UNIVERSIDAD NACIONAL AUTÓNOMA DE MÉXICO Instituto de Biotecnología Obtención y caracterización de cristales catalíticos de cloroperoxidasa T E S I S Que para obtener el título de Doctora en Ciencias Presenta I.Q. Marcela! Ayala Aceves Cuernavaca, Morelos. TFSIS rOM FALLA t í ORiGBI 2002 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. A mi querido esposo Andoni. A mi querido amigo Raúl. A la familia Ayala Aceves, con amor. Si sobrevives, si persistes, canta, sueña, emborráchate. Es el tiempo del frío: ama, apresúrate. El viento de las horas barre las calles, los caminos. Los arboles esperan: tú no esperes, éste es el tiempo de vivir, el único. Jaime Sabines Agradecimientos Al Dr. Rafael Vázquez Duhalt, por su orientación durante mi formación como investigadora. Gracias por el entusiasmo y las oportunidades que me brindaste para completar este trabajo. A los miembros de mi comité tutorah Dr. Agustín López-Munguía Canales, Dr. Eduardo Horjales Reboredo y Dr. Manuel Jiménez Estrada, por su comprometida participación durante el desarrollo de este proyecto. Al M. en C. Raunei Tinoco Valencia y a la Biól. Rosa Román Miranda por el entrenamiento y apoyo técnico proporcionado a lo largo de mi estancia en el Instituto de Biotecnología. A los grupos del Dr. Rafael Vázquez, Dr. Agustín López y Dr. Eduardo Horjales por su excelente disposición y su colaboración incondicional. Al Dr. Riccardo Basosi y a su grupo de investigación por su hospitalidad y cooperación durante mi estancia en la Universidad de Siena, Italia. A los miembros del jurado: Dr. Mario Soberón Chávez, Dr. Sergio Revah Moiseeu, Dr. Edmundo Castillo Rosales, Dr. Eduardo Bárzana García, Dr. Enrique Rudiño Pinera y Dr. Rodolfo Quintero Ramírez, por el tiempo dedicado a la revisión de esta tesis y las sugerencias aportadas para su perfeccionamiento. Al Instituto Mexicano del Petróleo (FIES 98-110-VI) y ai Consejo Nacional de Ciencia y Tecnología (33611-U) por el respaldo proporcionado a este trabajo de investigación. Al Consejo Nacional de Ciencia y Tecnología y a la UNAM por la beca y los apoyos para estudios de posgrado otorgados durante el periodo septiembre 1997-agosto 2002. ÍNDICE Resumen 1 Introducción 3 1. Antecedentes 1.1 Aplicaciones industriales de las enzimas 5 1.2 Estabilización de proteínas 11 1.3 Cristales entrecruzados de enzimas 17 1.4 La cloroperoxtdasa de Caldariomyces fumago 23 1.5 Biodesulfurización 34 2. Hipótesis y objetivos 40 3. Protocolo experimental 3.1 Reactivos 42 3.2 Equipo 42 3.3 Métodos 43 4. Resultados y discusión 4.1 Biodesulfurización 49 4.2 Cristalización de la cloroperoxidasa 58 4.3 Entrecruzamiento de cristales de cloroperoxidasa 64 4.4 Propiedades de los cristales entrecruzados de cloroperoxidasa 74 5. Conclusiones y perspectivas 81 6. Bibliografía 84 7. Apéndices 96 7.1 Biocatalytic oxidation of fuel as an alternative to biodesulfurization 7.2 Substrate specificity and ionization potential in chloroperoxidase- catalyzed oxidation of diesel fuel 7.3 Biocatalytic chlorination of aromatic hydrocarbons by chloroperoxidase of Caldariomyces fumago 7.4 Spectroscopic characterization of a manganese-lignin peroxidase hybrid isozyme produced by Bjerkandera adusta in the absence of manganese: evidence of a protein centred radical by hydrogen peroxide 7.5 Suicide inactivation of peroxidases and the challenge of engineering more robust enzymes 7.6 Cross-línked crystals of chloroperoxidase RESUMEN La cloroperoxidasa de Caldariomyces fumago cataliza la oxidación de compuestos azufrados tales como los que se encuentran presentes en el diesel. Por esta razón, la capacidad catalítica de esta enzima resulta atractiva en procesos como la biodesuífurización. Sin embargo, al igual que muchas proteínas, la cloroperoxidasa pierde rápidamente su actividad en condiciones típicas de este tipo de industria, como son altas temperaturas o presencia de solventes orgánicos. En este trabajo se reporta por primera vez la obtención de cristales entrecruzados de una peroxidasa, la cloroperoxidasa de Caldariomyces fumago. Los cristales de esta enzima fueron entrecruzados con glutaraldehído para producir cristales catalíticamente activos e insolubles. A diferencia de otras preparaciones inmovilizadas de cloroperoxidasa, los cristales entrecruzados fueron más termoestables que la enzima libre. Se propone que este aumento en la estabilidad se debe a la conservación de la estructura tridimensional en el arreglo cristalino. Adicionalmente, se encontró que ios cristales sin entrecruzar retienen más actividad que la enzima libre en presencia de un solvente orgánico con bajo contenido de agua. Al comparar con ia enzima libre se encontró que los cristales entrecruzados tienen una actividad específica menor. La eficiencia del entrecruzamiento depende de varios factores: los impedimentos estéricos que representa la glicosilación de la enzima; el reducido número de grupos amínos primarios en la superficie de la proteína y las condiciones de reacción poco favorables para el entrecruzamiento pero esenciales para la estabilidad de la enzima. Se estableció que un aumento en el número de grupos aminos en la superficie de la proteína favorece el entrecruzamiento de los cristales, aunque la actividad específica observada sigue siendo muy baja. Los resultados obtenidos en este trabajo nos permiten suponer que las modificaciones inespecíficas del glutaraldehído sobre algunos residuos y los problemas de accesibilidad de sustratos voluminosos hacia los sitios activos son responsables de este comportamiento. En conclusión, se obtuvieron cristales entrecruzados de cloroperoxidasa cuya principal ventaja sobre la enzima libre es la mayor estabilidad que el estado cristalino confiere a las moléculas de proteína dentro del cristal. Estos cristales son más estables que la enzima libre en presencia de solventes orgánicos con bajo contenido de agua y resisten altas temperaturas. ABSTRACT Chloroperoxidase from Caídariomyces fumago catalyzes the oxidation of sulfur- containing compounds üke those present in diesel fuel. The catalytic efficiency of the enzyme in sulfoxidation reactions makes it a potential catalyst for biodesulfurization. However, chloroperoxidase is rapidly inactivated when exposed to high temperatures or organic solvents. In this work, cross-linked crystals of a peroxidase have been obtained for the first time. Chloroperoxidase from Caídariomyces fumago was crystallized and the crystals were cross-linked with glutaraldehyde. The cross-ünked crystals were insoluble and catalytically active. An increase in the number of amino groups on the protein surface enhanced the cross-linking extent of the crystais. Unlike other immobilized preparations, cross-linked crystals of chloroperoxidase are more thermostable than the soluble enzyme. Probably the increased stability could be due to a better conservation of the protein tridimensional conformation inside the crystalline array. Moreover, non cross- linked crystals retained more activity than the soluble enzyme after exposure to an organic solvent. Cross-ünked crystals showed lower specific activity then the soluble enzyme. This might be due to glutaraldehyde-derived inespecific modifications of some residues that might be important for catalysis. Accesibility problems for bulky hydrophobic substrates to active sites may also explain the lower activity of cross-linked crystals. INTRODUCCIÓN La cloroperoxidasa de Caldariomyces fumago tiene aplicaciones potenciales en diversos campos de la biotecnología. Una de ellas es la biodesulfurízación de combustibles derivados del petróleo, como el diesel. Dentro de las características que hacen atractiva a la cloroperoxidasa para la industria petrolera están su eficiencia catalítica en la reacción de sulfoxidación y el amplio intervalo de sustratos azufrados que reconoce. El diseño de un catalizador estable y reciclable es esencial para el desarrollo de procesos económicamente viables. Se ha reportado que los cristales entrecruzados de enzimas representan un tipo de biocatalizador atractivo para procesos que involucran condiciones de operación tales como el uso de solventes orgánicos y alta temperatura. Esto se debe a que dentro de un cristal de proteína existen contactos intermoieculares que estabilizan la estructura de las moléculas, lo que por un lado desfavorece la desnaturalización y la agregación de las moléculas de proteína y por otro lado mantiene la actividad catalítica bajo condiciones altamente desfavorables para las enzimas libres. Tomando esto en consideración, en este trabajo se estudió la posible aplicación de la cloroperoxidasa en procesos de desulfurización. De esta manera se planteó como objetivo principal la obtención de cristales catalíticamente activos de cloroperoxidasa para el desarrollo de un biocatalizador insoluble y más estable que la enzima libre. 1. ANTECEDENTES Antecedentes 1.1 Aplicaciones industriales de las enzimas La función de una enzima es la de cualquier catalizador: acelerar la velocidad de una reacción química, permitiendo al sistema alcanzar más rápido el equilibrio de reacción. Una enzima de origen microbiano es un catalizador con una estructura compleja de carácter proteico que se produce a partir de recursos renovables por medio de una fermentación. Cabe mencionar que la complejidad estructural de las enzimas es el resultado de millones de años de evolución, lo cual les permite catalizar diversas reacciones con gran especificidad. La complejidad de las enzimas se traduce en ventajas muy particulares. Una de ellas es su alta selectividad química. Gracias a esta selectividad disminuyen los subproductos y las reacciones secundarias en el medio de reacción, lo que reduce costos y facilita la recuperación del producto. Además, una enzima puede ser muy selectiva para discriminar entre estereoisómeros y distinguir regiones y grupos funcionales dentro del sustrato. Las enzimas son capaces de funcionar a temperatura ambiente y presión atmosférica. Adictonalmente, las enzimas son biodegradabies. Por último, las enzimas son catalizadores muy eficientes, acelerando la velocidad de una reacción en varios órdenes de magnitud. Por sus ventajas las enzimas son cada vez más usadas en procesos industriales. Un ejemplo es la producción de más de 30,000 toneladas/año de acrilamida utilizando a la enzima nitrilo hidratasa (Figura 1.1.1) [1]. El proceso químico utiliza un catalizador de cobre que genera desechos tóxicos como HCN, opera a temperaturas entre 80° y 100°C y siempre produce ácido acrílico como subproducto. Por otro lado, el proceso enzimático opera a 10°C y produce acrilamida con un rendimiento del 100% [2]. Estas condiciones de reacción son ambientalmente deseables pues disminuyen el gasto en energía y la emisión de gases al ambiente. NH, H?O Nitrilo hidratasa Figura 1.1.1 Ejemplo de una reacción de producción de una sustancia química a escala industrial mediante una biotransformación Antecedentes Debido a que las enzimas han sido diseñadas a la perfección por la naturaleza para cumplir con una función fisiológica, su actividad y estabilidad no siempre satisfacen las expectativas de un industrial o de un químico orgánico. De hecho, existe cierta resistencia a la idea de utilizar enzimas en ciertos campos de la industria. Esto se debe a la percepción de que las enzimas son catalizadores costosos y poco estables en las condiciones de operación de ciertos procesos, como la presencia de solventes orgánicos y alta temperatura. Esta idea no siempre es correcta. Por ejemplo, los precios de los catalizadores biológicos y los químicos no son tan diferentes (Tabla 1.1.1). Tabla 1.1.1 Precios de catalizadores industriales [2] Enzima Deshidrogenasa láctica Esterasa de hígado porcino Penicilina amidasa Aspartasa Tripsina Lipasa Glucosa ¡somerasa Proteasa de detergente Glucoamilasa Dólares / kg 100,000 15,000 10,000 10,000 5,000 5,000 500 250 100 Catalizador BINAP Platino ChiraPhos Sharpless Pd(Diphos)2 Rh(PPh3)3CI Jacobsen Chirald Níquel Raney Dólares / kg 40,000 12,000 10,000 10,000 5,000 2,000 1,000 500 30 En términos económicos es más pertinente preguntarse cuánto contribuye la enzima al costo del producto final. La Tabla 1.1.2 muestra algunos ejemplos de productos cuya obtención involucra el uso de enzimas y nos permite afirmar que en ocasiones el costo de las enzimas representa solamente una pequeña parte del valor del producto [3]. Tabla 1.1.2 Costo de algunas enzimas en relación al valor del producto [3] Producto Detergentes Jarabe fructosado Etanol Queso Cantidad de enzima (ppm) 150-200 150-200 300-400 3-6 Costo de la enzima (% del valor del producto) 1-4 2-3 2-3 0.1-0.3 Antecedentes En cuanto a la estabilidad de las enzimas, el factor realmente importante es su estabilidad operacional. En la mayoría de los procesos biocatalíticos se utilizan preparaciones de enzimas inmovilizadas, que pueden ser recuperadas y reutilizadas. La estabilidad y estabilización de las enzimas serán discutidas en la siguiente sección. El desarrollo de las aplicaciones industriales de las enzimas ha sido espectacular. En la década de los 80's las enzimas industriales tenían un mercado de 300 millones de dólares. Como resultado de los avances en la tecnología de ADN recombinante, las compañías incrementaron su producción e introdujeron nuevas enzimas al mercado. En ia Tabla 1.1.3 se muestran ejemplos de algunas enzimas y el tipo de actividad en el que encuentran aplicación. La lista completa que incluye todas las enzimas que se utilizan en diagnóstico clínico, análisis de alimentos y en ingeniería genética, es larga y muy impresionante. Tabla 1.1.3 Ejemplos de enzimas y su aplicación [3] Enzima Amilasas Glucosa isomerasa Proteasas fi-1,4-Galactosidasa Pectínasa Lipasas Colesterol oxidasa Glucosa oxidasa Hexoquinasa Alcohol deshidrogenasa. ADN polimerasas Enzimas de restricción Sustrato Almidón Glucosa Proteína Lactosa Pectina Grasa y aceites Colesterol Glucosa Glucosa, fructosa Etanol ADN ADN Industria Jarabe fructosado, etanol Jarabe fructosado Detergentes, endulzantes, piel, lácteos, carne, bebidas, terapia Lácteos Bebidas Detergentes, lácteos Análisis de alimentos Diagnóstico clínico Análisis de alimentos (bebidas) Análisis de alimentos (bebidas) Técnicas moleculares Técnicas moleculares Actualmente la industria de las enzimas tiene ventas anuales por 1,600 millones de dólares y aplicaciones en el procesamiento de almidón y alimentos (45%), en la formulación de detergentes (34%), en la industria textil (11%), en la industria de la piel (3%) y en procesamiento del papel (1.2%) [4]. Un signo del avance de esta industria es que más del 60% de las enzimas industriales son productos recombinantes. Es Antecedentes interesante notar que las industrias de detergentes, aumentos y almidón representan e! 75% del mercado y las enzimas que se utilizan son principalmente hidroiasas: proteasas, amilasas, lipasas y celulasas. Adicíonalmente existe un mercado importante para las enzimas con aplicación en el diagnóstico clínico, las tecnologías analíticas y la industria farmacéutica. El volumen de ventas de algunas de las enzimas que se utilizan actualmente en diversos campos de la industria y la investigación se muestra en la Tabla 1.1.4. Tabla 1.1.4 Volumen de ventas de algunas enzimas [4] Enzima Aplicación Millones de dólares Subtilisina Detergentes 200 Quimosina Fabricación de queso 140 Enzimas de restricción Técnicas moleculares 100 Taq polímerasa Técnicas moleculares 80 Enzimas terapéuticas Farmacéutica 2,300 Pocas enzimas son vendidas directamente al público, como las enzimas incluidas en detergentes y en ablandadores de carne. En el sector químico y farmacéutico, las enzimas se utilizan para producir mejores o nuevos productos, los cuales son separados del medio de reacción y purificados. La Tabla 1.1.5 muestra el mercado mundial para algunos de estos productos obtenidos por reacciones enzimáticas. Tabla 1.1.5 Volumen de ventas de productos obtenidos enzimáticamente [2] Enzima Glucosa isomerasa Aminopeptidasa, termolisina Nitrilo hidratasa Penicilina amidasa Producto Jarabe fructosado Aspartamo Acrilamida Ácido 6-aminopenicilánico Millones de dólares 1000 800 300 200 Industria Alimentaria Aumentaría Química Farmacéutica La especificidad de una enzima en la síntesis y resolución de intermediarios racémicos permite añadir valor agregado al producto y ahorrar costos [5]. Los compuestos enantioméricamente puros ocupan un lugar cada vez más importante dentro de las industrias química y farmacéutica. De los 100 fármacos más importantes a Antecedentes nivel mundial, más de la mitad son compuestos quirales. El volumen mundial de ventas de fármacos quirales excedió los 123,000 millones de dólares en el 2000 [6]. Algunas de las enzimas que se utilizan para producir compuestos quirales con aplicación en diversas industrias se muestran en la Tabla 1.1.6 [7,8], Tabla 1.1.6 Compuestos quirales producidos enzimáticamente [8] Producto Aplicaciones, enzima, % ee, producción, compañía Intermediario en la síntesis de Trusopt para el tratamiento de glaucoma. Alcohol deshidrogenasa de Neurospora crassa. >98%, toneladas, Zeneca Life Science Molecules, Inglaterra (4S,6S)-5,6-Dihidro-4-hidroxi-6-metil-4H-tieno[2,3b]tiopiran-7,7-dÍóxido Intermediario en síntesis de herbicidas. Oxidasa de Beauviera bassiana. >98%, reactor de 120,000 L. BASF AG, Alemania Ácido (R)-2-(4'-hidroxifenoxi) propiónico NHj+ Aditivo nutricional. D-aminoácido transaminasa de R^^coo Bacillus. 100%, toneladas. NSC Technologies, Monsanto, EL) Intermediario para farmacéuticos y pesticidas. Lipasa de Burkholdeha platarii > 9 9% , >100 ton/año. BASF AG, Alemania D-aminoácido (S)-I-Feniletilamina o o Intermediario en la síntesis de paclitaxel usado en el tratamiento de cáncer. Lipasa de Burkholdería cepacia. >99.5%, kilogramos. Bristol Myers Squibb, EU Acetato de (3R.4S) cis-azotidinona Antecedentes Tabla 1.1.6 (continuación) O (2R,3S)-3-{4-Metox¡fenil) ácido glicídico metil éster Intermediario en la síntesis de diltiazem. Lipasa de Serratia marcescens. 99.9%. Tanabe Seiyaku Co. Ltd Japón y DSM, Holanda o • Endulzante bajo en calorías. Termolisina de o Bacillus proteolicus. >99.9%, >2000 ton/año Holland Sweetener Company, Holanda Aspartamo (a-L-aspartil-L-fenilalanina metil éster) Aditivo nutricional para niños, deportistas y ancianos. Carnitina deshidratasa de Escherichia 0 H ° coli, 99.9%, 300 ton/año. Lonza AG, Suiza 3-hidroxi-4-(trimetilam¡no) butanoato de (R)-L-Carnitina ee: exceso enantiomérico Las enzimas pueden competir con los catalizadores químicos. Las enzimas son biocatalizadores muy eficientes, son muy selectivas y son ambientalmente inocuas. Todas estas propiedades se traducen en beneficios económicos. Sin embargo las enzimas no siempre son estables bajo las condiciones de operación que requieren varios procesos industriales, lo cual limita su aplicación en este campo. Las enzimas pueden manipularse para modificar o mejorar su selectividad, estabilidad y actividad mediante la inmovilización, la modificación química o la ingeniería genética de proteínas. Es importante explorar aplicaciones novedosas y estrategias para el uso de las enzimas. En la siguiente sección se tratan algunas de las estrategias que permiten mejorar la estabilidad de ¡as enzimas. 10 Antecedentes 1.2 Estabilización de proteínas por inmovilización La estabilidad estructural de las proteínas se debe a la suma de múltiples interacciones débiles. Las interacciones hidrofóbicas, los puentes de hidrógeno, los puentes salinos, las interacciones dipolo-dipolo y otras interacciones electrostáticas entre los grupos de átomos que conforman una proteína contribuyen a su estabilidad. A pesar de que la energía de cada una de estas interacciones es pequeña, el gran número que existe en una proteína hace que su contribución energética a la conservación de la estructura sea significativa [9]. Cuando las proteínas pierden su estructura tridimensional, es decir se desnaturalizan, pierden también su actividad biológica. Hay varias maneras de estabilizar fa estructura tridimensional de una proteína, por ejemplo manteniendo la temperatura por debajo de la temperatura fisiológica, añadiendo sustancias estabilizadoras o manipulando las propiedades de la proteína mediante ingeniería genética, modificación química e inmovilización [10]. Mientras que mantener la temperatura baja es una buena opción para almacenar enzimas, no es una alternativa práctica desde el punto de vista de un proceso ya que significa un gasto de energía y una disminución en la velocidad de reacción. La adición de sustancias estabilizadoras como polioles implica la introducción de compuestos extraños al proceso que después es necesario remover. Los primeros métodos para modificar las propiedades de una enzima fueron métodos químicos. Actualmente con las técnicas de ingeniería genética de proteínas, es posible cambiar ciertas propiedades de una enzima y sobre todo determinar el papel que juegan residuos individuales en su función y estabilidad. Sin embargo, la modificación química sigue siendo una herramienta importante para influir en ciertas características de las enzimas [11]. Esta estrategia ofrece opciones únicas, ya que existe un gran número de agentes muy versátiles, de diferente tamaño, hidrofobicidad y reactividad que permiten realizar modificaciones químicas complejas. Se ha reportado que una de las modificaciones químicas que refuerzan la estructura compacta de la molécula de manera que no pierda su conformación es el entrecruzamiento intra e intermolecular con agentes bifuncionales. En ocasiones este entrecruzamiento químico confiere estabilidad térmica a las enzimas. [12-14]. Otro enfoque es introducir moléculas de alto peso molecular que contribuyen a la solubiiización y estabilización de las enzimas en solventes orgánicos [15-17]. Generalmente estas moléculas son anfifílicas y 11 Antecedentes es posible que su efecto protector se deba a !a rigidización de la estructura proteica y a la eliminación de repulsiones electrostáticas entre cargas superficiales [18]. La ingeniería de proteínas utiliza diferentes estrategias para modificar las propiedades de las enzimas [19]. El diseño racional es una de estas estrategias y requiere un conocimiento detallado de la estructura y mecanismo catalítico de la proteína para alterar sus propiedades de una manera dirigida [20]. Las interacciones dentro de la proteína regulan su estabilidad, actividad y especificidad. El conocimiento de esta compleja red de interacciones es incompleto y aún no se puede predecir confiablemente qué cambios estabilizarán a una enzima sin modificar otras propiedades [21]. En el otro extremo del abanico de estrategias se encuentra la mutación ai azar. Este enfoque evolutivo no requiere mucha información sobre la enzima y se basa en la mutación y recombinación acompañada de una selección de las vanantes con la propiedad deseada. Su éxito depende de la capacidad de realizar esta selección dentro de bibliotecas muy grandes de proteínas mutantes [22], Las enzimas son catalizadores, por lo que no son consumidas durante las reacciones que involucra un proceso. Sin embargo, durante su uso pierden actividad por inactivación o desnaturalización. La inmovilización afecta la estabilidad de las enzimas, generalmente protegiéndolas contra la desnaturalización. Adicionalmente, cuando las enzimas son utilizadas en forma soluble su recuperación y reutilización no son económicamente viables. Bajo estas condiciones la enzima residual contamina al producto y su remoción involucra gastos de purificación adicionales. Una forma de eliminar estos inconvenientes es la inmovilización de la enzima en un soporte adecuado que permita separarla del producto, recuperarla y reutilizarla. Durante la inmovilización las moléculas de proteína se fijan a un material inerte, formando otra fase sólida en el medio de reacción. Los sustratos y productos pueden difundir entre las dos fases. De esta manera, la enzima no contamina a la fase que contiene al producto y puede ser recuperada y reutiüzada. Debido a que la inmovilización supone un gasto adicional, tiene que haber una ventaja o beneficio económico para optar por la enzima inmovilizada. Generalmente este beneficio es una reducción en los gastos de purificación del producto. Algunas de las biotransformaciones industriales utilizan enzimas inmovilizadas, como se muestra en la Tabla 1.2.1. 12 Antecedentes Tabla 1.2.1 Aplicaciones industriales de enzimas inmovilizadas [23] Enzima Producto Aminoacilasa L-aminoácidos Cianidasa Ácido fórmico Glucoamilasa D-glucosa Glucosa ¡somerasa Jarabe fructosado p-1,4-Galactosidasa Lácteos sin lactosa Nitrito hidratasa Acrilamida Penicilina amidasa Ácido 6-aminopenicilánico Termolisina Aspartamo La inmovilización influye sobre la estabilidad y la actividad de la enzima, ya que el microambiente de la enzima inmovilizada depende de las propiedades del soporte. La inmovilización también puede generar problemas de difusión, por lo que la elección del soporte y del método de inmovilización es crítica. Dentro de las propiedades del soporte que son importantes considerar se encuentran la composición química, los grupos funcionales, la estabilidad mecánica y química, el tamaño de poro y el diámetro de partícula [24]. Los métodos de inmovilización pueden clasificarse en tres grupos (Tabla 1.2.2) dependiendo del tipo de interacción utilizada. Tabla 1.2.2 Métodos para inmovilizar enzimas Covalentes No covalentes Inclusión Entrecruza miento Cristalización Micelas Unión a un soporte Adsorción a un soporte Membranas Plásticos biocatal¡ticos Secado Polímeros Películas La unión covalente de una o varias zonas de la proteína a un soporte puede compactar la estructura y estabilizar a la enzima frente a la desnaturalización [25,26], Los llamados plásticos biocatalíticos consisten en la incorporación de moléculas de proteína a redes poüméricas, lo que rigidiza a la proteína y disminuye !a modificación de su estructura tridimensional [27]. En el caso de las enzimas en solventes orgánicos puede utilizarse métodos no covalentes, aprovechando el hecho de que la mayoría de las proteínas no son solubles 13 Antecedentes en estos solventes. La cristalización de enzimas, en la que la enzima es a la vez catalizador y soporte, será discutida en la siguiente sección. Una de las alternativas más simples y efectivas cuando se utilizan solventes inmiscibles en agua es la adsorción de la proteína a un soporte [28]. Otra alternativa es !a deshidratación, por ejemplo por liofilización, la cual genera un polvo enzimático seco que es insoluole en medios orgánicos y puede dispersarse y recuperarse al final de la reacción [29,30]. La inmovilización de la enzima por inclusión en una matriz polimérica, en una membrana o en micelas invertidas confina a la enzima a una fase diferente, permitiendo su separación del medio de reacción y su reutilización. Debido a la ausencia de interacciones covaientes entre la proteína y el soporte, su inclusión en una membrana no contribuye a estabilizar a la enzima. En el caso de las micelas invertidas la movilidad de las proteínas se ve restringida y puede aumentar su estabilidad [31]. La inclusión de una enzima en una película también genera un biocatalizador estable y reutilizable [32]. La mayoría de los métodos de inmovilización requiere la optimización de una serie de variables. Por ejemplo, ía relación enzima/soporte, el pH, el solvente utilizado y la presencia de aditivos protectores. En condiciones no optimizadas pueden existir problemas como son el bloqueo de sitios activos, la conformación inadecuada de la enzima, la baja afinidad entre la enzima y el soporte y la desnaturalización de la enzima. Las primeras biotransformaciones industriales se realizaron sobre sustratos solubles en agua [33]. Actualmente muchos de los compuestos orgánicos de interés para la industria química y farmacéutica son poco solubles en agua. La biocatálisis en medios orgánicos es un campo muy estudiado debido a los retos que supone [34,35]. Las proteínas son inestables y tienden a desnaturalizarse rápidamente en mezclas de agua y solvente orgánico. Sin embargo en solventes orgánicos anhidros o con bajo contenido de agua, la flexibilidad de las proteínas se ve reducida. Mientras que el agua actúa como lubricante, las moléculas de solvente orgánico no interactúan de la misma manera con Ía proteína ya que no tienen la misma capacidad que el agua de formar puentes de hidrógeno; además, debido a la baja constante dieléctrica de estos solventes las interacciones electrostáticas intra-proteína son más fuertes. En consecuencia, en un solvente orgánico anhidro la proteína tiene menos movilidad y su estructura es más rígida. La disminución en la movilidad de las proteínas en presencia de solventes orgánicos ha sido detectada mediante la difracción de rayos X de cristales de incubados en algunos solventes [36,37]. Aparentemente, la movilidad y por consiguiente la 14 Antecedentes desnaturalización de las moléculas de proteína es menos favorecida en solventes orgánicos con bajo contenido de agua [38,39]. Desde el punto de vista catalítico esta falta de flexibilidad tiene sus desventajas y puede disminuir la actividad enzimática drásticamente. Sin embargo, no todos los solventes interactuan de la misma manera con las proteínas. Los solventes orgánicos polares son más perjudiciales para la actividad enzimática que los solventes más hidrofóbicos, debido a que los primeros pueden secuestrar con más facilidad las moléculas de agua fuertemente unidas a la proteína. Estas moléculas de agua son esenciales para mantener cierta movilidad y sin este mínimo contenido de agua, la estructura enzimática podría ser demasiado rígida para efectuar la catálisis [34,40]. Existen varias maneras de mejorar la estabilidad y actividad de las enzimas en solventes orgánicos [39,41]. Por ejemplo, se pierde más actividad al liofilizar a una enzima que al ponerla en contacto con un solvente orgánico. Para resolver este problema, pueden añadirse lioprotectores y análogos de sustratos para mantener mejor la estructura tridimensional original de la proteína y la integridad del sitio activo [41,42]. También pueden añadirse aditivos que favorecen la flexibilidad en el medio orgánico, como sales, agentes desnaturalizantes o macromoléculas que reducen las interacciones electrostáticas [43-45]. Es muy importante considerar que la adición de algunos solventes orgánicos puede afectar e! estado de protonación de algunos residuos; estos cambios pueden desestabilizar a la enzima [46]. Se pueden seguir varias estrategias para controlar el estado de protonación de los aminoácidos cuando las proteínas se encuentran en un medio orgánico, por ejemplo adicionar amortiguadores de pH orgánicos, que son solubles en este tipo de solventes [47,48]. Recientemente se ha reportado el uso de cristales entrecruzados como una alternativa para la inmovilización de enzimas [24]. Dentro de los cristales, la desnaturalización de las moléculas de proteína es desfavorecida. Por esta razón los cristales retienen su actividad catalítica bajo condiciones que desnaturalizan a la mayoría de las proteínas, incluyendo la presencia de solventes orgánicos y alta temperatura. La biocatálisis en solventes orgánicos expande las posibilidades del uso industrial de las enzimas. También supone vencer una serie de obstáculos, puesto que la mayoría de las enzimas evolucionaron para realizar su función en un medio acuoso y su estabilidad en presencia de solventes orgánicos no siempre es la deseada. 15 Antecedentes Aunque la presencia de solventes orgánicos afecta a las propiedades de las enzimas, es posible usarlas de una manera eficiente si se identifican los factores clave que limitan la actividad y sistemáticamente minimizar o eliminar sus efectos. En la Tabla 1.2.3 se mencionan algunos de estos factores y las soluciones que permiten mejorar la actividad enzimática en solventes orgánicos. Tabla 1.2.3 Problemas de la Problema • Problemas de difusión • Sitios activos bloqueados • Cambios de estructura • Partición desfavorable de sustratos hidrofóbicos • Movilidad reducida • pH no óptimo biocatálisis en solventes orgánicos [38] Solución • Disminuir el tamaño de partícula; usar agitación vigorosa • No utilizar partículas amorfas • Adicionar aditivos (iioprotectores, polímeros) • Seleccionar un solvente que ¡nteractúe desfavorablemente con el sustrato • Optimizar la actividad de agua; adicionar co-solventes o aditivos. • Controlar el pH de la deshidratación; usar amortiguadores de pH orgánicos 16 Antecedentes 1.3 Cristales entrecruzados de enzimas En los cristales de proteínas, las moléculas están arregladas simétricamente con una orientación definida. Estos cristales son materiales porosos que poseen canales entre las moléculas de proteína. El solvente en los canales puede ocupar entre 27% y 65% del volumen del cristal y está disponible para la difusión de moléculas [49,50]. Los canales pueden tener un diámetro de entre 20 y 100 A, dependiendo de la proteína y del arreglo espacial del cristal [51]. La reactividad de las enzimas en su estado cristalino depende de la preservación de la estructura nativa y de la difusión de los sustratos y productos a través de los canales. Aunque algunas enzimas no retienen su actividad en el estado cristalino debido a efectos estéricos o rigidización generados por contactos intermoleculares, varias enzimas pueden cristalizar en una conformación catalíticamente activa. Los cristales de enzimas globulares pueden considerarse como arreglos abiertos de moléculas con algunos sitios de contacto intermolecular y en promedio 50% de su volumen constituido por solvente. En la Figura 1.3.1 se muestra una imagen generada por computadora de la carboxipeptidasa (PM=110,000) en estado cristalino. Como se puede apreciar en la imagen, las moléculas de proteína en los cristales no forman una estructura impenetrable sino que existen espacios entre las moléculas a través de todo el cristal. El sustrato de esta enzima, un dipéptido (PM=208, en rojo), puede fácilmente difundir a través de los poros para llegar al sitio activo de las moléculas de enzima en el interior del cristal. Figura 1.3.1 Representación de un cristal de carboxipeptidasa 17 Antecedentes La cristalización de proteínas es un proceso en el que participan muchos factores. La obtención de los cristales es el resultado de la precipitación controlada de una solución concentrada de proteína. Para que los cristales mantengan su integridad, la combinación de los factores que permitieron la cristalización no debe alterarse. Pequeños cambios en el pH, la temperatura y ia concentración de sales u otros componentes de la solución de cristalización son suficientes para destruir estos cristales. A diferencia de los cristales de sales, los cristales de proteína son mecánicamente muy frágiles debido a la naturaleza y al menor número de contactos que existen entre las moléculas. De hecho, esta característica se utiliza para averiguar si un cristal es de sal o de proteína: la prueba consiste en tocar al cristal con una aguja y observar al microscopio si el cristal se pulveriza. El entrecruzamiento químico de ias moléculas que componen un cristal de enzima sirve para estabilizar el arreglo cristalino, de manera que aumenta la resistencia del cristal fuera de la solución de cristalización. Quiocho y Richards fueron los primeros en reportar en 1964 la estabilización de un cristal mediante el entrecruzamiento con glutaraldehído [52]. Su intención fue estabilizar mecánicamente a! cristal para que resistiera por más tiempo la radiación de los rayos X y de esta manera obtener un mejor patrón de difracción. Estos investigadores encontraron que los cristales se volvían insolubles después del tratamiento y que retenían su actividad enzimática en solución acuosa [53,54]. Posteriormente, diversos autores reportaron algunas propiedades de cristales entrecruzados de otras enzimas, como las constantes catalíticas, la afinidad por sustratos, inhibición, el perfil de pH, los problemas de difusión y la resistencia a proteólisis y a agentes desnaturalizantes [55-58]. El interés por los cristales entrecruzados se renovó recientemente cuando una compañía patentó la tecnología para producir cristales entrecruzados de enzimas (CLEC) [59,60], Los cristales entrecruzados de enzimas tienen ciertas propiedades atractivas desde el punto de vista de la biocatálisis industrial, la más obvia es su carácter de partícula macroscópica compuesta por enzima. En efecto, el cristal entrecruzado es en sí mismo un conjunto de moléculas de enzima inmovilizadas que no requiere de soportes adicionales. Las partículas que conforman este biocatalizador pueden separarse del medio de reacción mediante procedimientos convencionales, como filtración o centrifugación. Actualmente estas preparaciones están disponibles comercialmente [61]. La Tabla 1.3.1 lista los cristales entrecruzados ofrecidos por Altus Biologics, Inc. 18 Antecedentes Tabla 1.3.1 CLEC disponibles comercialmente [61] Enzima Origen Alcohol deshidrogenasas Thermobacterium brokii, hígado de caballo Asparaginasa E. coli Elastasa Páncreas porcino Esterasas Hígado porcino, bacterias Glucosa Isomerasa Streptomyces rubiginosus Glucosa Oxidasa Aspergiilus niger Lipasas Páncreas porcino, bacterias, hongos Luciferasa Bacterias y luciérnagas Lisozima Clara de huevo Penicilina Acilasa E. coli Subtilisina Bacillus sp. Termolisina Thermoproteolyticus rokko Ureasa Extracto de frijol A menudo la cristalización de proteínas se concibe como un arte, debido a la dificultad metodológica que representa esta etapa en el proceso para obtener la estructura tridimensional de una proteína. Debido a la calidad que deben tener, la obtención de los cristales apropiados para obtener información estructural mediante la difracción de rayos X es un proceso laborioso que puede tomar mucho tiempo. Es importante que los cristales tengan un buen tamaño (> 0.2 mm ) con el fin de producir una señal detectable, deben ser lo suficientemente resistentes para soportar la manipulación y la radiación necesarias para colectar los datos y es importante que no tengan defectos que impidan la interpretación del patrón de difracción. Sin embargo, para aplicaciones biocatalíticas no es necesario contar con cristales tan grandes ni tan perfectos. Más aún, se requieren microcristales con dimensiones pequeñas para evitar problemas de difusión. Los cristales con menos de 100 um en alguna de sus dimensiones son preferidos para aplicaciones químicas porque las limitaciones de transferencia de masa son reducidas y el tamaño es adecuado para recuperarlos por filtración [24,62]. Microcristales con estas características se producen muy frecuentemente durante la cristalización de proteínas. Incluso existen procedimientos para producir cristales con estas características a gran escala [63], 19 Antecedentes La propiedad más interesante y atractiva desde el punto de vista de la biocatálisis es la sorprendente estabilidad que muestran los cristales entrecruzados. Estos cristales son entre 2 y 4 órdenes de magnitud más estables que las enzimas libres cuando son expuestos a solventes orgánicos, temperaturas altas y degradación enzimática [62,64- 70]. La Tabla 1.3.2 muestra la estabilidad de los cristales entrecruzados bajo diferentes condiciones. Tabla 1.3.2 Estabilidad de cristales entrecruzados de diferentes enzimas [62] Enzima Lipasa Termolisina Subtilisina Alcohol deshidrogenasa Medio Acuoso, 40°C 50% tetrahidrofurano, T amb. 50% dimetilsulfóxido, T amb. 50% dimetilformamida, T amb. Proteasa Acuoso, 60°C 50% dimetilformamida, 40°C 50% acetona, 40°C 50% tetrahidrofurano, 40°C Acetato de etilo, 55°C Proteasa Acuoso, 70°C 50% tetrahidrofurano, T amb. Acuoso, 30°C 25% isopropanol, 40°C t 1/2 enzima libre 5h 2h -s 18h 20 h <1 h > 6h > 8h 3h 1.5 h j < 3 días <5min >• 10 min 45 min 14 días 4h t 1/2 cristal 13 días > 5 días > 5 días > 18 días > 4 días 5h 170 h >3 meses > 4 días Los cristales entrecruzados también son estables a la agitación mecánica; experimentos realizados en pequeña escala que simulan las condiciones encontradas en un tanque agitado, sugieren que la velocidad de agitación necesaria para mantener una dispersión homogénea de los cristales no provoca su destrucción [71]. Adicionalmente los cristales entrecruzados mantienen su capacidad de catalizar reacciones con alta selectividad y pueden ser recuperados y reutilizados en varios ciclos de reacción, lo cual aumenta la productividad del proceso [72,73]. Todas estas 20 Antecedentes propiedades señalan a los cristales entrecruzados como catalizadores muy prometedores. El escalamiento a 100 L de un proceso biocatalítico para la resolución de una mezcla racémica utilizando cristales entrecruzados de lipasa indica que es una alternativa económicamente viable [73]. La aplicación de los cristales entrecruzados no se limita a la síntesis de compuestos ópticamente activos [74]. En la literatura pueden encontrarse ejemplos de otra índole. Por ejemplo, los cristales de la enzima organofósforo Nasa pueden ser de utilidad en la degradación de compuestos neurotóxicos [75]. También se ha discutido el potencial uso de los cristales en la aplicación de vacunas y otros fármacos de lenta liberación [76]. Incluso se ha sugerido que su estabilidad y porosidad los convierten en mallas moleculares bioorgánicas para separar moléculas pequeñas [51]. La estabilidad de los cristales entrecruzados proviene del arreglo simétrico de las moléculas y del entrecruzamiento de este arreglo. Si el cristal no se entrecruza, no es estable y se disuelve. Para algunas enzimas se ha encontrado que cuando se entrecruza a la enzima libre o un precipitado amorfo de la misma, su estabilidad no aumenta de la misma manera que cuando se parte de la enzima en forma cristalina [62]. En la matriz cristalina en donde el empaquetamiento de las moléculas de proteína se acerca al límite teórico, existe un gran número de interacciones intermoleculares de tipo hidrofóbico y electrostático que pueden mejorar la estabilidad de la proteína y desfavorecer la desnaturalización y la agregación. Adicionalmente la estructura de las proteínas puede ser estabilizada por el entrecruzamiento de manera equivalente a una enzima inmovilizada a un soporte mediante múltiples enlaces, previniendo de esta manera la desnaturalización [77]. La actividad específica de los cristales entrecruzados puede ser igual o menor a la que exhibe la enzima libre [77,78]. Entre los factores que pueden influir sobre la actividad de manera importante se encuentran el tamaño del cristal, el tamaño del sustrato y la conformación de la enzima en el cristal. El espesor del cristal debe mantenerse dentro de un límite para evitar problemas difusionales. De la misma manera, los sustratos demasiado voluminosos tendrán problemas para difundir al interior del cristal a través de los poros. Si el espesor del cristal o el tamaño del sustrato son demasiado grandes, la enzima que se encuentra en el centro del cristal no participará en la catálisis y solamente será productiva la enzima localizada en la superficie del cristal. El efecto macroscópico derivado de esta situación es una reducción en la velocidad de ia reacción. El espesor y el tamaño de los sustratos 21 Antecedentes apropiados varían según la enzima y el tipo de cristal que ésta forme. Por otro lado, la movilidad de la enzima puede estar restringida dentro del cristal y debido a la falta de flexibilidad, la velocidad de la reacción también puede verse disminuida. Otras variables que deben cuidarse ai utilizar cristales entrecruzados, sobre todo en solventes orgánicos, son el pH y la forma de deshidratar al cristal. Debido a que la concentración de proteína en los cristales es muy alta, debe utilizarse un amortiguador lo suficientemente fuerte para controlar el pH. En solventes orgánicos existe además la dificultad de mantener la protonación de los residuos en el estado óptimo para ia catálisis. Esto puede resolverse utilizando amortiguadores solubles en solventes orgánicos [79]. También es importante revisar el modo en que se deshidrata un cristal antes de utilizarlo en un solvente orgánico. Esto puede afectar profundamente el contenido del agua esencial para la catálisis y la conformación de las moléculas de enzima, perjudicando la actividad catalítica [80]. La actividad de los cristales entrecruzados en solventes orgánicos puede verse afectada por los mismos factores que las enzimas libres (ver Tabla 1.2.3) y puede verse disminuida en varios órdenes de magnitud [81]. Las soluciones son similares en ambos casos: optimizar el estado de protonación de los residuos, seleccionar un solvente que minimice la partición desfavorable del sustrato y mantener la actividad de agua en un nivel apropiado para la catálisis. La inmovilización de enzimas para su uso en un proceso biocatalítico se traduce en beneficios económicos. Permite que el catalizador sea separado del producto y recuperado y reutilizado. También representa beneficios adicionales, como la estabilización de las enzimas. Existen numerosos métodos y técnicas de inmovilización que pueden ser utilizados de acuerdo con las propiedades de la enzima y de la reacción que se desea catalizar. Los cristales entrecruzados de enzimas se presentan como una forma de inmovilización novedosa que confiere estabilidad a las moléculas de proteína, previniendo su desnaturalización e inactivación en presencia de solventes orgánicos y altas temperaturas. A diferencia de las proteínas inmovilizadas en soportes inertes, los cristales enzimáticos son catalizadores con altísima actividad específica en términos de velocidad por unidad de peso. Sin embargo, al igual que con las enzimas de uso industrial, actualmente la mayor parte de estos cristales se componen de hidrolasas, lo que limita su aplicación a otros procesos como los de óxido-reducción. 22 Antecedentes 1.4 La cloroperoxidasa de Caldariomyces fumago Las peroxiclasas son enzimas que utilizan peróxido para catalizar la oxidación de un sustrato. La mayoría contiene metales y de éstas, la mayor parte contiene un complejo fierro-porfirina llamado hemo [82]. Existe otro tipo de metalo-peroxidasas que contienen vanadio o selenio en lugar de hemo. Las hemo peroxidasas pueden clasificarse con base en su secuencia de aminoácidos en dos superfamilias: peroxidasas de mamíferos y peroxidasas de plantas [83]. La primera superfamilia incluye enzimas como la lactoperoxidasa, la mieloperoxidasa y la prostaglandina H sintasa. La segunda superfamilia se encuentra dividida en tres clases. La clase I incluye a las peroxidasas intracelulares, como la citocromo c peroxidasa de levadura y la ascorbato peroxidasa de cloroplasto. La clase II comprende a las peroxidasas extracelulares de hongos, como la lignino peroxidasa y la manganeso peroxidasa. La clase III incluye a las peroxidasas de plantas como la peroxidasa de rábano blanco. Con base en una comparación entre las secuencias de aminoácidos, se ha propuesto que las hemo peroxidasas de bacterias, hongos y plantas tienen un ancestro común [83]. En la Tabla 1.4.1 se listan algunas hemo peroxidasas y su función en la naturaleza. Tabla 1.4.1 Ejemplo de hemo peroxidasas y su función biológica [84] Peroxidasa Cloroperoxidasa Citocromo c peroxidasa Peroxidasa de rábano Lactoperoxidasa Lignino peroxidasa Mieloperoxidasa Origen Caidariomyces fumago Saccaromyces cerevisiae Armocracia rusticana Leche bovina Phanerochaete chrysosporium Leucocitos humanos Función Biosíntesis de caldariomicina Reducción de H2O2 y oxidación de citocromo c Biosíntesis de hormonas de plantas Antimicrobiana Degradación de lignina Anti microbiana Existe un tipo de peroxidasas llamadas haloperoxidasas. En la naturaleza las haloperoxidasas catalizan la oxidación de haluros utilizando H2O2, resultando en la halogenación de compuestos orgánicos. Mientras que existen haloperoxidasas sin ningún grupo prostético en su sitio activo, también existen metalo-haloperoxidasas que 23 Antecedentes pueden tener vanadio o un grupo hemo. Las primeras son de origen bacteriano mientras que las que contienen vanadio son producidas principalmente por algas y hongos. Por otro lado, la única hemo haloperoxidasa que ha sido caracterizada bioquímicamente es la cloroperoxidasa de Caldariomyces fumago. Otras hemo haloperoxidasas de algas y de bacterias han sido aisladas, aunque no han sido caracterizadas [85-87]. La cloroperoxidasa de C. fumago tiene una actividad específica de 1000 jxmol/mg min, la más alta de todas las haloperoxidasas, usando como sustrato una dicetona cíclica llamada monocorodimedona [86]. Las haloperoxidasas no muestran homología entre ellas ni con las hemo peroxidasas descritas anteriormente [88,89]. La cloroperoxidasa de C. fumago presenta un plegamiento único y es estructuraimente diferente a cualquier otra hemo peroxidasa conocida. Es decir, la cloroperoxidasa no parece estar relacionada evolutivamente con las demás hemo peroxidasas. Esta es sólo una de las características que hacen de la cloroperoxidasa una proteína poco común entre las de su clase. La cloroperoxidasa de C. fumago es una proteína extracelular que puede alcanzar una concentración de hasta 100 mg/L [90]. La enzima tiene un peso molecular aproximado de 42,000 Da y un grado de glicosilación de 20-35%. El patrón de glicosilación depende de la cepa y del medio de cultivo; en la enzima que se utilizó para este estudio los carbohidratos predominantes son mañosa y glucosa y en menor proporción glucosamina, xilosa y galactosa [91,92]. La cloroperoxidasa es una proteína monomérica con un grupo prostético hemo IX (Figura 1.4.1). Los sitios de glicosilación son principalmente treoninas y serinas, aunque también se han detectado modificaciones en algunas asparaginas [93]. OH Cl Cl o Figura 1.4.1 Hemo IX Figura 1.4.2 Caldariomicina 24 Antecedentes La cloroperoxidasa fue descubierta por el grupo de Lowell P. Hager en la década de los 60's mientras investigaba e! metabolismo de un compuesto clorado llamado caldariomicina, producido por el hongo C. fumago (Figura 1.4.2). La investigación de Hager estaba motivada por el interés de estudiar la producción en la naturaleza de compuestos halogenados [94,95]. Al descubrir que una sola enzima era !a responsable de la halogenación observada in vivo, Hager se dio a la tarea de purificarla y encontró a la cloroperoxidasa. Para su sorpresa Hager descubrió que además de tener capacidad de halogenar, in vitro esta enzima presentaba actividades similares a las de otras enzimas. La cloroperoxidasa tiene actividad de halogenasa, peroxidasa, catalasa y peroxigenasa (Tabla 1.4.2) [96]. Es la peroxidasa más versátil que se conoce y ha sido ampliamente estudiada y caracterizada. Tabla 1.4.2 Actividades de la cloroperoxidasa de C. fumago Actividad enzimática Halogenación (halogenasa) Deshidrogenación (peroxidasa) Descomposición de peróxido (catalasa) Inserción de oxígeno (peroxigenasa) Reacción catalizada RH + H2O2 + Cr + H+ -> R-CI + 2H2O 2RH + H2O2 -> R-R + 2H2O 2ROOH -> O2 + 2ROH R + H2O2 -> R-OH + H2O Sustratos Dicetonas cíclicas, fenoles.ésteres fenólicos Metoxifenoles (guaiacol), anilinas, indol Peróxido de hidrógeno, peróxidos orgánicos Ttoanisoles, alquenos, bencilos La evidencia experimental parecen indicar que la cloroperoxidasa sigue el ciclo de reacción que se muestra en la Figura 1.4.3. El ciclo catalítico de la enzima comienza con ia reacción de una molécula de peróxido con el sitio activo (ruta 1). El peróxido se une al hemo y después de una transferencia de protones facilitada por los aminoácidos que rodean a! sitio activo, la molécula se rompe para liberar agua y formar un intermediario enzimático conocido como Compuesto I. El rompimiento heterolítico del enlace oxígeno-oxígeno de la molécula de peróxido requiere dos electrones, por lo que debe ocurrir un rearreglo electrónico intramolecular. En la cloroperoxidasa, así como en la mayoría de las peroxidasas, un electrón proviene del fierro y el segundo electrón proviene de la porfirina. El resultado es la presencia de un radical libre deslocalizado en 25 Antecedentes la porfirina y la formación de un enlace dativo entre el átomo de fierro y un átomo de oxígeno [97]. Compuesto X ,-o-ci Estado basal Figura 1.4.3 Ciclo catalítico de la cloroperoxidasa El Compuesto I puede seguir tres caminos. En primer lugar puede sufrir una reducción de un electrón para formar el Compuesto II, mediante la oxidación de una molécula de un sustrato orgánico (ruta 2). Durante la formación de este intermediario, el electrón transferido desde el sustrato reduce al radical centrado en la porfirina. El Compuesto II es igual para todas las peroxidasas y sólo tiene un equivalente oxidativo que se encuentra localizado en el fierro. Este intermediario puede catalizar la oxidación de una segunda molécula de sustrato (ruta 3). De esta manera la enzima regresa a su estado basal y puede comenzar el ciclo de nuevo. Las rutas 1, 2 y 3 comprenden el ciclo clásico de las peroxidasas que utilizan peróxido para catalizar la deshidrogenaron de dos moléculas de sustrato formando radicales libres, los cuales generalmente reaccionan entre sí formando un polímero. La actividad peroxidasa de la cloroperoxidasa corresponde aproximadamente al 10% de la actividad de una peroxidasa clásica cuando se mide con guaiacol, un sustrato típico de ías peroxidasas [96]. 26 Antecedentes Otra alternativa es que el Compuesto I reaccione con una molécula de sustrato unida en el sitio activo, cerca del fierro del grupo hemo. En este caso el átomo de oxígeno se transfiere directamente al sustrato (rutas 4 y 5). Esta reacción de transferencia de dos electrones en un paso es típica de las monooxigenasas como el citocromo P450, con la diferencia de que estas enzimas pueden utilizar O2 como donador de oxígeno y las peroxidasas no. Las peroxidasas comunes no catalizan esta reacción. En tercer lugar, el Compuesto I puede reaccionar con una segunda molécula de peróxido para formar oxígeno molecular y el alcohol correspondiente (ruta 6). Las rutas 1 y 6 corresponden a la actividad catalasa. Las peroxidasas clásicas catalizan esta reacción con un 0.01% de la reactividad de una catalasa clásica. La cioroperoxidasa no es tan eficiente como una catalasa clásica, pero aún así en presencia de cloruro y peróxido de hidrógeno tiene aproximadamente 20% de la actividad detectada con una catalasa [96,98]. Por último, puede ocurrir que el Compuesto I reaccione en presencia de iones haluro para formar un intermediario halogenante (ruta 7). Existe controversia sobre la identidad de este agente en las reacciones catalizadas por la cioroperoxidasa, específicamente las de cloración y bromación. Los estudios sobre la naturaleza de los productos sugieren que la halogenación es química y favorecen las rutas 8 y 9. Esto se debe principalmente a que no se observa estereoselectividad en los productos y éstos son idénticos a los que se obtienen cuando la reacción se lleva a cabo químicamente adicionando HOCI [85, 99-101]. Por otro lado, los estudios sobre la velocidad de halogenación favorecen la ruta 10, ya que la velocidad enzimática de formación de un intermediario químico como HOCI es muy lenta en comparación con la velocidad de halogenación del sustrato [102,103]. Sin embargo, la enzima es muy compleja y es afectada por un gran número de variables, por lo que los datos cinéticos pueden ser interpretados de diferente manera [104,105], Por otra parte, recientemente se detectó un intermediario enzimático durante la halogenación (Compuesto X) [106]; aunque esta evidencia fortalece un mecanismo de reacción como el de la ruta 10, todavía quedan dudas por resolver y no existe un consenso en la comunidad científica al respecto. A pesar del enorme potencial de las peroxidasas como biocatalizadores, su aplicación comercial es limitada. Por ejemplo, la lactoperoxidasa se usa como aditivo antimicrobiano en pastas de dientes (Zendium™) y la peroxidasa de soya se utiliza 27 Antecedentes como aditivo en la preparación de pan (Quest International) [84]. La aplicación comercial de las peroxidasas que más se ha desarrollado es en el área de análisis clínicos [3]. Aunque las moléculas que son ha!ogenadas> por la cloroperoxidasa no son ópticamente activas, existe un número importante de reacciones catalizadas por esta enzima cuyos productos son interesantes desde el punto de vista sintético. En la Tabla 1.4.3 se muestran algunas de estas reacciones y ejemplos de las aplicaciones potenciales que tienen los productos [107,108,109]. Tabla 1.4.3 Reacciones enantioselectivas catalizadas por la cloroperoxidasa Reacción Producto Aplicación potencial Oxidación de sulfuras Sulfóxidos quirales Fármacos para el tratamiento de úlcera Epoxidación de olefinas Epóxidos quirales Intermediarios en la síntesis de famacéuticos contra el VIH Oxidación de indol 2-Oxoindol Intermediario en la síntesis de farmacéuticos antiinflamatorios La cloroperoxidasa también tiene aplicaciones potenciales en el área ambiental. Por ejemplo, la enzima cataliza la formación de complejos de ADN y compuestos poliaromáticos y puede utilizarse como un modelo para determinar la genotoxicidad de estos compuestos [110j. Adicionalmente, se sabe que la cíoroperoxidasa puede modificar las petroporfirinas presentes en la fracción de asfáltenos del petróleo pesado [111]. Este tratamiento podría ser eficiente para la remoción de metales de los asfáltenos. La versatilidad catalítica de la cloroperoxidasa ha sido objeto de diversos estudios y ha sido explicada de diversas formas. En las peroxidasas, el fierro está pentacoordinado con los cuatro nitrógenos pirrólicos y con un nitrógeno de una histidina. Esta histidina, llamada proximal, está muy conservada en todas las peroxidasas. Del otro lado del hemo se encuentran otros tres residuos muy conservados: una histidina, llamada distal, una arginina y un ácido glutámico. Estos tres residuos forman una red de puentes de hidrógeno y en conjunto facilitan la unión y el rompimiento de la molécula de peróxido [112]. El ambiente del hemo en la cloroperoxidasa es polar, al igual que en las demás peroxidasas. Sin embargo, una característica más que hace de la cíoroperoxidasa una 28 Antecedentes enzima atípica entre las de su clase es que el ligando distal es un ácido glutámico y el ligando proximal es una cisteína [93]. La coordinación entre el fierro y la cisteína es una característica que la cloroperoxidasa comparte con el citocromo P450 y que le confiere a estas dos proteínas propiedades espectroscópicas muy similares [113]. Anteriormente se pensaba que ei papel de la cisteína como ligando del hemo era muy importante para la catálisis de !a cloroperoxidasa. Se sugirió que la cisteína actuaba como un fuerte donador de electrones, promoviendo el rompimiento de la molécula de peróxido y permitiendo la formación de! Compuesto I de una manera muy eficiente [113]. También se propuso que la presencia de la cisteína podría explicar la asombrosa capacidad oxidativa de la enzima, dado que otras peroxidasas pueden oxidar yoduro y bromuro pero la cloroperoxidasa es la única que puede oxidar al cloruro. La mejor manera de explorar estas posibilidades es mediante mutagénesis sitio dirigida. Durante muchos años se intentó tener un sistema de expresión para la cloroperoxidasa recombinante y sólo recientemente se tuvo éxito [114,115], Obviamente, lo primero que se intentó fue cambiar la cisteína por histidina [114], Esta mutación no alteró drásticamente las propiedades catalíticas de la enzima y demostró que el papel de la cisteína no es tan crítico como se pensaba. Es posible que el inusual ligando distal estabilice al Compuesto I y sea el responsable del gran poder oxidativo de la cloroperoxidasa [116]. Esta hipótesis aún no se ha comprobado. Las características atípicas del sitio activo de la cloroperoxidasa no terminan aquí. Por un lado, la mayoría de las peroxidasas abstrae un electrón del sustrato, sin transferirle oxígeno. Por otro lado, a pesar de que también forman un intermediario tipo Compuesto I, las monooxigenasas como el citocromo P450 transfieren directamente un átomo de oxígeno a su sustrato. Se ha propuesto que en las peroxidasas solamente el borde del hemo es accesible al solvente, mientras que el acceso al fierro del hemo está restringido [117]. Esta propuesta se analizó utilizando un sustrato suicida, ia fenilhidrazina, que en todas tas peroxidasas forma un aducto con el grupo fenilo unido a uno de los carbonos del hemo. Sin embargo, en otras hemo proteínas como la mioglobina, la hemoglobina y el citocromo P450 en las que se sabe que el fierro se encuentra en un sitio activo abierto, se demostró la formación de un complejo con el fenilo unido al fierro. Una alineación de las secuencias de aminoácidos de las peroxidasas de plantas y hongos sugiere que en todas ellas existe un residuo aromático adyacente a la histidina 29 Antecedentes dista! [83]. En el caso de la peroxidasa de rábano blanco, este residuo es una fenilalanina. Se realizaron mutaciones puntuales para reemplazar a la fenilalanina por un residuo menos voluminoso como la alanina. Las mutantes mostraron una mayor velocidad de formación del Compuesto I y una velocidad de transferencia de oxígeno a un sustrato azufrado 10 veces mayor que la enzima nativa [118]. En el caso de la cloroperoxidasa, se sabe que el borde del hemo y el fierro son accesibles desde el solvente. Tanto el complejo fenilo-fierro como el aducto fenilo-hemo han sido detectados [119], lo que sustenta que la topología del sitio activo de la cloroperoxidasa es una especie de híbrido, compartiendo características de las peroxidasas comunes y de las monooxigenasas como el citocromo P450. La cloroperoxidasa también tiene un residuo aromático, una fenilalanina, en eí túnel de acceso al fierro del hemo. Este residuo está cerca del ácido glutámico catalítico y es flexible. Su función parece ser acomodar pequeños sustratos hidrofóbicos en la vecindad del hemo [116]. Será interesante analizar los resultados de una mutación puntual en este aminoácido para cambiarlo por un grupo más pequeño. En resumen, las peroxidasas suprimen la transferencia de oxígeno al sustrato bloqueando el acceso al fierro con una barrera de aminoácidos y de esta manera favorecen la transferencia de electrones a través de la porfirina. Por otro lado, en enzimas como el citocromo P450 la estructura del sitio activo permite la entrada del sustrato para reaccionar directamente con el fierro. La cloroperoxidasa cuenta con el acceso al borde del hemo y también con el acceso al fierro del hemo, lo que explica su capacidad de catalizar reacciones tanto de peroxidación como de peroxigenación. La Tabla 1.4.4. resume las propiedades del sitio activo de algunas hemo proteínas. Tabla 1.4.4 Propiedades del sitio activo de algunas hemo proteínas Proteína Cloroperoxidasa Peroxidasa de rábano Citocromo P450 Citocromo c peroxidasa Mioglobina Catalasa Ambiente del hemo Polar Polar No polar Polar No polar Polar Ligando axial Cisteína Histidina Cisteína Histidina Histidina Tirosina Acceso al borde . del hemo Sí Sí No Sí Sí No Acceso al Fe del hemo Sí Limitado Sí Sí Sí No 30 Antecedentes Como todas las peroxidasas, la cloroperoxidasa es inactivada por el peróxido durante el proceso catalítico. En ausencia de un sustrato reductor o en presencia de un exceso de peróxido, la proteína actúa como reductor. De esta manera, los equivalentes oxidativos son transferidos a la cadena peptídica o a la porfirina ocasionando modificaciones que provocan la inactivación de las enzimas. En la Tabla 1.4,5 se muestra el tiempo de vida media de la cloroperoxidasa en presencia de diferentes concentraciones de peróxido de hidrógeno y a diferentes valores de pH. Esta inactivación de las peroxidasas se traduce en una estabilidad operacional baja y limita su aplicación industrial. Tabla 1.4.5 Estabilidad de la cloroperoxidasa en presencia de H2O2 [120] PH 5 5 5 4 6.2 H2O2(nM 30 200 1000 1000 1000 ) t1í2(h) 1.1 1.0 0.5 0.5 0.1 Se ha propuesto un mecanismo de inactivación general para las hemo peroxidasas (Figura 1.4.4) [121] (ver Apéndice 7.5). En presencia de un exceso de peróxido y en ausencia de un sustrato reductor, puede formarse un intermediario enzimático con un radical superóxido llamado Compuesto III (ruta 1). Este intermediario es sumamente reactivo y puede inactivar a la proteína de varias maneras. Una posibilidad es que el radical oxide a la porfirina, con la consecuente destrucción del grupo hemo y la liberación del fierro de la porfirina (ruta 2). Otra posibilidad es que los residuos que conforman el sitio activo se oxiden al actuar como reductores donando electrones (ruta 3). Los radicales libres que se formen pueden desplazarse a través de la cadena peptídica hasta llegar al residuo con menor potencial de reducción. Por último, el Compuesto III puede liberar al radical superóxido, que no es estable en solución y se descompone para formar radicales hidroxilo (ruta 4). Estos radicales son especies oxidantes muy poderosas y pueden atacar varios sitios de la proteína, incluyendo el sitio activo. En la literatura existe evidencia que soporta la existencia de estas tres rutas para la inactivación suicida de las hemo peroxidasas. Debido a la rapidez con que ocurre la oxidación de ta proteína, es difícil discernir cuál es la secuencia de estas reacciones. 31 Antecedentes + Fe (III) Oxidación de la proteírta Figura 1.4.4 Mecanismo propuesto de inactivación suicida de las peroxidasas Debido a que no existe un sistema eficiente de expresión para la cloroperoxidasa recombinante, no ha sido posible utilizar la ingeniería genética de proteínas para reducir ia velocidad de la inactivación por peróxido. Sin embargo, existen otras alternativas para incrementar el tiempo de vida media de la enzima en presencia de peróxido. Una de las estrategias consiste en mantener la concentración de peróxido en niveles bajos, de tal manera que todos los equivalentes oxidativos se transfieran al sustrato y no a la proteína. Para mantener estas condiciones, se han ensayado diferentes maneras de adicionar el peróxido: continuamente, en adiciones sucesivas y alimentando según el progreso de la reacción, utilizando un sensor para peróxido [122]. También se ha intentado generar in situ el peróxido mediante la generación electrolítica del mismo [123] o la co-inmoviiización de una glucosa oxidasa. En presencia de glucosa y O2, la glucosa oxidasa produce peróxido de hidrógeno que puede ser utilizado por la cloroperoxidasa para oxidar al sustrato [124]. Como resultado de estos esfuerzos se ha incrementado el número de recambio (número de moléculas de producto generadas por una molécula de enzima) hasta 860,000 [122]. Además de la estabilidad al peróxido, es importante mejorar otras propiedades de la cloroperoxidasa tales como la estabilidad en presencia de solventes orgánicos y la termoestabilidad. Estas propiedades se vuelven indispensables al pensar en cualquier proceso biocatalítico. Como se mencionó anteriormente, una opción es la ingeniería de 32 Antecedentes proteínas [125]. Debido a que esta opción ha surgido recientemente, todavía no existen resultados en este sentido. Una manera de estabilizar a ías enzimas es mediante su inmovilización en un soporte inerte. En la literatura se encuentra poca información sobre la inmovilización de cloroperoxidasa. La mayoría de las estrategias de inmovilización se basa en la adsorción de la enzima a soportes como perlas de vidrio, talco y celita [126- 128]. Solamente existe un reporte en el que la enzima es covalentemente unida a un polímero, presumiblemente mediante tos grupos hidroxilo de los carbohidratos expuestos en la superficie de la proteína [129]. Algunas de estas preparaciones muestran ventajas sobre la cloroperoxidasa soluble tales como su aplicación en solventes orgánicos y el reuso en varios ciclos de reacción [124,126,130]. Sin embargo ninguna incrementa su termoestabilidad. La cloroperoxidasa en solución es inactivada rápidamente cuando se incuba a temperaturas mayores de 50°C [92]. La cloroperoxidasa de C. fumago es un catalizador muy versátil con aplicación en varios campos. El mayor reto consiste en obtener un biocatalizador que retenga su actividad enzimática bajo las condiciones que usualmente se encuentran en un proceso como altas temperaturas, agitación mecánica y presencia de solventes orgánicos. 33 Antecedentes 1.5 Biodesulfurización La biotecnología tiene un nicho de aplicación potencial en la industria petrolera. Se han identificado varios organismos capaces de oxidar o reducir azufre, remover nitrógeno, eliminar metales y efectuar biocraqueo del petróleo. Con estos tratamientos la calidad del petróleo puede mejorar al modificarse algunas de sus propiedades tales como el contenido de azufre y la viscosidad. Estos cambios permiten reducir costos e incrementar el valor del petróleo [131]. Hoy en día, cerca de 80 millones de barriles de petróleo se extraen diariamente. Los estimados de las reservas de combustibles fósiles indican que podemos seguir a este ritmo por 70 años [132]. La mayor parte de estos combustibles están contaminados con azufre, y al ser quemados liberan grandes cantidades de óxidos de azufre a la atmósfera. Estos óxidos son los responsables de la lluvia acida. Adicionalmente, estos óxidos envenenan a los convertidores de los automóviles, los cuales hacen más eficiente la combustión de la gasolina. Cuando los convertidores son envenenados, los hidrocarburos sin quemar llegan a la atmósfera, contribuyendo a la contaminación de! aire. Por estos motivos la legislación es cada vez más severa en cuanto al contenido de azufre en los combustibles. La concentración de azufre en el petróleo crudo varía entre 1,000 y 30,000 ppm. Actualmente el contenido de azufre en las gasolinas es de 500 ppm en Estados Unidos y Japón. Para el 2005 se legislará el uso de combustible con 50 ppm de azufre en Japón y Europa. Para el 2006, Estados Unidos restringirá el contenido de azufre en combustibles a 15 ppm [133]. La hidrodesulfurización (HDS) es el proceso químico que se utiliza en las refinerías para remover el azufre de los combustibles derivados del petróleo. Actualmente existen más de 35 unidades de HDS en el mundo con una capacidad total de 1.5 millones de barriles al día [134]. En este proceso se utiliza un catalizador inorgánico para convertir el azufre orgánico en ácido sulfhídrico. Para lograrlo se hace reaccionar al petróleo crudo con hidrógeno a altas presiones (10-200 atm) y temperaturas entre 290 y 455°C. 'El azufre orgánico en las fracciones ligeras de petróleo {p.e. la gasolina) se encuentra en forma de tioles, sulfuros y tiofenos. Estos compuestos son eficientemente removidos por la HDS. Sin embargo, en las fracciones más pesadas (p.e. el diesel) el azufre forma parte de heterociclos poliaromáticos como benzotiofenos, dibenzotiofenos y dibenzotiofenos sustituidos cerca del azufre (Tabla 1.5.1). Estos compuestos son difíciles de remover mediante la HDS. La biodesulfurización (BDS) se presenta 34 Antecedentes actualmente como una alternativa complementaria a la HDS, capaz de modificar los compuestos recalcitrantes para el proceso químico. Tabla 1.5.1 Compuestos organoazufrados presentes en combustibles fósiles [131] Tipo de combustible Compuesto azufrado Estructura química Temperatura de destilación ro O O H-,0 O OH OH O H7O OH 'OH OH OH CH O ' "O* CH Va P-NH; NhT •Q NH ,P- Figura 4.3.2 Estructuras del glutaraldehído y mecanismo de reacción en solución acuosa acida o neutra (P-NH2 = proteína) 65 Resultados y discusión Por otro lado, en condiciones alcalinas el giutaraldehído sufre una condensación aldólica intermolecular que produce oligómeros de aldehidos ot,p-insaturados (VI) como se muestra en la Figura 4.3.3. Al reaccionar con los grupos aminos de una proteína podrían formarse bases de Schiff que son estabilizadas por !a resonancia del doble enlace adyacente (Vil) [164], Esto explicaría la estabilidad de los aductos formados. Adicionalmente puede ocurrir una adición de Michael al doble enlace (VIII) [165]. Ambos productos son estables a la hidrólisis acida. La capacidad de formar oligómeros de diferente tamaño le confiere al giutaraldehído una ventaja como entrecruzante. Al existir polímeros de diferente longitud, aumenta la probabilidad del entrecruzamiento intermolecular dentro del cristal de proteína. VI VIII Figura 4.3.3 Mecanismo de reacción del giutaraldehído en soluciones alcalinas Considerando lo anterior, se realizaron pruebas de entrecruzamiento con la cloroperoxidasa libre. Cuando la cloroperoxidasa libre se puso en contacto con giutaraldehído a pH 5, no se detectó ningún cambio en el perfil de migración electroforético de la enzima en presencia o ausencia de un agente reductor (Figura 4.3.4, carriles 5 y 6). Esto se debe probablemente al reducido número de usinas superficiales de la enzima. Adicionalmente, bajo estas condiciones las usinas están 66 Resultados y discusión protonadas y son nucleófilos pobres. Por otro lado, al realizar la reacción de entrecruzamiento a pH 9 se observaron especies de mayor peso molecular que sugieren entrecruzamiento intermolecular de la proteína (carril 4 ). La adición de un agente reductor no modifica este perfil (carril 3). 1 Control con reductor a pH 5 2 Control con reductor a pH 9 3 Reacción con giutaraldehído + reductor, pH 9 4 Reacción con giutaraldehído, pH 9 5 Reacción con giutaraldehído + reductor, pH 5 6 Reacción con giutaraldehído, pH 5 7 Cloroperoxidasa nativa 8 Marcadores de peso molecular 107 kDa 67 kDa 44 kDa 3 4 5 6 7 Figura 4.3.4 SDS-PAGE de la reacción entre el giutaraldehído y la cloroperoxidasa libre Aunque un pH alcalino favorece la reacción con giutaraldehído, es imposible usar estas condiciones con la cloroperoxidasa debido a que ésta sufre una inactivación rápida e irreversible cuando el pH es mayor que 7.5 [166]. La Figura 4.3.5 muestra la actividad residual de la enzima después de incubarla durante diferentes intervalos de tiempo en soluciones con diferente pH a temperatura ambiente. 120 ra re s i TJ CO T3 100 i 80- 60 40 20 0 2 3 4 Tiempo (horas) Figura 4.3.5 Estabilidad al pH de la cloroperoxidasa libre 67 Resultados y discusión Para el caso de los cristales, se vanaron algunos parámetros de la reacción con glutaraldehído. El pH se modificó entre 5.5 y 6.5, la reacción se realizó a 4°C o 25°C y se utilizó un exceso molar de glutaraldehído sobre la proteína de entre 10 a 350,000 moles de glutaraldehído por moles de proteína. Todas las reacciones se monitorearon hasta 6 horas. En ningún caso se encontraron cristales que resistieran a la disolución y que conservaran su actividad. Después de reaccionar con el glutaraldehído los cristales ya no polarizaban, eran flexibles, se disolvían y presentaban baja o nula actividad. Es probable que el grado de entrecruzamiento no sea suficiente para mantener al cristal estable y activo al mismo tiempo, por las razones de baja disponibilidad de aminos primarios descritas arriba para la cloroperoxidasa soluble. La situación puede agravarse en los cristales, en donde las moléculas de proteína tienen una orientación espacial definida que puede afectar la probabilidad del entrecruzamiento. La Figura 4.3.6 muestra la distribución de las Usinas y de los grupos caboxilos de la cloroperoxidasa. La enzima libre en solución tiene 37 grupos carboxilos distribuidos homogéneamente, de los cuales alrededor de 25 están expuestos al solvente. Por otro lado, la enzima tiene cinco lisinas concentradas en una región de la proteína; solamente tres de ellas están expuestas al solvente. A B Figura 4.3.6 Grupos carboxilos y lisinas de la cloroperoxidasa de Caldariomyces fumago (carboxilos en azul, lisinas en magenta, hemo en rojo) B es la rotación de 180° de A TFSIS COH FALLA í.'E ORIGEN 68 Resultados y discusión Con la intención de favorecer la reacción de entrecruzamiento a un pH en que la cloroperoxidasa sea estable, se incrementaron los grupos aminos de la enzima libre utilizando una modificación química con carbodiimida sobre los grupos carboxilos superficiales. En presencia de una amina u otro nucleófilo, la carbodiimida promueve la formación de una amida en dos pasos (Figura 4.3.7) [25]. Inicialmente la carbodiimida reacciona con el carboxilo para formar un intermediario reactivo (O-acil-isourea). Posteriormente el intermediario reacciona con la amina para formar la amida correspondiente. NR2 O II II Ri-COOH + R2N=C=NR3 ^ - ^ Ri-COO- C ^ r *R r CO-NH-R 4 + R2NH-C-NHR3 HNRa ~NH2 Figura 4.3.7 Mecanismo de reacción de las carbodiimidas Para modificar a la cloroperoxidasa se utilizó la 1-eti!-3~(3-dimet¡!aminopropil) carbodiimida, que es soluble en agua, y una diamina como nucleófilo. De esta manera, cada grupo carboxilo modificado se convertiría en una amina primaria. Se encontró que la modificación química con hexandiamina en los carboxilos superficiales favorece el entrecruzamiento por glutaraldehído de la enzima soluble (Figura 4.3.8). La cloroperoxidasa libre modificada tiene una migración electroforética diferente a la cloroperoxidasa sin modificar (carriles 3 y 2, respectivamente). Mientras que la cloroperoxidasa sin modificar es ligeramente entrecruzada a pH 8 (carril 6), la reacción con glutaraldehído a pH 5 y 8 produce especies de mayor peso molecular con la cloroperoxidasa modificada químicamente (carriles 4 y 5, respectivamente). 1 Marcadores de peso molecular 2 Cloroperoxidasa (CPO) nativa 3 CPO-hexandiamina (CPO-HD) 4 Reacción con glutaraldehído + CPO-HD, pH 5 5 Reacción con glutaraldehído + CPO-HD, pH 8 6 Reacción con glutaraldehído + CPO, pH 8 44 kDa 1 2 3 4 5 6 Figura 4.3.8 SDS-PAGE de la reacción entre el glutaraldehído y la cloroperoxidasa libre modificada químicamente • TESIS COH FALLA DE ORIGEN 69 Resultados y discusión Para comprobar si la modificación química también favorecería el entrecruzamiento de los cristales, éstos se pusieron en contacto con carbodiimida y hexandiamina. Cualquier solución conteniendo compuestos con grupos aminos o carboxilos interferiría con la reacción de modificación. La solución de cristalización contenía acetato de zinc 0.1 M por lo que fue necesario modificarla sin perjudicar la integridad de los cristales para poder realizar la modificación química. En la Tabla 4.3.1 se muestran los compuestos que se probaron como sustitutos del acetato de zinc, tratando de reemplazarlo con compuestos que pudieran sustituir las interacciones del acetato (sulfatos) y manteniendo la fuerza iónica constante. El sulfato de zinc fue el único que permitió preparar una solución que mantuviera estables a los cristales durante el tiempo que dura la reacción de modificación. Tabla 4.3.1 Modificación de la solución de reacción Sal Acetato de zinc MES Sulfato de amonio Sulfato de zinc Cloruro de sodio Concentración (M) 0.1 0.3 0.1 0.075 0.3 Para comprobar que los cristales fueron realmente modificados químicamente con carbodiimida, se disolvió una parte de ellos antes de entrecruzarlos con glutaraldehído y se corrió una muestra en un SDS-PAGE. Se observó una migración aberrante como la obtenida para la cloroperoxidasa libre (Figura 4.3.8, carril 3). Adicionalmente se comparó el perfil de elución en una cromatografía de intercambio aniónico de la cloroperoxidasa nativa y modificada libre o en cristales; la proteína modificada se retiene menos en la columna, indicando una disminución en las cargas negativas. Sin embargo, esta información es sólo cualitativa y no permite determinar con precisión el grado de modificación química. Como alternativa se utilizó un método fluorométrico cuantitativo para determinar el grado de modificación de la proteína. Mientras que los grupos aminos primarios de la cloroperoxidasa libre aumentaron de 5 a 15 por molécula, para los cristales se detectaron hasta 10 grupos aminos primarios por molécula en promedio. Esta diferencia puede deberse a que la concentración de reactivos que se utiliza para modificar a los 70 Resultados y discusión cristales es menor a la concentración que puede utilizarse para modificar a la enzima libre. Debido a que la composición del medio afecta la estabilidad de los cristales, concentraciones altas de los reactivos los destruyen. Adicionalmente es posible que la modificación química de las moléculas dentro del cristal no sea homogénea ni ocurra en las mismas posiciones que para la cloroperoxidasa libre, ya que dentro del crista! existen impedimentos estéricos que no se encuentran en solución. Después de la modificación química los cristales siguen polarizando la luz, lo cual indica que el arreglo del cristal no ha sido afectado seriamente. Adicionalmente, al disolver los cristales modificados químicamente se observó que ia actividad de la enzima no fue afectada por la modificación. Después de realizar la modificación química sobre los cristales, éstos fueron entrecruzados con diferentes cantidades de glutaraldehído. La Figura 4.3.9 compara el perfil de solubilidad y actividad de los cristales sin (panel A) y con (panel B) modificación química después del entrecruzamiento. Es evidente que la modificación química favorece el entrecruzamiento, de tal manera que la cantidad de glutaraldehído necesaria para obtener cristales entrecruzados estables es menor. Sin embargo, los cristales todavía siguen perdiendo la mayor parte de su actividad. Aún en presencia de imidazol que actúa como ligando del fierro y puede proteger la integridad del sitio activo, los cristales siguen perdiendo actividad después del entrecruzamiento. Aunque los cristales no polarizan la luz, retienen actividad catalítica y son estables a !a disolución y a la temperatura. Las propiedades cinéticas de estos cristales entrecruzados se discuten más adelante. 0 2000 4000 6000 Exceso molar de giutaraldehído 0 2000 4000 6000 Exceso molar de glutaraldehído Figura 4.3.9 % proteína en cristal (•) y % actividad (A) de cristales entrecruzados (A) sin y (B) con modificación química 71 Resultados y discusión Un análisis de aminoácidos de los cristales entrecruzados utilizando este procedimiento muestra que el número de Usinas disminuye, lo que indica que durante el entrecruzamiento con giutaraldehído estos residuos reaccionan aunque su participación no es suficiente para estabilizar al cristal. El análisis no muestra el incremento de grupos aminos primarios debido a la modificación química con carbodiimida, ya que el enlace formado es de tipo amida y se hidroliza como un enlace peptídico. Además de las usinas, también disminuye el número de argininas y serinas (Figura 4.3.10). El giutaraldehído no es un reactivo específico para las Usinas, sino que reacciona con cualquier nucleófilo. Aunque a pH alcalino las usinas son los nucléofilos más potentes, bajo las condiciones de reacción descritas para obtener cristales entrecruzados de cloroperoxidasa las usinas están protonadas y no son muy reactivas. Es posible que el giutaraldehído, al reaccionar con otros aminoácidos como las argininas y las serinas, modifique grupos o zonas importantes para la catálisis. La cloroperoxidasa cuenta con 27 serinas, de las cuales 3 están localizadas cerca de la entrada al sitio activo. Estos resultados permiten suponer que estas modificaciones inespecíficas pueden afectar la catálisis al obstaculizar la unión de los sustratos ai sitio activo o al rigidizar la estructura de la proteína mediante la formación de enlaces intramoleculares. En consecuencia, es posible que estas modificaciones sean responsables en parte por la baja actividad observada para los cristales. g o c (O T3 C Figura 4.3.10 Análisis de aminoácidos de la cloroperoxidasa ( f¿ ) y de los cristales entrecruzados ( • ) 72 Resultados y discusión B) Otros entrecruzantes 1) Carboditmida-diamina/Cloruro de adipoilo El cloruro de adipoilo y la hexandiamina en proporción 1:1 se utilizan para producir Nylon 66. Los cristales de cloroperoxidasa modificados químicamente con hexandiamina fueron puestos en contacto con hexandiamina y cloruro de adipoilo 0.5:1 para formar redes que intercalaran cristales. Esta estrategia produciría un plástico biocatalítico con cristales inmovilizados. Aunque los cristales se incorporan al polímero, éste tiene una estuctura granulosa que lo hace muy frágil. Al poner en solución acuosa este polímero, los cristales se disuelven. Probablemente el entrecruza miento debe ser no sólo superficial sino también interno para que los cristales sean estables fuera de la solución de cristalización. 2) Periodato de sodio/Dihidrazida La idea detrás de esta estrategia era formar grupos aldehidos a partir de los cabohidratos superficiales de la cloroperoxidasa mediante su oxidación con periodato. Posteriormente se podrían formar enlaces intermoleculares con una diamina. La dihidrazida del ácido adípico resultaba un buen candidato para esta estrategia ya que posee dos grupos aminos en sus extremos y se comporta como un nucléofüo a pH neutro, debido a que tiene un pKa de 2.45. La reacción formaría una base de Schiff que podría ser estabilizada por reducción. Sin embargo, los cristales no soportaron la presencia del periodato, que es un oxidante muy fuerte y agresivo con las proteínas. 3) Carbodiimida-dihidrazida/Glutaraldehído Al modificar los carboxilos de la proteína con dihidrazida, se obtendrían grupos aminos superficiales desprotonados a pH neutro. Esto se traduciría en nucleófilos más poderosos y posiblemente la reacción de entrecruzamiento con glutaraldehído sería más favorable. Desafortunadamente la modificación química con esta diamina no pudo realizarse porque los cristales son extremadamente sensibles a la presencia de dihidrazida en el medio de reacción y se destruyen. 73 Resultados y discusión 4.4 Propiedades de los cristales entrecruzados de cloroperoxidasa Las fuerzas involucradas en el empaquetamiento cristalino de las macromoléculas son débiles en comparación con las que mantienen la cohesión en cristales de moléculas pequeñas. Las fuerzas involucradas son puentes salinos, puentes de hidrógeno, van der Waals, dipoio-dtpolo e hidrofóbicas. La débil cohesión entre las macromoléculas dentro del cristal se debe a que sólo una pequeña parte de la superficie participa en contactos intermoleculares, mientras que el resto permanece en contacto con el solvente. El estado cristalino mejora notablemente la estabilidad de la cloroperoxidasa en presencia de un solvente orgánico, como se muestra en la Tabla 4.4.1 para los cristales sin entrecruzar [167] (ver Apéndice 7.6). La inactivación por desnaturalización es más rápida para la enzima libre que para la enzima en estado cristalino, donde las moléculas están ordenadas y estabilizadas por las interacciones que existen entre ellas. La estabilidad que confiere el estado cristalino ha sido observada también para otras proteínas [168], Esta estabilidad se debe a que dentro del cristal las moléculas de proteína conservan mejor su estructura secundaria y tienen menos tendencia a agregarse. Tabla 4.4.1 Actividad residual de la cloroperoxidasa después de una hora de incubación en 99% t-butanol Temperatura (°C) 30 40 50 (% Enzima actividad 31 28 7 libre residual) Enzima cristalina sin entrecruzar (% actividad residual) 50 51 38 La mayor estabilidad del estado cristalino se ve reflejada también en una mayor resistencia a la exposición a temperaturas altas, como se muestra en la Figura 4.4.1 para los cristales entrecruzados. La cloroperoxidasa libre se inactiva rápidamente a temperaturas mayores que 50°C [92]. La actividad de los cristales entrecruzados se mantiene prácticamente sin cambio aún después de una hora de incubación a 70°C. La termoestabilidad de los cristales entrecruzados es consecuencia tanto del arreglo ordenado de las moléculas como de la rigidez de la estructura tridimensional que 74 Resultados y discusión ocasiona el entrecruzamiento con glutaraldehído. Por otro lado, en solución y en ausencia de estos contactos estabilizadores la conformación nativa de tas moléculas de proteína puede destruirse más fácilmente. 120 100 80 en ñ 60 "O ? 40 20 50 60 70 Temperatura (°C) 80 Figura 4.4.1 Actividad residual de cristales entrecruzados de cloroperoxidasa (D) y enzima libre (•) después de una hora de incubación a diferentes temperaturas en solución acuosa El estado físico de la proteína tiene un efecto negativo sobre la actividad. La Tabla 4.4.2 muestra la actividad de la cloroperoxidasa en dimetilformamida con un contenido de agua de 1%. El cambio de fase reduce a un tercio la actividad. Adicionalmente, la actividad observada para los cristales entrecruzados es todavía un orden de magnitud menor. Esta reducción ha sido documentada para otras enzimas y la actividad enzimática de los cristales puede ser entre uno y tres órdenes de magnitud menor que la actividad de la enzima en solución [24,53,55,56,57,169,170]. Tabla 4.4.2 Actividad enzimática para la oxidación del pinacianol Velocidad inicial (min -1) Cloroperoxidasa soluble Cloroperoxidasa cristalina Cristales entrecruzados 11 3.15 0.34 TESIS COH FALLA DE ORIGEN 75 Resultados y discusión Los factores más importantes que influyen sobre la actividad de los cristales son el tamaño del cristal, el tamaño del sustrato y la conformación de la enzima dentro del cristal [62], Es poco probable que la baja actividad enzimática de los cristales entrecruzados de cloroperoxidasa se deba a una conformación incorrecta de la enzima dentro del cristal, debido a que esta enzima no sufre cambios conformacionales importantes durante la catálisis. El tamaño de los cristales utilizados en este trabajo está dentro del intervalo adecuado para evitar problemas de difusión [24,62,169,170]. Para desechar la presencia de problemas difusionales en los cristales, se determinó la velocidad inicial para la oxidación del guaiacol a una [S]«KM . En estas condiciones, una relación lineal entre la velocidad y ta cantidad de enzima indica la ausencia de problemas de difusión al interior del cristal. En presencia de problemas difusionales, esta relación no es lineal y la velocidad inicial correlaciona con la raíz cuadrada de la cantidad de enzima. En la Figura 4.4.2 se presentan las dos correlaciones. La relación entre la velocidad inicial y la concentración de enzima es lineal (panel A), mientras que la relación entre la velocidad y la raíz cuadrada de la concentración de enzima no lo es (panel B). En cristales de este tamaño aparentemente la difusión de los sustratos no es una limitante para la actividad. 1000 2000 Cioroperoxidasa (nM) 3000 -i 10 0 20 40 Cloroperoxidasa (nM)1/2 60 Figura 4.4.2 Relación entre la velocidad inicial para la oxidación del guaiacol y la concentración de cloroperoxidasa La Tabla 4.4.3 contiene las constantes cinéticas de la cloroperoxidasa libre y de los cristales entrecruzados para la oxidación de dos sustratos azufrados y para la dismutación de H2O2. La reducción en la actividad de los cristales se debe principalmente a una disminución en la constante catalítica, pues la constante de 76 Resultados y discusión afinidad de los sustratos es del mismo orden de magnitud tanto para la enzima Ubre como para los cristales entrecruzados. Tabla 4.4.3 Constantes cinéticas de la cloroperoxidasa para la oxidación del tiantreno y tioanisot, y para la dismutación de H2O2 Sustrato Enzima libre Cristales entrecruzados Tiantreno Tioanisol H2O2 Kcat (s-1) 555 4170 1670 KM (mM) 0.0033 1.75 10 kcat/KM (s-1 NT1) 1.6x108 2.3x106 1.6x105 kcat (s-1) 0.85 28 55 KM (mM) 0.0073 2.2 4.6 kcat/KM (s'1 M"1) 1.1x106 1.2x104 1.2x104 En los tres casos la actividad de los cristales entrecruzados es menor que la actividad de la enzima libre. Sin embargo, la actividad es menor para el sustrato más grande y a medida que el tamaño del sustrato disminuye, la actividad aumenta (Figura 4.4.3). La eficiencia catalítica de los cristales entrecruzados es tres, dos y un orden de magnitud menor que la de la enzima libre para el tiantreno, tioanisol y H2O2, respectivamente. Estos resultados hacen suponer que los sustratos voluminosos tienen un acceso limitado a los sitios activos dentro del cristal. La accesibilidad de los sustratos está limitada por el tamaño de los poros dentro del cristal. Este tamaño varía con la proteína y con el tipo de arreglo espacial, por lo que no es una variable fácil de controlar. Figura 4.4.3 Estructura química del tiantreno, tioanisol y H2O2 Por otro lado, el perfil de pH y de temperatura de los cristales entrecruzados se muestran en la Figura 4.4.4. El pH óptimo es similar al de la cloroperoxidasa libre. Por otra parte, la temperatura óptima para la actividad de los cristales entrecruzados es ligeramente mayor. Esta diferencia probablemente refleja la estabilidad de la 77 Resultados y discusión estructura tridimensional de las moléculas dentro del cristal que retarda el proceso de desnaturalización térmica. 140 120 - 20 40 60 80 100 Temperatura ( C) Figura 4.4.4 Perfil de temperatura y de pH para los cristales entecruzados (•) y la cloroperoxidasa libre (•) La Tabla 4.4.4 muestra que los cristales de cloroperoxidasa son más activos en solvente orgánico cuando la actividad de agua aumenta. Este es un factor importante para la catálisis en solventes anhidros, pues se sabe que la actividad enzimática se ve afectada por el contenido de agua presente en el medio. Tabla 4.4.4 Velocidad inicial para la oxidación del pinacianol en dimetilformamida % H2O 1 2 5 10 aw 0.0469 0.0911 0.2089 0.3635 Cristales entrecruzados (min"1) 0.18 1.16 13 26 Cloroperoxidasa libre (min"1) 11 96 123 157 Los cristales también se probaron en presencia de diferentes solventes orgánicos para determinar su capacidad de oxidar compuestos organoazufrados. Se hicieron pruebas en presencia de 15% tetrahidrofurano utilizando tiantreno y diesel primario como sustratos. En ambos casos los cristales entrecruzados son capaces 78 Resultados y discusión de oxidar a los compuestos azufrados, al igual que la enzima libre. Sin embargo, al ensayar esta reacción de oxidación en medios más hidrofóbicos (sistema ternario tolueno-isopropanol-agua o solventes anhidros o con bajos porcentajes de agua) se encontró que no hay actividad. A pesar de que la actividad de agua se ajustó a niveles similares en los diferentes solventes, al parecer existen otros factores que afectan la actividad catalítica de los cristales en solventes con bajo contenido de agua. Por un lado, se sabe que la partición de un sustrato hidrofóbico entre el solvente y el sitio activo de una enzima se ve desfavorecida en medios no polares [171,172]. Al igual que otras peroxidasas, en el sitio activo de la cloroperoxidasa existen residuos cargados que forman una cavidad polar alrededor del hemo. De hecho, se ha observado que las constantes de afinidad entre la cloroperoxidasa libre y algunos sustratos aromáticos son más grandes a medida que aumenta el contenido de solvente orgánico en el medio de reacción [173], Adicionalmente, es razonable suponer que también existe un fenómeno de partición desfavorable del sustrato hacia el interior de los cristales entrecruzados, debido a que el microambiente dentro de los mismos exhibe una alta concentración de residuos polares y cargados que forman parte de las moléculas de proteína. Por otro lado, la estructura de las moléculas probablemente es más rígida en los cristales que en solución reduciendo la flexibilidad de la enzima en solventes orgánicos con bajo contenido de agua. En resumen, a pesar de que su actividad se ve afectada por el entrecruzamiento los cristales entrecruzados de cloroperoxidasa son una preparación superior que la enzima libre en términos de estabilidad. Otras propiedades de los cristales como el perfil de actividad a diferente pH o las constantes de afinidad por algunos sustratos son muy similares a los de la enzima libre. FSTA TESIS NO H'/:o 79 5. CONCLUSIONES Y PERSPECTIVAS 80 Conclusiones y perspectivas En este trabajo se demostró que la cloroperoxidasa es una enzima con aplicación potencial en la industria petrolera, específicamente en la biodesulfurización de combustibles derivados del petróleo. La cloroperoxidasa reconoce a un gran número de sustratos azufrados y cataliza eficientemente la oxidación de los mismos, lo que facilita su remoción. La industria petrolera requiere un biocatalizador estable y activo a altas temperaturas y en presencia de solventes orgánicos. Por esta razón, este proyecto planteó el objetivo general de obtener cristales entrecruzados de cloroperoxidasa, con el fin de desarrollar un biocatalizador con estas características. Se demostró que es posible obtener cristales catalíticos de cloroperoxidasa, los cuales fueron entrecruzados y caracterizados cinéticamente. Los resultados encontrados muestran que la principal ventaja de los cristales sobre la enzima libre es su mayor estabilidad al ser expuestos a solventes orgánicos y altas temperaturas. Algunas propiedades intrínsecas de la cloroperoxidasa obstaculizaron la obtención de los cristales entrecruzados. Por ejemplo, la inestabilidad de la enzima en condiciones alcalinas, el bajo número de residuos reactivos y la concentración de los mismos en una zona de la superficie de la proteína. Estos problemas fueron parcialmente resueltos al incrementar el número de aminos primarios en las moléculas dentro del cristal, de tal manera que disminuyó la cantidad de entrecruzante necesaria para mantener estable a los cristales. Sin embargo, la actividad específica observada de los cristales entrecruzados es menor que la actividad de la enzima libre. Una de las razones que puede explicar este comportamiento es la modificación con glutaraldehído de algunos residuos cercanos al sitio activo. Posiblemente estas modificaciones afectan la movilidad de la proteína o la funcionalidad del sitio activo. Una estrategia para determinar el número efectivo de sitios activos en un catalizador es la titulación de los mismos. Sin embargo para las peroxidasas no existe un procedimiento similar. A pesar de esto, los resultados obtenidos en este trabajo demuestran que la menor actividad de los cristales entrecruzados se debe en parte a problemas de accesibilidad de sustratos voluminosos a! interior del cristal. A pesar de que los cristales entrecruzados de cloroperoxidasa son más estables que la enzima libre en presencia de solventes orgánicos y también resisten altas temperaturas, no cumplen con la propiedad de catalizar la transformación de sustratos hidrofóbicos bajo estas condiciones. Probablemente en un medio no polar la partición de un sustrato hidrofóbico hacia un sitio activo polar se ve desfavorecida y por consiguiente la velocidad de la reacción disminuye. Por otra parte, es probable que el 81 Conclusiones y perspectivas tamaño de los poros internos de! cristal impida el acceso de los sustratos de interés. Estas limitaciones podrían superarse aplicando estrategias de ingeniería de solventes e ingeniería de proteínas para controlar la partición desfavorable del sustrato; y analizando el tamaño de los poros internos de cristales obtenidos bajo diferentes condiciones de cristalización, lo cual en ocasiones permite producir diferentes arreglos tridimensionales. Otras estrategias de entrecruza miento requieren un mayor conocimiento sobre el arreglo tridimensional de las moléculas dentro del cristal. La difracción de rayos X de estos cristales de cloroperoxidasa puede proveer información sobre la orientación de la proteína en el cristal y la localización de los contactos intermoleculares. Con estos datos sería posible diseñar un entrecruzante bifuncional del tamaño adecuado para formar enlaces covalentes intermoleculares lejos del sitio activo. Por ejemplo, la cloroperoxidasa tiene más de 20 grupos carboxilos distribuidos homogéneamente en su superficie. Es posible que pueda diseñarse una diamina o una mezcla de diaminas con la longitud apropiada para entrecruzar a las moléculas a través de los grupos carboxilos, evitando el bloqueo de los sitios activos o de los poros a través de los cuales difunde el sustrato. Los resultados de este trabajo sugieren que los cristales entrecruzados de cloroperoxidasa son un biocatalizador potencial para procesos que involucren altas temperaturas, sustratos pequeños de mediana hidrofobicidad y medios de reacción con bajo contenido de solventes orgánicos miscibles en agua. 82 6. BIBLIOGRAFÍA 83 Bibliografía 1 A. Líese y M.V. Filho, Production of fine chemicals using biocatalysis. Curr.Opin.Biotechnol., 70(1999) 595-603. 2 J.D. Rozzell, Commercial scale biocatalysis: myths and realities. Bioorg.Med.Chem., 7 (1999) 2253-2261. 3 W.Gerhartz (ed.), Enzymes in industry: production and applications, VCH, Germany, 1990. 4 A.L. Demain, Small bugs, big business: the economic power of the microbe. Biotechnol. Adv., 18 (2000) 499-514. 5 B. Schulze y M.G. Wubbolts, Biocatalysis for industrial production of fine chemicals. Curr.Opin.Biotechnol., 70(1999)609-615. 6 S.C. Stinson, Chiral Chemistry. Chem.Eng.News, May 14 (2001) 45-56. 7 R.N. Patel, Biocatalytic synthesis of intermediates for the synthesis of chiral drug substances. Curr.Opin.Biotechnol., 72(2001)587-604. 8 K.E. Jaeger, T. Eggert, A. Eipper y M.T. Reetz, Directed evolution and the creation of enantioselective biocatalysts. Appl.Microbio!.Biotechnol., 55 (2001) 519-530. 9 R. Jaenicke, Stability and stabilization of globular proteins in solution. J.Biotechnol., 79(2000) 193-203. 10 J.S. Dordick, Y.L. Khmelnitsky y M.V. Sergeeva, The evolution of biotransformation technologies. Curr.Opin.Microbiol., 7(1998)311-318. 11 R.E. Feeney, Chemical modification of proteins: comments and perspectives. IntJ.Peptide Protein Res., 29 (1987) 145-161. 12 S.S. Wong y L.C. Wong, Chemical crosslinking and the stabilization of proteins and enzymes. Enzyme Microb.Technol., 74(1992)866-874. 13 D. Kazan, H. Ertan y A. Erarslan, Stabilization of Escherichia coli penicillin G acylase against thermal inactivation by cross-linking with dextran dialdehyde polymers. Appl.Microbiol.Biotechnol., 48(1997) 191-197. 14 L. Gao, F. van Rantwijk y R.A. Sheldon, Cross-linked enzyme aggregates: a simple and effective method for the immobilization of penicillin acylase. Org.Lett., 2 (2000)1361-1364. 15 R. Vazquez-Duhalt, P.M. Fedorak y D.W.S. Westlake, Role of enzyme hydrophobicity in biocatalysis in organic solvents. Enzyme Microb.Technol., 14 (1992)837-841. 16 R. Vankatesh y P.V. Sundaram, Modulation of stability properties of bovine trypsin after in vitro structural changes with variety of chemical modifiers. Protein Engng., 77(1998)691-698. 84 Bibliografía 17 R. Tinoco y R. Vazquez-Duhalt, Chemical modification of cytochrome c improves their catalytic properties in oxidation of polycyclic aromatic hydrocarbons. Enzyme Microb.Technol., 22 (1998) 8-12. 18 G. DeSantis y J.B. Jones, Chemical modification of enzymes for enhanced functionality. Curr.Opin.Biotechnol., 10 (1999) 324-330. 19 U.T. Bornscheuery M. Pohl, Improved bioctalysts by directed evolution and rational protein design. Curr.Opin.Chem.Biol., 5 (2001) 137-143. 20 J.L. Harris y C.S. Craig, Engineering enzyme specificity. Curr.Opin.Chem.Biol., 2 (1998)127-132. 21 F.H. Arnold, P.L. Wintrode, K. Miyazaki y A. Gershenson, How enzymes adapt: lessons from directed evolution. Trenos Biochem.Sci., 26 (2001) 100-106. 22 F.H. Arnold y A.A. Volkov, Directed evolution of biocatalysis. Curr.Opin.Chem.Biol., 3 (1999) 54-59. 23 M.F. Chaplin y C. Bucke, Enzyme Technology, University of Cambridge, Cambridge, 1990. 24 W. Tischer y V. Kasche, Immobilized enzymes: crystals or carriers? Trends Biotechnoí., 77(1999)326-335. 25 S.S. Wong, Chemistry of protein conjugation and cross-linking, CRC Press, Boca Ratón, Florida, 1993. 26 P. Wang, S. Dai, S.D. Waezsada, A.Y. Tsao y B.H. Davison, Enzyme stabilization by covalent binding in nanoporous sol-gel glass for nonaqueous biocatalysis. Biotechnol.Bioeng., 74 (2001) 249-255. 27 S.J. Novick and J.S. Dordick, Investigating the effects of polymer chemistry on activity of biocatalytic plástic materials. Biotechnol.Bioeng., 68 (2000) 665-671. 28 A.M. Azevedo, D.M.F. Prazeres, J.M.S. Cabral y L.P. Fonseca, Stability of free and immobilized peroxidase in aqueous-organic solvents mixtures. J.Mol.Cat.B Enzymatic, í5 (2001) 147-153. 29 L. Dai y A.M. Klibanov, Striking activation of oxidative enzymes suspended in nonaqueous media. Proc.Natl.Acad.Sci.USA, 96 (1999) 9475-9478. 30 K. Griebenow, M. Vidal, C. Vaez, A.M. Santos y G. Barletta, Nativelike enzyme properties are important for optimum activity in neat organic solvents. J.Am.Chem.Soc, 723(2001) 5380-5381. 31 M. Tuena de Gómez-Poyou y A. Gómez-Poyou, Enzymes in low water systems. Crit.Rev.Biochem.Mol.Biol., 33(1998) 53-89. 32 S.J. Novick y J.S. Dordick, Protein-containing hydrophobic coatings and films. Biomateríals, 23(2002)441-448. 85 Bibliografía 33 M.D. Lilly, Biotransformations using immobilized biocatalysts: past, present and futura. Adv.Mol.CellBiol., 15A (1996) 141-149. 34 G. Bell, P.J. Halling, B.D. Moore, J. Partridge y D.G. Ress, Biocatalyst behaviour in low-water systems. Trends Biotechnoi, 73(1995)468-473. 35 P.J. Halling, Biocatalysis in low-water media:understanding effects of reaction conditions. Curr.Opin.Chem.Biol., 4 (2000) 74-80. 36 P.A. Fitzpatrick, D. Ringe y A.M. Klibanov, X-ray crystal structure of cross-linked subtilisin Carlsberg in water vs. acetonitrile. Biochem.Biophys.Res.Commun., 198 (1994)675-681. 37 Z. Wang, G. Zhu, Q. Huang, M. Qian, M. Shao, Y. Jia y Y. Tang, X-ray studies on cross-linked lysozyme crystals in acetonitrile-water mixture. Biochim.Biophys.Acta, 7384(1998)335-344. 38 A.M. Klibanov, Why are enzymes less active in organic solvents than in water? Trends Biotechnoi., 15 (1997) 97-101. 39 A.M. Klibanov, Improving enzymes by using them in organic solvents. Nature, 409 (2001)241-246. 40 J.A. Rupley y G. Careri, Protein hydration and function. Adv.Protein Chem., 41 (1991)37-172. 41 G. Carrea y S. Riva, Properties and synthetic applications of enzymes in organic solvents. Angew.Chem.Int.Ed., 39(2000) 2226-2254. 42 Y.L Khmelnitsky y J.O. Rich, Biocatalysis in nonaqueous solvents. Curr.Opin.Chem.BioL, 3 (1999) 47-53. 43 M.T. Ru, K.C. Wu, J.P. Lindsay, J.S. Dordick, J.A. Reimer y D.S. Clark, Towards more active biocatalysts in organic media: increasing the activity of salt activated enzymes. Biotechnoi.Bioeng., 75(2001) 187-196. 44 Y. Guo y D.S. Clark, Activation of enzymes for nonaqueous biocatalysis by denaturing concentrations of urea. Biochim.Biophys.Acta, 7546(2001)406-411. 45 D. van Unen, J.F.J. Engsbersen y D.N. Reinhoudt, Why do crown ethers actívate enzymes in organic solvents? Biotechnoi.Bioeng., 77(2002) 248-255. 46 S. Shubhada y P.V. Sundaram, The role of pH change caused by the addition of water-miscible organic solvents in the destabilization of an enzyme. Enzyme Microb.Technol., 77(1995)330-335. 47 E. Zacharis, B.D. Morre y P.J. Halüng, Control of enzyme activity in organic media by solid-state acid-base buffers. J.Am.Chem.Soc, 779(1997) 12396-12397. 86 Bibliografía 48 E, Zacharis, P.J. Halling y D.G. Rees, Volatile buffers can override the pH memory of subtilisin catalysis in organic media. Proc.Natl.Acad.ScLUSA, 96 (1999) 1201- 1205. 49 W.H. Bishop and F.M. Richards, Properties of liquids in small pores. J.Moi.Biol., 38(1968)315-328. 50 B.W. Matthews, Solvent content of protein crystals. J.Moi.Biol., 33 (1968) 491-497. 51 L.Z. Vilenchik, J.P. Griffith, N.L. St.Clair, M A Navia y A.L. Margolin, Protein crystals as novel microporous materials. J.Am.Chem.Soc, 120 (1998) 4290-4294. 52 F.A. Quiocho y F.M. Richards, Intermolecular cross linking of a protein in the crystalline state: Carboxipeptidase A. Proc.Natl.Acad.Sci.USA, 52 (1964) 833-839. 53 F.A. Quiocho y F.M. Richards, The enzymic behavior of carboxipeptidase A in the solid state. Biochemistry, 5 (1966) 4062-4076. 54 F.A. Quiocho, W.H. Bishop y F.M. Richards, Effects of changes in some solvent parameters of carboxipeptidase A in solution and in cross-ünked crystals. Proc.Natl.Acad.Sci.USA, 57(1967) 525-537. 55 P.J. Kavinsky y N.B. Madsen, Activity of glycogen phosphorylase in the crystalline state. J.Biol.Chem., 251 (1976) 6852-6859. 56 G.M. Alter, D.L. Leussing, H. Neurath y B.L. Valiee, Kinetic properties of carboxipeptidase B in solution and crystals. Biochemistry, 16 (1977) 3663-3668. 57 C.A. Spilburg, J.L. Bethune y B.L. Valiee, Kinetic properties of crystalline enzymes. Carboxipeptidase A. Biochemistry, 76(1977)1142-1150. 58 A. Yonath, A. Sielecki, J. Moult, A. Podjarny y W. Traub, Crystallographic studies of protein denaturation and renaturation. Biochemistry, 16 (1977) 1413-1417. 59 M.A. Navia, N.L. St.Clair y J.P. Griffith, Crosslinked enzyme crystals as immobilized enzyme particles. En W.J.J. van den Tweel, A. Harder y R.M. Buitelaar (eds.), Stability and Stabilization of enzymes., Elseveir Science Publishers B.V., Maastricht, The Netherlands, 1992, pp. 63-73. 60 M.A. Navia y N.L. St.Clair, Crosslinked enzyme crystals. U.S.Patent, 5,618,710 (1997) 61 Altus Biologics Inc. (En línea), (http://www.altus.com). Ultimo acceso: Mayo 2002. 62 A.L. Margolin, Novel crystalline catalysts. Trends Biotechnol., 14 (1996) 223-230. 63 T.S. Lee, J.D. Vaghjiani, G.J. Lye y M.K. Turner, A systematic approach to the large-scale production of protein crystals. Enzyme Microb.Technol., 26 (2000) 582-592. 64 N.L. StClair y M.A. Navia, Cross-linked enzyme crystals as robust biocatalysts. J.Am.Chem.Soc., 114 (1992) 7314-7316. 87 Bibliografía 65 R.A. Persichetti, N.L. St.Clair, J.P. Griffith, M.A. Navia y A.L Margolin, Cross- linked enzyme crystals (CLEC) of thermolysin ¡n the synthesis of peptides. J.Am.Chem.Soc, 777(1995) 2732-2737. 66 J.J. Lalonde, C. Govardhan, N. Khalaf, A.G. Martínez, K. Visuri y A.L. Margolin, Cross-linked crystals of Candida rugosa lipase: highly efficient catalysts for the resolution of chiral esters. J.Am.Chem.Soc, 777(1995)6845-6852. 67 N. Khalaf, C. Govardhan, J.J. Lalonde, R.A. Persichetti, Y. Wang y A.L Margolin, Cross-linked enzyme crystals as highly active catalysts in organic solvents. J.Am.Chem.Soc, 118 (1996) 5494-5495. 68 Y. Wang, K. Yakovlevsky y A.L. Margolin, An efficient synthesis of chiral amino acid and peptide alkylamides via CLEC-subtilisin catalyzed coupling and in situ resolution. Tetrahedron Lett, 37 (1996) 5317-5320. 69 N.L. St.Clair, Y. Wang y A.L. Margolin, Cofactor-bound cross-linked enzyme crystals (CLEC) of alcohol dehydrogenase. Angew.Chem.Int.Ed., 39 (2000) 380- 383. 70 D. Costes, E. Wehtje y P. Adlercreutz, Cross-linked crystals of hydroxynitrile lyase as catalyst for the synthesis of optically active cyanohydrins. J.Mol.Cat.B Enzymatic, 11 (2001) 607-612. 71 T.S. Lee, M.K. Turner y G.J. Lye, Mechanical stability of immobilized biocatalysts (CLECs) in dilute agitated suspensions. Biotechnol.Prog., (2001) 72 P. Pepin y R. Lortie, Influence of water actívity on the enantioselective esterification of (R,S)-ibuprofen by crosslinked crystals of Candida antárctica lipase B in organic solvent media. Biotechnol.Bioeng., 75(2001)559-562. 73 A.M. Collins, C. Maslin y R.J. Davies, Scale-up of a chiral resolution using cross- linked enzyme crystals. Org.Proc.Res.Dev., 2 (1998) 400-406. 74 D. Haring y P. Schreier, Cross-linked enzyme crystals. Curr.Opin.Chem.Biol., 3 (1999)35-38. 75 F.C.G. Hoskin, J.E. Walker y R. Stote, Degradation of nerve gases by CLECS and cells: kinetics of heterogenous systems. Chem.BioUnteract, 119-120 (1999) 439- 444. 76 N.L. St.Clair, B. Shenoy, L.D. Jacob y A.L. Margolin, Cross-linked protein crystals for vaccine delivery. Proc.Nati.Acad.Sci.USA, 96 (1999) 9469-9474. 77 C. Govardhan, Crosslinking of enzymes for improved stabiiity and performance. Curr.Opin.Biotechnol., 70(1999) 331-335. 78 F. Secundo, S. Spadaro, G. Carrea y P.L.A. Overbeeke, Optimization of Pseudomonas cepacia lipase preparations for catalysis in organic solvents. Biotechnol.Bioeng., 62 (1999) 554-561. Bibliografía 79 K. Xu y A.M. Klibanov, pH control of the catalytic activity of cross-linked enzyme crystals in organic solvents. J.Am.Chem.Soc, 118 (1996) 9815-9819. 80 J. Partridge, G.A. Hutcheon, B.D. Morre y P.J. Halling, Exploiting hydration hysteresis for high activity of cross-linked subtilísin crystals in acetonitrile. J.Am.Chem.Soc, 778(1996) 12873-12877. 81 J.L. Schmitke, C.R. Wescotty A.M. Klibanov, The mechanistic dissection of the plunge in enzymatic activity upon transition from water to anhydrous solvents. J.Am.Chem.Soc, 778(1996) 3360-3365. 82 S.K. Chapman, S. Daff y A.W. Munro, Heme: the most versatile redox centre in biology? En Structure and Bonding, Vo. 88, Springer Verlag, Berlín Heidelberg, 1997, pp. 39-70. 83 K.G. Welinder, Superfamily of plant, fungal and bacterial peroxidases. Current Topics in Structural Biology, 2 (1992) 388-393. 84 F. van de Velde, F. van Rantwijk y R.A. Sheldon, Improving the catalytic performance of peroxidases in organic synthesis. Trends Biotechnol., 19 (2001) 73-80. 85 M.C.R. Franssen y H.C. van der Pías, Haloperoxidases: their properties and their use in organic synthesis. En S.L. Neidleman y A.l. Laskin (eds.), Advances in Applied Microbiology Vol. 37, Academic Press, Inc., 1992, 86 M.C.R. Franssen, Halogenation and oxidation reactions with haloperoxidases. Biocatal., 70(1994)87-111. 87 J. Littlechild, Haloperoxidases and their role in biotransformation reactions. Curr.Opin.Chem.Biol., 3 (1999) 28-34. 88 A.T. Smith y N.C. Veitch, Substrate binding and catalysis in heme peroxidases. Curr.Opin.Chem.Biol., 2 (1998) 269-278. 89 A. Conesa, P.J. Punty C.A.M.J.J. van den Hondel, Fungal peroxidases: molecular aspects and applications. J.Biotechnol., 93 (2002) 143-158. 90 M.A. Pickard, A defined growth médium for the production of chloroperoxidase by Caldariomyces fumago. Can.J.Microbiol., 27(1981) 1298-1305. 91 M.A. Pickard y A. Hashimoto, Isoenzymes of chloroperoxidase from Caldariomyces fumago. Can.J.Microbiol., 28(1982) 1382-1388. 92 M.A. Pickard y A. Hashimoto, Stability and carbohydrate composition of chloroperoxidase from Caldariomyces fumago in a fructose-salts médium. Can.J.Microbiol., 34 (1988) 998-1002. 93 M. Sundaramoorthy, J. Temer y T.L. Poulos, The crystal structure of chloroperoxidase: a heme peroxidase-cytochrome P450 functional hybrid. Structure, 3 (1995) 1367-1377. 89 Bibliografía 94 K.A. Beckwith, R. Clark y L.P. Hager, Vil. Studies on the biosynthesis of caldariomycin. J.Biol.Chem., 238 (1963) 3086-3090. 95 S.L. Neidleman y J. Geigert, Biohalogenation. Principies, basic roles and applications., Ellis Horwood Limited, Chichester, England, 1986. 96 J.A. Thomas, D.R. Morris y L.P. Hager, Chloroperoxidase Vil: classical peroxidatic, catalatic and halogenating forms of the enzyme. J.Biol.Chem., 245 (1970) 3129- 3134. 97 J. Everse, The structure of heme proteins Compound I and II: some misconceptions. Free Rad.Biol.Med, 24 (1998) 1338-1346. 98 K. Muraii y L.P. Hager, Utilization of peroxide and its relevance in oxygen insertion reactions catalyzed by chloroperoxidase. Biochim.Biophys.Acta, 1547 (2001) 408- 417. 99 H. Yamada, N. Itoh y Y. Izumi, Chloroperoxidase-catalyzed halogenation of trans- cinnamic acid and its derivatives. J.Biol.Chem., 260 (1985) 11962-11969. 100 P A Cohén y G.H.N. Towers, Chlorination of anthraquinones by lichen and fungal enzymes. Phytochemistry, 44 (1997) 271-274. 101 I. Malnar y C.J. Sih, Chloroperoxidase-catalyzed chlorination of didechloroaglucovancomycin and vancomycin. J.Mol.Cat.B Enzymatic, 10 (2000) 545-549. 102 H.B. Dunford, A.M. Lambeir, M.A. Kashem y M.A. Pickard, On the mechanism of chlorination by chloroperoxidase. Arch.Biochem.Biophys., 252 (1987) 292-302. 103 H.B. Dunford, Peroxidase-catalyzed halide ion oxidation. Redox Rep, 5 (2001) 169-171. 104 R.D. Libby, A.L. Shedd, A.K. Phipps, T.M. Beachy y S.M. Gerstberger, Defining the involvement of HOCI or Cl2 as enzyme-generated intermediates in chloroperoxidase-catalyzed reactions. J.Biol.Chem., 267(1992) 1769-1775. 105 D.R. Libby, T.M. Beachy y A.K. Phipps, Quantitating direct chlorine transfer from enzyme to substrate in chloroperoxidase-catalyzed reactions. J.Biol.Chem., 271 (1996)21820-21827. 106 H. Wagenknecht y W. Woggon, Identification of intermediates in the cataiytic cycle of chloroperoxidase. Chem.Biol., 4 (1997) 367-372. 107 M.P.J. van Deurzen, F. van Rantwijk y R.A. Sheldon, Selective oxidations catalyzed by peroxidases. Tetrahedron, 53 (1997) 13183-13220. 108 S. Colonna, N. Gaggero, C. Richelmi y P. Pasta, Recent biotechnological developments in the use of peroxidases. Trends Biotechnol., 17 (1999) 163-168. 90 Bibliografía 109 L.P. Hager, F.J. Lakner y A. Basavapathruni, Chiral synthons via chloroperoxidase catalysis. J.Mol.CatB Enzymatic, 5 (1998) 95-101. 110 F.J. Márquez-Rocha, Y. Pica-Granados, A.M. Sandoval-Villasana y R. Vazquez- Duhalt, Determination of genotoxicity using chloroperoxidase-mediated model of PAH-DNA adduct formation. Buli.Environ.Contam.Toxicol., 59 (1997) 788-795. 111 P.M. Fedorak, K.M. Semple, R. Vazquez-Duhalty D.W.S. Westlake, Chloroperoxidase-mediated modifications of petroporphyrins and asphaltenes. Enzyme Microb.Technol., 15 (1993) 429-437. 112 L. Banci, Structural properties of peroxidases. J.BiotechnoL, 53 (1997) 253-263. 113 J.H. Dawson y M. Sonó, Cytochrome P450 and chloroperoxidase: thiolate-ligated heme enzymes. Spectroscopic determination on their active site structures and mechanistic implications of thiolate ligation. Chem.Rev., 87 (1987) 1255-1276. 114 M. Yi, M. Mroczko, K.M. Manoj, X. Wang y L.P. Hager, Replacement of the proximal heme thiolate ligand in chloroperoxidase with a histidine residue. Proc.Natl.Acad.Sci.USA, 96(1999) 12412-12417. 115 A. Conesa, F. van de Velde, F. van Rantwijk, R.A. Sheldon, C.A.M.J.J. van den Hondel y P.J. Punt, Expression of the Caldariomyces fumago chloroperoxidase in Aspergillus nigeranó characterization of the recombinant enzyme. J.Biol.Chem., 276(2001)17635-17640. 116 M. Sundaramoorthy, J. Temer y T.L. Poulos, Stereochemistry of the chloroperoxidase active site: crystallographic and molecular-modeling studies. Chem.Bioi., 5 (1998) 461-473. 117 P.R. Ortiz de Montellano, Catalytic sites of hemoprotein peroxidases. Annu.Rev.Pharmacol.Toxicol., 32 (1992) 89-107. 118 S.L. Newmyery P.R. Ortiz de Montellano, Horseradish peroxidase His42-Ala, His42-Val and Phe41-Ala mutants. J.Biol.Chem., 270 (1995) 19430-19438. 119 V.M. Samokyszyn y P.R. Ortiz de Montellano, Topology of the chloroperoxidase active site: regiospecificity of heme modification by phenyihydrazine and sodium azide. Biochemistry, 30 (1991) 11646-11653. 120 M.P.J. van Deurzen, Seiective oxidations cataiyzed by chloroperoxidase, Delft Universityof Technology, Delft, 1996, PhDThesis. 121 B. Valderrama, M. Ayala y R. Vazquez-Duhalt, Suicide inactivation of peroxidases and the challenge of engineering more robust enzymes. Chem.Bioi., 9 (2002) 555- 565. 122 K. Seelbach, M.P.J. van Deurzen, F. van Rantwijk, R.A. Sheldon y U. Kragl, Improvement of the total turnover number and space-time yield for chloroperoxidase cataiyzed oxidation. Biotechnol.Bioeng., 55 (1997) 283-288. 91 Bibliografía 123 L Stephan, W. Christian, S. Eberhard, L Andreas, Verfahren zur enzymatischen oxidation van substraten mit H2O2. Patente alemana DE10054082 A, 16 de mayo, 2002. 124 F. van de Velde, N.D. Lourenco, M. Bakker, F. van Rantwijk y R.A. Sheldon, Improved operational stability of peroxidases by coimmobilization with glucose oxidase. Biotechnol.Bioeng., 69 (2000) 286-291. 125 F. van de Velde, M. Bakker, F. van Rantwijk, G.P. Rai, LP. Hagery R A Sheldon, Engineering chloroperoxidase for activity and stability. J.Mol.Cat.B Enzymatic, 11 (2001)765-769. 126 T.A. Kadima y M.A. Pickard, Immobilization of chloroperoxidase on aminopropyl- glass. Appl.Environ.Microbiol., 56(1990)3473-3477. 127 S. Aoun, C. Chebli y M. Baboulene, Noncovalent immobiüzation of chloroperoxidase onto tale: catalytic properties of a new biocatalyst. Enzyme Microb.Technot., 23(1998) 380-385. 128 M. Andersson, B.K. Samra, H. Holmberg y P. Adlercreutz, Use of celite- immobilised chloroperoxidase in predominantly organic media. Biocatal.Biotrans., 17 (1999) 293-303. 129 M. Bakker, F. van de Velde, F. van Rantwijk y R.A. Sheldon, Highly efficient immobilization of glycosylated enzymes into polyurethane foams. Biotechnoi.Bioeng., 70(2000) 342-348. 130 F. van de Velde, M. Bakker, F. van Rantwijk y R.A. Sheldon, Chloroperoxidase- catalyzed enantioselective oxidations in hydrophobic organic media. Biotechnol.Bioeng., 72 (2001) 523-529. 131 B.L. McFarland, Biodesulfurization. Curr.Opin.Microbiol., 2 (1999) 257-264. 132 D.J. Monticello, Biodesulfurization and the upgrading of petroleum distillates. Curr.Opin.Biotechnol., 11 (2000) 540-546. 133 K. Yazu, Y. Yamamoto, T. Furuya, K. Miki y K. Ukegawa, Oxidation of dibenzothiophenes in an organic biphasic system and its application to oxidative desulfurization of light oil. Biotechnol.Prog., (2002) 134 A.M. Aitani, M.F. Aíi y H.H. At-Ali, A review of non-conventional methods for the desulfurization of residual fuel oil. Petroluem Science and Technology, 18 (2000) 537-553. 135 K.A. Gray, O.S. Pogrebinsky, GT. Mrachko, L. Xi, D.J. Monticello y C.H. Squires, Molecular mechanisms of biocatalytic desuifurization of fossil fuels. Nature Biotechnol., 14 (1996) 1705-1709. 136 MJ. Grossman, M.K. Lee, R.C. Prince, K.K. Garrett, G.N. George y I.J. Pickering, Microbial desulfurization of a crude oil middle-distillate fraction: analysis of the 92 Bibliografía extent of sulfur removal on remaining sulfur. Appl.Environ.MicrobioL, 65 (1999) 181-188. 137 L. Linguist y M. Pachecho, Enzyme-based diesel desulfurization process offers energy, CO2 advantages. OH and Gas Journal, February 22 (1999) 46-48. 138 Y. Izumi y T. Ohshiro, Purification and characterization of enzymes involved in desulfurization of dibenzothiophene in fossil fuels. J.Mol.Cat.B Enzymatic, 11 (2001)1061-1064. 139 L. Casella, M. Gullotti, R. Ghezzi, S. Poli, T. Beringhelli, S. Colonna y G. Carrea, Mechanism of enantioselective oxygenation of sulfides catalyzed by chloroperoxidase and horseradish peroxidase. Spectral studies and characterization of enzyme-substrate complexes. Biochemistry, 31 (1992) 9451- 9459. 140 S. Colonna, N. Gaggero, L Casella, G. Carrea y P. Pasta, Chloroperoxidase and hydrogen peroxide: an efficient system for enzymatic enantioselective sulfoxidations. Tetrahedron: Asymm., 3 (1992) 95-106. 141 P. Pasta, G. Carrea, E. Monzani, N. Gaggero y S. Colonna, Chloroperoxidase- catalyzed enantioselective oxidation of methyl phenyl suifide with dihydroxyfumaric acid/oxygen or ascorbic acid/oxygen as oxidants. Biotechnol.Bioeng., 62 (1999) 489-493. 142 M.A. Pickard, T.A. Kadimay R.D. Carmichael, Chloroperoxidase, a peroxidase with potential. J.Ind.Microbio!., 7(1991)235-242. 143 D.R. Morris y L.P. Hager, Chlorperoxidase I: Isolation and properties of the crystalline glycoprotein. J.Biol.Chem., 241 (1966) 1763-1768. 144 R.D. Libby, N.S. Rotberg, J.T. Emerson, T.C. White, G.M. Yen, S.H. Friedman, N.S. Sun y R. Goldowski, The chloride-activated peroxidation of catechol as a mechanistic probé of chloroperoxidase reactions. J.Biol.Chem., 264 (1989) 15284- 15292. 145 P.F. Hollenberg y L.P. Hager, Purification of chloroperoxidase from Caldariomyces fumago. Methods Enzymol., 52 (1978) 521-529. 146 M.B.Arnao, M.Acosta, J.A. del Río, R. Varón y F. García-Cánovas, A kinetic study on the suicide ¡nactivation of peroxidase by hydrogen peroxide. Biochim. Biophys. Acta. 1041 (1990)43-47. 147 J.R. Benson y P.E. Haré, o-phthalaldehyde: fluorogenic detection of primary amines in the the picomole range. Comparison with fluorescamine and ninhydrin. Proc.Nati.Acad.Sci.USA, 72 (1975) 619-622. 148 O.H. Lowry, N.J. Rosebrough, A.L. Farry R.J. Randall, Protein measurement with the Foiin phenol reagent. J.Biol.Chem., 193 (1951) 265-275. 93 Bibliografía 149 R. Vazquez-Duhalt, D.W.S. Westlake y P.M. Fedorak, Lignin peroxidase oxidation of aromatic compounds in systems containing organic solvents. Appi.Environ.Microbio!.; 60 (1994) 459-466. 150 A. Tuynman, M.K.S. Vink, H.L Dekker, H.E. Schemakery R. Wever, The sulphoxidation of thianisole catalysed by lactoperoxidase and Coprinus cinereus peroxidase: evidence for an oxygen-rebound mechanism. Eur.J.Biochem., 258 (1998)906-913. 151 S. Kobayashi, M. Nakano, T. KimurayA.P. Schaap, On the mechanism of the peroxidase-catalyzed oxygen transfer reaction. Biochemistry, 26 (1987) 5019- 5022. 152 M. Ortiz-Leon, L. Velasco y R. Vazquez-Duhalt, Biocatalytic oxidation of polycyclic aromatic hydrocarbons by hemoglobin and hydrogen peroxide. Biochem.Biophys.Res.Commun., 215 (1995) 968-973. 153 R. Vazquez-Duhalt, Cytochrome c as a biocatalyst. J.Mol.Cat.B Enzymatic, 7 (1999)241-249. 154 M. Ayala, R. Tinoco, V. Hernández, P. Bremauntz y R. Vazquez-Duhalt, Biocatalytic oxidation of fuel as an alternative to biodesulfurization. Fuel Process. Technol., 57 (1998) 101-111. 155 Ion energetics data. Mallard, W. G. y Linstrom, P. J. NIST Chemistry WebBook, NIST Standard Reference Datábase Number69 (En línea). 1998. National Institute of Standars and Technology (http://webbook.nist.gov) Ultimo acceso: Mayo 2002. 156 B.W. Bogan y R.T. Lámar, One-electron oxidation in the degradation of creosote polycyclic aromatic hydrocarbons by Phanerochaete chrysosporium. Appl.Environ.Microbio!., 61 (1995) 2631-2635. 157 M. Ayala, N.R. Robledo, A. Lopez-Munguia y R. Vazquez-Duhalt, Substrate specificity and ionization potential in chioroperoxidase-catalyzed oxidation of diesel fuel. Environ.Sci.Technol., 34 (2000) 2804-2809. 158 B.W. Bogan, R.T. Lámar y K.E. Hamrnel, Fluorene oxidation in vivo by Phanerochaete chrysosporium and in vitro during manganese peroxidase- dependent lipid peroxidation. Appl.Environ.Microbiol., 62 (1996) 1788-1792. 159 R. Vazquez-Duhait, M. Ayala y F.J. Márquez-Rocha, Biocatalytic chlorination of aromatic hydrocarbons by chloroperoxidase of Caldaríomyces fumago. Phytochemistry, 58 (2001) 929-933. 160 M.J. Axley, P. Kenigsberg y L.P. Hager, Fructose induces and glucose represses chloroperoxidase mRNA levéis. J.Biol.Chem., 261 (1986) 15058-15061. 161 P.C. Weber, Overview of protein crystallization methods. Methods Enzymol., 276 (1997)13-22. 94 Bibliografía 162 A. Ducruix y R. Giegé (eds.), Crystallization of Nucleic Acids and proteins. A practical approach, Oxford University Press, 1992. 163 D.R. Walt y V.l. Agayn, The chemistry of enzyme and protein immobilization with glutaraldehyde. Trenos Anal.Chem., 73(1994) 425-430. 164 P. Monsan, Optimization of glutaraldehyde activation of a support for enzyme ¡mmobiíization. J.Mol.Cat, 3(1978)371-384. 165 F.M. Richards y J.R. Knowles, Glutaraldehyde as a protein cross-linking reagent. J.Mol.Biol., 37 (1968) 231-233. 171 A.M. Lambeiry H.B. Dunford, A kinetic and spectral study of the alkaline transitions of chloroperoxidase. Arch.Biochem.Biophys., 220(1983)549-556. 172 M. Ayala, E. Horjales, M.A. Pickard y R. Vázquez-Duhalt, Cross-linked crystals of chloroperoxidase. Biochem. Biophys. Res. Commun., 295(2002) 828-831. 168 B. Shenoy, Y. Wang, W. Shan y A.L. Margolin, Stability of crystalline proteins. Biotechnol.Bioeng., 73 (2001) 358-369. 169 M.S. Doscher y F.M. Richards, The activity of an enzyme in the crystalline state: Ribonuciease S. J.Biol.Chem., 238 (1963) 2399-2406. 170 M.W. Makinen y A.L. Fink, Reactivity and cryoenzymology of enzymes in the crystalline state. Ann.Rev.Biophys.Bioeng, 6 (1977) 301-343. 171 E. Torres, R. Tinoco y R. Vazquez-Duhalt, Solvent hydrophobicity predicts biocatalytic behaviour of lignin peroxidase and cytochrome c in aqueous solution of water-miscible organic solvents. J. Biotechnol, 49 (1996) 59-67. 172 E. Torres, B. Siminovich, E. Barzana y R. Vazquez-Duhalt, Thermodynamic hydrophobicity of aqueous mixtures of water-miscible organic solvents predicts peroxidase activity. J.Mol.Cat.B Enzymatic, 4 (1998) 155-159. 173 M.P.J. van Deurzen, I.J. Remkes, F. van Rantwijk y R.A. Sheldon, Chloroperoxidase catalyzed oxidations in íert-butyl alcohol/water mixtures. J.Mol.Cat.A Chemical, íf7(1997) 329-337. 95 7. APÉNDICES 96 Apéndice 7.1 FUEL PROCESSING TECHNOLOGY ELSEVIER Fuel Processing Technology 57 (1998) lOI-l 11 • Biocatalytic oxidation of fuel as an alternative to biodesulfurization Marcela Ayala \ Raunel Tinoco a, Verónica Hernández a, Pilar Bremauntz b, Rafael Vazquez-Duhalt *•* * Instituto de Biotecnología, UNAM. Apañado Postal 510-3, Cuernavaca, Afórelos, 62250 México b Instituto Mexicano del Petróleo. México Reccivcd 26 Novcmbcr 1997; revísed 23 July 1998; accepted 23 July 1998 Abstract A biotechnological method for fuel desulfurizatíon is described. The method includes the steps of biocatalytic oxidation of organosulfides and tbiophenes, contained in the fuel, with hemo- proteins to form sulfoxides and sulfones, followed by a disüllation step in which these oxidizcd compounds are removed from the fuel. Straíght-ron diesel fuel contaíning 1.6% sulfur was biocatalytically oxidized with chloroperoxídase from Catdariomyces jumaga in the presence of 0.25 mM hydrogen peroxide. The reaction was carried out at room temperature and the organosulfur compounds were effectively transfonned to their respective sulfoxides and sulfones which were then removed by distillation. The resulting fraction after distiliation contained only 0.27% sulfur. Biocatalytic oxidation of fuels appcars as an interesting altcmative to biodesulfuriza- tion. © 1998 Elsevier Science B.V. A]l rights reserved. Keywords: Biocatalytic oxidation; Fuel; Biodesulfurization 1. Introduction The use of fóssil fuels for power generation and in the pelrochemical industry is expected to increase in the first decades of the next century. The demand for low-sulfur fossil fuels has been intensífíed by the increasing regulatory standards for reduced levéis of sulfur-oxides in atmospheríc emissions, by the decline of easily accessible sources of convenüonal and light crude oils, and by the high cost of physicochemical processes of * CoiTesponding author. Tel.: + 52-5-6227600; fax: +52-73-172388; e-mail: vazqduh@ibt.unam.mx 0378-3820/98/$ - see front matter © 1998 Elsevier Science B.V. All rights reserved. PII: SO378-3820(98)00076-9 Apéndice 7.1 102 M. Ayaia et al./Fuel Processing Technology 57 (1998) 101-111 hydrodesulfurization (HDS). It can be estimated that in the next decades 30% of oil should be desulfurized. The use of microorganisms for the biodesulfurization of high sulfur coals and oil has been proposed as an interesting altemative for the reduction of the organosulfur content of fossil fuels [1-3]. The number of laboratories involved in biodesulfurization research is rising rapidly. Different strains of aerobic microorganisms have been reported to non-selectively remove sulfur by a naphthalene degradative pathway, such as in Pseudomonas sp. [4,5] and Arthrobacter sp. [6,7]. Selective sulfur removal has also been reported by a pathway involving the conver- sión of dibenzothiophene (DBT) to 2-hydroxybiphenyl (2-HBP) and sulfate, as in the case of Corynebacterium sp. [8], Rhodococcus erythropolis [9-12], and Rhodococcus sp. strain IGTS8 [3,13]. Rhodococcus strain IGTS8 has, doubtless, been the most extensively studíed and it is the basis of the commercial biodesulfurization program of Energy Biosystems [14]. In addition to DBT, this strain is able to metabolize other thiophenes, sulfides, thíanthrene, sulfoxides and sulfones [131 and also to carry out a selective desulfurization of sterically hindered analogs of DBT [7]. Identification, cloning, characterization and overexpression of the genes involved ín the specific desulfurization have been completed [15-19], The plasmid-encoded pathway includes three genes, sox ABC, arranged in an operan residing in a 4 kb-region. These genes are responsible for the transformation of DBT to 2-HBP and sulfate. The soxA, soxB, and soxC genes encode for proteins wíth predicted molecular masses of 49.5, 38.9, and 45.1 kDa, respectively. The oxidation of DBT to DBT sulfone has been linked to the enzyme encoded by the soxC gene. This enzyme has been recently characterized as sulfite/sulfoxide monooxygenase [18] and requires reduced flavin mononucleotide for activity. Most of the microbial biodesulfurization studies have focused on the aerobic conver- sión of DBT, coal or fuels. Nevertheless, reductive desulfurization of fossil fuels is an idea proposed more than 25 years ago by Denis-Larose et al. [19], Mixed cultures containing sulfate-reducíng bacteria (SRB) desulfurized a variety of model compounds, including thiophenes [20], organosulfides [20,21] and petroleum preparations [22]. Reductive desulfurization of DBT to form hydrogen sulfide and biphenyl has been achieved by several species of SRB that are able to grow using DBT as solé source of sulfur and solé electrón acceptor [23-26]. Hydrogen gas is the normal source of reducing equivalent, however, electrochemícally generated reducing equivalents can be tncorporated ínto the normal electrón transport system of SRB [23]. This dissimilatory anaerobic process for sulfur removal would accumulate no wasteful biomass ñor introduce oxygen into the sulfur containing molecule or potential fuel. Desulfovibrio desulfuricans M6 is the best organism found so far in the anaerobic desulfurization [27]. Microbial desulfurization of petroleum derivatives has two main problems: Microbial activity is carry out in aqueous phase, thus a two phase system reactor with the intrinsic mass transfer limitations would be needed to metabolize the hydrophobic substrate. On the other h'and, the microbial biocatalyst must have a broad substrate specificiry for the various organosulfur compounds present in oil. These problems could be addressed by using broad specifícity enzymes instead of whole microorganisms. Enzymes are able to perform catalytic reactions in organic Apéndice 7.1 M. Ayala et al./Fuel Processing Technology 57(1998) 101-11 ¡ 103 solvente [28], in which the mass transfer Iimitations are reduced. The solvent could be the fuel itself. Under anhydrous conditions or at very low water activity, enzymes are generally more thermostable, and reactions could be performed at temperatures higher than 100°C [29]. Biocatalytic modificación of complex mixtures fróm petroleum, such as asphaltenes, have been performed in organic solvents [30]. Several enzymes have the ability to oxidize thiophenes and órganosulfur compounds in vitro; cytochromes P450 [31-37], lignin peroxidase from the white rot fungus Phanerochaete chrysosporium [38,39], lactoperoxidase [40,41], chloroperoxidase from Caldariomyces fumago [37,38,41-44], and horseradish peroxidase [37,40-42]. Non enzymatic hemoproteins are also able to perform the DBT oxidation in vitro, such as hemoglobin [37,45,46], cytochrome c [45,47,48], and microperoxídase [49,50], AJÍ the proteíns mentioned above are hemoproteins, and in all cases the producís of the biocatalytic oxidations are the respective sulfoxides. In this work, the biocatalytic oxidation of órganosulfur compounds which are contaíned in a diesel fuel, followed by a distillation that removes the oxidized com- pounds is shown as a two-steps alternative process for fuel desulfurízation. The possibHity of usíng a biocatalytic process in non-aqueous systems for the desulfurízation of fuels ís discussed. 2. Experimental 2.1. Chemicals Purified chloroperoxidase from C. fumago was a gift from Dr. M.A. Pickard from the University of Alberta, Canadá. Horse heart cytochrome c was obtained from Sigma (St. Louis, MO). Poly(ethylene glycol)-modified cytochrome c (PEG-Cyt) was prepared as previously reported [48] by using activated PEG wíth cyanuric chloride. Hydrogen peroxide and buffer salts were purchased from J.T. Baker (Phillipsburg, NJ). Dibenzoth- iophene, thianthrene, phenyl sulfide, phenyl disulfíde, benzothiophene, ethyl phenyl sulfide, and bithiophene were obtained from Aldrich Chemicai (Milwaukee, WI), and HPLC-grade organic solvents from Fisher Scíentífic (Sprigfield, NJ). Desulfurized diesel fuel and primary diesel fuel were obtained from Petróleos Mexicanos, PEMEX. 2.2. Biocatalytic reactions with diesel Diesel fuel oxidations with chloroperoxidase were carried out in a 10-ml reaction mixture containing 2 jjug of diesel fuel and 20 mM KC1 in a 20% acetonitrile-60 mM acétate buffer pH 3.0. The reaction mixture contained 2.3 nmol of chloroperoxidase and the reaction was started by addíng 0.25 mM of hydrogen peroxide. Reactions with cytochrome c (8 nmol) were performed in 20% acetonitrile-60 mM phosphate buffer pH 6.1 ín the presence of 1 mM hydrogen peroxide. After a 1 h reaction, the mixture was acidificó with nitric acid to pH 2, and extractcd three times with 2 mi of methylene Apéndice 7.1 104 M. Ayala a al./Fuel Processing Technology 57 (199&) ¡01-111 chloride. The organic extract was reduced under vacuum, and analyzed by gas chro- matography. 2.3. Biocatalytic oxidations of organosulfur compounds Organosulfur compounds reactions with chloroperoxidase were performed in a 15% acetonitrile, 20 mM KC1, 60 mM acétate buffer pH 3.0, 1-ml reaction mixture, containing 20 |xM substrate and frotn 0.2 to 5 nM of enzyme. Reactions were started by adding 1 mM of hydrogen peroxide and the disappearance of the substrate was monitored by HPLC after 10 min reaction. For producís identiñcation, 10-ml reactions were carried out. After I h reaction, the mixture was then acidified, extracted with methylene chloride and the extract reduced under nitrogen, before being analyzed by GC-MS. 2.4. Kinetic and inactivation constants determination The kinetic constants for the. oxídation of thianthrene with chloroperoxidase were estimated in I-ml reaction mixture containing 15% acetonitrile and 20 mM KC1 in a 60 mM acétate buffer pH 3.0, and with 1.5 nM of enzyme. The reaction was monitored spectrophotometrically by following the decreasing absorbance of thianthrene at 254 nm (e = 35 mM"1 cm"')- The specific activity is defined as the number of mol of substrate transformed by 1 mol of enzyme per minute. The inactivation of chloroperoxidase by hydrogen peroxide was measured by incubating 5 pmo! of enzyme in a 1-ml mixture containing 20 mM KC1, 60 mM acétate buffer pH 3.0 and different concentrations of hydrogen peroxide, for different periods of time. The reaction was started by adding 0.1 mM of monochlorodimedon (MCD) and the reaction was monitored by the decreasing absorbance of the substrate at 278 nm (e= 12.2 mM' 1 cm"1). 2.5. Analytical procedures Gas chromatography was carried out in a Hewlett-Packard GC chromatograph (model 5890 plus) coupled to two detectors: fíame ionization (FID) and fíame photometric (FPD) detectors. The GC was equipped with a 30 m X 0.25 mm SPB-20 column (Supelco), and the temperature program started at 90*0 for 2 min, then raised to 300°C at 8°C/min, and held for 10 min. Product identifícation was performed in a Hewlett- Packard GC (model 6890>MS (model 5972), with a 30 m X 0.25 mm SPB-20 column (Supelco). Infrared (IR) analyses were performed in a Fourier Transform Spectropho- tometer (Beckman FT 220) at the Faculty of Chemistry of the Morelos State University. UV-Vis measuremcnts were made in a Beckman Spectrophotometer (DU 530). Samples were analyzed by HPLC using a Perkin Elmer (series 200) system, with a Hypersyl 5 fi.m (100 X 2.1 mm) Hewlett-Packard column, elutíng with an acetonitrile-water phase. Microdistillations were carried out according to the standard test for boiling range distribution of petroleum fractions by gas chromatography, ASTM D 2887-89. Organic sulfur determinatíons on diesel fuel were carried out by'X-ray fluorimetry in a Horiba X-ray fluorimeter. Apéndice 7.1 M. Ayala et al./Fuet Processing Technology 57 (1998) ¡01 -III 3. Resulte and discussion 105 Desulfurized diesel fue! ( < 0.05% of sulfur) was enriched with 10 g 1"' of DBT and treated with poly(ethylene)glycol-modified cytochrome c (PEG-Cyt) and hydrogen peroxide. The gas chromatogram shows (Fig. 1) that the DBT is transformed to DBT sutfoxide, while the hydrocárbons seem to be not affected. DBT sulfoxide is an unstable compound which may be oxidized to fonn DBT sulfone. Cytochrome c ís a biocatalyst able to oxidize thiophenes and organosulfides [47] and has severa! advantages when compared with other hemoenzymes. It is active in a pH range from 2 to 11, has the heme prosthetic group covalently bound, exhibíting activity at high concentrations of organic solvents, and is not expensive [47,48]. In addition, this biocatalyst can be modified by site-directed mutagenesis [51] and by chemical modifícation [52] to improve both its catalytíc activity and range of substrates. PEG-modified cnzymes are soluble in v> c o Q. V> O O-- FID 10 FPD 20 DBT sulfoxide — 0BT-- i 30 40 -—-DBT sulfone O 10 20 30 40 Time (m¡n) Fig. 1. Gas chromatograms of desulfurized diesel fticl, enriched with dibenzomiophene, aftcr biocatalytic treaüncnt with poly(ethylene)glycol-modiFied cytochrome c. FID, Fíame ionization detector (genera! detector). FPD, Fíame photometric detector (sulfur selectíve detector). /¿Y Apéndice 7.1 106 M. Ayala et al. / Fue! Processing Technology 57 (1998) 101-111 30 0 -- Bíocatalytic oxidation FIO J JkJ i 10 20 Biocataiytic oxidation FPD . J J 1 b 30 20 Time (min) 30 Fig. 2. Gas chromatograms of primary diesel fuel (a) before and (b). after biocatalylic treatment with chlwoperoxidase from Cnldariomyces fianago. FID, Fíame ionization detector (general detector). FPD. Fíame photometric detector (sulfur selective detector). Apéndice 7.1 M. Ayaía et al./Fuel Processing Technology 57 (1998) I0I-U1 107 Table 1 Specifíc activíty of the oxidaüon of puré organosulfur compounds with chloropcroxidase from Caldariomyces fiímago Organosulfur compound Specific acüvity (mín ' ) Ethyl phcnyl sulfide Thianthrenc Bilhiophene Phenyl sulfide Benzothiophene Phenyl disulfide Dibcnzothiophene 1725 (±145) 1310 (±132) S40(±8) 831 (±32) 557 (±42) 352 (±10) 126 (±9) Standard deviations, in parcnlhescs, were calculated from three independent replícales. organic solvents and their acüvity in organic solvents is increased because of thc reduction of mass transfer limitations in the system [53]. Straight-run diesel fuel, obtained from primary distillation and containing 1.6% sulfur, was tested for oxidatíon with PEG-Cyt. Using this authentic diesel fuel, the modífied cytochrome c was able to oxidize most of the organosulfur compounds it contained. The oxidaüon was detected by the increase of boiling point (retention time) of these compounds on the gas chromatogram monitored with a Fíame Photometric Detector (FPD), which ís a sulfur selecüve detector. However, PEG-Cyt was not able to oxidize all of the organosulfur compound because some of them remained unchanged in the FPD chromatogram. This could be due to the fact that some of the organosulfur compounds found in the straíght-nm diesel fuel could be sterically hindered and do not bind with the acüve site. This has been observed in the oxidation of model compounds by cytochrome c [Al]. With the aim of increasing the biocatalytic oxidaüon of sulfur compounds, chloroper- oxidase from the imperfect fungus C. fumago was tested on primary diesel fuel. Fig. 2 shows that most of the organosulfur compounds in the primary diesel fuel were significantly oxidized and a considerable increase of the boiling points of all the sulfur compounds was found. Chloroperoxidase has been shown to be more active than other Tablc 2 Mass spectral dala of producís from enzymaüc oxidation of some organosulfur compounds with chloroperoxi- dase Substrate Product Mass spectral ions (m/z) Phenyl sulfide Dibenzotbiopbene Thtanthrene Phenyl sulfonc Dibenzothiophene sulfone 5-thianthrene oxide 5,10-thtanthrene dioxide 218(27) [M+ i 125(100), 97(26). 77(53), 51(47), 50(16) 216(100) [M+ I 187(46), 160(31), 150(16), 139(30) 232(16) IM+ 1184(100), 171(15), 139(14), 69(14) 248(77) [M+ i 200(86), 184(84), 171(100), 168(23). 139(30), 108(24X69(36) Valúes in parentheses are relative abundantes. IM+J, molecular ion. Apéndice 7.1 108 M. Ayala et ai./Fttel Processing Technology 57 {1998) ¡01-1} 1 Table 3 Kinetic constants of chloroperoxidase in the oxidation of Ihianthrene with hydrogen peroxide* *„, (s~ ' ) Thianthrene Hydrogen peroxide 64 1.45 44.2 133 0.482 ' Kinetic constants for hydrogen peroxide with a thianthrene saturating concentration of 20 p.M, and for thianthrene with a hydrogen peroxide saturating concentration of 1 mM. peroxidases, including lignin peroxidase from P. chrysosporium, in the oxidation of polycyclic aromatíc hydrocarbons [54]. In order to know the kinetic properties of chloroperoxidase with organosulfur compounds and the chemical nature of produets, enzymatic oxidations were performed in media containing puré substratos, such as thiophenes and organosulfides. Table 1 shows the specific activity of chloroperoxidase in the oxidatíon of puré organosulfur compounds. The producís of some of these reactíons, identified by GC-MS, are Usted in Table 2. These organosulfur compounds are good substrates for chloroperoxidase, because they are easily oxidized to form sulfoxides. Kinetic constants for the oxidation of thianthrene by chloroperoxidase were determined (Table 3). The JtcaI for the oxidation reaction was 64 s~' and the KM for thianthrene was 90 X lower than for hydrogen peroxide. On the other hand, hemoproteins are inactivated by hydrogen peroxide; the inactiva- tton constants for chloroperoxidase determined from a first-order equation, in the presence of different concentrations of hydrogen peroxide are shown in Table 4. As in the cases of horseradish peroxidase [55], lignin peroxidase [56], manganese peroxidase [57], lactoperoxidase [58], and other peroxidases, chloroperoxidase is inactivated by the presence of an excess of hydrogen peroxide. This substrate inactivation leads to the modífication of the heme prosthetic group and probably, to the formation of a verdo- haemoprotein as a final product [57]. So far, the inactivation mechanism has not been clearly elucidated [55-60]. Experimentation is currently being performed in order to improve the enzyme stabiliry against hydrogen peroxide by chemical and genetic modifications of chloroperoxidase. Obviously, in a one-step enzymatic reaction the removal of sulfur compounds was not envísaged, however, the oxidation of these compounds to sulfoxides permits their removal by a single distíllatíon. Microdistillation of both treated and untreated diesel Table 4 Inactivation constants (km) of chloroperoxidase in the presence df different concentrations of hydrogen peroxide Hydrogen peroxide (mM) kin (min " ' ) -—5 0.203 ; ~ ~ 0.50 0.248 1.00 0.287 Apéndice 7.1 M. Ayala et al./Fuel Processing Technology 57 (1998) I0I-1J1 109 fuels monitored by Fíame Ionization ¡Detector, FID and by FPD (Fig. 3) shows that the hydrocarbon distiUation profile monitored by FID (general detection) changes slightly after the biocatalytic treatment. On the other hand, the specific sulfur detector (FPD) shows a significant change of the distillation profile. The IR spectrum of oxidized diesel fuel showed the presence of two strong absorbance bands at 1385 and 1464 cm~' indicating the presence of suífoxides and sulfones. Oxidized sulfur compounds can be removed by a distillation step in which the final distillation point is 50°C lower than the starting fraction. When primary diesel fuel containing 1.6% sulfur is distilled in order to obtain a 100% distillation at a temperature 50°C lower than the original fraction, it produces a diesel fuel containing 1.27% of 20 15 10 v> b 40 3 0 o ^ 20 OT b 10 FID Control CPO treated FPD CPO treated Control 100 200 300 Boiling point (°C) 4 0 0 500 Fig. 3. MícrodistiUation profiles of untreated and enzymatically treated primary diesel fuel. FID, Fíame ionization detector (general detector). FPD, Fíame photometric detector (sulfur selective delector). CPO, chloroperoxidase. Apéndice 7.1 110 M. Ayala et al/Fuei Processing Technology 57 (1998) 101-111 sulfur and 83% of the original hydrocarbons. The undistilled heavy fraction (17% of starting hydrocarbons) contained 1.94% sulfur. If this petroleum fraction is previously oxídized by chloroperoxidase and hydrogen peroxide, and dislilled at the same condi- tions, the distiliate shows a sulfur content of only 0.27%, and 71% of total hydrocar- bons. Thus, a bíocatalytíc treatment of primary diesel fuel with chloroperoxidase from C. fumago, followed by a distillation is abíe to reduce the sulfur content by 80%. This mass balance was determined by GC integrations with Fíame íonization and Fíame Photometric detectors and by sulfur content detenninations on a X-ray fluorimeter. This approach for mass balance has some limitations, but it is useful for comparíng similar fractions tested under the same conditions. In addition, microdistillation usíng Fíame íonization Detector is currently a standard method (ASTM D 2887-89) for mass balance. In conclusión, biocatalytic oxidation of organosulfur compounds can be performed in complex hydrocarbon mixtures. Biocatalytic oxidation of organosulfur compounds found in fuels to less volatile products, which can then be removed by distillation, could be considered as a biodesulfurization process. So far, the reactions were carried out in aqueous mixtures of diesel, however our final goal is to perform the biocatalytic oxidation of organosulfur compounds in the diesel itself as reaction solvent without addition of water or organic solvent. Genetic and chemical modifications on different biocatalysts are in progress in our laboratory, focused on biocatalysis in organic solvent systems. Acknowledgements This work was funded by the Instituto Mexicano de Petróleo Grant FIES 95-137-11. We thank: Dr. M.A. Pickard from the University of Alberta for the chloroperoxidase supply. The authors thank Rosa Román for her technical assistance. References Hl D.J. MonliceUo, W.R. Fínnerty, Annu. Rcv. Microbiol. 39 (1985) 371. (2) A. Bhadra. J.M. Scharer, M. Moo-Yong, Biotechnot. Adv. 5 (1987) 1. I3J J.J. Kilbane, K. Jackowstci. Biotechnol. Biocng. 40 (1992) 1107. [4] DJ. Monticello. D. Bakker, W.R. Finncrty, Appl. Environ. Microbiol. 49 (1985) 756. [5] L. Setti, G. Lanzarini. P.G. Pifferi, Proc. Biochem. 30 (1995) 721. 16] M.D. Dahlbcrg, R.L. Rohere. DJ. Fauttb, R. Sprecher. GJ. Olson, Fuel 72 (1992) 1645. [7] M.K. Lee, J.D. Senius. M J. Grossman, Appl. Environ. Microbiol- 61 (1995) 4362. [8] T. Omori, L. Monna, Y. Saiki, T. Kodama, Appl. Environ. Microbiol. 49 (1992) 911. [9] Y. Izunií, T. Oshiro, Y. Hiñe, M. Shimao, Appl. Environ. Microbiol. 60 (1994) 223. liOl T. Ohshiro. T. Hirata, Y. Izumi. Appl. Microbiol. Biolechnol. 44 (1995) 249. {II] P. Wang. A.E, Humphrey, S. Krawiec, Appl. Environ. Microbiol. 62 (1996) 3066. [12] P. Wang, S. Krawiec, Appl. Environ. Microbiol. 62 (1996) 1670. (13] K.J. Kayser, B.A. Bielaga-Jones. K. Jackowski, O. Odusan, J.J. Kilbane, J. Gen. Microbio!. 139 (1993) 3123. [14] Biotreatment News, March 1994. 3. [15] S.A. Denpme. E.S. Olson, K.D. Young, Appl. Environ. Microbio!. 59 (1993) 2837. [16] S.A. Denome, E.S.C. Oldfield, L.J. Nash, K.D; Young, J. Bacteriol. 176 (1994) 6707. [17] C.S. Piddington, B.R. Kovaccvich, í. Rambosefc, Appl. Environ. Microbiol. 61 (1995) 468. Apéndice 7.1 M. Ayala et al/Fuel Processing Technology 57 (¡998) 101-111 111 {18] B. Leí. S.C. Tu, i. Bacterio!. 178 (1996) 5699. (19] C. Denis-Larosc, D. Labbe, H. Bcrgcron, A.M. Jones, C.W. Greer. J. Al-Hawari, MJ. Grossman, B.M. Sankey, P.C.K. b u . Appl. Environ. Microbiol. 63 (1997) 2915. [20] M. Kohler, I.L. Genz. B. Schicht, V. Eckart, Zenlralbl. Mikrobiol. 139 (1984) 239. [21] K.W. MiUcr, Appl. Environ. Microbiol. 58 (1992) 2176. [22] V. Eckart, M. Kohler, W. Hieke, Zcntralbl. Mikrobiol. 141 (1986) 291. [23] T.S. Kim, H.Y. Kim, B.H. Kim. Biotechnol. LetL 12 (1990) 757. [24] H.Y. Kim, T.S. Kim, B.H. Kim, Biotechnol. Lett. 12 (1990) 761. [25] H.M. Lizama. L.A. Wilkins, T.C. Scott, Biotechnol. Lett. 17 (1995) 113. [26] S.M. Amstrong. B.M. Sankey, G. Voordouw, Biotechnol. Lett. 17 (1995) 1133. [27] B.H. Kim, H.Y. Kim, T.S. Kim, Fuel Process. Technol. 43 (1995) 87. [28] J.S. Dordick. Enzyme Microb. Technol. 11 (1989) Í94. [29] V.V. Mozbaev, K.G. Poltevsky, V.I. Slepnev, GA. Badun, A.V. Lcvashov, FEBS Lett. 292 (1991) 159. [30) P.M. Fedoralt, K.M. Semple, R. Vazquez-Duhalt, D.W.S. Westlakc, Enzyme Microb. Technol. 15 (1993) 429. [31] W. Nastainzcyk, H.H. Ruf. V. Ullrich, Eur. i, Biocheia 60 (1975) 615. [32] D. Fukushima, Y.H. Kim, T. Iyanagi, S. Oac, J. Biochem. 83 (1978) 1019. [33] Y. Wantabe, T. Iyanagi, S. Oae, Tetrahedron Lett 21 (1980) 3685. [34] Y. Wantabe, T. Iyanagi, O. Shigeru, Tetrahedron Lett. 23 (1982) 533. [35] t . Takata, M. Yamazaki. K. Fujimori, Y.H. Kim, T. Iyanagi, S. Oae, Bull. Chem. Soc. Jpn. 56 (1983) 2300. [36] D. Mansuy, P. Valadon, Y. Erdelmcicr, P. Lopez-Garcia, C. Amar, J.P. Girault. P.M. Dansettc, J. Ara. Chem. Soc. 113(1991)7825. [37] J.C. Alvarez, P.R. Ortiz de Montellano, Biochemistry 31 (1992) 8315. [38] R.P. Scheiner, S.E. Stevcns. M. Tien, Appl. Environ. Microbio!. 54 (1988) 1858. [39] R. Vazquez-Duhall, D.W.S. Westlake, P.M. Fedorak, App!. Environ. Microbiol. 60 (1994) 459. [40] D.R. Doerge, Arch. Biochem. Biophys. 244 (1986) 678. [41] D.R. Doerge, N.M. Cooray, M.E. Brewster, Biochemistry 30 (1991) 8960. [42] S. Kobayashi, M. Nakano, T. Goto, T. Kimura, A.P. Schaap, Biochem. Biophys. Res. Commun. 135 (1986) 166. [43] S. Colonna, N. Gaggero, L. CascIJa, G. Carrea, P. Pasta, Tetrahedron Asymmctry 3 (1992) 95. |44] P. Pasta, G. Carrea, S. Colonna, N. Gaggero; Biochim. Biophys. Acta 1209 (1994) 203. [45} N.L. Klyachko, A.M. Klibanov, Appl. Biochem. Biotechnol. 37 (1992) 53. [46] M. Ortiz-Leon, L, Vclasco, R. Vazquez-Duhalt, Biochem. Biophys. Res. Commun. 215 (1995) 968. [47] R. Vazquez-Duhalt, D.W.S. WesUake, P.M. Fedorak, Enzyme Mícrab. Technol. 15 (1993) 494. [48] R. Vazquez-Duhalt, KLM. Semple, D.W.S. Westlake, P.M. Fedorak, Enzyme Microb. Technol. 15 (1993) 936. [49] T. Mashino, S. Nakamura, M. Hirobe, Tetrahedron Lett. 31 (1990) 3163. [50] Si Colonna, N. Gaggero. G. Carrea, P. Pasta, Tetrabedron Lett. 35 (1994) 9103. [51) E. Torres, J.V. Sandoval, F.I. Rosell, A.G. Mauk, R. Vazquez-Duhalt, Enzyme Mkrob. Technol. 17 (1995) 1014. [52] R. Tinoco, R. Vazquez-Duhalt, Enzyme Microb. Technol. 22 (1998) 8. [53] R. Vazquez-Duhalt, P.M. Fedorak, D.W.S. Westlake, Enzyme Microb. Technol. 14 (1992) 837. [54] R. Vazquez-Duhalt, F-Torres, R. Tinoco, in: B.C. Alternan, A. Leeson (Eds.), In Situ and On-Sitc Bioremediation, Vol. 2, Battelle Press, Colombus, OH, 1997, p. 225. [55J M.B. Arnao, M. Acosta, J.A. del Rio, R. Varón, F. García-Cánovas, Biochim. Biophys. Acta 1041 (1990) 43. [56J Z.C. Hu, R.A. Koms, C.R. Venkataramu, R i . Crawford, Enzyme Microb. Technol. 15 (1993) 567. [57] M. Mylrajan. K. Valli, H. Wariishi, M-H. Gold, T.M. Lochr, Biochemistry 29 (1990) 9617. [58] M. Huwiler, H. Jenzer, H. Kohler, Eur. J. Biochem. 158 (1986) 609. [59J A.N.P. Hiner, J. Hernándcz-Ruiz, K García-Cánovas, A.T. Smith, M.B. Arnao. M. Acosta, Eur. J. Biochem. 234 (1995) 506. [60] J.N. Rodriguez-Lópcz, i. Hernández-Ruiz, F. García-Cánovas, R.N.F. Thomeley, M. Acosta, M.B. Arnao, J. Biol. Chem. 272 (1997) 5469. Environ. Sci. Technol. 2000, 34, 2804-2809 Apéndice 7.2 Substrate Specificity and lonization Potential in Chloroperoxidase-Catalyzed Oxidation of Diesel Fuel MARCELA AYALA,'-' NORMA R. ROBLEDO,1 AGUSTÍN LOPEZ-MUNGUIA,T AND RAFAEL VAZQUEZ-DUHALT' Institute of Biotechnology, UNAM. Apartado Postal 510-3, Cuemavaca, Morelos 62271, México, and CEPROBi, ¡PN. Yautepec, Morelos, México Straight-run diesel fuel contaíníng 1.6% of sulfur was enzymatically oxidized with ctiloroperoxidase from Caldariomyces fumago. Most organosulfides and thiophenes were transformed to form suffoxides and suífones. The oxidized organosulfur compounds can be effectively removed by distiliation. The resulting fraction after distillation contained only 0.27% sulfur, while the untreated straight- run diesel fuel after the same distillation process still showed 1.27% sulfur. To know the Chemical nature of the producís, nine organosulfur compounds and 12 polycyclic aromatic compounds (PACs) were transformed by chloroperoxidase in the presence of chloride and hydrogen peroxide. Organosulfur compounds were only oxidized to form sulfoxides and suífones, and no chlorinated derivatives were detected, except for bithiophene. In contrast, PACs were exclusívely chlorinated, and no oxidized derivatives could be found. No enzymatic activity was detected on PACs with an ¡onization potential higher than 8.52 eVr while in the lower región it was found that the higher the ionization potential of the PAC the lower the specific activity. On the other hand, the substrate ionization potential did not seem to influence chloroperoxidase activity in the oxidation of organosulfur compounds. Ai! organosulfur compounds tested were oxidized by chíoroperoxidase. From double-substrate experiments, it appears that organosulfur compounds are oxidized by both compound I and compound X enzyme intermedíales, while PACs react only with the halogenating intermedíate, compound X. Introduction The environmental driver for diesel sulfur reduction is well- established. Meeting sulfur regulations on petroleum prod- ucís is driving up the cost of refining, because conventional hydrodesulfurization becomes increasely expensive and less efficient in handling sulfur removal as lower and lower sulfur levéis are reached (1). In the United States, there are plans to greatly reduce motor-vehicle emissions and sulfur contení * Corresponding author phone: +52 73 291619; fax: +52 73 172388; e-mail: maa@ibt.unam.mx. Mailing address: Apartado Postal 510-3. Cuernavaca. Morelos 62271, México. ' Institute of Biotechnology. UNAM. ! CEPROBI, IPN. 2804 - ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34, NO. 13, 2000 in gasoline (2). The plan would cut the average sulfur level in gasoline by 90%, from an average of about 330 ppm to 30 ppm, by 2004. Microbial desulfurization of fossil fuels has been under active investigaron for several decades and has been recently reviewed (3-5). Research groups and companies worldwide are developing the technology for fuel biodesulfurization, the most successful being a unique refinery process using bacteria to selectively remove sulfur from diesel. The patented bacteria, Rhodococcus IGTS8, has beengenetically engineered to increase both activity and stability (6). Recently, we have described an enzymatic method for fuel desulfurization (7). The method includes the steps of biocatalytic oxidation of organosulfur compounds contained in straight-run diesel fuel by chloroperoxidase from Cal- dariomyces fumago, followed by a distillation process in which the oxidized compounds are removed. Chloroperoxidase from Caldariomyces fumago (CPO) (EC1.11.1.10) is a versatile heme enzyme because of its catalytic diversity. CPO is a peroxide-dependent chlorinating enzyme, and it also cata- lyzes peroxidase-. catatase-, and cytochrome P450-type reactions of dehydrogenation, H2O2 decomposition, and oxygen insertion, respectively. This unusual combination of enzymatic activities is the origin of a number of studies involving CPO as a catalyst with potential applications, including the petroleum industry. It has been demonstrated that CPO is able to remove nickel and vanadium from asphaltene fractions (8). CPO is also able to perform interesting reactions. like the enantíoselective epoxidation of alkenes (9), oxidation of phenolic pollutants {10, 11), oxygenation of sulfides (7. 12), oxidation of organophos- phorus pesticides (13), and the determinaron of genotoxic potential of pollutants (14). to give only a few examples. However, the catalytic mechanism of CPO has not been completely established, and the exact role of chloride and the identity of the halogenating species remains a subject of controversy (¡5-18). The ability of fungal peroxidases to biotransform petro- leum compounds, such as polyaromatic hydrocarbons (PAHs), has been investigated before, specially lignin per- oxidase {LiP) and manganese peroxidase (MnP) from Phan- erochaete chrysosporium. These nonspecifíc extracellular enzymes are believed to be ¡nvolved in pollutant biotrans- formation. Interestingly, the activity of LiP and MnP correlates with the ionization potential (1P) of the PAHs. A threshold IP valué was found for each enzyme. LiP oxidizes PAHs with IP < 7.55 eV as well as some heterocyclic compounds with IP < 8 eV (19, 20). while MnP oxidizes PAHs with IPs as high as 8.1 eV (21, 22). With this evidence it was possible to distinguish whether a substrate was transformed vía an electrón subtraction process. Considering that both organosulfur and PACs are con- tained in diesel fuel, in the present work the enzymatic activity of CPO toward a group of several organosulfur compounds (thiophenes and organic sulfides) and PACs was determined. The chemical nature of reaction producís and the role of the substrate ionization potential were analyzed. Experimental Section Chemicals. Purified CPO from Caldariomyces fumago was produced in a fructose médium and purified according to Pickard (23); all preparations used in this study had an Rz = 1.36. which corresponds to 95% purity. Hydrogen peroxide and buffer salts were obtained from J. T. Baker (Phillisburg, NJ). Polycyclic aromatic compounds and aro- matic thiophenes and sulfides were purchased from Aldrich 10.1021/es99l270o CCC: $19.00 ©2000 American Chemical Society Publistied on Web OS/31/2000 /OJ TABLE 1. Sulfur Content of Straighl-Run Diesel Fuel after Enzymatic Oxídation with Chloroperoxidase (rom Caldariomyces fumago Followed by a Distillation lo 325 C as Final Distillation Point distíllate residue dtstillatton TPH3(%) sulfur(%) 83 1.27 17 3.21 " Total petroleum hydrocarbons. enzymatic TPH (%) 71 29 + distillation sulfur (%) 0.27 5.51 Chemical (Milwaukee, WI). HPLC-grade acetonitrile and methylene chioride were purchased from Fisher Scientific (Springfield. NJ). Reaction Conditions. Diesel fuel oxidations with chlo- roperoxidase were carried as previously reported (7). Oxida- tion reactions of individual organosulfur and aromatic compounds were carried out in a 1-mL reaction mixture containing 20/*M substrate and 15% acetonitrile in a 60 mM acétate buffer, pH 3.0. with or without 20 mM KC1 at room temperature. From 0.4 pmol to 0.2 nmol of the purifíed enzyme were used in the mixtures. Reactions were started by addition of 1 mM H2O2. Reaction rates were estimated by monitoring the substrate peak in a HPLC system equipped with a diode array detector. Enzyme activities were obtained from the differences in peak área after 10 min of reaction. transformed by a standard curve, and adjusted for protein concentration. Reported valúes are the mean of three replicates. Specific reaction rates are given as mol of substrate converted per mol of enzyme per minute or simply in min"'. For producís identification, 10-mL reactions were performed; after 1 h, the mixture was acidified and extracted with methylene chioride, and the extract was reduced under nitrogen, before being analyzed by GC-MS. Two-Substrate Reactions. Reaction mixtures contained either 20 fiM thianthrene or 30 /¿M pyrene and 100 /iM monochlorodimedone (MCD) in 15% acetonitrile in a 20 mM KC1, 60 mM acétate buffer pH 3.0. The reaction was started by addition of 0.25 mM H2O2 and monitored spectropho- tometrically at 288 nm (MCD) and eÍthér$fS1nme(tfiÍan- threne) or 335 nm {pyrene). Kinetic Constants Determination. Reactions were per- formed in 1 mL of 60 mM acétate buffer pH 3.0. 20 mM KC1 and either 15% for thianthrene or 20% acetonitrile for MCD and pyrene. Reaction was started by addition of 1 mM H2O2. The initial reaction rates were obtained by following the decrease in absorbance at 254 nm for thianthrene (e = 35 mM'1 CITT!) and at 335 nm for pyrene (É = 32.6 mM ' cm*1). Analytical Methods. Substrate concentration was mea- sured in a Perkin Elmer (series 200) HPLC system, using a Cis Hypersyl 5/ím Hewlett-Packard column and eluted with an acetonitrile-water (70:30 v/v) solvent mixture. Substrate and producís detection was carried out using a diode array detector coupled to the HPLC system. The used wavelengths for detection (/ldet) are Usted in Tables 3 and 4. Other UV measurements were made in a Beckman Spectrophotometer (DU 530). Product identification was performed in a Hewlett- Packard GC (model 6890)-MS (modei 5972) equipped with a SPB-20 column (30m x 0.25 mm, Supelco). The GC system was coupled to both a fíame ionization detector (FID, general detector) and a fíame photometric detector (FPD, specific sulfur detector). The temperature program started at 100 °C for 2 min; the temperature was raised to 290 °C at a rate of 8 °C/min and kept at 290 °C for 10 min. Microdistillations were carried out according to the standard test for boiling range distributíon of petroleum fractions by gas chromatography, ASTM D 2887-89. Organic sulfur content on diesel fuel were determined by X-ray fluorimetry in a Horiba X-ray fluorimeter. Total petroleum hydrocarbons (TPH) were estimated by the USEPA 8OJ5 method (modified). Results Sraight-run diesel fuel, obtained from primary distillation and containing 1.6% sulfur, was oxidized with chloroper- oxidase in the presence of 20 mM KC1 and 1 mM hydrogen peroxide. The gas chromatographic analysis with both fíame TABLE 2. Mass Spectral Data of Products" substrate product benzothiophene benzothiophene sulfone diphenyl sulfide diphenyl sulfone dibenzothiophene di benzothiophene sulfone thianthrene 5-thianthrene oxide 5,10-thianthrene dioxide acenaphthene dichloroacenaphthene trichloroacenaphthene anthracene 9,10-dichIoroanthracene biphenylene dichlorobiphenyiene trichlorobíphenylene fluorene dichlorofluorene phenanthrene chlorophenanthrene pyrene chloropyrene dichloropyrene triphenylene chlorotriphenylene bithiophene dichlorobithiophene trichlorobithiophene tetrachlorobithiophene "Valúes in parenlheses are relative abundances. mass spectral ions [miz) 166 (42) [M-J, 138 (9), 137 (100), 118 (15), 109 (48), 90 (14), 89 (16), 76 (15), 75(15), 74 (14), 65 (9), 63 (13) 218 (27) [MI, 125 (100), 97 (26), 77 (53), 51 (47), 50 (16) 216 (100) [WT], 187 (46), 160 (31), 150 (16), 139 (30) 232 (16) [M*L 184 (100), 171 (15), 139 (14), 69 (14) 248 (77) [M'L 200 (86), 184 (84), 171 (100), 168 (23), 139 (30), 108 (24), 69 (36) 224 (39), 222 (64) [M+], 187 (100), 152 (95), 93 (17), 75 (24) 258 (57), 256 (81) (M+], 221 (66), 186 (100), 150 (50), 110 (27), 98 (18), 75 (23) 248 (68), 246 (100) [M+], 176 (43), 87 (10) 222 (64), 220 (100) [M+], 185 (17), 150 (45), 75 (11) 258 (30), 256 (93), 254 (100) [MÍ, 219 (13), 184 (49), 149 (14), 74 (10) 238 (25), 237 (7), 236 (40) [Mi, 201 (31), 199 (18), 166 (63), 165 (100), 164 (17), 163(24), 100 (11), 82 (35) 214 (32), 213 (16) [M+J, 212 (100), 177 (20), 176 (55), 175 (11), 174 (10), 151 (14), 150 (14), 106 (17), 88 (33), 87 (11), 75 (13) 238 (31), 236 (100) ÍM+], 200 (34), 100 (12) 272 (62), 270 (100) IJVH, 235 (11), 200 (53), 135 (12), 100 (23) 265 (7), 264 (34), 263 (21) [M^, 262 (100), 227 (14), 226 (63), 225 (16), 224 (23), 200 (11), 132 (9), 131 (20), 113 (56), 112 (43), 100 (21), 99 (12), 87 (9) 238(15), 236 (72) [M+], 234 (100), 201 (36), 164 (45), 157 (28), 155 (76), 142 (10), 119 (21), 93 (17), 82 (19), 79 (14), 69 (30) 272 (37), 271 (11), 270 (100) [M'J. 233 (58), 198 (81), 191 (53), 189 (82), 163 (15), 154 (33), 135 (18), 119 (37), 103 (19), 93 (39), 81 (46), 79 (58), 69 (36), 58 (12) 308 (13), 306 (49), 304 (96) [M~], 302 (71), 267 (52), 232 (44), 223 (39), 197 (22), 188 (30), 162 (12), 153 (76), 117 (72), 98 (11), 93 (36), 81 (74), 79 (100), 69 (26) VOL. 34, NO. 13, 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY • 2805 S&f TABLE 3. Specilic Activity of CPO with Organosulfur Compounds specific i(nm) IP(eV) activity (min' 1 thianthrene 2 2,2'-b¡thiophene 3 diphenyl sulfide 4 dibenzothiophene 5 benzothiophene 6 ethyl phenyl sulfide 7 benzenethiol 8 thioanisote 9 diphenyl disulfide 254 300 248 232 226 254 238 256 240 7.80 7.83a 7.88 8.39 8.73 8.80 8.90 8.95 9.40 1310 (± 132) 840 (± 8) 831 (± 32) 126 (±9) 557 (± 42) 1725 (±145) 116 (±5) 2917 (± 58) 352 (± 10) * IP measured by charge transfer (25). TABLE 4. Specific Activity of CPO with Aromatic Compounds 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 azufene 9-methylanthracene anthracene biphenylene 2-methyIanthracene pyrene acenaphtene ftuorene fluoranthene phenanthrene triphenylene naphthalene biphenyl dibenzofuran anthrone Adei (nm) 270 254 250 248 248 236 226 260 236 250 256 220 250 280 260 IP (eV) 7.43a 7.46 7.51 7.56a 7.70 7.72 7.73a 7.91 7.95a 8.07 8.10 8.18 8.64 8.77 9.43 IP measured by photoelectron spectroscopy (25). detected. specific activity Imin"1) 676 (± 34) 758 t±27) 134 (±14) 10 (±0.5) 107 (± 8) 53 (± 6) 65 (± 8) 1.9 (±0.13) 3 (± 0.2) 7 (±0.1) 0.8 (± 0.09) 0.6 (± 0.01) NR" NR* NR" 6 NR: no teactton ionization (FID) and fíame photometric (FPD) detectors showed that chloroperoxidase was able to oxidize most of organosulfur compounds contained in the diesel fuel. The oxidation was detected by the increase of boiling point (retention time) of these compounds on the gas chromato- gram monitored by the sulfur selective detector (FPD). Microdistillation of both chloroperoxidase-oxidized and untreated diesel fuels monitored by FID (general detection) and FPD (sulfur selective detection) (Figure 1) shows that the hydrocarbon distillation profile changes slightly after enzymatic treatment. In contrast. the specific sulfur detector (FPD) shows a significant change of the distillation profile, in which most of organosulfur compounds were effectively oxidized and theír boiling points increased after enzymatic treatment. Oxidized sulfur compounds can be removed by a distil- lation process (Table 1). After distillation, the sulfur contení in the enzymatically oxidized diesel fue] is only 0.27%, while for the untreated fuel is 1.27%. The distillation of the straight- run diesel fuel (1.6% sulfur) to a final distillation point of 325 °C produced a distillate containing 66% of the total sulfur, while if the diesel fuel is previously oxidized with chloro- peroxidase, the obtained distillate contained only 12% of the total sulfur. Thus, by using an enzymatic oxidation with chloroperoxidase coupled with a distillation process it is possible to obtain a diesel fue! with six times lower sulfur concentration than straight-run diesel fuel. Few hydrocar- bons are also transformed during the enzymatic treatment, and after distillation an additional 12% of them remain in the residue (Table 1). To know the chemical nature of the producís from the enzymatic reaction, nine organosulfur compounds, including 2 8 0 6 • ENVIRONMENTAL SCIENCE & T6CHNOLOGY / VOL. 34, NO. 13, 2000 100 - 80 - 60 40 • 20 • 0 Apéndice 7 2 Total Petroleum Hydrocarbons Untreated diesel fuel \ CPO oxidized diesel fuel 2 O 100 80 S° 60 Organosulfur compounds Untreated diesel fuel 04 200 CPO oxidi2ed diesel fuel 250 300 350 400 Boiling Point Í°C) 450 500 FIGURE 1. Microdistillation of untreated and chloroperoxidase- oxidized straight-run diesel fuel: FID, fíame ionuation detector (general detector) and FPD, fíame photometric detector {sulfur selective detector). thiophenes, organic sulfides, and thiols, and 15 aromatic compounds were tested for chloroperoxidase transformation. Table 2 shows the products identífied by GC-MS. Products from all the organosulfur compounds were their respective sulfoxides and sulfones, except for biothiophene for which chlorinated derivatives were detected. Sulfones are the final product of CPO reactions; successive additions of both enzyme and H2O2 to complete substrate modification did not change the chemical nature of the products. In addition, sulfone standard compounds were not substrate for CPO as determined by GC and HPLC methods. Aromatic hydrocarbons are also important constituents of diesel fuel. It is well know that chloroperoxidase is able to transform some PAHs (14, 24). Twelve of the 15 PACs tested were transformed by CPO in the presence of 1 mM H2OZ and 20 mM KCf, as monitored by HPLC. GC-MS analysis of the reaction products showed that the substrates were exclusiveiy chlorinated during the reaction (Table 2). Fur- thermore, in the absence of chloride there was not observable reaction. Tables 3 and 4 show the specific activity of CPO and IP valúes for the organosulfur and PACs compounds assayed. The ionization potentials (IP) taken are measured by electrón impact, except for azulene, biphenylene, fluoranthene, and bithiophene (25). Figure 2 shows the correlation between IP valúes and specific activity for PACs and organosulfur compounds. To determine the eflfect of the presence of a good substrate for halogenation, such as monochlorodimedone (MCD), thianthrene oxidation and pyrene halogenation reactions were performed in the presence of 0.1 mM MCD (Figure 3). Under these conditions, thianthrene was inítially oxidized to form a sulfoxide with a significantly low rate (Table 3). Once MCD was exhausted, thianthrene oxidation rate became similar to that found in the absence of MCD (Figure 3a). In the case of pyrene, halogenation did not start until all MCD was halogenated, suggesting a strong affinity of MCD for the enzyme (Figure 3b). Under our experimental conditions, the specific activity of halogenation of MCD is 3480 min1 . The specific reaction rate is only slightly affected in the presence of pyrene (3200 min"1). On the other hand. the presence of thianthrene decreases the specific reaction rate for haloge- nation of MCD (2830 min"1). while the initial rate for 2- 1- o 3 s .2 2- 1 i-I o. W o> o O Organosulfur compounds Polycydic aromatic compounds 7.5 8.5 9.5 lonizalion Potential (eV) FIGURE 2. Influence of the subslrate íonization potenttal on the specífic activity of CPO. Substrates are organosulfur compounds (•) and polycydic aromatic compounds (O). Numbers in superior and inferior paitéis correspond to those in Tables 3 and 4, respectively. 100 * 2 4 , 6 Time (min) 100 *• • MCD A Pyrene 1 2 3 4 5 Time (min) FIGURE 3. Competition between (A) MCD and thianthrene and (B) MCD and pyrene. thianthrene oxidation is 300 min" ' , until MCD is exhausted. The addition of the MCD halogenation and thianthrene oxidation rates results in a valué cióse to that obtained with MCD alone. Kinetic constants for pyrene halogenation and thianthrene oxidation were determined (Table 5). Chloroperoxidase is a more efficient catalyst in the reaction of oxidation of thianthrene than in the reaction of halogenation of pyrene. 7 7 MCD thianthrene pyrene 94 64 37 1.4 1.5 32 TABLE 5. Kinetic Constants for Thianthrene Oxidation and Pyrene Halogenation substrate kcat (s"1) KM (/*M) kcaJKM {/itA* i 67 44 1.2 as can be seen from the catalytic efficiencies IÍC^/KM. Though both the affinity and the catalytic constant are higher for thianthrene, the main effect comes from the affinity of the enzyme for the substrate, whichis 1 orderofmagnitudelower for pyrene. Discussion Enzymatic oxidation of diesel fuel allows the organosulfur compounds to be separated by a single distillation process. Chloroperoxidase from C. fumago is a very active enzyme able to perform transformaüon of complex oil fractions, such as dieseí (7) and asphaltenes (8). Chloroperoxidase shows three different catalytic activities: halogenase, peroxidase, and catalase (26-28).Jn addition, some reports nave claimed that chloroperoxidase catalyzes two-electron reactions (per- oxygenase), which could be considered a kind of monooxy- genase activity (29-32). Nevertheless, when organosulfur compounds such as thiophenes and organosulfides are substrates, mainly sulfoxides are formed by the peroxidase activity (Table 2 and Figure 1). All nine organosulfur compounds tested were oxidized by chloroperoxidase, even when the reaction system contained 20 mM KCI {Table 3), except for 2,2'-bithiophene from which halogenated deriva- tives were detected. These results are in agreement with previous work reporting that sulfoxides are produced from chloroperoxidase activity (30, 32, 33). On the other hand, polycydic aromatic compounds (PACs) are halogenated (Table 2). Other peroxidases, such as Ügnin peroxidase (19, 20) and manganese peroxidase (21, 22) and even hemoproteins with peroxidase activity (24,34), produce mainly quiñones from PAHs oxidation. Specific activity of chloroperoxidase on PACs halogenation shows a clear correlation with the substrate ionization potential {Figure 2). Because ionization potential could be defined as the energy involved in taking out one electrón from the substrate molecule, this correlation suggests a one-electron mechanism with a free radical-mediated reaction. Only PACs with ionization potentiai lower than 8.52 eV were halogenated (Table 4). In general, the lower the ionization potential of the PAC, the higher the specific activity of the chloroperoxidase for that substrate (Figure 2). The ionization potential valué of 8.52 eV appears to be a threshold, as none of the compounds tested having higher ionization potentials were transformed by chloroperoxidase. This threshoid valué is significantly higher than those reponed for other peroxidases. Lignin peroxidase is able to oxidize PAHs and form quiñones up to a PAHs ionization potential of 8.0 eV (20), and manganese peroxidase from P. chrysosporium shows a threshold valué for PAHs substrates of 8.1 eV (22). Interest- ingly. no clear correlation could be found between the ionization potential and the specific activity for organosulfur compounds (Figure 2). In fact, we were not abie to found a single organosulfur compound, thiophene or sulfide, which is not transformed by chloroperoxidase. A possible production of chlorinated derivatives from PAHs by CPO reactions is an undesirable side of the process. However. as shown in Tables 3 and 4 (enzyme activity on single substrates} and in Table 5 (affinity constants. KM, for thianthrene and pyrene), organosulfur compounds are better substrates and therefore can compete favorably with PAHs. This means that in a mixture containíng both types of VOL. 34, NO. 13. 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY • 2807 Compound Apéndice 7.2 A*+OH AH Compound 1 RH + H FIGURE 4. Proposed catalytic cycle of chloroperoxidase. compounds, the sulfur compounds would be preferenüally oxidized fay CPO. Nevertheless, the reaction conditions and the biocatalyst preparation should be designed to minimize the halogenation reactions because of the environmental implications of chlorínated aromatic compounds. So far, the catalytic cycle of chloroperoxidase is not completely elucidated {15, 16. 18, 35). The proposed mech- anism (Figure 4) includes a first activation step, in which hydrogen peroxide transforms the (Felu)porphyrin group (native state) to oxofFe1^ porphyrin radical catión (compound I). Then compound I can follow two ways: the oxidation of a substrate molecule to form an oxofFe'̂ O porphyrin without the associated porphyrin ^-radical catión (compound II) or the reaction with a chlorine ion to form a CIO (Fe"1) porphyrin group. called compound X, which is the only responsifale for the enzymatic reaction of halogenation. In addition, this compound X seems also to be able to perform oxidation reactions liberating a chlorine ion. After both reactions, compound X returns to the native (Fe111) porphyrin state. From this proposed mechanism, it seems that the organosulfur compounds are able to react with both compound I and compound X. while PACs are only reactive to compound X. This is in agreement with our results, as when a high affinity halogenation substrate (MCD) is present in the médium, a slow thianthrene oxidation is found (Figure 3a). The oxidation rate is lower because most of compound I is rapidly transformed to compound X, due to rapid compound X turnover by MCD reaction. Thus the observed thianthrene oxidation is mainly mediated by compound X. When MCD is exhausted, thianthrene competes more favorably with the chlorine ions for compound I, its transformation involving both forms: compound I and compound X. This competition between a halogenation substrate (MCD) and a peroxidase substrate (catechol) has been previously reported (35). In this case, MCD quantitatively replaces catechol as a substrate for part of the enzymatic reaction. In contrast, and as expected. chloroperoxidase is not able to react with pyrene when MCD is present in the médium (Figure 3b), a situation that can be explained by the significant differences between the catalytic efficiencies of MCD S~') and pyrene (Acal/iÍM = 1.2 fiM~l s"1)- The main effect comes from the different affinity of chloroperoxidase for the substrates, whereas for MCD KM = 1.2 /ÍM, and for pyrene /ÍM = 32,«MT RCI + H;O 1 order of magnitude lower. Under these conditions the available halogenating active sites are readily saturated by MCD; pyrene, which is unable to compete for compound X, is transformed only until MCD is exhausted. Chloroperoxidase from C. fumago catalyzes the oxidation of most of organosulfur compound found in straight-run diesel fuel. This oxidation allows the desulfurization of diesel fuel by distillation. Sulfoxides and sulfones are the main producís from CPO reaction on organosulfur compounds, while halogenated aromatic compounds are the only prod- ucís from PACs reactions. Furthermore, PACs halogenation by chloroperoxidase seems to be dependent on the substrate ionization potential. In general, PACs with an ionization potential of < 8.52 eV were halogenated. Our results support a free radical mechanism for enzymatic halogenation and a catalytic cycle in which compound X [C10(Felll)porphyrin] could be responsible for both substrate halogenation and oxidation in a chloríne-dependent process. The broad specificity and high activity of chloroperoxidase encourage further investigation in the use of this enzyme as an efficient catalyst in a desuifurization process, including an enzymatic treatment followed by a fractional distülation step. Sulfur removal from a very complex mixture, such as petroleum fractions, is far from being accomplished. Con- ventional hydrodesulfurization becomes expensive and less efficient as lower and lower sulfur levéis are reached. The biotechnological process could be applied after a conven- tional desulfurization process in order to reach these new regulatory low-levels for sulfur contení in fuels. At the moment, the use of chloride as an aclivator in this process seems unavoidable, since in this hydrophobic médium, chloroperoxidase presents very iow activity and chloride greatly improves the reaction rate. Unfortunately, the pres- ence of halogens would yield some environmentally unde- sirable products. Our research is currently focused on the protein engineering of chloroperoxidase in order lo reduce ifs halogenase activity, maintaining or increasing the per- oxidase activity. In addition, different approaches to improve the stability of chloroperoxidase are under research, such as genetic engineering and cross-linking of enzyme crystals. The stabilization of enzymes in non-conventional low-water contení médium is a priority for the successful development of industrial enzymatic processes. 2808 • ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 34. NO. 13, 2000 Acknowledgments This work was supported by DGPA-UNAM grant IN 220598 from the National Autonomous University of México (UNAM). Lrterature Cited (1) Hunt, P. OH CasJ. 1996, 94. 53. (2) Grisham, J. L. Chem. Eng. News 1999. 77, 21. (3) Monticello, D. J.; Finnerly. W. R. Annu. Rev. Microbio!. 1985.39, 371. (4) Monticello D. J. CHEMTECH 1998, 28, 38. (5) McFarland. B. L; Boron, D. J.; Deever, W.; Meyer, J. A; Johnson A. R.; Atlas. R. M. Crít. Rev. Microbio!. 1998. 24, 99. (6) Pacheco. M. A.; Lange, E. A.; Pienkos, P. T.; Yu, L.Q.; Rouse. M. P.; Lin, Q.; Linguist. L. K. National Petrochemical and Refíners Association Annual Meeting, March 21-23. San Antonio. TX, 1999. (7) Ayala, M.; Tinoco, R.; Hernández, V.; Bremauntz, P.; Vazquez- Duhalt. R. Fue! Processing Technol 1998, 57. 101. (8) Fedorak. P. M; Semple, K. M.; Vazcjuez-Duhalt. R.; Westlake. D. W. S. Enzyme Microb. Technol. 1993. ¡5, 429. {9) Zaks. A.; Dodds, D. R. / Am. Chem. Soc. 1995. 117. 10419. (10) Aitken. M. D.; Massey, I. J.; Chen. T.; Heck, P. E. Water Res. 1994, 28, 1879. (11) Carmichael. R.; Fedorak, P. M.; Pickard M. A. Biotechno!. Leu. 1985. 7, 289. (12) Caseila, L; Colonna. S.; Carrea. G. Biochemistiy 1992.31,9451. (13) Hernández,].; Robledo. N. R.; Velasco, L.;Quintero, R.; Pickard. M. A.; Vazquez-Duhalt, R. Pest. Biochem. Physiol, 1998. 61,87. (14) Márquez-Rocha. F. J.; Pica-Granados. Y.; Sandoval-Villasana, A. M.; Vazquez-Duhalt. R. Bull. Environ. Contam. Toxico!. 1997. 59. 788. (15) Dunford. H. B.; Lambeir, A.; Kashem, M.; Pickard M. A. Arch. Biochem. Biophys. 1987, 252. 292. (16) Libby, R. D.; Beachy, T. M.; Phipps A. K. / Biol Chem. 1996.271, 21820. (I?) Sundaramoorthy,M.;Temer. J.: Poulos.T. L. Chem. Biol. 1998, 5. 461. (18) Wagenknecht. H. A.; Woggon. W. D. Chem. Biol. 1997. 4. 367. (19) Hammel. K. E.; Kalyanaraman. B.; Kirk, T. K. /. Biol. Chem. 1986. 261. 16948. (20) Vazquez-Duhalt. R.; Westiake. D. W. S.; Environ. Microbiol. 1994. 60. 459. (21) Bogan, B. W.; Lámar, R. T. App!. Environ. Microbio!. 1995, 61, 2631. (22) Bogan. B. W.; Lámar. R. T.; Hammel, K. E. App!. Environ. Microbio!. 1936, 62. 1788. (23) Pickard. M. A.; Kadima. T. A.; Carmichael, R. D. /. índ. Microb. 1991. 7. 235. (24) Torres. E.; Tinoco, R.; Vazquez-Duhalt. R. Wat. Sci. Tech. 1997, 36, 37. (25) Ion EnergeticsData. In NISTChemhtry WebBook, NISTStandard Reference Datábase Number 69 (Online); Mallard, W. G., Linstrom, P. J.. Eds.; National Institute of Standards and Technology: Gaithersburg, MD. 1998 (http://webbook.nist.gov) (September 28. 1999. last date accessed). (26) Hager, L. P.; Morris, D. R.; Brown, F. S.; Eberwein. H. / Biol. Chem. 1966, 241, 1769. (27) Thomas. J. A.; Morris. D. R.; Hager, L. PJ. Biol. Chem. 1970.245. 3! 29. (28) Thomas, J. A.; Morris. D. R.; Hager. L. PJ. Biol Chem. 1970,245. 3135. (29) Kedderis. G. L; Rickert. D. E.; Pandey. R. N.; Hollenberg. P. F. J. Biol Chem. 1986. 261, 15910. (30) Kobayashi,S.; Nakano,M.; Kimura.T.; Schaap, A. P. Biochemistry 1987. 26, 5019. (31) McCarthy, M. B.; White R. E.J. Biol Chem. 1983. 258. 9153. (32) Ortiz de Montellano, P. R.; Choe. Y. S.; DePiliis, G.; Catalano. C. E. /. Biol. Chem. 1987, 262. 11641. (33) Pasta. P.; Carrea, G.; Colonna.S.; Gaggero, N. Biochim. Biophys. Acta. 1994. 1209. 203. (34) Tinoco. R.; Vazquez-Duhalt. R. Enzyme Microb. Technol. 1998. 22.8. (35) Libby. R. D.; Rotberg, N. S.;. Emerson. J. T.; White. T. C.; Yen, G. M.; Friedman, S. R; Sun, N. S.; Goldowski. R. / Biol. Chem. 1989, 264, 15284. Received íor review November 10, 1999. Revised manuscript received March 16, 2000. Accepted April 5. 2000. ES991270O VOL. 34. NO. 13. 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY • 2 8 0 9 PERGAMON Phytochcmistry 58 (2001) 929-933 Apéndice 7.3 P H YTOC HEMISTRY www.clscvicr.com/locate/phytochcm Biocatalytic chlorination of aromatic hydrocarbons by chloroperoxidase of Caldariomyces fumago Rafael Vázquez-Duhalta>*, Marcela Ayalaa, Facundo J. Márquez-Rochab *Biotechnology Inslitule UNAM, AP510-3, Cuemavaca, Morelos 6227i, México bMarine Bioprocess Enginering Laboratory, CICESE, Baja California, 22860, México Receivcd 19 March 2001; rcceivcd in rcvised form 5 July 2001 Abstract Chloroperoxidase from Caldariomyces fumago was able to chlorinate 17 of 20 aromatic hydrocarbons assayed in the presence of hydrogen peroxide and chloride ions. Reaction rates varied from 0.6 min"1 for naphthalene to 758 min"1 for 9-methylanthracene. Mono-, di- and tri-chlorinated compounds were obtained from the chloroperoxidase-mediated reaction on aromatic compounds. Dichloroacenaphthene, trichloroacenaphthcne, 9,10-dichloroanthracene, chloropyrene, dichloropyrene, dichlorobiphenylene and trien!orobiphenylene were identified by mass spectrai analyses as produets from acenaphthene, anthracene, pyrene and biopheny- lene respectively. Polycyclic aromatic hydrocarbons with 5 and 6 aromatic rings were also substrates for the chloroperoxidase reaction. The importance of the microbial chlorination of aromatic pollutants and its potentia! environmental impact are discussed. © 2001 Elsevier Science Ltd. All rights reserved. Keywords: Fungal chloroperoxidasc; Enzymatic chlorination; Polycyclic aromatic hydrocarbons 1. Introduction Organochlorine industrial compounds, chlorinated pesticides and polychlorinated biphenyls (PCBs) are considered among the most important pollutant xeno- biotics. Their toxic effeets have been extensively studied (Evangelista de Duffard and Duffard, 1996; Giesy and Kannan 1998; Tilson and Kodavanti, 1998). Because of their potential publie health risk, some organochlorine compounds, such as PCBs, have been banned in western countries, but many are still manufactured and used as pesticides, plasticizers, paint and printing-ink compo- nents, adhesives, fíame retardants, hydraulic and heat transfer fluids, refrigerants, solvents, additives for cut- ting oils, and textile auxiliarles. Thus, contamination with organochlorine compounds still oceurs and this is of great publie concern due to potential toxicity to humans and wíldlife. Microbiological studies have been almost entireíy focused on the degradation of these toxic compounds (Robinson, 1998; Wiegel and Wu, 2000; Chaudhry and Chapalamadugu, 1991). * Corresponding author at: Instituto de Biotecnología UNAM, Aparto Postal 510-3, Cucrnavaca, Morelos 62250, Mcxico. Tel.: + 52- 5622-7655; fax: +52-7317-2388. E-mail address: vazqduh(a;ibt.unam.mx (R. Vázqucz-Duhall). In spite of the microbial capacity to produce haloge- nated compounds (Neidleman, 1975; de Jong and Field, 1997), information is scarce on the microbial production of toxic organochlorine compounds. ín addition to de novo synthesis of chlorinated compounds, microorgan- isms transform some non-halogenated xenobiotics into organochlorine compounds with a possible increase in their toxicity. In this work, we show the capacity of chloroperoxidase from the fungus Caldariomyces fumago to chlorinate aromatic hydrocarbons, including polycyclic aromatic hydrocarbons (PAHs). PAHs are widely dispersed in the environment, and they are con- sidered to be a potential health risk because of their possible carcinogenic and mutagenic activitíes. Chloro- peroxidase (CPO) is a 42,000 Da extracellular heme glyco- enzyme containing ferriprotoporphyrin IX as the prosthetic group (Sundaramoorthy et al., 1995). CPO exhibits a broad spectrum of chemical reactivities, it is a peroxide-dependent chlorinating enzyme and it also cata- lyzes peroxidase-, catalase- and cytochrome P450-type reactions of dehydrogenation, H2O2 decomposition and oxygen insertion, respectively (Yi et al., 1999). CPO is only one of a variety of halogenase enzymes that can be found in nature, other enzymes such as vanadium and non-heme halogenases (van Pee et al., 2000; Vollenbroek et al., 1995) are also potentially able to chlorinate organic pollutants. 0031-9422/01/$ - scc front matter © 2001 Elsevier Science Ltd. A!l rights reserved. P I1 : S0031-9422(Ol)00326-O Apéndice 7.3 930 2. Results and discussion R. Vázquez-Duhait et al. / Phytochemistry 58 (2001) 929-933 Chloroperoxidase was able to utilize 17 of the 20 aromatic hydrocarbons assayed as substrate (Table 1). Only biphenyl, and the oxygen-eontaining dibenzofuran and anthrone, were not substrates for CPO under our reaction condítions. The specific activity valúes were from 0.6 min~' for naphthalene to 758 min~' for 9- Table l Specific activiiy of chloroperoxidase from Caldariomyces futnago against aromatic compounds Aromatic compound Specific activity (mirT1) 9-Mcthytanthracenc Azulenc Anthraccnc 2-Methylanthraccne 7,12-Dimethylbenzanthracene Benzo[a]pyrcne 7-Methylbcnzo[a]pyrcne Accnaphthcnc Pyrcne Benzo[gh¡]pcry!cnc Pcrylcnc Biphcnylcnc Phenanthrenc Fluoranthcnc Fluorcne Triphenylenc Naphthalcnc Biphenyl Dibenzofuran Anthrone 758 (±27) 676 (±34) 134 (±14) 107 (±8) 87 (±8) 84 (±6) 81 (±7) 65 (±8) 53 (±6) 45 (±7) 25 (±10) 10 (±0.5) 7 (±0.1) 3 (±0.2) 1.9 (±0.13) 0.8 (±0.09) 0.6 (±0.01) NRa NR" NR" NR, no rcaction detected. methylanthracene. Interestingly, recalcitrant and carci- nogenic 5- and 6-aromatic rings PAHs were substrates for chloroperoxidase in the presence of hydrogen per- oxide and chloride ions. Chloroperoxidase is able to perform a broad range of reactions, like the enantio- selective epoxidation of alkenes (Zaks and Dodds, 1995), oxidation of phenolic pollutants (Aitken et al., 1994; Carmichael et al., 1983), oxygenation of sulfides (Colonna et al., 1990), oxidation of organophosphorus pesticides (Hernández et al., 1998), and the PAH-DNA adduct formation (Márquez-Rocha et al., 1997), to give only a few examples. In addition, CPO is able to oxidize very complex molecules, such as asphaltenes (Fedorak et al., 1993) and complex mixtures such as petroleum distillates (Ayala et al., 1998). Thus this halogenating enzyme is the most versatile enzymatic hemoprotein. Other halogenases are also able to perform halogenations, mainly chlorination, on a variety of substrates (Neidle- man, 1975; Vollenbroek et al., 1995). The chemical nature of the reaction producís was determined by gas chromatography-mass spectrometry (GC-MS). The mass spectra of the producís from the enzymatic reaction with acenaphthene, anthracene, biphenylene, fluorene, phenanthrene, pyrene, and tri- phenylene are shown in Table 2. Monochlorinated, dichlorinated and trichlorinated compounds were found (Fig. 1), and they showed major ions at m¡z= [M + ]-35, m/z = [M+]-70, and m/z = [M+]-105, respectively. No oxygen incorporation was detected in any of the pro- duets from the CPO-mediated reactions. In addition, no reaction could be detected with any PAH tested under peroxidase activity conditions (pH 5.0 and in the absence of chloríne ions). Significant information is available on the toxicity of chlorinated aromatic com- pounds, such as polychlorophenols (International Table 2 Mass spectral data of the producís from the chloropcroxidasc-mediated reaction on aromatic compounds Substrate Product Mass spectral ions {m¡zf-b Acenaphthene Anthracene Biphenylene Fluorene Phenanthrene Pyrene Triphcnylcne Dichloroacenaphthenc Trichloroacenaphthene 9,10-Dichloroanthracene Dichlorobiphcnylcnc Trichlorobiphenylcne Dichlorofliiorcne Chlorophenanthrcnc Chloropyrenc Dichloropyrcnc Chlorotriphenylenc 224 (39), 222 (64) (M +], 187 (100), 152 (95), 93 (17), 75 (24). 258 (57), 256 (81) (M ' ], 221 (66), 186 (100), 150 (50), 110 (27), 98 (18), 75 (23). 248 (68), 246 (100) [M + ], 176 (43), 87 (10). 222 (64), 220 (100) [M ' ], 185 (17), 150 (45), 75 (11). 258 (30), 256 (93), 254 (100) \M ' ], 219 (13), 184 (49). 149 (14), 74 (10). 238 (25), 237 (7), 236 (40) [M*]. 201 (31), 199(18), 166(63). 165(100), 164(17), 163(24), 100 til), 82(35). 214(32), 213 (16) [M f J, 212 (100). 177(20). 176(55), 175(11), 174(10), 151 (14), 150(14). 106(17), 88 (33), 87 (11). 75(13). 238 (31). 236 (100) [M + ], 200(34), 100(12). 272 (62). 270 (100) [M *], 235 (11), 200 (53), 135 (12), 100 (23). 265 (7), 264 (34), 263 (21) [Mf ], 262 (100), 227 (14), 226 (63), 225 (16), 224 (23), 200(11), 132(9), 131 (20), 113(56). 112(43), 100 (21), 99 (12). 87 (9). " Valúes in parentheses are rclativc abundances. h [M ' ] molecular ion. Apéndice 7.3 R. Vázquez- Duhalt el al. I Phytochemistry 58 (2001) 929-933 CPO 931 Cl Cl Cl C!, Phenanthrene CPO Triphenylene Fig 1. Chlorinated producís from thc cnzymaiic halogenation of aromatic compounds by chloropcroxidasc in thc presence of hydrogen pcroxide and chioridc ions. Agency for Research on Cáncer, 1999) and PCBs (Robinson, 1998; Wiegel and Wu, 2000), and to a less extent, on the toxicity of chlorinated derivatives of polycyclic aromatic hydrocarbons. Chlorinated PAHs have been tested for mutagenic activity by the Ames test on Salmonella typhimurium (Colmsjo et al., 1984; John- sen et al., 1989). Chlorinated fluorene, flouranthene and benzo(a)pyrene acted as strong mutagens both in the presence and in the absence of metabolic activation, while only benzo(a)pyrene showed mutagenic activity as parent hydrocarbon. Mono- and di-chloropyrene iso- mers showed from 40 to 4000 times higher mutagenic activity than 1-nitropyrene and pyrenoquinones, respec- tively. On the other hand, a mixture of chlorinated chrysene isomers was consíderably more potent than the párent hydrocarbon in terms of embryolethality and cytochrome P450 induction (7-ethoxyresorufin-O-de- ethyfase and aryl hydrocarbon hydroxylase) (Gustafsson et al., 1994). The chlorinated chrysene caused anoma- lies, including edema and beak defects, similar to those reported after treatment of chick embryos with coplanar PCBs. These effects of the chlorinated mixture were mainly accounted for by 6-chlorochrysene and 6,12- dichlorochrysene. Chloronaphthalene was 5000-times Apéndice 7.3 932 R. Vázquez-Duhah et al. ¡ Phytochemislry 58 (2001) 929-933 more potent than naphthalene for the inhibition of mitochondrial respiration in molar basis (Beach and Harmon, 1992). Monochloronaphthalene represent a risk for human health (Tsunenar et al., 1982) and for aquatic organisms (Ward et al., 1981). 3. Conclusions Chloroperoxidase from the imperfect fungus Caldar- iomyces fumago is the most versatile enzyme in the hemoprotein family (Yi et al., 1999). CPO performs halogenase, peroxidase, catalase and cytochrome P450- like reactions. However, under our reaction conditions and with PAHs as substrates, CPO only acts as halo- genase and no oxygenated producís could be detected. In contrast, peroxidase actívity on aromatic compounds produces mainly quiñones, such as in the case of lignín peroxidase (Hammel et al., 1986; Vazquez-Duhalt et al., 1994) and manganese peroxidase (Bogan et al., 1996). Caldariomyces fumago has been isolated from damp sites and also has been reported as a marine fungus (Dawson and Sonó, 1987; Colonna et al., 1999). Chloro- peroxidase, as other halogenases, is an extracellular en2yme that can react with a variety of substrates in the microbial environment. This enzyme is able to catalyze PAH-DNA adduct formation in vitro (Márquez-Rocha et al., 1997), suggesting the production of genotoxic aromatic intermediates. The present work shows that the transformation of aromatic pollutants into chlori- nated derivatives by microbial enzymes may occur in polluted sites. This biocatalytíc process should be con- sídered because the toxicity and environmental impact of aromatic compounds may be increased. 4. Experimental 4.1, Chemicals Purified CPO [EC 1.11.1.10] from C. fumago 89362 (Commonwealth Mycological Institute, Kew, Surrey, UK) was produced in a fructose médium and purified according to Pickard et al. (1991); all preparations used in this study had an Rz=l36, which corresponds to 95% purity. Hydrogen peroxide and buñer salts were obtained from J.T. Baker (Phillisburg, NJ). Polycyclic aromatic hydrocarbons were purchased from Aldrich Chemical (Milwaukee, WI). HPLC-grade acetonitrile and methylene chloride were purchased from Fislier Scientífic (Springfield, NJ). 4.2. Reaction rate measurements Reactions were carried out in a I-mi reaction mixture containing 20 uM substrate and 15% acetonitrile in a 60 mM acétate bufíer, pH 3.0, with 20 mM KC1 at room temperature. From 0.4 pmol to 0.2 nmol of the purified enzyme were used in the mixtures. Reactions were star- ted by addition of 1 mM H2O2 and monitored by HPLC. Reaction rates were measured from the differences in peak área after 10 min and referred to the purified pro- tein concentraron for specific activity calculations. Reported valúes are the mean of three replicates. Specific reaction rates are given as mol of substrate converted per mol of enzyme per minute or simply in min"1. Reac- tions for 5 and 6 aromatic rings PAHs 7,12-dimethyIbenz- anthracene, benzo[a]pyrene, 7-methyIbenzo[a]pyrene, benzo[ghi]perylene and perylene were also monitored by fluorescence spectrum, with an excitation at 300 nm, in a Luminescence spectrometer, Perkin-Elmer, Model LS 50, during 2 min at 25 °C. The CPO activity was mea- sured as the disappearance of the respective máximum emission peak for each PAH over a two minute reaction. 4.3. Analytical methods Substrate concentration was measured in a Perkin- Elmer (series 200) HPLC system, using a C|8 Hypersyl 5 um Hewlett-Packard column eluted with an acetoni- trile-water (70:30 v/v) solvent mixture. Substrate and product detection was carried out using a diode array detector coupled to the HPLC system. Product identifi- cation was performed in a Hewlett-Packard GC (model 6890)-MS (model 5972) equipped with a SPB-20 column (30 mx0.25 mm, Supelco). The temperature program started at 100 °C for 2 min; the temperature was raised to 290 °C at a rate of 8 °C/min and kept at 290 °C for 10 min. Acknowledgements This work was supported by The National Council of Science and Technology of México (CONACYT grant 33611-U) and by The Mexican Oil Institute (grant FIES 98-I10-VI). References Aitkcn, M.D., Masscy. I.J., Chen, T , Hcck, P.E., 1994. Characteriza- tion of reaction products from the enzyme catalyzcd oxidation of • phenolie pollutants. Wat. Res. 28, 1879. Ayala, M., Tinoco, R., Hernández, V., Bremauntz, P., Vazquez- Duhalt, R., 1998. Biocatalytic oxidation of fucl as an altcrnative to biodcsulfurization. Fucl Processing Technol. 57, 101-111. Beach, A.C, Harmon, H.J., 1992. Additive cffeets and potential inhi- bitory mechanism of some common aromatic pollutants on in vitro mítochondrial respiration. J. Biochem. Toxicol. 7, 155-161. Bogan, B.W., Lámar, R.T., Hammel, K..E., 1996. Fluorene oxidation in vivo by Phanerochaete chrysosporium and in vitro during man- ganese peroxidasc-dependant lipid peroxidation. Appl. Environ. Microbiol. 62, 1788-1792. Apéndice 7.3 R. Vázquez-Duhalt er al./ Phylochemistry 58 (2001) 929-933 933 Carmichacl, R., Fedorak, P.M., Pickard, M.A., 1983. Oxidation of phcnols by chloroperoxidasc. Biotcchnol. Lctt. 7, 284-294. Chaudhry, G.R., Chapalamadugu, S., 1991. Biodcgradaiion of halo- gcnatcd organic compounds. Microbiol Rcv. 55, 59. -79. Colmsjo, A., Rannug, A., Rannug, U., 1984. Some chloro derivativos of polycyclic aromatic hydrocarbons are potent mmagcns in Sal- monella íyphimurium. Mutat. Res. 135, 21. -29. Colonna, S., Gaggero, N., Richelmi, C , Pasta, P., 1999. Recent bio- tcchnological devclopments in thc use of peroxidases. Trends Bio- tcchnol. 17, 163-168. Colonna, S., Gaggero, N., Manfredi, A-, Casella, L., Gullotti, M., Carrea, G., Pasta, P., 1990. Enantioselective oxidations of sulfidcs catalyzed by chloropcroxidase. Biochemistry 29, 10465. -10468. Dawson, J.H., Sonó, M., 1987. Cytochrome P-450 and chloropcrox- idase: thiolatc-ügand heme enzymes. Spectroscopic detemínation of their active site structures and mechanistic ¡mpltcations of thiolate lígand. Chcm. Rev. 87, 1255-1270. de Jong, E., Ficld, J.A., 1997. Sulfur tuft and turkey tail: biosynthcsis and biodegradation of organohalogens by Basidiomycctcs. Annu. Rcv. Microbio!. 51,375-414. Evangelista de Duffard, A.M., Duífard, R., 1996. Behavioral toxicology, risk assessment, and chlorinated hydrocarbons. Environ Health Perspect. 104 (2), 353-360. Fedorak, P.M., Scmple, K.M., Vazquez-Duhalt, R., Westlakc, D.W.S., 1993. Chloropcroxidase-mediated modifications of petro- porphyrins and asphaltenes. Enzyme Microb. Technol. 15, 429-437. Giesy, J.P., Karman, K., 1998. Dioxin-likc and non-dioxin-like toxic efleets of polychlorinatcd biphenyls (PCBs): implications for risk assessment. Crit. Rcv. Toxico!. 28, 5! 1-569. Gustafsson, E., Brunstrom, B., Nilsson, U., 1994. Lcthality and EROD-inducing potcney of chlorinated chryscnc in chick embryos. Chcmosphcrc 29, 2301-2308. Hammel, K.E.. Kalyanaraman, B., Kirk, T.K., 1986. Oxidation of polycyclic aromatic hydrocarbons and dibenzo(p)díoxins by Pha- nerochaete chrysosporium ligninasc. J. Biol. Chcm. 261, 16948- 16952. Hernández, J., Robledo, N.R., Velasco, L., Quintero, R., Pickard, M.A., Vazqucz-Duhalt, R., 1998. Chloroperoxidase-mediated oxi- dation of organophosphoms pesticides. Pcst. Biochem. Physiol. 61, 87-94. International Agency for Research on Cáncer, 1999. Poly- chlorophenols and their sodium salts. IARC Monographs on the Evaluation of thc Carcinogenic Risks of Chemicals on Humans. 71 (2), 769-816. Johnsen, S,, Gribbestad, I.S., Johansen, S., 1989. Formation of chlorinated PAH: a possibte health hazard from water chlorination. Sci. Total Environ. 81-82, 231-238. Márquez-Rocha, F.J., Pica-Granados. Y., Sandoval-Villasana, A.M., Vázquez-Duhalt, R-, 1997. Determination of genotoxicity using a chtoroperoxidase-mediated model of PAH-DNA adduct formation. Bull. Environ. Contam. Toxicol. 59, 788-795. Neidleman, S.L., 1975. Microbial halogenation. CRC Crit. Rcv. Microbiol. 3, 333-358. Pickard, M.A., Kadima, T.A., Carmichacl, R.D.J., 1991. Chloroper- oxidase a peroxidasc with potential. Ind. Microb. 7, 235-242. Robinson, G.K., 1998. (Bio)remcdiation of polychlorinatcd biphenyls (PCBs): problems, perspectives and solutions. Biochem. Soc. Trans. 26, 686-690. Sundaramoorthy, M., Temer, J., Poulos, T.L., 1995. The crystal stmeture of chloropcroxidase: a heme pcroxidasc-cytochromc P450 functional hybrid. Structure 3, 1367-1377. Tilson, H.A., Kodavanti, P.R., 1998. The neurotoxicity of poly- chlorinated biphenyls- Neurotoxicology 19. 517-525. Tsuncnar, S., Yonemitsu, K., Uchimura, Y., Takacsu, H., Kamisato, M., 1982. A rarc fetal case of wood preservative, mono- chloronaphthalenc (MCN), poisoning. Forcnsic Sci. Int. 20, 173— 178. van Pee, K.H., Keller, S., Wage, T., Wynands, I., Schncrr, H., Zehner, S., 2000. Enzymatic halogenation catalyzed via a catalytic triad and by oxidoreductases. Biol. Chem. 381, 1-5. Vazquez-Duhalt, R., Wcstlake, D.W.S., Fedorak, P.M., 1994. Lignin peroxidasc oxidation of aromatic compounds in systcms containing organic solvents. Appl. Environ. Microbiol. 60, 459-466. Voltenbrock. E.G., Simons, L.H., van Schijndcl, J.W., Barnctt, P., Balzar, M., Dckker, H., van der Linden, C , Wcvcr, R., 1995. Vanadium chloropcroxidascs oceur widcly in naturc. Biochem. Soc. Trans. 23, 267-271. Ward, G.S., Parrish, P.R., Rigby, R.A., 1981. Early life stage toxicity test with a saltwatcr lísh: effeets of eight chemicals on survival growth, and devclopment of sheepshead minnows (Cyprinodon var- iegatus). i. Toxicol. Environ. Health 8, 225-240. Wicgel, J., Wu, Q., 2000, Microbial reductívc dchalogenation of poly- chlorinated biphenyls. FEMS Microbio!. Eco!. 32, 1-15. Yi, X., Mroczko, M., Manjol, K.M., Wang, X., Hager, L.P., 1999. Replaccment of thc proximal heme thiolate ligand in chloroperox- idasc with a histidinc residue. Proc. Nat. Acad. Sci. 96, 12412- 12417. Zaks, A., Dodds, D.R., 1995. Chloroperoxidase asyrnmetric oxida- tions: Substratc specificity and mechanistic study. J. Am. Chem. Soc. 117, 10419-10424. Apéndice 7.4 JOURNAL OF MOLECULAR CATALYSIS 8: ENZYMATIC ELSEVIER Journal of Molecular Catalysis B: Enzymatic 16 (2001) 159-167 www.elsevier.com/locate/molcatb Spectroscopic characterization of a manganese-lignin peroxidase hybrid isozyme produced by Bjerkandera adusta in the absence of manganese: evidence of a protein centred radical by hydrogen peroxide M. Ayala Acevesb, M.C. Baratío3, R. Basosia ? R. Vazquez-Duhaltb, R. Pognia'* a Department of Chemistry, University of Siena, Via A. Moro, 53IÓ0-I Siena, Italy h Instituto de Biotecnología, UNAM, Apartado Postal 510-3, Cuernavaca 62271, Momios, México Received 29 May 2001; received in revised form 31 August2001; accepted 31 August 2001 Abstract Electronic absorption and electrón paramagnetic resonance (EPR) spectra are reported for a novel manganese-lignin peroxidase (MnUP) hybrid isozyme produced by Bjerkandera adusta in the absence of manganese at pH 5. The room temperatura absorption and the low temperature (10 K) EPR spectra indícate that the same coordination and spin states are present at both temperatures: maínly six coordinate high spin containing low percentage six coordínate low spin ferric heme, the latter probably with a bis-imidazoie coordination. A protein centred radical was detected in the presence of an excess of hydrogen peroxide and assumed to be a tryptophanyl radical. The catalytic significance of this site was addressed by specific chemical modification of the tryptophan residues that revealed a marked effect on the specific activity of the enzyme. It is proposed that substrate oxidation might proceed through a iong range-eiectron transfer process. © 2001 Elsevier Science B.V. All rights reserved. Keywords: EPR; Spin state; Heme protein; Peroxidase; Protein radical 1. Introduction The biodegradation of lignin, a major constituent of wood, is far from being fully elucidated. White rot fungi, with a complex enzymatic system, contribute in the recycling of íhis biopolymer in nature. Several Abbreviations: LiP, lignin peroxidase; MnP, manganese peroxi- dase; MnLiP, manganese-lignin peroxidase hybrid; HS, high spin; LS, low spin; 6c-HS, hexacoordinated high spin; 5c-HS, perita- coordinated high spin; NBS, A'-bromosuccinimide; BSA, bovine serum albumin; VA, veratryl alcohol • Corresponding author. Tel.: +39-0577-23-4258; fax: +39-0577-23-4239. E-mail address: pogni@unisi.it (R. Pogni). extracellular enzymes from these ligninolytic fungi are involved in the degradation of lignin polymer. Among these enzymes, lignin peroxidase (LiP) and manganese peroxidase (MnP) from Phanerochaete chrysosporium have been extensively studied [1,2]. The catalytic cycle for both enzymes is similar to that of other heme peroxidases and begins with a two-e!ectron oxidation of the heme prosthetic group by hydrogen peroxide yielding compound I. These two oxidizing equivalents are then sequentially trans- ferred to substrate molecules. One of the oxidizing equivalents of compound I is located on the heme ¡ron atom, leading to the formation of an oxyferryl iron(IV) centre. However, differences have been noted with 1381-1177/01/$ - see front matter © 2001 Elsevier Science B.V. All rights reserved. PII: S1381-1 I77(OI)OOO56-X Apéndice 7.4 160 M. Ayala Aceves el al./Journal of Molecular Catalysis B: Enzymaüc 16 (2001) 159-167 respect ío the second one. In all, peroxidases charac- terized up to date, the UV-VIS and resonance Raman (RR) spectroscopy have provided evidence that the second oxidizing equivalent is in fact a porphyrin n radical catión [3] excepí in cytochrome c peroxidase (CcP). In this case the radical resides on a trypto- phan residue in the vjcinity of the proximal histidine [4]. Although the exact mechanism of lignin degrada- tion is not fully understood, it has been proposed that in vivo the process by which LiP and MnP attack the lignin polymer is different. LiP directly interacts with its substrates to form radical cations whereas MnP oxidizes Mn(II) to Mn(III), which in turn acts as a diffusable oxidizing intermedíate [5]. Recently, two novel fungal peroxidases belonging to class II of the piant peroxidase superfamily [6] were detected in the white rot fungi Bjerkandera adusta and Pleurotus eryngü [7,8]. Both enzymes are able to catalyze manganese-dependent as well as manganese-independent reactions. The manganese- dependent activity proceeds optimally at pH 5, while the manganese-independent reactions require more acidic conditions, showing máximum rates at pH 3. The catalytic intermedíales seem to be the same in these enzymes, as they oxidize Mn(II) and aromatic substrates with comparable valúes of Vmax and Km [8]. The biotechnological interest for the manganese- lignin peroxidase hybrid (MnLiP) from B. adusta lies in its ability to oxidize large substrate molecules in the absence of Mn(II). Camarero et al. [9] suggested that MnLiP might be a structural hybrid, based on a comparison of a molecular model of the MnLiP structure from P eryn- gü with the known structures of LiP and MnP from P. chrysosporiwn. This hybrid possesses a MnP-like Mn(II) binding site as well as a long range-electron transfer mechanism similar to that recently found for LiP [10]. Therefore, depending on the reaction con- ditions this enzyme would oxidize substrates unable to reach the heme pocket thxough the mediation of a diffusable, highly reactive intermedíate or through an activated surface amino acid residue in the case of less accessible substrates [II] . The present results show that MnLiP from B. adusta exhibits a prevalent 6c-high spin ferric form with a low percentage of 6c-Iow spin probably due to a bis-imidazole co-ordination at both room and low temperature. A protein centred radical was detected at room temperature in the presence of hy- drogen peroxide, supporting the hypothesis of a long range-electron transfer mechanism for the oxidation of large molecules. 2. Experimental Lignin peroxidase (LiP) and manganese peroxidase (MnP) from P. chrysosporium were obtained from Tienzyme, Inc. (Sait Lake City, UT). The partially purifíed hybrid peroxidase (MnLiP) from B. adusta UAMH 7308 (Rz (A4o3/A28o) = 1.3) was kindly provided by Prof. M.A. Pickard, from the University of Alberta, Canadá. Protein concentration was deter- mined with BioRad reagent using BSA as standard. The molecular weight of the protein was estimated from a Coomasie-stained 10% SDS-PAGE gel and it is equal to 45kDa. The single band in the gel shows that there is no contamination from other proteins in the preparation for this work (data not shown). 2-Methyl-2-nitroso propane (MNP) was purchased from Sigma (St. Louis, MO) and H2O2 from Fluka (Germany). Ail reagents were used without further purification. EPR and UV-VIS measurements were performed on the native enzyrne and the enzyme modified with N-bromo succinimide (NBS) [12]. EPR solutions of the native enzyme were prepared with a final con- centration of 0.21 mM MnLiP and 2mM H2O2 at pH 5. The chernical modification of Trp residues was done using NBS recrystallized from hot water just before use. To follow the effect on activity of the chemical modification of Trp, 0.2 mi of a 0.24 mM MnLiP solution were mixed with different molar ex- cesses of NBS dissolved in water. After 2min, the reaction was stopped by adding a 600-fold molar ex- cess of Trp in water. To follow the decrease of Trp fíuorescence at 315 nm, 2 mi of a 0.0015 mM Mn- LiP solution were mixed with increasing amounts of NBS dissolved in water, and after 1 mín the emission spectra were recorded. For the EPR measurements, 1 mi of a 0.17mM MnLiP solution was mixed with a 60-fold molar excess of NBS dissolved in water. The reaction was allowed to proceed for 2min and stopped by adding 0.1 g tryptophan dissolved in wa- ter. The reaction mixtures were extensively washed Apéndice 7.4 M. Ayaia Aceves e¡ al. /Journal of Molecular Catalysis B: Enzymaüc 16 {2001) 159-167 161 with lOOmM acétate buffer pH 4.5 in an ultrafiltra- tion celi using a lOkDa cutoff membrane, in ortíer to remove substances that could interfere with the EPR and enzymatic assays. Absorption spectra of a 0.005 mM solution of MnP, LiP and MnLiP were measured in 60 mM phosphate buffer pH 5. The Mn(II)-independent activity at differ- ent pH was estimated using hydroquinone as substrate. Reactions were performed in 60 mM phosphate buffer, pH 3 or 5, in the presence of 1 mM of hydroquinone and 0.1 mM H2O2. The reaction was started by adding H2O2 and followed by measuring the changes in ab- sorbance at 247nm (e = 21mM"1cm~ I). LiP and MnP activity were measured as reported elsewhere [13,14]. For LiP acíivity, the reaction mixture con- tained 40 mM succinate buffer pH 4, 4mM veratryl alcohol (VA) and 0,4 mM hydrogen peroxide. The pro- duction of veratryl aldehyde was followed at 310 nm (e = 9.3 mM"1 cm"1). For MnP activity, the reaction consisted of 50 mM malonate buffer pH 4.5, 0.1 mM manganese sulfate and O.imM hydrogen peroxide. The production of the malonaíe-Mn3+ complex was monitored at 270nm (é = ll.59m.M~1 cm"1). For all reactions, the enzyme concentration was in the pH range. UV-VIS measurements were performed on a Hewlett-Packard 8453 spectrophotometer. Fluores- cence measurements were performed on a Perkin Elmer LS 50 Lumínescence spectrometer, with an excitaíion wavelength of 285 nm, excitation slit = 8nm and emission slit = 5nm. EPR spectra were recorded on a Bruker 200D SRC instrument equipped with a microwave frequency counter XL (Jagmar, Krakow, Poland). The spectrometer was interfaced with a PS/2 technical instrument hardware com- puter and the data acquired using the EPR data system CS-EPR produced by Stelar Inc. (Mede, Italy). An Oxford instrument ESR 900 cryostat was used to obtain low temperatures. The spectra were recorded under non-saturating conditions at 10K, 9.466GHz microwave frequency, lOmW microwave power and 1 mT modulation amplitude. Spectra were also recorded at 20 K and 80 mW. Al! samples were frozen rapidly by immersion of the EPR tube into liquid nitrogen. The spectrum at 120K was recorded using an ER4111VT Bruker variable temperature unit, 9.565 GHz microwave frequency and 63 mW microwave power. 3. Resulte Fig. 1 shows the electronic absorption spectrum at room temperature and pH 5 of the hybrid MnLiP from B. adusta compared with those of LiP and MnP from P. chrysosporium. The spectrum shows a Soret band at 409 nm, a. and p bands at 579 (shoulder) and 541 nm, and two charge transfer bands at 501 nm (CT2) and 632nm(CTl). All the electronic absorption spectra reported in Fig. 1 show a CT1 band at 632 nm. The wavelength of the band in the 600-650 nm región is sensitive to the heme pocket environment [15]. This band, observed only in the high spin (HS) heme proteins, was as- signed to a charge transfer transition (CT1) from the porphyrin to the ¡ron [a^iir) ~* eg(d^)J- For the hexacoordinate (6c) proteins containing an imidazole as the fifth ligand, the CT1 ranges from 600-637 nm [16]. It has been suggested [15,17-19] that the room temperature electronic absorption spectra of MnP and LiP indícate that the heme iron is predominantly in the hexacoordinated HS state (6c-HS) with a water molecule bound at the sixth coordination position of the Fe atom. The cióse similarity of the absorption spectrum from MnLiP with those from MnP and LiP suggests that the major heme state of MnLiP is also 6c-HS. Furthermore, the presence of a low percentage of 6c-LS heme in MnLiP of bis-imidazole coordina- tion is revealed by the pronounced 3 band at 541 nm and the a band at 579 nm, the latter being absent in the spectra of LiP and MnP. Fig. 2a shows the X-band EPR spectrum of MnLiP al pH 5, recorded under non-saturating conditions (temperature 10K, microwave power lOmW). The spectrum is characterized by two distinct signáis which indicate the coexistence in MnLiP of a HS species (g = 6.00, 2.00) and a small amount of LS species (g = 3.15, 2.07) at liquid helíum temperature. The third valué of the latter species is too weak to be observed. The feature at g = 4.3 corresponds to a non-heme iron impurity often seen in protein samples [20]. The LS signal was optimized at a temperature of 20 K and a microwave power of 80 mW as shown in Fig. 2b. Particular attention should be given to íhe signa! at g — 3.15. Many model compounds and heme proteins which have bis-imidazole coordination have g valúes cióse to 3.0 or slightly lower [21,22]. /ZZ Apéndice 7.4 162 M. Ayala Aceves et al./Journal of Molecular Catalysis B; Enzymatic 16 (2001) 159-167 409 a) 380 400 420 440 460 475 500 525 550 575 600 625 650 675 700 Wavelenght (nm) Fíg. 1. Electronic absorption spectra of ferric (a) MnLiP, (b) LiP, (c) MnP obtained at room temperature in 60mM phosphate buffer solution at pH 5. The región between 475-700nm has been expanded. Furthermore, a wide variety of membrane bound cy- tochromes b, which have bis-imidazole coordínation, have g > 3.0. These large g valúes have been shown to result from a non-parallel orientation of the imidazole ligands [23-25]. The LS signal at g = 3-15 observed ín our case supports the assignment of the LS species noted in the absorption spectra to a bis-imidazole complex from an endogenous ligand. Thus, the low temperature EPR and absorption measurements indi- cate that MnLiP has a similar heme environment at both temperatures, in which HS and LS species coex- ist in similar proportions. Consequently, it seems that the protein does not undergo a conformational change upon lowering the temperature, which would iead to a significant modífication of the heme spin state proportions. The addition of H2O2 to the enzyme solution shows the presence of an enzyme intermediate at both EPR iiquid helíum temperature and UV-VIS room temperature región (data not shown). The UV-VIS spectrum of this intermedíate is similar to those reported for compound I from other peroxidases [26,27]. The signal of this intermediate is stable for at least 4min, after which the spectrum spontaneously begins to slowly revert to that of the native enzyme. The room temperature EPR spectrum of a protein centred radical with a giS0 =. 2.005 is reported ín Fig. 3a. Although the signal is clearly evident, it is characterízed by a very rapid decrease in intensity, indicative of a short lifetime species. The spectrum reveáis superhyperfine structure that becomes more evident in the second derivative display, as evidenced /ZJ Apéndice 7.4 M. Ayaía Aceves el al. /Journal of Molecular Catalysis B: Enzytnatic 16 (2001) ¡59-167 6.O0 163 a) 4.3 3.15 60 100 140 3.15 180 220 260 300 340 380 420 Magnetic Field / mT 200 240 280 320 360 400 440 480 Magnetic Field / mT Fig. 2. X-band EPR spectraof the fenic iron MnLiP inóOmM phosphate buffer solution at pH 5 (a) recorded at 10K wíth I mT modulation amplitude, lOmW power and v = 9.466; (b) recorded at 20K with I mT modulation amplitude, 80mW power and i> = 9.464. previously by others [28]. In Fig. 3b the same radical at nitrogen temperatura ¡s reported. The reaction was performed in the absence of Mn(II) ions ai acidic pH. To investígate the role of a Trp residue in the protein centred radical, a chemical modification of the Trp residues of MnLiP was performed. Table I shows the specific activity of MnLiP with hydroquinone at two Apéndice 7.4 164 M. Ayala Aceves el al./Journal of Molecular Caialysis B: Enzymatic 16 (2001) 159-167 a) 2.009 336 338 340 342 344 Magnetic Fiek) / mT Fig. 3. X-band EPR spectra of the protein centred radical of MnLiP in óOmM buffer solution at pH 5 recorded at (a) 298 K, v = 9.5716 and (b) 120K, v =9.5652. The spectra were recorded with 0.02mT modulation amplitude and 63mW power. Table 1 Enzymatic activity of the hybrid peroxidase from B. adusta Substrate Specific activity (min ') Hydroquinone, pH 3 Hydroquinone, pH 5 Veratryl alcohol Manganese(II) 16.4 (±0.8) 4.5 (±0.2) 263 (±13) 2847 (±142) different pH. As expected, the Mn(II)-independent activity is higher at low pH valúes. The MnP-Hke ac- tivity and the LiP-like activity are reported as well. Al- though the enzyme catalyzes both Mn(II)-dependent and Mn(II)-independent reactíons, it is more efficient when Mn(II) is used as substrate. Fig. 4 shows the changes in fíuorescence at 315 nm and the decrease in Apéndice 7.4 M. Ayala Aceves et al./Journal of Molecular Catalysis B: Enzymctlic 16 (2001) ¡59-167 120 165 Fluorescence at 315 nm MnP-Iike acíivity LiP-like activity 20 40 60 80 mol NBS/mol MnLiP 100 Fig. 4. Effect of the enzymatic activity and fluorescence reduciion of the NBS modified Trp residues in MnLíP. enzymatic activity as the molar excess of NBS over MnLiP increases. Although the chemicaf modificaron of Trp lowers both the MnP-like and the LiP-like ac- tivities, a more pronounced effect is observed on the Mn(II)-independent activity. This indirect evidence also supports the hypothesis of a tryptophanyl radical involved in the cataiytic cycle of this enzyme. 4. Discussion The crystal structures and the available sequence alignments show that the heme active sites of perox- idases share many common features. The conserved functional residues on the dista! (His, Arg, Asn) and proximal (His, Asp) sides are linked by a hydrogen bond network, mediated by polar residues and water molecules, connecting the proximal and the distal sides of the heme. Nevertheless, small structural vari- ations from one type of peroxidase to another which modulate the strength of the axial histidine hydrogen bond could easily account for the considerable range of midpoint potentiais found for this class of enzymes [29]. Normally the heme iron is pentacoordinate at room temperature, an exception being LiP and its var- íous isozymes which have a water molecule located cióse to the heme iron, giving rise tb a 6c-HS heme [15,18,19]. The percentage of a 6c-LS form in MnLiP distin- guishes this enzyme wíth respect to LiP and MnP. The presence of a 6c-LS species is confirmed by EPR at liquid helium temperature (Fig. 2a and b), thus revealing the same coordinaron at both temper- atures. The signal at g = 3.15 indicates that this LS species might be assigned to a bis-imidazole complex. Such large g valúes have been shown to aríse from a non-parallel orientation of the imidazole ligands [23,30}. Crystal field and thermodynamic analysis, based upon reasonable estimates of the tetragonal and rhombic splñtings of the d orbitals of low spin Fe(III) porphyrins, suggest that perpendicular alignment of planar axial ligands could lead to a positive shift in redox potential of about 50 mV over that observed for parallel alignment, all other structural and environ- mental factors being equal [30]. The ability of MnLiP to oxidize large substrates prompted the hypothesis that different sites in the pro- tein were involved in the oxidation of the substrate. In a recent work by Doyle et al. [10] two distinct substrate interaction sites in LiP were postulated. One was a heme-edge site typical of those encountered in other peroxidases. The second one was a novel site centred around Trp 171 which is required for the oxidation of VA even if, until now, this latter protein radical has not been directly detected. Considering the structural homology of MnLiP from Pleurotus Apéndice 7.4 166 M. Ayala Aceves el al./Journal of Molecular Catalysis B: Enzymatic 16 (2001) 159-167 eryngü with LiP and MnP from P. chrysosporium for which crystal structure data are available [18,19,31], the substrate access channel and heme appear rel- atively inaccessible suggesting that the classical heme-edge site may not be the only site for substrate interaction [9]. The presence of a tryptophan residue exposed to the solvent and in cióse proximity to the heme might be a probable site for the oxidation of substrates. A long-range electrón transfer pathway in the pres- ence of H2O2 was envisaged based on the structural homology with LiP, where Trp 171 participates in the oxidation of VA [8]. Our results show that at room temperature a radical is formed (g — 2.005) in the presence of excess hydrogen peroxide (Fig. 3a). The signal intensity decreases very rapidly, indicative of a short lifetime. The spectrum has been obtained at room temperature performíng the reaction with an excess of hydrogen peroxide, even if the formation of the radical is evident also for an equivalen! quantity of hydrogen peroxide with respect to the protein. Our attempts to highlight the protein radical adduct with the 2-methy!-2-nitroso propane (MNP) spin trap were prevented [11,29,32]. The protein radical might be a Trp with a similar role to that of Trp 171 in LiP. The C3 of the indolyl radical catión is generaliy believed to be the prime target of a spin trap like MNP or dioxygen, since this is the site of the highest spin density in tryptophan radicáis. In LiP the positions of the pyrrole moiety of the Trp 171 Índole ring are hardly accessible by the solvent in contrast with the benzene ring that becomes more accessible site giving an adduct that promptly becomes diamagnetic [11]. The structural model of MnLiP from P eryngü was examined and two Trp residues (Trp 250 and Trp 170) were identified. Both tryptophans are on the proximal side of the protein, the Trp 250 is buried inside while the Trp 170 is solvent exposed via the benzene ring, in cióse proximity of the heme and Ser 174 that is bound to the proximal His 175. To probé the localization of the radical on MnLiP tryptophan residues a chemical modification of the peroxidase from B. adusta was undertaken using the tryptophan-specifie agent NBS. For LiP, it was reported [12] that the activity decreased when from 10- to 25-fold molar excess of NBS was added to reach a constant valué of ~15% residual activity at "-30-fold excess of reagent. The change in activity was accompanied by a reduction of tryptophan fluores- cence because oxindoles are non-fluorescent. The loss of about one-third of the initial fluorescence intensity was achieved with five equivalents of NBS while the activity was slightly affected. This could be attributed to the modification of a surface tryptophan residue (Trp 170). The second stage of the modification re- quired higher amounts of NBS, and resulted in the loss of most of the activity and roughly another third of the initial fluorescence. The remaining emission was attributed to a tryptophan buried inside the protein (Trp 250) [12], With MnLiP frorn B. adusta the fiuorescence at 315 nm diminished as higher amounts of NBS were added (Fig. 4). This is indicative of chemical mod- ification of the solvent exposed Trp residues in the protein. The enzymatic activity decreases when a 60-fold molar excess of NBS reaets with the pro- tein. The Mn(II)-independent activity is drastically reduced and consequently the enzyme retains only 15% of its original activity. On the other hand, the Mn(II)-dependent activity is reduced to 55% of its original valué. No EPR radical signáis were detected with this chemically modified sample, showing that solvent exposed tryptophan residues are important for the formation of the protein radical. These findings strongly suggest a mechanism in which a protein radical is formed on a Trp residue of MnLiP in the presence of H2O2 and in the absence of Mn(II). When Trp residues are chemically oxidized, no electrons can be extracted from the Trp residue and thus the radical cannot be formed. Consequently, the oxidation of the substrate at this site cannot be accomplished. Nevertheless, the oxidation pathway at the Mn(II)-binding site is still available. The EPR low temperature (120K) spectrum of the MnLiP radical (Fig. 3b) differs from that at room temperature (Fig. 3a) displaying a slightly axial symmetry with an apparent g\\ > gj_, which is similar to that obtained for the Trp 191 free radical species of wild-type CcP [28]. The radical signal is difficult to satúrate even at 10K, which is again similar to the behaviour of the Trp 191 radical of wild-type CcP, and thus it is presumably coupled to the S ~ 1 oxyferryl heme iron centre [33]. In conclusión, the UV-VIS absorption and EPR spectra show the presence of a percentage of 6c-LS species for MnLiP that might result from a /Z? Apéndice 7.4 M. Ayala Áceves el al./Journal of Molecular Catalysis B: Enzymatic 16 (2001) ¡59-167 167 bis-imidazole complex, very likely due to the distal histidine becoming bound to the sixth coordination site of the iron atom. MnLiP from B. adusta described here is truly a hybrid peroxidase, as it possesses two oxidation pathways. One pathway resembles that of MnP and it proceeds through the oxidation of Mn(H) to the highly reactive Mn(III). A second different pathway resembles that of LiP and it involves the formation of a solvent exposed protein radical. To our knowledge, it is the first time that a protein centred radical has been directiy detected by EPR in the pres- ence of hydrogen peroxide at room temperature for a ligninolytic peroxidase, demonstrating the key role played by this site in the oxidation of substrates by a long range-electron transfer. The importance of Trp residues in both peroxidase activity and protein radi- cal formation has been demonstrated by the specific chemical modification of Trp residues, even if further characterization of the free radical site is under way. This is the fourth enzyme, together with CcP, DNA photolyase [34] and LiP [11], in a group of unique enzymes that display an active Trp required for the transformation of natural substrates. Acknowledgements Authors gratefully acknowledge Prof. M.A. Pickard, for generous supply of MnLiP by Bjerkan- dera Adusta and Dr. B.D. Howes for technical assis- tance and useful discussion. This work was funded by Cofín MURST CFSIB 97 and C.N.R. "Biotecnolo- gie e Biología Molecolare" n. 98.01066.CT14, and by the Mexican Council of Sciences and Technology (Conacyt33611-U). References [I] M. Tien, T.K. Kirk. Science 221 (1983) 661. [2] M. Kuwahara. J.K. Glenn, M.A. Morgan, M.H. Gold, FEBS Lett. 169 (1984) 247. [3] A. Khindaria, S.D. Aust, Biochemistry 35 (1996) 13107. [4] M. Sivaraja, D.B. Goodin, M. Smith, B.M. Hoffman, Science 245 (1989) 738. [5] L. Banci, S. Cioffi-Baffoni, M. Tien, Biochemistry 38 (1999) 3205. [6] K.G. Welinder, Curr. Opin. Stnict. Biol. 2 (1992) 388. [7] T. Mesler, J.A. Field. J. Biol. Chem. 273 (1998) 15412. [8] A. Heinfiing, F,J. Ruiz-Dueñas, M.J. Martínez, M. Bergbauer, U. Szewzyk, A.T. Martínez. FEBS Lett. 428 (1998) 141. [9] S. Camarero, S. Sakar, F.J. Ruiz-Duenas, M.J. Martínez. A.T. Martínez, J. Bioi. Chem. 274 (1999) 10324. [10] W.A. Doyle, B. Blodig, N.C. Veitch, K. Piontek, A. Smith, Biochemistry 37 (1998) 15097. [II] W. Blodig, A.T. Smith, K. Winterhalter, K. Piontek, Arch. Biochem. Biophys. 370 (1999) 86. [12] W. Blodig, W.A. Doyle, A.T. Smith, K. Winterhalter. T. Choinowski, K. Piontek, Biochemistry 37 (1998) 8832. [13] M. Tien, K. Kirk. Meth. Enzymol. 161 (1988) 238. [14] H. Wariishi, K. Valhi, M. Gold. J. Biol. Chem. 267 (1992) 23688. [15] G. Smulevich, Biospectroscopy 4 (1998) S3. [16] G. Smulevich, F. Neri, M.P. Marzocchi, K.G. Welinder. Biochemístry 35 (1996) 10576. [17] J.K. Glenn, M.H. Gold, Arch. Biochem. Biophys. 242 (1985) 329. [18] T.L. Poulos, S.L. Edwards, H. Wariishi, M.H. Goid. J. Biol. Chem. 268(1993) 4429. [19] K. Piontek. T. Glumoff, K. Winterhalter, FEBS Lett. 315 (1993) 119. [20] W.E. Blumberg. J. Peisach. B.A. Wittenberg. J.B. Wiítenberg. J. Biol. Chem. 243 (1968) 1854. [21] W.E. Blumberg, J. Peisach. in: C.B.T. Yonetani, A.S. Mitdvan (Eds.), Probes of Structure and Function of Macromolecules and Membranes. Vol. 2, Academic Press, New York. 1971, p. 215. [22] B.D. Howes, A. Feis, C. Indiani, M.P. Marzocchi. G. Smulevich. J. Biol. Inorg. Chem. 5 (2000) 227. [23] F.A. Waiker, B.H. Huynh, W.R. Scheidt, S.R. Osvath, J. Ara. Chem. Soc. 108 (1986)5288. [24] F.A. Waiker, Coord. Chem. Rev. 185/186 (1999) 471. [25] C. Indiani, A. Feis, B.D. Howes, M.P. Marzocchi, G. Smulevich, J. Am. Chem. Soc. 112 (2000) 7368. [26] A. Tuynman, M.K.S. Vink, H.L. Dekker, H.E. Schoemaker, R. Wever, Eur. J. Biochem. 258 (1998) 906. [27] W.D. Hewson,' L.P. Hager, J. Biol. Chem. 254 (1979) 3182. [28] H. Hori, T. Yonetani, J. Biol. Chem. 260 (1) (1985) 349. [29] D.B. Goodin, D.E. McRee, Biochemistry 32 (1993) 3313. [30] G.T. Babcock, W.R. Widger, W.A. Cramer, W.A. Oertling, J.G. Metz, Biochemistry 24 (1985) 3638. [31] M. Sundaramoorthy, K. Kishi. M.H. Gold, T.L. Poulos, J. Biol. Chem. 269 (Í994) 32759. [32] R. Pogni, G. Della Lunga, E. Busi, R. Basosi, Int. J. Quant. Chem. 73 (1999) 249. [33] J.E. Huyett, PE. Doan, R. Gurbiel, A.L.P. Houseman, M. Sivaraja, D.B. Goodin, B.M. Hoffman, J. Am. Chem. Soc. 117 (1995) 9033. [34] J. Stubbe, W.A. Van der Donk, Chem. Rev. 98 1998) 705. /Zf Chemistry & Biology, Vol. 9, 555-565, May, 2002, ©2002 Elsevier Science Ltd. Al! rights reserved. Pll S107 -7 Suicide Inactivation of Peroxidases and the Challenge of Engineering More Robust Enzymes Review Brenda Valderrama,1 Marcela Ayala, and Rafael Vázquez-Duhalt Instituto de Biotecnología Universidad Nacional Autónoma de México AP 510-3 Cuernavaca Morelos 62250 México As the number of industrial applications for proteins continúes to expand, the exploitation of protein engi- neering becomes critical. tt is predicted that protein engineering can genérate enzymes with new catalytic properties and créate desirable, high-value, products at lower production costs. Peroxidases are ubtquitous enzymes that catalyze a variety of oxygen-transfer re- actions and are thus potentially useful for industrial and biomedical applications. However, peroxidases are unstable and are readily inactivated by their sub- strate, hydrogen peroxtde. Researchers rely on the powerful tools of molecular biology to improve the stability of these enzymes, either by protecting resi- dues sensitive to oxidation or by devising more effi- cíent intramolecular pathways for free-radical alloca- tion. Here, we discuss the catalytic cycle of peroxidases and the mechanism of the suicide inacti- vation process to establish a broad knowledge base for future rational protein engineering. Introduction Redox reactions require a small but steady supply of oxidative species. In particular, physiological amounts of hydrogen peroxide, and of other reactive oxygen spe- cies, are generated during normal cefíular activity either as a toxic by-product of respiration {1, 2] or from other oxidative reactions [3, 4]. Oxidative species attack a variety of cedular constituents, but proteins are welí known to be especially sensitive, with the most suscepti- ble amino acids being methtonine, cysteine, tryptophan, tyrosine, and histidine [5, 6], afthough damage ío other residues has also been observed [6,7]. Protein oxidation is considered a cause or contributory factor to many diseases, such as Refsum's disease and Alzheimer's disease, and has also been related to the aging process [8-12]. Besides the global eff ects of oxidative damage to proteins, as evidenced by the overalf tncrease of protein carbonyl content [13], certain enzymes nave been found to be specificalfy inactivated by the irreversible oxida- tion of catalytically important residues [14-18]. For ex- ample, cysteine residues, once oxidized, disrupt the overall protein structure and facilítate further damage [19]. Additionally, protein glutathione adducts can form through cysteine residues [20]. In particular, the revers- ible oxidation of methionine residues has been sug- gested to play an important role in the regulation of 'Correspondence: brenda@ibt.unam.mx various cellular processes [21,22]. Although protein oxi- dation is generally independent of the catalytic activity of any given protein, the oxidative inactivation of heme- peroxidases is mechanism based. Here we present an integrative review of the literature related to the oxidative inactivation of different groups of peroxidasic hemepro- teins with emphasis upon the general nature of this phe- nomenon and the proposal of a consensus mechanism. Peroxidases can be divided into three classes that are defined according to their specific active center: hemeperoxidases are a subset of the hemeproteins, which contain the prosthetic group iron porphyrin and control a wide variety of fundamental biological pro- cesses in most living organisms [23]; vanadium peroxi- dases contain a vanadate ion at their active site and are most commonly found in marine environments [24, 25]; and finally, non-metal peroxidases require acétate or propionate buffer for activity and are found in bacteria [26]. The crystal structures of twelve hemeperoxidases and seven cytochrome P450 hemeproteins have been re- ported to date {see Table 1 and references therein). In each of these structures, a coordinated heme prosthetic group is present in the form of iron-protoporphyrin IX, generally coordinated by a histidine residue as the proxi- mal ligand, with the exception of chloroperoxidase from Caldariomyces fumago and cytochrome P450 from ai! species examined to date, in which the proximal ligand is a cysteine residue [27], The hemeproteins in Table 1 are distributed within five different folding groups ac- cording to structure-structure alignments [28]. The chlo- roperoxidase from Caldariomyces fumago is the only member of its group. Fo!d similarity is highly significant among members of the same group, despite the low identity of their amino acid sequences, which can be as little as 20% idéntica). The classical reaction cataiyzed by hemeperoxidases is oxidative dehydrogenation, although they also cata- lyze a variety of related reactions, inctuding oxygen transfer, hydrogen peroxide cleavage, and peroxidative halogenations. These reactions are described in more detaii in the following paragraphs. Oxidative dehydrogenation involves one-electron transfer processes between an oxo-iron(IV)porphyrin- based ir-free radical (pathway 2 in Figure 1) or an oxo- iron(IV)porphyrin {pathway 3 in Figure 1) and a diversity of organic and inorganic substrates, with hydrogen per- oxide, organic hydroperoxides, peracids, or inorganic oxides, such as periodate and chlorite, as electrón do- nors. An example of this reaction is the spontaneous polymerization of. phenol and aniline free radicáis [29]. 2 RH + H2O2 — 2 R• + 2H2O - R-R Oxygen transfer is, from a synthesis point of view, the most interesting oxidative transformation cataiyzed by peroxidases. This oxidation is comparable to those per- formed by monooxygenases, such as cytochrome P450. Such oxidations include hetero-atom oxidation (S-oxi- /Zf Chemistry & Biology 556 Apéndice 7.5 Table 1. Peroxidases and Cytochromes P450 wíth Available Crystal Structures Grouped by Fold Similarity Folding Group Hemeprotein So urce Plant peroxidases Soybean Arabidopsis Baiiey grain Horseradish Peanut Arabidopsis Manganese and lignin peroxidases Phanerochaete chrysosporíum Arthromyces ramosus Phanerochaete chrysosporíum Cytochrome c peroxidases Yeast Pisum sativum Chloroperoxidase Caldariomyces fumago Cytochromes P450 Mammalian microsomal Bacilius megateríum Pseudomonas sp. Fusaríum oxysporium Mycobacterium tuberculosis Sulfolobus sulfactarícus Pseudomonas putida Function peroxidase peroxidase A2 peroxidase peroxidase peroxidase peroxidase N Mn -peroxidase peroxidase lignin-peroxidase cytochrome c peroxidase ascorbate peroxidase chloroperoxidase cytochrome P450 monooxygenase cytochrome P450 cytochrome P450 nitríc oxide reducíase «-sterol deinethylase cytochrome P450 cytochrome P450cam PDB Account Number 1FHF 1PA2 1BGP 1ATJ 1SCH 1OGJ 1MNP 1ARV 1LLP 1RYC 1APX 1CPO 1DT6 1BU7 1CPT 1R0M 1EA1 1F4T 1PHD Reference [153] [154] [155] [156] 1157] [158] [159] [160] [130] [1611 [162] [163] [164] [165] [166] [167] 1168] [169] [170] dation and N-oxidation), epoxidation, and C-H bond oxi- dation (pathways 9 and 10 in Figure 1) [30]. RH + H2O2 -+ ROH +• H2O Hydrogen peroxide decomposition is achieved through the heterolytic cleavage of H2O2 to form water (pathways 1 and 8 in Figure 1). In particular, chloroperoxidase ex- hibits a substantial catalase activity when hydrogen or organic peroxides are the only reductants present in the reaction mixture [31, 32]. In recent years, a new class of peroxidase-related enzymes, the catalase-peroxidase group, has been discovered. In addition to catalase ac- tivity, they exhibit a significant classical-peroxidase ac- tivity [33-36]. Compound X ,0-CI RH + H* - * RCI + H,0 Ground state Figure 1. Summary Mechanistic Cycle of Peroxidases Review 557 Apéndice 7.5 2 ROOH — 2 ROH + Oa Finally, peroxidative halogenation is catalyzed by a spe- cial class of peroxidases callee! haloperoxidases, which medíate the haíogenation of organic substrates {path- ways 4-7 in Figure 1). Peroxidative halogenation is not limited to hemeperoxídases but is also catalyzed by vanadium haloperoxidases and other non-heme en- zymes [24-26, 37]. Chloroperoxidase from Caldario- myces fumago is the most active heme-containing ha- loperoxidase [31]. RH + H,O, H RX + 2 H,0 Non-enzymatic hemeproteins are also able to catalyze peroxidase-like reactions. Hemoglobin [38,39], myoglo- bin [40], cytochrome c [41], and microperoxidase, a heme-bound octapeptide derived from the enzymatic proteolysis of cytochrome c [42], are able to oxidize organtc substrates. Hydrogen peroxide enables the ¡ron atoms of these proteins, whether high- or low-spin and whether hexa- or penta-coordinated, to perform one- electron oxidations, most likely by the peroxidase eyele. The development of ecoefficient technological inno- vations witl be critical in this new century and will propel us toward sustainable industrial practices [43]. Green chemistry seeks to develop and deploy chemica! prod- ucís and processes that reduce or elimínate the use and generation of hazardous substances [44]. Valid con- cerns about the effeets of current practices on the envi- ronment and energy sources are the driving forcé behind these unavoidable and necessary changes. Biotechnol- ogy witl, without a doubt, play an important role in this transformation. The industrial application of enzymes is growing, and the industrial processes in which they are involved are considered clean and low in energy de- mand. Peroxidases have potentially interesting applica- tions in diverse fields. The most important application so far is in the analytical diagnostte field, where peroxi- dases are used as a key component of biosensors and immunoassays [45-47]. Peroxidases are currently ex- tensively studied for their use in industrial processes such as Kraft puip bleaching [48-50], in which they can substitute for the large amounts of chlorine that are currently used and thus prevent the formation of toxic halogenated compounds during the process. Further, these enzymes are involved in the degradation of aro- matic compounds and other xenobiotics, including pesticides, polycycllc aromatic hydrocarbons, and diox- íns [51] and thus can be developed for the removal of phenolic and aromatic poHutants [52, 53], as antioxi- dants [54], as indicators for food processing [47], in bioelectrodes [55}, in the production of pharmaceuticals [56], and in the synthesis of conducting plastics [29]. In addition,. peroxidases could also be used in the synthe- sis of fine chemicals and opticalty and biologically active compounds [30, 57]. Despite the obvious valué of peroxidases, their pres- ent commercial uses are limited, primarily by the low stability of peroxidases in the presence of their natural substrate, hydrogen peroxide. All hemeproteins, includ- ing peroxidases, are inactivated in the presence of cata- lytic concentrations of hydrogen peroxide. This process, which can be described as a suicide inactivation, is especially important in the absence of reducing sub- strates, and its mechanism has not been fulfy elucidated. Suicide Inactivation of Peroxidases Classical Peroxidases Extensive investigatíons intothe mechanism of function of classical peroxidases resulted in a consensus cata- lytic network (Figure 1) that proceeds via the establish- ment of a sixth-coordination bond between hydrogen peroxide and the heme iron and yields Compound I, a high-valent oxo-iron{IV)porphyrin-based -rr-free radical (pathway 1 in Figure 1). Electron paramagnetic reso- nance (EPR) studies established that the second oxida- tion equivatent in Compound I is initially present as a porphyrtn-based free radical, but in some cases, elec- trón abstraction from the protein results in formation of a second Compound I species with an unpaired electrón based in a residue cióse to the porphyrin [58, 59]. The presence of 0.5 equivalents of a two electron-reducing agent, such as an aromatic compound, generates Com- pound II, which ¡s an oxo-iron(IV)porphyrín without the associated porphyrin ir-free radical (pathway 2 in Figure 1). Compound II oxidizes a second moleculeof substrate via the peroxidase shorteut to form the resting-state íron(l!l)porphyrin (pathway 3 in Figure 1). Classical per- oxidases are irreversibty inactivated by exposure to high concentrations of hydrogen peroxide [60-63]. Ligninolytic microorganisms secrete two extracellular types of peroxidases, lignin peroxidase [64-67] and manganese peroxidase [68]. Although these enzymes are structurally very similar (see Table 1), their reaction mechanisms are significantly different. Lignin peroxi- dase presents an unusually low optimum pH and is able to cataíyze the oxidation of a variety of compounds with reduction potentials exceeding 1.4 V [69], although the most important substrate is veratryl alcohol (3,4-dimeth- oxybenzyl alcohol) [70]. Like the classical peroxidases, ferric tignin peroxidase fotlows the peroxidase eyele (pathways 1-3 in Figure 1). In the absence of substrate, the addition of hydrogen peroxide resufts in the forma- tion of Compound III (pathway 1 in Figure 2) [71 ]. Further addition of hydrogen peroxide to Compound III drives the enzyme toward bleaching, and irreversible inactiva- tion [71 ]. Although not dtrectly related to the inactivation process, the surface tryptophan 171 residue is hydroxyt- ated and functions as the endogenous electrón donor for Compound I reduction, revealing the existence of múltiple electrón transfer pathways between the protein and the porphyrin [67, 72]. In contrast, manganese per- oxidase produces the oxidant Mn(fll) ion from Mn(H), which behaves as a low-molecular-weight mediatorthat diffuses to remote regions into the lignin molecule and initiates its oxidation. In the presence of hydrogen per- oxide, manganese peroxidase forms Compound I, which is in turn reduced by a bound Mn(ll) atom to form Com- pound II {pathway 2 in Figure 1). Compound II then oxidizes another Mn(ll) ion, driving the enzyme back to theground state. As with other peroxidases, the addition of excess hydrogen peroxide drives manganese peroxi- dase into Compound III [73], which can be further oxi- dized until bleaching and irreversible inactivation [74]. This fungal peroxidase is, so far, the only known enzyme Chemistry & Siology 558 Apéndice 7.5 o + Fe (III) protein oxidation -4 • OH Figure 2. Alternatív© Inactivation Pathways from Compound III Intermedíate system that utilizes soluble Mn(ll)/Mn(lll) as a redox couple. Chloroperoxidase is the most unusual type of peroxi- dase described so far, and its catalytic cycle has not been completely elucidated [23, 75-77]. The proposed mechanism includes a first activation step, in which hy- drogen peroxide transforms the iron{lil)porphyrin group to an oxo-iron(IV)porphyrin-based ir-free radical (Com- pound I) (pathway 1 in Figure 1). Subsequently, Com- pound I can follow three alternative pathways: first, the oxidation of a substrate molecuie to form an oxo-iron- (lV)porphyrin without the associated porphyrin -n radi- cal (Corripound II) (pathway 2 in Figure 1); second, the reaction with a chlorine ion to form a Cl0-iron(lll)por- phyrin group, called Compound X {pathway 4 in Figure 1), which seems to be the only enzymatic activity responsi- ble forthe enzymatic reaction of halogenation. tn addi- tion, this Compound X appears also to be able to per- form oxidative reactions by liberating a chlorine ion [75, 78] (pathway 7 ¡n Figure 1). After both reactions occur, Compound X returns to the ground iron(lll)potphyrin state (pathways 5 and 6 in Figure 1). Finally, chloroper- oxidase shows a significant catalatic activity (pathway 8 in Figure 1), in which Compound I reacts with a second peroxide molecuie to form a three-oxygen-containing enzyme intermedíate, which decomposes to yield stoi- chiometric amounts of oxygen, water, and ground-state chloroperoxidase [32]. Chloroperoxidase is significantly resistant to inactivation by hydrogen peroxide or by or- ganic hydroperoxides because of its intrinsic catalase activity [79]. Nevertheless, exposure to high concentra- tions of hydrogen peroxide (30 mM) irreversibly inacti- vates chloroperoxidase with a half-life of about 1 min [80]. However, very high turnover numbers can be reached with fed-batch reactors coupled to a hydrogen peroxide-sensor controller [81]. No information is avail- abfe about the chloroperoxidase ¡ntermediates during the hydrogen peroxide-mediated inactivation process. Other Hemeproteins Cytochrome P450 is widely distributed among living or- ganisms and has been found in mammals, fish, yeast, bacteria, and plants [82]. This enzyme is part of multien- zymatic systems called monooxygenases that catalyze the activation of dioxygen and the transfer of one of its oxygen atoms into substrates with the consumption of NADPH (or NADH). The monooxygenase cycle of cyto- chrome P450 has been thoroughly studied [27, 83, 84]. The low-spin hexa-coordínated ferric resting form of iron is converted into a high-spin penta-coordinated state upon substrate binding. Further reduction leads to a high-spin penta-coordinated ferrous form capáble of binding dioxygen or carbón monoxide. One-electron re- duction of the oxygen adduct, the last stable intermedí- ate in the reduction cycle, leads to reléase of water and the hydroxylated product. Cytochrome P450 is also able to carry out oxidations with exogenous single-oxygen atom donors such as H2O2, al ky I-hydroperoxides, iodo- sobenzene, amine oxide, and peracids [85-87]. Oxida- tions with these single-oxygen donors are observed in vitro without any need for cytochrome P450 reducíase or any other electrón transfer protein and without con- sumption of NADPH [85,86]. In this case, it seems possi- ble that the enzyme follows a peroxtdase-like catalytic cycle as showed in pathways 9 and 10 in Figure 1. The major producís formed from the oxidation of benzo(a)- pyrene by the monooxygenase pathway with NADPH are phenols, whereas in the presence of cumene hydro- peroxide, the producís are quiñones, supporting the existence of a peroxldasic activity for cytochrome P450 [85]. Treatment of cytochrome P450 with excess cu- mene hydroperoxide leads to protein inactivation and to destruction of the porphyrin into reactive fragments that irreversibly bind to the protein [86]. Self-inactivation of cytochrome P450 during benzphetamine oxidation is accompanied by heme degradation and apo-enzyme modification, followed by protein aggregation, probably Review 559 Apéndice 7.5 through cross-linked oligomers, as evidenced by the appearance of novel carbonyl groups [87). The peroxidasic activity of cytochrome c has been previously reviewed [41], and the hydrogen peroxide- medíated production of peroxyl and alkoxyl radicáis has been investigated by ESR-trapping techniques [88, 89]. Exposure of cytochrome c to excess hydrogen peroxide leads to heme bleaching and protein inactivation [90] by a mechanísm presumably involving the formation of a reactive species equivalent to Compound III. Free- radical species such as tyrosine-based free radicáis [91, 92], as well as alkoxyl and peroxyl reactive species [88, 93], have been detected after exposure of cytochrome c to an excess of hydrogen peroxide. The metabolic role of myoglobin and hemogtobin in- volves the reversible binding of molecular oxygen to an iron(ll)porphyrin group [94]. Part of the oxygenated iron(ll)porphyrin form of the protein is spontaneously auto-oxidized into the iron(lll)porphyrin state known as the met form, which is unable to bind oxygen, with the generation of a superoxide anión [94]. Auto-oxidation is pH dependent, and aithough the half-life of oxygenated myoglobin at physiological conditions was found to be 3.3 days, at low pH it became less than 30 min [95]. Aithough the met form could be reduced back to the ground iron(ll)porphyrin state, the superoxide anión pro- duced can easily be converted into hydrogen peroxide by spontaneous dismutation [96, 97]. Hydrogen perox- ide can induce very raptd oxidation of the deoxyiron(ll)- porphyrin into the met form through the formation of an iron(IV)porphyrin species. On the other hand, the iron(III)- porphyrin-met form further reacts with hydrogen perox- ide via the cyclic formation of an oxo-iron(IV)porphyrin radical catión, analogousto Compound I of peroxidases. That species is very unstable and decays nearly immedi- ately to form a tryptophanyl radical that rapidly reacts with oxygen to form a peroxyl radical [98-102]. It is unclear from the data if a second tyrosyl radical ob- served is formed simultaneously with the tryptophanyl radical or if it is formed by subsequent rescue of the former radical [98-102]. Reaction with hydrogen perox- ide also results ¡n covalent dimerization of sperm whale myoglobin [103] artd in the oxidation of the porphyrin moiety [104, 105J. A Consensus Inactivation Mechanism The oxidative inactivation of hemeproteins is mecha- nism based. The molecular mechanism undertying this hydrogen peroxide-mediated inactivation is extraordi- narily complex because of the fact that a multitude of reactions can occur subsequent to the reaction of the heme iron with the hydroperoxide {Figure 1). Despite peculiarities among different hemeproteins, a common inactivation mechanism comprising several stages can be proposed. In the absence of substrate, or when ex- posed to high concentrations of hydrogen peroxide, per- oxidases show the kinetic behavior of a suicide inactiva- tion, in which hydrogen peroxide is the suicide substrate that converts Compound II into a highly reactive peroxy- iron(IH)porphyrin free-radical called Compound III (path- way 1 in Figure 2) [106]. Compound III is not part of the peroxidase cycle, but it is produced under excessive exposure of protonated Compound II to oxidative spe- cies in a reaction partially mediated by superoxide free radical [60, 106, 107]. ESR spin trapping and spectral analyses have demonstrated the occurrence of this spe- cies after the oxidative treatment of cytochrome c [89], horseradish peroxidase [60], prostaglandin H synthase [108], ügnin peroxidase [71], and manganese peroxi- dase [73]. Despite representing different structurai groups, the kinetic models for the hydrogen peroxide-mediated in- activation of horseradish peroxidase [109] and ascor- bate peroxidase [61] are similar in that they are time dependent and show saturatíon kinetics. tn both cases, the addition of a reducing substrate protected the en- zyme from inactivation. From the stoichiometry of the inactivation, it was concluded that for ascorbate peroxi- dase only two molecules of hydrogen peroxide are re- quired per active site to genérate the inactíve form [61] in contrast to 265 molecules required for horseradish peroxidase [109]. This difference arises from the fact that horseradish peroxidase exhibits a low, albeit signifi- cant, catalattc activity that is absent in ascorbate per- oxidase [110]. For ascorbate peroxidase, inactivation correlated with enzyme bleaching, suggesting heme de- struction [61]. The addition of excess substrate wouid preclude the suicide inactivation by competing with hydrogen perox- ide for Compound II, as has been previously suggested [61, 109]. Once formed, Compound III might follow at least three alternative decomposition pathways (path- ways 2, 3, and 4 in Figure 2). First, given the vicinity of the bound peroxyl radical of Compound III to the porphyrin ring, it is reasonable to suspect that once formed, this reactive species would potentially reach the tetrapyrrole structure and oxidize the porphyrin moiety (pathway 2 in Figure 2). This speculation is supported by the existence of an inactive species, different from but related to Compound III and characterized by heme bleaching, which has been observed after the treatment of ascorbate peroxidase [61 ], hemoglobin [111], myoglo- bin [112], horseradish peroxidase [62], prostaglandin H synthase-1 [108], microperoxidase-11 [113], cytochrome P450 [86], chloroperoxidase [114], and peroxidase from Coprínus cinéreas [63], as well as in prostaglandin H synthase [115] with excess hydrogen peroxide. Heme compounds are particularly susceptible to the formation of biliverdin ring systems by oxidative attack at the meso positions, and the dependence of this process on exog- enous peroxide has been recently demonstrated [116]. Such oxidation readify leads to rupture or elimination of the carbón bridges linking the pyrrole rings and results in cleavage of the porphyrin macrocycle and formation of an open-chain tetra-pyrrole structure [62, 117, 118]. The reléase of heme iron during the formation of these species confirms that they are associated with heme degradation. Second, Compound III might return to the ground state after catalyzing the oxidation of the sur- rounding protein, yielding an oxidized amino acid side chain group in a reaction similar to that previously de- scribed for ügnin peroxidase (Pathway 3 in Figure 2) [119]. Alternatively, the electrón donor might be a sub- strate molecule, in which case the porphyrin moiety would be repaired, and a ground state enzyme would Chemistry 8 Biology 560 Apéndice 7.5 Table 2. Redox Potential oí Some of the Reactions Involved in the Oxidative Inactivation of Peroxidases Redox Reaction NO- + H+ + e" — H2O Mn (III) + e - - M n ( l l ) Met-T + e" — Met HOO- + H* + e" - H2O2 VA-* + 6" — VA Trp-+ + e" — Trp TyrO' + e" + H+ — TyrOH CysS- + e" + H' — CysS Em(mV) 2200 1540 1500 1480 1400 1200 930 900 Reference [151] [171] [156] 1151] [69] [151] [151] [151] result [120]. Finally, the spontaneous tiberation of free radicáis by the unimolecular decay of Compound III is feasible because the peroxyl radical is not covalently bound to the porphyrin. This assumption is supported by experimental evidence demonstrating that in the presence of excess hydrogen peroxide and no reduc- tant, Compound III decays irreversibly because of the formation of reactive oxygen species [109, 121-124]. Once released, two molecules of superoxide free radi- cáis might undergo spontaneous rearrangement into a short-lived tetraoxide species that decomposes into two moiecules of hydroxyl free radicáis and one of oxygen [123]. Hydroxyl free radicáis are more reactive than per- oxyl free radicáis (Table 2), and given their solubility, they are potentially able to oxidize remote amino acid side chains. Amino acid-based free radicáis have been detected after the treatment of cytochrome c [92], met- myoglobin [99], prostaglandin H synthase [125, 126], microperoxidase-8 [127], ascorbate peroxidase [58], and lignin peroxidase [128-130] with excess hydrogen peroxide. In some cases the amino acid-based free radi- cal is located tn the vicinity of the porphyrin, whereas in other cases it is stabilized in the outskirts of the pro- teín. It is well known that free-radical damage may be propagated within protein structures and that, in most cases, transient short-lived species react rapidly with a range of targets to yield other radicáis [5]. Once formed, protein-based free radicáis might travel back and forth between the protein backbone and proximal side chains until they reach the lowest reduction potential site avail- able. The ultímate sink for oxidízing equivalents in pro- teins is cysteine residues (Table 2), although tryptophan- and tyrosine-based free radicáis were observed after the examination of a number of peptide radicáis [131]. Transfer reactions within hemeproteins have been ob- served in myoglobin [132], hemoglobin (133], and leghe- moglobin [134]. These hemeprotein-derived radicáis gen- érate intermolecular crosslinks through the formation of di-tyrosine links. Tyrosine-mediated oligomerization has been observed during the oxidative inactivation of myoglo- bin [103], cytochrome c peroxidase [135], cytochrome P450 [87], lactoperoxidase [136], and myeloperoxidase [1371. Improving Stability through Protein Engineering Protein engineering is generally understood as the use of site-directed or random mutagenesis to alter the properties of a protein. Due to the delicate balance be- tween stabilizing and destabilizing interactions, proteins are only stable under physiologtcal conditions, and two different approaches have been used to increase their stability under non-natural conditions: (1) random muta- genesis followed by selection, also known as directed or molecular evolution and (2) the rational introduction of possibly stabilizing amino acids based on the knowl- edge derived from the three-dimensional structure. Comprehensive reviews of these strategies are available elsewhere [138,139]. It may be possible to harness the powerful tools of molecular biology to directly replace low-redux-potential residues around the active site. This would alter the intermolecular radical transfer pathways with the goal of reducing undesirable paths. As a conse- quence, the inactivation process would be delayed or even suppressed. 1n an interesting structure-based molecular-modeling approach, ascorbate peroxidase variants with increased activity were recovered by engineering of the active site [140]. Stabilization of non-heme proteins has been ob- tained by substitution of amino acids, mainly methionine residues, that are susceptible to oxidative damage [141 ]. Significant results increasing oxidative stability in heme- proteins have resulted from site-directed mutagenesis. In the case of cytochrome P450 BM-3, the F87A substi- tution significantly increased the stability toward hydro- gen peroxide [142], whereas in hemoglobin the forma- tion of a stable thiyl radical decreased the rate of autoxidation and reduced heme degradation attributed to the reaction of superoxide with the heme [143]. Sub- stitution of the active site residues N52I and Y67F abol- ished heme destruction but not protein inactivation in cytochrome c [90], whereas substitution of residue N81 significantly increased the resistance toward hydrogen peroxide apparently by reducing the protein afinity for this compound in the case of recombinant manganese peroxidase [144]. In light of the inactivation mechanism proposed here, these results can be interpreted as evi- dence of electrón abstraction pathway reconfigurations that lead to an altered electron-allocation equilibrium between the substrate molecule and secondary sinks. Despite the wide acceptance of directeci-evotution methods in protein engineering, only a few hemepro- teins have been subjected to this approach and, in most cases, the aim was to modifiy their catalytic properties. In some pioneering studies, a recombinant horse heart- myoglobin quadruple variant with increased peroxidastc activity was obtained after successive rounds of random mutagenesis and screening [145]. In a more recent publi- cation, yeast cytochrome c peroxidase mutants with increased activity toward guaiacol were recovered after several rounds of DNA shuffling and screening [146]. Interestingly, all these mutants contained the multi-spe- cies conserved arginine 48 residue changed to histidine, in addition to other changes related to the general stabil- ity of the apo-protein, which is believed to play a key role in the active site of the enzyme as a gatekeeper controlling the access of small molecules to the ferry! oxygen and the distal histidine [146]. Pseudomonas put- ida cytochrome P450 was successfully evolved toward the hydroxylation of naphthatene with hydrogen perox- ide as the electrón acceptor. A first round of mutagene- sis and a single round of recombination were foltowed by a second series of experiments, in which five first- generatton variants were used as parents in a novel Review 561 Apéndice 7.5 shuffling reaction yielding mutants with as much as a 20- fold improvement in activity over the wild-type protein [147]. Finally, random mutagenesis has been used in the Caldaríomyces fumago chloroperoxidase locus to develop a mutant that resisted the suicidal inactivation by allylbenzene [148] as well as a mutant with increased activity in organic cosolvents [149]. tn the only example of molecular evolution of a heme- protein being aimed to increase protein stability toward hydrogen peroxide, a combination of approaches were used to develop a funga! peroxidase for activity in the highly alkaline and oxidative conditions of laundry wash water [150]. Using the crystal structure as a guide, the researchers first used site-directed mutagenesis to tar- get amino acid residues susceptible to oxidation by hy- drogen peroxide. Three amino acids were identified and combined by site-directed mutagenesis to genérate a variant with an oxidative stability 5-fold higher than that of the wild-type enzyme and thermostability that had impoved by more than 100-fold. Further random muta- genesis and selection identified a series of mutants with greater improvements in thermostability and peroxide stability, but such improvements carne at the cost of reducing the overall activity of the enzyme. To overeóme this obstacle, the authors then shuffled in vivo clones with improved thermostability with clones encoding en- zymes with high activity. The output of these experi- ments was mutant enzymes with higher specific activity and greater stability than any of the input parentai gene produets. Afinal round of in vivo shuffling resulted in two distinct mutants with substitutions in the same position, I49. The best of these produets was 174 times more thermally stable and 100 times more stable in the pres- ence of hydrogen peroxide than the native enzyme. Most of the changes selected in the former mutant were lo- cated inside the active site, mainly in the contact potnt between the two hélices that coordínate the binding of peroxide. The replacement of ¡soleucine at position 49 with amino acids that can establish hydrogen bonding was espectally important. The I49S substitution in- creased oxidative stability 50-fold, probably by promot- ing the estabtishment of an alternative hydrogen bond- ing network. Protein-based free radicáis are known to be allocated through covalent bonds but also through hydrogen bonding networks [151 ], and therefore the sta- bifizing effect of a novel hydrogen bonding architecture might allow alternative free-radical pathways, enabling other sources of electrons to provide the reducing power instead of the protein. In cases in which neither the crystallographic struc- ture ñor a recombinant expressíon system are available, disruption of the electrón tunnel pathway could be at- tained by chemical means. An example of this approach is the electrón abstraction source switching from the substrate to the porphyrin ring, as observed in the hydro- gen peroxide-treated prostaglandin H synthase upon the addition of cyclooxygenase inhibitors [152]. Future Prospects From the evidence presented here, we conclude that in the case of hemeperoxidases substrate oxidation is a naturally imperfect process, and we hypothesize that to some extent, the porphyrin ring or the protein backbone may become the alternative electrón sources. Protein destruction seems to arise ultimately as a consequence of the of nonproductive electrón abstraction pathways in the reaction pathway, whereas protection by the sub- strate comes from the favorable partition of the oxidative equivalents toward the substrate. Although we do not know whether these alternative pathways opérate se- quentially or símultaneously, overall stabitization of per- oxidases against hydrogen peroxide might be achieved through the rational reorganizaron of low-reduction- potential residues within the active site. The aim of this approach would be to orient the electrón abstraction pathways toward the substrate instead of the porphyrin or the protein. We are confídent that the systematic study of the mechanism-based inactivation of hemeper- oxidases wíll undoubtedly provide the knowledge re- quired for the rational design of site-directed variants to prevent suicide inactivation. Acknowledgments This work was funded by Conacyt grant 33611 and by Mexican Institute of Petroleum (IMP) grant FIES 98-110-VI. We thank Grant Mauk and Michael Pickard for their comments on the manuscript. References 1. Branchaud, B.P. (1999). Free radicáis as a resuft of dioxygen metaboltsm. Metal lons Biol. Sys. 36, 79-102. 2. Boveris, A., and Cadenas, E. (2000). Mitochondrial production of hydrogen peroxide regulation by nitric oxide and the role of ubisemiquinone. IUBMB Ufe 50, 245-250, 3. Salazar, J.J., and Van Houíen, B. (1997). Preferential mito- chondiral DNA injury caused by glucose oxidase as a steady generator of hydrogen peroxide in human fibroblasts. Mut. Res. DNA Repair. 385, 139-149. 4. Bockle, B., Martínez, M.J., Guillen, F., and Martínez, A.T. (1999). Mechanism of peroxidase inactivation ¡n liquid cultures of the ligninolytic fungus Pleurotus pulmonaríus. Appl. Environ. Mi- crobiol. 65, 923-928. 5. Hawkins, C.L., and Davies, M.J. (2001). Generation and propa- gation of radical reactions on proteins. Biochim. Biophys. Acta 7504,196-219. 6. Stadtman, E.R. (1993). Oxidatíon of free amino acids and amíno acid residues in proteins by radiolysis and by metal-catalyzed reactíons. Annu. Rev. Biochem. 62, 797-821. 7. Fu, S.L., and Dean, R.T. (1997). Structural characterization of the produets of hydroxyl-radical damage to leucine and their detection on proteins. Biochem. J. 324, 41-48. 8. Rana, S., and Robinson, B.H. (2000). Mitochondria, oxygen free radicáis, disease and ageing. Trends Biochem. Scí. 25, 502-508. 9. Muller, F. (2000). The nature and mechanism of superoxide production by the electrón transport chaín: is relevance to aging. J. Am. Aging Asoc. 23, 227-253. 10. Anson, R.M., and Bohr, V.A. (2000). Mitochondria, oxidative DNA damage, and aging. J. Am. Aging Asoc. 23,199-218. 11. Dean.R.T., Fu, S.L., Stocker, R-, and Davies, M.J. (1997). Bio- chemistry and pathoíogy of radical-mediated protein oxida- tion. Biochem. J. 324, 1-18. 12. Rúan, H., Tang, X.D., Chen, M.-L, Joiner, M.A., Sun, G., Brot, N., Weissbach, H., Heinemann, S.H., Iverson, L, Wu. C.-F., et al. (2002). High-quality life extensión by the enzyme peptide methionine sulfoxide reducíase. Proc. Nati. Acad. Sci. USA 99, 2748-2753. 13. Chevion, M., Berenshtein, E., and Stadtman, E.R. (2000). Hu- man studies related to protein oxidation: Protein carbonyl con- tení as a marker of damage. Free Radie. Res. 33, S99-S108. 14. Alien, J.D., Bentley, D., Stringer, R.A., and Lowther, N. (1999). /ir Chemistry & Biology 562 Apéndice 7.5 Hybrid (BDBB) interferon-alpha: preformulation studies. Int. J. Pharm. 187, 259-272. 15. Liu, J.L., Lu, K.V., Katta, V., Westcott, K.R., Narhi, L.O., and Lu, H.S. (1998). In vitro methionine oxidation of recombinant human leptin. Pharm. Res. 75, 632-640. 16. Taggart, C , Cervantes-Laurean, D., Kim, G., McElvaney, N.G., Wehr, N., Moss, J., and Levine, R.L. (2000). Oxidation of either methionine 351 or methionine 358 in-alpha(i}-antitrypsin causes loss of anti-neutrophii elastase activity. J. Biol. Chem. 275, 27258-27265. 17. Milzani, A., Rossi, R., Di Simplicio, P., Giustarini, D., Colombo, R., and DalleDonne, I. (2000). The oxidation produced by hy- drogen peroxide on Ca-ATP-G-actin. Protein Sci. 9, 1774- 1782. 18. DePaz, R.A., Barnett, C.C., Dale, D.A., Carpenter, J.F., Gaert- ner, A.L., and Randolph, T.W. (2000). The excluding effects of sucrose on a protein chemical degradation pathway: methio- nine oxidation in subtilisin. Aren. Biochem. Biophys. 384, 123-132. 19. Kim, B.M., Schultz, L.W., and Raines, R.T. (1999). Variants of ribonuclease inhibitor that resist oxidation. Protein Sci. 8, 430-434. 20. Hanson, S.R.A., Chen, A.A., Smith, J.B., and Lou, M.F. (1999). Thiolation of the gamma B-crystailins ín intact bovine lens exposed to hydrogen peroxide. J. Biol. Chem. 274,4735-4742. 21. Levine, R.L., Moskovitz, J., and Stadtman, E.R. (2000). Oxida- tion of methionine in proteins: roles in antioxidant defense and cellular regulation. IUBMB Life 50, 301-307. 22. Levine, R.L., Berlett, B.S., Moskovitz, J-, Mosoni, L, and Stadt- man, E.R. (1999). Methionine residues may protect proteins from critical oxidative damage. Mech. Ageing Dev. 107, 323-332. 23. Dunford, H.B. (1999). Heme Peroxidases. (New York: John W¡- iey and Sons). 24. Símons, B.H., Barnett, P., Vollenbroek, E.G.M., Dekker, H.L., Muijesers, A.O., Messerschmidt, A., and Wever, R. (1995). Pri- mary structure and characterisation of the vanadium chloro- peroxidase from the fungus Curvularia inaequalis. Eur. J. Bio- chem. 229, 566-574. 25. Shimonishi, M., Kuwamoto, S., Inoue, H., Wever, R., Ohshiro, T., Izumi, Y., and Tanabe, T. (1998). Cloning and expression of the gene for a vanadium-dependent bromoperoxidase from a marine macro-algae, Corallina pilulfera. FEBS Lett. 362, 105-110. 26. Wolfframm, C, Lingens, F.M.R., and van Pee, K.-H. (1993). Chloroperoxidase-encoding gene from Pseudomonas pyrroci- nia—sequence, expression in heterologous hosts, and purifi- cation of the enzyme. Gene 130,131-135. 27. Dawson, J.H., and Sonó, M. (1987). Cytochrome P-450 and chloroperoxidase: thiolate-ligated heme eníymes. Spectro- scopic determinaron of their active site structures and mecha- nistic implications of thiolate ligation. Chem. Rev. 87, 1255- 1276. 28. Holm, L., and Sander, C. (1996). Fold classification based on structure-structurealignmentof proteins— European Bioinfor- matics Institute. Science 273, 595-603. 29. Liu, W., Kumar, J., Tripathy, S., Senecal, K.J., and Samuelson, L. (1999). Enzymatically synthesized conducting polyaniline. J . Am. Chem. Soc. 121, 71-78. 30. van Rantwijk, F., and Sheldon, R.A. (20QO). Selective oxygen transfercatalysed by heme peroxidases: synthetic and mecha- nistic aspeets. Curr. Opin. Biotechnol. 11, 554-564. 31. Thomas, J.A., Morris, D.R., and Hager, L.P. (1970). Chloroper- oxidase. Vil. Classical peroxidase, catalattc, and halogenating forms of the enzyme. J. Biol. Chem. 245, 3129-3134. 32. Sun, W., Kadima, T.A., Pickard, M.A., and Dunford, H.B. (1994). Catalasé activity of chloroperoxidase and its interaction with peroxidase activity. Biochem. Cell Biol. 72, 321-331. 33. Hochman, A., and Shemesh, A. (1987). Purification and charac- terization of a catalase-peroxidase from the photosynthetic bacterium Rhodopseudomonas capsúlala. J. Biol. Chem. 262, 6871-6876. 34. Amemura-Maekawa, J., Mishima-Abe, S., Kura, F., Takahashi, T., and Watanabe, H. (1999). Identification of a novel periplas- mic catalase-peroxidase KatA of Legíoneila pneumophila. FEMS Microbiol. Lett. 176, 339-344. 35. Zamocky, M., Janecek, S., and Koller, F. (2000). Common phy- logeny of catalase-peroxidases and ascorbate peroxidases. Gene 256,169-182. 36. Chouchane, S., Lippai, I., and Magliozzo, R.S. (2000). Catalase- peroxidase {Mycobacterium tuberculosis KatG) catalysis and isoniazid activation. Biochemistry 39, 9975-9983. 37. Pfeifer, O., Pelletier, I., Altenbuchner, J., and von Pee, K.-H. (1992). Molecular cloning and sequencing of a non-haem bro- moperoxidase gene from Streptomyces aureofaciens ATCC 10762. J. Gen. Microbiol. 138,1123-1131. 38. Ortíz-León, M., Velasco, L, and Vazquez-Duhalt, R. (1995). Biocatalytic oxidation of polycyclic aromatic hydrocarbons by hemoglobin and hydrogen peroxide. Biochem. Biophys. Res. Comm. 2/5, 968-973. 39. Lenk, W., and Sterzl, H. (1984). Peroxidase activity of oxyhae- moglobin in vitro. Xenobiotica 14, 581-588. 40. Rao, S.I., Wilks, A., Hamberg, M., and Ortiz de Montellano, P.R. (1994). The lipoxygenase activity of myoglobin. Oxidation of íinoleic acid by the ferryl oxygen ratherthan protein radical. J. Biol. Chem. 269, 7210-7216. 41. Vazquez-Duhalt, R. (1999). Cytochrome c as a biocatalyst. J. Mol. Cat. B Enz. 7, 241-249. 42. Adams, P.A., and Louw, J. (1995). Dioxygen bond scission and heme degradation in haemproteins: a kinetic-study of chemi- cal-model systems using ferrimyoglobin and haempeptide- non-haempeptide complexes as catalysts for peroxidasic re- duction of hydrogen-peroxide. J. Chem. Soc. [Perkin 1] 2, 1683-1690. 43. DeSimone, L.D., and Popoff, F. (1997). The Business Link to Sustainable Development. (Cambridge, MA: MIT Press). 44. Anastas, P.T., and Wamer, J.C. (1998). Green Chemistry: The- ory and Practice. (New York: Oxford Universíty Press). 45. Ramos, M.C., Torijas, M.C., and Diaz, A.N. (2001). Enhanced chemiluminiscence biosensor for the determination ol pheno- lic compounds and hydrogen peroxide. Sensors and Actuators B-Chemical. 73, 71-75. 46. Iwuoha, E.I., Joseph, S., Zhang, Z., Smyth, M.R., Fuhr, U., and Ortiz de Montellano, P.R. (1998). Drug metabolism biosensors: electrochemical reactivities of cytochrome P450cam ¡mmobi- lised in synthetic vesicular systems. J. Pharm. Biomed. Anal. 17, 1101-1110. 47. Ingram, D.T., Lamichhane, C.M., Rollins, D.M., Carr, LE., Mal- línson, E.T., and Joseph, S.W. {1998). Development of a colony lift ¡mmunoassay to facilítate rapid detection and quantification of Escheríchia cotí 0157:H7 from agar plates and füter monitor membranes. Clin. Diagnos. Lab. Immunol. 5, 567-573. 48. Paice, M.G., and Jurasek, K.L. (1984). Peroxidase catalyzed color removal from bleach plant effluent. Biotechnol. Bioeng. 26, 477-480. 49. Michel, F.C., Dass, S.B., Grulke, E.A., and Reddy, C.A. (1991). Role of manganese peroxidase and lignin peroxidase of Pha- nerochaete chrysosporium in the decolorization of Kraft bleach plant effluent. Appl. Environ. Microbiol. 57, 2368-2375. 50. Ferrer, I., Dezotti, M., and Duran, N. (1991). Decolorization of Kraft effluent by free and immobilized lignin peroxidases and horseradish peroxidase. Biotechnol. Lett. 13, 577-582. 51. Hammel, K.E., Kayanaraman, B., and Kirk, T.K. (1986). Oxida- tion of polycyclic aromatic hydrocarbons and dibenzo(p)diox- ins by Phanerochaete chrysosporium ligninase. J. Biol. Chem. 36, 16948-16952. 52. Klibanov, A.M., Tu, T.-M., and Scott, K.L. (1983). Peroxidase catalyzed removal from coal-conversion waste waters. Sci- ence 22/ , 259-260. 53. Aitken, M.D. (1993). Waste treatment applications of enzymes: opportunities and obstacles. Chem. Eng. J. 52, B49-B58. 54. Patel, M., and Day, B.J. (1999). Metalloporphyrín class of thera- peutic cataíytic antioxidants. Trends Pharm. Sci. 20, 359-364. 55. Gaspar, S-, Popescu, I.C., Gazaryan, I.G., Bautista, A.G., Sak- harov, I.Y., Mattiasson, B., and Csoregi, E. (2000). Biosensors based on novel plant peroxidases: a comparative study. Elec- trochimica Acta. 46, 255-264. Review 563 Apéndice 7.5 56. Yilmaz, E. (2001). Biotechnological production oí prostagían- din. Biotechnol. Adv. 19, 387-397. 57. van Deurzen, M.P.J., van Rantwijk, F.,and Sheldon, R.A. (1997). Selective oxidations catalyzed by peroxidases. Tetrahedron 53,13183-13220. 58. Hiner, A.N.P., Martínez, J.I., Arnao, M.B., Acosta, M., Tumer, D.D., Raven, E.L., and Rodríguez-Lo pez, J.N. (2001). Detection oí a tryptophan radica! in the reacíion of ascorbate peroxidase with hydrogen peroxide. Eur. J. Biochem. 268, 3091-3098. 59. Miller, V.P., Goodin, D.B., Frtedman, A.E., Hartmann, C, and Ortiz de Montellano, P.R. (1995). Horseradish peroxidase Phe172—Tyr mutant. Sequential formation of compound I with a porphyrin radical catión and a protein radical. J. Biol. Chem. 270,18413-18419. 60. Adediran, S.A., and Lambeir, A.M. (1989). Kinetics of the reac- tion of compound fl of horseradish peroxidase with hydrogen peroxide to form compound III. Eur. J. Biochem. 186,571-576. 61. Hiner, A.N.P., Rodriguet-Lopez, J.N., Arnao, M.B., Lloyd Ra- ven, E., Garcia-Canovas, F., and Acosta, M. (2000). Kinetic study of the inactivatíon of ascorbate peroxidase by hydrogen- peroxide. Biochem. J. 348, 321-328. 62. Nakajima, R., and Yamazaki, t. (1980). The conversión of horse- radish peroxidase c to a verdohemoprotein by a hydroperoxide derived enzymatically from indole-3-acetic acid and by m-nitro- peraxybenzoic acid. J. Biol. Chem. 255, 2067-2071. 63. Chang, H.C., Holland, R.D., Bumpus, J.A., Churchwell, M.I., and Doerge, D.R. (1999). Inactivation of Coprínus cinéreas per- oxidase by 4-chloroaniline during turnover: comparison with horseradish peroxidase and bovine lactoperoxidase. Chem. Biol. Interact. 123,197-217. 64. Gold, M.H., Kuwahara, M., Chiu, A.A., and Glenn, J.K. (1984). Purificaron and characterization of an extracellular HtO2- requiring diarylpropane oxygenase from the white rot basidio- mycete, Phanerochaete chrysosporíum. Arch. Biochem. Bio- phys. 234, 353-362. 65. Kuwahara, M., Glenn, J.K., Morgan, M.A., and Gold, M.H. (1984). Separation and characterization of rwo extracellular H;O2-dependent oxidases from ligninolytic cultures of Phaner- ochaete chrysosporium. FEBS Lett. 169, 247-250. 66. Tien, M., and Kirk, T.K. (1984). Lignin-degrading enzyme from Phanerochaete chrysosporium: purificaron, characteriíation, and catalytic properties of a unique H2O2-requÍring oxygenase. Proc. Nati. Acad. Sci. USA 8/ , 2280-2284. 67. Píontek, K-, Smith, A.T., and Blodig, W. (2001). ügnin peroxi- dase structure and function. Biochem. Soc. Trans. 29,111-116. 68. Pease, E.A., and Tien, M. (1992). Heterogeneity and regulation of manganese peroxidase from Phanerochaete chrysospor- ium. J. Bacterio!. 174, 3532-3540. 69. Bietti, M., Baciocchi, E., and Steenken, S. (1998). üfetime, reduction potential and base-induced tragmentation of the ver- atryl alcohol radical catión ¡n aqueous soluiion. Pulse radioly- sis studies on a ligninase "mediator." J. Phys. Chem. A. 102, 7337-7342. 70. Harvey, P.J., Floris, R., Lundell. T., Palmer, J.M., Schoemaker, H.E., and Wever, R. (1992). Catalytic mechanisms and regula- tion of lignin peroxidase. Biochem. Soc. Trans. 20, 345-349. 71. Warlishi, H., and Gold, M.H. (1990). ügnin peroxidase com- pound III. Mechanism of formation and decomposition. J. Biol. Chem. 265, 2070-2077. 72. Blodig, W., Smith, A.T., Doyte, W.A., and Piontek, K. (2001). Crystal structure of pristine and oxidatively processed lignin peroxidase expressed in Escherichia coli and of the W171F variant that eliminates the redox active tryptophan 171. Impli- cationsforthe reaction mechanism. J. Mol. Biol. 305,851-861. 73. Wariishi, H., Aktleswaran, L, and Gold, M.H. (1988). Manga- nese peroxidase from the basidiomycete Phanerochaete chrysosporium: spectral characterization of the oxidized states and the catalytic cycle. Biochemistry 27, 5365-5370. 74. Timofeevski, S.L., Reading, N.S., andAust, S.D. (1998). Mecha- nisms for protection against inactivation of manganese peroxi- dase by hydrogen peroxide. Arch. Biochem. Biophys. 356, 287-295. 75. übby, R.D., Beachy, T.M., and Phipps, A.K. (1996). Quantitat- ing direct chlorine transfer from enzyme to substrate in chloro- peroxidase-catalyzed reactions. J. Biol. Chem. 277, 21820- 21827. 76. Wagenknecht, H.A., and Woggon, W.D. (1997). Identification of ¡ntermediates in the catalytic cycle of chloroperoxidase. Chem. Biol. 4, 367-372. 77. Libby, R.D., Rotberg, N.S., Emerson, J.T., White, T.C., Yen, G.M., Friedman, S.H., Sun, N.S., and Goldowski, R. (1989). The chloríde-activated peroxidation of catechol as a mechanistic probé of chloroperoxidase reactions. Competitive activation as evidence for a catalytic chloride binding site on compound I. J. Biol. Chem. 264,15284-15292. 78. Libby, R.D., Shedd, A.L., Phipps, A.K., Beachi, T.M., and Gerst- berger, S.M. (1992). Defining the involvement of HOCI or Cl2 as enzyme-generated ¡ntermediates ¡n chloro peroxidase-cata- lyzed reactions. J. Biol. Chem. 267, 1769-1775. 79. Manoj, K.M., and Hager, L.P. (2001). Utiüzation of peroxide and its relevance in oxygen insertion reactions catalyzed by chloroperoxidase. Biochim. Biophys. Acta 7547, 408-417. 80. Shevelkova, A.N., and Ryabov, A.D. (1996). Irreversible inacti- vation of Caldariomyces fumago chloroperoxidase by hydro- gen peroxide. A kinetic study in chloride and bromide systems. Biochem. Mol. Biol. Int. 39, 665-670. 81. Seelbach, K., van Deurzen, M.P.J., van Rantwijk, F,, Sheldon, R.A., and Kragl, V. (1997). Improvement of the total turnover number and space-time yield for chloroperoxidase catalyzed oxidation. Biotechnol. Bioeng. 55, 283-288. 82. Ortiz de Montellano, P.R. (1986). Cytochrome P450, Structure, Mechanism and Biochemistry. (New York: Plenum Press). 83. Sonó, M., Roach, M.P., Coulter, E.D., and Dawson, J.H. (1996). Heme-containing oxygenases. Chem. Rev. 96, 2841-2887. 84. Schlichting, I., Berendzen, J., Chu, K., Stock, A.M., Maves, S.A., Benson, D.E., Sweet, R.M., Ringe, D., Petsko, G.A., and Sligar, S.G. (2000). The catalytic pathway of cytochrome P450cam at atomic resolution. Science 287,1615-1622. 85. Capdevila, J., Estabrook, R.W., and Prough, R.A. (1980). Differ- ences ín the mechanism of NADPH- and cumene hydroperox- ¡de-supported reactions of cytochrome P-450. Arch. Biochem. Biophys. 200, 186-195. 86. He, K-, Bornheim, L.M., Falick, A.M., Maítby, D., Yin, H.Q., and Correia, M.A. (1998). Identification of the heme-modified peptides from cumene hydroperoxide-inactivated cytochrome P450 3A4. Biochemistry 37,17448-17457. 87. Karuzina, I.I., Zgoda, V.G., Kuznetsova, G.P., Samenkova, N.F., and Archakov, A.l. (1999). Heme and apoprotein modification of cytochrome P450 2B4 during its oxidative ¡nactivation in monooxygenase reconstituted system. Free Radie. Biol. Med. 26, 620-632. 88. Davies, M.J. (1988). Detection of peroxyl and alkoxyl radicáis produced by reaction of hydroperoxides with heme-proteins by electrón spin resonance spectroscopy. Biochim. Biophys. Acta 964, 28-35. 89. Barr, D.P., and Masón, R.P. (1995). Mechanism of radical pro- duction from the reaction of cytochrome c with organic hydro- peroxides—an ESR spin-trappíng investigation. J. Biol. Chem. 270, 12709-12716. 90. Villegas, J.A., Mauk, A.G., and Vazquez-Duhalt, R. (2000). A cytochrome c variant resistant to heme degradation by hydro- gen peroxide. Chem. Biol. 7, 237-244. 91. Deterding, L.J., Barr, D.P., Masón, R.P., and Tomer. K.B. (1998). Characterization of cytochrome c free radical reactions with peptides by mass spectrometry. J. Biol. Chem. 273, 12863-12869. 92. Barr, D.P., Gunther, M.R., Deterding, L.J., Tomer, K.B., and Masón, R.P. (1996). ESR spin-trapping of a protein-derived tyrosyl radical from the reaction of Cytochrome c with hydro- gen peroxide. J. Biol. Chem. 271, 15498-15503. 93. Busi, E., Howes, B.D., Pogni, R., Basosi, R., Tinoco, R., and Vazquez-Duhalt, R. (2000). Modified cytochrome c/HzO2 sys- tem: spectroscopic EPR investigation of the btocatalytic be- haviour. J. Mol. Cat. B Enz. 9, 39-48. 94. Shikama, K. (1998). The molecular mechanism of autoxidation for myoglobin and hetnoglobín: a venerable puzzle. Chem. Rev. 98,1357-1373. 95. Sugawara, Y., and Shikama, K. (1980). Autoxidation of native /J? Chemistry & Biology 564 Apéndice 7.5 oxymyoglobin. Thermodynamic analysis of the pH profile. Eur. J. Biochem. Í70, 241-246. 96. Rabani, J., and Nielsen, S.O. (1969). Absortion spectrum and decay kinetics of O2~ and HO2 ¡n aqueous solutions by pulse radiolysis. J. Phys. Chem. 73, 3736-3744. 97. Fridovich, I. (1975). Superoxide dismutases. Annu. Rev. Bio- chem. 44,147-159. 98. Gunther, M.R., Sturgeon, B.E., and Masón, R.P. (2000). A long- lived tyrosyl radical from the reaction between horse metmy- oglobin and hydrogen peroxide. Free Radie. Biol. Med. 28, 709-719. 99. Gunther, M.R., Tschirret-Guth, R.A., Witkowska, H.E., Fann, Y.C., Barr, D.P., Ortiz de Montellano, P.R., and Masón, R.P. (1998). Site-specific spin trapping of tyrosine radicáis in the oxidation of metmyoglobin by hydrogen peroxide. Biochem. J. 330,1293-1299. 100. Gunther, M.R., Kelman, D.J., Corbett, J.T., and Masón, R.P. (1995). Self-peroxidation of metmyoglobin results in formation of an oxygen-reactive tryptophan-centered radical. J. Biol. Chem. 270, 16075-16081. 101. Svístunenko, D.A. (2001). An EPR study of the peroxyl radicáis induced by hydrogen peroxide in the haem proteins. Biochim. Bíophys. Acta 1546, 365-378. 102. Patel, R.P., Svistunenko, D.A., Darley-Usmar, V.M., Symons, M.C., and Wilson, M.T, (1996). Redox eyeling of human met- haemoglobin by H2OS yields persistent ferryl iron and protein based radicáis. Free Radie. Res. 25, 117-123. 103. Tew, D., and Ortiz de Montellano, P.R. (1988). The myoglobin protein radical. Coupling of Tyr-103 to Tyr-151 in the H2O2- mediated cross-linking of sperm whale myoglobin. J. Biol. Chem. 263, 17880-17886. 104. Sugiyama, K., Highet, R.J., Woods, A., Cotters, R.J., and Osawa, Y. (1997). Hydrogen peroxide-mediated alteration of the heme prosthetic group of metmyoglobin to an iron chlorin product: evídence for a novel oxidative pathway. Proc. Nati. Acad. Sci. USA 94, 796-801. 105. Cantoni, L, Gibbs, A.H., and De Matteis, F. (1981). Loss of haem and haemoproteins during the generation of superoxide anión and hydrogen peroxide: a pathway not involving produc- tion of carbón monoxide. Int. J. Biochem. 13, 823-830. 106. Nakajima, R., and Yamazaki, I. (1987). The mechanism of oxy- peroxidase formation from ferryl peroxidase and hydrogen peroxide. J. Biol. Chem. 262, 2576-2581. 107. Keilin, D., and Mann, T. (1937). On the haematin compound of peroxidase. Proc. R. Soc. Lond. B. 122,119-133. 108. Wu, G., Wei, C , Kulmacz, R.J., Osawa, Y., and Tsai, A. (1999). A mechanistic study of self-inactivation of the peroxidase ac- tivity in prostaglandin H synthase-1. J. Biol. Chem. 274,9231- 9237. 109. Hiner, A.N., Hernandez-Ruiz, J., Garcia-Canovas, F., Smith, A.T., Arnao, M.B., and Acosta, M. (1995). A comparative study of the inactivatíon of wüd-type, recombinant and two mutant horseradish peroxidase isoenzymes C by hydrogen peroxide and m-chloroperoxybenzoic acid. Eur. J. Biochem. 234, 506-512. 110. Hernandez-Ruiz, J., Amao, M.B., Hiner, A.N.P., Carcia-Cano- vas, F., and Acosta, M. (2001). Catalase-like activity of horse- radish peroxidase: relationship to enzyme inactívation by H2O2. Biochem. J. 354,107-114. 111. Nagababu, E., and Rifkind, J.M. (2000). Heme degradation dur- ing autoxidation of oxyhemoglobin. Biochem. Biophys. Res. Comm. 273, 839-845. 112. Catalano, CE., Choe, Y.S., and Ortiz de Montellano, P.R. (1989). Reactions of the protein radical in peroxide-treated myoglobin. Formation of a heme-protein cross-link. J. Biol. Chem. 264,10534-10541. 113. Spector, A., Zhou, W., Ma, W.C., Chignell, C.F., and Reszka, K.J. (2000). lnvestigation of the mechanism of action of micro- peroxidase-11, (MP11), a potential anti-cataract agent, with hydrogen peroxide and ascorbate. Exp. Eye Res. 71,183-194. 114. Lambeir, A.M., and Dunford, H.B. (1985). Oxygen binding to dithionite-reduced chloroperoxidase. Eur. J. Biochem. Í47, 93-96. 115. Mahy, J.P., Gaspard, S., Delaforge, M., and Mansuy, D. (1994). Reactions of prostaglandin-H synthase with monosubstituted hydrazines and diazenes—formation of iron(ll)-diazene and iron(lll)-o-alkyl or iron(lll)-a-aryl complexes. Eur. J. Biochem. 226, 445-457. 116. Sigman, J.A., Wang, X., and Lu, Y. (2001). Coupled oxidation of heme by myoglobin ¡s mediated by exogenous peroxide. J. Am. Chem. Soc. 123, 6945-6946. 117. O'Carra, P. (1975). Heme-cleavage: biological systems and chemical analogs. In Porphyrin and Metalloporphyrins, K. Smith, ed., (Amsterdam, The Netherlands: Elsevier Scientific Publishing). pp. 123-153. 118. Brown, S.B. (1976). A model for the formation of bile-pigments isomers in vivo and in vitro. Biochem. J. 159, 23-27. 119. Cai, D.Y., and Tien, M. (1989). On the reactions of lignin peroxi- dase compound ill (isozyme H8). Biochem. Biophys. Res. Comm. 162, 464-469. 120. Dordick, J.S., Klibanov, A.M., and Marletta, M.A. (1986). Horse- radish peroxidase catalyzed hydroxylations: mechanistic stud- ies. Biochemistry 25, 2946-2951. 121. Baynton, K.J., Bewtra, J.K., Biswas, N., and Taylor, K.E. (1994). Inactivation of horseradish peroxidase by phenol and hydro- gen peroxide: a kínetic investigation. Biochim. Biophys. Acta 7206, 272-278. 122. Amao, M.B., Acosta, M., del Rio, J.A., and Garcia-Canovas, F. (1990). Inactivation of peroxidase by hydrogen peroxide and its protection by a reductant agent. Biochim. Biophys. Acta 1038, 85-89. 123. Adediran, S.A. (1996). Kinetics of the formation of p-670 and of the decay of Compound III of horseradish peroxidase. Aren. Biochem. Biophys. 327, 279-284. 124. Ximenes, V.F., Catalani, L.H., and Campa, A. (2001). Oxidation of melatonin and tryptophan by and HRP eyele involving Com- pound III. Biochem. Biophys. Res. Comm. 287,130-134. 125. DeGray, J.A., Lassmann, G., Curtís, J.F., Kennedy, T.A., Marnett, L.J., Eling, T.E., and Masón, R.P. (1992). Spectral analysis of the protetn-derived tyrosyl radicáis from prosta- glandin H synthase. J. Biol. Chem. 267, 23583-23588. 126. Shi. W., Hoganson, C.W., Espe, M., Bender, C.J., Babcock, G.T., Palmer, G., Kulmacz, R.J., and Tsai, A. (2000). Electron paramagnetic resonance and electrón nuclear double reso- nance spectroscopic identification and characterizaton of the tyrosyl radicáis in prostaglandin H synthase 1. Biochemistry 39, 4112-4121. 127. Primus, J.L., Grunenwald, S., Hagedoorn, P.-L, Albrecht-Gary, A.-M., Mandón, D., and Veeger, C. (2002). The nature of the intermediates in the reactions of Fe(lll)- and Mn(lll)-microper- oxidase-8 with H2O2: a rapid kinetics study. J. Am. Chem. Soc. 724, 1214-1221. 128. Blodig, W., Doyle, W.A., Smith, A.T., Winterhalter, K., Choinow- ski, T., and Piontek, K. (1998). Autocatalytic formation of a hydroxy group at C beta of Trp171 in lignin peroxidase. Bio- chemistry 37, 8832-8838. 129. Doyle, W.A., Blodig, W,, Veitch, N.C., Piontek, K., and Smith, A.T. (1998). Two substrate interaction sites in lignin peroxidase revealed by site-directed mutagenesis. Biochemistry 37, 15097-15105. 130. Choinowski, T., Blodig, W., Winterhalter, K.H., and Piontek, K. (1999). The crystal structure of lignin peroxidase at 1.70 Á resolution reveáis a hydroxyl group on the C-beta of trypto- phan 171: a novel radical site formed during the redox eyele. J. Mol. Biol. 2S6, 809-827. 131. Prutz, W.A. (1990). Free radical transfer invoMng sulphur pep- tide functions. In Sulfur-Centered Reactive Intermediates in Chemistry and Biology, C. Chatgilialoglu and K.D. Asmus, eds. (New York: Plenum Press), pp. 389-399. 132. Wilks, A., and Ortiz de Montellano, P.R. (1992). Intramolecular translocation ol the protein radical formed in the reaction of recombinant sperm whale myoglobin with H¡O2. J. Biol. Chem. 267, 8827-8833. 133. Giulivi, C , and Cadenas, E. (1998): Heme protein radicáis: for- mation, fate, and biological consequences. Free Radie. Biol. Med. 24, 269-279. 134. Davies, M. J., and Puppo, A. (1992). Direct detection of a globín- Review 565 Apéndice 7.5 derived radical in leghaemoglobin treated with peroxides. Bio- chem. J. 287,197-201. 135. Pfister, T.D., Gengenbach, A.J., Syn, S., and Lu, Y. (2001). The role of redox-active amino acids on compound I stability, substrate oxidation, and protein cross-linking in yeast cyto- chrome c peroxidase. Biochemistry 40,14942-14951. 136. Lardinois, O.M., Medzihradszky, K.F., and Ortiz de Montellano, P.R. (1999). Spin trapping and protein cross-linking of (he |ac- toperoxidase protein radical. J. Biol. Chem. 274,35441-35448. 137. Lardinois, O.M., and 0rti2 de Montellano, P.R. (2000). EPR spin-trapping of a myeloperoxidase protein radical. Biochem. Biophys. Res. Comm. 270,199-202. 138. Tobin, M.B., Gustafsson, C , and Huisman, G.W. (2000). Di- rected evolution; the 'rational' basis for 'irrational' design. Curr. Opin. Struct. Biol. 10, 421-427. 139. Bornscheuer, U.T., and Pon!, M. (2001). Improved btocatalysts by directed evolufíon and rational protein design. Curr. Opin. Chem. Biol. 5,137-143. 140. Celfk, A., Cuílis, P.M., Sutcíiffe, M.J., Sangar, R., and Raven, E.L. (2001). Engineering the active site oí ascorbate peroxi- dase. Eur. J. Biochem. 268, 78-85. 141. Kim, Y.H., Berry, A.H., Spencer, D.S., and Stites, W.E. (2001). Comparing the effect on protein stabiíity of methionine oxida- tion versus mirtagenesis: steps toward engineering oxidative resistance in proteins. Protein Eng. 14, 343-347. 142. ü , Q.S., Ogawa, J., and Shimizu, S. (2001). Critica! role of the resídue size at position 87 in H¡O2-dependent substrate hydroxylation activity and H?O? inactivation of cytochrome P450BM-3. Biochem. Biophys. Res. Comm. 280, 1258-1261. 143. Balagopalakrishna, C , Abugo, O.O., Horsky, J., Manoharan, P.T., Nagababu, E., and Rifkínd, J.M. (1998). Superoxide pro- duced in the heme pocket of the beta-chain of hemoglobin reacts with the beta-93 cysteine to produce a thiyl radical. Biochemistry 37, 13194-13202. 144. Miyazaki, C , and Takahashi, H. (2001), Engineering of the H2O2-binding pocket región of a recombinant manganese per- oxidase to be resistant to HaO2. FEBS Lett. 509, 111-114. 145. Wan, L., Twitchett, M.B., Eltis, L.D.. Mauk, A-G., and Smith, M. (1990). In vitro evolution of horse heart myoglobin to increase peroxidase activity. Proa Nati. Acad. Sci. USA 95, 12825- 1283í, 146. Iffland, A., Tafelmeyer, P., Saudan, C, and Johnsson, K. (2000). Directed molecular evolution of cytochrome c peroxidase. Bio- chemistry 39, 10790-10798. 147. Joo, H., Lin, Z., and Arnold, F.H. (1999). Laboratory evolution of peroxide-mediated cytochrome P450 hydroxylation. Nature 399, 670-673. 148. Dextef, A.F., and Hager, L.P. (1995). Transient heme n-alkyl~ ation of chloroperoxidase by terminal alkenes and alkynes. J. Am. Chem. Soc. 117, 817-818. 149. van deVelde, F., Bakker, M., van Rantwijk, F., Rai, G.P., Hager, L.P., and Sheldon, R.A. (2001). Engineering chloroperoxidase for activity and stability. J. Mol. Cat. B Enz. 11, 765-769. 150. Cherry, J.R., Lamsa, M.H., Schneider, P., Vind, J., Svendsen, A., Jones, A., and Pedersen, A.H. (1999). Directed evolution of a funga! peroxidase. Nat Biotechnol. / 7, 379-384. 151. Davies, M.J., and Dean, R.T. (1997). Radical-Mediated Protein Oxidation. From Chemistry to Medicine. (New York: Oxford Science Publications). 152. Wu, G., Vuletich, J.L., Kulmacz, R.J., Osawa, Y., and Tsai, A. (2001). Peroxidase seif-inactivation in prostaglandin H syn- thase-1 pretreated with cyclooxygenase inhibitors or substi- tuted with mangano porphyrin IX. J- Biol. Chem. 276,19879- 19888. 153. Henriksen, A., Mirza, O., Indiana, C Teilum, K., Smulevich, G., Welinder, K., and Gajhede, M. (2001). Structure of soybean seed coat peroxidase: a plant peroxidase with unusual stability and haem-apoprotein interactions. Protein Sci. 10, 108-115. 154. Ostergaard, L, Teilum, K., Mirza, O-, Mattsson, O., Petersen, M., Welinder, K.G., Mundy, J-, Gajhede, M., and Henriksen, A. (2000). Arabidopsis ATP A2 peroxidase. Expression and high- resolutíon structure of a plant peroxidase with implicatíons for lignification. Plant Mol. Biol. 44, 231-243. 155. Henriksen, A., Welinder, K.G., and Gajhede, M. (1998), Struc- ture of barley grain peroxidase refined at 1.9 Á resolution. A plant peroxidase reversibly inactivated at neutral pH. J. Biol. Chem. 273, 2241-2248. 156. Gajhede, M., Schuller, D.J., Henriksen, A., Smith, A.T., and Poulos, T.L. (1997). Crystal structure of horseradish peroxidase C at 2.15 Á resolution. Nat. Struct. Biol. 4, 1032-1038. 157. Schuller, D.J., Ban, N., Huystee, R.B., McPherson, A-, and Poulos, T.L. (1996). The crystal structure of peanut peroxidase. Structure 4, 311-321. 158. Mirza, O., Oestergaard, U Welinder, K.G., Henríksen, A., and Gajhede, M. (2000). Arabidopsis thaliana peroxidase N: struc- ture of a novel neutral peroxidase. Acta Crystallogr. D 56, 372-375. 159. Sundaramoorthy, M., Kishi, K., Gold, M.H., and Poulos, T.L, (1994). Preliminary crystallographic analysis of manganese peroxidase frorn Phanerochaete chrysosporium. J. Mol. Biol. 238,845-848. 160. Fukuyama, K., Kunishima, N., Amada, F., Kubota, T., and Mat- subara, H. (1995). Crystal structures of cyanide- and triiodide- bound forms of Arthromyces ramosus peroxidase at different pH valúes. Perturbations of active site residues and their impli- cation in enzyme catalysís. J. Biol. Chem. 270, 21884-21892, 161. Fitzgerald, M.M., Musah, R.A,, Mcree, D.E., and Goodin, D.B. (1996). A ligand-gated, hinged loop rearrangement opens a channel to a buried artificial protein cavity. Nat. Struct. Biol. 3, 626-631. 162. Patterson, W.R., and Poulos, T.L. (1995). Crystai structure of recombinant pea cytosolfc ascorbate peroxidase. Biochemis- try 34, 4331-4341. 163. Sundaramoorthy, M., Temer, J., and Poulos, T.L. (1995). The crystal struclure of chloroperoxidase: a heme peroxidase- cytochrome P450 functional hybrid. Structure 3,1367-1377. 164. Wüíiams, P.A., Cosme, J-, Sridhar, V., Johnson, E.F., and Mcree, D.E. (2000). Mammalian microsomal cytochrome P450 monooxygenase: structurai adaptatíons for membrane binding and functional diversity. Mol. Cell 5, 121-131. 165. Sevrioukova, I.F., Li, H,, Zhang, H., Peterson, J.A., and Poulos, T.L. (1999). Struclure of a cytochrome P450-redox partner electron-transfer complex. Proc. Nati. Acad. Sci. USA 96, 1863-1868. 166. Hasemann, C.A., Ravichandran, K.G., Peterson, J.A., and Deisenhofer, J. (1994). Crystal structure and refinement of cy- tochrome P450terp at 2.3 Á resolution. J. Mol, Biol, 236,1169- J185. 167. Park, S.Y., Shimizu, H., Adachi, S., Nakagawa, A., Tanaka, I., Nakahara, K., Shoun, H., Obayashi, E., Nakamura, H., lizuka, T., et al. (1997). Crystal structure of nitric oxide reducíase from denitrifying fungus Fusarium oxysporum. Nat. Struct. Biol. 4, 827-832, 168. Podust, L.M., Poulos, T.L., and Waterman, M.R- (2001). Crystal structure of cytochrome P450 14alpha-sterol demethylase (Cyp51) from Mycobacterium tuberculosis in complex with azole inhibitors. Proc. Nati. Acad. Sci. USA 98, 3068-3073. 169. Yano, J.K., Koo, L.S., Schuller, D.J., Li, H., Ortiz de Montellano, P.R., and Poulos, T.L. (2000). Crystal structure of a thermophilic cytochrome P450 from the archaeon Sulfolobus soifactaricus. J. Biol. Chem. 275, 31086-31092. 170. Poulos, T.L., and Howard, A.J. (1987). Crystal structures of metyrapone- and phenylimidazole-inhibited complexes of cy- tochrome P-450cam. Biochemistry 26, 8165-8174. 171. Lide, D.R. (2000). Handbook of Chemistry and Physics. (Boca Ratón, FL: CRC Press LLC). YBBRC 7370 DISK7 2/7/02 ARTICLE 1N PRESS ACADEMIC PRESS Biochemica! and Biophysícat Research Communications 295 (2002) 828-831 Cross-linked crystals of chloroperoxidasé Apéndice 7.6 No. of pages: 4 DTD 4.3.1/SPS BBRC www. acad emicp ress.com Marcela Ayala,* Eduardo Horjales, Michael A. Pickard, and Rífael V%quez-Duhalt Institute ofBiotechnology, UNAM. Apartado Postal 510-3, Cuernavaca, Received 14 June 2002 ),"ffl¿xico 6 Absfract íj 7 Chloroperoxidasé from Caldariomyces fumago was crystallized. The cry t̂ttĵ v/eiriií modifi'ed wíth several cross-linkers, but only 8 glurataldehyde was able to produce catalytically active and insoluble crystals.''tíí||Üce other immobilized chloroperoxidase prepa- 9 ratíons, these catalytic crystals are more thermostable than the unmodiñed soluble áríi^me. The enhanced stability is probably due 10 to the stmcture conservaüon in the crystalline matrix. In addition, non:cr.o.ss-linked chloroperoxidase crystals retained more activity 11 than the soluble enzyme after incubation in an organic solvent wiíM^iÓW^ater content. Although the cross-linked crystals were 12 catalytically active, they showed lower specific activity than the soluble enzy^he. This low activity may be due to non-speciflc re- 13 actions between the cross-Hnker and essenüal residues for cat¿lysig¿^Jterri|iitive cross-Hnking strategíes are discussed. © 2002 14 Elsevier Science (USA). All rights reserved. J?'''':\> '::l|;# 15 Keywords: Biocatalysis; Chloroperoxidasé; CLEC; Cross-linked 16 Protein stability is one of the major challeñges for 17 large-scale use of enzymes. Cross-linked enzyme ctystals,. 18 are a suitable preparation that confers s.|nicílír^....resi^;: 19 tance to proteins by stabíHzation of the cJyst^Jíné:MMx 20 [1,2]. In crystals, protein molecules aíi||5#nleliically 21 arranged and their native conformalbn is áíajbjjiî éd [3]. 22 These crystalline biocatalysts are i^óre É|ble tha-fí soluble 23 enzymes when exposed to orgánícllspj^iii&'and hígh 24 temperatures, condítions normaljy fou#d:;;:;in many in- 25 dustrial processes. Moreove;rí:cró'áS-linkeá'enzyme crys- 26 tais are mechanically resistánTi-Sgl'clrFbe recycled and 27 reused. Although there..:a :&;;py;erS£i£íí>ss-iÍnked crystals 28 commercially available|.most@Í|thern are prepared with 29 hydrolytic enz3Tnes;.bui|few are áyailable for other reac- 30 tions such as redoj-firbces&éfe;:;:;̂ ' 31 Chloroperoxid;|se txoríi-£aldariomyces fumago is the 32 most versatile^an%iínusua|:henie-peroxidase. In vitro, 33 cUoroperoxiátáe,;..snti^s:.;:li:álogenase-, peroxidase-, cata- 34 lase-, and:J::c:j^c:h||^ne P450-like activities [4j. This en- 35 zyme haá '^^ 'po té í í t i a l applications, ranging from 36 synthesisi^^f op'tyíajly puré compounds [5] to environ- 37 menjally relatad processes [6]. In spíte of thisf few re- 38 poílts-%a].ing i|ith cbJoroperoxidase immobilization are * Corrcsponding author. Fax: +52-777-317-2388. E-mail address: maa@ibt.unam.mx(M. Ayala). available [7-9], Immobilized forms of chloroperoxidasé have shown some advantages over the soluble form [10- 12], nevertheless no improved stability was conferred by the immobilization procedure. Chloroperoxidasé in both soluble and immobilized preparations is readily inacti- vated abo ve 50 °C [13], limíting its use in many fields. This is the flrst report of a cross-linked crystal of a peroxidase. The chloroperoxidasé cross-linked crystals showed lower activity but increased thermostabüity when compared to the soluble enzyme. Materials and methods Chemicals, 1,6-Hexandiamine, Í-ethyl-3-(3-dimethylaminopropyl) carbodiimide hydrochloride, glutaraldehyde, cacodyüc acid, thioani- sole, and thianthrenc were purchased from Sigma-AIdtich (St. Louis, MO). Zinc acétate was obtained from Fluka Chemie GmbH (Buchs, Switzerland). Polyethylene glycol 8000 was purchased from Hampton Reáearch (Laguna Niguel, CA). Buffer salts and organic solvents were obtained from J.T. Baker (Phillipsburg, NJ). Chloroperoxidasé from C. fumago CMI 89362 was obtained and purified as pteviously re- ported [14], The RJZ (^OS/^JBO) of the enzyme was 1.42. Crystallization of chloroperoxiáase. The enzyme was crystaliized in sitting drops using the vapor diflusion method. Thirty microliters of a 7 mg/mL chloroperoxidasé solution in 10 mM phosphate buffer, pH 5 was mixed with 30 uL of the crystallization solution. The crystalliza- tion solution contained 14% polyethylene glycol 8000, 0.1 M zinc ac- étate, and 0.1 M sodium cacodylate, pH 5.5. The reservoir contained 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 0006-291X/02/S - see front matter © 2002 Elsevier Science (USA). All rights reserved. PII: S0006-291X(02)00766-0 Apéndice 7.6 YBBRC 7370 DISK / 2/7/02 ARTICLE IN PRESS M. Ayala et al. I Biochemkal and Biophysical Research Communications 295 (2002) 828-331 No. of pages: 4 DTD4.3.1/SPS 829 65 1 mL of the crystallization solution. The drops were equilibrated for 66 15days at 18°C and microseeded with 1 \iL oí crushed crystals. Fine 67 needles 25-50 um wide grew under these conditions and after lOdays 68 the crystallization was complete. The crystals were recovered and 69 washed with a solution containing 25% polyethylene giycol 8000,0.1 M 70 zinc acétate, and 0.1 M sodium cacodylate, pH 5.5. 71 Chemical modification of crystals. To increase the number of amino 72 groups of chloroperoxidase moiecules inside the crystals, a chemical 73 modification was perfonned using carbodiimide chemistry. The crys- 74 tais were washed ín a solution containing 25% poíyethylene giycol 75 8000, 0.075M zinc sulfate, and 0.1 M sodíum cacodylate, pH 5.5. Zinc 76 acétate was replaced by the sulfate salt as carboxylate groups interfere 77 with the reaction. Hexandiamine (25OM excess) and carbodiimide 78 (500 M excess) were prepared separately and the pH was adjusted to 79 pH 5,5. The reactants were added to the crystals and the reaction 80 mixture was gently stirred for 2 h at room temperature. Crystals were 81 recovered by gentle centrifugation and washed with the original solu- 82 tion containing acétate salt. 83 Cross-linking of crystals. The crystals were washed in a solution 84 containing 25% polyethylene giycol 8000,0. í M zinc acétate, and 0.1 M 85 sodium cacodylate, pH 6.5. Different proportions of glutaraldehyde 86 were added and the reaction was gently stirred for 1 h at room tem- 87 perature. Crystals were recovered and washed as described above. 88 Enzymatic activity measurements. The oxidation of thioanisole and 89 thíanthrene was followed spectrophotometrically at 250 nm 90 (e = 8O5OM~lcnr') and 254nm (F, = 35,000M~ lcnr l) respectively. 91 Reactions were performed in lmL containing 20mM KC1 and 20% 92 rerí-butanol in 60mM citrate buffer, pH 3. The reaction was started by 93 adding 1 mM HiO2. The reaction was stopped after 1 mín by adding an 94 excess of Na2SC>3 and ínitial rates were calculated using the difference; 95 between the initial and final absorbance. H2C>2 dismutation due tejí, 96 catalase-like activity of chloroperoxidase was measured at pH 4 :by::' 97 foilowing the loss of absorbance at 240nm (R ~ 43.6M~1corl)ífÍ5]. 98 Residual activity was measured with monochlorodimedone.:i :ás.;:;de- 99 scribed by Hollenberg and Hager [16], Reported valúes are th&imeáivof;:; 100 three replicates. Crystals were recovered from the reactioníjiiixiure by' 101 centrifugation. For protein determination of the crystafé, a modified 102 versión of Lowry's method was used according to Spilbífeget aí. [17);:.. 103 Determination offree amino groups. A fluorescent.metwiii3iw.as usejj' 104 to measure free amino groups in chloroperoxidase Bnefly, aftet-:tne 105 chemical modification procedure the crystals wefe dissoí^éd in buffer. 106 The same amount of unmodified and modified protein Was treated with 107 o-phthaldialdehyde and the resulting complex was detecfcd as de- 108 scribed by Benson and Haré [18]. The ifegree •&£modification was 109 calculated based on the diiference betweeñ tn:&ñu.órescer¿ce' intensity of 110 the unmodified and modified enzyme. :... :iv;::,.,: 1 \ 1 Results and díscussion 112 Protein crystals aré;:.poroiis::::ínaterials in whích the 113 diffusion of small r#olé{|ites is póssible. To avoid mass 114 transfer limitatiotí's, micro^íysíals less than 100 um in 115 one-diraension al | . preferid [1,19]. Chloroperoxidase 116 was crystalLiz$| an%$ñj .ny^rocrystals retained catalytíc 117 activity whicn w6r.e sttitáble in size for biocataíytic ap- 118 plication^i^gyi^iííSymmetrical arrangement of mole- 119 cules cj3pfers:estructural stability to proteins [3]. The 120 beaefit'ÓF::an on%.éd matrix is reñected by the hígher 121 residual activity of non-cross-linked crystals after incu- 122 batíoMíjo/erí-bltitanol with low water content, as shown 123 in Tabte 'rB#" 124 Several bifunctional agents such as glutaraldehyde, 125 adipoyl chloride, and carbodiimide plus diamines of dif- KM-KMWW-. KÍ-̂ -iK-: .i**' Fig. 1. Scanning electrón micrograph of chloroperoxidase microcrys- tals. Table 1 '*%&& ,;•• Residual activity QÍ!:soluble and crystalline native chloroperoxidase after•J,h incubation in 99% íerí-butanol and 1% acétate buífer, pH 6 TempérS|bre (°Q Soluble enzyme Unmodified crystals :;;:;:. ::;:;: (% residual activity) (% residual activity) 30': 50 31 28 7 50 51 38 ferent length were assayed to stabilize the crystalline 126 matrix by intermolecular cross-linking. However, only 127 glutaraldehyde produced water-insoluble crystals. This 128 may be due to the ability of glutaraldehyde to form dif- 129 ferent sized polymers in solution, with the advantage that 130 the distance between two amino acids can be covered by 131 eme of those species [20]. In contrast, other cross-linkers 132 lacking this property might not have the appropriate 133 length to form bonds between two amino acids. 134 Glutaraldehyde is not a specific reagent for a par/tic- 135 ular amino acid. The reaction involves a nucleophilic 136 attack and generally lysines are the target for glutaral- 137 dehyde as they are strong nucleopbiles under alkaline 138 pH. However, because chloroperoxidase suffers an ir- 139 reversible inactivation at pH above 7.5, the cross-linking 140 reactions were performed at pH 6.5 under conditions 141 where lysines are protonated and are not strong nucle- 142 ophiles. The combination of these ínherent properties of 143 the enzyme resulted in a requírement of high concen- 144 trations of glutaraldehyde in the reaction to produce 145 insoluble crystals. Fig. 2A shows the decreased activity 146 of the crystals as the concentration of glutaraldehyde 147 increased. An amino acid analysis of cross-iinked crys- 148 tais indicated that not only lysines but also serines and 149 arginines were modified (data not shown). The enzyme 150 seems to be inactivated by glutaraldehyde reactions with í 51 amino acid residues which are essential for activity. 152 Apéndice 7.6 YBBRC 7370 DISK / 2/7/02 ARTICLE IN PRESS No, of pages: 4 DTD4.3.1/SPS 830 M. Avala ei al I Biochemica! andBiophyska! Research Communications 295 (2002) 828-831 c B"o Q. 2000 4000 6000 Molar excess of glutaraidehyde 'o o a. 03 cu c tu o 03 0 2000 4000 Molar excess of glutaraidehyde 6000 Fig. 2. Activity (A) and solubility (•) of unmodified (A) and chemicaliy modified (B) crystais after cross-linking with different amounts of> glutaraldehydc. ... 153 154 155 156 157 158 p 159 surface. Hexandiamine moÍetie1i^re::¿oValently coupled 160 to free carboxylic gro^|^li^:..cariil!^e[iímide chemistry. 161 The activity and soluü|J.íty pirSíiles ":bf the crystals after 162 treatment with differ:gn&|mpuntsji>f glutaraidehyde were 163 significantíy alterad* oncetM^emical modification was 164 performed (Fig. ;|B). Thé^chemical modification with There are five lysines ín chíoroperoxiBs|^.onlyjithree of them are on the surface of the .píi||e,in áft^^fey are not homogéneously distributed;:;[|l]/||g,,,e.^¡í:Snce the intermolecular cross-linking and reai^rín§i:íñactivation of the enzyme, a chemical m$¡iy|ic.ation::;:%as performed to increase the amount of pf|ino'^Pjjj!§3n the protein Table 2 ^ Kinetic constante for tite oxidation of thianthrene and thioanisole, and the dismutation of H22 hexandiamine increased the number of free amino 165 groups per chloroperoxidase molecule to 10, and the 166 activity retained in the crystals increased; nevertheless, a 167 major portion of the enzyme was still inactivated. 168 Kinetic constants of cross-linked crystals and soluble 169 chloroperoxidase for the sulfoxi4ation of thianthrene and 170 thioanisole and the dismutatioj!. of'p^C^ are shown in 171 Tabie 2. Thianthrene and thíóSn||§le are model com- 172 poundsforsulfur-containing^g^nic'nt^leculespresentin 173 fuels such as diesel; the ox|.datioh%£.suc:h compounds is a 174 key step in enzymatic bíoá|sulfurization processes [6,22]. 175 The activity of the sr^^lmli'lil^E^ystals was lower than 176 that of the solublejenzyti^.fíowever, the activity was 177 íowest for the more -tfoluminófs substrate and it increased 178 as the size of íh^suflli^íáécreased. For thianthrene, 179 thioanisole, |fíd : ' í | |p2 th'e cataíytic efficiency of cross- 180 ünked ci^stáÍ^:íi|a§:píie^tWo, and one order of magnitude 181 lower th^rííí3|^í: oíll^Dluble enzyme, respectively. These 182 results,;|^gjhí-|reflecfaccessibility problems for bulky 183 substrates i^l^e interior of the crystals, thus lowering the 184 cataíytic efficíéi|y of the cross-linked crystals. 185 These cross-linked crystals of chloroperoxidase 186 showed improved stability to temperature. Fig. 3 shows 187 ,,the; residual activity of the crystalline and soluble en- 188 zyía'|::::af|e'r 1 h of incubation at different temperatures. 189 i;;Cte:aríy;::the crystals retained most of its cataíytic activity 190 '..at ||mperatures up to 70 °C. This results from the 191 Estructural stability of protein molecules inside the crys- 192 tálline array. The increased stability may come from 193 :both the pre-ordered arrangement of the raolecules and 194 the rigidity of the three-dimensional structure of mole- 195 cules caused by cross-linking. In solution and in the 196 absence of such stabilizing factors, the native confor- 197 mation is readíly lost and enzyme is inactivated. 198 A more refined cross-linking strategy could be envis- 199 aged as an alternative to stabilize chloroperoxidase crys- 200 tais. Chloroperoxidase contains a large number of 201 superficial aspartic and gíutamic amino acids. It could be 202 feasible to design a bifunctional amine to form intermo- 203 lecular bonds between carboxylic groups using carbodü- 204 mide as a zero-length cross-linker. The three-dimensional 205 arrangement of protein molecuíes in the crystal musí be 206 known to carry out such procedure. Based on the infor- 207 mation given by X-ray díffracüon, it might be possible to 208 find an appropriate diamine to cover the distance between 209 two carboxylic groups in adjacent protein molecules. 210 Substrate :S(JÍuble Cross-linked crystals (mM) Thíantfiréíiéí Thioanisole 555 4170 1670 0.0033 1.75 10 1.6 x 10* 2.3 x I06 1.6 x 105 0.85 28 55 0.0073 2.2 4.6 1.1 x 10s 1.2 x 104 1.2 x 10" Apéndice 7.6 YBBRC 7370 DISK / 2/7/02 ARTICLE IN PRESS M. Ayala et al. I Biockemical and Blophysical Research Commttnicaüons 295 (2002) 828-831 No. of pages: 4 DTD4.3.1/SPS 83 i 60 70 Temperatura (°C) Fig. 3. Residual activity of cross-ünked crystals of diloroperoxidase (G) and soluble cozyiac {•) after a lh iucubation at differeat tein- peratures. 211 In conclusión, this is the first report of cross-linked 212 crystals of a peroxidase. Unmodified crystals of chlor- 213 operoxidase retained more activity than soluble enzyme 214 upon exposure to a water-miscible organic solvent with 215 low water contení, which refleets the beneñt of an or- 216 dered arrangement of molecules. In contrast to other| 217 immobilization procedures, crystal cross-linking yielddáíí 218 a chloroperoxidase preparatíon with enhanced therjiial 219 resistance. Substrate accessibility probleras may.r;é%c_.e 220 the nuraber of effective active sites available fqf %taípi 221 sis, thus lowering the observed specific actiyííy'bf the 222 cross-linked crystals. The crystals lost activity|as a residí. 223 of the cross-linking treatment with | f 224 Chloroperoxidase crystal structure 225 rently used to design a better c r o s ^ ; j ^ 226 the activity loss. ..-,:,:. '::-:'%-,,# 227 Acknowledgments "'•^••Í:.. 228 This work was supported by a¡íg%#t fiÉí^lthfeltfexican Petroleum 229 Instituto (FIES 98-ílO-VI) and from the^tioi íal Council for Science 230 and Technology of México Í : 231 232 233 234 235 236 237 [3] B. Shenoy.Y. Wang, W. Shan, A.L. Margolin, Stabüity of crystaliine proteins, Biotechnol. Bioeng. 73 (2001) 358-369. [4] S. Colonna, N. Gaggeto, C. Richelmi, P. Pasta, Recent biotech- nological developments in the use of peroxidases, Trends Bio- technol. 17 (1999) 163-168. [5] F. van Rantwijk, R.A. Sheldon, Selective oxygen transfer catal- ysed by heme peroxidases: synthgífe.:,and mechanistic aspeets, Curr. Opin. Biotechnol. II (20OO)|54-5^. [6] M. Ayala, R. Tinoco, V. Hemaá&ésigi?;,;'Srémauntz, R. Vázquez- Duhalt, Eiocatalytic oxidation. of 'fiííife.as .an altcrnative to biodesulfurization, Fuel Prpc^v:t%:hnoÍ;::3í:(1998) 101—111. [7] T.A. Kadima, M.A. Picka^ Immobíti^ation of chloroperoxidase on aminopropyl-glass, Apptegnviron. JÍJícrobiol. 56 (1990) 3473- 3477. ,;; . ;^:v' : :%i;:,:::/ [8] S. Aoun, C. Cheblíi :;M'."éái|.Qulehé, Ñoncovalent immobilization of chloroperoxida^.. onto talg:; catalytic properties of a new biocatalyst, Enzymé%;(grob. T^-hnol. 23 (1998) 380-385. [9] M. Bakker, Ji&san iÉ::^dB¿í:F. van Rantwijk, R.A. Sheldon, Highly effieiént íSmobilization of glycosylated enzymes into polyurethá^:%!a^i^pihnol . Bioeng. 70 (2000) 342-348. [10] S. Aoun,£;jM. Ba-üífjjlgn?, Regioselective bromohydroxylation of alkenej ca'tat^?ed by^nloroperoxídase: advantages of the immo- biliza:t%::í?f|nzyine on tale, J. Mol. Catal. B: Enzymatic 4 (1998) Reférences ..:;:•:.... : ' -# s . : ^ [1] A.L. Margolin, l%yel crystai|ine catalysts, Trends Biotechnol. 14 (1996) 223-2|0. %,,. . . f' [2] M.A. Naviá;::M.J-;. 'sÍÉÉÍMf J.P. Griffith, Crosslinked enzyme crystals as.immobiliKd enzyme particles, in: W.J. J. van den Tweel, A. Haráéííi;R^ÍíiÜuí£elaaE (Eds.), Stabüity and Stabilization of Enzyjies, Éís'éve.ir, Maastricht, The Netherlands, 1992, pp. 63-73. [11] F. van de Véíp,' M. Bakker, F. van Rantwijk, R.A. Sheldon, ... ...Qhloroperoxidáse-catalyzed enantioselective oxidations in hydro- .f mf&¿>Mc organic media, Biotechnol. Bioeng. 72 (2001) 523-529. ¿y 2] F. 'v3|;.de Velde, N.D. Lourenco, M. Bakker, F. van Rantwijk, i:íífeííP..A. Sheldon, Improved operational stabüity of peroxidases by "eQimjnbbilization with' glucose oxidase, Biotechnol. Bioeng. 69 •#>.. ,(2¿p) 286-291. [13] 'ÍSÍÍ.Á. Pickard, A. Hashimoto, Stability and carbohydrate com- ;::;:;;:•::,;-íposiÜon of chloroperoxidase from Caldariomyces fumago in a '!F fructose-salts médium, Can. J. Microbiol. 34 (1988) 998-1002. [14] M.A. Pickard, T.A. Kadima, R.D. Cannichael, Chloroperoxidase, ;; a peroxidase with potential, J. Ind. Microbiol. 7 (1991) 235-242. [15] M.B. Araño, M. Acosta, J.A. del Río, R. Varón, F. García- Cánovas, A kinetic study on the suicide inactivation of peroxidase by hydrogen peroxide, Biochim. Biophys. Acta 1041 (1990) 43-47. [16] P.F. Hollenberg, L.P. Hager, Purification of chloroperoxidase from Caldariomyces fumago, Methods Enzymol. 52 (1978) 521-529. [17] C.A. Spilburg, J.L. Bethune, B.L. Vallce, Kinetic properties of crystaUine enzymes. Carboxypeptidase A, Biochemistry 16 (1977) 1142-1150. [18] J.R. Benson, P.E. Haré, o-Phthataldehyde: fluorogenic detection of primary amines in the picomole range. Comparison with fluorescamine and ninhydrin, Proc. Nati. Acad. Sci. USA 72 (1975) 619-622. [19] W. Tischer, V. Kasche, Immobilized enzymes: crystals or carriers? Trends Biotechnol. 17 (1999) 326-335. [20] D.R. Walt, V.I. Agayn, The chemistry of enzyme and protein immobilization with glutaraldehyde, Trends Anal. Chem. 13 (1994) 425-430. [21] M. Sundaramoorthy, J. Temer, T.L. Poulos, The crystal structure of chloroperoxidase: a heme peroxidase-cytochrome P450 func- tional hybrid, Structure 3 (1995) 1367-1377. [22] M. Ayala, N.R. Robledo, A. Lopez-Munguia, R. Vázquez- Duhalt, Substrate specificity and ionization potential in chlorop- eroxidase-catalyzed oxidation of diesel fuel, Environ. Sci. Tech- nol. 34 (2000) 2804-2809. 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297