusted in the interval described in Table 2. 
h
 Moreno-Sánchez, Marín-Hernández, Encalada & Saavedra, unpublished results. 
i
 recalculated from the ΔGº´ = - 14.2 KJ mol
-1
. 
j
 taken from [15] 
k
 recalculated from the ΔGº´ = - 31.4 KJ mol
-1
. 
m
 adjusted in the interval described in Table 2. 
n 
adjusted 
 p 
values reported for AS-30D in Table 1.
  
 q 
value used for 1 mM glucose condition 
r 
values used for hypoxia condition 
s 
taken from K. Imamura, T. Tanaka, Pyruvate kinase isoenzymes from rat, Methods Enzymol. 90 (1982) 
150–165.  
 
 
 
 
 
Table S3. Enzyme activity re-calculation per mg of total cellular protein 
Ka F26BP 
β 
α 
L 
Keq 
KADP 
KF16BP 
1.8 x10
-4 h 
2.35 
h
 
4.47 
h
 
13 
h
 
247 
i
 
5 
b
 
5 
b
 
8.4 x10
-4 h 
0.98 
h
 
0.32 
h
 
4.1 
h
 
247 
i
 
5 
b
 
5 
b 
 
L 1
b
 1
b
 
LDH Km Pyr 
Km Lac 
Km NADH 
Km NAD 
α 
β 
 
0.13 
e
 
4.7 
e
 
0.002 
e
 
0.07 
e 
1
b
 
1
b
 
0.3 
e
 
4.7 
p
 
0.002 
p
 
0.07 
e, p 
1
b
 
1
b
 
Glycogen 
degradation 
v= 1.2x10 
-3 m 
 
6x10 
-3 m 
 
 
Glycogen 
synthesis 
 
v= 2.2x10
-3 m 
1 x 10
-3 q
 
1.1x10 
-3 m 
 
ATPases k= 4.2x10
-3 n
 0.003 
n 
0.0045
 r
 
ALDO Km F16BP 
Km G3P 
Km DHAP 
0.01 
e
 
0.16 
e
 
0.08 
e
 
0.009 
e
 
0.16 
p
 
0.08 
p 
 
AK k1= 
k2= 
1 
n
 
2.26 
n
 
1 
n
 
2.26 
n 
 
DHases k1= 
k2= 
250 
n
 
1 
n
 
250 
n
 
1 
n
 TPI Km DHAP 
Km G3P 
1.9 
e
 
0.41 
e
 
1.6 
e
 
0.51 
e 
 
GAPDH Km G3P 
Km 1,3BPG 
Km NAD+ 
Km NADH 
Km Pi 
 
 
0.29 
e
 
0.02 
e
 
0.08 
e
 
0.004 
e
 
11
e
 
 
0.19 
e
 
0.022 
e
 
0.09 
e
 
0.01 
e
 
29 
e
 
 
 PPP v= 9.5x10
-5 m
 9.5 x 10
-5 m
 
MPM v= 5 x 10
-4 m
 1 x 10
-4 m
 
TK v= 9.5x10
-5 n
 9.5x 10
-5 n
 
 87 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Values from 
a 
[10]; 
b 
Present study (Table 1); 
c 
activity in the absence of activators at pH 7.5 (Moreno-
Sánchez, Marín-Hernández, Encalada & Saavedra, unpublished results) and 
d
 [24]. 
e
 By determining the 
Vm value in the indicated direction in the total cellular extract (fourth column), the Vm value in the other 
Enzyme  
activity 
U (mg cytosolic 
protein) 
-1
 
U (mg total 
 cellular protein) 
-1
 
Vm value used  
in the model 
  AS-30D 
HK Vmf 0.46 
a
 0.35 ± 0.17 (3) 0.35 
HPI 
 
Vmf 
Vmr 
4.9 
b
 
3.4 
b
  
NM 
0. 98 ± 0. 46 (3) 
1.42 
e
 
0.98 
PFK-1 Vmf 0.273 
c
 NM 0.066
f
 
ALDO 
Vmf      
Vmr 
0.23 
a
 
0.18 
b
 
0.056 (1) 
NM 
0.056 
0.044 
e
 
TPI 
Vmf      
Vmr 
5.6 
b
 
56 
a
 
NM 
13.9 (1) 
1.39 
e
 
13.9 
GAPDH 
Vmf      
Vmr 
1 
a
 
0.9
a
 
NM 
0.34 (1) 
0.38 
e
 
0.34 
PGK 
 
Vmf 
Vmr 
27 
a
 
4.3 
b
 
NM 
NM 
10.8 
g
 
1.72 
g
 
PGAM 
 
Vmf 
Vmr 
20 
a
 
1.3 
b
 
NM 
NM 
8 
g
 
0.52 
g
 
ENO 
 
Vmf 
Vmr 
0.51
a
 
0.74 
b
 
0.29 (1) 
NM 
0.29 
0.42 
e
 
PYK Vmf 6.6 
a
 4.2 (2) 4.2 
LDH 
Vmf     
Vmr 
13.4 
b
 
1.8 
b
 
2.0 (2) 
NM 
2.0 
0.27 
e
 
  HeLa 
HK Vmf 0.06 
d
 0.04 ± 0.01 (3) 0.04 
HPI 
 
Vmf 
Vmr 
1.2 
b
 
2.8 
b
 
NM 
0.9 ± 0.1 (3) 
0.4 
e
 
0.9 
PFK-1 Vmf 0.078 
c
 NM 0.031
f
 
ALDO 
Vmf      
Vmr 
0.2 
a
 
NM 
0.08 ± 0.02 (3) 
NM 
0.08 
0.063 
i
  
TPI 
Vmf      
Vmr 
5 
b
 
42 
a
 
NM 
NM 
3.4 
h
 
28 
h
 
GAPDH 
Vmf      
Vmr 
2 
a
 
2.5 
a
 
0.58 
NM 
0.58  
0.72 
e
 
PGK  
 
Vmf 
Vmr 
13 
a
 
3.8 
b
 
NM 
NM 
8.7 
h
 
2.5 
h
 
PGAM  
 
Vmf 
Vmr 
1.4 
a
 
0.53 
b
 
NM 
NM 
0.94 
h
 
0.36 
h
 
ENO 
 
Vmf 
Vmr 
0.36 
a
 
0.4 
b
 
0.34 ± 0.07 (3) 
NM 
0.34 
0.38 
e
 
PYK Vmf 3 
a
 1.9 ± 0.4 (3) 1.9 
LDH 
Vmf     
Vmr 
11.4 
b
 
NM 
3.4 ± 0.5 (3) 
NM 
3.4 
0.54 
i
 
 88 
direction (fifth column) was calculated from the Vmf/Vmr ratio determined in the cytosolic protein (first 
column).  
f 
The PFK-1 Vm was estimated from the ratio Vm cytosolic protein/Vm in total protein of the Vmf 
of ALDO.  
g
 The activities were estimated from the average ratio Vm cytosolic/ Vm total protein of the 
GAPDH and ENO in AS-30D cells and 
h 
ALDO and ENO for HeLa cells.  
i
.
 
These activities were based on 
the ratio Vmf/Vmr of ALDO and LDH in the cytosolic fraction of AS-30D. For the pathway modeling under 
hypoxia the Vm values used for GLUT, HK and HPI were 0.043, 0.055 and 0.52/1.17 U* mg of cellular 
protein, respectively. 
 
 
Table S4. Fixed metabolite concentrations used in Gepasi for 
glycolysis kinetic models 
 
 
 
Values (in mM) were taken from 
a
[10], 
b
 Table 2 , 
c
 Tables 3 and 4, 
d 
 For modeling, free Pi 
was used instead of total Pi: 53% of the total Pi concentration determined in AS-30D 
(Table 3) and HeLa cells (Table 4) was considered as free Pi; 
e
 [39] and 
f
 adjusted. 
 
Figure S1. Simulation of PFK-1 activity in the 
presence and absence of modulators 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
The values for simulation were taken from Table 3, Table S2 and Table S3.  Simulation of the PFK-1 
equation (see section 2.6 rate equations) was performed in the OriginPro 7.5 software with the following 
concentration values: 5.6 mM ATP + 1.7 mM citrate (+ inhibitors); 6 µM F26BP (+ activator).  
Metabolites AS-30D HeLa 
Gluout 5 5 
Lac 27 
a
 33 
a
 
Glycogen 26 
b
 135 
b
 
6PG 0.35 
b
 0.39 
b
 
Citrate 1.7 
a
 1.7 
a
 
F26BP 0.006 
c
 0.0043 
c
 
Pi 2.5 
d
 4 
d
 
Ery4P 1 
b
 0.016 
e
 
Xy5P 0.54 
f
 0.016 
e
 
0 20 40 60 80 100
0.00
0.03
0.06
0.09
0.12
V
F6P (mM)
+ activator
(+ modulators)
No modulators
+ Inhibitors
 89 
Table S5. Effect of some metabolites on the activity of 
HPI, ALDO, ENO and GAPDH 
 
Metabolite Concentration (mM) HPI 
a
 ALDO 
b
 ENO
c
 GAPDH
d
 
  Remaining activity (%) 
AMP  5 100 90 * * 
ATP 5 + MgCl2 5mM 101 103 87 53 
PPi  5 + MgCl2 5mM 87 73 * * 
Pi  5 + MgCl2 5mM 98 96 * * 
G6P   5 * 98 * * 
F6P 5 * 103 * * 
FBP 5 35 * * * 
F26BP 0.006 
e
 86 96 * * 
6PG 0.093 56 * * * 
Ery4P 0.055 16 84 * * 
Citrate  2 101 * * * 
DHAP 2.2 93 * * * 
3 * * 85 * 
G1P 1
f
 74 * * * 
L-lactate 25 95 * * * 
 
Inhibitions were assayed in 50 mM Mops pH 7 at 37
o
 C. 
a
 F6P concentration was 1.5 mM. 
b
 FBP 
concentration was 0.2 mM. 
c
 2PG concentration was 1 mM. 
d
 G3P and NAD
+
 concentrations were 1.5 mM 
and 1 mM, respectively. 
e 
physiological concentration. 
f
 12 times the physiological concentration. * not 
measured. 
 
Table S6.  Inhibition of HPI from different biological sources  
Cell type Ki (μM) 
FBP 6PG Ery4P 
Forward reaction 
AS-30D NM 12 (2) 2.5 ± 0.8 (3) 
HeLa NM 17.6  1± 0.07 (3) 
hepatocytes NM NM 0.9 
Reverse reaction 
AS-30D 170 ± 110 (3) 6.8 ± 1.5 (3) 0.86 ± 0.3 (3) 
HeLa 60  15.5 ± 4 (3) 0.79  
hepatocytes 210 ± 190 (3) 8  1.2  
yeast * NM 47  6.7  
Entamoeba 
histolytica** 
NM 15 5.9  
The values represent the mean ± SD with the number of different preparations assayed in parentheses; 
the absence of parenthesis indicates one preparation assayed.  NM, not measured.  * commercial 
enzyme; ** recombinant purified enzyme [13].  Ki values were determined by nonlinear regression 
analysis of the experimental data, according to the competitive inhibition equation, by using the computer 
program Microcal Origin 5.0.  
 
 90 
Figure S2.  Hexose phosphate isomerase (HPI) inhibition by Ery4P  
 
 
 
 
 
 
 
 
 
The concentration of Ery4P were 0 (■), 13.8 (●) and 28 (▲) μM.  The inset shows a re-plot of the slopes 
obtained from a Dixon plot ([inhibitor] against 1/v at different substrate concentrations).  The linear fitting, 
which intersects the origin, indicates that Ery4P is an HPI competitive inhibitor [25].  
 
 
 
 
Figure S3. GAPDH affinity by Pi 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
GAPDH activity in AS-30D cytosolic extracts was determined in 50 mM Mops pH 7 at 37
o
C and in the 
presence of 1 mM NAD
+
, 2.9 mM G3P and 5 mM cysteine.  The experimental points were fitted to a 
Michaelis-Menten equation by using the software  Microcal Origin v 5. 
   
0 2 4 6 8 10
0
1000
2000
3000
H
P
I a
ct
iv
it
y 
(n
m
o
l/m
in
*m
g
)
F6P (mM)
0 5 10 15 20
0.0
0.2
0.4
0.6
sl
o
p
e
1/[F6P]
H
P
I a
ct
iv
it
y 
(n
m
o
l/m
in
*m
g
)
sl
o
p
e
0 20 40 60 80 100
0
500
1000
1500
Vm 1800  ± 105
Km 10 ± 2G
A
P
D
H
 a
ct
iv
it
y 
(n
m
o
l/m
in
*m
g
)
Pi (mM)
 91 
Table S7.  Robustness for Pi concentration values in the AS-30D  
glycolysis kinetic model 
 Pi (mM) 
Metabolite (mM) 2.5* 1.3 2 2.8 
Glu in 3.4 3.5 3.4 3.3 
G6P 6.5 7.3 6.6 6.4 
F6P 0.03 0.02 0.02 0.02 
FBP 5.2 35 9.7 3.8 
DHAP 14 39 19.4 11.8 
G3P 0.3 0.8 0.41 0.25 
1,3BPG 0.01 0.008 0.01 0.01 
Pyr 0.84 0.82 0.83 0.84 
ATP 7.9 7.2 7.8 8 
ADP 2.1 2.4 2.2 2.1 
AMP 1.3 1.7 1.4 1.3 
NADH 0.005 0.005 0.005 0.005 
NAD
+
 1.34 1.34 1.34 1.34 
 
glycolytic flux 29 26 29 29 
Flux control coefficients     
GLUT 0.2 0.11 0.18 0.2 
HK 0.44 0.27 0.4 0.45 
HPI 0.4 0.25 0.37 0.41 
PFK-1 0.02 0.01 0.02 0.02 
GAPDH 0.05 0.33 0.1 0.04 
*Pi concentration used in the main model.  
 
 
 
 92 
Table S8. Elasticity coefficients for substrates (S), products (P) and modulators (M) 
 εEi
S  εEi
P  εEi
M 
  AS-30D HeLa   AS-30D HeLa   AS-30D HeLa 
GLUT Glu out 2.2 0.8  Glu in -2.1 -0.18     
HK Glu in 
ATP 
0.95 
0.17 
0.85 
0.18 
 G6P 
ADP 
-0.94 
-0.06 
-0.82 
-0.07 
    
 
HPI G6P 
 
1.0 
 
1.4 
 
 
F6P 
 
-0.04 
 
-0.42 
 
 FBP 
6PG 
Ery4P 
-0.05 
-0.06 
-0.89 
-0.05 
-0.55 
-0.34 
 
PFK-1 F6P 
ATP 
 
0.65 
0.71 
 
0.66 
0.04 
 
 FBP 
ADP 
 
-0.004 
-0.003 
 
-4 x 10
-5
 
-3 x 10
-5
  
 
 F26BP 
CIT 
0.001 
-1 x 10
-5
 
0.003 
-1.8 x 10
-8
 
ALDO FBP 
 
1.4 
 
2.1 
 
 DHAP 
G3P 
-1.4 
-1.2 
-2.1 
-1.6 
    
TPI DHAP 75 177  G3P -75 -177     
GAPDH G3P 
NAD
+
 
Pi 
0.70 
0.29 
1.1 
0.74 
0.1 
0.99 
 
13BPG 
NADH 
-0.23 
-0.26 
-0.03 
-0.06 
    
PGK 13BPG 
ADP 
5.7 
5.7 
3.0 
2.3 
 3PG 
ATP 
-4.8 
-5.6 
-2.0 
-2.3 
    
PGAM 3PG 8.9 1.3  2PG -8.4 -0.38     
ENO 2PG 1.5 1.0  PEP -0.75 -0.06     
PYK PEP 
ADP 
0.99 
0.16 
0.98 
0.16 
 Pyr 
ATP 
-8 x 10
-5
 
-9 x 10
-5
 
-4 x 10
-5
 
-9 x 10
-5
 
 FBP  2 x 10
-12
 
LDH Pyr 
NADH 
7.1 
7.1 
19.6 
19.6 
 Lac 
NAD
+
 
-7.0 
-7.0 
-19.5 
-19.5 
    
ATPases ATP 1 1         
DHases NADH 2690 13447  NAD
+
 -2689 -13446     
 
 93 
 
Table S9.  Concentration control coefficients 
 Model 
 AS-30D  HeLa 
 G6P F6P PEP ATP NADH  G6P F6P PEP ATP NADH 
GLUT 0.25 0.12 0.24 0.18 *  0.5 0.57 0.47 0.3 * 
HK 0.54 0.26 0.52 0.39 *  0.11 0.12 0.1 0.06 * 
HPI -0.48 0.24 0.48 0.36 *  -0.66 0.07 0.06 0.04 * 
PFK-1 -0.03 -1.5 0.03 0.02 *  -0.42 -1.4 0.04 0.02 * 
ALDO -0.02 * 0.02 0.01 *  -0.02 * * * * 
GAPDH -0.06 -0.01 0.06 0.04 *  -0.08 0.01 0.01 * * 
ENO * * 0.01 * *  * * * * * 
PYK * * -1 * *  * * -1 * * 
            
Glycogen degradation 0.07 -0.01 0.07 0.08 *  0.7 0.82 0.74 0.69 * 
ATPases  -0.15 1.0 -0.35 -1.1 *  0.02 0.03 -0.26 -1.0 * 
PPP * * -0.01 * *  -0.01 -0.01 -0.01 -0.01 * 
Glycogen synthesis -0.13 0.08 -0.15 -0.22 *  -0.13 -0.15 -0.15 -0.17 * 
MPM 0.03 -0.2 0.07 0.21 *  * * 0.01 0.05 * 
TK * * 0.01 0.01 *  * 0.01 0.02 0.02 * 
 * The values were lower than 0.01.  The respective TPI, PGK, PGAM, LDH, AK and DHases elasticity coefficients were lower than 0.01 and hence they were 
not included in the table.  
 
 
 94 
 
----- Original Message ----- From: "BBA - Bioenergetics (ELS)" <bbabio@elsevier.com> 
To: <morenosanchez@hotmail.com>; <rafael.moreno@cardiologia.org.mx> 
Sent: Tuesday, September 21, 2010 5:20 PM 
Subject: BBABIO-10-99: Decision 
 
Manuscript No.: BBABIO-10-99 
 Title: Modeling cancer glycolysis 
 Article Type: Special Issue:  Bioenergetics of Cancer 
 BBA Section: BBA - BBA - Bioenergetics 
 Corresponding Author: Dr. Rafael Moreno-Sanchez 
 All Authors: Alvaro Marín-Hernández, Ph. D.; Juan  C Gallardo-Pérez, Ph. D.; Sara Rodríguez-Enríquez, 
Ph. D.; Rusely Encalada, B.S.; Rafael Moreno-Sanchez; Emma Saavedra, Ph. D. 
 Submit Date: Jun 04, 2010 
  
 Dear Dr. Moreno-Sanchez: 
  
 On behalf of the Executive Editors of Biochimica et Biophysica Acta, I would like to thank you for 
submitting the above-named article to BBA - Bioenergetics. 
  
 I am pleased to inform you that your paper may become acceptable for publication. The reviewers, whose 
comments are enclosed below, indicate that the manuscript is a valuable contribution to the field and 
recommend publication after modifications. However, before returning your revised paper, you should 
answer the points raised by the reviewers. We look forward to receiving your prompt response. 
 
Thank you, and we look forward to receiving your revised manuscript. 
  
 Yours sincerely, 
  
 Rodrigue Rossignol, PHD (Guest Editor) 
 BBA - Bioenergetics 
  
 Please see below Reviewer(s') comments: 
   
 Reviewer #1: Cancer glycolysis is a subject of interest as a better knowledge of altered cancer 
metabolism could help to identify new drug targets. The work presented in this paper is a continuation of 
the previous work and propose  possible therapeutic targets to inhibit tumour energy metabolism. The 
paper is suitable for publication in BBA provided that the authors take into account the following points: 
  
 - The HeLa cells are cultured at 25 mM and experiments are performed at 5 mM glucose. The authors 
discuss that the high flux of glycogen degradation observed in these cells can be influenced by this fact. 
The authors not discuss if the  high control coefficients observed for glycogen degradation in HeLa cells 
could be influenced also by this change in glucose concentration. A discussion on how the change of 
glucose concentration can result in the high control coefficients observed for glycogen need to be 
included. 
  
 - The authors use an equation "random bi-substrate Michaelis-Menten" to describe the kientics of 
pyruvate kinase (PK). It has been well described that PK is an allosteric enzyme and that in tumors an 
specific isoform is overexpressed. The regulation of this enzyme by dimer-tetramer formation has also 
been reported in several papers (see:  Robust metabolic adaptation underlying tumor progression. Vizán 
P., Mazurek S., Cascante M. Metabolomics (2008) 4:1-12.  and also Mazurek S. Ernst Schering Found 
Symp Proc. 2007;(4):99-124.). Inclusion of this regulatory mechanism in the kinetic equation or a  
justification of why this mechanism of regulation has not been taken into account is necessary. 
  
 - Page 15: the authors comment that: "The low PPP flux and intermediary concentrations compared to 
 
 
 95 
glycolytic flux and metabolites suggested low activity of the PPP indicating that tumor cells were not in 
proliferative stage".  The authors compute the PPP from changes in 6PG which can result in an 
underestimation of the flux. It should be discussed the error associated with this method as well as why 
tracers as 14C or 13C labelled glucose has not been used to verify that this flux is so low. 
  
 -  Page 18: The authors say that: "ALDO activity was not modulated by any of these metabolites. 
Although GAPDH was inhibited by ATP (with non-competitive inhibition, Ki = 4.3 mM) and ENO was 
inhibited by DHAP and ATP (non-competitive inhibition, Ki = 31 mM, and uncompetitive inhibition, Ki = 14 
mM, respectively), these Ki values were far from the physiological concentrations, and their effects were 
not included in the rate-equations. However, GAPDH from AS-30D cells has low affinity for Pi (Km =10 
mM) (Fig. S3) and the Pi concentration in the cells is 0.6 mM [23], suggesting low GAPDH activity. By 
decreasing the GAPDH activity to 15% of the Vmf and Vmr values showed in Table S3, because Pi was 
not included in the rate equation, accumulation of FBP was promoted reaching a near-physiological value 
of 11.4 mM (Table 3)". It should be justified why Pi is not included in the rate equation of GAPDH if the 
concentrations are available. 
   
 Page 24:  The authors say that : "Increased enzyme activities under hypoxia for HK and HPI were also 
determined (37% and 30%, respectively) whereas a 225% increased protein content for GLUT was 
determined by western-blot analysis and then a corresponding increase in transport activity was assumed 
for modeling (data not shown)". Western-blot must be included in the paper instead of mention as data not 
shown. 
  
 - Figure 1: The positive and negative metabolite regulatory loops considered in the kinetic equations must 
be included in the figure. 
  
 - Pyruvate dehidrogenase reaction has not been included in the model. A justification for this fact need to 
be provided. 
  
 - HPI is a reversible enzyme described in general as in rapid equilibrium and Vmf and Vmf and Vmr  
arround 200 times higher than Vmf of HK have been described in muscle (Puigjaner J, et al.  M. FEBS 
Lett. 1997 Nov 24;418(1-2):47-52.). Vmf and Vmr are difficult to determine due to the reversibility of 
enzyme and the high rates in both directions of this reaction. It is sourprising that here Vmr are  the same 
order of magnitude for both enzymes in AS-30D.  In HeLa the differences are more notorious. It will be 
interesting to verify in the model how much dependent is the control coefficient of HPI on the magnitudes 
of Vmf and Vmr for HPI (i.e. increase in an order of magnitude both Vmf and Vmi for HPI and see how 
much this fact affect the control coefficient). 
   
 Reviewer #2: Review  BBABIO-10-99 
   
 The authors submitted an important paper dealing with the application of metabolic control theory on 
glycolysis of tumor cells. They measured in two tumor cells enzyme kinetics and steady state 
concentrations as well as fluxes of all relevant glycolytic enzymes put these data into a computer model 
calculated flux control and concentration control steps. This is a large and interesting work. Authors were 
able to detect the most limiting steps which could be used as targets for cancer therapy, also the 
approach to detect the extent of single enzyme inhibition which could the reduce glycolytic ATP formation 
by 50 %. They compared these new data with data from non tumor cells obtained from others using not 
such an elegant but laborious approach. I agree with the paper. However, I have a little problems with the 
concept of the paper. It would be more interesting to compare tumor data with data obtained with the 
same method from non tumor cells. 
 Moreover we know, that in tumors also the interaction between mitochondria and glycolysis is impaired. 
The tumor metabolism is also characterized by impaired mitochondria. Limitation on the glycolysis only, is 
therefore a limitation of possibilities. I know that is a large bigger task but the problem is known since 
more than 40 years ago. (The reviewer was to that time involved in the task to bring together the glycolytic 
and mitochondrial pathway of reticulocytes). Therefore I think it should at least discussed in the present 
paper, why the authors concentrated on the glycolysis only and only on two tumor cell lines. 
 
 
 
 96 
 
Rodrigue Rossignol, Ph. D.                                                                                            November 1, 2010 
BBA-Bioenergetics 
Guest Editor 
Dear Dr. Rossignol, 
 
Thank you for your letter of September 21, 2010 regarding the submission of the manuscript (No. 
BBABIO-10-99) entitled “Modeling cancer glycolysis”.  We were encouraged to resubmit a revised paper, 
in which the observations made by the two reviewers should be answered.  Therefore, several 
modifications were made throughout the manuscript most of which required further experimentation and in 
silico modeling.  We would like to also thank the reviewers, particularly reviewer 1, for their insightful 
observations and suggestions, which have enabled us to strengthen the paper on several points (enzyme 
kinetics, pathway modeling, and paper focus).  The new experimental data, included in the revised 
manuscript, added further support to our findings and conclusions.  The changes made to the manuscript 
were indicated in bold letter.  A point-by-point response to the observations raised by the reviewers is 
shown below.  Trusting that the manuscript will now be acceptable for publication in BBA-Bioenergetics 
special issue on Bioenergetics of Cancer, we remain. 
 
Sincerely yours, 
                                       Emma Saavedra, Ph.D.  Rafael Moreno-Sánchez, Ph. D. 
                                                               Corresponding authors 
Response to Reviewer 1 
1.  Discussion on how the change in glucose concentration may affect the contribution of glycogen 
degradation to the control of glycolysis was included, as suggested, on p. 25 2
nd
 paragraph.  
 
2.  Full PYK kinetic description. 
The PYK rate equation was changed to that of a cooperative enzyme regarding [PEP], [F1,6BP] and 
[ATP] (the Monod et al. equation for exclusive binding; p. 14).  However, as expected from saturation with 
allosteric activator (KA(F1,6BP) 
presence of a relatively high concentration of an allosteric inhibitor (Ki(ATP) = 2.5 mM in Table S2; [ATP] = 
7.9 mM in Table 3), no changes on PYK flux control, and on flux control distribution and metabolite 
concentrations were attained.  Under maximal activation (e.g., all enzyme molecules are in the active 
tetrameric conformation), an enzyme with sigmoidal kinetics shifts to hyperbolic behavior and therefore a 
simplified Michaelis-Menten equation can be used.  However, for the sake of accuracy, the more complex 
Monod equation for PYK (instead of the previous Michaelis-Menten equation) was included in the pathway 
models, as suggested by the reviewer.   
 
3.  Pentose Phosphate Pathway flux. 
In agreement with the reviewer’s observation, we have re-written the section describing and discussing 
the results of the PPP flux (p. 17, 1
st
 paragraph).  In there, we have (i) recognized the possibility of 
underestimating the PPP flux when evaluated from changes in the 6PG concentration and (ii) discussed 
an improvement in the PPP flux determination by using radio-labelled glucose, as suggested.   
Moreover, the robustness of the pathway models was tested for variation in the PPP flux to 10-20 times 
the original value (and assuming that our values were underestimated).  However, no significant changes 
on fluxes or control coefficients were attained, and only a 20-40% decrease in FBP and DHAP 
concentrations was observed.  These observations were concisely summarized on p. 24, 2
nd
 paragraph.  
Thus, it seems that under the conditions used in the present study, the PPP flux is indeed low.  
 
4.  Inclusion of Pi in the GAPDH rate equation. 
The GAPDH rate-equation was modified to a rapid equilibrium ordered Ter-Bi reversible Michaelis-Menten 
equation (p. 13) now including Pi.  The inclusion of this metabolite prompted an experimental re-
evaluation of the GAPDH affinity for Pi (KmPi) and the determination of the total Pi intracellular content 
under the conditions of the present study; the description of the method used for determination of Pi was 
now added on p. 8.  These new data confirmed the high GAPDH KmPi (Table 1), described in the previous 
version of the manuscript, and showed a high total Pi level, even in the presence of glucose in both 
cancer cell lines (p. 20, 3
th
 paragraph; Tables 3 and 4).  For pathway modeling, free Pi was used which 
 
 
 97 
was stated in the same paragraph and in the legends of Tables 3 and 4. The effect of varying the Pi 
concentration on the GAPDH control coefficient and FBP concentration was modeled and the results were 
included in Table S7 in supplementary material and discussed on section 4.4 of the main text.  
 
5.  GLUT content under hypoxia. 
As suggested, the Western-blot analysis for the glucose transporter under normoxic and hypoxic HeLa 
cells was included in the revised paper as Figure 2 and mentioned on p. 27, 1
st
 paragraph. 
 
6.  Figure 1 was modified, as suggested, including the positive and negative metabolite regulatory loops 
and new substrates in the modified rate-equations. 
 
7.  PDH was not included in the model. 
The rate-reaction for this step was not specifically defined in the model.  However, this enzyme was 
implicitly included as part of mitochondrial pyruvate oxidation, which was experimentally evaluated as the 
steady-state rate of O2 consumption.  A statement in this regard was added on p. 9, last four lines.   
 
8.  HPI Vmf and Vmr values.   
HPI certainly catalyzes a near-equilibrium reaction under physiological conditions but its kinetic 
parameters are not difficult to determine using the appropriate conditions.  In our assays we made sure to 
be under initial velocity conditions by using an excess of the coupling system (i.e., G6PDH for the reverse 
reaction; recombinant PFK from Entamoeba histolytica
forward reaction) and making rate-limiting the HPI activity by adjusting the amount of cellular extract. 
Regarding the magnitude of the HPI Vmf and Vmr, and HK Vm, values in AS-30D cells: In muscle cells 
and in general in non-tumor cells, indeed very high HPI Vmf (or Vmr)/HK Vm ratios are commonly found.  
This activity ratio is still high in tumor cells, but lower than that found in non-tumor cells, as a consequence 
of the enhanced over-expression of all glycolytic enzymes, particularly HK, in tumor cells [see 10, 32, 40].  
On the other hand, similar HPI Vmf and Vmr values (Table 1, AS-30D) have been also documented in 
other biological systems (see 40, 50, Blood Cells Mol. Dis. 1998, 24: 54-61). To address the reviewer 
concern regarding HPI Vm values, the AS-30D model was subjected to variations in this enzyme activity.  
Increasing HPI activity from 0.5 to 10 times (i.e., simultaneously increasing both Vmf and Vmr) brought 
about a proportional decrease in the HPI flux control coefficient and increase in flux and the FBP and 
DHAP levels (see the Table S10 included in this letter but not in the paper).  These data indicated that the 
model was not so robust when varying HPI, suggesting that HPI high activity has to be tightly modulated.  
However, this observation had already been noted in the previous version of the manuscript (section 4.8). 
Table S10. Model Robustness regarding HPI Vm values. 
Variation Fold 1X 0.5X 2X 10X 
Vmf 1.42 0.71 2.84 14.2 
Vmr 0.98 
0.49 1.96 9.8 
Metabolite 
    Glu in 3.4 3.8 3 2.7 
G6P 6.5 9 4.9 4 
F6P 0.03 0.02 0.03 0.03 
FBP 5.2 1.3 26 374 
DHAP 14 7.3 30 109 
G3P 0.3 0.16 0.6 2.3 
1,3BPG 0.01 0.005 0.03 0.15 
Pyr 0.84 0.86 0.86 0.89 
ATP 7.9 6 9.7 10.8 
ADP 2.1 2.6 1.3 4.2 
AMP 1.3 2.6 3.8 0.04 
glycolytic flux 29 
21.5 36.7 41.7 
Flux control coefficients 
 
 
 98 
GLUT 0.2 
0.16 0.15 0.02 
HK 0.44 0.51 0.25 0.03 
HPI 0.4 0.49 0.22 0.02 
Glycogen degradation 0.05 0.07 0.02 -0.02 
PFK-1 0.02 0.02 0.02 0.003 
ATPases -0.1 -0.15 -0.18 0.88 
Response to Reviewer 2 
-  “It would be more interesting to compare tumor data with data obtained with the same method from non-
tumor cells”. 
We have now extended the comparison of the present results (cf. Table 5) with data from glycolysis in 
non-tumor cells, in which MCA was applied (p. 21, last 3 lines of 2
nd
 paragraph; p. 14, 2
nd
 paragraph from 
bottom; p. 22, lines 6-9; p. 22, last 4 lines of 2
nd
 paragraph; p. 28, 3
th
  paragraph).  The main conclusion 
drawn from this analysis was that HPI was a more specific target for AS-30D cancer cells than GLUT or 
HK. 
-  “…in tumors the interaction between mitochondria and glycolysis is impaired.  The tumor metabolism is 
also characterized by impaired mitochondria”. 
The second part of the Warburg hypothesis, that mitochondria in cancer cells are impaired (Science 1956, 
123: 309-314), is by no means an established and universal fact, and is still a matter of controversy.  
Undoubtedly, the mitochondrial function is damaged or diminished in many tumors, whereas in some 
others it is well-documented that the mitochondrial function is not impaired and actually may surpass the 
activity found in non-tumor cells (see references 1, 2, 47 for reviews).  Therefore, impaired mitochondria 
do not represent a generalized finding within the different cancers.  Indeed, in recent studies with different 
cancer cell models, oxidative phosphorylation is active and may predominate for ATP supply (Toxicol Appl 
Pharmacol 2006, 215:208; Amino Acids 2008, 35:681; J Cell Physiol 2008, 216:189; BBA-Bioenergetics 
2009, 1787: 1433-1443; Int J Biochem Cell Biol 2010, 42:1744; J Bioenerg Biomembr 2010 42:55 ). 
Regarding the impaired interaction between mitochondria and glycolysis, the reviewer is probably referring 
to the lower pyruvate dehydrogenase complex (PDH) activity found in some tumor cells and to its 
inactivation by a hypoxia-upregulated pyruvate dehydrogenase kinase (PDK) (see reference 5 for review).  
However, the proposal that PDH inactivation by PDK is the mechanism by which mitochondria become 
impaired in cancer cells incorrectly assumes that pyruvate is the main oxidizable substrate in cancer cells 
and that PDH is the rate-limiting step of the Krebs cycle and oxidative phosphorylation.  Moreover, full 
PDH inactivation has not been observed.  On the other hand, the aspartate/malate shuttle, which is also 
involved in the interaction between mitochondria and glycolysis, is similar in some cancer cells (Ehrlich 
hepatoma) with respect to normal cells (hepatocytes) (Biochem J 1995, 310: 665-671); decreased activity 
of this shuttle in cancer cells has not been described. 
 
-  “why the authors concentrated on the glycolysis only”. 
One of the most prominent metabolic alterations in cancer cells, in comparison with normal cells, is the 
enhanced glycolytic capacity (see references 1 and 2 for reviews).  In fact, glycolysis in many cancer cells 
is the main ATP supplier, surpassing and/or replacing oxidative phosphorylation.  Therefore, it is highly 
relevant, from the mechanistic and clinical standpoints, to fully understand how this pathway is controlled 
in cancer cells to thus facilitating the identification of the most controlling enzymes (and transporters) 
which, as pointed out by the reviewer, could be used as targets for cancer therapy.  This was originally 
stated on p. 3, lines 1-6; p. 4, 2
nd
 and 4
th
 paragraphs; p. 6, 1
st
 paragraph; p. 28, 2
nd
 paragraph. Extension 
of the glycolytic model to include Krebs cycle and oxidative phosphorylation in an oxidative-type tumor cell 
can be a natural next step in modeling energy tumor metabolism. 
 
-  “why the authors concentrated…only on two tumor cell lines” 
We decided to initiate the MCA of cancer glycolysis with the experimental AS30D hepatoma and the 
human cervix HeLa carcinoma because these cancer cell lines provide relatively high biomass yields for 
experimentation.  It could have been any other cancer cell line with fast growth characteristics.  One 
further development of the present study is that the validated cancer glycolysis kinetics models may now 
serve as platforms on which glycolysis from other cancer cells can be analyzed and predictions be made, 
as far as sufficient data from the respective tumors (i.e., relative enzyme contents; fluxes; metabolite 
concentrations) can be provided and poured into the models.  Statements on this last regard were 
included in the main text on p. 26 3
rd
  paragraph; p. 28 2
nd
 paragraph.   
 
 
 99 
5.2.1 Comentarios manuscrito No. 3 
En este trabajo se construyeron los modelos cinéticos de la glucólisis de las 
líneas tumorales AS-30D y HeLa, a partir de los parámetros cinéticos de cada enzima y 
transportador en el programa Gepasi. Ambos modelos se validaron al lograr predecir el 
flujo glucolítico y las concentraciones de metabolitos en diferentes condiciones, que se 
determinaron experimentalmente. De esta manera, se identificaron a los principales 
sitios de control en la glucólisis de AS-30D (la HK, la HPI y el GLUT) y HeLa 
(degradación de glucógeno, el GLUT y la HK). En consecuencia, todos estos sitos de 
control deben inhibirse simultáneamente para reducir drásticamente el flujo glucolítico y 
la concentración de ATP en células cancerosas. Los modelos desarrollados en este 
trabajo lograron también predecir con gran precisión tanto los flujos y las 
concentraciones de metabolitos bajo hipoxia e hipoglucemia y nos permitieron 
comprender la importancia que pueden tener tanto la HPI en el control del flujo 
glucolítico, como la GAPDH en el control de la concentración de la F1,6BP.  A partir de 
estos resultados concluimos que nuestros modelos son robustos pues predijeron con 
precisión los flujos y las concentraciones de metabolitos en diversas condiciones 
experimentales y que el GLUT, la HK y la HPI pueden ser considerados como posibles 
blancos terapéuticos para inhibir la glucólisis tumoral. 
 
 
 
 100 
5.3 Inhibición de la hexocinasa  
Una vez que habíamos establecido cuales eran los principales sitios de control 
en la glucólisis tumoral (manuscritos 2 y 3), el siguiente paso fue determinar si la 
inhibición de alguno de estos sitios reducía significativamente el flujo glucolítico en 
células tumorales. 
En el siguiente trabajo (Manuscrito No. 4), se determinó el efecto que tiene la 
inhibición de la HK sobre el flujo de la glucólisis. Para inhibirla se empleo un nuevo 
fármaco antineoplásico la Casiopeína II-gly (CasII-gly) que pertenece a una familia de 
compuestos de coordinación con cobre. A continuación se muestra la primera versión 
del manuscrito en preparación con estos resultados. 
 
 
 
 101 
 Casiopeina II-gly and BromoPyruvate inhibition of tumor 
hexokinase  
Alvaro Marín-Hernández1, Juan Carlos Gallardo-Pérez1, Sayra Yoselin López-Ramírez1, 
Jorge Donato García-García1, Lena Ruiz-Ramírez2, Isabel Gracia-Mora2, Alejandro 
Zentella-Dehesa3, Rafael Moreno-Sánchez1 and Sara Rodríguez-Enríquez1*.   
 
1 Departamento de Bioquímica, Instituto Nacional de Cardiología, MEXICO. 
2 Departamento de Química Inorgánica y Nuclear, Facultad de Química, UNAM.  
3Departamento de Bioquímica INNSZ y Departamento de Medicina Genómica y 
Toxicología Ambiental, Instituto de Investigaciones Biomédicas, UNAM. 
Running Title: Casiopeina II-gly inhibits tumor hexokinase  
*Author for correspondence:  Sara Rodriguez-Enríquez, PhD 
Instituto Nacional de Cardiología 
Departamento de Bioquímica 
Juan Badiano No. 1, Col. Sección XVI 
Tlalpan, México D.F. 14080 
MEXICO 
Phone: 5552-55732911 ext. 1422 
Fax: 5552-55730926 
E-mail: saren960104@hotmail.com 
 
Abbreviations: ALDO, Aldolase; 3BrPyr, 3-bromopyruvate; CasII-gly, Casiopeina II-gly; 
F6P, fructose-6-phosphate; F1,6BP, fructose-1,6-bisphosphate; F2,6BP, fructose-2,6-
bisphosphate; GAPDH, glyceraldehyde 3-phosphate dehydrogenase; GK, glucokinase; 
G6P, glucose-6-phosphate; -GPDH, -glycerophosphate dehydrogenase; HK, 
hexokinase; HPI, hexosephosphate isomerase; OxPhos, oxidative phosphorylation; 
PFK-1, phosphofructokinase type 1; PGK, phosphoglycerate cinase ; PYK, pyruvate 
kinase; TPI, triosephosphate isomerase; TriP, trioses phosphate.  
 
Key Words: Cisplatin; fast-growing tumor cells; glycolysis; hexokinase; 3-
bromopyruvate. 
 
 
 102 
Abstract  
The copper-based drug Casiopeina II-gly (CasII-gly) shows potent anti-neoplastic effect 
on several human and rodent fast-growth tumors.  At the metabolic level, CasII-gly 
affects the cellular ATP level and energy mitochondrial metabolism.  In order to 
elucidate whether CasII-gly also affects tumor glycolysis, the effect of CasII-gly was 
assayed on flux through the complete glycolytic pathway and through the initial glycolytic 
segment as well as on the activities of glycolytic enzymes of AS-30D hepatocarcinoma 
cells.  CasII-gly was 13-91 times more potent than 3-BromoPyruvate (3BrPyr), HK 
inhibitor, decreasing tumor cells proliferation, whereas no tumor cells were insensitive to 
CasII-gly. In tumor cells incubated with 100 µM CasII-gly for 30-60 min, lactate 
production was stimulated by 2- (glucose medium) and 14- (glucose-free medium) fold.  
However, glycogen breakdown also increased by 2.8 (glucose medium) and 2-fold 
(glucose-free medium), suggesting that CasII-gly induced increase in lactate production 
derived from glycogenolysis.  Moreover, in CasII-gly treated cells soluble HK activity and 
the flux through the first segment of glycolysis (HK-TPI) were significantly inhibited (58- 
and 41-60 %, respectively), whereas PFK-1, GAPDH, PGK, PYK activities and the flux 
through the HPI-TPI segment were not affected.  In extracts from non-treated cells, 10 
M CasII-gly inhibited both soluble and mitochondrial-bound HK activity, and flux though 
the HK-TPI segment by 50 %; whereas flux through the HPI-TPI segment was not 
affected.  Furthermore, CasII-gly was a 1.3-21 times more potent inhibitor of tumor HK 
than 3BrPyr, whereas, cisplatin did not affect HK maximal rate. These data suggested 
that CasII-gly is a  HK inhibitor and may be considered as an alternative 
chemotherapeutic drug for glycolytic tumors.   
 
 
 
 103 
Introduction 
For ATP supply, fast-growing tumor cells depend on glycolysis (human 
oligodendroglioma, meningioma and medulloblastoma; rat Novikoff hepatoma), OxPhos 
(human colon sarcoma, breast cancer melanoma, breast, lung, ovarian, cervix and 
uterus carcinomas; rat hepatomas (Reuber H-35, Morris, AS-30D) or both pathways 
(human glioblastoma multiforme; rat astrocytoma C6; rat hepatomas (Ehrlich lettré, 
Walker-256, Morris 3683 and Dunings LC18; mouse (ascites, sarcoma 27) (reviewed in 
Moreno-Sanchez et al., 2007; 2009).     
In human carcinomas and lymphomas with severe mitochondrial defects (i.e., 
decreased content and activity in respiratory chain complexes), and in tumors subjected 
to prolonged hypoxia, high resistance to chemo- and radiotherapy is usually 
encountered (Carew and Huang, 2002; Simonnet et al., 2002; Xu et al., 2005; Pelicano 
et al., 2006).  In these tumors, the glycolysis rate is exacerbated, suggesting that tumor 
resistance might be associated with the accelerated glycolysis, a metabolic 
characteristic related to the development of invasive, metastatic and aggressive 
phenotypes (Brizel et al., 2001; Simonnet et al., 2002; Robey et al., 2005). Thus, 
inhibition of glycolysis has been suggested as an alternative target to diminish tumor 
progression in circumstances where traditional therapy is not effective (Pedersen et al., 
2002; Pelicano et al., 2006; Gatenby and Gillies, 2007).  
Pedersen et al.,(2002) proposed HK as target for cancer treatment by considering 
that this enzyme regulates the G6P levels that feed synthesis of ribose 5-phosphate and 
glycogen synthesis and inhibits apoptosis in malignant cells.  Accordingly, several 
groups have tried to manipulate tumor growth by using anti-neoplastic drugs that 
presumably inhibits HK.  For example, lonidamine (5-75 μM) inhibits the HK activity by 
17-85 % in rat Ehrlich ascites (Floridi et al., 1981), although doses 6-30 times higher are 
required to diminish human and rodent fibrosarcomas growth (IC50 = 155-467 μM) 
(reviewed in Rodríguez-Enríquez et al., 2009), thus suggesting that inhibition of HK 
activity is not relevant for the sarcoma development.  The strong inhibition of HK by the 
alkylating agent 3-bromopyruvate (3BrPyr, 5 mM), also reduced the growth and viability 
in AS-30D hepatocarcinoma through ATP cellular depletion (Ko et al., 2001).   
 
 
 104 
Moreover, both lonidamine and 3BrPyr are non-specific as they also affect the 
activity of other enzymes (PGK, GAPDH, pyruvate decarboxylase, isocitrate lyase and 
vacuolar H+-ATPase) and metabolic pathways (acetylcholine synthesis) (Arendt et al., 
1990; Dell’ Antone, 2006; Pereira da Silva et al., 2009); and their toxicity profiles have 
not been evaluated in non-tumorigenic cells. In the case of the lonidamine has been 
reported numerous side effects in humans (Robustelli della Cuna and Pedrazzoli, 1991). 
Although the use of these drugs has shown inconveniences, the basic idea of 
considering HK as the principal target in anticancer therapy is supported by the 
observation that this enzyme is one of the steps controlling tumor glycolysis in human 
and rodent models (Marín-Hernández et al., 2006; 2010).   
In the search for new anti-neoplastic drugs with high selectivity for specific 
enzymes in tumor cells but with null effect in the host cells, the copper-based drugs 
Casiopeinas® have shown strong anti-tumor activity in several human carcinoma lines 
(HeLa, SiHa, CaSki, C33-A, CH1, ovarian) and rodent tumors (leukaemia L1210, glioma 
C6) (De Vizcaya-Ruiz et al., 2000; Marín-Hernández et al., 2003; Trejo-Solís et al., 
2005; Rodríguez-Enríquez et al., 2006) and, encouragingly, a moderate toxic effect on 
lymphocytes (Rodríguez-Enríquez et al., 2006) and heart (Hernandez-Esquivel et al., 
2006).  Similarly to cisplatin, carmustine, mitomycin c, melphalan, cyclophosphamide 
and dacarbazine (reviewed in Lawley and Phillips, 1996), casiopeinas interfere with 
DNA replication through the formation of DNA-adducts.  However, at the doses used to 
block DNA replication (0.4 nM- 30 μM), casiopeina II-gly (CasII-gly) also severely affects 
tumor OxPhos (< 10 μM) through inhibition of mitochondrial 2 oxoglutarate-, pyruvate- 
and succinate-dependent dehydrogenases (Marín-Hernández et al., 2003; Hernández-
Esquivel et al., 2006).  In isolated human and rodent carcinomas and perfused-heart 
(Rodríguez-Enríquez et al., 2006; Hernández-Esquivel et al., 2006) CasII-gly also 
diminished the glycolytic flux.  Therefore, the aim of this work was to determine the 
effect of CasII-gly on tumor glycolysis and HK activity.  The results indicated that CasII-
gly may indeed be considered as a promising multi-site metabolic drug that also targets 
the bioenergetics of glycolytic-dependent tumors.  
 
Materials and methods 
 
 
 105 
Chemicals 
HK, G6P dehydrogenase, HPI, ALDO, GPDH, TPI, lactate dehydrogenase, G6P, F6P 
and F1,6BP were purchased from Roche Co. (Mannheim, Germany). Glucose, G3P, 
MgCl2, ATP, NADH, NAD+, NADP, 3BrPyr and amyloglucosidase were from Sigma 
Chemical (St Louis, MO, USA).  Recombinant HK from Entamoeba histolytica was kindly 
provided by Dr. E. Saavedra, Instituto Nacional de Cardiología, México.  Casiopeina II-
gly (aqua (4-7 dimethyl-1, 10-phenanthroline) (glycine) copper (II) nitrate) was 
synthesized as previously described (Kachadourian et al., 2010) and dissolved in 
distilled water.  
 
Isolation of tumor and no tumor cells 
Human tumor cells (HeLa, PC-3, MDA-MB-231, TD47, HTC15, SK-LU, U87MG, glioma 
C6 and MCF-7) were grown in Dulbecco-MEM medium supplemented with 5% fetal 
serum and 10,000 U/ ml streptomycin/ penicillin at 37°C in 95% air/ 5% CO2.  After  90% 
confluence, cells were removed from each culture flask by adding 0.25% trypsin/ 1mM 
EDTA and washed once by centrifugation with Krebs Ringer buffer (125 mM NaCl, 5 
mM KCl, 1 mM MgCl2, 1.4 mM CaCl2, 1 mM KH2PO4, 25 mM HEPES, pH 7.4).  AS-30D 
hepatocarcinoma cells (4 X 108 cells/ml) were propagated in female 200-250 g Wistar 
rats by intraperitoneal inoculation.  AS-30D cells were isolated by centrifugation as 
previously described (López-Gómez et al., 1993).  
Human platelets were isolated from healthy donor blood.  The platelet rich-plasma was 
separated by differential centrifugation at 200 g for 20 min.  Platelets were washed twice 
with Krebs-Ringer medium and centrifuged at 1800 g for 10 min.  Afterwards, cellular 
pellet was re-suspended in Krebs Ringer medium at a final concentration of 15-20 mg/ml 
(Vasta et al., 1987). Lymphocytes were isolated from donor volunteers as described 
elsewhere (Rodríguez-Enríquez et al., 2006). Primary human endothelial cells 
(HUVECs) were isolated from umbilical cords according to Jaffe et al (1973).  
 
Cellular protein was determined by biuret assay using bovine serum albumin 
(BSA) as standard (Gornall, et al., 1949).  Viability assessed by the trypan blue 
 
 
 106 
exclusion assay. revealed less than 10-15% cellular death under all experimental 
conditions for tumor cells and normal cells .   
 
Cell growth in cells exposed to CasII-gly and 3BrPyr   
 Cell growth inhibition was determined by the crystal violet method (Gallardo-
Pérez et al., 2009).Briefly, cells (tumor and normal cells) were seeded at 2 x 105 
cells/well in 96 and 24-well plates (for 24 and 72 hrs respectively), left for 24 h prior to 
the treatment with drugs. After, cells were exposed for 24 and 72 hrs to different 
concentrations (0.1-100 µM) of CasII-gly and 3BrPyr.  
Isolation of rat hepatocarcinoma mitochondria 
 AS-30D mitochondria were isolated by differential centrifugation with digitonin (15 
μg/ mg cellular protein; ICN, Ohio, USA) as previously described by López-Gómez et al 
(1993). 
 
Determination of glycolytic fluxes and metabolite concentrations  
 Tumor cells (15 mg protein/ml) were incubated in Krebs-Ringer buffer under 
orbital shaking (150 rpm) at 37°C with or without 100 μM CasioII-gly.  After 1 min pre-
incubation, 5 mM glucose (+GLU) was added.  One ml of cellular sample was taken 
immediately (time 0), and the rest was further incubated for 10, 20, 30 and 60 min.  
Each sample was mixed with perchloric acid (3%, v/v final concentration) and 
centrifuged at 1800 g by 1 min.  The supernatant was neutralized with 3 M KOH/ 0.1 M 
Tris and centrifuged once to eliminate potassium salts.  Supernatants were kept on ice 
for determination of G6P, F6P, F1,6BP, Trioses phosphate (TriP), ATP and L-lactate 
concentrations by using standard enzymatic methods (Bergmeyer, 1974).  In parallel, 
cellular samples (1 ml) were taken after 30 and 60 min incubation, and frozen in liquid 
nitrogen and kept at -70 °C for determination of HK and PFK-1 activities, and fluxes from 
HK to TPI (first section of glycolytic pathway).   
 For glycogen determination, the cellular samples withdrawn after 30 and 60 min 
incubation were centrifuged at 20800 g for 20 s.  The pellet was frozen in liquid nitrogen 
and kept at -70 °C until its use.  The thawed cellular pellet was re-suspended in 30% 
KOH and heated for 30 min at 90°C.  Afterwards, the sample was mixed with 6% 
 
 
 107 
Na2SO4 and 100% ethanol and centrifuged at 20800 g for 5 min.  Pellet was washed 
once with 1 ml of 80% ethanol and incubated at room temperature.  Finally, the dried 
sample was re-suspended in 0.2 M acetate buffer, pH 4.8, plus 5 U amyloglucosidase 
and incubated for 1 h at 37°C (Hue et al., 1975).  Glucose units derived from glycogen 
breakdown were determined by enzymatic analysis (Bergmeyer, 1974). 
 
Enzyme activities and initial glycolytic segment fluxes 
 For determination of HK and PFK-1 activities, and fluxes from HK to TPI (first 
section of glycolytic pathway), aliquots of 1 ml (15 mg protein) were taken from cells 
incubated with CasII-gly at 30 and 60 min and centrifuged at 20800 g for 20 s.  Cellular 
extracts were obtained from pellets  re-suspended in 0.3 ml 25 mM Tris/HCl buffer plus 
1 mM EDTA, 5 mM DTT and 1 mM PMSF (Marín-Hernández et al., 2006) and frozen in 
liquid nitrogen.  Afterwards, three cycles of freezing/thawing at 37°C were carried out to 
obtain cellular lysates, which were centrifuged at 39000 g for 20 min at 4°C.  Collected 
supernatants were stored in 10% (v/v) glycerol at -20°C until determination of enzyme 
activities and fluxes in KME buffer (100 mM KCl, 50 mM Mops and 1 mM EGTA) pH 7.0, 
at 37°C (Marín-Hernández et al., 2006).  
For HK activity determinations in liver, brain and heart rat tissues, the organs 
were homogenized in 2 mL of SHE buffer (250 mM sucrose, 10 mM Hepes, 1 mM 
EGTA, pH 7.3) plus 1 mM EDTA, 0.5 mM DTT and 1 mM PMSF.  Homogenates were 
centrifuged at 39 000 g for 20 min and 4 °C.  Afterwards, the supernatant was collected 
for determination of enzyme activity. 
 The assays for both cytosolic and mitochondrial HK were carried out in 1 mL of 
KME (120 mM KCl, 20 mM Mops, EGTA 0.5 mM, pH 7) or KM (100 mM KCl, 50 mM 
Mops, pH 7) buffer plus 10 mM ATP, 15 mM MgCl2, 1 mM NADP, 1 U/ml G6PDH and 
cellular extract (0.03 - 0.1mg).  The reaction was started by adding 3 mM glucose.  PFK-
1 activity was determined in KME buffer plus 2 mM ATP, 5 mM MgCl2, 5 µM F2,6BP, 
0.15 mM NADH, 0.36 U aldolase/ml, 9 U triosephosphate isomerase/ml, 3 U α-GPDH/ml 
and cellular extract (0.03 – 0.1 mg protein).  The reaction was started with 2 mM F6P. 
 GAPDH, PGK and PYK activities were determined according to Saavedra et al. 
(2004).   
 
 
 108 
 The flux through the first section of the glycolytic-pathway, i.e., from HK to TPI 
was determined by the method described by Torres et al.,(1988), modified as follows: 
The cellular extract (0.016- 0.5 mg protein) was pre-incubated for 5 min in KME buffer 
plus 10 mM ATP, 15 mM MgCl2, 5 µM F2,6BP, 0.15 mM NADH and 6 U α- GPDH /ml; 
the reaction started by adding 3 mM glucose.  To determine flux through the HPI-TPI 
segment, the reaction started by adding 10 mM G6P; for flux through the PFK-1-TPI 
segment, the reaction started by adding 5 mM F6P; and for flux through the ALDO-TPI 
segment, the reaction was started with 1 mM F1,6BP.  In all assays, control experiments 
were undertaken to assess linearity of fluxes with respect to protein used.  When the 
specific substrate was omitted, the flux rate was negligible.  Also, coupling enzymes 
were insensitive to CasII-gly at the range of concentration used.   
 
Statistical analysis 
 Student´s t-test for non-paired samples was used considering P < 0.05 as 
criterion of significance. 
 
Results 
Effects of CasII-gly and 3BrPir on the proliferation of normal and cancer cells  
Due to that in preliminary essays were observed that CasII-gly inhibited HK 
activity in extracts of tumor cells, we decided to compare the effects of CasII-gly with 3-
Bromo pyruvate (3BrPyr, aparently HK inhibitor) on the proliferation of several no-tumor 
and tumor cells lines. Both drugs decrease cells proliferation, in spite of presenting 
distinct  effectiveness (Table1). It can be observed that CasII-gly was 13-91 times more 
potent than 3BrPyr decreasing tumor cells proliferation in 24 and 72 hrs. In contrast, 
normal cells (lymphocytes and endothelial cells) tested were relatively resistant to CasII-
gly and 3BrPyr in comparison with tumor cells (Table1). Similar IC50 values of both drugs 
have been reported in other tumor and normal cells (Xu et al., 2005; Rodriguez-
Enríquez, et al., 2006; Cao et al., 2008). These results are relevant because CasII-gly 
showed specificity by tumor cells over normal cells. 
Effect of CasII-gly on tumor glycolytic flux and high energy-metabolites content  
 
 
 109 
Treatment of tumor cells with CasII-gly and 3BrPyr induced a dramatic decrease 
in intracellular levels of gluthation (GSH) in function of protein concentration (data no 
shown). In consequence, we decided to determine optimal cellular protein concentration 
in which it was not observed modifications in the levels of GSH and this concentration 
was 15 mg /mL. 
Titration of tumor glycolytic flux with CasII-gly (range 1-100 M) (data not shown) 
revealed that 100 M was an adequate concentration to induce changes in the glycolytic 
flux and glycolytic enzyme activities, without cellular viability alterations. 
Therefore,  the effect of 100 M CasII-gly on the glycolytic flux was assayed in  
AS-30D carcinoma cells   incubated in glucose-free medium (-GLU) or glucose-medium 
(+GLU) for 10-60 min (Fig. 1A).  The incubation for 60 min without exogenous glucose 
did not change the lactate (Fig. 1A) and glycogen contents (Table 2), indicating 
depressed glycolysis and glycogenolysis rates.  In the presence of CasII-gly for 60 min, 
the endogenous lactate increased 3.6-fold (Fig. 1A), suggesting drug activation of tumor 
glycolysis.  In the presence of glucose, an increase of 13.7-times in lactate production 
was observed, in agreement with a previous report (Marín-Hernández et al., 2006), and 
the addition of CasII-gly did not modify significantly the production of lactate (Fig. 1A).  
Thereby, a significant  increase in the glycolytic flux (12.2- and 1.9-fold) was attained in 
both –GLU  and  +GLU media between 30 and 60 min (Fig. 1B), indicating an active 
glycogen breakdown induced by the anti-neoplastic drug.    To add support to the last 
hypothesis, the content of glycogen was determined.  As expected, the drug  
significantly decreased (2-2.8times)  the glycogen content (Table 2; Fig. 1C), which 
correlated well with the enhanced production of lactate (Fig. 1B) in cells stimulated by 
exogenous glucose, i.e., when glycolysis is active; this correlation was less clear in cells 
incubated with no glucose.   
Casiopeina II-gly did not modify the steady-state concentration of some glycolytic 
intermediates such as G6P, F6P, F1,6BP and TriP in glucose-incubated cells (Table 2), 
except for ATP, that was severely reduced by  60 % after 60 min incubation.  In the 
absence of added glucose, the concentration of the glycolytic intermediates was below 
the limit of detection.  However, although glycolysis was operating at a low rate, a high 
ATP level was still detected, which most likely derived from OxPhos; under these 
 
 
 110 
conditions, CasII-gly also induced a significant diminution in the ATP content (78 %), 
indicating an inhibition of  OxPhos as previously described (Marín-Hernández et al., 
2003) (Table 2). 
 
Effect of CasII-gly on the HK and PFK-1 activities and HK-TPI flux in drug-pretreated 
carcinoma cells 
The observation that G6P levels were not modified by Cas-IIgly suggested (i) that 
the G6P pool was maintained by an activated glycogen breakdown (Table 2; Fig. 1C) or 
(ii) that HK was not a target for CasII-gly.  Therefore, the activity of the soluble HK 
fraction (cytosolic enzyme)  was determined in extracts of AS-30D tumor cells exposed 
for 60 min to 100 M CasII-gly.  Independently of the presence of glucose, CasII-gly 
drastically reduced the HK activity by 40-60%; in contrast, the PFK-1 activity (as control 
enzyme) was not affected by the drug (Fig. 2A).  To further examine whether  CasII-gly 
affects other enzymes of tumor  glycolysis, the effect of the drug was assayed in both 
upper- and lower pathway segments (Fig. 2B).  The flux of the first segment  was 
assayed  by feeding HK, HPI, PFK-1 or ALDO with its specific substrate (Fig. 2B) and 
measuring the -GP produced under each experimental condition.  CasII-gly inhibited 
the HK-TPI flux by 58%.  On the contrary, the drug was ineffective on (i) the HPI-TPI (by 
adding G6P), PFK1-TPI (with F6P), and ALDO-TPI (with F1,6BP) fluxes, and (ii) the 
GAPDH, PGK and PYK activities (data no shown) .  These observations clearly 
indicated that HK was the principal and exclusive CasII-gly target of tumor glycolysis.   
 
Effect of CasII-gly on the HK activity and HK-TPI flux in non-drug treated  tumor cells  
To assess whether CasII-gly directly interacts with HK, mitochondrial and soluble 
HK fractions were prepared from non-drug treated AS-30D hepatoma.   . To evaluate 
the effect of CasII-gly on the HK activity, the kinetics assays were performed in free-
EGTA medium (Table 2).  HK derived from different tumor sources was highly sensitive 
to CasII-gly or 3BrPyr, as well as HK derived from non-neoplastic tissues (Table 2).  In 
the presence of EGTA, CasII-gly also promoted a significant inhibition of both soluble- 
and mitochondrial-HK AS-30D fractions at slightly higher concentrations (IC50 = 8 ± 3 
 
 
 111 
M; n=3; and 10 ± 2 M; n=4; respectively) after 5 min incubation; in turn, 3BrPyr IC50 
values for soluble-HK activity was similar in absent or in presence of EGTA (28 ± 8 (4) 
and 28 ± 10 (4) M respectively). 
Flux through the HK-TPI segment in non previously-treated cells was 48% 
inhibited by CasII-gly (10 μM), whereas flux through the HPI-TPI was unaffected (Fig. 
3B), indicating that HK is a specific and direct CasII-gly target.   
In order to evaluate whether other metal-based anti-neoplastic drugs, or the ion 
copper, have similar effect on HK activity as CasII-gly, cellular extracts were incubated 
for 5 min with 100 M cisplatin (Fig. 3C,a) or 10-20 M CuCl2 (Fig 3C,b,c).  Non-
significant diminution in HK activity was attained with cisplatin, at same doses used to 
inhibit  rabbit muscle Aldo (85%) and GAPDH (45%) (Aull, et al., 1979).  
Copper is the metal involved in the coordination of the dimethyl-phenantroline 
rings in CasII-gly.  In human erythrocytes, copper (15-100 M) reacts with thiol-groups 
of several glycolytic enzymes promoting activity inhibition (Boulard et al., 1972).  
However, only at 20 µM Cu2+ HK activity was partially inhibited (Fig. 3C,c). 
The activities of HK and other glycolytic enzymes (PFK, GAPDH, PGK and PYK) 
were also evaluated in extracs of cells incubated 60 min in the presence of 100 M 
3BrPyr or CuCl2.  However, none of the drugs, or ion, tested significantly modified the 
activity of these enzymes and lactate production (data not shown).Only 3BrPyr reduced 
GAPDH activity 25% (data not shown).  
 
Discussion 
Metabolic drugs have recently emerged as a suitable alternative therapeutic 
approach against malignant chemo- and radiotherapy-resistant tumors (Rodriguez-
Enriquez et al., 2009).  The action mechanism of the copper-based drug Casiopeina II-
gly is to disturb DNA stabilization in HeLa, SiHa, Caski and C33-A tumors (Mejía and 
Ruiz, 2008; reviewed in Rodríguez-Enríquez et al., 2009) and to promote apoptosis in a 
process mediated by reactive oxygen species (Trejo-Solís et al., 2005; Mejía and Ruiz, 
2008;).  Moreover, CasII-gly affects mitochondrial ATP synthesis in rat AS-30D 
hepatocarcinoma, a highly OxPhos-dependent tumor (Rodríguez-Enríquez et al., 2006) 
 
 
 112 
by interacting with the reactive thiol groups of Krebs cycle (pyruvate, 2-OG, succinate) 
dehydrogenases (Marín-Hernández et al., 2003; Hernández-Esquivel et al., 2006).   
 
Glycogen breakdown-induced by CasII-gly is responsible of the increment in tumor 
glycolysis   
 After prolonged incubation with CasII-gly, lactate production significantly 
increased suggesting glycolysis activation.  A similar pattern was observed in normal 
(mouse fibroblast 3T3-L1 adipocytes, L6 rat skeletal myoblasts) and HepG2 tumor cells 
after treatment with the anti-hyperglycemic drugs biguanides (buformin, phenformin), 
nefazodone (antidepressant) or the alkaloid berberine (Dykens et al., 2008a, 2008b; Yin 
et al., 2008).  It was suggested that biguanides-, nefazodone- and berberine-glycolysis 
activation was the result of mitochondrial dysfunction (Dykens et al., 2008a, 2008b; Yin 
et al., 2008).  However, glycogen degradation increased significantly in the presence of 
CasII-gly (cf., Fig 1C), thus indicating that the increase in lactate production (cf., Fig 1B) 
was the result of the drug-activated glycogen-breakdown.  Although, the mechanism 
involved in the CasII-gly-activated glycogenolysis was not analyzed here, in rat isolated 
hepatocytes was demonstrated that non-steroidal anti-inflammatory drugs (Ibuprofen, 
indomethacin, meclofenamate) may promote glycogen degradation by interfering with 
the intracellular Ca2+ homeostasis (Brass and Garrity, 1985) and probably increasing 
glycogen phosphorylase A activity.    
The glycogen breakdown after 60 min incubation with CasII-gly was 16 (-GLU) 
and 72 (+GLU) nmol/mg protein (cf. Table 1), which should theoretically generate 32 
and 144 nmol lactate/mg protein, respectively.  These predicted values matched the 
lactate production (cf. Fig. 1A, B) of 61 (-GLU) and 86 (+GLU) nmol/mg protein.  
 
Mitochondrial and soluble-hexokinase are targets of CasII-gly   
The majority of fast-growth tumor cells over-express HK-II, which is 30-50% 
located in the cytosol (Ko et al., 2001; Marín-Hernández et al., 2006) and the rest bound 
to the external mitochondrial membrane (Pastorino and Hoek, 2003) in a close vicinity to 
the adenine nucleotide translocator and VDAC (Pastorino and Hoek, 2003).  Its 
 
 
 113 
mitochondrial location is considered as a metabolic strategy of cancer cells to channel 
mitochondrial ATP for glycolysis (Pastorino and Hoek, 2003) and to inhibit apoptosis by 
blocking cytochrome c release (Pedersen et al., 2002; Pastorino and Hoek, 2003).  
Furthermore, HK is one of the main controlling steps of tumor glycolysis (Marín-
Hernández et al., 2006; 2010).  These characteristics make this enzyme an attractive 
target for anti-neoplastic therapy.  
3BrPyr has shown potent inhibitory effect on the cellular proliferation of leukemia 
HL-60 cancer and AS-30D hepatocarcinoma (IC50≈ 10- 50 μM) and Hep 3B and 
SNU449 hepatocarcinomas (50-200 M) (Ko et al., 2004; Xu et al., 2005;  Kim et. al., 
2008).  In VX-2 epidermoid rabbit and rat AS-30D tumors, 3BrPyr bolus of 1-25 ml of 
0.5-2 mM significantly reduced the tumor progression in vivo (Geschwind et al., 2002; 
Ko et al., 2004). Although it has been claimed that 3BrPyr effect on tumor proliferation 
(including tumor progression, tumor toxicity) was the result of HK inhibition, it seems that 
other sites are involved as HK inhibition is attained at very high 3BrPyr concentration (2-
10 mM) (Ko et al., 2001).  In addition, several groups have described non-specific 
effects of 3BrPyr.  For example, in HepG2 cells, 150 μM 3BrPyr does not affect HK 
activity but severely decreases (90 and 70 %) GAPDH and PGK activities (Pereira da 
Silva et al., 2009). Similar specific inhibition on GAPDH activity is reported in other 
human (Hep3B and SK-Hep1) and rabbit (Vx-2) hepatocellular carcinoma cell lines 
exposed to concentrations of 10 to 200 µM (Ganapathy-Kanniappan et al., 2009). In our 
study with 100 µM 3BrPyr only was observed 25% in the reduction of GAPDH activity 
(data no shown). These differences in results can be attributed to time of exposition 
(0.5-2 hr) and DTT concentration used in each case (0.5-5 mM). At 3BrPyr doses 
assayed to arrest tumor progression, pyruvate decarboxylase, isocitrate lyase, H+ 
vacuolar ATPase, and acetylcholine synthesis are also affected (Arendt et al., 1990; 
Dell’ Antone, 2006).  
Several highly malignant tumors such as human melanoma, lymphoma, head and 
neck, breast, liver and colorectal cancers are apparently dependent on the glycolysis 
pathway to survive (Okada, et al, 1992; ;Lapela, et al, 1995; Abdel-Nabi, et al, 1998; 
Schwimmer, et al, 2000; Czernin and Phelps, 2002).  In this context, the search for a 
potent, specific and permeable anti-glycolytic drug appears as a novel and promising 
 
 
 114 
task.  Encouragingly, CasII-gly inhibited both soluble (AS-30D, HeLa, MCF-7, MDA-MB-
231, SK-LU, PC-3) and mitochondrial (AS-30D) HK isoforms (cf. Fig. 2A, 3A; Table 2). 
However, CasII-gly potency was reduced 2 times in presence of EGTA, probably 
because the copper is sequestrated by this agent. Therefore, tricarboxylic acid cycle 
intermediates as citrate and malate (physiological chelators) may reduce potency of 
CasII-gly in vivo.   
The HK inhibition by Cas-IIgly was highly specific as other glycolytic enzymes 
were not affected.  However, HKI, HKII and GK of normal rat tissues and human 
platelets were also inhibited (Table 2). But when normal and tumor cells were exposed 
to CasII-gly, this drug was more potent in reduced cellular proliferation on tumor cells 
than normal cells (see Table 1)  This selective effect is not achieved by lonidamine, 
clotrimazol or 3BrPyr (Floridi et al., 1981; Penso et al., 2002; Pereira da Silva et al., 
2009).  The apparent higher selectively of CasII-gly towards tumor cells over normal 
cells seems related to its physico-chemical nature of lipophilic cation, which facilitates its 
entrance to tumor cells and mitochondria that maintain higher transmembrane electrical 
gradients (negative inside) across the plasma and inner mitochondrial membranes, in 
comparison with normal cells and mitochondria (Moreno-Sánchez et al., 2007; 2009; 
Rodríguez-Enríquez et al., 2009). 
 
CasII-gly inhibits HK at lower doses than 3BrPyr and cisplatin  
 CasII-gly (IC50= 4-10 M) resulted to be a more potent HK inhibitor than 3BrPyr 
(IC50=28 M), cisplatin (cf, Fig 3B) and others anticancer drugs (clotrimazol or 
lonidamine) strongly suggesting that CasII-gly may be considered as a potent metabolic 
anti-cancer drug affecting energy metabolism.  It is reported that 3BrPyr have the ability 
to interact with sulfhydryl groups (Ganapathy-Kanniappan et al., 2009). These groups 
apparently are critical for structural stability and catalytic activity of hexokinase isoforms 
(Hutny and Wilson, 1990;  Tiedge et al., 2000) as consequence are higher susceptibility 
to several alquilant agents such as 3BrPyr (Connolly and Trayer, 1978; Fiorani et al., 
2000; Tiedge et al., 2000). However in this study was not analyzed the mechanism of 
CasII-gly inhibition of HK, this inhibition can   be atribute to cysteine interaction how has 
been establish for inhibition of mitochondrial 2 oxoglutarate-, pyruvate- and succinate-
 
 
 115 
dependent dehydrogenases (Marín-Hernández et al., 2003; Hernández-Esquivel et al., 
2006).  
However, we demonstrated that CasII-gly inhibits HK activity, it was observed a 
negligible effect on glycolysis. The results suggest that CasII-gly enhances glucose 
metabolism by stimulation of glycogen degradation, which is related to a mechanism to 
compensate the HK inhibition and lower ATP. The level of glycogen is function of the 
both its production and its breakdown, which are mediated by the enzymes glycogen 
synthase (GS) and glycogen phosphorylase (GP), respectively. The activity of these 
enzymes is regulated by phosphorylation and several effectors. GP catalizes the first 
step in intracellular degradation of glycogen, which may activated by AMP and is 
inhibited by ATP and other metabolites so that the enzyme is response to the levels of 
metabolites in the cell (Stalmans et al., 1974; Ercan et al, 1996). As the effect of ATP is 
counteracted by AMP concentration, we can propose that the glycogenolysis is activated 
because CasII-gly lead to ATP depletion (possibly as result of mitochondrial 
impairment), increasing availability of AMP. These changes in ATP and AMP 
concentrations could support the activation of the GP.Finally, the results suggested that 
CasII-gly may be considered as a potent and specific metabolic drug that affect energy 
metabolism and also indicated that the inhibition of solely one controlling step brought 
about negligible effects on glycolysis. Only the simultaneous inhibition of the other 
controlling steps (GLUT and HPI) may have significant impact on glycolytic flux (Marín-
Hernández et al., 2010).  
 
Acknowledgements 
This work was partially supported by a CONACyT-Mexico grant (No 80534 ).  AMH was 
supported by a CONACyT fellowship (No. 59991).   
 
References  
Abdel-Nabi, H., Doerr, R.J., Lamonica, D.M., Cronin, V.R., Galantowicz, P.J., Carbone, G.M., Spaulding, 
M.B., 1998. Staging of primary colorectal carcinomas with fluorine-18 fluorodeoxyglucose whole-body 
PET: correlation with histopathologic and CT findings. Radiology. 206, 755-60. 
Aull, J.L., Allen, R.L., Bapat, A.R., Daron, H.H., Friedman, M.E., Wilson, J.F., 1979. The effects of 
platinum complexes on seven enzymes. Biochem. Biophys. Acta 571, 352-358. 
 
 
 116 
Arendt, T., Schugens, M.M., Marchbanks, R.M., 1990. Reversible inhibition of acetylcholine synthesis and 
behavioural effetcs caused by 3-bromopyruvate. J. Neurochem. 55, 1474-1479. 
Bergmeyer, H.U. (Ed.), 1974. Methods of Enzymatic Analysis, second English Ed. Verlag Chemie-
Academic Press, New York. 
Boulard, M., Blume, K.G., Beutler, E., 1972. The effect of copper on red cell enzyme activities. J. Clin. 
Inves. 51, 459-461. 
Brass, E.P., Garrity, M.J., 1985. Effect of nonsteroidal anti-inflammatory drugs on glycogenolysis in 
isolated hepatocytes. Br. J. Pharmacol. 86, 491-496 
Brizel, D.M., Schroeder, T., Scher, R.L., Walenta, S., Clough, R.W., Dewhirst, M.W., Mueller-Klieser, W., 
2001. Elevated tumor lactate concentrations predict for an increased risk of metastases in head-and-
neck cancer. Int. J. Radiat. Oncol. Biol. Phys. 51, 349-53. 
Cao, X., Bloomston, M., Zhang, T., Frankel, W.L., Jia, G., Wang, B., Hall, N.C., Koch, R.M., Cheng, H., 
Knopp, M.V., Sun, D., 2008. Synergistic antipancreatic tumor effect by simultaneously targeting 
hypoxic cancer cells with HSP90 inhibitor and glycolysis inhibitor. Clin. Cancer Res. 14, 2008. 
Carew, J.S., Huang, P., 2002. Mitochondrial defects in cancer. Mol. Cancer 1:9.Connolly, B.A., Trayer, 
I.P., 1979. Affinity labelling of rat-muscle hexokinase type II by a glucose-derived alkylating agent. Eur 
J Biochem. 93, 375-83. 
Czernin, J., Phelps, M.E., 2002, Positron emission tomography scanning: current and future applications. 
Annu. Rev. Med. 53, 89-112. 
Dell’ Antone, P., 2006. Inactivation of H+
- vacuolar ATPase by the energy blocker 3-bromopyruvate, a new 
antitumor agent. Life Sci. 79, 2049-2055. 
De Vizcaya-Ruiz, A., Rivero-Mueller, A., Ruiz-Ramírez, L., Kass, G.E., Kelland, L.R., Orr, R.M., 2000.  
Induction of apoptosis by a novel cupper-based anticancer compound, casiopeina II, in L1210 murine 
leukaemia and  CH1 human ovarian carcinoma cells. Toxicol. In Vitro 14,1-5. 
Dykens, J.A., Jamieson, J.D., Marroquin, L., Nadanciva, S., Xu, J.J., Dunn, M.C., Smith, A.R., Will, Y., 
2008a. In vitro assessment of mitochondrial dysfunction and cytotoxicity of nefazodone, trazodone, and 
buspirone. Toxicol. Sci. 103, 335-345.  
Dykens, J.A., Jamieson, J., Marroquin, L., Nadanciva, S., Billis, P.A., Will, Y., 2008b. Biguanide-induced 
mitochondrial dysfuntion yields incresed lactate production and cytotoxicity of aerobically-poised 
HepG2 cells and human hepatocites in vitro. Toxicol. Appl. Pharmacol. 233, 203-210.  
Ercan, N., Gannon, M.C., Nuttall, F.Q.,1996. Arch Biochem Biophys. 328, 255-64.Fiorani, M., De Sanctis, 
R., Scarlatti, F., Vallorani, L., De Bellis, R., Serafini, G., Bianchi, M., Stocchi, V., 2000. Mol Cell 
Biochem. 209, 145-53. 
Floridi, A., Paggi, M.G., D’ Atri, S., Martino, C.D., Marcante, M.L., Silvestrini, B., Caputo, A., 1981. Effect 
of lonidamie on the energy metabolism of Erlich ascites tumor cells. Cancer Res. 41, 4661-4666. 
Gallardo-Pérez, J.C., Espinoza, M., Ceballos-Cancino, G., Daniel, A., Rodríguez-Enríquez, S., Aviles, A., 
Moreno-Sánchez, R., Melendez-Zajgla, J., Maldonado, V., 2009. NF-kappa B is required for the 
development of tumor spheroids. J. Cell. Biochem. 108, 169-180. 
Ganapathy-Kanniappan, S., Geschwind, J.F., Kunjithapatham, R., Buijs, M., Vossen, J.A., Tchernyshyov, 
I., Cole, R.N., Syed, L.H., Rao, P.P., Ota, S., Vali, M., 2009. Anticancer Res. 29, 4909-18. 
Gatenby, R.A., Gillies, R.J., 2007. Glycolysis in cancer: a potential target for therapy. Int. J. Biochem. Cell 
Biol. 39, 1358-66.  
Geschwind, J.F.H., Ko, Y.H., Torbenson, M.S., Magee, C., Pedersen, P.L., 2002.Novel therapy for liver 
cancer: direct intraarterial injection of a potent inhibitor of ATP production. Cancer Res. 62, 3909-3913. 
Gornall, A.G., Bardwill, C.J., David, M.M., 1949. Determination of serum proteins by means of biuret 
reaction. J. Biol. Chem. 177, 751-766. 
Hernández-Esquivel, L., Marín-Hernández, A., Pavon, N., Carbajal, K., Moreno-Sánchez, R., 2006. 
Cardiotoxicity of copper based antineoplastic drugs casiopeinas is related to inhibition of energy 
metabolism. Toxicol. Appl. Pharmacol. 212, 79-88. 
 
 
 117 
Hue, L., Bontemps, F., Hers, H., 1975. The effects of glucose and of potassium ions on the 
interconversion of the two forms of glycogen phosphorylase and of glycogen synthetase in isolated rat 
liver preparations.  Biochem. J. 152, 105-14.  
Hutny, J., Wilson, J.E., 1990.Effect of ligand-induced conformational changes on the reactivity of specific 
sulfhydryl residues in rat brain hexokinase. Arch Biochem Biophys. 283, 173-83. 
Jaffe, E., Nachmen, R.L., Becker, G.C., Minick, C.R., 1973. Culture of human endothelial cells from 
umbilical veins. J. Clin. Invest. 52, 2745-2756. 
Kachadourian, R., Brechbuhl, H. M., Ruiz-Azuara, L., Gracia-Mora, I., Day, B.J., 2010. Casiopeína IIgly-
induced oxidative stress and mitochondrial dysfunction in human lung cancer A549 and H157 cells. 
Toxicology, 268, 176-183.  
Kim, J.S., Ahn, K.J., Kim, J.A., Kim, H.M., Lee, J.D., Lee, J.M., Kim, S.J., Park, J.H., 2008. Role of 
reactive oxygen species-mediated mitochondrial dysregulation in 3-bromopyruvate induced cell death 
in hepatoma cells. J. Bionerg. Biomembr. 40, 607-618.       
Ko, Y.H., Pedersen, P.L., Geschwind, J.F., 2001. Glucose catabolism in the rabbit VX2 tumor model for 
liver cancer: characterization and targeting hexokinase. Cancer Lett. 173, 83-91. 
Ko, Y.H., Smith, B.L., Wang, Y., Pomper, M.G., Rini, D.A., Torbenson, M.S., Hullihen, J., Pedersen, P.L., 
2004. Advenced cancers: eradication in all cases using 3-bromopyruvate therapy to deplete ATP. 
Biochem. Biophys. Res. Commun. 324, 269-275.  
Lapela, M., Grénman, R., Kurki, T., Joensuu, H., Leskinen, S., Lindholm, P., Haaparanta, M., 
Ruotsalainen, U., Minn, H., 1995. Head and neck cancer: detection of recurrence with PET and 2-[F-
18]fluoro-2-deoxy-D-glucose. Radiology. 197, 205-11. 
Lawley, P.D., Phillips, D.H., 1996. DNA adducts from chemotherapeutics agents. Mut. Res. 355, 13-40. 
López-Gómez, F., Torres-Márquez, M.E., Moreno-Sánchez, R., 1993. Control of oxidative phosphorylation 
in AS-30D hepatoma mitochondria. Int. J. Biochem. 25, 373-377. 
Marín-Hernández, A., Gracia-Mora, I., Ruiz-Ramírez, L., Moreno-Sánchez, R., 2003. Toxic effects of 
cupper-based antineoplastic drugs (Casiopeinas) on mitochondrial fuctions. Biochem. Pharmacol. 65, 
1979-1989. 
Marin-Hernandez, A., Rodriguez-Enriquez, S., Vital-Gonzalez, P.A., Flores-Rodriguez, F.L., Macias-Silva, 
M., Sosa-Garrocho, M., Moreno-Sanchez, R., 2006.  Determining and understanding the control of 
glycolysis in fast-growth tumor cells. Flux control by an overexpressed but strongly product-inhibited 
hexokinase. FEBS J. 273, 1975–1988. 
Marín-Hernández, A., Gallardo-Pérez, J.C., Rodríguez-Enríquez, S., Encalada, R., Moreno-Sánchez, R., 
Saavedra E., 2010. Modeling cancer glycolysis. Biochim Biophys Acta. 2010 “In Press”. 
Mejia, C., Ruiz-Azuara, L., 2008. Casiopeinas IIgly and IIIia induce apoptosis in medulloblastoma cells. 
Pathol. Oncol. Res. 14, 467-472.    
Moreno-Sánchez, R., Rodríguez-Enríquez, S., Marín-Hernández, A., Saavedra, E., 2007. Energy 
metabolism in tumor cells. FEBS J. 274, 1393-1418. 
Moreno-Sánchez, R., Rodríguez-Enríquez, S., Saavedra, E., Marín-Hernández, A., Gallardo-Pérez, J.C., 
2009. The bioenergetics of cancer: Is glycolysis the main ATP supplier in all tumor cells. Biofactors 35, 
209-225. 
Okada, J., Yoshikawa, K., Itami, M., Imaseki, K., Uno, K., Itami, J., Kuyama, J., Mikata, A., Arimizu, N., 
1992. Positron emission tomography using fluorine-18-fluorodeoxyglucose in malignant lymphoma: a 
comparison with proliferative activity. J Nucl Med. 33, 325-9. 
Pastorino, J.G., Hoek, J.B., 2003. Hexokinase II: The integration of energy metabolism and control of 
apoptosis. Curr. Med. Chem. 10, 1535-1551. 
Pedersen, P.L., Mathupala, S., Rempel, A., Geschwind, J.F., Ko, Y.H., 2002. Mitochondrial bound type II 
hexokinase: a key player in the growth and survival of many cancers and an ideal prospect for 
therapeutic intervention. Biochim. Biophys. Acta 1555, 14-20. 
Pelicano, H., Martin, D.S., Xu, R.H., Huang, P., 2006. Glycolysis inhibition for anticancer treatment. 
Oncogene 25, 4633-4646. 
 
 
 118 
Penso, J., Beitner, R., 2002.Detachment of glycolytic enzyme from cytoskeleton of Lewis lung carcinoma 
and colon adenocarcinoma cells induced by clotrimazole and its correlation to cell viability and 
morphology. Mol. Genet. Metab. 76, 181-188.    
Pereira da Silva, A.P., El-Bacha, T., Kyaw, N., dos Santos, R.S., da Silva, W.S., Almeida, F.C., Da Poian, 
A.T., Galina, A., 2009. Inhibition of energy –producing patways of HepG2 cells by 3-bromopyruvate. 
Biochem. J. 417, 717-726. 
Robey, I.F., Lien, A.D., Welsh, S.J., Baggett, B.K., Gillies, R.J., 2005. Hypoxia-inducible factor-1alpha and 
the glycolytic phenotype in tumors. Neoplasia. 7, 324-30. 
Rodríguez-Enríquez, S., Vital-González, P.A., Flores-Rodríguez, F.L., Marín-Hernández, A., Ruiz-Azuara, 
L., Moreno-Sánchez, R., 2006. Control of cellular proliferation by modulation of oxidative 
phosphorylation in human and rodent fast-growing tumor cells. Toxicol. Appl. Pharmacol. 215, 208-
217. 
Rodríguez-Enríquez, S., Marín-Hernández, A., Gallardo-Pérez, J.C., Carreño-Fuentes, L., Moreno-
Sánchez, R., 2009. Targeting of cancer energy metabolism. Mol. Nutr. Food Res. 53, 29-48.       
Robustelli della Cuna, G., Pedrazzoli, P., 1991. Toxicity and clinical tolerance of lonidamine. Semin. 
Oncol. 18, 18-22. 
Saavedra, E., Encalada, R., Pineda, E., Jasso-Chávez, R., Moreno-Sánchez, R., 2004. Glycolysis in 
Entamoeba histolytica.Biochemical characterization of recombinant glycolytic enzymes and flux control 
analysis. FEBS J. 272, 1767-1783. 
Schwimmer, J., Essner, R., Patel, A., Jahan, S.A., Shepherd, J.E., Park, K., Phelps, M.E., Czernin, J., 
Gambhir, S.S., 2000. A review of the literature for whole-body FDG PET in the management of patients 
with melanoma. Q. J. Nucl. Med. 44, 153-67. 
Simonnet, H., Alazard, N., Pfeiffer, K., Gallou, C., Beroud, C., Demont, J., Bouvier, R., Schagger, H., 
Godinot, C., 2002.  Low mitochondrial respiratory chain content correlates with tumor aggressiveness 
in renal cell carcinoma.  Carcinogenesis 23, 759-768. 
Stalmans, W., Laloux, M., Hers, H.G., 1974. The interaction of liver phosphorylase a with glucose and 
AMP. Eur J Biochem. 49, 15-27.Tiedge, M., Richter, T., Lenzen, S., 2000. Importance of cysteine 
residues for the stability and catalytic activity of human pancreatic beta cell glucokinase. Arch Biochem 
Biophys. 375, 251-60. 
Torres, N.V., Mateo, F., Sicilia, J., Meléndez-Hevia, E., 1988. Distribution of the flux control in convergent 
metabolic pathways: theory and application to experimental and simulated systems. Int. J. Biochem. 
234, 169-174. 
Trejo-Solís, C., Palencia, G., Zuñiga, S., Rodríguez-Ropon, A., Osorio-Rico, L., Sánchez-Torres, L., 
Gracia-Mora, I., Marquez-Rosado, L., Sánchez, A., Moreno-García, M.E., Cruz, A., Bravo-Gómez, 
M.E., Ruiz-Ramírez, L., Rodríguez-Enríquez, S., Sotelo, J., 2005. Cas II gly induces apoptosis in 
glioma C6 cells in vitro and in vivo through caspase-dependent and caspase-independent 
mechanisms. Neoplasia 7, 563-574. 
Vasta, V., Bruni, P., Farnararo, M., 1987. Mechanism of thrombin-induced rise in platelet fructose 2,6-
bisphosphate content. Biochem. J. 244, 547-551.   
Xu, R., Pelicano, H., Zhou, Y., Carew, J.S., Feng, L., Bhalla, K.N., Keating, M.J., Huang, P., 2005. 
Inhibition of glycolysis in cancer cells: a novel strategy to overcome drug resistance associated with 
mitochondrial respiratory defect and hypoxia. Cancer Res. 65, 613-621. 
Yin, J., Gao, Z., Liu, D., Liu, Z., Ye, J., 2008. Berberine improves glucose metabolism througth induction 
of glycolysis. Am. J. Physiol. Endocrinol. Metab. 295, E148-156. 
 
 
 
 119 
Figure legends 
 
Figure 1. Effect of CasII-gly on glycolysis in AS-30D hepatocarcinoma cells  
(A) AS-30D carcinoma cells (15 mg protein/ ml) were incubated in the presence (●, ■) or in the absence 
(○, □) of 100 µM CasII-gly in Krebs-Ringer buffer containing 5 mM glucose (+ GLU) or without glucose (-
GLU).  After 10, 20, 30, and 60 min, lactate (Lac) content was enzymatically determined.  (B) Net lactate 
(Lac) production and (C) net variation in glycogen (Glyc) content were determined by subtracting values 
attained in the absence (Lac, Glyc) from values in the presence of 100 µM CasII-gly (LacCasII-gly; Glyc CasII-
gly) in AS-30D cells incubated for 30 and 60 min.  Lactate is expressed in nmol/mg cellular protein and 
glycogen is expressed in nmol of glucose equivalents/mg cellular protein.  Data show the mean ± SD of 
three different cellular preparations.  * P< 0.05, **P<0.01,***P< 0.005 vs. LacCasII-gly-Lac or GlycCasII-gly- Glyc 
values is cero. 
 
Figure 2. Effect of CasII-gly on HK and PFK-1 activities and on flux through the initial glycolytic segment  
AS-30D tumor cells were incubated for 60 min in the presence or in the absence of 100 M CasII-gly in 
Ringer-Krebs buffer with (+Glu) or without (-Glu) 5 mM glucose.  Afterwards, cellular extracts were 
obtained as described in the Material and Methods section.  (A) 100 % activities of hexokinase (HK) and 
phosphofructokinase-1 (PFK-1) corresponds to 515 ± 140 and 152 ± 16 nmol/min/mg protein in the 
condition + GLU; 429 ± 96 and 140 ± 33 nmol/min/mg protein in condition - GLU, respectively. (B)  
Metabolic fluxes from HK to TPI in AS-30D cells (+GLU) treated with or without 100 µM CasII-gly.  For the 
HK-TPI flux, 2 mM glucose was added; for the HPI-TPI, PFK-1-TPI and ALDO-TPI fluxes, the reaction 
was started with 10 mM G6P, 5 mM F6P, and 1 mM F1,6BP, respectively.  100 % control corresponds to 
35 ± 25 (HK-TPI, GLU), 72 ± 57 (HPI-TPI, G6P), 153 ± 106 (PFK-TPI, F6P) and 438 ± 86 (ALD-TPI, 
F1,6BP) nmol/min/mg protein, respectively.  Data show the mean ± SD of three different preparations.*P< 
0.005 vs. no CasII-gly addition.  
 
Figure 3. Activities of mitochondrial and soluble HK and fluxes of the first glycolytic segment at low CasII-
gly concentration and 3BrPyr 
A) Effect of CasII-gly and 3BrPyr on mitochondrial bound (●) and cytosolic HK activity (□, ■).B) Inhibition 
of flux from HK to TPI and from HPI to TPI by 10 μM CasII-gly (black bars) C) Effect of 100 μM cisplatin 
(a); 10 μM CuCl2 (b) ; 20 μM CuCl2 (c); 10 μM CasII-gly (d) on cytosolic tumor HK activity.*P<0.005 vs A) 
CasII-gly condition, B) 100% flux; **P<0.05 vs 100% HK activity. 
 The extracts were pre-incubated for 5 min with the indicated drug concentration. Afterwards, 
mitochondrial and cytosolic HK activities and segment fluxes were determined by adding 5 mM glucose 
(HK-TPI) or 5 mM G6P (HPI-TPI).  In the absence of drug, the rates of mitochondrial bound- and cytosolic 
HKs were 1150 ± 996 and 648 ±  242 nmol NAD(P)H/min/mg protein; and the fluxes from HK to TPI and 
HPI to TPI were 55 ± 39 and 201 ± 103 nmol NAD(P)H/min/mg protein. Data show the mean ± SD of at 
least three different preparations. 
 
 
 120 
 
Table 1.- Potency of Cas-IIgly and 3BrPyr on tumor and normal cells. 
 IC50 (µM) 
Time 24 hr 72 hr 
 CasII-gly 3BrPyr CasII-gly 3BrPyr 
Rodent Tumor cells 
Glioma C6 0.75 (1) >100 (1) ND ND 
Human     
HeLa 5 ± 2 (4) > 100 (2) 0.66 (2) 48 (2) 
MCF-7 5.7 ± 1 (3) 78 ± 15 (3) 0.76 ± 0.09 (3) 69 ± 27 (3) 
PC3 6.7 (2) >100 (2) 0.52 (2) >100 (2) 
SKLU 3.6 (2) 79 (2) 0.86 (2) 45 (2) 
U87MG 5.7 (2) 74 (2)  1.9 (2) 79 (2) 
HTC15 5.1 (1) 68 (1) 0.74 (1) 58 (1) 
Human Normal Cells 
Lymphocytes >100 (3)  >100 (2) >10 (3) >10 (3) 
Endothelial cells from 
umbilical cords  
>100 (2) 
 
>100 (3) 27 (1) 
 
>100 (2) 
 
 
 
 
 121 
Table 2 Effect of CasII-gly on the content of glycogen, glycolytic intermediates and ATP in AS-30D 
hepatocarcinoma  
Values are expressed in nmol/mg protein and  represent the mean ± SD; number of different preparations assayed is 
shown in parentheses.  N.D., not detected *P < 0.01 or **P< 0.005  vs. no CasII-gly at  30 or 60 min. . 
 
 + GLU - GLU 
 30 min 60 min 30 min 60 min 
  + Cas-IIgly  + Cas-IIgly  + Cas-IIgly  + Cas-IIgly 
Glycogen 133 ± 8 (4) 105 ± 6* (4) 152 ± 14 (4) 81 ± 18* (4) 73 ± 7 (3) 65 ± 6 (3) 69 ± 2 (3) 53 ± 5 * (3) 
G6P 0.4 ± 0.2 (3) 0.5 ± 0.3 (3) 0.5 ± 0.4 (6) 0.9 ± 0.6 (6) N.D. N.D. N.D. N.D. 
F6P 0.4 ± 0.3 (5) 0.7 ±0.5(5) 0.2 ± 0.1 (6) 0.4 ± 0.2 (6) N.D. N.D. N.D. N.D. 
F1,6BP 1 ± 0.8 (5) 2 ± 1.1(3) 0.4 ± 0.1(4) 0.6 ± 0.3 (3) N.D. N.D. N.D. N.D. 
TriP 2 ± 1.1 (4) 1 ± 0.4(3) 0.6 ± 0.3 (3) 0.7 ± 0. 2(4) N.D. N.D. N.D. N.D. 
ATP 9 ± 2.2 (3) 7 ± 1.1(3) 10 ± 1.5 (4) 4 ± 1.3**(4) 11 ± 0.7 (3) 7 ± 0.8**(3) 9± 2.6 (4) 2 ± 0.3 **(3) 
 
 
 122 
Table 3. Potency of anti-neoplastic drugs on tumor and non-tumor cytosolic HK  
 IC50 (μM) 
CasII-gly 3BrPyr 
                               Neoplasia 
              Rodent    
AS-30D 4 ± 1 (4) 28 ± 8 (4)*  
Glioma C6 17 ± 2 (3) 41 ± 8 (3)** 
              Human   
HeLa 16 (2) 34 (2) 
PC-3  13 (2) 49 (2) 
MDA-MB-231 19 (2) >100 (3) 
MCF-7   14 (2) 24 (2) 
SK-LU      6 (2) 7.5 (2) 
TD47 20 (1) 50 (1) 
No tumorigenic tissues 
Rat brain (HKI) 6 ± 0.6 (3) 47 ± 13 (3)** 
Rat heart (HKII) 9 ± 0.6 (3) 36 ± 13 (3)*** 
Rat liver  (HKIV) 2 ± 0.8 (3) 43 ± 20 (3)*** 
Human platelets 4  ± 2 (3) 25  ± 8 (3)** 
 
The values represent the mean ± SD; number of different preparations assayed is 
shown in parentheses.   *P< 0.005, **P< 0.01 and ***P< 0.05 vs. CasII-gly.   
 
 
 
 
 
 123 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
10 20 30 40 50 60
0
100
200
300
400 *
- GLU
+ GLU
 Time (min)
L
a
c
A
B
Figure 1
-80
-60
-40
-20
0
**
 
 
 
Time (min)
***
***
3060 6030
G
ly
c
C
as
II-
g
ly
-G
ly
c
C
0
25
50
75
100 ***
+ GLU- GLU
****
L
ac
C
as
II-
g
ly
-L
ac
 
:::::::::::::::::-; 
 
 
 124 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
0
50
100
*
*
      GLU  -   -   +   +       -   -   +   +
CasII-gly -   +   -   +       -   +   -   +  
PFKHK
 
A
ct
iv
it
y 
(%
)
A
B
0
25
50
75
100
*
 + CasII-gly
 F1,6BPF6PG6PGLU
F
lu
x 
(%
)
Figure 2
..
 
 
 125 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Figure 3
0
25
50
75
100
CasII-glyCu 2+
**
*
 
a b
c
d
H
K
 a
ct
iv
it
y 
(%
)
Cisplatin
A
B
C
0
25
50
75
100
*
HK-TPI
F
lu
x
 (
%
)
HPI-TPI
0 3 5 8 10
40
60
80
100
+ CasII-gly
+ BrPyr
*
H
K
 a
ct
iv
it
y 
(%
)
Drug (M)
H
K
 a
ct
iv
it
y 
(%
)
F
lu
x
 (
%
)
H
K
 a
ct
iv
it
y 
(%
)
-~ 
---i 
I 
1 
--, 
-
.L 
~ .. 
 
 
 126 
5.3.1 Comentarios manuscrito No. 4 
Nuestro objetivo principal del manuscrito No. 4 era el determinar que la inhibición 
de la HK se traducía en una reducción en el flujo glucolítico y el de proponer a la CasII-
gly como un fármaco antineoplásico que pueda específicamente reducir la proliferación 
celular en las células tumorales debido a que afecta su metabolismo energético. Sobre 
todo porque fue mas potente que el 3-bromopiruvato (inhibidor de la HK) entre 1-21 
veces al inhibir la actividad de la HK y entre 13 -91 veces al inhibir  la proliferación 
celular.  Sin embargo, a pesar de haber demostrado que el fármaco inhibía específica y 
significativamente la actividad de la enzima, no se observó una reducción en la 
producción de láctico, debido a que se inducía una activa degradación de glucógeno. 
Estos resultados nos condujeron a concluir que con solo inhibir a esta enzima no se iba 
a lograr reducir el flujo glucolítico debido a que la vía metabólica, y la célula, tienen 
alternativas que disminuyen los estragos que pueda producir su inhibición. Por 
consiguiente, se tendrán mejores resultados si se inhiben al mismo tiempo al resto de 
las enzimas que controlan la glucólisis (GLUT y HPI). 
 
 
 
127 
 
Capítulo 6 
 
6.1 Discusión general  
 
En esta tesis, nuestra propuesta inicial consideraba que el control de la glucólisis 
tumoral no debería recaer en la HK y la PFK-1, a diferencia de lo que ocurre en las 
células normales en donde son los principales sitios de control, debido a los cambios en 
sus mecanismos de regulación y a sus niveles de sobre-expresión en células tumorales. 
Sin embargo, nuestros resultados (obtenidos en AS-30D y HeLa; Manuscrito No. 2) 
apoyaron parcialmente nuestra propuesta original, ya que únicamente en el caso de la 
PFK-1 se observó una reducción en su control, lo que atribuímos a que se encuentran 
concentraciones intracelulares saturantes de la F2,6BP (4-6 µM, Ka= 0.2-0.8 µM), el 
AMP y el Pi, activadores de la enzima que, además, bloquean la inhibición que ejercen 
el citrato y el ATP, y moderadamente el H+ (datos no mostrados). Lo anterior concuerda 
con observaciones previas que sugerían que la enzima se encuentra muy activa en las 
células tumorales (Loiseau et al., 1985; Staal et al., 1987; Marín-Hernández et al., 
2006), motivo por el cual se observa una pérdida en el efecto Pasteur (Eigenbrodt et al., 
1985; Rodríguez-Enríquez et al., 2001). 
Por otra parte, la HK mantiene un control significativo sobre la glucólisis tumoral 
debido a que, independientemente de su sobre-expresión (hasta en 300 veces en AS-
30D) o su localización (membrana externa mitocondrial), su producto la G6P la inhibe 
fuertemente (Marín-Hernández et al., 2006; Manuscrito No. 2); el aumento en la 
actividad de la HK en células cancerosas trae como consecuencia un aumento 
concomitante en su producto, la G6P. Aparentemente, el que esta enzima se mantenga 
regulada de esta manera evita la disminución en los niveles de ATP intracelular.  
 
 
128 
 
 Además, en esta tesis logramos establecer que en el control de la glucólisis 
pueden participan otros sitios como el GLUT, la HPI y la degradación de glucógeno, los 
cuales fueron visualizados por medio del modelado de la glucólisis, una vez que fueron 
construidos los modelos cinéticos de esta vía (ver sección 5.2; Manuscrito No. 3). Los 
modelos lograron predecir los flujos y las concentraciones de metabolitos encontradas 
en las células intactas. Sin embargo, su principal debilidad fue la predicción de la 
concentración de F6P, la cual fue 50 veces menor con respecto a lo detectado 
experimentalmente. Esto lo atribuimos a que en los modelos se simula a una PFK-1 
muy rápida que reduce la concentración de F6P. Entonces consideramos que puede 
existir un freno cinético que aún no se ha identificado. Por ejemplo, en la PFK de 
músculo de conejo se ha reportado que el PEP, el 3PG y la creatina fosfato pueden 
incrementar el efecto inhibitorio del ATP (Colombo et al., 1975). Esto último sugiere la 
necesidad de caracterizar el efecto de éstos y otros metabolitos sobre la enzima 
tumoral. 
Por un trabajo previo de nuestro grupo de trabajo se determinó que en las células 
tumorales AS-30D predomina el GLUT3 mientras que en HeLa predomina el GLUT1 
(Rodríguez-Enríquez et al., 2009). En ambos tipos de células el GLUT ejerce un control 
significativo debido a su baja actividad con respecto al resto de los componentes de la 
vía, pero además, a ésto se suma la expresión de una isoforma del GLUT con una baja 
afinidad por la glucosa (Km = 9 mM) en el caso de las células HeLa.  
Otra aportación importante de nuestro trabajo de tesis (Manuscrito No. 3) fue el 
de proponer a la HPI (una de las enzimas más rápidas de la glucólisis) como un sitio 
que controla el flujo glucolítico, debido a la fuerte y múltiple inhibición competitiva que 
ejercen sobre ella la F1,6 BP (intermediario de la glucólisis), la eritrosa-4P y el 6-
 
 
129 
 
fosfogluconato (ambos intermediarios de la vía de las pentosas). Aunque cabe señalar 
que el efecto inhibitorio de estos metabolitos ya se había descrito previamente (Zalitis 
and Oliver, 1967; Chirgwin et al., 1975; Gaitonde et al., 1989), no se había evaluado la 
importancia que puede tener la inhibición de esta enzima en el control de la glucólisis. 
Al mismo tiempo, sugerimos que la distribución de control puede verse 
influenciada por las condiciones en las que crecen las células tumorales, que pueden 
influir en la expresión de las enzimas y los transportadores de la glucólisis (Yun et al., 
2005; Natsuizaka et al., 2007). Así, la degradación de glucógeno y el GLUT mantienen 
un alto coeficiente de control en las células HeLa, debido a que al cultivarse en 25 mM 
de glucosa, las células mantienen un alto contenido de glucógeno y además expresan 
una isoforma del GLUT con baja afinidad por glucosa (GLUT1, Km= 9 mM). En cambio, 
las células AS-30D que crecen en el espacio intraperitoneal de la rata, en donde se ven 
sometidas a una baja concentración (0.026 mM) de glucosa (Rodríguez-Enríquez et al., 
2000) expresan un GLUT con alta afinidad (GLUT3, Km= 0.5 mM) y mantienen una baja 
concentración de glucógeno, que propicia que la glucólisis dependa únicamente de la 
glucosa exógena y  por lo tanto el GLUT, la HK y la HPI controlen.  
Otro aspecto interesante que debe señalarse es que aunque el control que ejerce 
la GAPDH sobre el flujo glucolítico no es significativo, éste depende de la concentración 
de fosfato libre presente en la célula, ya que puede limitar su actividad debido a la baja 
afinidad que tiene la enzima por éste (Heinz and Kulbe, 1970; Patra et al., 2009), lo cual 
no se ha considerado al analizar la glucólisis en otros organismos (levadura y 
tripanosoma) (Bakker et al., 1999; Teusink et al., 2000),  en los que se supone que el 
fosfato no tiene ninguna relevancia en la actividad de esta enzima porque no limita su 
actividad. 
 
 
130 
 
Una vez que se han analizado los resultados obtenidos en esta tesis, la pregunta 
obvia es ¿Cuál es la enzima que se necesita inhibir para bloquear la glucólisis en las 
células tumorales? Como parte de la respuesta tenemos que el GLUT, la HK y la HPI 
pueden ser considerados como sitios adecuados para inhibir la glucólisis. En este caso 
no se propone inhibir la degradación de glucógeno porque el glucógeno se agotara 
eventualmente. Sin embargo, como único blanco específico para la glucólisis tumoral, la 
HK y el GLUT quedarían descartadas, ya que estas proteínas participan controlando la 
glucólisis en las células normales junto con la fosfofructocinasa tipo I (PFK-I), lo que 
deja como única opción de un posible blanco terapéutico a la HPI. Aunque al inhibir sólo 
a esta enzima la reducción en el flujo glucolítico no sería tan potente como lo sería 
inhibir al mismo tiempo a esta enzima junto con el GLUT y la HK, pero sí mayor con 
respecto a inhibir a enzimas que no controlan como la PFK-1, como lo predicen 
nuestros modelos. A esto se suma que experimentalmente demostramos que sólo 
inhibiendo a una enzima como la HK no basta para reducir la glucólisis (Manuscrito No. 
4), porque la célula utiliza rutas alternativas para disminuir los estragos que pudo 
ocasionar la inhibición de esta enzima, lo cual se hubiera evitado al inhibir al mismo 
tiempo al GLUT y/o a la HPI. 
Por lo tanto debe considerarse la necesidad de buscar en otras vías (metabólicas 
o de señalización) blancos terapéuticos potenciales que permitan una terapia multisitio, 
es decir buscar aquellos sitios (enzimas o transportadores) que contribuyan 
importantemente en el control de sus vías pero que en las células normales su control 
no sea significativo (Moreno-Sánchez et al., 2010).  
En el campo de la biología del cáncer son pocos los estudios en los cuales se 
aplica el análisis de control metabólico para identificar blancos terapéuticos. Esto es 
 
 
131 
 
debido a que en general para el diseño de fármacos los grupos de investigación se 
enfocan en la búsqueda de un paso limitante, olvidando que este concepto no es 
aplicable en la mayoría de las vías metabólicas en las cuales más de una enzima puede 
controlar el flujo de la vía (Fell, 1997; Moreno-Sánchez et al., 2008; 2010). A esto se 
suma que en algunos casos la elección del blanco terapéutico se basa en la suposición 
de que es esencial en el crecimiento tumoral. Por lo tanto, se eligen sitios que no 
ejercen control, los cuales tienen que ver disminuida su actividad al menos un 80% para 
lograr observar cambios en el flujo de la vía que se desea inhibir. Esto provoca el 
empleo de concentraciones masivas de inhibidores que pueden conducir a  afectos no 
deseados. Por este motivo el uso de inhibidores glucolíticos no ha tenido éxito en el 
tratamiento del cáncer. Esto se evitaría al inhibir sitios que controlan y que tienen la 
ventaja de que no es necesario abatir por completo su actividad para causar estragos 
en el flujo de una vía, con lo cual se emplearían concentraciones bajas de los 
inhibidores reduciendo los efectos adversos (Moreno Sánchez et al., 2010). Por todas 
estas razones, nuestro trabajo de tesis resulta relevante. 
Finalmente, podemos indicar que el estudio de la glucólisis tumoral nos ha 
permitido determinar y entender los mecanismos bioquímicos involucrados en el control 
de esta vía, lo que nos permitirá extender nuestro estudio a otros tipos de cáncer para 
poder ayudar en la selección de nuevos blancos terapéuticos, y con ello el diseño de 
inhibidores más potentes y específicos. 
 
 
 
132 
 
6.2 Conclusiones generales 
     
1. La HK ejerce un control significativo en la glucólisis tumoral debido a que sigue 
siendo inhibida fuertemente por su producto la G6P. 
2. La PFK-1 no controla a causa de que la concentración intracelular de F2,6BP es 
suficiente para bloquear el efecto inhibitorio del citrato y el ATP. 
3. El GLUT, la HPI y la degradación de glucógeno pueden participar en el control de 
la glucólisis. 
4. La distribución de control depende del grado de expresión de las enzimas que 
controlan y de sus mecanismos de regulación así como del flujo de síntesis- 
degradación de glucógeno. 
5. El análisis de control metabólico (mediante el empleo del análisis de 
elasticidades y el  modelado cinético) nos permitió identificar  a las enzimas que 
controlan la glucólisis así como entender los mecanismos bioquímicos 
involucrados. Entre estos últimos se destaca la inhibición de la HPI por 
metabolitos de la glucólisis y de la vía de las pentosas.       
6. La inhibición de la HK no conduce a una reducción en el flujo de la glucólisis 
debido a que la célula usa el glucógeno para evitar los estragos que ocasiona la 
inhibición, a que la HK controla sólo parcialmente, y a que las otras enzimas 
controladoras compensan la disminución en la actividad de la HK y ajustan el 
flujo.  
 
 
133 
 
6.3 Perspectivas 
En nuestro estudio establecimos que la distribución del control puede verse 
modificada como consecuencia de las condiciones en las que crecen las células y 
aunque con nuestros modelos cinéticos hicimos algunas aproximaciones al estudiar el 
efecto de la hipoxia y la hipoglucemia, una perspectiva de nuestro trabajo de tesis es el 
de explorar detalladamente los efectos que tiene sobre las células tumorales la 
exposición prolongada a la hipoxia y a la hipoglucemia. Esto es relevante si recordamos 
que en los tumores sólidos las células se ven sometidas a bajas concentraciones de 
oxígeno y de nutrientes (glucosa) por largos periodos de tiempo. Experimentalmente se 
han simulado estas condiciones en el laboratorio y otros grupos de investigación han 
observado un incremento en la expresión (RNAm) de los genes de algunas de las 
enzimas que hemos demostrado controlan la glucólisis como el GLUT y la HK, como 
consecuencia de la activación del HIF-1α y de la AMPK (Yun et al., 2005; Natsuizaka et 
al., 2007). Pero lamentablemente en estos estudios no se consideró evaluar los 
cambios en la actividad de las enzimas y el flujo de la glucólisis, por lo cual la pregunta 
que nos hacemos es: ¿Cómo se modifica el control de la glucólisis en hipoxia e 
hipoglucemia prolongada? ¿O no se modifica?  
Para contestar la pregunta sería necesario someter a células HeLa en monocapa 
a periodos prolongados de hipoxia e hipoglucemia. En estas células habría que 
determinar las actividades del GLUT, HK y HPI, así como las concentraciones de sus 
moduladores y la concentración de glucógeno. Estos valores se vaciarían en los 
modelos de la glucólisis que hemos construido para establecer cuál es la distribución de 
control.  Dado que los reportes indican que bajo estas condiciones  el GLUT3 
(transportador con alta afinidad por glucosa) llega a incrementar su expresión junto con 
 
 
134 
 
la HKII y si el contenido de glucógeno es bajo, se esperaría que las enzimas que 
controlen el flujo bajo estas condiciones sean, otra vez, el GLUT, la HK y la HPI como lo 
determinamos en las células AS-30D que están sometidas a bajas concentraciones de 
glucosa durante su crecimiento. Estos posibles resultados demostrarían que el control 
de una vía metabólica es robusto o, en otras palabras, el control es constante bajo 
diversas condiciones experimentales.  Esta conclusión es relevante para el diseño de 
terapias anti-cancer, pues permitiría establecer con certeza los mejores sitios para 
inhibición específica.  
Otro punto interesante es determinar cómo influyen en la glucólisis y en la 
distribución de control la activación de oncogenes y la falta de genes supresores de 
tumores,  considerando que cada tipo de cáncer es una enfermedad diferente debido a 
que se han desarrollado por haberse apagado algunos genes supresores de tumores y 
al mismo tiempo se han encendido algunos oncogenes que sabemos pueden modificar 
la expresión de los componentes de la glucólisis. Por ello  una perspectiva interesante 
es determinar si en cada tipo de cáncer existe una diferencia en la distribución del 
control y si esta diferencia es atribuida a la expresión de factores oncogénicos 
específicos.  
Con este tipo de estudios parecería difícil conseguir resultados claros en líneas 
de células tumorales establecidas, pues varios oncogenes pueden encontrase activos al 
mismo tiempo y varios genes supresores de tumores inactivos. Entonces, la estrategia 
experimental sería emplear una línea celular no tumoral en la cual, por medio del  
empleo de siRNAs, apagar genes supresores de tumores o en su caso activar 
oncogenes (como se ha hecho para c-Myc o H-Ras en fibroblastos de rata) y evaluar 
los efectos sobre la glucólisis y la distribución de control. Se puede esperar que 
 
 
135 
 
aquellos factores que modifiquen la expresión de la HK y el GLUT puedan modificar la 
distribución de control.       
Por otra parte, como nuestros  resultados (Manuscritos No. 3 y 4) sugieren que 
para reducir el flujo glucolítico drásticamente es necesario inhibir al mismo tiempo a las 
enzimas que controlan (GLUT, HK y HPI), y como  hasta el momento no se conocen  
inhibidores específicos de estas  enzimas, se plantea el diseño de oligonucleótidos 
antisentido (que son secuencias específicas de un gen de interés que se unen al RNA 
mensajero de este gen, de tal forma que es inhibida la traducción y no hay proteína) 
para inhibir específicamente al GLUT3, a la HKII y a la HPI.  Nuestro modelo para 
implementar estos experimentos serían las células AS-30D en las cuales esperaríamos 
una reducción drástica en el flujo glucolítico. Al mismo tiempo podríamos emplear estos 
oligonucleótidos antisentido en un modelo de célula no tumoral (hepatocitos) en el que 
se esperaría que debido a que el control  sobre el flujo glucolítico no lo ejercen las 
mismas proteínas (GLUT, HPI y HK) no se vería disminuida la glucólisis.    
Consideramos además que debe caracterizarse la glucólisis en modelos como 
las células rho y las células troncales. Esto debido a que las células rho son 
predominantemente glucolíticas ya que su mitocondria es disfuncional al carecer de 
DNA mitocondrial y podrían simular aquellas células tumorales que mantienen una 
reducción en la actividad de la mitocondria por el elevado número de mutaciones que 
presentan en su DNA mitocondrial (Carew y Huang, 2002 ). Mientras que el estudio de 
la glucólisis en las células troncales es relevante porque estas células son consideradas 
las precursoras de algunos tipos de cáncer (leucemias y cáncer de mama). Por lo cual,  
la caracterización de esta vía en ambos modelos con la aplicación del análisis de 
control metabólico podría permitir la identificación de blancos terapéuticos.   
 
 
136 
 
Finalmente, considerando que la muerte de la célula tumoral puede lograrse al 
disminuir los niveles de ATP y ésto sólo puede ocurrir al inhibir a la glucólisis y la 
fosforilación oxidativa, y aunque podría pensarse que en los tumores sólidos la 
inhibición de la fosforilación oxidativa no sería tan importante, hay reportes que indican 
que en los tumores sólidos pueden encontrarse poblaciones de células en las que 
predomina el metabolismo glucolítico por encontrarse lejos de los vasos o en zonas 
hipóxicas y otras en las que predomina el metabolismo mitocondrial por localizarse 
cerca de los vasos sanguíneos (Sonveaux et al., 2008). Por lo tanto nuestro modelo 
cinético se puede extender e incluir al ciclo de Krebs y a la fosforilación oxidativa para 
poder determinar los sitios que controlan el flujo de cada una de estas vías metabólicas 
y proponerlos para una terapia multisitio. El modelo experimental idóneo para poder 
encontrar estas dos poblaciones de células al mismo tiempo es el esferoide, que simula 
a un tumor en sus primeras etapas de desarrollo. En este modelo se pueden separar 
ambas poblaciones, lo que nos puede permitir hacer la caracterización de cada una de 
las vías metabólicas y obtener los datos suficientes para construir los modelos. En este 
modelo se ha reportado un incremento en la expresión de la HK y el GLUT en la 
población glucolítica con respecto a la población oxidativa (Rodríguez-Enríquez et al., 
2008), por lo cual en las células oxidativas podría esperarse que el control de la 
glucólisis lo ejercieran el GLUT y la degradación de glucógeno (como ocurre en las 
células HeLa) mientras en la población glucolítica se podría esperar que fueran el 
GLUT, la HK y la HPI (como ocurre en AS-30D). Por otra parte aunque se ha estudiado 
el control de la fosforilación oxidativa en tumores es difícil establecer qué enzimas 
podrían ejercer el control sobre la fosforilación oxidativa dado que existen pocos datos 
sobre los flujos y actividades de las enzimas que participan en esta vía.   
 
 
137 
 
6.4 Referencias 
 Arora KK, Pedersen PL (1988) Functional significance of mitochondrial bound 
hexokinase in tumor cell metabolism. J. Biol. Chem. 263, 17422-17428. 
 Bakker BM, Walsh MC, Ter Kuile BH, Mensonides FIC, Michels PAM, Opperdoes 
FR, Westerhoff HV (1999) Contribution of glucose transport to the control of the 
glycolytic flux in Trypanosoma brucei, Proc. Natl. Acad. Sci. USA 96, 10098–10103. 
 Bartrons R, Caro J (2007) Hypoxia, glucose metabolism and the Warbur´s effect. J. 
Bioenerg. Biomembr. 39, 223-229. 
 Bauer DE, Hatzivassiliou G, Zhao F, Andreadis C, Thompson CB (2005) ATP citrate 
lyase is an important component of cell growth and transformation. Oncogene 24, 
6314–6322. 
 Brown GC, Lakin-Tomas L, Bran MD (1990) Control of respiration and oxidative 
phosphorylation in isolated rat liver cells. Eur. J. Biochem. 192, 355-362. 
 Bustamante E, Pedersen PL (1977) High aerobic glycolysis of rat hepatoma cells in 
culture: role of mitochondrial hexokinase. Proc. Natl. Acad. Sci. USA 74, 3735-3739. 
 Carew JS, Huang P (2002) Mitochondrial defects in cancer.  Mol. Cancer 1, 1-9. 
 Chirgwin JM, Parsons TF, Ernst AN (1975) Mechanistic implications of the pH 
independence of inhibition of phosphoglucose isomerase by neutral sugar 
phosphates, J. Biol. Chem. 250, 7277-7279. 
 Clem B, Telang S, Clem A, Yalcin A, Meier J, Simmons A, Rasku MA, Arumugam S, 
Dean WL, Eaton J, Lane A, Trent JO, Chesney J (2008) Small-molecule inhibition of 
6-phosphofructo-2-kinase activity suppresses glycolytic flux and tumor growth. Mol. 
Cancer Ther. 7, 110-1120. 
 Colombo G, Tate PW, Girotti AW, Kemp RG (1975) Interaction of inhibitors with 
muscle phosphofructokinase. J. Biol. Chem. 250, 9404-12. 
 Colomer D, Vives-Corrons JL, Pujades A, Bartrons R (1987) Control of 
phosphofructokinase by fructose 2,6 bisphosphate in B-lymphocytes and B-chronic 
lymphocytic leukemia cells. Cancer Res. 47, 1859-1862. 
 Dang CV, Semenza GL (1999) Oncogenic alterations of metabolism. TIBS. 24, 68-
72. 
 DeBerardinis RJ, Lum JJ, Hatzivassiliou G, Thompson CB (2008) The biology of 
cancer: metabolic reprogramming fuels cell growth and proliferation. Cell Metab. 7, 
11-20.  
 Dunaway GA, Kasten TP, Sebo T, Trapp R (1988) Analysis of the 
phosphofructokinase subunits and isoenzymes in human tissues. Biochem. J. 251, 
677-683.  
 Eigenbrodt E, Fister P, Reinacher M (1985) New perspectives in carbohydrate 
metabolism in tumor cells. In. Regulation of carbohydrate metabolism (Reitner, R, 
ed.), vol. 2, pp 141-179. CRC Press, Boca Raton, FL. 
 
 
138 
 
 El-Bacha T, Sampaio de Freitas M, Sola-Penna M (2003) Cellular distribution of 
phosphofructokinase activity and implications to metabolic regulation in human 
breast cancer. Mol. Gen. Metab. 79, 294-299. 
 Evans A, Bates V, Troy H, Hewitt S, Holbeck S, Chung YL, Phillips R, Stubbs M, 
Griffiths J,  Airley R (2008) GLUT1 as a therapeutic target: increased 
chemoresistance and HIF-1-independent link with cell turnover is revealed through 
compare analysis and metabolomic studies. Cancer Chemother. Pharmacol. 61, 
377-393.  
 Fanciulli, M., Bruno, T., Giovannelli, A., Gentile, F.P., Di Padova, M., Rubiu, O. and 
Floridi, A (2000) Energy metabolism of human LoVo colon carcinoma cells: 
correlation to drug resistance and influence of Lonidamine. Clin. Cancer Res. 6, 
1590-1597. 
 Fell D (1997) Understanding the control of metabolism. Plenum Press, London. 
 Fisher K, Hoffmann P, Voelkl S, Meidenbauer N, Ammer J, Edinger M, Gottfried E, 
Schwarz S, Rothe G, Hoves S, Renner K, Timischl B, Mackensen A, Kunz-Schughart 
L, Andreesen R, Krause SW, Kreutz M (2007) Inhibitory effect of tumor cell-derived 
lactic acid on human T cells. Blood 109, 3812-3819.  
 Gaitonde MK, Murray E, Cunningham VJ (1989) Effect of 6-phosphogluconate on 
phosphoglucose isomerase in rat brain in vitro and in vivo. J. Neurochem. 52, 1348-
52. 
 Gatenby R, Gillies RJ (2004) Why do cancers have high aerobic glycolysis?. Nature 
Rev. Cancer  4, 891-899. 
 Giatromanolaki A, Sivridis E, Gatter KC, Turley H, Harris AL, Koukourakis MI (2006) 
Lactate dehydrogenase 5 (LDH-5) expression in endometrial cancer relates to the 
activated VEGF/VEGFR2(KDR) pathway and prognosis. Gynecol. Oncol. 103, 912-8. 
 Gillies RJ, Robey I, Gatenby RA (2008) Causes and consequences of increased 
glucose metabolism of cancer. J. Nucl. Med. 49, 24S-42S.   
 Groen AK, van Roermund CW, Vervoom RC, Tager JM (1986) Control of 
gluconeogenesis in rat liver cells. Flux control coefficients of the enzymes in the 
gluconeogenic pathway in the absence and presence of glucagon. Biochem. J. 237, 
379-389. 
 Heinz F, Kulbe KD (1970) Glyceraldehydephosphate dehydrogenase from liver. I. 
Isolation and characterization of the bovine liver enzyme. Hoppe Seylers Z Physiol 
Chem. 351, 249-62. 
 Horberg JJ, Bruggeman, FJ, Bakker BM, Westerhoff HV (2007) Metabolic control 
analysis to identify optimal drug targets. Progress in Drug Research, Vol 64. 
 INEGI (2009) Estadísticas a propósito del día mundial contra el cáncer. 
www.inegi.org.mx. 
 
 
139 
 
 Jannaschk KD, Burgos M, Centerlles JJ, Ovadi J, Cascante M (1999) Application of 
metabolic control analysis to the study of toxic effects of copper in muscle glycolysis. 
FEBS Lett. 445, 144-148.  
 Jones RG and Thompson CB (2009) Tumor suppresors and cell metabolism: a 
recipe for cancer growth. Genes Dev.  23, 537-548. 
 Kashiwaya Y, Sato K, Tsuchiya N, Thomas S, Fell DA, Veech RL, Passonneau JV 
(1994) Control of glucose utilization in working perfused rat heart. J. Biol. Chem. 269, 
25502-25514. 
 Kim WJ, Dang CV (2006) Cancer´s molecular sweet tooth and the Warburg effect. 
Cancer Res. 66, 8927-8930.  
 Kosow DP, Rose I (1968) Ascites tumor mitochondrial hexokinase II. Effect of 
binding on kinetic properties. J. Biol. Chem. 243, 3623-3630. 
 Kumagai S, Narasaki R, Hasumi K (2008) Glucose-dependent active ATP depletion 
by koningic acid kills high-glycolytic cells. Biochem Biophys. Res. Commun. 365, 
362-368. 
 Lardner A (2001) The effects of extracellular pH on immune function. J. Leukoc. Biol. 
69, 522-530. 
 Lee K A, Roth RA, LaPres JJ (2007) Hypoxia, drug therapy and toxicity. Pharmacol. 
Ther. 113, 229-46. 
 Liu Y, Zuckier LS, Ghesani NV (2010) Dominant uptake of fatty acid over glucose by 
prostate cells: a potential new diagnostic and therapeutic approach. Anticancer Res. 
30, 369-74. 
 Loiseau AM, Rousseau GG, Hue L (1985) Fructose 2,6-bisphosphate and the control 
of glycolysis by glucocorticoids and by other agents in rat hepatoma cells. Cancer 
Res. 45, 4263-9. 
 Marín-Hernández A, Gallardo-Pérez JC, Ralph SJ, Rodríguez-Enríquez S, Moreno-
Sánchez R (2009) HIF-1alpha modulates energy metabolism in cancer cells by 
inducing over-expression of specific glycolytic isoforms. Mini Rev. Med. Chem. 9, 
1084-101. 
 Marín-Hernández A, Rodríguez-Enríquez S, Vital-González PA, Flores-Rodríguez 
FL, Macías-Silva M, Sosa-Garrocho M, Moreno-Sánchez R (2006) Determining and 
understanding the control of glycolysis in fast-growth tumor cells. Flux control by an 
over-expressed but strongly product-inhibited hexokinase. FEBS J. 273, 1975-88. 
 Maschek G, Savaraj N, Priebe W, Braunschweiger P, Hamilton K, Tidmarsh GF, De 
Young LR, Lampidis TJ (2004) 2-deoxy-D-glucose increases the efficacy of 
adriamycin and paclitaxel in human osteosarcoma and non-small cell lung cancers in 
vivo. Cancer Res. 64, 31-4. 
 Meldolesi MF, Macchia V, Laccetti P (1976) Differences in phosphofructokinase 
regulation in normal and tumor rat thyroid cells. J. Biol. Chem. 251, 6244-6251. 
 
 
140 
 
 Moreno-Sánchez R, Rodríguez-Enríquez S, Marín-Hernández A, Saavedra E (2007) 
Energy metabolism in tumor cells. FEBS J. 274,1393-1418. 
 Moreno-Sánchez R, Saavedra E, Rodríguez-Enríquez S, Olín-Sandoval V (2008) 
Metabolic control analysis: a tool for designing strategies to manipulate pathways. J. 
Biomed. Biotech. 2008: 597913.  
 Moreno-Sánchez R, Saavedra E, Rodríguez-Enríquez S, Gallardo-Pérez JC, 
Quezada H, Westerhoff HV (2010) Metabolic control analysis indicates a change of 
strategy in the treatment of cancer. Mitochondrion 10, 626-39. 
 Nakashima RA, Paggi MG, Scott LJ, Pedersen PL (1988) Purification and 
characterization of a bindable form of mitochondrial bound hexokinase from the 
highly glycolytic AS-30D rat hepatoma cell line. Cancer Res. 48, 913-919.  
 Natsuizaka M, Ozasa M, Darmanin S, Miyamoto M, Kondo S, Kamada S, Shindoh M, 
Higashino F, Suhara W, Koide H, Aita K, Nakagawa K, Kondo T, Asaka M, Okada F, 
Kobayashi M (2007) Synergistic up-regulation of Hexokinase-2, glucose transporters 
and angiogenic factors in pancreatic cancer cells by glucose deprivation and 
hypoxia. Exp. Cell Res. 313, 3337-48. 
 Nelson DL, Cox MM (2005) Lehninger Principles of Biochemistry, 4 edición, Editorial 
Freeman. 
 Nishijima T, Nishina M, Fujiwara K (1997) Measurement of Lactate Levels in serum 
and bile using proton nuclear magnetic resonance in patients with hepatobiliary 
diseases: its utility in detection of malignancies. Jpn. J. Clin. Oncol. 27, 13-17. 
 Olovnikov IA, Kravchenko JE, Chumakov PM (2009) Homeostatic functions of the 
p53 tumor suppressor: Regulation of energy metabolism and antioxidant defense. 
Semi. Cancer Biol. 19, 32-41. 
 Oskam R, Rijksen G, Staal GEJ, Vora S (1985) Isozymic composition and regulatory 
properties of phosphofructokinase from well-differentiated and anaplastic medullary 
thyroid carcinomas of the rat. Cancer Res. 45, 135-142. 
 Osthus RC, Shim H, Kim S, Li Q, Reddy R, Mukherjee M, Xu Y, Wonsey D, Lee LA, 
Dang CV (2000) Deregulation of glucose transporter 1 and glycolytic gene 
expression by c-Myc. J. Biol. Chem. 275, 21797–21800. 
 Parry DM, Pedersen PL (1983) Intracellular localization and properties of particulate 
hexokinase in the Novikoff ascites tumor. J. Biol. Chem. 258, 10904-10912. 
 Patra S, Ghosh S, Bera S, Roy A, Ray S, Ray M (2009) Molecular characterization of 
tumor associated glyceraldehyde-3-phosphate dehydrogenase. Biochemistry (Mosc). 
74, 717-27. 
 Pedersen PL, Mathupala S, Rempel A, Geschwind  JF, Ko YH (2002) Mitochondrial 
bound type II hexokinase: a key player in the growth and survival of many cancers 
and an ideal prospect for therapeutic intervention. Biochem. Biophys. Acta 1555, 14-
20. 
 
 
141 
 
 Pedersen PL (1978) Tumor mitochondria and the bioenergetics of cancer cells. Prog. 
Exp. Tumor Res. 22, 190-274.   
 Pelicano H, Martin D S, Xu RH, Hung P (2006) Glycolysis inhibition for anticancer 
treatment. Oncogene 25, 4633-4646. 
 Plathow C and Weber WA (2008) Tumor cell metabolism imaging. J. Nucl. Med. 49, 
43S-63S. 
 Poolman MG, Assmus HE, Fell DA (2004) Applications of metabolic modeling to 
plant metabolism. J. Exp. Bot. 55,1177-1186.  
 Rapoport T, Heinrich R, Jacobash G, Rapoport S (1974) A linear steady-state 
treatment of enzymatic chains. A mathematical model of glycolysis of human 
erythrocytes. Eur. J. Biochem. 42, 107-120. 
 Robey RB, Hay N (2009) Is Akt the “Warburg kinase”?-Akt-energy netabolism 
interactions and oncogenesis. Semin. Cancer Biol. 19, 25-31.  
 Rodríguez-Enríquez S, Gallardo-Pérez JC, Avilés-Salas A, Marín-Hernández A, 
Carreño-Fuentes L, Maldonado-Lagunas V, Moreno-Sánchez R (2008) Energy 
metabolism transition in multi-cellular human tumor spheroids. J. Cell Physiol. 216, 
189-97.  
 Rodríguez-Enríquez S, Juárez O, Rodríguez-Zavala JS, Moreno-Sánchez R (2001) 
Multisite control of the Crabtree effect in ascites hepatoma cells. Eur. J. Biochem. 
268, 2512-9.  
 Rodríguez-Enríquez S, Marín-Hernández A, Gallardo-Pérez JC, Moreno-Sánchez R 
(2009) Kinetics of transport and phosphorylation of glucose in cancer cells. J. Cell 
Physiol. 22, 1552-9. 
 Rodríguez-Enríquez S, Torres-Márquez ME, Moreno-Sánchez R (2000) Substrate 
oxidation and ATP supply in AS-30D hepatoma cells. Arch. Biochem. Biophys. 375, 
21-30. 
 Shim H, Dolde C, Lewis BC, Wu CS, Dang G, Jungmann RA, Dalla-Favera R, Dang 
CV (1997) c-Myc transactivation of LDH-A: Implications for tumor metabolism and 
growth. Proc. Natl. Acad. Sci. 94, 6658–6663. 
 Shonk CE, Arison RN, Koven BJ, Majima H, Boxer GE (1964) Enzyme patters in 
human tissues. III. Glycolytic enzymes in normal and malignant tissues of the colon 
and rectum. Cancer Res. 25, 1965. 
 Smith  TAD (2000) Mammalian hexokinase and their abnormal expression in cancer. 
Br. J. Biomed. Sci. 57, 170-178. 
 Sonveaux P, Végran F, Schroeder T, Wergin MC, Verrax J, Rabbani ZN, De 
Saedeleer CJ, Kennedy KM, Diepart C, Jordan BF, Kelley MJ, Gallez B, Wahl ML, 
Feron O, Dewhirst MW (2008) Targeting lactate-fueled respiration selectively kills 
hypoxic tumor cells in mice. J. Clin. Invest. 118, 3930-42. 
 
 
142 
 
 Staal GEJ, Kalff A, Hessbeen EC, van Veelen CWM, Rijksen G (1987) Subunit 
composition, regulatory properties and phosphorylation of phosphofructokinase from 
human gliomas. Cancer Res. 47, 5047-5051.     
 Stubbs M, Bashford CL, Griffiths JR (2003) Understanding the tumor metabolic 
phenotype in the genomic era. Curr. Mol. Med. 3, 49-59.  
 Teusink B, Passarge J, Reijenga CA, Esgalhado E, van der Weijden CC, Schepper 
M, Walsh MC, Bakker BM, van Dam K, Westerhoff HV, Snoep JL (2000) Can yeast 
glycolysis be understood in terms of in vitro kinetics of the constituent enzymes? 
Testing biochemistry. Eur. J. Biochem. 267, 5313-29. 
 Tysnes B, Bjerkvig R (2007) Cancer initiation and progression: involvement of stem 
cells and the microenvironment. Biochim. Biophys. Acta 1775, 283-297. 
 Volgelstein B, Kinzler KW (2004) Cancer genes and the pathways they control. 
Nature Med. 10, 789-799.  
 Vora S, Halper JP, Knowles DM (1985) Alterations in the activity and prolife of 
human phosphofructokinase during malignant transformation in vivo and in vitro: 
transformation and progression-linked discriminants of malignancy. Cancer Res. 45, 
2993-3001. 
 Vora S, Oskam R, Staal EJ (1985)b isoenzymes of phosphofructokinase in the rat. 
Biochem. J. 229, 333-341. 
 Walenta S, Mueller-Klieser WF (2004) Lactate: mirror and motor of tumor 
malignancy. Semin. Radiat. Oncol. 14, 267-274. 
 Widjojoatmodjo MN, Mancuso A, Blanch HW (1990) Mitochondrial hexokinase 
activity in a murine hybridoma. Biotech. Lett. 12, 551-556.  
 Wilson JE (2003) Isozymes of mammalian hexokinase: structure subcellular 
localization and metabolic function. J. Expert. Biol. 206, 2049-2057. 
 Xu RH, Pelicano H, Zhou Y, Carew JS, Feng L, Bhalla KN, Keating MJ, Huang P 
(2005) Inhibition of glycolysis in cancer cells: a novel strategy to overcome drug 
resistance associated with mitochondrial respiratory defect and hypoxia. Cancer 
Res. 65, 613-21. 
 Yalcin A, Telang S, Clem B, Chesney J (2009) Regulation of glucosa metabolism by 
6-phosphofructo-2-kinase/fructosa-2,6-bisphosphatases in cancer.  Exp. Mol. Pathol. 
86, 174-179.  
 Yeung SJ, Pan J, Lee MH (2008) Roles of p53, MYC and HIF-1 in regulating 
glycolysis - the seventh hallmark of cancer. Cell. Mol. Life Sci. 65, 3981-99. 
 Yun H, Lee M, Kim SS, Ha J (2005) Glucose deprivation increases mRNA stability of 
vascular endothelial growth factor through activation of AMP-activated protein kinase 
in DU145 prostate carcinoma. J. Biol. Chem. 280, 9963-72. 
 Zalitis J, Oliver IT (1967) Inhibition of glucose phosphate isomerase by metabolic 
intermediates of fructose. Biochem. J. 102, 753-9. 
 
 
143 
 
6.5 Apéndice 
 
 
 
 
 
 
 
Otras publicaciones durante el doctorado 
 
 
144 
 
 
REB 25(4): 121. 2006 12 1 
PROBLEMA BIOQuíMICO 
Aná l is is de con t ro l me t abó li co . 
Control de la glucólisis en cél ulas tu morales 
La glllcólisis en las células tumorales se encuentra 
incrementada con respecro a los tej idos 1l00ulales. Esto 
se debe a un aumento en la actividad de todas las enzimas 
que forman parte de esta vía y a cambios en los mecanis-
mos de regulación de las dos enzimas consideradas como 
sitios de control en células l1omlales. la hexocinasa (HK) 
y la fosfofructocinasa tipo 1 (PFK-I). que las llacel1 me-
nos sensibles a sus inhibidores fi siológicos (glucosa-6-
fosfato. cin'ato y ATP) (1 ) . Para evaluar la hipótesis de 
que estas enzimas no controlan la glucólisis en las células 
AS-30D (hepatoma ascítico). se utilizó el análisis de elas-
ticidades. que pennite calcular los coeficientes de control 
de flujo (CJ) a partir de los coeficientes de elasticidad. El 
coeficiente de elasticidad ( E) es la capacidad de un blo-
que de enzimas de variar su velocidad al cambiar las C011-
centraciol1es de su producto o sustrato (2). Experimel1-
talmeme se calcula a partir de la variación de la COllcen-
tI"ación de intelmediarios (susu"atos o productos de las 
enzimas de la vía) y del flujo de la vía en estado estac io-
nario. Esto se logra mediante el empleo de inhibidores o 
al variar la concentración del sustra to de la 11Ita metabólica 
(glucosa en el caso de la glucólisis). En la Tabla 1 se 
muestran las velocidades de generación de lác tico (velo-
cidad de glucólisis) y las concenn'aciones en estado esta-
cionario de glucosa -6-fosfato (G6P) y fru ctosa-l.6-
bifosfato (F- l.6BP) a diferentes concentraciones de glu-
cosa y en la Tabla 2. las velocidades de glucólisis y las 
TABLA 1 
Velocidad d p glucólisis, concentr aciones de G6P y de 
F -1,6BP a diferentes concentraciones de glucosa 
Glucosa Velocidad de G6P Velocidad de F-l ,6BP 
(mM) glucólisis (lwlol!mg) glucólisis (nmollmg) 
(runollminlmg) (nwollminlwg) 
4 1.8 2.7 1.8 30 
4 5 5 4 2.8 5 4 27 
5 13.4 32 13.4 31 
5.5 16.4 3.6 16.4 34 
6 20.5 3.4 20.5 46 
Alvaro Marín Hernánd ez y Rafae l Moreno Sánchez 
Co rreo E: marin hernnd ez@yahoo .com.mx 
TABLA 2 
Velocidad de glucólisis, concentraciones de G6P y de 
F-l ,6BP a diferentes concentraciones de oxalato 
0xa1alo Velocidad de G6P Velocidad d ~ F1,6BP 
(mM) g1ucólisis (nmollmg) glucólisis (runollmg) 
(runollminlmg) (wllollmiu/ mg) 
O 15 3 16 28 
05 10.7 31 14.2 29.4 
1.0 9.6 3.4 12.8 30.8 
1.5 8 3.5 12.6 35.3 
2.0 6.5 3.8 10.1 37.2 
concentraciones de ambos intelulediarios. pero en pre-
senc ia de diferentes concenu'ac iones de oxalato (+ 5 mM 
de glucosa) (3) . 
Con estos datos: 
l. D eterminar los coeficientes de control de flujo. 
2. ¿Disminuyó el cOlmol que ejercen la HK y la PFK-I en 
la glucólisis de células nllllorales. considerando que en 
etiu'Ocito cOllU'olan 0.7 y 0.3. respectivamente (4)? 
3. ¿Qué enzimas son las que ejercen mayor COlltTOl? 
El volumen itm-acelular el1 las células nUllorales es de 2.28 
~1 l/ 1. 8mg de proteína. 
REFERENCIAS 
1. Moreno-Sánchez R. Rodriguez- Enriquez S. Marin-
Hemández A. Saavedra E. (2007) Energy metabolism in 
hunor cells. FEBS J. en Prensa. 
2. Fell D (1997) Understanding the control of metabolismo 
Plellum Press. London. 
3. Malill-HemándezA. Rodriguez-Ellliquez S. Vital-González 
PA. Flores-Rodliguez FL. Macías-Silva M. Sosa-Gan ocho 
M. Nloreno-Sánc hez R. (2006) Detenn ini ng and 
understallding the control of glycolysis in fa se-growth tu-
morcells. FEBSJ. 273: 1975-1988. 
4. RapopoI1 T. Hei1ll'iclt R. Jacobasch G & RapOpOtl S (1974) 
A linear steady-state treatment of enzylllatic chains. A 
mathematical model of glycolysis of human elytrocytes. 
EmJBiochem42: 107-120. 
 
 
145 
 
42 REB ] 8(2): 42-51. 2009 
EL FACTOR INDUCIDO POR LA HIPOXIA-l (HIF-l) 
y LA GLUCÓLlSIS EN LAS CÉLULAS TUMORALES* 
RESUMEN 
El fa ctor inducido por la hipoxia 1 (HIF -1) tiene un papel 
fundamental en la respuesta a la baja tensión del oxigeno. 
ya que regula la expresión de una gran variedad de genes. 
cuyo s produc tos part icipan en procesos co mo la 
angiogénesis. el metabolismo energético. la eritropoyesis 
y la proliferación celular. Diversos estudios indican que 
existe una relación estrecha entre el cáncer y el HIF- l. 
debido a que los mecanismos que regulan su expresión se 
encuentran alterados en las células tumorales. El HIF- I 
es uno de los factores involucrados en el incremento de 
la glucólisis en las células nUllora les. ya que aumenta la 
ac tividad de ciertas isofonnas de las enzimas glucolíticas 
(GLUTI. GLUT3. HKI. HKIT. PFK-L. ALD-A. ALD-C. 
PGK I. E O -a. PYK-M2. LDH -A. PFKFB-3) . 
promoviendo un aumento del flujo glucolitico (producción 
de lacta to. H+ ATP e intennediarios de la glucólisis). Por 
otra parte . a lguna s de e s tas ¡so foIma s participan 
activa mente en otros procesos C01110 son la inhibición de 
la apoptosis (HKI y HKIT). la transcripción de hi stonas 
(LDH-A) y la migración celular (EN O-a ). las cuales 
favorecen el desanollo nunora!. 
PALABRAS CLAVE : Hipox ia . g lucó lisis. HIF -1. 
isoenzimas glucolítica s. células tumorales. 
Alvaro Marín-Hernández 
ABSTRACT 
The hypoxia-inducible fac tor 1 (HIF -1) is an imponalll 
key mediator during low-oxygen cellular adap ta tion. 
Severa l genes i ll vo lve d 111 311giogeneSlS. 
erythropo iesis. energy metabolism and cell surviva l 
are up-regulated by HIF- l. In IUIl10ur cells. HIF- I 
overexpress ion 3 11d ac tiva rí on is assoc iated \V irh 
maligllancy. developmelll 311d angiogenesis: wherea s 
HIF -1 low express ion has been detennined in normal 
tissues and benign nUllors. Atmetabolic level. HIF-
I -activation stimulates transporters (GLUTi. GLUT3) 
and glycolytic enzymes (HKI. HKII . PFK-L. ALD-A. 
ALD-C. PGK I. E O-a. PYK-M2. LDH-A. PFKFB-
3) transc ription . In conseqllellce. a signifi cant 
incremelll in the glycolyt ic flux (lactate .ATP. and H+) 
and g lyco ly tic intermediares levels are arta illed. An 
additional role ofsoll1e oftbese HIF-induced iso fonlls 
as act iva tors o f surv iva l. h istones transcript iolla l 
ac tivat ion (LDH-A). apopto tic inh ib it ion (HKI y 
HKII) and cellular migration (ENO-o) pathways is 
discussed . 
KEY WORDS: Hypoxia. glycolysis. HIF-1. glycolytic 
enzymes. tumonr ceUs . 
INTRODUCCION 
La baja tensión de oxigeno altera la 
hOll1eostasis de la célula. lo que con-
duce a la activación del factor induci-
do por la hipoxia 1 (HIF- I). El HIF- I 
tiene como función incrementar la 
transcripción de genes cuyos produc-
tos son proteínas que participan eIl la 
angiogénesis. la erirropoyesis. la pro-
liferación celular. la remode lación 
vascular y el metabolislllo energético: 
permitiendo a la célula adaptarse a 
estas condiciones tan adversas (1 ). 
tensión de oxígeno presente en los nl-
mores y las alteraciones en los meca -
nismos que regulan su expresión. Por 
lo anterior. el HIF- I puede conside-
rarse como un marcador hUlloral y un 
blanco terapéutico. 
En la mayoria de los nUllores pri-
marios de cerebro. páncreas. mama. 
colon. ovaIio. pulmón y próstata. y sus 
metástasis. se detectan altos niveles 
del HIF-I comparados con los tej idos 
nOl1nales de los cuales provienen o 
tumores benignos (2). Las causas que 
promueven la sobreexpresión del HIF-
l en las células hl1110rales son la baja 
-Recibido: 19 de agosto de 2008 Aceptado: 14 de abril de 2009 
En esta revisión se analiza el pa -
pe l que juega el HIF- I en la regula-
ción de la glucólisis. así como las 
ventajas que ofrece a las cé lula s 
mmorales la expresión de las enzimas 
Instimta acianal de Cardialagia Ignacio Chávez". Depa11amenta de Bioquimica. Juan Badiana No. l . Col. Sección :A..rvI. México DE 
14080. Coneo E: marillhel1111dez@yahoo.calllllL'X 
 
 
146 
 
• : 
• • 
REV I EW A RT ICL E 
Energy metabolism in tumor cells 
Rafael Moreno-Sánchez, Sara Rodríguez-Enriquez, Alvaro M arín-Hernández and Emma Saavedra 
Instituto Nacional de Cardiología. Departamento de Bioquímica. Tlalpan, México, Mexico 
Keywords 
casiopeinas; chemotherapy; glycolysis; 
m etabolic control analysis; mitochondrial 
metabolism; PET; rhodamines 
Co rrespo nde nce 
R. Moreno-Sánchez. Instituto Nacional de 
Cardiología, Departamento de Bioquímica. 
Juan Badiana no. 1. Tlalpan. México DF 
14080, Mexico 
Fax: +5255 55730926 
Tel: +5255 55732911. ext. 1422. 1298 
E-mail: rafael.moreno@cardiologia.org.mx or 
morenosanchez@hotmail.com 
(Received 31 Octaber 2006, revised 2 Janu-
ary 2007, accepted 10 January 2007) 
doi:, O., 11 1/j.17424658.2007.05686.x 
In ea rly studies on ene rgy melabolism of t umor cells, il \Vas proposed lhat 
lhe enhanced glyco lysis was induced by a decreased oxidative ph osphoryla -
tion . Since then it has been indi scrimina tely applied to all types of tumor 
cells that the A TP supply is mai nly or only provided by glycolysis, without 
a n appro priate experimental evalua tion . In th is review, the different genetic 
a nd biochemical mechanisms by which tumor cells achieve an enha nced 
glycolytic flux a re ana lyzed . Furthermore, lhe proposed mechanisms tha t 
a rgua bly lea d to a decreased oxidative phosphorylation in tumor cells are 
discussed . As the O2 concentratio n in hypoxic regio ns o f tumors seems not 
to be limit ing for the functio ning o f oxidative ph osphorylation, this pa th-
way is re-evalua ted rega rding oxidizable subst rate ut iliza tion a nd its contri-
b ution to AT P supply versus glycolysis. In the tumor cell lines where the 
oxidative metaboli sm preva ils over the glycolytic meta bolism fo r AT P sup-
ply, the flu x control d istribution o f both pa thways is described. The efTect 
of glycolytic and mitochondrial drugs o n tumor energy metabolism a nd cel-
lular prolifera tion is descri bed a nd d iscussed . Simi larly, lhe energy meta -
bolic changes associated wi th inherent and acq ui red resista nce to 
radiothera py a nd chemolherapy of tumor cells, a nd th ose determined by 
positron emissio n tomography, are revised. It is proposed lha t energy 
metabolisrn may be an altem ati ve therapeutic targe t for both hypoxic 
(glyco lytic) and oxidative tumors. 
In bioche mical and physio logical stud ies, tumo r cells a re 
usually c1assifi ed accord ing to their ra te of growth : low; 
intermedi ate; or fast [1]. For tumo rs in experimenta l ani-
mals, the growth ra le is determined by size and volume, 
mitotic count , degree of di fTerent ia tion and thyrnid ine 
incorpora tion [2]. Examples of fas t-growth tumors in 
mice inc1ude several experimental cancers, such as Ehrl-
ich asciles tumor, fi brosa rcoma 1929 and Iymphocytic 
leukemia L1 210; a nd in ra ts, fas t-growth l umors inc1 ude 
the hepatomas of Morris (3924A, 7793, 7795, 7800, 
7288C, 73 168 ,3683), Reu ber H-35, Novikoff, AH 130 
and AS-30D, breast carcinosarcoma Walker 256, 
hepatoce ll ula r carcinoma HC-252, hepatoma induced 
by dimethylazo benzene and DS-carcinosarcoma [1]. 
In human tumors, c lassifica tion is based on their 
his tological characteristics a nd stage of c1inical pro-
gression. By their ad vanced develo pmental stage a nd 
metasta tic p roperties, sorne human tumors considered 
to be of fas t growth are breast ca rcin oma, ovarian ca r-
cinoma, melanoma, thyro id carcinoma, uterine carci-
noma a nd lung carcino ma [1 ,3 ]. Human primary bra in 
tumors, such as gliomas, glioblas tomas and medu lo-
blaslOmas, a re also considered as fast-growth tumors 
beca use of their high ra te o f pro li fe ration (average 
transfer in days or weeks) and their conversion 10 a 
poorly different ia ted status [4,5]. 
Tumor ce lls ex hi bit profound genetic, biochemica l 
and histological differences wi th respecl to the origi nal, 
Abbreviati ons 
ALD, ak:lolase; ANT, adenine nucleotide translocase; COX, cyclooxygenase; CT, computed l omography; F2 ,6BP. fructose-2.6-bisphosphate; 
FDG, 18fluoro-deoxyglucose; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; GLUT, glucose transporter; G6P, glucose-&-phosphate; 
HIF, hypoxia inducible factor; HK, hexokinase; LDH, lactate dehydrogenase; NSAID, nonsteroidal anti-inflammatory drug; PDH, pyruvate 
dehydrogenase complex; PET. positron emission tomography; PFK-', phosphofructokinase type '; PFK-2. phosphofruclokinase type 2; 
Pyr, pyruvate 
FEBS Journal 274 120071 1393-141 8 e 2007 The Authors Journal ccmpilat ion e 2007 FEBS 1393 
 
 
147 
 
ORIGINAL ARTICLI: 
Energy Metabolism 
Transition in Multi-Cellular 
Human Tumor Spheroids 
SARA RODRíGUEZ-ENRíQUEZ, 1* JUAN CARLOS GALLARDO-PÉREZ,I 
ALEJANDRO AVILÉS-SALAS,2 ALVARO MARíN-HERNÁ N DEZ,1 LlLIANA CARREÑO-FUENTES,I 
VILMA MALDONADO-LAGUNAS,2 AND RAFAEL MOREN O-SÁNCHEZ I 
I Departamento de Bioquímica, Instituto Nadonal de Cardiolog{o, Mexico, Mex;co 
2 Departamentos de Bio/agio Molecular y de Pato/agio. Instituto Nadonal de Con cero/agio, Mexico, Mex;co 
le is thought mar glycolysis is che predominant energy pathway in caneer, partkularly in salid and poorly vascularized tumors where 
hypoxic regions develop. To evaluate whether glycolysis does effectively predominate far ATP supply and to identify the underlying 
biochemical mechanisms. the glycolytic and oxidative phosphorylation (OxPhos) fluxes. ATP/A DP ratio, phosphorylat ion patemial, and 
expression and activity of relevant energy metabolism enzymes were determined in multi·cellular tumor spheroids, as a model of human 
sol id tumors. In HeLa and Hek293 young-spheroids, the OxPhos fluxand cytochrome e oxidase protein content and activitywere similar 
ro chose observed in monolayer cultured cells, whereas che glycolytic flux inc reased two- ro fourfold; the contribution of OxPhos to ATP 
supply was 60%. In contrast, in old-spheroids, OxPhos, ATP content, ATP/ADP ratio, and phosphorylation potential dimin ished 50-70%, 
as well as che activity (88%) and content (3 times) of cytochrome e oxidase. Glycolysis and hexokinase increased significandy (both, 
4 times); consequently glycolys is was che predom inant pathway for ATP supply (80%). These changes were associated with an increase 
(3 .3 times) in che HIF-I (X content. After chronic exposure, boch oxidative and glycolytic inhibitors blocked spheroid growch, although the 
glycolytic inhibitors, 2-deoxyglucose and gossypol (ICso of 15-17 nM), were more potent chan the mitochondrial inhibitors. cas io peina 11 -
gly, laherradurin, and rhodamine 123 (ICso > 100 nM). These results suggest that glycolysis and OxPhos might be considered as mecabolic 
targets to diminish cellular proliferation in poorly vascularized, hypoxic solid tumors. 
J. Cell. Physiol. 2 16: 189- 197,2008. <!:> 2008 Wiley-Liss, Ine. 
The fie ld of tumor energy metabolism seemed resolved by 
1956, when Warburg proposed that an energy deficiency 
caused by an irreversible damage of the mitochondria l function 
induced increased glycolysis, which presumably supported the 
energy demand for tumor formation and growth. Since men, it 
has been thought that Warburg's hypomesis applied to all or 
most cancer cell cypes (Weber. 200 1; Stubbs et aL, 2003; Xu 
et al. . 2005). Such an asseveration may indeed be particularly 
valid for sol id tumors, in which the lower inner oxygen 
concentration activates, via the hypoxia inducible factor I 
(H IF- I), the uanscription and translation of glycolytic genes 
(G LUT-I and -3, HK, phosphofruetokinase type 1, aldolase, 
GAPDH, phosphoglycerate kinase, enolase, pyruvate kinase, 
and lactate dehydrogenase) (Dangand Semenza, 1999). which in 
turn, stimulates the glycolytic flux. However, oxidative 
phosphorylation (enzyme activities and flux) has not been 
ana lyzed in parallel. 
In this regard, recent data from numerous sources on the 
role of oxidative phosphorylation (OxPhos) during tumor 
developmenc (Lampidis et al., 1983; Guppy et al .. 2002; Zu 
and Guppy, 2004; Rodríguez-Enríquez et aL, 2oo6a; 
Moreno-Sánchez et aL, 2007) have prompted a re-evaluation 
of Warburg's hypothesis. For instan ce, it has been shown 
in human (HeLa, MCF-7, HL-60) and rodent (hepatoma 
AS-30D) tumors that mitochondrial ATP may indeed 
support me acce lerated cellular proliferation rate (Sweet and 
Singh, 1995; Guppy et al .. 2002; Rodriguez-Enriquez et al .. 
2006a) ando in con sequen ce. anti-mitochondrial drugs are 
able to abolish tumor growth (Rodriguez- Enriquez et al.. 
2006a). For several aggressive metastasic human carcinomas 
(HeLa, CRL 1420 panereas, MCF-7 breast, MB49 bladder) 
and osteosarcomas, OxPhos inhibitors alone. such as 
casiopeina II-gly and rhodamines (123, 6G). or in 
combination with glycolytic drugs have been effectively 
used tO decrease the proliferation rate (reviewed in 
Moreno-Sánchez et al.. 2007). 
C2oo8WIL!:::Y LISS.INC . 
It should be noted. however. that most of the 
above-mentioned studies have been carried out with cellular 
suspensions or monolayer culture cells. It is usually assumed 
that energy metabolism of monolayer cultures (i.e .. enzyme 
activity. metabolic pathway flux. and sensitivity to metabolic 
drugs) is simi lar to mat found in solid tumors. Solid tumors 
maintain a complex physiological structure derived from their 
three-dimensional organization (i.e .• cell-matrix interactions. 
variations in nutdent supply due tO glucose. lactate. pH. and 
oxygen gradients; as well as variations in the expression of 
HIF- Ia and c-Myc), which might result in gene regu lation 
changes and. hence. in enzyme expression with respect to 
monolayer cultures and cell suspensions (Sutherland. 1988; 
Khaitan et al., 2006). 
This article includes Supplementary Material available from the 
auttlors upon request or via the Internet at http:// 
www.interscience.wiley.comljpages/OO21-9S4I/suppmat. 
Abbreviations: ANT, adenine nucleotide translocator; casll -gly. 
casiopeina II-gly; COX, cytochrame c oxidase; GAPDH, 
glyceraldehyde 3P-dehydrogenase; GLUT -1, glucose transponer 1: 
HK, hexokinase; rhod 123, rhodamine 123; 2-DOG, 
2-deoxyglucose. 
Contract grant sponsor: CONACyT-Mexico; 
Contract grant numbers: SaJud-2002-COI-7677. SEP-2007-60517. 
*Carrespondence to: Sara Rodríguez-Enriquez. Departamento de 
Bioquimica, Instituto Nadonal de Cardiología, Juan Badiana No. I 
Seccion XVI. Talpan, Mexico D.F. 14000, Mexico. 
E-mail: saren960104@hotmail.com 
Received H October 2007; Accepted II December 2007 
DOI: 10. 1 002ljcp.21392 
189 
 
 
148 
 
Mol. Nutr. Food Res. 2009, 53, 29-48 OOI10.10021mnfr.20070Q470 
Review 
Targeting of cancer energy metabolism 
Sara Rodriguez-Enríquez, Alvaro Marín-Hernández, Juan Carlos Gal lardo-Pérez, 
Li liana Carreña-Fu entes and Rafae l Moreno-Sánchez 
Departamento de Bioquímica, Instituto acional de Cardiología, Mexico, Mexico 
The main purpose of thi s rcv iew is 1'0 update and analyze the effcct of several antineoplastic drugs 
(adriamycin, apoptodi lin , casiopeinas, cisplatin, c1 otrimazole, cyclophospharnidc, diterca linium, 
NSAIDs, tarnoxifen, taxol , 6-mcrcaptopurinc, and a-tocopheryl succinate) and energy mctaboli sm 
inhibitors (2-DOG, gossypol , dclocalizcd lipophilic cations, and uncouplcrs) on tumor dcvclopmcnt 
and progress ion. Thc possibility thal Ihese antineoplast ic drugs currcntly used in i" vitro cancer mod-
els, in chemo- therapy, or mujer study in phase 1 to I11 c1inical trials induce tumor cellular death by 
altering also mctabo litc concentration (i.e. , ATP), enzymc activitics, andlor cncrgy mctabolism nuxcs 
is assessed. It is proposed that the use of energy metabo lic therapy, as an alternative or complemen-
tary strategy, might be a promising novel approach in the trcatment of canccr. 
Kl'ywo rds: Anticanccr drugs I Glycolysis l Mitochondria I Oxidat i"c phosphorylation 
Rccc i"cd: Novcmbcr 19, 2007; rcvi scd: Junc 11 , 2008; acccpted: July 27, 2008 
1 Introduction 
The fi eld of caneer energy mctabolism sccmed resolved by 
1956 when Warourg [1] proposed that lhe prime cause of 
cancer was an energy defi ciency caused by an irreversible 
damage to the mitochondria l function mat induced an 
increased glycolysis. Since then, severa I researchers have 
thought that the Warburg hypothesis app lies to all or most 
cancer cell typcs (2- 14), beca use one ofLh e most notorious 
and well-kno\Vn alterations in tumor ce ll s is certain ly an 
increased glycolytic eapacity, even in lhe presence of high 
O, eoneentration [3, 4, 15- 18]. 
The observat"Íon that tumors have a higher glycolytic 
capacity than nomlal ec ll s has fo und application in the use 
o fpositron emission tomography (PET) for diagnosis, mon-
itoring, and treatmcnt o f canccr [19, 20). With 'S Auo re_ 
deoxyglueose (FDG) as traeer, PET has established lhat lhe 
vast majority of metastati c tumors (>90%) are highly gly-
colyr-ic; and has al so a 1I0wed for lhe accurate detec tion 
CO!"rl'spo lldl' II(,l': Dr. Sara Rodrigucz-Enríqucz, Departamento de Bi-
oquímica, Instituto Nacional de Cardio logía, Juan Badiano No. 1, Col. 
Sccción 16, Tlalpan , Mcxico, Mcx ico 
E-mail: sara.rodrigucZ@ca rdiolog ia.org.mx 
liax : +5255-55- 7) -09-26 
Ahhl"c\' j:l1 ioJls: AlOO. aldolase; A¡\ IE autocrinc motililY factor; C as 
Il gly. casiopcina 11 gly; cccr, carbonyl cyanidc-J-chlorophcnylhy-
drazonc; CM L. chronic mycJogcnous leukemia; CI'. cyclophospha-
mide ; OLC, dcloca lized lipophilic cation; 2-00G, 2-deox yglucose; 
ENO. cnolase; F t.6 U1'. fructose 1,6 bisphosphalc; f"2.6 Ur. fructose 
© 2009 WI LEY -VCH Verlag GmbH & Co. KGaA, Weinheim 
(> 90%) of solitary pulmonary nodules, mediastinal and 
axillary Iymph nodes, colorecta l canccrs, Iymphomas, me l-
ano mas, breast eanccrs, and head and neck cancers [19] . 
2 Glycolysis in tumor cells 
2.1 Ge neral 
Fast-g rowth cancers from human (Ieukemia, 1-leLa) and 
rodent (AS-30D, Morris 7800, Dunings LC I8, Novikolf 
hepatomas) show an increased g lycolytic rate (2- to 17-
times) in comparison \Vith nontllmorigenic cell s (21- 231-
This metabolic featllre is the consequence of glyco lytic 
enzymes over-ex pression indllced by (i) oncogene c-myc 
(glueose transporter 1 (GLUT I), hexosephosphate isomer-
ase (HPll , phosphofTueto kinase type 1 (pFK I l, glyeeralde-
hyde 3-phosphate dehydrogenase (GAPDH), phosphog ly-
eerate kinase (pGK), and enolase (ENO)) [5] and (ii ) HIF-
la (hypoxia indueible fu etor la), GLUT I, glueose trans-
2,6 bi sphosphalc; Foc" l'fluoro-deoxyglucosc ; GA I'OI-I, glyccralde-
hydc ) ·phosphate dc hydrogcnase; G LUl" l. glucose transponer 1; 
Gl UT J. gtucose transponer J; G6 1'. glucosc 6-phosphatc; 1-1 K 1. hcxo· 
kinase typc 1; HK II. hcxokinase typc 11 ; 111' 1, hcxoscphosphatc iso-
merase; LO II , lactalc dchydrogcnasc; 61\11', 6-mcrcaptopurinc; J1\II'A, 
J .mcrcaptopicolinic acid; I\ I I'T. mitochondrial pcrmcability tran. .. ilion; 
NSAID. nonsteroidal anti · inflanunatory drug; I'ET. positron emission 
lomography; I'FK 1, phosphofructo kinase type 1; I' f·K2. phospho-
fructo kinasc typc 2; r GK, phosphoglyccmte kinasc; I '- g l~ ' . glycopro-
tein typc P; u-TO S, u-tocophcryl succinate 
( - ~ 1 :~ .. Science' "- ,,,.,,, , , .. , ..... "" www.mnf-joumal. com 
--
 
 
149 
 
ORIGINAL ARTICLE 
Kinetics of Transport and 
Phosphorylation of Glucose 
Cancer Cells 
• In 
SARA RODRíGUEZ-ENRíQUEZ*, ALVARO MARíN-HERNÁNDEZ, . . 
JUAN CARLOS GALLARDO-PEREZ, ANO RAFAEL MORENO-SANCHEZ 
Departamento de Bioquimica, Instituto Nacional de Cardio/agio Ignacio Chávez, Tia/pan. México Oty, Mexico 
Metabolic control analysis of tumor glycolysls has indicated dlat hexokinase (HK) and glucose transporter (GLUT) exert (he majn flux 
control (71 %). T o understand why t hey are che main controlling steps. (he G LUT and HK kineti cs and (he contenes of GLUT 1, GLUT2. 
GLUT3. GLUT 4 . HKI, and HKIl were analyzed in rat hepatocarcino ma AS· 30D and HeLa human cervix cancer. An improved protocol to 
determine the kinetic parameters af GLUT was developed with o- [2-3H-glucose] as physiologica l substrate. Kinetic analysis revealed 
two components at low- and h ¡ g h - ~Iucose concentrations ln bom tumor cell s. At low glucose and 37 C. the Vmax was SS ± 20 and 
17.2 ± 6 nmol (min X mg protein) • whereas the Km was 0.52 ± 0.7 and 9.3 ± 3 mM for hepatoma and HeLa cells. respectively. 
GLUT activity was partially inhibited by cytochalasin B (ICso = 0.44 ± 0.1; Kt = 0.3 ± 0.1 ~ M ) and phloretin (ICso = 8.7 ~M ) in AS-30D 
hepatocarcinoma. At physiologica l glucose. GLUTI and GLUT3 were the predominant active isoforms in HeLa cells and AS-30D cells. 
respectively. HK activity in He La cells was much lower (60 mU/mg protein) than that in AS-30D cells (700 mU/mg protein) . but both HKs 
were st ro ngly inhibited by G6P. HKII was the predominam isoform in AS-30D carcinoma and HeLa cells. The much lower GLUT Vm;¡x and 
catalytic efficiency fI/ mJ Km) values in comparison to those of G6P-sensitive HK suggested che transporter exerts higher control on che 
glycolytic flux than HK in cancer cells. Thus . GLUT seems a more adequate therapeutic target. 
J. Cell. Phys iol. 221: 552- 559. 2009. © 2009 Wiley-liss . Ine. 
The glycolytic rate of human and rodent fast-growing rumor 
cell s. including those cumors depending on mitochondrial 
metabolism. is substamiallyhigher than in non-tumorigenic cells 
(reviewed in Moreno-Sánchez et al.. 2(07). The activation 
of cereain oncogenes (c-myc. H-Ros. v-src) and transcription 
factors. such as H IF-I a (hypoxia-inducible factor I a). induces 
the over-expression and increase the activity of most glycolytic 
enzymes and glucose transporters (Dang et al.. 1997). 
Metabolic control analysis applied to non-rumorigenic 
mammalian cells (hepatocytes. erytrocytes. skeleta l and cardiac 
muscle) shows that the main controlling steps of glycolysis 
(8(}- IOO%) are the glueose eransport (GLlfT). hexokinase 
(HK). and phosphofructokinase type-I (PFK- I). whereas the 
remaining flux control is distributed among the other steps 
(Rapoport et al.. 1976; Torres et al .• 1989; Kashiwaya et al .• 
1994). However. in cancer cells a re-distribution of flux control 
might occur because such controlling stepS are the most 
over-expressed glycolytic proteins (Moreno-Sanchez et al.. 
2007.2008). In this regard. it was recently determined by using 
metabolic control analysis (Marin-Hernández et al .. 2(06) that 
A5-30D glyeolysis was mainly eonero ll ed by HK and GLUT 
(up to 71 % of flux control). The rest of the glycolytic enzymes 
(from HP1 eo LDH) and glyeolysis branehes (peneoses 
monophosphate. glycogen synthesis. pyruvate branches. and 
ATP-demand pathways) exerted the remaining flux control; 
PFK- I showed negligible flux control (6%). Most of the flux 
control of the G6P-producing block in tumor cells might reside 
in HK beca use (i) the intracellular glucose concentration is high 
and saturating for HK (i.e .. HK elasticity is low) and (i i) the 
enzyme is st rongly inhibited by G6P. 
HKII is the predominant isoform expressed in most cancer 
cell s (Pedersen et al.. 2(02); this isoform may bind to che 
external mitochondrial membrane to rapidly consume the 
mitochondrial-derived ATP and generate G6P (Arora and 
Pedersen. 1988). It has been widely documented that che 
HK-maximal rate of human and rodent tumors is higher 
(60-420 mU/mg protein) chan that of normal t issues 
(0.00 1- 170mU/mg proeein) (Oskam ee al.. 1985; Mandarino 
et al.. 1995; Marín-Hernandez et al .• 2006; Moreno-Sánchez 
e lOOQ WILEY LI5S. INe . 
et al.. 2007). However. in me majority of the kinetic studies 
in cancer cells. the H K activity has been calculated from 
the free or cytosolic isoform. whereas me contribution of 
membrane-bound HK has not always been evaluated. thus. 
underestimating total HK activity. 
HK is massively over-expressed by AS-30D 
hepatocarcinoma (300-times vs. hepatocytes); however. both 
HK isoforms. cytosolic and membrane-bound. are strongly 
inhibited by their product. G6P (Marín-Hernandez et al. . 2006). 
In con sequen ce. HK beco mes one ofthe main controlling steps 
of tumor glycolysis. To make well-informed statements and 
generalizations about the behavior of this relevant glycolytic 
step it is then required (i) te establish what HK isoform is 
expressed by each particular tumor cell line and (ii) to 
determine ies kinetic properties. 
GLUT I is the most frequently isoform over-expressed in 
fast-growing and metastatic rumors (Medina and Owen. 2(02). 
However. other GLUT isoforms (GLUT3 and GLUT5) noe 
usua ll y found in the original tissue may also be over-expressed 
(Higashi et al.. 1997). GLUT over-expression has usually been 
determined as an mRNA or protein in crease (Nagamaesu et al. . 
"'brwwlatb.: G6P. "tcOIl , p~ GlUT. atumse 
tranIpOrtlIr; HK. ~ HPI. NI 1111 p ........... IIom1 ..... 
• ddldol'll SupponIna Inforn'ldol. ma)' be bnlln che onIlne 
_ of tHs 1I1kIo. 
Contnct _t sponsor. CONACyT-MÓIdCO¡ 
Contrxt IJW1t nJmbJrs: 80534. 60085. 
· CorrespondJnce lO: Sara R odrfCU1z- En rf~Z. o.partamlNO de 
Bioquímica, InstiunD Nadonal de Cardiología. Juan &adiano No. l . 
Colonia Sección, xv~ llalJ:8n 14<80, M6xieo. E ~¡j l : 
sararodriguez@cardiolocia.org.mx 
Received I J Hay 200'J; Accepted 221"'. 2009 
Published online in Wiley InterScience 
(www.interscience.wiley.eom.), 13 August 2009. 
om lo.IOO2/jcp.2 188S 
552 
 
 
150 
 
Rafael Moreno-Sánchez, * Sara Rodríguez-Enríquez, Emma Saavedra, Alvaro Marín-Hernández, luan Carlos Gallardo-Pérez 
Instituto Nacional de Cardiología, Departamento de Bioquímica, luan Badiana 1, Tlalpan, México DF, Mexico 
Abstract. 
The molecular mechanisms by which tumor cells achieve an 
enhanced glycolytic flux and, presumably, a decreased 
oxidative phosphorylation are analyzed. As the O, 
concentration in hypoxic regions of tumors seems not 
Iimiting for oxidative phosphorylation , the role of this 
mitochondrial pathway in the ATP supply is re-evaluated. 
Drugs that inhibit glycoysis and oxidative phosphorylation 
© 2009 Intemationa l Union of Biochemistry and Molecular Biology, Ine. 
Volume 35. Number 2, March/ April 2009, Pa ges 209-22 5 • 
E-mall: rafael.moreno@C ardiologia.org.mx or morenosancheZ@hotmail.com 
l. Introduction 
AII tumor cell types show an altered energy metabolism in 
comparison to their tissue of origin (Figs. 1 and 2). The best 
characterized energy metabolism modification in tumor cells 
is an increased glycolytic capacity even in the presence of a 
high O, concentration a'I, reviewed in ref. 2) . Several mech-
anisms for the enhanced glycolysis in tumor cells have been 
proposed (Table 1). However, each particu lar tumor cell line 
has its own combination of mechanisms to increase glycoly-
sis and, therefore, generalizations should be avoided. 
2. Genetic regulation of glycolysis 
In tumor ce 115, there is an enhanced transcription of genes 
of several or all glycolytic pathway enzymes and transport-
ers that is accompanied by an enhanced protein synthesis 
Abbreviations: AlD, aldolase; k.(oA. acetyl CoA; Cit, citrate; DHAP, dihydroxyacelone 
phosphate; ENO, enolase; F6P, fruclOse-6·phosphate; Ft,6BP, ITuctose-l. 6· 
bisphosphate: F2,6BP. fructose-2, 6-bisphosphate; Fum, fumarate; GAPDH, 
glyceraldehyde.3-phosphate dehydrogenase; o:GPDH, a-glycerophosphate 
dehydrogenase; GLUT, glucos@ transport@r; G6P, glucos@-6-phosphat@; Gln, glutamin@; 
Glut, glutamat@; HK, hexokinas@; HPI, huos@-6-phosphalE! isom@ras@; lDH, lactat@ 
d@hydrog@nas@; OxPhos, oxidativ@phosphorylation; 2-oXO, 2-oxoglutaral@; PDH, 
pyruvalE! d@hydrog@nas@ complu; PET, positron @n@rgy transf@r; PFK-t, 
phosphofructokinas@ Iyp@ 1; PGK, phosphoglyc@rat@ kinas@; PGAM, phosphoglyc@ral@ 
mUlas@; PVK, pyruvat@ kinas@; PEp, phosphoenolpyruvat@; t ,3 BPG, t.3-
bisphosphogtyc@ral@; 2PG, 2-phosphoglyc@rate; 3PG, 3 phosphoglycerate; Pyr, 
pyruvate;TPI, trioSf'phosphat@ isom@ras@_ 
*Addr@ss for correspondenc@: Rafa@l Mor@no·S~nch@z, Ph.D., Instituto Nacional d@ 
(ardiologfa, Departamt>nlo de Bioqulmica, Juan Badiano No. 1, (01. Seccion XVI, 
Tlalpan, México DF 14080, Mexico. T@I: + S2SS SS73 2911 (@Xt. 1422, 129B); 
E-mail: rafael.moreno@cardiologia.org.mxormorenosanchez@tlotmail.com. 
Received 3 December 2008; acC@Pled 18 January 2009 
DOI: 10.1oo2/ biof.31 
Published online 11 March 2009 in Wiley Int@r$ci@nce 
(www.interscienc@.wilt!Y.com) 
are analyzed for their specificity toward tumor ce 115 and 
effect on proliferation. The energy metabolism mechanisms 
involved in the use of positron emission tomography are 
revised and updated. It is proposed that energy metabolism 
may be an alternative therapeutic target for both hypoxic 
(glycolytic) and oxidative tumors. 
Keywords: glycolysis, oxidative phosphorylation, mitochondrial 
function, PET, metabolic control analysis 
(21. It is usually assumed that these transcriptional changes 
lead to increased activity and pathway flux. Unfortunately, 
the required biochemical (metabolic, kinetic) experimenta-
tion is considered " old -fashioned" and , only in few of the 
molecu lar biology studies, transporter and enzyme activities 
and flux rate have been determined. 
In comparison with normal rat hepatocytes, the activity 
of all glycolytic enzymes are overexpressed in rat AS-30D 
hepatoma by 2-4 times (HPI, ALD, TPI, GAPDH, PGK, PGAM, 
ENO, and LDH), 8- to 10-fold for PYK, and '7-300 times for 
PFK-l and HK (Fig. 1) (101. For human cervix HeLa cells, all 
enzymes including HK and PFK-l are overexpressed by 2-7 
times, with the exception of PGAM and LDH which are 2-7 
times lower than in rat hepatocytes (101. However, for this 
last case a more rigorous comparison should be made with 
normal uterine cervix epithelial cells, the original source, 
when data become available. In rat Morris hepatomas, the 
activities of HK, PFK and PVK are 5- to soo-fold higher than 
in liver (261 whereas in human breast cancer, the activities 
of HK, ALD, PYK, and LDH are 3.7-7 times higher than in 
normal tissue (271. In several human carcinomas, overexpres-
sion (Table 1) of GLUT (31, GAPDH, ENO-l , PYK, and LDH 
(28,291 correlates with a high glycolysis rate. In contrast, in 
human cartilage and bone marrow cancer, glycolysis is high 
but only sporadic expression of GAPDH, ENO-l, and PYK is 
observed (291. Unfortunately, the expression pattern of the 
whole set of glycolytic enzymes is not always analyzed, 
which frequently leads to extend these observations to all 
types of cancer. 
In cancer cells, the hypoxia inducible factor ' " (HIF-1Cl) 
is a transcriptional factor that upregulates the expression of 
specific isoforms of GLUT, HK, PFK-l, PFK-2, ALD, GAPDH, 
209 
 
 
151 
 
 
 
ELSEVLER 
The ImernationalJournal of Biochemistry & CeH Biology 42 (2010) 1744-1751 
Contents lists available at ScienceDirect 
The InternationaI journaI of Biochemistry 
& CeH Biology 
jou rna l ho mepage: www.e l sevier.comll ocate / b i oce 1 
Oxidative phosphorylation is impaired by prolonged hypoxia in breast and 
possibly in cervix carcinoma 
Sara Rodríguez-Enríquez a,< , Lili a na Ca rreño- Fuentes a , Juan Carlos Ga llardo-Pérez a , Emma Saaved ra a , 
Héctor Quezad a a, Al icia Vega b, Alvaro Marín-Herná ndez a, Viridiana Olín-Sandova l a, 
M. Eugen ia Torres- Má rquez b , Rafael Moreno-Sánchez a 
i Departamento de Bioquímica. Instituto Nacional de Cardiología Ignacio Chávez. México D.F. 14080. Mexico 
b Departamento de Bioquímica. Facultad de Medicino. UNAM, México D.F. 04510. Mexico 
A R TICLE IN F O ABSTRAC T 
Artic/e history: 
Received 26 M.uch 2010 
Received in revised form I SJune 2010 
Accepted 12July201O 
Available online 21 July 2010 
Keywords: 
Glycolysis 
Hypoxia 
Mitochondria 
Oxidative phosphorylarion 
Breas[ cancer 
It has been assumed that oxidative phosphorylation (OxPhos) in solid tumors is severely reduced due to 
cytochrome e oxidase substrate restriction. although the measured extracel lular oxygen concentration 
in hypoxic areas seems not limiting for this ac tivity. To iden tify alternative hypoxia-induced OxPhos 
depressing mechanisms. an integral ana lysis of transcription. translation. enzyme activities and paeh-
way fl uxes was performed on glycolysis and OxPhos in HeLa and MCF-7 carcinomas. In both neoplasias 
exposed to hypox ia. an ea rly transcriptiona l response was observed after 8 h (two times increased 
glycolysis-re lated mRNA synthesis promoted by increased HIF- Io: levels ). However. major metabolic 
re modeling was observed only after 24 h hypoxia: increased glycolytic protein content (t-S-times), 
enzyme activities (2-times ) and fluxes (4-6-times ).lnte restingly, in MC F-7 cells, 24 h hypoxia decreased 
OxPhos flux (4-6-fold). and 2-oxoglutarate dehydrogenase and gl utaminase activities (3-fold). with no 
changes in respiratory complexes I and IV activi ties. In contrast, 24 h hypoxia did not significan tly affect 
HeLa OxPhos flux: neither mitochondria relaeed mRNAs, protein contents or enzyme activities, although 
the enhanced glyco lysis beca me the main ATP supplier. Thus, prolonged hypoxia (a) targeted sorne mito-
chondrial enzymes in MCF-7 but not in HeLa ceUs, and (b) induced a transition from mitochondrial 
towards a glycolytic-dependent energy metabolism in both MCF-7 and HeLa carcinomas. 
1. Introduction 
For growth, solid turnors have developed strategies to rnai n-
tai n the carbon source and oxygen supplies. Thus, turnors exhibit 
an act ive angiogenesis w hich, however, is high ly inefficient gen-
erating chaotic networks, with unorganized, frag ile , and leaky new 
vessels ( reviewed by Nagy et al.. 2009). In consequence. rhe dynam-
ics of the blood flow is affecred, leading to hypoxic regions a t 
100-200 J.lm away rrom a ru nctional blood supply (Vaupel et al., 
1989; Helmlinger et al., 1997; Nagy et al.. 2009 ). 
It has been derermined rhat the oxygen concentratio n inside 
well-oxygenated a reas of several carcinomas ranges fro m 30 to 
Abbreviations: cOX,cytochromecoxidase: HIF- I a. hypoxia-inducible factor l a: 
HK, hexokinase: GA. glutaminase: OxPhos. oxidative phosphorylation: POH. pyru-
vate dehydrogenase complex: POK1, pyruvate dehydrogenase kinase-l; 2-OGOH, 
2-oxoglutarare dehydrogenase. 
• Corresponding .1urhor ar: Oeparr.1menro de Bioquímica. Instituto Nacional de 
C.1rdiología, Juan Bad iano No. 1, Col. Sección 16, T1al pan, México D.F. 140S0, Mexico. 
Tel.: +52 55 55 73 29 11. 
E-mail address: SClren960104@hotmaíl.com (S. Rodr íg u ez-Enríquez ~ 
1357-2725{S - see front matter e 2010 Elsevier Ltd. AII ríghrs reserved. 
doi: 1 0.1016fj.bioceL201 0.07.010 
" 2010 Elsevier Ud. AII rights reserved. 
80 mm Hg w hereas in the ir hypoxic regions, it may reach a value of 
2.5-10 mm Hg(equivalent to 3-13 J.lM O,. caJcu lated by Horan and 
Koch. 2001 )(Kallinowski et al., 1989; Vaupel et al., 199 1; Hunjan et 
al., 1998; Erickson et al., 2003). The developmentor hypoxic regions 
in solid turnors is a typ ical characteristic linked ro rnalignant pheno-
rype, metastasis, chemo-, immuno- and radio-therapy resistance, 
high genet ic instab ility and apoptosis tole rance (Graeber et al.. 
1996; Bristow and Hill, 2008 ). Sorne of these processes a re rnodu-
lated by HIF-1a, a key transc riptiona l factor that regulates the gene 
transcriptio n of prote ins involved in angiogenes is, ce llu la r prolif-
era rio n, eryrhro poietic and vascu la rization pathways (Weidemann 
and Johnson, 2008). allowing t issues to adjust to scarce oxygen 
availab ili ty. At rnetabo lic leve!, HIF- l a increases the gene t ran-
sc ript ion of speci ftc isoenzyrnes of alrnost all g lyco lyt ic enzyrnes 
and transporte rs (reviewed in Marín· Hernández et al., 2009); in 
consequence. the glycolyt ic flux increases by at least 2-ti rnes in the 
rnajority of neoplasias (Altenberg and Greulich , 2004; Walenta e t 
al.. 2004) . 
In spite of the hypoxic-glyco lytic activation, the tumor intracel-
lular ATP pool drastically diminishes (30-60%) w hereas phosphate 
augrnents (Muelle r-Kl ieser et al., 1990; Heerle in et al., 2005;