( PDF ) Chiquimato desidrogenase de Mycobacterium tuberculosis: mecanismos cinético e químico de enzima recombinante

Download PDF

ads:

UNIVERSIDADE FEDERAL DO RIO GRANDE DO SUL

PROGRAMA DE PÓS-GRADUAÇÃO EM CIÊNCIAS BIOLÓGICAS:

BIOQUÍMICA

Chiquimato Desidrogenase de Mycobacterium

tuberculosis: mecanismos cinético e químico da enzima

recombinante

Isabel Osorio da Fonseca

Orientador: Prof. Dr. Luiz Augusto Basso

Tese apresentada ao Programa de Pós-Graduação em Ciências

Biológicas: Bioquímica da Universidade Federal do Rio Grande do Sul

como pré-requisito para obtenção do grau de Doutor.

Porto Alegre, agosto de 2006.

ads:

Livros Grátis

http://www.livrosgratis.com.br

Milhares de livros grátis para download.

Esta tese foi julgada e aprovada para a obtenção do grau de

doutor em Ciências Biológicas: Bioquímica pelo orientador e pela

comissão examinadora no programa de Pós – Graduação em

Ciências Biológicas: Bioquímica.

Orientador:

Prof. Dr. Luiz Augusto Basso

Faculdade de Biociências

Pontifícia Universidade Católica do Rio Grande do Sul

Comissão Examinadora:

Dr. Carlos Alberto Saraiva Gonçalves

Membro do Programa de Pós-Graduação em Ciências Biológicas:

Bioquímica

Universidade Federal do Rio Grande do Sul

Dr. Arthur Germano Fett-Neto

Membro do Programa de Pós-Graduação em Biologia Celular e

Molecular

Universidade Federal do Rio Grande do Sul

Dr. Walter Filgueira de Azevedo Jr.

Departamento de Ciências Fisiológicas da Faculdade de Biociências

Pontifícia Universidade Católica do Rio Grande do Sul

ads:

Este trabalho foi realizado no Centro de Pesquisas em Biologia

Molecular e Funcional, Instituto de Pesquisas Biomédicas,

TECNOPUC, Pontifícia Universidade Católica do Rio Grande do Sul,

sob orientação do Prof. Dr. Luiz Augusto Basso.

iii

Este trabalho é dedicado...

ao meu marido, Kiko, por todo o amor, compreensão, força,

carinho e apoio incondicional.

ao meu filho, Jota, pela compreensão, carinho, paciência e apoio

que me deram forças para vencer mais essa etapa.

A eles, meu MMMUUUUIIIIITTTTOOOO OBRIGADO por me

amarem e estarem ao meu lado...

Agradecimentos

Ao meu orientador Dr. Luiz Augusto Basso, pelos ensinamentos, pela

confiança... sinceramente... MUITO OBRIGADO!!!

Ao Dr. Diógenes Santiago Santos, pela oportunidade, pela confiança, pelos

ensinamentos, pelo exemplo... MUITO OBRIGADO POR TUDO!!!

Ao meu colega Rafael pela enorme força para o desenho dos experimentos,

análise dos resultados, por todos os ensinamentos... MMUUIIITTTTOOOOO

OBRIGADO!!!

A colega Cláudia pelos ensinamentos em bioinformática... MUITO OBRIGADA!!!

Aos meus pais pela força, pelo exemplo, pelo amor, por acreditarem em mim...

AMO MMUUIIITTTTOOOOO VOCÊS!!!

Aos meus colegas, Cristopher, Clarissa, Jotinha, Bruna, Raquel, Zé, Fer,

Isabelinha... pela amizade, pelas contribuições, pelo apoio, pelas brincadeiras,

pelas conversas... que tornaram ainda mais agradáveis meus dias... MUITO

OBRIGADO!!!

À Clotilde, Renilda e Gleci pela dedicação zelando sempre a manutenção e

funcionamento do laboratório... MUITO OBRIGADO!!!

Ao Programa de Pós-Graduação em Ciências Biológicas: Bioquímica da

Universidade Federal do Rio Grande do Sul pela oportunidade que me foi dada.

À CAPES pelo apoio financeiro que me auxiliou nesse período.

Índice

Parte I 1

Resumo 2

Abstract 3

Lista de Abreviaturas 4

Introdução 5

Tuberculose 5

Via do Chiquimato e Chiquimato desidrogenase 11

Objetivo 19

Parte II 21

Capítulo I: 22

Functional shikimate dehydrogenase from Mycobacterium tuberculosis

H37Rv: purification and characterization.

Capítulo II: 23

Shikimate Dehydrogenase from Mycobacterium tuberculosis H37Rv:

Kinetic and Chemical Mechanisms.

Parte III 24

Discussão

Capítulo I: 25

Functional shikimate dehydrogenase from Mycobacterium tuberculosis

H37Rv: purification and characterization.

Capítulo II: 35

Shikimate Dehydrogenase from Mycobacterium tuberculosis H37Rv:

Kinetic and Chemical Mechanisms.

Conclusão 50

Referências 51

Anexos

Anexo I 55

1. AZEVEDO, WF; SILVA, RG; FONSECA, IO; RENARD, G; BASSO, LA;

SANTOS, DS. (2003) Docking and small angle X-ray scattering studies of

purine nucleoside phosphorylase. Biochemical and Biophysical Research

Communications 309, 928-933.

Anexo II 56

2. POLETTO, SS; FONSECA, IO; CARVALHO, LP; BASSO, LA; SANTOS, DS.

(2004) Selection of an Escherichia coli host that expresses mutant forms

of Mycobacterium tuberculosis 2-trans enoyl-ACP(CoA) reductase and 3-

Ketoacyl-ACP (CoA) reductase enzymes. Protein Expression and Purification

34, 118-125.

Anexo III 57

3. RIZZI, C; FONSECA, I.O.; GALLAS, M; WEBER, P; FRAZZON, J; SANTOS,

DS; BASSO, LA. (2005) DAHP synthase of Mycobacterium tuberculosis

H37Rv: Cloning, overexpression, and purification of the functional

enzyme. Protein Expression and Purification 40, 23-30.

Anexo IV 58

4. PATENTE (2005) SANTOS, D. S.; BASSO, L.A.; RENARD, G.; FONSECA,

I.O.; CHIES, J.M. Método de Obtenção de Seqüências Nucleotídicas

Quiméricas e Seqüência Nucleotídica Quimérica - INPI Registro PI0506047-

Anexo V 59

5. LEOPOLDINO, A.M.; CANDURI, F.; CABRAL, H.; JUNQUEIRA, M.;

MARQUI, A.B.T.; APPONI, L.H.; FONSECA, I.O.; DOMONT, G.B.; SANTOS,

D. S.; VALENTINI, S. (2006) Expression, purification, and circular

dichroism analysis of human CDK9. Protein Expression and Purification 47,

614-620.

vii

PARTE I

Resumo

O aumento na prevalência da tuberculose (TB), a emergência de cepas resistentes a múltiplas

drogas de Mycobacterium tuberculosis, o agente etiológico da TB, e o efeito devastador da co-

infecção pelo HIV têm enfatizado a urgente necessidade no desenvolvimento de novos agentes

antimicobacterianos. A análise completa da seqüência genômica do M. tuberculosis H37Rv

mostrou a existência de genes envolvidos na via de biossíntese de aminoácidos aromáticos, a

via do chiquimato, e evidências experimentais indicam que esta via é essencial para o M.

tuberculosis e apresenta-se ausente no homem. Os produtos dos genes que são essenciais

para o crescimento dos microrganismos fazem deles alvos atrativos para a ação de drogas,

pois a inibição de sua função pode matar o bacilo. Previamente, nosso grupo relatou a

clonagem do gene aroE de M. tuberculosis e a expressão do seu produto na forma solúvel, a

enzima chiquimato desidrogenase (MtbSD), que catalisa o quarto passo na via do chiquimato.

Neste trabalho apresentamos, no primeiro artigo, intitulado “Functional shikimate

dehydrogenase from Mycobacterium tuberculosis H37Rv: purification and characterization”, a

purificação da MtbSD solúvel e ativa por cromatografia líquida, o seqüenciamento do N-

terminal, a espectrometria de massas, a determinação do estado oligomérico por cromatografia

em gel filtração, a estabilidade térmica, a determinação de parâmetros cinéticos aparentes em

estado estacionário para as reações direta e reversa, a constante de equilíbrio aparente e

energia de ativação para a reação química catalisada pela MtbSD. No segundo artigo, intitulado

“Shikimate Dehydrogenase from Mycobacterium tuberculosis H37Rv: Kinetic and Chemical

Mechanisms”, descrevemos os parâmetros de velocidade inicial na reação direta, os estudos de

inibição pelos produtos, os efeitos isotópicos cinéticos primários do deutério, os efeitos

isotópicos cinéticos do solvente, os efeitos isotópicos cinéticos múltiplos, proton inventory, o

efeito de pH na química ácido/base para a ligação e catálise dos substratos, e a estrutura 3D da

MtbSD obtida in silicon pela modelagem por homologia. Esses resultados servem como base

para os estudos cinéticos e estruturais que podem auxiliar no desenho racional de inibidores a

serem testados como agentes antimicobacterianos.

Abstract

The increasing prevalence of tuberculosis (TB), the emergence of multidrug-resistant strains of

Mycobacterium tuberculosis, the causative agent of TB, and the devastating effect of co-

infection with HIV have highlighted the urgent need for the development of new

antimycobacterial agents. Analysis of the complete genome sequence of M. tuberculosis H37Rv

shows the presence of genes involved in the aromatic amino acid biosynthetic pathway, the

shikimate pathway, and experimental evidence showed that this pathway is essential for M.

tuberculosis and it is absent in humans. The gene products that are essential for the growth of

the microorganisms make them attractive drug targets since inhibiting their function may kill the

bacilli. Our group has previously reported the cloning of M. tuberculosis aroE gene and the

expression of the product in the soluble form, the shikimate dehydrogenase (MtbSD) enzyme,

that catalysis the fourth reaction in the shikimate pathway. Currently, in the first manuscript

“Functional shikimate dehydrogenase from Mycobacterium tuberculosis H37Rv: purification and

characterization”, we present the purification of soluble and active MtbSD, N-terminal

sequencing, mass spectrometry, assessment of the oligomeric state by gel filtration

chromatography, thermal stability, determination of apparent steady-state kinetic parameters for

both forward and reverse directions, apparent equilibrium constant, and energy of activation for

the enzyme-catalyzed chemical reaction. In the second manuscript “Shikimate Dehydrogenase

from Mycobacterium tuberculosis H37Rv: Kinetic and Chemical Mechanisms”, we describe the

kinetic mechanism, initial velocity patterns in the forward direction, product inhibition studies,

primary deuterium kinetic isotope effects, solvent kinetic isotope effects, multiple isotope effects,

proton inventory, pH rate profile and the MtbSD 3D structure obtained in silicon by homology

modeling. These results should be useful as a solid base for structural and kinetic studies, which

can aid in the rational design of inhibitors to be tested as antimycobacterial agents.

Lista de Abreviaturas

DAHP

3-desoxi-D-arabinoheptulosonato-7-fosfato

DHS

3-desidrochiquimato

EcoliQSD quinato-chiquimato desidrogenase de Escherichia coli

EcoliSD chiquimato desidrogenase de Escherichia coli

EPSP 5-enolpiruvilchiquimato-3-fosfato

E4P eritrose-4-fosfato

HinflSD chiquimato desidrogenase de Haemophilus influenzae

IPTG β-D-tiogalactopiranosídeo

MjanSD chiquimato desidrogenase de Methanacoccus jannaschii

MtbSD chiquimato desidrogenase de Mycobacterium tuberculosis

PEP Fosfoenolpiruvato

PsatSD chiquimato desidrogenase de Pisum sativum

SD chiquimato desidrogenase

SHK

chiquimato ou ácido chiquímico

TB tuberculose

Introdução

Tuberculose

A tuberculose (TB) é a doença infecto-contagiosa responsável pela maior parte

da mortalidade humana causada por um único agente infeccioso, o Mycobacterium

tuberculosis (ENARSON & MURRAY, 1996).

No século XIX, a TB já era uma doença devastadora, com taxa de transmissão e

índice de morbidade altíssimos; período que corresponde ao começo da revolução

industrial, que trouxe como conseqüência a formação de aglomerados urbanos com

péssimas condições de higiene e habitação (DORMANDY, 2000).

A infecção ocorre por meio da inalação do bacilo, presente em partículas

infectivas, expelidas por pacientes com tuberculose ativa através da tosse ou espirro.

Elas apresentam dimensão suficiente para permanecerem suspensas no ar de

ambientes fechados e assim alcançar os alvéolos pulmonares de outros indivíduos

(DUNLAP e cols., 2000; GLICKMAN & JACOBS, 2001).

Após a infecção, a doença é regulada fundamentalmente pelo sistema imune do

hospedeiro, que determina se o microrganismo é eliminado, condicionado a um estado

de latência, ou se está propenso a desenvolver a doença ativa no seu hospedeiro.

Assim que a micobactéria chega aos bronquíolos e alvéolos, os macrófagos

fagocitam o invasor. A partir daí, a infecção irá ou não se estabelecer dependendo da

virulência bacteriana e da capacidade bactericida dos macrófagos do hospedeiro. Caso

a micobactéria não seja eliminada pela primeira linha de defesa do organismo, a

erosão da bactéria no interstício pulmonar estabelecerá uma inflamação local; logo, o

dano tissular, a formação dos granulomas e a agregação destes formarão o tubérculo -

a característica nodular da lesão de TB (MILBURN, 2001). Aproximadamente 4 meses

após a infecção, ocorre a eliminação da maior parte dos bacilos e cessa a infecção

primária.

A tuberculose conhecida como tuberculose extra-pulmonar ou miliar, representa

15% dos pacientes com a infecção ativa. Ela é causada pelo crescimento bacteriano

excessivo no granuloma, que atinge então a corrente sangüínea e disseminando-se

para a pleura, linfonodos, fígado, baço, ossos e articulações, coração, cérebro, sistema

genito-urinário, meninges, peritônio ou na pele.

Uma outra possibilidade é a coexistência pacífica entre invasor e hospedeiro

humano, que recebe a denominação de infecção latente. Nesse caso se estabelece um

reservatório bacteriano no indivíduo infectado, onde o metabolismo bacteriano está

reduzido como conseqüência da ação do sistema imunológico do hospedeiro. No

momento em que o sistema imune do hospedeiro apresentar falhas, como subnutrição,

câncer, diabetes, quimioterapia, entre outros, pode ocorrer a reativação do bacilo,

levando o indivíduo a desenvolver TB ativa (PARRISH e col., 1998).

O M. tuberculosis adapta-se a uma variedade de condições ambientais, dentro e

fora do hospedeiro, sendo então considerado o patógeno de maior sucesso da história.

Esta característica peculiar do bacilo faz com que a TB permaneça como uma das

principais causas de morte no mundo (ENSERINK, 2001). A exata forma como o

microrganismo permanece silencioso no corpo humano, em trégua com o sistema

imune, tem sido um longo mistério (WICKELGREN, 2000).

A TB começou a ser esclarecida no meio científico em 1882, quando no IV

Congresso Mundial de TB, o médico e bacteriologista alemão Robert Koch tornou

pública a identificação do agente etiológico da doença. Este marco da história científica

tornou possível a procura de terapias eficazes no tratamento de pacientes com TB.

A descoberta da estreptomicina, em 1944, marcou a era de ouro no

desenvolvimento de drogas para TB. Outras drogas, ainda hoje utilizadas para o

tratamento da doença, foram sendo introduzidas: o ácido p-aminosalicílico (1946),

isoniazida (1952), ciclosserina (1955), canamicina (1957), rifampicina (1965),

etionamida (1966), etambutol (1968) e pirazinamida (1970) (DUNCAN, 2003). A

administração de estreptomicina, isoniazida e ácido para-amino-salicílico contra a

doença ativa reduziu consideravelmente a mortalidade por TB (BLOOM & MURRAY,

1992; DANIEL, 1997). Essas drogas e as vacinas de prevenção da infecção trouxeram

um período de relativa tranqüilidade com respeito à doença.

Em 1993, a TB foi declarada “emergência de saúde global” pela Organização

Mundial da Saúde (OMS). Estima-se que anualmente ocorram 8 milhões de novos

casos e 3 milhões de mortes devido a esta doença (DUNLAP e cols., 2000). Em torno

de um terço da população mundial está hoje infectada pelo M. tuberculosis e 10%

deste número desenvolverá a doença ao longo da vida (ENSERINK, 2001). Além disso,

a OMS e o Centro de Controle e Prevenção de Doenças dos Estados Unidos indicam a

TB como responsável por 27% de todas as mortes evitáveis em todo o mundo

(SEPKOWITZ e cols., 1995). Estima-se que até 2020, um bilhão de pessoas se

tornarão infectadas, 200 milhões de pessoas desenvolverão a doença e 70 milhões

morrerão devido à TB se a atual tendência permanecer inalterada (PASQUALOTO &

FERREIRA, 2001).

A argumentação de que a tuberculose é uma doença reemergente é válida

apenas para alguns países europeus e para os Estados Unidos; pois no Brasil, a TB

não é um problema reemergente e sim um problema persistente (RUFFINO-NETTO,

2002).

Os fatores responsáveis pelo ressurgimento da TB, até então considerada

vencida em países desenvolvidos são: (1) advento da Síndrome da Imunodeficiência

Adquirida (AIDS); (2) deterioração dos programas de saúde pública visando o controle

da TB; (3) aumento da transmissão do M. tuberculosis em hospitais e prisões; e (4)

surgimento de cepas resistentes a múltiplas drogas (MDR, resistentes a pelo menos

isoniazida e rifampicina) (ESPINAL, 2003).

A TB e a AIDS são referenciadas como co-epidemia, visto que nesta situação a

probabilidade de desenvolvimento da doença aumenta até 30 vezes. A taxa de

reativação de TB latente em pacientes HIV-positivos é de 2,3 a 13,3%, dependendo do

nível de imunossupressão, contra 0,2% em pessoas imunocompetentes (KNIGGE e

cols., 2000). Pacientes HIV-positivos infectados com M. tuberculosis progridem para TB

ativa em uma taxa de 37% nos primeiros 6 meses, ao contrário de pacientes

imunocompetentes cuja taxa é de 2 a 5% nos primeiros dois anos (SEPKOWITZ,

1995). Esse aumento considerável na probabilidade leva a uma aceleração da

transmissão do agente etiológico (KNIGGE e cols., 2000) e, portanto, a co-infecção TB

- HIV representa um problema de efeito devastador tanto para pacientes infectados

como para a população em geral.

O aumento da transmissão do M. tuberculosis em hospitais e prisões ocorre

devido aos aglomerados humanos com condições de higiene e habitação péssimas.

Um agravante deste quadro decorre da deterioração dos sistemas de saúde visando o

controle da TB (FÄTKENHEUER e col., 1999).

A TB resistente a múltiplas drogas (MDR-TB – “multidrug-resistant tuberculosis”)

está relacionada à resistência aos medicamentos para combater a doença, visto que

ocorre a resistência à pelo menos rifampicina e isoniazida. Em 2000 foi estimado que

3,2% dos novos casos são de MDR-TB (Espinal, 2003). Segundo a OMS, ocorrem

300.000 novos casos de MDR-TB anualmente, onde 79% são de “super cepas”, que se

refere às cepas resistentes a pelo menos três das quatro principais drogas usadas no

tratamento de TB (WHO, 2004).

Através de um estudo realizado em 35 países pela OMS e a União Internacional

Contra a Tuberculose e Doenças Pulmonares (IUATLD – “International Union Against

Tuberculosis and Lung Disease”) foi demonstrado que a doença consiste em um

problema global, pois todas as regiões analisadas apresentaram linhagens de M.

tuberculosis com resistência à pelo menos uma das drogas antituberculose, sendo

normalmente a isoniazida ou estreptomicina (PABLOS-MENDEZ e col., 1998).

O MDR-TB surge sob pressão seletiva da quimioterapia inapropriada ou da não

adesão ao tratamento (RILEY, 1993). Terapia inadequada compreende o uso de uma

única droga, combinações inadequadas de drogas, má-absorção de drogas

administradas, etc. Nesses casos o M. tuberculosis estará exposto a concentrações

sub-letais de compostos bactericidas, quando as cepas mais resistentes na população

bacteriana sobreviverão e superarão a cepa sensível (PETRINI & HOFFNER, 1999).

No caso da co-epidemia de HIV-TB, o aparecimento das linhagens MDR-TB foi

favorecido pelo maior índice de abandono do tratamento (BRENNAN, 1997).

Atualmente o tratamento recomendado pela OMS consiste na administração

combinada dos antibióticos isoniazida, rifampicina, pirazinamida e estreptomicina (ou

etambutol) durante 2 meses, seguida pela combinação de isoniazida e rifampicina por

pelo menos mais 4 meses.

Mesmo com um tratamento bem estabelecido, a eficácia do tratamento contra o

bacilo de Koch através da utilização de antibióticos tem sido dificultada, em parte,

principalmente pela oposição do paciente em cumprir com o tratamento de 6 meses

(YOUNG, 1998). A longa duração e os desagradáveis efeitos colaterais do tratamento

levaram a OMS ao desenvolvimento de um programa conhecido como DOTS (directly

observed treatment short-course) que visa acompanhar a adesão dos pacientes ao

tratamento, monitorar seu progresso e observar a ingestão de cada dose da medicação

(NSB Editorial comment, 2000).

O programa DOTS combina cinco elementos: compromisso político, serviços de

microscopia, suprimento de medicação, sistemas de monitoramento e observação

direta do tratamento (PASQUALOTO & FERREIRA, 2001). Entretanto, em alguns

países super populosos o programa não consegue alcançar 100% dos indivíduos

infectados, não chegando aos objetivos primeiramente propostos. Uma alternativa é a

obtenção de drogas que simplifiquem o tratamento da TB facilitando a execução do

programa DOTS (DUNCAN, 2003).

Desenvolvimento de Agentes anti-TB

Frente a essa exposição de fatos que envolvem a TB, faz-se necessária à

introdução de novas drogas anti-TB, pois desde a metade da década de 70 não houve

a descoberta, nem tampouco o desenvolvimento de novas drogas de primeira linha que

possam ser utilizadas no combate ao bacilo causador da TB (PETRINI & HOFFNER,

1999).

As novas drogas teriam como objetivo a efetividade no tratamento de MDR-TB e

TB latente, além da redução no tempo do tratamento (O´BRIEN & NUNN, 2001). Isto

representaria vantagem significativa em relação às drogas atualmente utilizadas.

DUNCAN (2003) estima que o desenvolvimento completo de uma nova droga anti-TB

custe de 100 a 800 milhões de dólares. Considerando-se que 95% dos novos casos de

TB ocorram em países em desenvolvimento, poucos pacientes poderiam pagar os altos

custos necessários para o lucro da indústria farmacêutica (O´BRIEN & NUNN, 2001). O

provável retorno financeiro é baixo e, conseqüentemente, o interesse neste setor

também é baixo (DUNCAN, 2003).

Mais de um século após a descoberta de Robert Koch, está disponível uma nova

perspectiva para o desenvolvimento de novas drogas, em decorrência da seqüência

completa do genoma do M. tuberculosis, marcando uma nova era na batalha contra

esse patógeno (COLE e cols., 1998).

Utilizando a análise de homologia de seqüências, foi possível a identificação de

genes estruturais envolvidos em rotas metabólicas. A identificação e validação de rotas

essenciais ao microrganismo, ausentes no hospedeiro, são o primeiro passo no

desenvolvimento de inibidores específicos de baixa toxicidade (DUNCAN, 2003).

Alguns genes expressos na infecção latente foram descritos, e suas proteínas podem

ser bons alvos para o desenvolvimento de drogas para o tratamento da TB latente

(WICKELGREN, 2000).

Uma possibilidade para a obtenção de novas drogas é o desenho racional que

pode ser alcançado quando existe informação detalhada sobre a interação entre

proteína alvo - ligante, fornecida pela estrutura tridimensional do alvo (DUNCAN, 2003).

Com a obtenção da estrutura 3D do alvo é possível fazer uma triagem em biblioteca

virtual de compostos químicos que possam ser utilizados como drogas naturais ou que

sirvam como base para o desenho racional de novas drogas sintéticas. As informações

sobre os mecanismos cinético e químico também são importantes para o desenho

racional de inibidores.

Uma alternativa atraente para obter novos inibidores específicos para alvos

definidos é a busca desses compostos na biodiversidade brasileira, e para isso é

necessária à obtenção de um alvo isolado e funcional. Utilizando a tecnologia de

detecção através da Ressonância Plasmônica de Superfície (SPR), imobiliza-se o alvo

e após faz-se a triagem de ligantes (possíveis inibidores) presentes em extratos de

plantas oriundos da biodiversidade brasileira e/ou seus compostos purificados. Essa

tecnologia monitora as interações bimoleculares entre compostos e o alvo de interesse

em tempo real, detectando a ligação de compostos ao alvo pelo aumento de massa em

relação ao estado inicial. Esse tipo de análise e metodologia é ideal para a descoberta

de novas drogas por ter alta sensibilidade, e pelo fato de que os compostos químicos

analisados não necessitam de marcação radioativa nem de fluorescência. Em posse

dos possíveis inibidores, segue-se para a determinação das constantes de inibição (K

)

para a caracterização da natureza da inibição (BASSO e col., 2005).

Dentre as vias metabólicas identificadas no genoma do M. tuberculosis está a

via do chiquimato (Figura 1) que representa um alvo validado para o desenvolvimento

de herbicidas e agentes antimicrobianos, pois estão presentes em algas, vegetais,

bactérias, fungos (BENTLEY, 1990) e parasitas do filo Apicomplexa (ROBERTS e cols.,

1998), mas ausente em vertebrados. Foi sugerido que esta via é essencial para o

crescimento do M. tuberculosis mesmo na presença de suplementos exógenos

(PARISH & STOKER, 2002). Com isso, inibidores específicos das enzimas presentes

nessa via apresentam potencial como agentes terapêuticos no tratamento da TB.

A via do chiquimato é composta de uma seqüência de sete reações que vão

desde a condensação da D-eritrose 4-fosfato (E4P) e do fosfoenolpiruvato (PEP) até a

formação do corismato. O corismato leva à biossíntese de precursores de aminoácidos

aromáticos, vitaminas E e K, ácido fólico (BENTLEY, 1990), naftoquinonas,

menaquinonas e micobactinas (em micobatérias) (Figura 2) (RATLEDGE, 1982).

Figura 1: A Via do Chiquimato. Esta via é composta por sete etapas enzimáticas que

culminam com a formação do corismato, o precursor chave para formação de

compostos aromáticos essenciais aos microrganismos.

Figura 2: Importância da Via do Chiquimato. O produto da via do chiquimato,

corismato e os compostos essenciais sintetizados a partir dele.

O trabalho aqui descrito mostra o estudo sobre os mecanismos cinético e químico

feito com um alvo enzimático, a chiquimato desidrogenase de M. tuberculosis (MtbSD),

presente na quarta etapa de catálise da via do chiquimato (Figura 1). A MtbSD catalisa

a redução reversível do 3-desidrochiquimato (DHS) a chiquimato (SHK), utilizando

especificamente o cofator NADPH

(Figura 3) (BENTLEY, 1990; CHAUDHURI e cols.,

1985).

chiquimato

desidrogenase

Figura 3: Reação enzimática catalisada pela enzima chiquimato desidrogenase. A

redução do 3-desidrochiquimato dependente de NADPH leva à formação do D-

chiquimato.

A proteína MtbSD, codificada pelo gene aroE (Rv2552c), possui 269

aminoácidos e pertence à superfamília das oxiredutases dependentes de NAD(P)H,

que atuam em rotas anabólicas e catabólicas (BENACH e cols., 2003).

Em 2002, a primeira descrição sobre a enzima MtbSD foi publicada

(MAGALHÃES e cols.). O artigo trata da clonagem do gene aroE da cepa M.

tuberculosis H37Rv em vetor para expressão de proteínas heterólogas pET23a(+)

(Novagen

) e expressão da enzima recombinante em E. coli BL21(DE3). O trabalho

relatou que o sistema pET23a(+)-aroE em E. coli BL21(DE3) é excelente para a

expressão do alvo MtbSD recombinante sem adição do indutor IPTG e a obtenção da

MtbSD recombinante solúvel e funcional foi feita por meio da utilização do

procedimento de congelamento e descongelamento em ciclos repetidos da E. coli

BL21(DE3) com pET23a(+)-aroE para o extravasamento do conteúdo celular

(MAGALHÃES e cols.).

A expressão da MtbSD também foi obtida em vetor de expressão pET28b

(Novagen

). A purificação da MtbSD foi feita em coluna de níquel por meio de ligação

da cauda de histidina da proteína recombinante. Os valores de K

aparente para os

substratos ácido chiquímico e NADP

foram de 0,030 mM e 0,063 mM,

respectivamente, e o k

cat

foi de 399 s

-1

. Esses valores foram determinados sem retirar

da cauda de histina da MtbSD recombinante, o que pode influenciar os dados obtidos

(ZHANG e cols., 2005).

A estrutura da chiquimato desidrogenase codificada pelo gene aroE de E. coli

(EcoliSD) foi determinada com resolução de 1,5 Å. A enzima apresenta uma arquitetura

com dois domínios

separados por um sulco grande que corresponde ao sítio

catalítico da enzima (MICHEL e col., 2003). O sítio ativo da EcoliSD foi identificado

pelos resíduos Ser14, Ser16, Lys65, Asn86, Thr101, Asp102 e Gln244, conservados na

família das chiquimato desidrogenase. A caracterização cinética aparente da proteína

EcoliSD, determinação de K

e k

cat

, foi feita utilizando a reação reversa que

corresponde à catálise de oxidação do ácido chiquímico. A EcoliSD trata-se de um

monômero em solução (CHAUDURI & COGGINS, 1985).

Um estudo semelhante foi feito com a chiquimato desidrogenase codificada pelo

gene aroE de Methanococcus jannaschii (MjanSD). A estrutura foi determinada em

2,35 Å e a arquitetura descrita é semelhante à EcoliSD. A proteína MjanSD apresenta-

se na forma dimérica em solução (PADYANA e BURLEY, 2003).

Objetivo

Considerando que a tuberculose é problema de saúde pública global, o quadro

alarmante com o advento da AIDS, a inexistência de programas eficazes de saúde

publica com relação à TB, o surgimento de cepas resistentes, o interesse na obtenção

de novos antimicobacterianos, a via do chiquimato é um alvo promissor, e que

inibidores específicos das enzimas presentes nessa via apresentam potencial como

agentes terapêuticos no tratamento da TB, “o presente trabalho visa elucidar os

mecanismos cinético e químico da enzima chiquimato desidrogenase de M.

tuberculosis (MtbSD), que servirão como base para o alcance de um inibidor

específico com possível utilização no tratamento da tuberculose.”

Para atingir tal objetivo foram feitas:

9 Caracterização da MtbSD

Determinação dos parâmetros cinéticos aparentes em ambas as direções da

reação, da constante de equilíbrio da reação, da estabilidade térmica da enzima

recombinante e da energia de ativação para a reação química.

9 Determinação dos mecanismos cinético e químico para a MtbSD

Com o intuito de investigar os mecanismos cinético e químico foram feitos

experimentos de velocidade inicial em estado estacionário, de estudos de

inibição pelos produtos, de efeitos isotópicos cinéticos primário, de efeitos

isotópicos cinéticos de solvente, de efeitos isotópicos cinéticos múltiplos, de

“proton inventory” e de efeito de pH na química ácido/base na reação.

9 Estudos estruturais da MtbSD utilizando modelagem por homologia

Os estudos estruturais in silicon da MtbSD são úteis na interpretação dos

resultados obtidos nos experimentos cinéticos e possibilita a identificação dos

possíveis resíduos envolvidos na ligação e catálise dos substratos. Para tal foi

modelada a estrutura 3D para a MtbSD em silicon utilizando como molde a

chiquimato desidrogenase de E. coli.

PARTE II

Capítulo I

Fonseca, I. O., Magalhães, M. L. B., Oliveira, J. S., Silva, R. G.,

Mendes, M. A., Palma, M. S., Santos, D. S., Basso, L. A. (2006)

“Functional shikimate dehydrogenase from Mycobacterium

tuberculosis H37Rv: purification and characterization” Protein Expr.

Purif. 46, 429-437.

Arquivo nomeado:

Functional shikimate dehydrogenase from Mycobacterium tuberculosis H37Rv.pdf

Functional shikimate dehydrogenase from Mycobacterium

tuberculosis H37Rv: Puriﬁcation and characterization

Isabel O. Fonseca

, Maria L.B. Magalha

, Jaim S. Oliveira

, Rafael G. Silva

Maria A. Mendes

, Mario S. Palma

, Dio

genes S. Santos

, Luiz A. Basso

Centro de Pequisas em Biologia Molecular e Funcional, TECNOPUC, Programa de Po

s-Graduac¸a

o em Cie

ncias Biolo

gicas: Bioquı

´mica,

Programa de Po

s-Graduac¸a

o em Biologia Celular e Molecular (PUCRS-UFRGS), Pontifı

´c

ia Universidade Cato

lica do Rio Grande do Sul,

Porto Alegre—RS 90619-900, Brazil

Laborato

rio de Biologia Estrutural e Zooquı

mica, Centro de Estudos de Insetos Sociais, Departamento de Biologia, Instituto de Biocie

ncias,

Universidade Estadual Paulista, Rio Claro—SP 13506-900, Brazil

Received 18 July 2005, and in revised form 30 September 2005

Available online 27 October 2005

Abstract

Tuberculosis (TB) poses a major worldwide public health problem. The increasing prevalence of TB, the emergence of multi-drug-

resistant strains of Mycobacterium tuberculosis, the causative agent of TB, and the devastating eﬀect of co-infection with HIV have

highlighted the urgent need for the development of new antimycobacterial agents. Analysis of the complete genome sequence of M. tuber-

culosis shows the presence of genes involved in the aromatic amino acid biosynthetic pathway. Experimental evidence that this pathway is

essential for M. tuberculosis has been reported. The genes and pathways that are essential for the growth of the microorganisms make

them attractive drug targets since inhibiting their function may kill the bacilli. We have previously cloned and expressed in the soluble

form the fourth shikimate pathway enzyme of the M. tuberculosis, the aroE-encoded shikimate dehydrogenase (mtSD). Here, we present

the puriﬁcation of active recombinant aroE -encoded M. tuberculosis shikimate dehydrogenase (mtSD) to homogeneity, N-terminal

sequencing, mass spectrometry, assessment of the oligomeric state by gel ﬁltration chromatography, determination of apparent

steady-state kinetic parameters for both the forward and reverse directions, apparent equilibrium constant, thermal stability, and energy

of activation for the enzyme-catalyzed chemical reaction. These results pave the way for structural and kinetic studies, which should aid

in the rational design of mtSD inhibitors to be tested as antimycobacterial agents.

Keywords: Mycobacterium tuberculosis; Shikimate pathway; Shikimate dehydrogenase; aroE; Enzyme kinetics; Protein puriﬁcation; Drug target

The causative agent of tuberculosis (TB),

Mycobacteri-

um tuberculosis, infects approximately 32% of the worldÕs

human population. TB remains the leading cause of mor-

tality due to a bacter ial pathogen. Currently, there are 8

million new cases and 2 million deaths annually from

tuberculosis, and it is predicted that a total of 225 million

new cases and 79 million deaths will occur between 1998

and 2030 [1]. Approximately 2 billion individuals are

believed to harbor latent TB based on tuberculin skin test

surveys [2], which represents a considerable reservoir of

bacilli. Possible factors underlying the resurgence of TB

worldwide include the HIV epidemic, increase in the home-

less population, and decline in health care structures and

national surveillance [3]. The pandemic of AIDS has had

doi:10.1016/j.pep.2005.10.004

Corresponding authors. Fax: +55 51 33203629.

E-mail addresses: [email protected] (D.S. Santos), luiz.basso@

pucrs.br (L.A. Basso).

Abbreviations used: DHS, 3-dehydroshikimate; ESI-MS, electrospray

ionization mass spectrometry; IPTG, isopropyl b-

D-thiogalactopyrano-

side; LB, Luria–Bertani; MDR-TB, multi-drug-resistant tuberculosis;

mtSD, Mycobacterium tuberculosis shikimate dehydrogenase; NADP

oxidized b-nicotinamide adenine dinucleotide phosphate; NADPH,

reduced b-nicotinamide adenine dinucleotide phosphate; SHK,

D-shikim-

ate; SDS–PAGE, sodium dodecylsulfate–polyacrylamide gel electropho-

resis; TB, tuberculosis.

www.elsevier.com/locate/yprep

Protein Expression and Puriﬁcation 46 (2006) 429–437

a major impact on the TB problem, owing not only to

increased reactivation of latent TB but also to acceleration

of transmission in HIV/TB co-infected patients following

the increase in the number of smear-positive infectious pul-

monary TB cases [4]. Another contributing factor is the

evolution of multi- drug TB (MDR-TB), deﬁned as resis-

tant to isoniazid and rifampicin, which are the most eﬀec-

tive ﬁrst-line drugs [5]. MDR-TB is more diﬃcult and

more expensive to treat, and more likely to be fatal [6].

According to the 2004 Global TB Control Report of the

World Health Orga nization, there are 300,000 new cases

per year of MDR -TB worldwide, and 79% of MDR-TB

cases are now ‘‘super-strains,’’ resistant to at least three

of the four main drugs used to treat TB [7]. The factors that

most inﬂuence the emergence of MDR-TB strains include

inappropriate treatment regimens, and patient non-compli-

ance in completing the prescribed courses of therapy due to

the lengthy standard ‘‘short-course’’ treatment or when the

side eﬀects become unbearable [8]. Thus, there is a need for

the development of new antimycobacterial agents to both

treat M. tuberculosis strains resistant to existing drugs

and shorten the duration of short-course treatment to

improve patient compliance [9].

The shikimate pathway is an attractive target for the

development of herbicides and antimicrobial agents

because it is essential in algae, higher plants, bacteria,

and fungi, but absent from mammals [10]. The mycobacte-

rial shikimate pathway leads to the biosynthesis of choris-

mic acid, which is a precursor of aromatic amino acids,

naphthoquinones, menaquinones, and mycobactins [11].

The salicylate-derived mycobactin siderophores have been

shown to be essential for M. tuberculosis growth in macro-

phages [12]. In addition, the shikimate pathway has more

recently been shown to be essential for the viability of

M. tuberculosis [13]. Accordingly, the essentiality of myco -

bacterial shikimate pathway and its absence from human

host indicate that any enzyme of this pathway represents

a promising target for the developm ent of non-toxi c anti-

mycobacterial agents.

Analysis of the complete genome sequence of M. tuber-

culosis shows the presence of seven aro genes predicted to

be involved in the shikimate pathway [14]. Amongst them,

the aroE-encoded shikimate dehydrogenase (SD) has been

predicted by DNA sequence homology to be present in

M. tuberculosis H37Rv strain. Shikimate dehydrogenase

(EC 1.1.1.25) catalyzes the fourth reaction in the shikimate

pathway. We have previously reported the cloning and

expression of M. tuberculosis SD (mtSD) [15]. In addition,

measurements of the NADPH-dependent redu ction of

3-dehydroshikimate to shikimate catalyzed by mtSD con-

ﬁrmed the correct assignment to the structural ge ne encod-

ing SD in M. tuberculosis [15]. Here, we report the

puriﬁcation to homogeneity of recombinant and functional

mtSD. The puriﬁcation protocol yielded approximately

11 mg of homogeneous recombinant mtSD from 14 L of

Escherichia coli cell culture. We also present N-terminal

amino acid sequencing and electrospray ionization mass

spectrometry (ESI-MS) data that unambiguously demon-

strate the identity and purity of homogenous recombinant

mtSD protein. The estimated molecular mass of native

homogeneous recombinant protein determined by gel ﬁl-

tration indicates that mtSD enzyme is a dimer in solution

with a subunit molecular mass value of 27,207 Da deter-

mined by ESI-MS. The apparent kinetic parameters for

mtSD were determined for all substrates in both forward

and reverse reactions. The mtSD thermal stability was eval-

uated, and an estimate for the activation energy (E

) was

obtained from a linear plot of log k versus 1/T (K

À1

). A

comparison of polypeptide sequences of SDs from

M. tuberculosis, E. coli, Haemophilus inﬂuenzae,and

Methanococcus jannaschii allowed identiﬁcation of amino

acid residues that are likely to be involved in 3-dehydros-

hikimate/shikimate binding.

The results presented here will pave the way for structur-

al and functional eﬀorts currently underway in our labora-

tory. It is hoped that these studies will provide a framew ork

on which to base the design of new agents with antituber-

cular activity and low toxicity to humans.

Materials and methods

Overexpression and release of mtSD

The recombinant plasmid pET23a(+)::aroE was trans-

formed into E. coli BL21 (DE3) host cells by electropor-

ation, and selected on LB agar plates containing

50 lgmL

À1

carbenicillin. Single colonies were used to

inoculate 14 L of LB medium containing 50 lgmL

À1

car-

benicillin with no addition of isopropyl b-

D-thiogalacto-

pyranoside (IPTG), an d grown for 24 h at 37 °Cat

180 rpm as described elsewhere [15]. Cells (49 g) were

harvested by centrifugation at 14,900g, for 30 min at

4 °C, and stored at À20 °C. Cells expressing recombinant

mtSD were placed into metal containers to allow fast

temperature equilibrium to be reached, which is neces-

sary for increased eﬃciency of cell disruption by the

freeze–thaw method [15]. Cell paste was placed into a

dry-ice/ethanol bath for 2 min and immediately trans-

ferred to an ice-water bath for no longer than 8 min; this

cycle was repeated 10 times. The cells were dissolved in

196 mL of 50 mM Tris–HCl, pH 7.8 (buﬀer A). After

incubating the mixture for 30 min on ice, cell debris

was removed by centrifugation (48,000g for 1 h) and

the supernatant containing soluble mtSD was collected.

Puriﬁcation of recombinant mtSD

All steps of the puriﬁcation protocol of recombinant

mtSD were performed on ice or at 4 °C. The supernatant

containing soluble mtSD was incubated with 1% (w/v) of

streptomycin sulfate for 30 min and centrifuged at

48,000g for 30 min. The supernatant was dialyzed twice

against buﬀer A, using a dialysis tubing with molecular

weight exclusion limit of 6000–8000 Da. This sample was

430 I.O. Fonseca et al. / Protein Expression and Puriﬁcation 46 (2006) 429–437

clariﬁed by centrifugation (48,000g for 30 min) and loaded

on a Q-Sepharose fast ﬂow (26 cm · 9.5 cm) column

(Amersham Biosciences) pre-equilibrated with the same

buﬀer. The column was washed with 5 column volumes

of buﬀer A and the absorbed material was eluted with a lin-

ear gradient (0–100%) of 20 column volumes of 50 mM

Tris–HCl, pH 7.8, 0.5 M NaCl (buﬀer B) at 1 mL min

À1

The fractions containing mtSD were pooled (55 mL) and

ammonium sulfate was added to a ﬁnal concentration of

1 M, and clariﬁed by centrifugation (48,000g for 30 min).

The supernatant was loaded on a Phenyl-Sepharose High

Performance (Amersham Biosciences) column pre-equi li-

brated with 50 mM Tris–HCl, pH 7.8, 1 M (NH

)

(buﬀer C). The column was washed with 5 column volumes

of buﬀer C and the bound proteins were eluted with a 20-

column volume linear gradient (0–100%) of buﬀer A at

1 mL min

À1

. The mtSD-containing fractions were pooled

(43 mL), concentrated to less than 4 mL using an Amicon

ultraﬁltration cell (MWCO 10,000 Da), and loaded on a

Sephacryl S-200 (26 cm · 60 cm) (Amersham Biosciences)

column pre-equilibrated with buﬀer A. The recombinant

mtSD protein was eluted in a total volume of 29 mL at a

ﬂow rate of 0.5 mL min

À1

, and loaded on an anion-ex -

change Mono-Q column (Amersham Biosciences) pre-

equilibrated with buﬀer A. The column was washed with

5 column volumes of buﬀer A and the absorbed material

was eluted with a 20-col umn volume linear gradient

(0–100%) of buﬀer B at 1 mL min

À1

. Elution proﬁles were

followed at 280 and 215 nm. Homogeneous mtSD was elut-

ed in a total volume of 12.5 mL and stored at À20 °C. Pro-

tein puriﬁcation was monitored by SDS–PAGE [16], and

the protein concentration was determined by the method

of Bradford et al. [17] using the Bio-Rad protein assay

kit (Bio-Rad) and bovine serum albumin as standard.

Enzyme activity assay of mtSD

The screening of fractions of the puriﬁcation protocol

containing shikimate dehydrogenase was performed by

assaying enzyme activity in the reverse direction in

100 mM Tris–HCl, pH 9.0, at 25 °C. The 500 lL assay

mixture contained 2 mM NADP

and 4 mM D-shikimate

(SHK) [18], and the reaction was initiated with addition

of 1 lL of the chromatographic fractions. Measurement

of the NADP

-dependent oxidation of SHK to form

NADPH and 3-dehydroshiki mate (DHS) catalyzed by

mtSD was continuously monitored by the increase in

absorbance at 340 nm (e

NADPH

= 6.18 · 10

À1

One unit of enzyme activity (U) is deﬁned as the amount

of enzyme catalyzing the conversion of 1 l mol of NADP

per minute at 25 °C.

Determination of apparent kinetic parameters

Determination of the apparent steady-state kinetics

parameters, V

max

and K

, for DHS and NADPH in the

forward reaction, was carried out at varying concentration

of one substrate (5, 10, 20, 30, 50, 100, and 200 lM) while

the other was maintained at constant saturation level. The

reverse reaction was performed in the same conditions of

pH (100 mM Tris–HCl buﬀer, pH 7.0), temperature

(25 °C) and range of concentrations of substrates. The

reaction was initiated with addition of 6 pmol of homoge-

neous M. tuberculosis SD enzyme and was monitored for

1 min. The kinetic data were analyzed by double reciprocal

plots.

Shikimate dehydrogenase activity measurements were

based on decreasing concentration of NADPH upon

DHS reduction or on increasing concentration of NADPH

upon SHK oxidation. The reaction catalyzed by mtSD was

continuously monitored by measuring change in absor-

bance at 340 nm (e

NADPH

= 6.18 · 10

À1

Determination of the energy of activation and thermo

stability of recombinant mtSD

The energy of activation (E

) was estimated for recom-

binant mtDS by the following Arrhenius equation:

k ¼ Ae

ÀE

=RT

, where k is the rate constant of the react ion

at temperature T (in Kelvin), A is a pre-exp onential factor

(related to collision frequency and a steric factor); E

is the

activation energy; R is universal gas constant

(8.3145 J K

À1

mol

À1

)ande

ÀE

=RT

is the fraction of molecu-

lar collisions that have energy equal to or greater than the

measured energy of activation (E

) of the system at a par-

ticular temperature T. The E

value was calculated from

the slope of the linear plot logk versus 1/T (K

À1

) ﬁtting

the data to the following equation: log k =(E

2.3RT) + log A. Measurements of mtSD enzyme activity

were in triplicates and performed in 100 mM Tris–HCl

buﬀer, pH 7.0, in the presence of saturating concentrations

of SHK (200 lM) and NADP

(100 lM) at the following

temperatures: 15, 20, 25, 30, and 37 °C.

For thermal stability determination, mtSD was incubat ed

at 15, 25, 37, and 55 °C and the remaining enzyme activity

was measured at diﬀerent times of incubation up to 1 h,

monitoring the mtSD reverse reaction in an assay mixture

containing saturati ng concentrations of substrates (200 lM

SHK and 100 lM NADP

) in 100 mM Tris–HCl, pH 7.0,

at 25 °C.

Mass spectrometry analysis

The homogeneity of recombinant protein preparation

was assessed by mass spectrometry (MS), employing some

adaptations made to the syst em described by Chassaigne

and Lobinski [19]. Samples were analyzed on a triple quad-

rupole mass spectrometer, model QUATTRO II, equipped

with a standard electrospray (ESI) probe (Micromass,

Altrinchan), adjusted to c a. 250 lL min

À1

. The source of

temperature (80 °C) and needle voltage (3.6 kV) were main-

tained constant throughout the experimental data

collection, applying a drying gas ﬂow (nitrogen) of

200 L h

À1

and a nebulizer gas ﬂow of 20 L h

À1

. The mass

I.O. Fonseca et al. / Protein Expression and Puriﬁcation 46 (2006) 429–437 431

spectrometer was calibrated with intact horse heart myo-

globin and its typical cone-voltage induced fragments.

The subunit molecular mass of M. tuberculosis SD was

determined by ESI-MS, adjusting the mass spectrometer

to give a peak width at half-height of 1 mass unit, and

the cone sample to skimmer lens voltage controlling the

ion transfer to mass analyzer was set to 38 V. About

50 pmol sample was injected into electrospray transport

solvent. The ESI spectrum was obtained in the multi-chan-

nel acquisition mode, scanning from m/z 500 to 2000 at

scan time of 7 s. The mass spectrometer is equipped with

MassLynx and Transform softwares for data acquisition

and spectra handling.

N-terminal amino acid sequencing

The N-terminal amino acid residues of puriﬁed recombi-

nant mtSD were identiﬁed by automated Edman degrada-

tion sequencing using a PPSQ 21A gas-phase sequencer

(Shimadzu).

Determination of native mtSD molecular mass

The molecular mass of native mtSD homogenous pro-

tein was estimated by gel-permeation chromatography on

a Superdex 200 HR column (1.0 cm · 30 cm) (Amersham

Biosciences). The column was eluted with 50 mM Tris–

HCl containing 0.2 M NaCl, pH 7.8, at a ﬂow rate of

0.4 mL min

À1

. The eluate was monitored at 215 and

280 nm and the column was calibrated with the following

protein standar ds (Amersham Biosciences): ribonuclease

A (13,700 Da) from bovine pancreas, chymotrypsin ogen

(25,000 Da) from bovine pancreas, ovalbumin

(43,000 Da) from hen egg, and albumin (67,000 Da) from

bovine serum. Blue Dextran 2000 was used to determine

the void volume (V

). The K

value was calculated for

each protein using the equation (V

À V

)/(V

À V

), where

is the elution volume for the protein and V

is the total

bed volume, and K

was plotted against the logarithm of

standard molecular weights.

Results and discussion

Expression in E. coli BL21 (DE3) of recombinant

M. tuber culosis shikimate dehydrogenase (mtSD) and the

method of disruption of transformed host cells were as

described elsewhere [15]. Recombinant mtSD was puriﬁed

as described under Materials and methods and the samples

of each chromatographic step were analyzed by SDS–

PAGE with Coom assie blue staining and assayed for

enzyme activity in the reverse direction, following the

increase in absorbance at 340 nm due to the NADP

-de-

pendent oxidation of

D-shikimate to form NADPH and

3-dehydroshikimate. The recombinant protein was puriﬁed

8.5-fold (Table 1) to electrophoretic homogeneity (Fig. 1).

The relative mobility of the polypeptide chain in SDS–

PAGE indicates a homogeneous protein with a subunit

molecular mass value of approximately 27 kDa (Fig. 1).

Even though a large amount of cells were needed for recov-

ery of homogeneous target protein in quantities necessary

for kinetic and structural studies, the freeze–thaw method

[20] was previously shown to be, amongst a num ber of

experimental protocols tested to reduce insoluble protein

production, the method of choice to obtain soluble mtSD

in its active form [15]. Approximately 11 mg of homoge-

neous recombinant mtSD could be obtaine d from 49 g of

E. coli BL21 (DE3) host cells followi ng the puriﬁcation

protocol presented in Table 1, which required four chroma-

tographic steps to obtain homogeneous mtSD. The recom-

binant protein eluted with approximately 50% of 50 mM

Tris–HCl, pH 7.8, containing 0.5 M NaCl (buﬀer B), con-

sistent with the theoretical value of 5.11 for the mtSD iso-

electric point. Since the theoretical isoelectric point values

for mtSD and E. coli SD (5.12) are quite similar, the

anion-exchange chromatographic step using Q-Sepharose

Fast Flow resin is unlikely to have separated these

Table 1

Puriﬁcation of M. tuberculosis shikimate dehydrogenase from E. coli BL21 (DE3) [pET23a(+)::aroE] cells

Puriﬁcation step Total protein (mg) Total activity (U) Speciﬁc activity

(U mg

À1

) Puriﬁcation (fold) Yield (%)

Crude extract 1040.4 475.3 0.46 1.0 100.0

Q-Sepharose 82.5 110.0 1.33 2.9 23.1

Phenyl-Sepharose 24.1 69.2 2.87 6.2 14.6

Sephacryl S-200 17.4 65.3 3.75 8.2 13.7

Mono-Q 10.8 42.1 3.90 8.5 8.9

The results presented are for a typical puriﬁcation protocol from 49 g of E. coli host cells.

UmL

À1

mg.

À1

Fig. 1. SDS–PAGE (12.5%) analysis of pooled fractions from the various

steps of the puriﬁcation protocol of mtSD. Lane 1, crude extract (50 lg);

lane 2, Q-Sepharose fast ﬂow ion exchange (30 lg); lane 3, Phenyl-

Sepharose hydrophobic interaction (10 lg); lane 4, Sephacryl S-200 gel

permeation (10 lg); lane 5, Mono-Q ion exchange (17 lg); lane 6, MW

marker High Range (Gibco).

432 I.O. Fonseca et al. / Protein Expression and Puriﬁcation 46 (2006) 429–437

enzymes. Nevertheless, this chromatographic step resulted

in 2.9- fold protein puriﬁcation (Table 1) and removal of

some noticeable contaminants with subunit molecular

weight values larger than 29 kDa (Fig. 1). The recombinant

mtSD protein in 50 mM Tris–HCl, pH 7.8, containing 1 M

(NH

)

adsorbed to a Phenyl-Sepharose High Perfor-

mance column and was eluted with approximately 57% of

50 mM Tris–HCl, pH 7.8, buﬀer resulting in a 6.2-fold

puriﬁcation with removal of substantial amount of con-

taminants (Fig. 1). Gel ﬁltration on Sephacryl S-200 resin

was performed to desalt and further purify the recombi-

nant protein. Elution of fractions containing mtSD from

an anion-exchange Mono-Q column with approximately

50% of 50 mM Tris–HCl, pH 7.8, buﬀer containing 0.5 M

NaCl resulted in homogeneous mtSD, with a protein yield

of approximately 9% (Table 1). It is noteworthy that there

was no need for elution with coenzyme from an aﬃnity col-

umn as described for E. coli shikimate dehydrogenase [21].

Accordingly, the puriﬁcation protocol here described rep-

resents an eﬃcient and low-cost method to obtain homoge-

neous mtSD. The puriﬁcation of M. jannaschii shikimate

dehydrogenase by glutathione and Sepharose Q chroma-

tography yielded, after proteolytic removal of the aﬃnity

tag, a protein with ﬁve N-terminal residues resulting from

cloning artifact [22]. The strategy followed to obtain homo-

geneous H. inﬂuenzae shikimate dehydrogenase provided a

protein with a C-terminal hexahistidine tag [23]. It ha s been

shown that N- and C-terminal hexahistidine tags have a

noticeable negative eﬀect on protein solubility of recombi-

nant proteins expressed in E. coli [24]. In addition,

His-tagged proteins may have diﬀerent structure [25] or

biological activity [26] as compared to their native form.

The mtSD cloning strategy we described elsewhere [15]

and protein puriﬁcation we presented here yielded homoge-

neous polypeptide chain with no extra amino acid residues.

The subunit molecular mass of active mtSD was deter-

mined to be 27,076 Da by electrospray ionization mass

spectrometry (ESI-MS), indicating removal of the N-ter-

minal methionine residue (predicted molecular mass:

27,207 Da). The ESI-MS result also revealed a peak at

54,150 Da, indicating that the enzyme could have a

dimeric form. No peak could be detected at the expected

molecular mass for E. coli SD (29,413 Da) and a degree

of purity of 98% could be estimated by ESI-MS, thus

providing evidence for the identity and purity of the

recombinant protein. The ﬁrst 11 N-terminal amino acid

residues of mtSD were identiﬁed as SEGPKKAGVLG

by the Edman degradation method. This result unambig-

uously identiﬁes the homogeneous recombinant protein

as mtSD and conﬁrms removal of the N-terminal methi-

onine. Modiﬁcation at the N-termini is a common type

of co-/post-translational alteration of proteins synthe-

sized in prokaryotic cells. Methionine aminopeptidase-

catalyzed cleavage of initiator methionine is usually

directed by the penultimate amino acid residues with

the smallest side chain radii of gyration (glycine, alanine,

serine, threonine, proline, valine, and cysteine) [27].

Removal of N-terminal methionine from mtSD polypep-

tide chain conforms to this rule since serine is the penul-

timate amino acid residue.

The enzymatic activity of homogeneous recombinant

mtSD puriﬁed was assayed in the reverse direction by con-

tinuously monitoring the increase in absorbance at 340 nm

upon NADP

-dependent oxidation of D-shikimate to form

NADPH and 3-dehydroshikimate. The activity of mtSD

was linearly dependent on sample volume added to the

reaction mixture (Fig. 2), thereby showing that the initial

velocity is proportional to total enzyme concentration

and that true initial velocities are being measured. The

M. tuberculosis SD was stable at À20 °C for at least 1 year.

A value of 58,367 Da for the molecular mass of

homogeneous mtSD protein was estimated by gel ﬁltra-

tion (data not shown). This result suggests that mtSD

is a dimer in solution, in agreement with the ESI-MS

results. Shikimate dehydrogenase from M. jannaschii

has recently been shown to be a dimer in solution [22].

Whereas dehydrogenases usually form oligomers, shikim-

ate dehydrogenase is present as a monomer in both

E. coli [18] and H. inﬂuenzae [23].

The apparent kinetics parameters obtained are present-

ed in Table 2. The plots ﬁtted to a hyperbolic equation,

indicating that the recombinant enzyme-catalyzed chemical

Fig. 2. Linear dependence of mtSD activity on homogeneous protein

volume. The rates of mtSD enzyme activity were followed in the reverse

reaction by continuously monitoring the increase of NADPH concentra-

tion at 340 nm. Reactions were started by addition of varying volumes of

homogeneous mtSD protein solution.

Table 2

Apparent kinetic parameters for M. tuberculosis shikimate

dehydrogenase

Substrates V

max

(U mg

À1

) K

(lM) k

cat

À1

) k

cat

À1

)

DHS 108 ± 5 31 ± 2 49 ± 2 1.6 (±0.1) · 10

NADPH 100 ± 5 10 ± 1 45 ± 2 4.5 (±0.5) · 10

SHK 18 ± 1 50.18 ± 0.01 8.2 ± 0.5 1.63 (±0.01) · 10

NADP

12.9 ± 0.7 22 ± 2 5.9 ± 0.3 2.68 (±0.03) · 10

All constants were measured in Tris–HCl 100 mM (pH 7.0) at 25 °C.

I.O. Fonseca et al. / Protein Expression and Puriﬁcation 46 (2006) 429–437 433

reaction obeys Michaelis–Menten kinetics for all substrates

(Fig. 3). The K

and V

max

values for DHS were found to

be, respectively, 31 lM and 108 U mg

À1

; and for NADPH

they were 10 lM and 100 U mg

À1

. The k

cat

for DHS at

saturating NADPH was 49 s

À1

yielding a k

cat

1.6 · 10

À1

. The K

value for DHS is lower than

the value for SD puriﬁed from Pisum sativum (340 lM);

but the K

value for NADPH is in agreement with the val-

ue of 4.3 lM determined at pH 7.4 [28]. For the reverse

reaction, the K

and V

max

values for SHK were found to

be, respectively, 50.18 lM and 18 U mg

À1

; and for NADP

they were 22 lM and 12.9 U mg

À1

. The k

cat

for SHK at

saturating NADP

was 8.2 s

À1

yielding a k

cat

1.63 · 10

À1

. For the reverse reaction catal yzed by

P. sativum SD, the K

values were 10.3 lM for NADP

and 600 lM for SHK [28]. The K

value for SHK is thus

10 times larger for P. sativum SD than for mtSD. A com-

parison of the SHK and NADP

values for E. coli

SD [29] with those obtained for mtSD shows that these

two en zymes catalyze the reverse reaction with similar

kinetic constants. A value of 19.6 for the apparent

equilibrium constant (K

) under the experimental

conditions given in the Materials and methods section

was estimated by the Haldane equation using the apparent

steady-state kinetic parameters for mtSD. An apparent

equilibrium constant value of 10.3 for the reaction at pH

7.4 has been determined by ﬁnding a mixture of substrate

concentrations that showed no measurable chan ge in

optical density at 340 nm [28]. However, it should be

pointed out that the Haldane equation we used here is

the kinetic relationship for rapid-equilibrium random Bi

Bi system, and it was implicitly assumed that the dissocia-

tion constant value for a substr ate binding second is not

changed by the substrate binding ﬁrst [30].

The temperature eﬀects on recombinant homogeneous

mtSD stability are presented in Fig. 5. Aliquots were removed

for assay of residual enzyme activity after heating for 1, 5, 10,

15, 30, 45, and 60 min at temperature values of 15, 25, 37, and

55 °C. The homogeneous mtSD is very stable, maintaining

the speciﬁc activity, measured at 25 °C, unchanged in the

reverse reaction up to 37 °C for at least 1 h of incubation

(Fig. 4). However, at 55 °C, there is a gradual loss of recombi-

nant mtSD biological activity, with only 20% of the initial

enzyme activity remaining after 1 h of incubation.

Fig. 3. Reciprocal plots with 3-dehydroshikimate (DHS), NADP

, D-shikimate (SHK) or NADPH as variable substrate (5, 10, 20, 30, 50, 100, and

200 lM) while the concentration of the other substrate was maintained at constant saturation level in the forward and reverse reaction. The enzyme

activity was assayed at 25 °C in 100 mM Tris–HCl buﬀer, pH 7.0. The reaction catalyzed by mtSD was continuously monitored by the absorbance at

340 nm (e

NADPH

= 6.18 · 10

À1

). The kinetics data were analyzed by linear regression ﬁt Michaelis–Menten kinetics. (A) 3-dehydroshikimate as

variable substrate (5, 10, 20, 30, 50, 100, and 200 lM) while the NADPH was maintained at constant saturation level (200 lM) in the forward reaction. (B)

NADPH as variable substrate (5, 10, 20, 30, 50, 100, and 200 lM) while 3-dehydroshikimate was maintained at constant saturation level (200 lM) in the

forward reaction. (C) Shikimate as variable substrate (5, 10, 20, 30, 50, 100, and 200 lM) while NADP

was maintained at constant saturation level

(100 lM) in the forward reaction. (D) NADP

as variable substrate (5, 10, 20, 30, 50, 100, and 200 lM) while shikimate was maintained at constant near-

saturation level (100 lM) in the forward reaction.

434 I.O. Fonseca et al. / Protein Expression and Puriﬁcation 46 (2006) 429–437

From the linear plo ts of logk versus 1/T (K

À1

), a val-

ue of 35.2 kJ mol

À1

for E

was obtained for mtSD

(Fig. 5). It should be pointed out that the E

value cal-

culated from the Arrhenius plot is an apparent or ‘‘aver-

age value,’’ and that the pre-exponential factor (A) was

considered as temperature-independent in the tempera-

ture range used in our experiments. Keeping that in

mind, 35.2 kJ mol

À1

can be considered as the minimal

amount of energy required to initiate the mtSD-catalyzed

chemical reaction, since we measured the enzyme activity

at saturating concentrations of substrates (200 lM SHK

and 100 lM NADP

). Interestingly, the Arrhenius plot

is linear indicating no change in the rate-limiting step

of the mtSD-catalyzed chemical reaction at diﬀerent tem-

peratures. In addition, there was no su dden drop in the

Arrhenius plot at low 1/T (high T) va lues that could

indicate protein denaturation, which is consistent with

mtSD thermal stability up to 55 °C.

The three-dimensional structures of shikimate dehydro-

genase from E. coli [29,31], H. inﬂuenzae [23,32],and

M. jannaschii [22] have been determined. Sequence align-

ment of shikimate dehydrogenas e from M. tuberculosis

H37Rv strain, E. coli , H. inﬂuenzae, M. jannaschii was car-

ried out using ClustalW [33,34]. The identity between the

SD sequences from M. tuberculosis and E. coli is 24%. The

substrate-binding site in the E. coli SD has been identiﬁed

by the position of the nicotinamide ring of the cofactor and

was delineated almost entirely by residues from the N-termi-

nal domain [29]. This binding site is in a pocket where most of

the residues absolutely conserved in the shikimate dehydro-

genase (SDH) family are located, i.e., Ser14, Ser16, Lys65,

Asn86, Thr101, Asp102, and Gln244 (E. coli SD numbering).

These residues are conserved in the mtSD polypeptide

sequence (Fig. 6) corresponding to the following residues:

Ser18, Ser20, Lys69, Asn90, Thr104, Asp105, and Gln243

(M. tuberculosis SD numbering). How ever, the 3-dehydros-

hikimate-binding site has not unambig uously been shown

for E. coli shikimate dehydrogenase since it was inferred

from DTT and sulfate ions bound to the crystal and position

of the nicotinamide ring of NADP

[29]. The sequence

identity (24%) between mtSD and M. jannaschii SD is similar

to the identity betw een the sequences of M. tuberculosis and

E. coli SDs. The N-terminal domain (domain I) of M. jann-

aschii SD is responsible for 3-dehydroshikimate substrate

binding [22]. The putative active site residues within domain

I that are likely to be involved with substrate binding are

invariant polar residues include Lys70, Asn91, and Asp106

(M. jannaschii SD numbering). Further examination of the

active site cleft of M. jannaschii SD revealed another strictly

conserved residue, Gln254. These residues correspond to

Lys69, Asn90, Asp105, and Gln244 in M. tuberculosis SD.

It has been proposed that these residues are likely to be

involved in catalytic reduction of DHS to SHK catalyzed

by M. jannaschii SD [22]. The amino acids Ser14, Ser16,

Lys65, Asn85, Asp102, and Gln245 (H. inﬂuenzae SD

numbering) have been suggested as some of the potential

residues involved in 3-dehydroshikimate-binding site in

H. inﬂuenzae SD [23]. All these residues are conserved in

the mtSD polypeptide sequence (Fig. 6), and corres pond to

the same residues observed in the E. coli SD sequence. More

recently, the crystal structure of a newly characterized

shikimate dehydrogenase-like protein (HI0607) from

H. inﬂuenzae has been determined [32], and the conserved

residues Lys67, Asn88, Asp103, and Gln242 have be en

proposed to be involved in either catalysis or substrate bind-

ing. It should be pointed out that the three-dimensional

structures of shikimate dehydrogenases determined were

in complex with the cofactor and no three-dimensional

structure in complex with 3-dehydroshikimate/shikimate

molecules have been reported to date.

Fig. 5. The activation energy (E

) estimated for homogeneous

M. tuberculosis SD enzyme by the Arrhenius equation. All data were

obtained in the reverse reaction in 100 mM Tris–HCl, pH 7.0, containing

200 lM SHK, 100 lM NADP

, and 0.24 nM of homogeneous

M. tuberculosis SD enzyme and the reaction was monitored for 1 min at

the following temperatures: 288, 293, 298, 303, and 310 K.

Fig. 4. Heat stability of mtSD activity as a function of time (minutes) of

incubation at diﬀerent temperatures. Incubation temperatures were as

follows: 15 °C(d), 25 °C(m), 37 °C(.), and 55 °C(n). All reactions were

carried out in 100 mM Tris–HCl, pH 7.0, assay mixture containing

200 lM SHK, 100 lM NADP

, and 0.24 nM of homogeneous

M. tuberculosis SD enzyme and the reaction was measured for 1 min at

25 °C.

I.O. Fonseca et al. / Protein Expression and Puriﬁcation 46 (2006) 429–437 435

The work presented here describes, to the best of our

knowledge, the ﬁrst puriﬁcation protocol of recombinant

M. tuberculosis SD to homogeneity and determ ination of

its oligomeric state. In addition, here we present a detailed

characterization of the homogeneous mtSD. It should be

pointed out that expression of mtSD in soluble and active

form proved to be laborious to achieve [15]. Protein pro-

duction and crystallization must be optimized if structural

genomics will ever reach its goal of solving the three-di-

mensional structure of the whole proteome encoded by a

given genome [35]. Unfortunately, even when a genome

can be sequenced, only up to 20% of the protein targets

can produce soluble proteins under very basic experimental

conditions [36]. Thus, expression of proteins in soluble

form has been identiﬁed as an important bottleneck in

eﬀorts to determine biological activity and crystal structure

of M. tuberculosis proteins [37].

The protocol for mtSD puriﬁcation presented here

should provide protein in qua ntities necessary for further

enzymological studies and three-dimensional structure

determination eﬀorts. Dete rmination of mtSD kinetic

mechanism by steady-state and pre-steady-state kinetics,

and isotope eﬀects, site-directed mutagenesis, and chemical

rescue will allow elucidation of its chemical and catalytic

mechanism. The three-dimensional structures of shikimate

dehydrogenases from E. coli [29,31], H. inﬂuenzae [23,32],

and M. jannaschii [22] should facilitate screening of exper-

imental con ditions to obtain crystals of mtSD protein and

may provide templates for mtSD structure determination

by molecular replac ement. All of the enzymes that make

up the shikimate pathway are potential targets for the

design of novel drugs directed against pathogenic bacteria.

The struc ture of M. tuberculosis shikimate dehydrogenase

and the understanding of the enzyme mode of action will

be used as a platform for the design of eﬀective inhibitors

of this pathway aiming at the development of antitubercu-

lar agents.

Acknowledgments

Financial support for this work was provide by Millen-

nium Initiative Program MCT-CNPq, Ministry of Health-

Department of Science and Technology-UNESCO (Brazil)

to D.S.S. and L.A.B. D.S.S. and L.A.B. also acknowledge

grants awarded by PRONEX, CNPq/FAPERGS. D.S.S.

(CNPq, 304051/1975-06), L.A.B. (CNPq, 520182/99-5),

and M.S.P. (CNPq, 500079/90 -0) are researchers awardees

from the National Research Council (CNPq) of Brazil. We

Fig. 6. Multiple sequence alignment of SD from M. tuberculosis using ClustalW. The boxes denote residues conserved in the SDH family involved in

catalytic reduction of DHS to SHK. This sequence alignment was constructed using the following sequences from GenBank: E. coli K12 AroE

(NP_417740, residues 1–172), H. inﬂuenzae AroE (YP_248346, residues 1–268), M. jannaschii AroE (NP_248077, residues 1–282), and M. tuberculosis

AroE (CAB06186, residues 1–269). The symbols are as follows: * = identity, : = strong similarity, . = week similarity. These nomenclature is based on the

properties of aminoacids using the substitution matrix BLOSUM [34].

436 I.O. Fonseca et al. / Protein Expression and Puriﬁcation 46 (2006) 429–437

thank Professor John W. Frost, Department of Chemistry

of Michigan State University, for his generous gift of

3-dehydroshikimate substrate.

References

[1] World Health Organization, World Health Report on Infections

Diseases: Removing Obstacles to Healthy Development. Atar,

Switzerland. WHO/CDS/99.1, 1999.

[2] C. Dye, S. Scheele, P. Dolin, V. Pathania, M.C. Raviglione, Global

burden of tuberculosis: estimated incidence, prevalence, and mortality

by country, JAMA 282 (1999) 677–686.

[3] B.R. Bloom, C.J.L. Murray, Tuberculosis: commentary on a

reemergent killer, Science 257 (1998) 1055–1064.

[4] A. Knigge, H. Morr, A. Kilian, Tuberculosis and HIV/AIDS epidem-

ics: opportunities for cross support, Deutsche Gesellschaft Fu

r Tech-

nische Zusammenarbeit (GTZ) GmbH (2000) 1–7 (<www.gtz.de/>).

[5] L.A. Basso, J.S. Blanchard, Resistance to antitubercular drugs, Adv.

Exp. Med. Biol. 456 (1998) 115–144.

[6] J.D. McKinney Jr., W.R. Jacobs, B.R. Bloom, Persisting problems in

tuberculosis, in: R.M. Krause (Ed.), Emerging Infections, Academic

Press, New York, 1998, pp. 51–146.

[7] World Health Organization, Anti-tuberculosis drug resistance in the

world: Third Global Report (2004).

[8] K. Duncan, Progress in TB drug development and what is still

needed, Tuberculosis 83 (2003) 201–207.

[9] R.J. OÕBrien, P.P. Nunn, The need for new drugs against tuberculosis,

Am. J. Respir. Crit. Care Med. 162 (2001) 1055–1058.

[10] R. Bentley, The shikimate pathway—a metabolic tree with many

branches, Crit. Rev. Biochem. Mol. Biol. 25 (1990) 307–384.

[11] C. Ratledge, Nutrition, growth and metabolism, in: C. Ratledge, J.L.

Stanford (Eds.), The Biology of the Mycobacteria, vol. 1, Academic

Press, London, 1982, pp. 185–271.

[12] J.J. de Voos, K. Rutter, B.G. Schroder, H. Su, Y. Zhu, C.E. Barry III,

The salicylate-derived mycobactin siderophores of Mycobacterium

tuberculosis are essential for growth in macrophages, Proc. Natl.

Acad. Sci. USA 97 (2000) 1252–1257.

[13] T. Parish, N.G. Stoker, The common aromatic amino acid biosyn-

thesis pathway is essential in Mycobacterium tuberculosis, Microbiol-

ogy 148 (2002) 3069–3077.

[14] S.T. Cole, R. Brosch, J. Parkhill, T. Garnier, C. Churcher, D. Harris,

S.V. Gordon, K. Eiglmeier, S. Gas, C.E. Barry Jr., F. Tekaia, K.

Badcock, D. Basham, D. Brown, T. Chillingworth, R. Connor, R.

Davies, K. Devlin, T. Feltwell, S. Gentles, N. Hamlin, S. Holroyd, T.

Hornsby, K. Jagels, B.G. Barrell, Deciphering the biology of

Mycobacterium tuberculosis from the complete genome sequence,

Nature 393 (1998) 537–544.

[15] M.L. Magalha

es, C.P. Pereira, L.A. Basso, D.S. Santos, Cloning and

overexpression of functional shikimate dehydrogenase (EC 1.1.1.25)

from Mycobacterium tuberculosis H37Rv, Protein Expr. Purif. 26

(2002) 59–64.

[16] U.K. Laemmli, Cleavage of structural proteins during the assembly of

the head of bacteriophage T4, Nature 227 (1970) 680–685.

[17] M.M. Bradford, R.A. McRorie, W.L. Williams, A rapid and sensitive

method for the quantitation of microgram quantities of protein

utilizing the principle of protein-dye binding, Anal. Biochem. 72

(1976) 248–254.

[18] S. Chaudhuri, J.R. Coggins, The puriﬁcation of shikimate

dehydrogenase from Escherichia coli, Biochem. J. 226 (1985)

217–223.

[19] H. Chassaigne, R. Lobinski, Characterization of horse kidney

metallothionein isoforms by electrospray MS and reversed-phase

HPLC–electrospray MS, Analyst 123 (1998) 2125–2130.

[20] B.H. Johnson, M.H. Hecht, Recombinant proteins can be isolated

from E. coli cells by repeated cycles of freezing and thawing, Bio/

Technology 12 (1994) 1357–1360.

[21] J. Maclean, S.A. Campbell, K. Pollock, S. Chackrewarthy, J.R.

Coggins, A.J. Lapthorn, Crystallization and preliminary X-ray

analysis of shikimate dehydrogenase from Escherichia coli, Acta

Crystallogr. Sect. D. Biol. Crystallogr. 56 (2000) 512–515.

[22] A.K. Padyana, S.K. Burley, Crystal structure of shikimate 5-

dehydrogenase (SDH) bound to NADP: insights into function and

evolution, Structure 11 (2003) 1005–1013.

[23] S. Ye, F.v. Delft, A. Brooun, M.W. Knuth, R.V. Swanson, D.E.

McRee, The crystal structure of shikimate dehydrogenase (AroE)

reveals a unique NADPH binding mode, J. Bacteriol. 185 (2003)

4144–4151.

[24] E.A. Woestenenk, M. Hammarstro

m, S. van dn Berg, T. Ha

rd, H.

Berglund, His tag eﬀect on solubility of human proteins produced in

Escherichia coli: a comparison between four expression vectors, J.

Struct. Funct. Genomics 5 (2004) 217–229.

[25] A. Chant, C.M. Kraemer-Pecore, R. Watkin, G.G. Kneale, Attach-

ment of a histidine tag to the minimal zinc ﬁnger protein of the

Aspergillus nidulans gene regulatory protein AreA causes a confor-

mational change at the DNA-binding site, Protein Expr. Purif. 39

(2005) 152–159.

[26] I. Fonda, M. Kenig, V. Gaberck-Porekar, P. Prostovaek, V. Menart,

Attachment of histidine tags to recombinant tumor necrosis factor-

alphadrasticallychangesitsproperties,Sci.WorldJ. 2(2002)1312–1325.

[27] P.-H. Hirel, J.-M. Schmitter, P. Dessen, G. Fayat, S. Blanquet, Extent

of N-terminal methionine excision from

Escherichia coli proteins is

governed by the side-chain length of the penultimate amino acid,

Proc. Natl. Acad. Sci. USA 86 (1989) 8247–8251.

[28] D. Balinsky, A.W. Dennis, W.W. Cleland, Kinetic and isotope-

exchange studies on shikimate dehydrogenase from Pisum sativum,

Biochemistry 10 (1971) 1947–1952.

[29] G. Michel, A.W. Roszak, V. Sauve

, J. Maclean, A. Matte, J.R.

Coggins, M. Cygler, A.J. Lapthorn, Structures of shikimate dehy-

drogenase AroE and its paralog YdiB, J. Biol. Chem. 278 (2003)

19463–19472.

[30] W.W. Cleland, An analysis of Haldane relationships, Methods

Enzymol. 87 (1982) 366–369.

[31] J. Benach, I. Lee, W. Edstron, A.P. Kuzin, Y. Chiang, T.B. Acton,

G.T. Montelione, J.F. Hunt, The 2.3 A

crystal structure of the

shikimate 5-dehydrogenase orthologue YdiB from Escherichia coli

suggests a novel catalytic environment for na NAD-dependent

dehydrogenase, J. Biol. Chem. 278 (2003) 19176–19182.

[32] S. Singh, S. Korolev, O. Koroleva, T. Zarembinski, F. Collart, A.

Joachimiak, D. Christendat, Crystal structure of a novel shikimate

dehydrogenase from Haemophilus inﬂuenzae, J. Biol. Chem. 280

(2005) 17101–17108.

[33] BLAST: S.F. Altschul, T.L. Madden, A.A. Scha

ﬀer, J. Zhang, Z.

Zhang, W. Miller, D.J. Lipman, Gapped BLAST and PSIBLAST: a

new generation of protein database search programs, Nucleic Acids

Res. 25 (1997) 3389–3402.

[34] CLUSTALW: J.D. Thompson, D.G. Higgins, T.J. Gibson, CLUS-

TAL W: improving the sensitivity of progressive multiple sequence

alignment through sequence weighting, position-speciﬁc gap penalties

and weight matrix choice, Nucleic Acids Res. 22 (1994) 4673–4680.

[35] W.R.A. Taylor, A Ôperiodic tableÕ for protein structures, Nature 415

(2002) 657–660.

[36] S.A. Lesley, P. Kuhn, A. Godzik, A.M. Deacon, I. Mathews, A.

Kreusch, G. Spraggon, H.E. Klock, D. McMullan, T. Shin, J.

Vincent, A. Robb, L.S. Brinen, M.D. Miller, T.M. McPhillips, M.A.

Miller, D. Scheibe, J.M. Canaves, C. Guda, L. Jaroszewski, T.L.

Selby, M.A. Elsliger, J. Wooley, S.S. Taylor, K.O. Hodgson, I.A.

Wilson, P.G. Schultz, R.C. Stevens, Structural genomics of the

Thermotoga maritima proteome implemented in a high-throughput

structure determination pipeline, Proc. Natl. Acad. Sci. USA 99

(2002) 11664–11669.

[37] R. Vicentelli, C. Bignon, A. Gruez, S. Canaan, G. Sulzenbacher, M.

Tegoni, V. Campanacci, C. Cambillau, Medium-scale structural

genomics: strategies for protein expression and crystallization, Acc.

Chem. Res. 36 (2003) 165–172.

I.O. Fonseca et al. / Protein Expression and Puriﬁcation 46 (2006) 429–437 437

Capítulo II

Este artigo será submetido para a revista ARCHIVES OF

BIOCHEMISTRY AND BIOPHYSICS

Fonseca I.O., Silva R.G., Fernandes C., Souza O.N., Basso L.A.,

Santos D.S. “Shikimate Dehydrogenase from Mycobacterium

tuberculosis H37Rv: Kinetic and Chemical Mechanisms.

Kinetic and chemical mechanisms of shikimate dehydrogenase from

Mycobacterium tuberculosis

Isabel O. Fonseca

‡π

, Rafael G. Silva

‡π

, Claudia Fernandes

†

, Osmar Norberto de Souza

†

, Luiz A.

Basso

‡*

, and Diógenes S. Santos

‡*

‡

Centro de Pequisas em Biologia Molecular e Funcional, Pontifícia Universidade Católica do Rio

Grande do Sul (PUCRS), Porto Alegre - RS 90619-900, Brazil.

Programa de Pós-Graduação em Ciências Biológicas: Bioquímica, Universidade Federal do Rio

Grande do Sul (UFRGS).

†

Laboratório de Bioinformática, Modelagem e Simulação de Biossistemas – LABIO, Faculdade

de Informática, PUCRS, Porto Alegre - RS 90619-900, Brazil.

Short Title: M. tuberculosis shikimate dehydrogenase mode of action

Address correspondence to: Luiz A. Basso or Diógenes S. Santos, Centro de Pesquisas

em Biologia Molecular e Funcional, Pontifícia Universidade Católica do Rio Grande do Sul

(PUCRS), Porto Alegre - RS - 90619-900, Brazil. Phone/Fax: +55 51 33203629; E-mail:

[email protected] or [email protected]

ABSTRACT

Mycobacterium tuberculosis shikimate dehydrogenase (MtbSD) catalyzes the fourth

reaction in the shikimate pathway, the NADPH-dependent reduction of 3-

dehydroshikimate. To gather information on the kinetic mechanism, initial velocity

patterns, product inhibition, and primary deuterium kinetic isotope effect studies were

performed and the results suggested a steady-state ordered bi-bi kinetic mechanism. The

magnitudes of both primary and solvent kinetic isotope effects indicated that the hydride

transferred from NADPH and protons transferred from the solvent in the catalytic cycle are

not significantly rate limiting in the overall reaction. Proton inventory analysis indicates

that one proton gives rise to solvent isotope effects. Multiple isotope effect studies indicate

that both hydride and proton transfers are concerted. The pH profiles revealed that

acid/base chemistry takes place in catalysis and substrate binding. The MtbSD 3D model

was obtained in silico by homology modeling. Kinetic and chemical mechanisms for

MtbSD are proposed on the basis of experimental data.

Keywords: shikimate dehydrogenase, tuberculosis, drug target, 3-dehydroshikimate,

shikimate, kinetic mechanism, chemical mechanism, Mycobacterium tuberculosis.

Introduction

Tuberculosis

(TB) is a pandemic, which even today, remains a major global health

concern. Its causative agent, Mycobacterium tuberculosis, is one of the most prolific

infectious agents affecting humans. A third of the world’s population is thought to host M.

tuberculosis and approximately 30 million people have died from the disease in the past

decade [1]. The treatment of multidrug-resistant TB (MDR-TB), defined as resistant to at

least isoniazid and rifampicin [2], requires the administration of second-line drugs that are

more toxic and less effective, and are given for at least three times as long as, and 100 times

as expensive as basic chemotherapy regimens [3]. More recently, a survey of the frequency

and distribution of extensively drug-resistant (XDR) TB cases, which are defined as cases

in persons with TB whose isolates were resistant to isoniazid and rifampicin and at least

three of the six main classes of second-line drugs, showed that during 2000-2004, of 17,690

TB isolates, 20% were MDR and 2% were XDR, and that XDR-TB has a wide geographic

distribution [4]. New antimycobacterial agents are thus needed to improve the treatment of

MDR- and XDR-TB, and to provide more effective treatment of latent tuberculosis

infection [5].

The availability of the complete genome sequence of M. tuberculosis allowed the

identification of molecular targets against which new therapeutical agents may be

developed [6]. Homologues to the seven enzymes of the shikimate pathway were identified

in the genome sequence of M. tuberculosis. This pathway has been shown to be essential

for the viability of M. tuberculosis [7]. Accordingly, the essentiality of the mycobacterial

shikimate pathway and its absence from human host indicate that any enzyme of this

pathway represents a promising target for the development of non-toxic antimycobacterial

agents. However, an understanding of the kinetic mechanism and availability of structural

models are needed to help the rational inhibitor design to progress more rapidly [8].

Shikimate dehydrogenase (EC 1.1.1.25), the fourth enzyme in the shikimate

biosynthesis pathway, catalyzes the NADPH-dependent reduction of 3-dehydroshikimate

(DHS) to shikimate (SHK) (Fig. 1). We have previously reported cloning and expression of

M. tuberculosis SD (MtbSD) [9]. In addition, measurements of the NADP

-dependent

oxidation of shikimate to 3-dehydroshikimate catalyzed by recombinant MtbSD confirmed

the correct assignment to the structural gene encoding SD in M. tuberculosis [9]. More

recently, we have reported purification to homogeneity of recombinant functional MtbSD,

N-terminal amino acid sequencing, electrospray ionization mass spectrometry analysis,

chromatographic size exclusion, and enzyme activity measurements that unambiguously

demonstrated the identity of recombinant MtbSD, its oligomeric state, the apparent kinetic

parameters for all substrates, the thermal stability, and the activation energy for the

enzyme-catalyzed chemical reaction [10].

In the present work, initial velocity patterns in the forward direction, product

inhibition studies, primary deuterium kinetic isotope effects, solvent kinetic isotope effects,

proton inventory, multiple kinetic isotope effects, pH rate profiles, and homology modeling

were performed to investigate the MtbSD kinetic and chemical mechanisms. The data here

presented provide a framework on which to base the rational design of inhibitors with

possible antitubercular activity.

Materials and Methods

Enzymatic Assay for MtbSD - All enzyme activity assays were carried out at 25°C in 100

mM potassium phosphate (pH 7.3) by monitoring the decrease in absorbance at 340 nm (ε

= 6220 M

-1

) accompanying the conversion of NADPH to NADP

in the presence of

DHS and MtbSD. The final enzyme concentration was 1.8 nM. To obtain initial velocity

measurements, rate values were determined under experimental conditions in which less

than 5% of substrate was consumed.

Initial Velocity and Product Inhibition - To determine the steady-state kinetic parameters

and initial velocity patterns, MtbSD activity was measured in the presence of varying

concentrations of one substrate and several fixed-varied concentrations of the other.

Product inhibition patterns were determined by measuring initial rates at five

concentrations of one substrate, fixed non-saturating concentration of the co-substrate, and

fixed-varying levels of product (either NADP

or SHK).

Kinetic Isotope Effects and Proton Inventory - The synthesis of [4S-

H]NADPH was

performed as described by Ottolina et al. (1989) [11]. Both NADPH and NADPD

substrates were purified on an FPLC Mono-Q column (GE Healthcare) as previously

described, and the fractions with absorbance ratios A

260nm

340nm

≤ 2.3 were pooled [12].

All measurements were performed in duplicate. Primary deuterium kinetic isotope effects

were determined by measuring initial rates in the presence of varying concentrations of one

substrate and five fixed concentrations of the other, with either NADPH or NADPD.

Solvent kinetic isotope effects were determined by measuring initial velocities using a

saturating level of one substrate and varying concentrations of the other in either H

O or 90

atom % D

O. Multiple kinetic isotope effects were obtained by determining the primary

isotope effects using D

O as solvent. The proton inventory was determined using saturating

concentrations of DHS and NADPH at various mole fractions of D

O. Each proton

inventory individual initial rate datum is the average of triplicate measurements.

pH-Rate Profiles - To determinate the pH dependence of k

cat

and k

cat

, initial velocities

were measured in the presence of varying concentrations of one substrate and a saturating

level of the other at different pHs in a mixture of potassium phosphate and boric acid. The

mixed buffers were used over the following pH values: 5.5, 6.0, 6.5, 7.0, 7.5, 8.0, 8.5, 9.0,

9.5, and 10.0. The MtbSD enzyme had been incubated in buffers at all the pH values

mentioned above and assayed under standard conditions to ensure enzyme stability at the

tested pHs.

Data Analysis - Initial velocity kinetic data were fitted to appropriate equations by using the

nonlinear regression function of SigmaPlot 2000 (SPSS, Inc.). Substrate hyperbolic

saturation curves at a single concentration of the fixed substrate and varying concentrations

of the other substrate were fitted to equation 1. Intersecting initial velocity patterns were

fitted to equation 2, which describes a sequential mechanism. For equations 1 and 2, v is the

measured reaction velocity, V is the maximal velocity, A and B are the concentrations of

substrates (DHS and NADPH), K

, and K

are the corresponding Michaelis- Menten

constants, and K

is the dissociation constant for substrate A.

v = VA / (K

+ A) (Eq. 1)

v = VAB / (K

B + K

A + K

+ AB) (Eq. 2)

Competitive and noncompetitive inhibition data were fitted to, respectively,

equation 3 and 4. For equations 3 and 4, I represent the inhibitor concentration, and K

and

are the slope and intercept inhibition constants, respectively.

v = VA / [K

(1 + I/K

) + A] (Eq. 3)

v = VA / [K

(1 + I/K

) + A (1 + I/K

)] (Eq. 4)

The kinetic isotope effect data were fitted to equations 5 or 6, which describe,

respectively, effects on both V and V/K and on V only, where Fi represents the fraction of

isotopic label, and E

is the isotope effect on V/K minus one and E

is the isotope effect on

V minus one.

v = VA / [K (1 + FiE

V/K

) + A (1 + FiE

)] (Eq. 5)

v = VA / [K + A (1 + FiE

)] (Eq. 6)

Data for pH profiles data that showed a decrease in logk

cat

and logk

cat

with a

slope of -1 as the pH values increased were fitted to equation 7, where y is the apparent

kinetic parameter, C is the pH-independent plateau value of y, H is the hydrogen ion

concentration, and K

is the apparent base dissociation constant for ionizing groups.

log y = log [C/(1 + K

/H)] (Eq. 7)

Homology Modeling - Primarily, homology modeling is the identification and selection of

template proteins from the Protein Data Bank (PDB) [13] that are related to the target

sequence. The program Blastp [14] was used to search for templates. Multiple sequence

alignment comparisons were carried out to improve the sensitivity of the search and to find

regions with high similarity using ClustalW [15]. The program MODELLER6v2 [16] was

used to build the protein models, using the standard protocol of the comparative protein

structure modeling methodology [17]. The best model of each enzyme was evaluated and

selected according to their stereochemical quality analyzed with PROCHECK [18].

Validation of the 3D profiles of the models was performed with VERIFY 3D [19]. Here,

the structure prediction of MtbSD was based on 3D structure for the homologous SD from

Escherichia coli (E. coli) (PDB ID: 1NYT), experimentally determined by X-ray

diffraction at 1.5Å resolution [20]. Root-mean square deviation (RMSD) between template

and model structures was calculated using SwissPdbViewer [21]. Drawing of protein

structures were performed with the program SwissPdbViewer [21].

Results

Initial Velocity and Product Inhibition Patterns - The constants for steady-state

kinetics of MtbSD were determined under initial velocity experimental conditions. When

DHS was varied at fixed-varied NADPH concentrations, the lines intersected on the x-axis

and on the left of the y-axis (Fig. 2A); when NADPH was varied with fixed-varied DHS

concentrations, the lines intersected below the x-axis and on the left of the y-axis (Fig. 2B).

The data were fitted to equation 2, and yielded the following values: K

DHS

= 44 ± 3

iaDHS

= 30 ± 9, K

NADPH

= 34 ± 2

M, k

cat

= 78 ± 2 s

-1

, k

cat

DHS

= 1.8 x 10

-1

, and

cat

NADPH

= 2.3 x 10

-1

Product inhibition experiments were performed and the data fitted to equation 3 or

4. The results are given in Table 1. NADP

is a competitive inhibitor (C) and a

noncompetitive inhibitor (NC) versus NADPH and DHS, respectively. For SHK product,

the inhibition pattern is C and NC versus, respectively, DHS and NADPH.

Kinetic Isotope Effects and Proton Inventory - The primary deuterium kinetic

isotope effects on V/K and V for both substrates are presented in Table 2. The primary

kinetic isotope effect values of 1.8 for

DHS

and 1.5 for

NADPH

using [4S-

H]NADPH

indicate that the C

-proS hydride (B side) is transferred to DHS in the oxy-reduction

reaction catalyzed by MtbSD.

The

(V/K

app

)

DHS

values decreased to a limiting value of 1.0 ± 0.03 as the

concentrations of either NADPH or NADPD increased (Fig. 3), whereas

(V/K

app

)

NADPH

(1.4 ± 0.3) was independent of fixed-varied concentrations of DHS (Fig. 3 - inset).

Solvent kinetic deuterium isotope effects were determined at pH 7.3 (Fig. 4A and

Fig. 4B), a pH region in which the kinetic parameters are independent of small changes in

pH. Small solvent kinetic isotope effects on V were observed for DHS (1.5 ± 0.3) and

NADPH (1.3 ± 0.2), whereas no solvent effects on V/K were observed (Table 2). The

proton inventory, a relationship between V and the mole fraction of D

O, was linear for

MtbSD (Fig. 4A - inset).

In order to distinguish between stepwise and concerted mechanisms, multiple

isotope effects were evaluated measuring the primary isotope effects for both substrates in

either D

O (Fig. 5A and 5B) or H

O (Fig. 5A and 5B - insets), and the results are

summarized in Table 2. The

V/K

NADPH

value in H

O (1.4 ± 0.3) was within standard error

of the

V/K

NADPH

value in D

O (1.5 ± 0.1), whereas

NADPH

in D

O (2.5) increased as

compared to the value obtained in H

O (1.5). The magnitude of the

DHS

value in D

(2.6) also increased, and, most important, running the reaction in D

O allowed the

observation of a small, but significant,

V/K

DHS

, (1.3 ± 0.1) in contrast to the unity obtained

in H

pH Rate Profiles - To probe the role of acid/base chemistry in the mechanism of

MtbSD, the pH dependence of k

cat

and k

cat

for DHS and NADPH was determined over

the pH range of 5.5 ≤ pH ≤ 10.0. The pH profiles of k

cat

for both substrates are very similar

and decrease at high pHs with a slope of -1 (Fig. 6A), demonstrating that deprotonation of a

single residue with apparent pK

value of 8.9 ± 0.1 abolishes MtbSD catalytic activity. The

cat

profiles for DHS and NADPH show a decrease at high pH values with a slope of -1

(Fig. 6B and Fig. 6C), suggesting that deprotonation of a single ionizable group with pK

value of 9.1 ± 0.1 diminishes DHS and NADPH binding.

Homology Modeling -In the search for templates to perform the homology modeling

we found three candidate structures, namely, E. coli SD (PDB ID: 1NYT) [20], E. coli

YdiB (PDB ID: 109B) [20] and Methanococcus jannaschii SD (PDB ID: 1NVT) [22]. The

Blastp results listed M. jannaschii SD (PDB ID: 1NVT) as the best template for MtbSD.

However, the structure coordinates were not available at the time of this modeling work.

The second best result was YdiB (PDB ID: 109B), although its function is different from

that of MtbSD. YdiB was characterized as a dual specificity quinate/shikimate

dehydrogenase that utilizes either NAD

or NADP

as cofactor. Multiple sequence

alignment, with the insertion of two gap regions into the MtbSD sequence, illustrates this

difference (Fig. 7). Segments of the target sequence which have no equivalent in the

template are the most difficult regions to model. Hence, the E. coli SD (PDB ID: 1NYT),

solved experimentally by X-ray diffraction at 1.5Å resolution [20], was used as a template

to model the 3D structure of MtbSD. They both have the same function and a sequence

identity of approximately 25%, the limit usually allowed for comparative protein structure

modeling [14].

Ten models of the enzyme were built, and evaluated by PROCHECK [18] and

VERIFY 3D [19] to choose the best one. MtbSD contains 269 amino acid residues. Out of

220 non-glycine and non-proline residues, 199 or 90.5% were located in the most favored

regions of the Ramachandran plot. The best model of MtbSD is illustrated in Figure 8. The

backbone RMSD between template and model structures is 1.93 Å. Considering the

sequence divergence between these ortholog sequences, approximately 25% sequence

identity, this RMSD value is expected and the modeling results, overall, demonstrate that

the MtbSD model obtained is highly satisfactory and can thus be used to infer structure-

activity relationships for this enzyme (Fig. 8).

Discussion

Kinetic Mechanism of MtbSD - The families of intersecting double-reciprocal plots

observed in the initial velocity for the forward direction for both substrates are consistent

with a sequential mechanism and thus both substrates must attach to the enzyme forming a

ternary complex before any product is released. Accordingly, a ping-pong mechanism

could be discarded. In addition, a rapid equilibrium ordered mechanism was ruled out

because the families of double reciprocal plots obtained intersect on the left of the y-axis

(Fig. 2A and Fig. 2B).

The steady-state kinetic constant values given here are larger than the apparent

kinetic parameters we have previously reported [10], whose data were collected in Tris-HCl

100mM, pH 7.0. We verified a significant reduction in the catalytic rate for this enzyme

when the assay mix was buffered using Tris-HCl instead of potassium phosphate (data not

show). In addition, here we report true steady-state kinetic parameters instead of apparent

kinetic parameters.

Determination of steady-state kinetic parameters for Pisum sativum SD [23] in 100

mM potassium phosphate, pH 7.4, yielded values of 340 ± 40

M for K

DHS

, which is

approximately 8-fold larger than the value for MtbSD (44

M), and 4.3 ± 0.5

M for

NADPH

, which is approximately 8-fold smaller than the value here reported for MtbSD (34

M).

The first step in determining the mechanism of action of an enzyme is to establish

its kinetic mechanism, the order of substrate binding and release of products. This is most

commonly achieved by product inhibition studies in which we observe the patterns

obtained when substrate concentrations are varied at fixed levels of other substrate(s) in the

presence of products, yielding information not available from initial velocity studies.

Product inhibition data for MtbSD (Table 1) were similar to those for P. sativum

SD. For the reaction of P. sativum SD [23], NADPH was found to give linear competitive

inhibition with NADP

as variable substrate, and NADPH was a noncompetitive inhibitor

versus SHK at low NADP

concentrations. The results for SHK inhibition versus DHS

indicated slightly noncompetitive inhibition, thereby suggesting that the mechanism is

ordered [23]. Product inhibition data presented here suggest two possible kinetic

mechanisms for MtbSD, steady-state ordered and rapid equilibrium random with two dead-

end ternary complexes (MtbSD-NADPH-SHK and MtbSD-NADP

-DHS). An application

of primary deuterium isotope effects is to distinguish among possible kinetic mechanisms.

Accordingly, to differentiate between steady-state ordered and rapid equilibrium random,

analysis of primary kinetic isotope effects was carried out.

Measurements of the steady-state primary deuterium kinetic isotope effects on

apparent V/K for one substrate were carried out at various cosubstrate concentrations (Fig.

3). Based on the mechanistic deductions from isotope effects for multireactant enzymes

developed by Cook and Cleland [24], the primary isotope effect on

(V/K

app

)

independent of the concentration of A and

(V/K

app

)

decreases as the concentration of B

increases and reaches a limiting value of 1.0 at infinite concentration of B for a steady-state

ordered mechanism. A and B refer to, respectively, the first and second substrates to bind to

the enzyme. The kinetic isotope effects on V/K results (Fig. 3) are in agreement with a

steady-state ordered bi-bi mechanism with DHS binding first followed by NADPH binding

to MtbSD active site (Fig. 9). In addition, the DHS substrate is sticky, hence it reacts to

give SHK as fast as, or faster than, it dissociates from the enzyme [24]. Based on initial

velocity patterns, product inhibition, analysis of Haldane relationships and isotope-

exchange studies, an ordered kinetic mechanism with NADPH binding first has been

proposed for P. sativum SD [23]. Initial velocity and dead-end inhibition studies showed

that the kinetic mechanism is ordered bi-bi for mouse class II alcohol dehydrogenase with

coenzyme binding first [25]. The NAD-dependent mouse class II alcohol dehydrogenase is

structurally homologous to E. coli SD dinucleotide-binding domain [20].

Stereospecificity and Rate-limiting Steps - The primary deuterium kinetic isotope

effects indicate that the C

-proS hydride (B side) is transferred to DHS in the oxy-reduction

reaction catalyzed by MtbSD. In contrast, E. coli SD has been shown to transfer the C

proR hydride (A-side) of NADPH [26], in agreement with the crystal structure of the

enzyme [20]. Isotope effects on V report on events following formation of the ternary

complex capable of undergoing catalysis, which include the chemical steps, possible

enzyme conformational changes, and product release [27]. Isotope effects on V/K report on

steps in the reaction mechanism from the binding of the isotopically labeled substrate to the

first irreversible step, usually the release of the first product. The apparent classical limit for

primary deuterium kinetic isotope effects on the maximal velocity is around 8 if carbon-

hydrogen bond cleavage is the rate-determining step, even though values as small as 2 have

been used, in a less rigorous practice, as evidence for rate-limiting steps [27]. The primary

kinetic isotope effect values of 1.8 for

DHS

and 1.5 for

NADPH

using [4S-

H]NADPH as

reductant indicate that the hydride transfer is only partly rate limiting for MtbSD enzyme

catalysis. Isotope-insensitive steps such as enzyme isomerization and/or product release are

probably contributing to the rate-limiting steps of the MtbSD reaction. The

(V/K

app

)

NADPH

and

(V/K

app

)

DHS

values indicate that substrate binding steps make a small contribution to

the rate-limiting steps.

Steady-state solvent isotope effects were evaluated to assess the contribution of

solvent proton transfer to a step in the enzymatic mechanism. The effect on V arises from

solvent-exchangeable protons being transferred during catalysis. The values for solvent

isotope effect on V for DHS (1.5) and NADPH (1.3) indicate that solvent-exchangeable

protons being transferred during catalysis is only partly rate limiting in the overall reaction.

No solvent isotope effect on V/K was observed for DHS and NADPH.

The number of protons transferred during the oxy-reduction reaction catalyzed by

MtbSD was determined by the proton inventory technique. Measurements of V in different

isotopic solvent mixtures (V relative versus mol fraction of D

O) showed a linear

relationship (Fig. 4A – inset), suggesting that a single proton is transferred in the step that

exhibits the solvent isotope effect [28]. A similar result was observed for the M.

tuberculosis NADPH-dependent mycothione reductase [29].

Chemical Mechanism of MtbSD - Double isotope effect studies are able to

distinguish whether two different isotopic substitutions affect the same or different

chemical steps. Assuming that the primary deuterium kinetic isotope effect is expressed

only on the hydride transfer reaction (Fig. 5A and 5B – insets) and solvent isotope effects

affect only the alkoxide protonation, double isotope effects can distinguish whether these

isotopic substitutions affect the same or different chemical steps (Fig. 5A and 5B). Theory

predicts that if protonation and hydride transfer occur in the same transition state, the

primary isotope effects will be larger or unchanged with D

O as compared to H

O. On the

other hand, if hydride transfer and protonation occur in distinct steps, the primary isotope

effects will be smaller with D

O as solvent, as proton transfer will become more rate

limiting [30, 31]. The increased values for the primary isotope effects on V/K measured in

O (

V/K

DHS(D

= 1.3 and

V/K

NADPH(D

= 1.5) as compared to values measured in H

(

V/K

DHS

= 1.0 and

V/K

NADPH

= 1.4) are in agreement with both hydride and proton

transfer taking place in the same step (concerted mechanism), thus implying a single

transition state. It should be pointed out that use of D

O allowed the observation of a

significant value for

V/K

DHS

(1.3), even though this substrate is sticky and reaction

proceeds through a steady-state ordered kinetic mechanism.

The

-elimination catalyzed by bovine liver crotonase is concerted [32], as well as

are the reactions catalyzed by pig liver acyl-CoA dehydrogenase [33] and isocitrate

dehydrogenase [34]. Concerted reactions are a common strategy utilized by enzymes to

avoid unstable intermediates, such as some enolate anions [35, 36].

The role of acid/base chemistry in the mechanism of MtbSD was determined using

the pH dependence of k

cat

and k

cat

for DHS and NADPH in pH range of 5.5 ≤ pH ≤ 10.0

(Fig. 6). In this experiment the k

cat

for both substrates are very similar and decrease at high

pHs with a slope of -1 (Fig. 6A), indicating that deprotonation of a single residue with

apparent pK

value of 8.9 ± 0.1 abolishes MtbSD catalytic activity. The k

cat

profiles for

DHS and NADPH show that deprotonation of a single ionizable group with pK

value of

9.1 ± 0.1 abolishes DHS and NADPH binding (Fig. 6B and 6C). It is likely that the

ionization behavior of the same group is being observed in the pH profiles for k

cat

and

cat

, and the difference in the pK values may reflect perturbation of the pK value upon

substrate(s) binding to MtbSD. The pK values for k

cat

and k

cat

lie in the normal pK range

for

-amino of lysine, thiol of cysteines and phenolic hydroxyl of tyrosine amino acids. The

pH-rate profiles of H. influenzae SD for the reverse reaction indicated that a group with pK

value of ~ 8.1 needs to be deprotonated for activity, and site-directed mutagenesis results

were consistent with Lys-67 playing a key role in catalysis [37].

Based on the double isotope effects and pH-rate profiles, we propose a chemical

mechanism for MtbSD (Fig. 10), in which hydride transfer and solvent proton transfer are

concerted, and an amino acid residue with pK

value of 8.9 is involved in catalysis.

Analysis of the structure of E. coli SD shows that DHS is positioned for

stereospecific reduction to SHK and the side chain of Lys65 has been proposed to hydrogen

bond to the C-4 hydroxyl group of DHS/SHK (E.coli SD numbering) [20]. This amino acid

residue has been proposed to be the acid/base catalytic group that donates a proton to the

carbonyl of DHS during reduction and that removes a proton during oxidation of SHK [20].

In agreement, previous studies of P. sativum SD showed that substrate-like inhibitors of the

enzyme require a C-4 hydroxyl, whereas either a C-5 hydroxyl or carboxylate group is

needed for strong binding [38]. Analogs of DHS substrate that lack the C-4 and C-5

hydroxyls were used to demonstrate the role of both C-5 hydroxyl and C-4 hydroxyl on the

substrate specificity for the E. coli SD, and the results showed that the C-4 hydroxyl has a

very significant effect on the specificity of the substrate [39]. It has been suggested that the

C-4 hydroxyl group hydrogen bonds to a charged group in the E. coli SD active site based

on an estimation of the binding energy given by k

cat

[39]. However, it should be kept in

mind that participation of a lysine side chain residue in the chemical reaction catalyzed by

E. coli SD is based on a molecular model, since the crystal structure is for E. coli SD-

NADP

binary complex, and E. coli SD transfers the C

-proR hydride (A-side) of NADPH.

Based on pH rate profiles of P. sativum SD, it has been postulated that a group with pK

value of 9.4, possibly or an

-amino group of lysine, binds the carboxylate ion of the

substrate, while a group of pK

8.6, possibly a sulphydryl residue, interacts with the C-4

hydroxyl group of DHS/SHK [40]. The crystal structure of M. jannaschii SD has been

determined and the residues involved in DHS binding and its catalytic reduction were

identified, amongst them Lys70 (M. jannaschii SD numbering) [22]. The side chain of

Asp102 of E. coli SD has also been suggested to form a hydrogen bond to C-4 hydroxyl

group of the substrate [20]. In addition, analysis of the crystal structure of H. influenzae SD

and site-directed mutagenesis studies have shown that Asp103 and Lys67 (H. influenzae

SD numbering) may function as a catalytic pair involved in acid/base catalysis for this

enzyme [37]. However, the pH-rate profiles for MtbSD did not show any residue whose

protonation would abolish enzyme activity at low pH values. However, it could be argued

that the lowest pH value (5.0) of the pH-rate profiles presented here would not allow

detection of carboxylic groups of aspartate side chains since its pK value is approximately

4. Notwithstanding, the pK for the carboxylate group of DHQ/SHK is approximately 4.1

and would be a difficult task to discriminate between substrate and/or amino acid ionizing

groups. We have previously shown by analysis of multiple sequence alignment of MtbSD,

E. coli SD, H. influenzae SD-like, and M. jannaschii SD that Lys69 and Asp105 (M.

tuberculosis numbering) are conserved [10]. The pH-rate profiles described here show

participation of a group whose deprotonation abolishes binding (pK

= 9.1) and catalytic

activity (pK

= 8.9), consistent with participation of Lys69. At any rate, site directed

mutagenesis of these residues and measurements of the steady-state kinetic parameters for

MtbSD enzyme reaction are currently underway to evaluate the role, if any, of Lys69 and

Asp105 may play in binding and/or catalysis.

MtbSD Structure - The MtbSD structure belongs to the homologous superfamily

denominated NAD(P)-binding Rossmann-fold domain [41]. The SDH family provides a

new example of a protein family displaying the dinucleotide binding-fold, without

significant sequence homology with other Rossmann-fold families [20]. The structure has

an elongated shape with two domains (Fig. 8). Each domain has an α/β architecture. The

domains are linked by an α-helix and a turn that keep them together and form a deep groove

into which the NADP cofactor binds. The NADP-binding domain, in the C-terminus,

adopts a nearly canonical Rossmann fold, a six-stranded parallel

-sheet with six

-helices,

three on each side of the

-sheet. The N-terminus consists of a six-stranded

-sheet and six

-helices, however the arrangement is irregular, since the near central

5 strand is in an

antiparallel orientation with respect to the other strands [20].

In the analysis of the E. coli SD structure utilized as template, the substrate-binding

site was identified by the position of the nicotinamide ring of the cofactor [20]. The

following residues conserved in the SD family were identified: Ser14, Ser16, Lys65,

Asn86, Thr101, Asp102, and Gln244 [20]. Both carboxyl oxygens from DHS form two

hydrogen bonds to the protein using probably the conserved serine residues at positions 14

and 16 [20]. These residues were identified in the MtbSD model (Ser18, Ser20, Lys69,

Asn90, Thr104, Asp105, and Gln243; MtbSD numbering), and they are present in

equivalent positions as showed in Fig. 11. The conserved residues Ser14, Ser16 and Tyr215

have been suggested to be involved in substrate carboxylate binding for E. coli SD [20].

These residues were found in the MtbSD model (Fig. 11) to be in equivalent positions and

appear to play a role in binding of the carboxylate group of DHS/SKH.

In the multiple sequence alignment between the target (MtbSD) and templates (E.

coli YdiB and E. coli SD) we identified the sequence Gly124-Ser125-Gly126-Gly127-

Thr128-Ala129 as MtbSD sequence pattern (G [A,s,g] G G [A,t] [A,S,g]), corresponding to

the diphosphate-binding loop, which is conserved in the entire SDH family [20] (Fig. 7).

Interaction between the diphosphate-binding loop and the active site in the E. coli SD

occurs through Gly129 and Ala130 residues, which correspond to, respectively, Thr128 and

Ala129 in MtbSD (Fig. 7). For the binding of the nicotinamide in E. coli SD it was verified

that the amide group N-7 of the nicotinamide ring is hydrogen-bonded to the carbonyl

group of Met213 and of the invariant Gly237 [20]. In the MtbSD model, the Gly236 (the

correspondent position for Gly237) is hydrogen-bonded to the amide group N-7. On the

other hand, the MtbSD model presents Ala213 residue in equivalent position of the E. coli

Met213. Notwithstanding, an equivalent bond with the amide group N-7 of nicotinamide

ring is maintained with MtbSD Ala213. The E. coli SD Arg150 and Arg154 play a crucial

role in adenine phosphate binding as they form an “electrostatic clamp” that sandwiches the

phosphate substituent of NADP

[20]. In the MtbSD model it was found that Arg149 and

Lys153 residues occupy equivalent positions, and it is thus tempting to suggest that these

residues play a role in adenine phosphate binding in MtbSD.

It has been suggested that NAD

binding in E. coli YdiB is favored by the

substitution of residues Thr151 and Arg154 of E. coli SD (E. coli SD numbering) by

Asp158 and Phe160 (E. coli YdiB numbering), respectively [20]. The hydrophobic residue

Phe160 creates a neutral environment, which is less discriminating than the basic binding

pocket with Arg154 from the SD structure. The Asn157 and Lys160 residues of NADP

dependent MtbSD, which correspond to Thr151 and Arg154 of NADP

-dependent E. coli

SD [20], occupy equivalent positions and appear to play a similar role in creating a more

discriminating environment in favor of NADP

substrate.

The work here presented is, to the best of our knowledge, the first detailed report on

the steady-state velocity patterns, product inhibition, primary deuterium kinetic isotope

effects, solvent kinetic isotope effects, proton inventory, double isotope effects, pH-rate

profiles, and molecular modeling of MtbSD. Initial velocity patterns in the forward

reaction, product inhibition studies, and primary deuterium kinetic isotope effects allowed

us to propose a steady-state ordered bi-bi kinetic mechanism for catalysis, with DHS

binding first followed by NADPH binding to the MtbSD enzyme catalytic site. The primary

deuterium kinetic isotope effects indicated that the C

-proS hydride is transferred from the

NADPH co-substrate in a step that is only partly rate limiting. Solvent kinetic isotope

effects demonstrated that proton transfer from the solvent is only partly rate limiting.

Proton inventory results indicated that a single proton is transferred in the solvent-sensitive

step. Double isotope effects showed that transfer of hydride and proton occur in concert.

The pH-rate profiles revealed that a charged group, probably the ε-amino group of Lys69

plays an important role in catalysis and substrate binding. The homology 3D model for

MtbSD identified the probable residues in the substrate binding site, amongst them Lys69,

in agreement with pH studies, and Asp105. These results allowed us to propose a kinetic

and chemical mechanism for MtbSD. The enzyme kinetics and modeling studies provide a

framework on which to base the design of enzyme inhibitors with potential use as

antitubercular agents. Site-directed mutagenesis, equilibrium binding spectrofluorimetry,

and pre-steady state experiments are currently underway to improve our understanding of

the mechanism of action of MtbSD.

Acknowledgements

Financial support for this work was provided by Millennium Initiative Program MCT-

CNPq, Ministry of Health-Department of Science and Technology (DECIT)-UNESCO

(Brazil) to D.S.S. and L.A.B. D.S.S. and L.A.B. also acknowledge grants awarded by

CNPq, FINEP, and PRONEX/FAPERGS/CNPq. D.S.S. (CNPq, 304051/1975-06) and

L.A.B. (CNPq, 520182/99-5) are researchers awardees from the National Research Council

of Brazil. We thank Professor John W. Frost, Department of Chemistry of Michigan State

University, for his generous gift of 3-dehydroshikimate substrate.

References

1. C. Dye, S. Scheele, P. Dolin, V. Pathania, M. C. Raviglione, JAMA 282 (1999)

677-686.

2. L. A. Basso, J. S. Blanchard, Adv. Exp. Med. Biol. 456 (1998) 115-144.

3. A. Plabos-Méndez, D.K. Gowda, T.R. Frieden, World Health Organ. 80 (2002)

489-495.

4. CDC (Centers for Disease Control and Prevention), Morb. Mortal. Wkly. Rep. 55

(2006) 301-305.

5. R. J. O’Brien, P. P. Nunn, Am. J. Respir. Crit. Care Med. 162 (2001) 1055-1958.

6. S. T. Cole, R. Brosch, J. Parkhill, T. Garnier, C. Churcher, D. Harris, S. V. Gordon,

K. Eiglmeier, S. Gas, C. E. Barry 3

, F. Tekaia, K. Badcock, D. Basham, D. Brown,

T. Chillingworth, R. Connor, R. Davies, K. Devlin, T. Feltwell, S. Gentles, N.

Hamlin, S. Holroyd, T. Hornsby, K. Jagels, A. Krogh, J. McLean, S. Moule, L.

Murphy, K. Oliver, J. Osborne, M. A. Quail, M. A. Rajandream, J. Rogers, S.

Rutter, K. Seeger, J. Skelton, R. Squares, S. Squares, J. E. Sulston, K. Taylor, S.

Whitehead, B. G. Barrell, Nature 393 (1998) 537-544.

7. T. Parish, N. G. Stoker, Microbiology 148 (2002) 3069-3077.

8. J.R. Coggins, C. Abell, L.B. Evans, M. Frederickson, D.A. Robinson, A.W. Roszak,

A.P. Lapthorn, Biochem. Soc. Trans. 31 (2003) 548-552.

9. M. L.Magalhães, C. P. Pereira, L. A. Basso, D. S. Santos, Protein Expr. Purif. 26

(2002) 59-64.

10. I. O. Fonseca, M. L. B Magalhães, J. S. Oliveira, R. G. Silva, M. A.Mendes, M. S.

Palma, D. S. Santos, L. A. Basso, Protein Expr. Purif. 46 (2006) 429-437.

11. G. Ottolina, S. Riva, G. Carrea, B. Danieli, A. F. Buckmann, Biochem. Biophys.

Acta 998, (1989) 173-178.

12. G. A.Orr, J. S. Blanchard, Anal. Biochem. 142, (1984) 232-234.

13. H.M. Berman, J. Westbrook, Z. Feng, , G. Gilliland, T.N. Bhat, H. Weissig,

I.N.Shindyalov, P.E. Bourne, Nucleic Acids Res. 28, (2000) 235-242.

14. S.F. Altschul, T.L. Madden, A.A. Schäffer, J. Zhang, Z. Zhang, W. Miller, D.J.

Lipman, Nucleic Acids Res. 25 (1997) 3389-3402.

15. J.D. Thompson, D.G. Higgins, T.J. Gibson, Nucleic Acids Res. 22 (1994) 4673-

4680

16. A Šali, T.L. Blundell, J. Mol. Biol. 234 (1993) 779-815.

17. M.A. Martí-Renom, A.C. Stuart, A. Fiser, R. Sánchez, F. Melo, A. Šali, Annu. Rev.

Biophys. Biomol. Struct. 29 (2000) 291-325.

18. R.A. Laskowski, M.W. MacArthur, D.S. Moss, J.M. Thornton, J. Appl. Cryst. 26

(1993) 283-291.

19. R. Lüthy, J.U. Bowie, D. Eisenberg, Nature 356 (1992) 83-85.

20. G. Michel, A. W. Roszak, V. Sauvé, J. Maclean, A. Matte, J. R. Coggins, M.

Cygler, A. J. Lapthorn, J. Biol. Chem. 278 (2003) 19463-19472.

21. N. Guex, M. C. Peitsch, Electrophoresis 18 (1997) 2714-2723.

22. A.K Padyana, S.K. Burley, Structure 11 (2003) 1005-1013.

23. D. Balinsky, A. W. Dennis, W. W. Cleland, Biochemistry 10 (1971) 1947-1952.

24. P. F. Cook, W. W. Cleland, Biochemistry 20 (1981) 1790-1796.

25. P. Strömberg, S. Svensson, K. B. Berst, B. V. Plapp, J. O. Höög, Biochemistry 43

(2004) 1323-1328.

26. P. Dansette, R. Azerad, Biochimie 56 (1974) 751-755.

27. D. B. Northrop, Biochemistry 14 (1975) 2644-2651.

28. D. M. Quinn, L. D. Sutton, in: P. F. Cook (Ed.), Enzyme Mechanism from Solvent

Isotope Effects, CRC Press, Florida, 1991, pp 73-126.

29. M. P. Patel, J. S. Blanchard, Biochemistry 40 (2001) 5119-5126.

30. J. D. Hermes, C. A. Roeske, M. H. O’Leary, W. W. Cleland, Biochemistry 21

(1984) 5106-5114.

31. J. G. Belasco, J. Albery, J. R. Knowles, J. Am. Chem. Soc. 105 (1983) 2475-2477.

32. Gerlt, J. A. Bioorganic Chemistry: Peptides and Proteins, Oxford University Press,

New York, 1998.

33. B. J. Bahnson, V. E. Anderson, Biochemistry 30 (1991) 5894-5906.

34. B. Pohl, T. Raichle, S. Ghisla, Eur. J. Biochem. 160 (1986) 109-115.

35. P. F. Cook, W. W. Cleland, Biochemistry 20 (1981) 1797-1805.

36. A. Thibblin, W. P. Jencks, J. Am. Chem. Soc. 101 (1979) 4963-4973.

37. S. Singh, S. Korolev, O. Koroleva, T. Zarembinski, F. Collart, A. Joachimiak, D.

Christendat, J. Biol. Chem. 280 (2005) 17101-17108.

38. D. Balinsky, D. D. Davies, Biochem. J. 80 (1961) 296-300.

39. T. D. Bugg, C. Abell, J. R. Coggins, Tetrahedron Lett. 29 (1988) 6779-6782.

40. A. W. Dennis, D. Balinsky, Int. J. Biochem. 3 (1972) 93-102.

41. A. G. Murzin, S. E. Brenner, T. Hubbard, C. Chothia. J. Mol. Biol. 247 (1995) 536-

540.

Figure Legends

Fig. 1. Shikimate dehydrogenase-catalyzed reaction.

Fig. 2. Initial velocity patterns for MtbSD with both substrates DHS (A) and NADPH (B) as

variable substrate. Each curve represents varied-fixed levels of the cosubstrate. Both DHS

and NADPH concentrations varied from 5 to 200

M. One unit of enzyme activity (U) is

defined as the amount of enzyme catalyzing the conversion of 1

mol of NADP

per

minute at 25 °C.

Fig. 3. Dependence of the apparent

V/K

DHS

values on the concentration of the cosubstrate.

The data were fitted to an equation describing a hyperbolic decay, which yielded a limiting

value of 1.0 ± 0.03 for the

V/K

DHS

. The inset shows that

V/K

NADPH

values do not depend

on the concentration of the cosubstrate.

Fig. 4. Solvent isotope effects for MtbSD. (A) DHS as varied substrate, with saturating

concentration of the cosubstrate. Reaction mix contained either 0 (●) or 90 (▼) atom %

O. Inset represents the proton inventory on MtbSD, with both substrates at saturating

concentrations. (B) NADPH as varied substrate, with saturating concentration of the

cosubstrate. Reaction mix contained either 0 (●) or 90 (▼) atom % D

Fig. 5. Multiple isotope effects for MtbSD. (A) NADPD as varied substrate, with saturating

concentration of DHS. Reaction mix contained either 0 (•) or 90 (▼) atom % D

O. Inset

represents the primary isotope effect, NADPH (•) or NADPD (▼) as varied substrate with

saturating concentration of the DHS. (B) DHS as varied substrate, with saturating

concentration of NADPD. Reaction mix contained either 0 (•) or 90 (▼) atom % D

Inset represents the primary isotope effect, DHS as varied substrate with saturating

concentration of NADPH (•) or NADPD (▼).

Fig. 6. Dependence of MtbSD kinetics parameters on pH. (A) pH dependence of log k

cat

(B) pH dependence of log k

cat

DHS

, and (C) pH dependence of log k

cat

NADPH

Experimental data were fitted to equation 7.

Fig. 7. ClustalW multiple sequence alignment between the target (MtbSD) and templates

(E. coli YdiB and E. coli SD). YdiB is more similar to MtbSD than to E. coli SD and

optimal alignment between the three sequences requires the insertion of two gap regions in

the C-terminus of MtbSD and E. coli SD sequences (underlined in YdiB).

-helices and

strands are represented as H and S, respectively. This sequence alignment was created

using the following sequences from GeneBank

: M. tuberculosis SD (CAB06186,

residues 1–269), E. coli K12 YdiB (NP 416207, residues 1-288), and E. coli K12 SD

(NP_417740, residues 1-272).

Fig. 8. Stereo ribbon representation of the MtbSD structure. The MtbSD structure is

characterized by a nearly canonical Rossmann fold with a six-stranded parallel

sheet (C-

terminal domain on top), and a

N-terminal domain (bottom) with an antiparallel

strand (

5). The α-helices are shown in light gray and the

-strands in black.

Fig. 9. Proposed kinetic mechanism for M. tuberculosis shikimate dehydrogenase.

Fig. 10. Proposed chemical mechanism for M. tuberculosis shikimate dehydrogenase-

catalyzed reaction. R represents ribose, adenosine diphosphate, and 2’-phosphate moieties

of NADPH.

Fig. 11. The conserved residues in the nicotinamine ring pocket current in the structures of

both E. coli SD (black) and in the MtbSD (gray). The tyrosine and serine residues involved

in hydrogen-bond with the DHS carboxylate group are in equivalent positions in both E.

coli SD (Tyr215, Ser14, and Ser16; black) and in the MtbSD (Tyr215, Ser18, and Ser20;

gray).

Table1: Product Inhibition Patterns for M. tuberculosis Shikimate Dehydrogenase

Varied

substrate

Product

inhibition

Inhibition type

(μM)

(

DHS SHK C 62.9 ± 10.5

NADPH SHK NC 256.3 ± 35.4 142.5 ± 74.0

DHS NADP

NC 30.4 ± 1.3 28.3 ± 7.6

NADPH NADP

C 8.3 ± 1.3

At 25°C and 100 mM potassium phosphate, pH 7.3.

C = competitive, NC-MT =

noncompetitive mixed-type.

is the slope inhibition constant.

is the intercept

inhibition constant.

Table 2: Kinetic Isotope Effects for M. tuberculosis Shikimate Dehydrogenase

Parameter Isotope Effect*

V/K

DHS

1.0 ± 0.03

DHS

1.8 ± 0.09

V/K

NADPH

1.4 ± 0.3

NADPH

1.5 ± 0.1

V/K

DHS

1.0 ± 0.01

DHS

1.5 ± 0.3

V/K

NADPH

1.0 ± 0.03

NADPH

1.3 ± 0.2

V/K

DHS(D

1.3 ± 0.1

DHS(D

2.6 ± 0.01

V/K

NADPH(D

1.5 ± 0.1

NADPH(D

2.5 ± 0.1

* value ± standard error obtained from fitting the data to

the appropriated equation

Abbreviations used: DHS, 3-dehydroshikimate; MDR-TB, multidrug-resistant

tuberculosis; MtbSD, Mycobacterium tuberculosis shikimate dehydrogenase; NADP

oxidized

-nicotinamide adenine dinucleotide phosphate; NADPH, reduced

-nicotinamide

adenine dinucleotide phosphate; NADPD, deuterated

-nicotinamide adenine dinucleotide

phosphate; SD, shikimate dehydrogenase; SHK, D-shikimate; TB, tuberculosis; XDR-TB,

extensively drug-resistant tuberculosis.

PARTE III

Resultados e Discussão

ARTIGO 1: “Functional shikimate dehydrogenase from

Mycobacterium tuberculosis H37Rv: Purification and

characterization”

Obs.: As figuras e tabelas em negrito, mencionadas na discussão a seguir,

estão identificadas de acordo com o artigo a que se referem.

Purificação e Análise da enzima recombinante MtbSD

No artigo intitulado “Functional shikimate dehydrogenase from

Mycobacterium tuberculosis H37Rv: Purification and characterization”, o

protocolo de purificação para a enzima MtbSD utiliza quatro etapas

cromatográficas, onde a proteína recombinante MtbSD foi purificada 8,5 vezes.

Foram necessárias 49g de células E. coli BL21(DE3) com vetor de expressão

pET23a(+) contendo o gene aroE de M. tuberculosis para obtenção de ~ 10 mg

de MtbSD pura. Esse protocolo de purificação e a análise da proteína

recombinante pura foram feitos na dissertação de mestrado da Maria de

Lourdes Borba Magalhães pelo Programa de Pós-Graduação em Biologia

Molecular e Celular da Universidade Federal do Rio Grande do Sul.

Neste protocolo a primeira etapa cromatográfica, utilizando-se da Q-

Sepharose Fast Flow, resultou na maior perda de todo o processo. Apesar

disso, essa etapa foi interessante por possibilitar a redução da quantidade das

proteínas totais em 90% e concentrar a proteína de interesse. Na etapa

subseqüente utilizou-se uma coluna de interação hidrofóbica, a Phenyl

Sepharose High Performance, que resultou em uma taxa de purificação de

quatro vezes e um rendimento de 60% em relação à etapa anterior. A terceira

etapa foi feita em gel filtração, através da matriz Sephacryl High Resolution S-

200 e por fim, a coluna de troca aniônica, Mono Q-HR, foi necessária para o

refinamento da purificação da MtbSD (Tabela 1 e Figura1).

A proteína purificada foi ensaiada monitorando a absorbância (340 nm)

da redução do cofator NADP

a NADPH por meio da oxidação do D-chiquimato

(SHK) que leva a formação do 3-desidrochiquimato (DHS) (Figura 3). É

possível verificar que a atividade enzimática da proteína purificada foi

linearmente dependente do volume de proteína homogênea à mistura de

reação, mostrando que a velocidade inicial é proporcional à concentração total

de enzima (Figura 2).

A análise do produto da purificação, por espectrometria de massas em

ESI-MS, mostra dois picos (Figura 4). Um pico em 27.207 Da que corresponde

à massa molecular da seqüência primária prevista para a MtbSD, com a perda

da metionina N-terminal; e o segundo pico em 54.150 Da que corresponde à

presença da forma dimérica da MtbSD. O grau de pureza obtido da proteína

pura foi de 98%. A ausência de um pico em 29.413 Da no espectro de massas

confirma que a MtbSD foi separada da enzima EcoliSD (CHAUDURI &

COGGINS, 1985), hospedeiro utilizado na expressão da proteína alvo.

Figura 4: Espectro de massas da MtbSD. O espectro mostra a massa molecular da enzima

na forma monomérica e dimérica. O eixo de valores x representa a massa molecular e o eixo

de valores y a intensidade do pico.

A comprovação da natureza dimérica da enzima em solução foi feita por

meio da cromatografia líquida em gel filtração utilizando-se a coluna Superdex-

200 e uma curva de calibração (Figura 5). As SDs de E. coli (MICHEL e cols.,

2003) e H. influenzae (YE e col., 2003) apresentam-se na forma monomérica

em solução, em contrapartida a MjanSD é dimérica (PADYANA & BURLEY,

2003).

FIGURE 5: Determinação da massa molecular da MtbSD em solução. A amostra foi

aplicada na coluna Superdex 200 HR 10/30 e a corrida foi feita em um fluxo de 0.4 mL min

-1

FPLC. Os padrões de massa molecular foram Ribonuclease A (13.700 Da), Chymotrypsinogen

(25.000 Da), Ovalbumin (43.000 Da) e Albumin (67.000 Da).

A seqüência dos resíduos de aminoácidos no N-terminal da proteína

purificada, SEGPKKAGVLG, foi determinada pelo método de química

degradativa de Edman, confirmando assim que a proteína pura trata-se da

MtbSD e que realmente ocorre a remoção da metionina N-terminal, um tipo

comum de modificação pós-traducional de proteínas sintetizadas em sistemas

procariotos.

Posteriormente este protocolo de purificação foi otimizado, reduzindo-se

a purificação para três etapas cromatográficas. Em torno de 5 g de células

foram necessárias para obtenção de 130,3 mg de proteínas totais, cujo

rendimento de 13,5 % resultou em 1,5 mg de MtbSD pura. Essa preparação

protéica foi utilizada para todos os ensaios de caracterização e de cinética.

Figura 6: Análise por SDS-PAGE das frações oriundas das etapas cromatográficas da

purificação da MtbSD. Canaleta 1: marcador de peso molecular Perfect Marker (GIBCO);

canaleta 2: extrato bruto (40 μg de proteínas); canaleta 3: pool das frações da Phenyl

Sepharose FF (40 μg de proteínas), canaleta 4: pool das frações da Sephacryl-S200 (20 μg de

proteína), canaleta 5: pool das frações da Mono-Q (10 μg de proteína).

Tabela1: Purificação da enzima chiquimato desidrogenase de M. tuberculosis H37Rv.

Etapa de

Purificação

Proteínas

totais (mg)

Unidades

totais (U)

Atividade

específica

(U/mg)

Taxa de

purificação

Rendimento

(%)

Extrato bruto

130,3 156,5 1,2 1 100

Phenyl

Sepharose FF

25,1 113,0 4,5 3,8 72,2

Sephacryl

S200

8,9 81,9 9,2 7,7 52,3

MonoQ

1,5 21,2 14,1 11,8 13,5

Analisando a tabela acima de purificação, a otimização da purificação

ocorreu pela mudança da primeira etapa cromatográfica, onde a coluna de

troca iônica, a Q-Sepharose Fast Flow foi substituida pela coluna de interação

hidrofóbica Phenyl Sepharose Fast Flow, visto que a coluna cromatográfica de

interação hidrofóbica foi bem sucedida no primeiro protocolo de purificação

proposto e a coluna Q-Sepharose Fast Flow resultou na maior perda do

processo anterior. Este novo protocolo de purificação reduz uma etapa na

purificação da MtbSD recombinante, melhorando os valores de rendimento e

taxa de purificação.

Os protocolos de purificação estabelecidos aqui, não usam a adição de

seqüências extras para o alcance da proteína purificada; como é o caso da

HinfluSD que utiliza uma cauda de histidinas (YE e cols., 2003). Já foi

demonstrado que esse tipo de artefato pode exercer um efeito negativo em

posteriores estudos da proteína recombinante, alterando à nível de atividade

biológica ou até mesmo alterações estruturais (WOESTENENK e cols., 2004).

Na purificação da enzima MjanSD utilizaram duas etapas cromatográficas e

após mais uma etapa de remoção proteolítica da cauda (PADYANA &

BURLEY, 2003).

Caracterização e Determinação dos Parâmetros Cinéticos Aparentes

O artigo em questão também descreve os parâmetros cinéticos

aparentes no sentido direto e reverso da reação catalisada pela enzima MtbSD

recombinante. É interessante ressaltar que esses resultados para a reação

direta foram os primeiros descritos na literatura para a família das SDs de

procariotos.

Os dados foram coletados e analisados em equação hiperbólica,

indicando que a reação catalisada pela MtbSD segue a cinética de Michaelis-

Menten para todos os substratos (Figura 3). Os valores de K

e V

max

encontrados para 3-desidrochiquimato (DHS) foram de 31 μM e 108 U mg

-1

respectivamente; e para o cofator NADPH foram de 10 μM e 100 U mg

-1

(Tabela 2). O valor de K

para o substrato DHS é bastante inferior ao valor

encontrado para a SD purificada de Pisum sativum (PsatSD), cujo valor

descrito é de 340 μM (BALINSKY e cols., 1971). Entretanto, o valor de K

encontrado para o cofator NADPH é próximo ao determinado para a enzima

PsatSD em condições semelhantes de pH (4,3 μM). Na reação reversa, os

valores aparentes de K

e V

max

para os substratos SHK e NADP

foram de

50,18 μM e 18 U mg

-1

; e de 22 μM e 12,9 U mg

-1

, respectivamente (Tabela 2).

O valor de K

para o SHK é 10 vezes maior para PsatSD em relação ao de

MtbSD. No entanto, os valores de K

para os substratos SHK e NADP

das

enzimas MtbSD e EcoliSD são bastante próximos, indicando que as duas

enzimas catalisam a reação reversa com constantes cinéticas semelhantes.

Os valores de V

max

para os substratos DHS e SHK rederam k

cat

e k

cat

para esses substratos de 49 s

-1

e 1,6 x 10

-1

, e 8,2 s

-1

e 1,63 x 10

-1

respectivamente (Tabela 2).

A constante de equilíbrio aparente (K

) também foi estimada utilizando

os parâmetros cinéticos aparentes em estado estacionário. Por meio da

equação de Haldane o valor de 19,6 foi alcançado para K

. A K

determinada

em pH 7.4 para PsatSD foi de 10,3. É importante ressaltar que a metodologia

utilizada para essa determinação considera o mecanismo cinético como sendo

aleatório em rápido equilíbrio. Nesse caso é assumido que a constante de

dissociação para o segundo substrato não é alterado pela ligação do primeiro

substrato.

O efeito da temperatura na estabilidade da proteína MtbSD foi verificado.

Foram incubadas amostras da proteína recombinante em temperaturas

diferentes (15, 25, 37 e 55°C) e alíquotas foram retiradas em tempos diferentes

de incubação para realização da atividade enzimática (25°C). A MtbSD se

mostra uma proteína bastante estável nas temperaturas de 15, 25 e 37°C

durante 1 hora de incubação (Figura 4). Entretanto, a atividade enzimática foi

reduzida gradativamente em 1 hora de incubação a 55°C, chegando a 20% da

atividade inicial no último ponto analisado. A estabilidade da MtbSD também é

constatada a temperatura de -20°C, onde a proteína mantém sua atividade

enzimática por pelo menos 1 ano.

Pela análise do gráfico logk verso 1/T (K

-1

), chegou-se a um valor de

35,2 kJ mol

-1

de energia de ativação (E

) para MtbSD (Figura 5). É importante

ressaltar que a E

calculada pelo gráfico de Arrhenius é um valor aparente e

que o fator pré-exponencial (A) foi considerado como independente da

temperatura no intervalo de temperatura utilizado nos experimentos. Logo,

esse valor para a E

pode ser considerado a quantidade mínima de energia

necessária para iniciar a reação química catalisada pela MtbSD em

concentrações de saturação de SHK e NADP

testadas. Outra observação feita

foi que o gráfico se apresenta linear, indicando que a etapa limitante de

velocidade não altera quando a temperatura é mudada. É importante ressaltar

que a veracidade dos dados sobre a E

da MtbSD é constatada pela

inexistência de queda repentina do gráfico que indicaria uma possível

desnaturação da enzima, consistente com os dados de estabilidade da

proteína.

Análise da Seqüência Proteica da MtbSD

Através de um alinhamento múltiplo de seqüências, a seqüência de

resíduos de aminoácidos da MtbSD foi comparada com outras três seqüências

de SDs, a de E. coli (MICHEL e col., 2003), a de H. influenzae (YE e col., 2003)

e M. jannaschii (PADYANA & BURLEY, 2003), cujas estruturas tridimensionais

já foram determinadas por cristalografia (Figura 6).

A maior identidade encontrada foi de 24% entre as seqüências de

MtbSD e EcoliSD, e entre MtbSD e MjanSD. Apesar desse valor baixo,

considerado o limite para realizar estudos de modelagem molecular, vários

resíduos conservados da família das SDs foram encontrados nas seqüências

analisadas.

O sítio de ligação do substrato na EcoliSD, localizado no domínio N-

terminal, foi identificado pelo posicionamento do anel nicotinamida do cofator

NADPH na estrutura resolvida. Este sítio está em um sulco encontrada na

estrutura da EcoliSD, onde a maior parte dos resíduos altamente conservados

da família das SDs estão presentes, como por exemplo, Ser14, Ser16, Lys65,

Asn86, Thr101, Asp102 e Gln244 (numeração pela seqüência de EcoliSD)

(MICHEL e col., 2003). Esses resíduos estão conservados na seqüência

polipeptídica de MtbSD, correspondendo a Ser18, Ser20, Lys69, Asn90,

Thr104, Asp105 e Gln243 (numeração pela seqüência de MtbSD).

O domínio N-terminal (domínio I) da estrutura da MjanSD é responsável

pela ligação do substrato DHS e os resíduos presentes neste sítio,

aparentemente envolvidos nessa ligação, Lys70, Asn91 e Asp106 (numeração

pela seqüência de MjanSD), são resíduos polares invariáveis em SD. Outra

revelação feita pela estrutura da MjanSD, foi o resíduo Gln254 (numeração

pela seqüência de MjanSD), que também aparece fazendo parte da estrutura

deste sítio e está conservado na família das SDs. Foi proposto que esses

resíduos estão envolvidos com a redução catalítica de DHS a SHK pela

MjanSD (PADYANA & BURLEY, 2003). Pelo alinhamento estes resíduos na

seqüência da MtbSD correspondem a Lys69, Asn90, Asp105 e Gln244

(numeração pela seqüência MtbSD).

Pelo estudo estrutural da HinflSD, foi sugerido que os aminoácidos

Ser14, Ser16, Lys65, Asn85, Asp102 e Gln245 (numeração pela seqüência de

HinflSD) estão envolvidos no sítio de ligação do DHS (YE e cols., 2003). Pelo

alinhamento múltiplo esses resíduos correspondem aos resíduos mencionados

anteriormente. Na enzima SD-like de H. influenzae foi proposto que os resíduos

conservados Lys67, Asn88, Asp103 e Gln242 estão envolvidos tanto na

catálise quanto na ligação do substrato (SINGH e cols., 2005). Todos os

resíduos mencionados nos trabalhos da HinfluSD e da SD-like de H. influenzae

estão conservados na seqüência de MtbSD e estão presentes nas demais

seqüências analisadas.

Este primeiro trabalho descreve o primeiro protocolo de purificação para

a enzima MtbSD, que é capaz de obter quantidade suficiente de enzima pura

para fazer experimentos de cinética e de cristalografia. Além disso, apresenta

uma caracterização detalhada informando o estado oligomérico e as

constantes cinéticas aparentes para ambas reações, direta e reversa, entre

outras. As constantes cinéticas da reação direta foram as primeiras descritas

na literatura para as SDs de procariotos. Como perspectiva tem-se a

determinação do mecanismo cinético em estado estacionário e pré-

estacionário, os efeitos isotópicos, a mutagênese sítio direcionada e

“salvamento químico” que permitirão a elucidação dos mecanismos cinético e

químico. A estrutura e o conhecimento do modo de ação da MtbSD serão

utilizados como “conhecimento base” para o desenho de inibidores efetivos da

via do chiquimato, pela inibição específica da SD. Esses inibidores serão

potenciais agentes anti-TB.

ARTIGO 2: “Shikimate Dehydrogenase from Mycobacterium

tuberculosis H37Rv: Kinetic and Chemical Mechanisms”

Obs.: As figuras e tabelas em negrito, mencionadas na discussão a seguir,

estão identificadas de acordo com o artigo a que se referem.

Parâmetros de Velocidade Inicial e Inibição pelos Produtos

O primeiro passo para esclarecer o mecanismo de ação de uma enzima

é estabelecer seu mecanismo cinético, determinando assim a ordem de ligação

dos substratos ao sítio ativo da enzima e posteriormente de liberação dos

produtos. O inicio dessa caracterização foi feita pela determinação das

velocidades iniciais e posteriormente plotando esses valores no gráfico de

Lineweaver-Burk para visualização do perfil das famílias de retas alcançadas.

Para determinar os parâmetros cinéticos em estado estacionário e os

padrões de velocidade inicial, a atividade da MtbSD foi determinada variando a

concentração de um dos substratos em algumas concentrações fixas do outro

substrato e vice-versa.

Esses dados foram plotados no gráfico de Lineweaver-Burk, e as

famílias de retas observadas na reação direta para os dois substratos são

consistentes com o mecanismo seqüencial e como conseqüência ocorrendo a

formação de um complexo ternário no decorrer da reação. Com isso foi

descartada a possibilidade do mecanismo ping-pong para a MtbSD, que

apresenta no plote de Lineweaver-Burk uma família de retas paralelas. Nos

ensaios onde DHS foi variado em sete concentrações fixas de NADPH (10, 25,

50, 100 e 200

M), as retas se interceptaram sobre o eixo x e a esquerda do

eixo y (Figura 2A). Quando o substrato NADPH foi variado, seguindo as

mesmas condições de ensaio para DHS, as retas se interceptaram abaixo do

eixo x e a esquerda do eixo y (Figura 2B). Com estes gráficos foi descartado

também o mecanismo ordenado em equilíbrio rápido, pois nesse mecanismo

se espera uma família de retas que se interceptam no eixo y.

Os dados coletados foram ajustados na equação geral para mecanismo

seqüencial v = VAB / (K

B + K

A + K

+ AB), onde v é a velocidade medida

da reação, V é a velocidade máxima, A e B são as concentrações dos

substratos, K

e K

são as constantes de Michaelis-Menten dos substratos A e

B e K

é a constante de dissociação para o substrato A. Os valores para as

constantes de Michaelis-Menten dos substratos DHS e NADPH é de 44 ± 2,8

M e 34 ± 2,2

M, respectivamente. A constante catalítica (k

cat

) foi de 77,8 ±

2,3 s

-1

alcançando então as constantes de especificidade de 1,8 x 10

-1

cat

DHS

) e de 2,3 x 10

-1

cat

NADPH

). Esses valores são maiores que os

parâmetros cinéticos aparentes apresentados no artigo “Functional shikimate

dehydrogenase from Mycobacterium tuberculosis H37Rv: Purification and

characterization”, onde os dados foram coletados em tampão Tris HCl 100 mM

pH 9,0. Foi verificado um aumento significativo na velocidade catalítica da

MtbSD quando no ensaio de reação utilizou-se tampão fosfato de potássio 100

mM pH 7,3 ao invés de tampão Tris HCl 100 mM pH 9,0. Os parâmetros

cinéticos em estado estacionário para a PsatSD, em tampão fosfato de

potássio 100 mM pH 7,3, são: K

DHS

340 ± 40

M e K

NADPH

4.3 ± 0.5

(BALINSKY e col., 1971). Observando esses valores, se constata que K

DHS

para PsatSD é 8 vezes maior e que K

NADPH

é 8 vezes menor em relação aos

valores determinados para MtbSD.

O estudo de inibição pelos produtos foi o segundo passo para a

determinação do mecanismo cinético da MtbSD. Os dados coletados foram

ajustados nas seguintes equações: v = VA / [K

(1 + I/K

) + A] ou v = VA / [K

+ I/K

) + A (1 + I/K

)], onde v é a velocidade medida da reação, V é a

velocidade máxima, A é a concentração do substrato, K

é a constante de

Michaelis-Menten do substrato A, K

é a constante de inibição angular, K

é a

constante de inibição linear e I é a concentração do inibidor.

O substrato NADP

mostrou-se um inibidor competitivo para NADPH e

não competitivo para DHS. O produto SHK é um inibidor competitivo e não

competitivo para os substratos DHS e NADPH, respectivamente (Tabela 1).

Esse perfil de inibição pelos produtos é compatível com os mecanismos

cinéticos ordenado em estado-estacionário e aleatório em rápido equilíbrio.

Dados similares para a inibição pelos produtos foram descritos para a PsatSD

(BALINSKY e col., 1971).

Esses dados de inibição pelos produtos indicam que no decorrer da

reação dois possíveis complexos não produtivos podem ser formados, o

complexo MtbSD-NADPH-SHK e o complexo MtbSD-NADP

-DHS, caso o

mecanismo cinético tratar-se de aleatório em rápido equilíbrio.

Efeitos Cinéticos Isotópicos Primário do Deutério

A diferenciação entre os dois possíveis mecanismos para a MtbSD foi

feita através de experimentos de efeitos cinéticos isotópicos primário. Para isso

foram determinados os parâmetros cinéticos em velocidade inicial variando um

dos substratos em cinco concentrações fixas do outro substrato, utilizando

NADPH ou

-nicotinamida adenina dinucleotídeo fosfato deuterado ([4S-

H]NADPH /NADPD). Os efeitos isotópicos cinéticos primário foram observados

sobre o V/K e o V em relação à reação não marcada.

O efeito isotópico sobre V/K descreve o efeito a partir da ligação do

substrato marcado até a primeira etapa irreversível da reação, que geralmente

está representada pela liberação do primeiro produto. O efeito isotópico sobre

V exibe o efeito nos eventos seguintes após a formação do complexo ternário

produtivo, incluindo as etapas químicas, possíveis isomerizações da enzima

até a liberação dos produtos (NORTHROP, 1975).

Os valores coletados foram ajustados nas seguintes equações: v = VA /

[K (1 + FiE

V/K

) + A (1 + FiE

) ou v = VA / [K + A (1 + FiE

)], as quais descrevem

o efeito sobre o V e V/K, e apenas sobre V, respectivamente. Os termos Fi

representa a fração isotópica marcada, Evk o efeito isotópico sobre V/K menos

1 e Ev o efeito isotópico sobre V menos 1 (Tabela 2).

O valor do efeito isotópico cinético primário de 1,8 para

DHS

e de 1,5

para

NADPH

utilizando [4S-

H]NADPH como redutor, indica que a etapa de

transferência do hidreto, que compreende parte da etapa química é

parcialmente limitante para a velocidade catalítica da MtbSD. Isso sugere que

outras etapas insensíveis ao efeito isotópico, tais como isomerização da

enzima e/ou liberação dos produtos, estão envolvidas com as etapas limitantes

de velocidade da reação catalisada pela MtbSD. Os valores baixos para

V/K

NADPH

V/K

DHS

indicam que a etapa de ligação dos substratos até a

formação do complexo irreversível tem uma pequena contribuição na

velocidade de catalise pela MtbSD.

A magnitude dos efeitos isotópicos cinéticos utilizando o [4S-

H]NADPH

indica que o hidrogênio proS do C

(lado B) do cofator é transferido na forma

de hidreto para o substrato DHS na reação de catálise pela MtbSD. Estudo

similar feito com a EcoliSD demonstrou que o hidreto transferido na mesma

reação de oxiredução é o proR do C

(lado A) do NADPH (DANSETTE &

AZERAD, 1974), dado que confere com o estudo da estrutura 3D da proteína

EcoliSD (MICHEL e col., 2003).

A dedução de mecanismo cinético utilizando os efeitos isotópicos para

enzimas com multireagentes foi desenvolvida por Cook & Cleland (1981),

utilizando os valores dos efeitos isotópicos cinéticos aparentes sobre V/K para

um substrato em várias concentrações do cosubstrato.

Os valores do

(V/K

app

)

NADPH

(1,4 ± 0,3) são independentes da

concentração de DHS, entretanto os valores de

(V/K

app

)

DHS

diminuem até o

valor de 1,0 ± 0,03 quando a concentração de NADPH ou NADPD é

aumentada (Figura 3). Segundo as deduções mecanísticas de Cook & Cleland

quando o efeito isotópico primário

(V/K

app

)

é independente da concentração

de A e o efeito isotópico primário

(V/K

app

)

diminui com o aumento da

concentração de B, chegando a 1,0 em concentração infinita de B, o

mecanismo cinético da reação em estudo é ordenado em estado estacionário.

A e B refere-se ao primeiro e ao segundo substrato, respectivamente, que se

liga a enzima. Frente a esses dados chega-se a conclusão de que o

mecanismo cinético para a MtbSD é o mecanismo ordenado em estado

estacionário, e que a ordem de ligação ao sítio catalítico da enzima é DHS

primeiro, seguida da ligação do cofator NADPH (Figura 10).

Utilizando os dados dos parâmetros de velocidade inicial, de inibição

pelos produtos, da análise da relação de Haldane e dos estudos isotope-

exchange, o mecanismo cinético ordenado com NADPH ligando primeiro foi

proposto para a PsatSD (BALINSKY e col., 1971). A enzima álcool

desidrogenase NAD-dependente (classe II) de camundongo, estruturalmente

homóloga ao domínio de ligação do cofator da EcoliSD (MICHEL e col., 2003),

apresenta o mecanismo seqüencial ordenado bi-bi com o cofator NADPH

ligando antes do outro substrato (Strömberg e col., 2004).

Efeito Isotópico Cinético do Solvente e “Proton Inventory”.

Para analisar a contribuição da etapa de transferência de próton pelo

solvente na velocidade da reação enzimática e para determinar o número de

prótons que são transferidos pelo solvente foram feitos experimentos de efeito

isotópico cinético do solvente e “próton inventory”, respectivamente.

O efeito isotópico cinético do solvente foi determinado medindo as

velocidades iniciais usando concentrações saturantes de um substrato variando

a concentração do outro em H

O e em 90% de D

O (Figura 4A e 4B). Os

valores coletados foram ajustados nas equações utilizadas para determinação

dos efeitos isotópicos primário. Os efeitos cinéticos isotópicos do solvente

observados sobre o V/K e o V estão mostrados na Tabela 2.

Os valores dos efeitos isotópicos do solvente sobre V (

V) foram

pequenos, atingindo para DHS a magnitude de 1,5 ± 0,3 e para NADPH de 1,3

± 0,2. O efeito sobre V/K (

V/K) não foi observado. Esses dados sugerem

que a transferência de próton pelo solvente não se trata da etapa limitante de

velocidade, e sim um contribuinte parcial para a velocidade da reação

catalisada pela MtbSD.

“Proton inventory” foi determinado utilizando a relação entre a velocidade

inicial em concentrações saturantes de NADPH e DHS e várias concentrações

de D

O (20, 40, 60, 80 e 90%). Esta relação foi linear, sugerindo que ocorra a

transferência de um único próton pelo solvente na reação de oxiredução

catalisada pela MtbSD (QUINN & SUTTON, 1991). Um resultado similar foi

observado para a enzima micotiona redutase dependente de NADPH de M.

tuberculosis (PATEL & BLANCHARD, 2001) (Figura inserida 4A)

Efeitos Isotópicos Cinéticos Múltiplos

A fim de caracterizar o mecanismo químico da reação catalisada pela

MtbSD, foram feitos os efeitos isotópicos cinéticos múltiplos através da

determinação dos efeitos isotópicos cinéticos do solvente utilizando NADPD

como substrato variado. Deste modo distingui-se entre o mecanismo químico

concertado, onde a transferência do próton e do hidreto ocorrem na mesma

etapa e o mecanismo químico stepwise, onde a transferência do próton e do

hidreto ocorrem em etapas distintas.

A teoria diz que, se caso a protonação e a transferência de hidreto

ocorrerem na mesma etapa de transição, o efeito isotópico primário será maior

ou igual com o uso de D

O como solvente quando comparado com a H

O. Por

outro lado, se caso a transferência de hidreto e a protonação ocorrerem em

etapas distintas, o efeito isotópico primário será menor com o uso de D

O como

solvente, visto que a transferência de próton se tornará limitante na velocidade

da reação (HERMES e col., 1984 e BELASCO e col., 1983).

Os resultados dos efeitos isotópicos cinéticos múltiplos estão mostrados

na tabela 2. O aumento nos valores de todos os efeitos isotópicos primários

medidos em D

O indicam o mecanismo químico concertado para a MtbSD

(Figura 5A e 5B). Um dado importante é o efeito isotópico sobre

V/K

DHS

, que

não era observado pelo efeito isotópico cinético primário e passou para 1,3 ±

0,1 quando foi determinado em D

O como solvente.

O mecanismo concertado é uma estratégia comum utilizada pelas

enzimas para evitar a formação de intermediários instáveis (COOK &

CLELAND, 1981 e THIBBLIN & JENCKS, 1979). A

-eliminação catalisada pela

crotonase é do tipo concertado (GERLT, 1998), bem como as reações

catalisadas pela acil-CoA desidrogenase (BAHNSON & ANDERSON, 1991) e a

isocitrato desidrogenase (POHL e col., 1986) de fígado suíno.

Estudos de pH

O papel da química ácido/base no mecanismo da MtbSD foi determinado

através das constantes k

cat

e k

cat

em diferentes pHs (5,5 – 10,0).

Previamente foi verificada a estabilidade da enzima MtbSD em cada um dos

pHs testados, para a certificação da veracidade dos dados coletados. Em cada

pH testado, as velocidades iniciais foram medidas variando a concentração de

um dos substratos mantendo o outro em concentração saturante. Os dados

coletados foram ajustados na equação log y = log [C/(1 + K

/H)] onde y é o

parâmetro cinético aparente, C é valor de y no platô independente do pH, H é a

concentração de H

e K

é a constante de dissociação aparente para os grupos

ionizados.

Os efeitos do pH na k

cat

para os substratos DHS e NADPH são similares,

diminuindo os valores em pHs elevados com uma inclinação de -1 (Figura 6A).

Isso demonstra que a deprotonação de um único resíduo com pK

aparente de

8,9 ± 0,1 anula a atividade catalítica da MtbSD.

A constante k

cat

para ambos os substratos sofreu um decréscimo em

altos pHs com uma inclinação de -1, sugerindo que um único grupo

deprotonado com valor de pK

aparente

igual a 9,1 ± 0,1, reduz a ligação entre

substrato-enzima (Figura 6B e 6C).

Provavelmente é a ionização do mesmo grupo que esta sendo

observado pelo efeito do aumento de pH sobre as constantes analisadas. O

valor de pK

para as constantes k

cat

e k

cat

encontra-se na faixa normal de

para os grupos

-amino da lisina, tiol da cisteína e hidroxila fenólica da

tirosina. Frente aos resultados de efeitos isotópicos cinéticos múltiplos e

estudos de pH foi proposto um mecanismo químico para a MtbSD mostrado na

figura 11.

A análise da estrutura da EcoliSD propõe que a cadeia lateral do resíduo

Lys65 estabelece uma ponte de hidrogênio ao grupo hidroxila do C-4 do

DHS/SHK (MICHEL e col., 2003). Os autores propuseram ainda que tal resíduo

seja o grupo catalítico ácido/base que doa um próton para o grupo carbonil do

substrato DHS durante a sua redução (MICHEL e col., 2003). Compostos

análogos ao substrato DHS que não continham as hidroxilas no C-4 e C-5

foram utilizadas para demonstrar o papel dessas hidroxilas na especificidade

da enzima EcoliSD pelo substrato. Os resultados mostraram que a hidroxila do

C-4 tem um efeito bastante significativo na especificidade da enzima pelo

substrato (BUGG e col., 1983). A avaliação feita na energia de ligação através

do k

cat

sugere que a hidroxila no C-4 estabelece uma ponte de hidrogênio

com um grupo carregado no sítio ativo da EcoliSD (BUGG e col., 1983). É

interessante ressaltar que a participação da cadeia lateral da lisina na reação

química catalisada pela EcoliSD está baseada em um modelo molecular, e a

estrutura determinada para EcoliSD por cristalografia é o complexo binário

enzima-NADP

e que a transferência do hidreto nessa reação é dada pelo

hidrogênio proR do C-4 do cofator. A cadeia lateral do resíduo Asp102 da

EcoliSD se propõe que também esteja envolvida com a hidroxila no C-4 do

DHS por meio de uma ponte de hidrogênio (MICHEL e col., 2003).

Estudos prévios da PsatSD mostraram que inibidores do tipo substrato-

like necessitam da hidroxila no C-4, ao passo que tanto uma hidroxila quanto

um grupo carboxila no C-5 é necessário para uma ligação forte entre inibidor-

enzima (BALINSKY & DAVIES, 1961). Os efeitos de pH sobre a velocidade da

PsatSD sugerem que uma cisteína ou um grupo

-amino está protonado para

interagir com o grupo hidroxila no C-4 do DHS/SHK (DENNIS & BALINSKY,

1972).

A estrutura da MjanSD foi determinada e os resíduos envolvidos na

ligação e na sua redução catalítica do DHS foram identificados, dentre eles

está a Lys 70 (PADYANA & BURLEY, 2003). A HinflSD tem sua estrutura

resolvida pela análise cristalográfica e estudos de mutagênese sítio direcionada

mostrou que o par Asp103 - Lys67 pode estar envolvido no mecanismo de

catálise desta enzima (SINGH e col., 2005).

Em nossos estudos de efeitos de pH na reação catalisada pela MtbSD

não foi identificado qualquer resíduo cuja protonação anule a atividade

enzimática em pH baixo. Entretanto, os efeitos dos pHs abaixo de 5,5 não

foram testados e com isso não foi permitida a detecção do grupo carboxila

presente na cadeia lateral do aspartato, visto que o valor de pK

para esse

grupo é de aproximadamente 4. Os testes em pHs inferiores a 5,5 foram

evitados pelo fato de que o pK do grupo carboxila do DHS/SHK é ~ 4,1 e isso

dificultaria a interpretação dos resultados.

Anteriormente foi feita uma análise da seqüência de MtbSD através do

alinhamento múltiplo de seqüências utilizando: a EcoliSD, a HinflSD-like e a

MjanSD, cujas estruturas já foram determinadas (FONSECA e col., 2006).

Dentre os resíduos identificados como conservados para a família das SDs

foram identificados os resíduos Lys69 e Asp105 da MtbSD que correspondem

aos resíduos discutidos nas demais estruturas (FONSECA e col., 2006). Os

dados descritos aqui no estudo de pH para MtbSD mostrou a participação de

um grupo cuja deprotonação desfavorece a ligação da MtbSD ao substrato

(pK

= 9.1) e reduz a atividade catalítica (pK

= 8.9), que é consistente com a

participação da Lys69 identificada na seqüência primária da enzima alvo.

Modelagem por Homologia

O primeiro passo para execução da técnica de modelagem por

homologia é buscar um molde para os estudos. Foram encontrados três

candidatos: a EcoliSD de (PDB ID: 1NYT) (MICHEL e col., 2003), a

quinato/chiquimato desidrogenase (QSD) de E. coli (PDB ID: 109B) (MICHEL e

col., 2003) and a MjanSD (PDB ID: 1NVT) (PADYANA & BURLEY, 2003). O

resultado do “BLASTp” identificou a MjanSD (PDB ID: 1NVT) como melhor

modelo para a MtbSD. Entretanto, as coordenadas da estrutura não estavam

disponíveis para o estudo. O segundo candidato pelos dados obtidos no

“Blastp” foi a QSD de E. coli (PDB ID: 109B), entretanto sua função é diferente

da exercida pela MtbSD e além disso a QSD tem a capacidade de utilizar como

cofator tanto o NADH quanto o NADPH. Outro ponto desfavorável em relação

ao molde QSD é a inserção de dois gaps na seqüência da MtbSD observados

no alinhamento múltiplo de seqüências (Figura 7). Esses segmentos que não

apresentam equivalência no molde são regiões difíceis de resolver na estrutura

a ser modelada.

Frente a essas justificativas, a EcoliSD (PDB ID: 1NYT) foi eleita como

molde para a MtbSD, pois apresentam a mesma função biológica, a sua

estrutura foi resolvida em 1,5Å (MICHEL e col., 2003) e a identidade entre as

duas seqüências é de 25%, limite permitido para os estudos de modelagem

estrutural de proteínas sejam feitos (ALTSCHUL e col., 1997).

Partindo desse molde, dez modelos foram construídos para a MtbSD.

Todos os modelos foram avaliados pelos programas PROCHECK

(LASKOWSKI e col. 1993) (Tabela 4) e VERIFY 3D (LUTHY e col., 1992) para

eleger o melhor modelo. Dos 269 resíduos de aminoácidos da MtbSD na

melhor estrutura modelada, 90,5% dos resíduos estão localizados na região

mais favorável no gráfico de Ramachandran (Figura 8 e Tabela 3). O RMSD

da cadeia principal entre as estruturas do molde e da modelada é de 1,93 Å.

Considerando as divergências entre as duas seqüências utilizadas no estudo, o

valor para RMSD é plausível e os resultados gerais da estrutura modelada

mostram que o molde eleito é bastante satisfatório e pode ser utilizado para

analisar a relação estrutural e funcional da MtbSD.

A estrutura da MtbSD pertence a superfamília denominada domínio

“Rossmann-fold“ de ligação a NAD(P) (MURZIN e col., 1995). A estrutura

apresenta uma forma alongada com dois domínios, que apresentam uma

arquitetura α/β (Figura 9). Esses domínios estão ligados por uma α-hélice e

uma volta que os mantém juntos e formando um sulco onde o cofator NADPH

liga-se. O domínio de ligação do cofator NADPH, localizado na porção C-

terminal, apresenta uma estrutura do tipo “Rossmann-fold“, que compreende

em seis fitas

paralelas e seis

hélices, onde cada três

hélices estão

dispostas sobre cada lado do leque central de fitas

. A porção N-terminal é

bastante semelhante ao domínio C-terminal descrito, entretanto ocorre um

arranjo irregular que compreende a orientação antiparalela da fita

-5 em

relação às demais fitas (MICHEL e col., 2003).

Na análise da estrutura da EcoliSD, utilizada como molde, foi identificado

o sítio de ligação dos substratos pela posição do anel nicotinamida do cofator,

e neste sulco os resíduos de aminoácidos, Ser14, Ser16, Lys65, Asn86,

Thr101, Asp102 e Gln244 foram identificados. Ambos os oxigênios carboxil do

substrato DHS formam pontes de hidrogênio com o sítio ativo da proteína,

provavelmente pelos resíduos conservados de serina nas posições 14 e 16

(MICHEL e col., 2003). Tais resíduos foram identificados no modelo para

MtbSD, Ser18, Ser20, Lys70, Asn91, Thr106, Asp107 e Gln263, em posições

estruturalmente equivalentes (Figura 12). Na estrutura da EcoliSD foi

identificado o resíduo Tyr215 que também está envolvido na ligação do

substrato DHS no sítio ativo da enzima através de ponte de hidrogênio. No

modelo da MtbSD o resíduo equivalente é Tyr215 (Figura 12).

No alinhamento múltiplo de seqüências entre o alvo (MtbSD) e os

moldes (EcoliQSD and EcoliSD) foi identificada a seqüência Gly124-Ser125-

Gly126-Gly127-Thr128-Ala129 como a seqüência padrão da MtbSD (G [A,s,g]

G G [A,t] [A,S,g]), correspondendo ao loop de ligação ao difosfato, que está

conservado na família das SDs (MICHEL e col., 2003) (Figura 7).

Para a ligação da nicotinamida na EcoliSD foi verificado que o N-7 do

grupo amida do anel nicotinamida é ligado por pontes de hidrogênio ao grupo

carbonil da Met213 e da Gly237 (MICHEL e col., 2003). A Gly236 (posição

correspondente para a Gly237) é conservada no modelo da MtbSD,

estabelecendo o mesmo tipo de interação com o cofator; em contrapartida na

posição equivalente a Met213 encontra-se a Ala213 que mantém a mesma

natureza de ligação com o N-7 do anel (Figura 7).

As argininas 150 e 154 na EcoliSD têm um papel crucial na ligação do

sítio ativo ao fosfato da adenine pela formação de um “grampo eletrostático”

(electrostatic clamp) que envolve o fosfato (MICHEL e col., 2003). No modelo

da MtbSD são encontrados os resíduos correspondentes Arg149 e Lys153.

Provavelmente no modelo em estudo estes resíduos têm o mesmo papel, visto

que aparecem pontes de hidrogênio entre eles e o fosfato da adenina (Figura

7).

O resíduo de Ser189 na EcoliSD é ligado ao oxigênio do anel da ribose

fosfato do cofator por ponte de hidrogênio. Analisando o modelo MtbSD essa

interação é dada pelo resíduo de Val189. A interação entre o grupo fosfato da

ribose e o sítio ativo está sendo feito por ponte de hidrogênio com Asn149 e

Asn150 na Ecoli SD e na MtbSD, respectivamente (Figura 7).

A interação entre a porção difosfato do cofator e o sítio ativo da EcoliSD

ocorre através da Gly129 e da Ala130, que corresponde a Thr128 e a Ala129

no modelo em estudo, respectivamente.

Michel e col. (2003) mostraram que a ligação do NAD

na EcoliQSD é

favorecida pela substituição dos resíduos Thr151 e Arg154, presentes na

EcoliSD, pelos resíduos Asp158 e Phe160, respectivamente. A explicação para

tal, é o ambiente hidrofóbico criado pelo resíduo apolar Phe160, que reduz a

especificidade pelo NADPH e permitindo assim, que tanto NADH quanto

NADPH sejam utilizados como cofatores pela EcoliQSD. Em posição

equivalente aos resíduos Thr151 e Arg154 na estrutura da EcoliSD estão os

resíduos Asn157 e Lys160 na estrutura modelada da MtbSD. Tanto no molde

quanto na estrutura modelada, esses resíduos estabelecem ponte de

hidrogênio e é essa interação que resulta em uma orientação similar das

cadeias laterais e como conseqüência um ambiente equivalente para a ligação

do NADPH (Figura 13).

O trabalho apresentado é a primeira descrição detalhada sobre os

mecanismos cinético e químico para a MtbSD, propondo um resíduo envolvido

na catálise e na etapa de ligação dos substratos.

Como perspectivas próximas há os estudos de mutagênese sítio

direcionada para confirmar a importância do resíduo Lys69 na ligação dos

substratos e na velocidade de catálise pela MtbSD e os estudos de

fluorescência de proteína que fornecerão as constantes de associação e

dissociação em equilíbrio para ambos substratos. A etapa subseqüente será a

cinética em estado pré-estacionário que possibilitará a determinação de

constantes de velocidade de primeira e segunda ordens o que possibilitará um

melhor detalhando do mecanismo cinético.

Estes estudos juntamente com os estudos estruturais do sítio ativo da

enzima fornecerão a base sólida para o desenho racional de inibidores para a

MtbSD, que poderão representar possíveis agentes quimioterápicos para

tratamento da tuberculose.

Conclusões

 A proteína recombinante purificada é a MtbSD na forma dimérica;

 A reação catalisada é reversível, com uma Keq de 19,6, uma Ea = 35,2

kJ mol-1 e segue a cinética de Michaelis-Menten;

 O mecanismo cinético é ordenado em estado estacionário, com DHS

ligando primeiro à enzima;

 As transferências de hidreto e próton contribuem modestamente para

determinação da velocidade da reação e ocorrem de forma concertada;

 Há catálise ácido/base na reação e a protonação de um grupo é

essencial para a catálise e a ligação dos substratos;

 Todos os resíduos sugeridos importantes para catálise na SD de E. coli

estão conservados e em posições semelhantes no sítio ativo da MtbSD.

Referências

[1]. ALTSCHUL, S.F., MADDEN, T.L., SCHÄFFER, A.A., ZHANG, J.,

ZHANG, Z., MILLER, W., LIPMAN, D.J. (1997) Nucleic Acids Res. 25,

3389-3402

[2]. BALINSKY, D., DAVIES, D. D. (1961) Biochem. J. 80, 296-300

[3]. BALINSKY, D., DENNIS, A.W., CLELAND, W.W. (1971) Biochemistry

10 1947–1952.

[4]. BASSO, L.A., SILVA, L.H.P., FETT-NETO, A.G., AZEVEDO, W.F.,

MOREIRA, I.S., PALMA, M.S., CALIXTO, J.B., ASTOLFI, S., SANTOS,

R.R., SOARES, M.B.P., SANTOS, D.S. (2005) Mem Inst Oswaldo Cruz

100, 475-506.

[5]. BELASCO, J. G., ALBERY, J., AND KNOWLES, J. R. (1983) J. Am.

Chem. Soc. 105, 2475-2477.

[6]. BUGG, T. D., ABELL, C., COGGINS, J. R. (1988) Tetrahedron Lett. 29,

6779-6782

[7]. BENACH, J., LEE, I., EDSTRON, W., KUZIN, A. P., CHIANG, P.,

ACTON, T. B., MONTELIONE, G. T. AND HUNT J. F. (2003) The

Journal of Biological Chemistry 278, 21, 19176-19182.

[8]. BENTLEY, R. (1990) Critical Reviews in Biochemistry and Molecular

Biology 25, 307-384.

[9]. BLOOM, B.R. & MURRAY, C.J.L. (1992) Science 257, 1055-1064.

[10]. BRENNAN, P.J. (1997) FEMS Immunology and Medical Microbiology

18, 263-269.

[11]. CHAUDHURI, S. AND COGGINS J. R. (1985) Biochemical Journal

226, 217-223.

[12]. COLE, S.T., BROSCH, R., PARKHILL, J. et al. (1998) Nature 393,

537–544.

[13]. COOK, P. F., CLELAND, W. W. (1981) Biochemistry 20, 1790-1796

[14]. DANSETTE, P., AZERAD, R. (1974) Biochimie 56, 751-755

[15]. DANIEL, T.M. Captain of death: the story of tuberculosis. 1

ed. New

York. University of Rochester Press, 1997.

[16]. DENNIS, A. W., BALINSKY, D. (1972) Int. J. Biochem. 9, 93-102

[17]. DORMANDY, T. (2002) The white death: a history of tuberculosis, New

York University Press.

[18]. DUNCAN, K. (2003) Tuberculosis 83, 201-207.

[19]. DUNLAP, N. E., BASS, J., FUJIWARA, P., HOPEWELL, P.,

HORSBURGH JR., C. R., SLAFINGER, M. AND SIMONE, P. M. (2000)

American Journal of Respiratory and Critical Care Medicine 161, 1376-

1395.

[20]. ENARSON, D.A. & MURRAY, J.F. (1996) Global epidemiology of

tuberculosis. In: Tuberculosis (Rom, W.M. & Garay, S., Eds.), pp. 57-75.

Little, Brown and Co., Boston, MA.

[21]. ENSERINK, M. (2001) Science 293, 5528, 234-236.

[22]. ESPINAL M. A. (2003) Tuberculosis 83, 44-51.

[23]. FÄTKENHEUER, G., TAELMAN, H., LEPAGE, P., SCHWENK, A. &

WENZEL, R. (1999) Diagn. Microbiol. Infect. Dis. 34, 139-146.

[24]. FONSECA, I. O., MAGALHÃES, M. L. B., OLIVEIRA, J. S., SILVA, R.

G., MENDES, M. A., PALMA, M. S., SANTOS, D. S., BASSO, L. A.

(2006) Protein Expr. Purif. 46, 429-437

[25]. GLICKMAN, M.S. & JACOBS, W.R. Jr. (2001) Cell 104, 477-485.

[26]. HERMES, J. D., ROESKE, C. A., O’LEARY, M. H., AND CLELAND, W.

W. (1984) Biochemistry 21, 5106-5114.

[27]. KNIGGE, A., MORR, H. AND KILLIAN, A. (2000) Tuberculosis and

HIV/AIDS. Epidemics: Opportunities for cross support. Disease Control

(Tuberculosis) and Health Promotion AIDS Control in Developing

Countries. Deutsche Gesellschaft für Technische Zusammenarbeit (GTZ)

GmbH.

[28]. LASKOWSKI, R.A., MACARTHUR, M.W., MOSS, D.S., THORNTON,

J.M. (1993) J. Appl. Cryst. 26, 283-291

[29]. LÜTHY, R., BOWIE, J.U., EISENBERG, D. (1992) Nature 356, 83-85

[30]. MURZIN A. G., BRENNER S. E., HUBBARD T., CHOTHIA C. (1995) J.

Mol. Biol. 247, 536-540

[31]. MAGALHÃES, M. L. B., PEREIRA, C. P., BASSO, L. A. AND SANTOS

D. S. (2002) Prot. Exp. and Purif. 26, 59-64.

[32]. MICHEL, G., ROSZAK, A.W., SAUVE, V., MACLEAN, J. MATTE, A.,

COGGINS, J.R., CYGLER, M., LAPTHORN, A.J. (2003) J. Biol. Chem.

278, 19463-19472.

[33]. MILBURN, H. J. (2001) Current Opinion in Pulmonary Disease 7, 133-

141.

[34]. NORTHROP, D. B. (1975) Biochemistry 14, 2644-2651

[35]. NSB Editorial comment. Taming tuberculosis-again. Nature Structural

Biology 7, 87-88 (2000).

[36]. O´BRIEN, R. J. AND NUNN P. P. (2001) American Journal of

Respiratory and Critical Care Medicine 162, 1055-1058.

[37]. PADYANA, A.K., BURLEY, S.K. (2003) Structure 11, 1005-1013.

[38]. PABLOS-MENDEZ, A., RAVIGLIONE, M.C. et al. (1998) The New

England Journal of Medicine 338, 1641-1649.

[39]. PARISH, T. AND STOKER, N. G. (2002) Microbiology 148, 3069-3077.

[40]. PARRISH, N.M., DICK, J.D. & BISHAI, W.R. (1998) Trends in

Microbiology 6, 107-112.

[41]. PASQUALOTO, K.F.M. & FERREIRA, E.I. (2001) Current Drug Targets

2, 427-437.

[42]. PATEL, M. P., AND BLANCHARD, J. S. (2001) Biochemistry 40, 5119-

5126.

[43]. PETRINI, B. & HOFFNER, S. (1999) International Journal of

Antimicrobial Agents 13, 93-97.

[44]. QUINN, D. M., AND SUTTON, L. D. (1991) Enzyme Mechanism from

Solvent Isotope Effects (Cook, P. F., Ed.) 1st ed., pp 73-126, CRC Press,

Florida.

[45]. RATLEDGE, C. (1982) Nutrition, Growth and Metabolism in “The

Biology of the Mycobacteria” (Ratledge, C. and Stanford, J. L. Ed.) vol. 1,

pp. 185-271, Academic Press, London.

[46]. RILEY, L.W. (1993) Clinical Infectious Diseases 17, 442-446.

[47]. ROBERTS, F., ROBERTS, C. W., JOHNSON, J. J., KYLE, D. E.,

KRELL, T., COGGINS, J. R., COOMBS, G. H., MILHOUS, W. K.

TZIPORI, S., FERGUSON, D. J. P., CHAKRABARTI, D. AND MCLEOD,

R. (1998) Nature 393, 801-805.

[48]. RUFFINO-NETTO, A. (2002) Rev. Soc. Bras. Med. Trop. 35, 51-58.

[49]. STRÖMBERG, P., SVENSSON, S., BERST, K. B., PLAPP, B. V.,

HÖÖG, J. O. (2004) Biochemistry 43, 1323-1328

[50]. SINGH, S., KOROLEV, S., KOROLEVA, O., ZAREMBINSKI, T.,

COLLART, F., JOACHIMIAK, A., CHRISTENDAT, D. (2005) J. Biol.

Chem. 280, 17101-17108

[51]. SEPKOWITZ, K. A., RAFFALI, J., RILEY, L., KIEHN, T. E. AND

ARMSTRONG, D. (1995) Clinical Microbiology Reviews 180-199.

[52]. ZHANG, X., ZHANG, S., HAO, F., LAI, X., YU, H., HUANG, Y. AND

WANG, H. (2005) J Biochem Mol Biol. 38, 624-31.

[53]. Ye, S., Delft, F.V., Brooun, A., Knuth, M.W., Swanson, R.V., McRee,

D.E. (2003) J. Bacteriol. 185, 4144–4151.

[54]. YOUNG, D.B. (1998) Nature 393, 515-516.

[55]. WICKELGREN, I. (2000) Science 288, 5470, 1314-1315.

[56]. WOESTENENK, E.A., HAMMARSTRO, M., BERG, S., HÄRD, T.,

BERGLUND, H. J. (2004) Struct. Funct. Genomics 5, 217–229.

ANEXO I

Docking and small angle X-ray scattering studies

of purine nucleoside phosphorylase

Walter Filgueira de Azevedo Jr.,

a,b,

Giovanni C



eesar dos Santos,

Denis Marangoni dos Santos,

Johnny Rizzieri Olivieri,

a,b

Fernanda Canduri,

a,b

Rafael Guimar

aaes Silva,

Luiz Augusto Basso,

Gaby Renard,

Isabel Os



oorio da Fonseca,

Maria Anita Mendes,

b,d



aario S



eergio Palma,

b,d

and Di



oogenes Santiago Santos

Departamento de F



ıısica, UNESP, S

aao Jos



ee do Rio Preto, SP 15054-000, Brazil

Center for Applied Toxinology, Instituto Butantan, S~aao Paulo, SP 05503-900, Brazil

Rede Brasileira de Pesquisas em Tuberculose, Departamento de Biologia Molecular e Biotecnologia, UFRGS, Porto Alegre, RS 91501-970, Brazil

Laboratory of Structural Biology and Zoochemistry, Department of Biology, Institute of Biosciences, UNESP, Rio Claro, SP 13506-900, Brazil

Received 19 August 2003

Abstract

Docking simulations have been used to assess protein complexes with some success. Small angle X-ray scattering (SAXS) is a

well-established technique to investigate protein spatial conﬁguration. This work describes the integration of geometric docking with

SAXS to investigate the quaternary structure of recombinant human purine nucleoside phosphorylase (PNP). This enzyme catalyzes

the reversible phosphorolysis of N-ribosidic bonds of purine nucleosides and deoxynucleosides. A genetic deﬁciency due to mu-

tations in the gene encoding for PNP causes gradual decrease in T-cell immunity. Inappropriate activation of T-cells has been

implicated in several clinically relevant human conditions such as transplant rejection, rheumatoid arthritis, lupus, and T-cell

lymphomas. PNP is therefore a target for inhibitor development aiming at T-cell immune response modulation and has been

submitted to extensive structure-based drug design. The present analysis conﬁrms the trimeric structure observed in the crystal. The

potential application of the present procedure to other systems is discussed.

Keywords: Geometric docking; SAXS; Purine nucleoside phosphorylase; Bioinformatics

Recent developments in the algorithm for protein

docking allowed the prediction of the conformation of

quaternary structures of several proteins. Among all

available algorithms for docking of biological macro-

molecules the geometric docking has been proved to

generate reasonable models of several macromolecular

assemblies [1,2]. Small angle X-ray scattering (SAXS)

technique provides information on the structural char-

acteristics of macromolecules in solution at a super-

atomic scale. One of the procedures to obtain structural

information from SAXS results is based on the

comparison between structure functions of proposed

models with diﬀerent conﬁ gurations of monomers or

subdomains with those determined from experiments.

Even though this technique cannot guarantee the

uniqueness of the model, it is widely used and was

demonstrated to yield useful information on the struc-

ture, on structural variations, and on the quaternary

structure of a number of macromolecules of biological

interest [3]. The Guinier analysis of SAXS intensity

provides a structural parameter, the radius of gyration

of the macromolecule in solution, which is independent

of any a priori mod el. The SAXS method also yields

information on the spatial conﬁguration of the macro-

molecular subdomains but ignores inter nal structural

details and dynamics features such as, vibration, rota-

tion or internal conformational changes [4].

Purine nucleoside phosphorylase (PNP) catalyzes the

reversible phosphorolysis of the ribonucleosides and

Biochemical and Biophysical Research Communications 309 (2003) 923–928

www.elsevier.com/locate/ybbrc

BBRC

Corresponding author. Fax: +55-17-221-2247.

E-mail address: [email protected] (W.F. de Azevedo Jr.).

doi:10.1016/j.bbrc.2003.08.093

-deoxyribonucleosides of guanine, hypoxanthine, and

a number of related nucleoside compounds [5], except

adenosine. Human PNP is an attractive target for drug

design and it has been submitted to extensive structure-

based design. PNP inhibitors could be used in the fol-

lowing applications: (1) treatment of T-cell leukemia;

(2) suppression of the host versus graft response in

organ transplantation recipients; (3) treatment of sec-

ondary or xanthine gout by restricting purine catabo-

lites to the more soluble nucleosides; and (4) in

combination with nucleosides to prevent their degra-

dation by PNP metabolism [6]. More recently, the

structure of human PNP has been solved using cryo-

crystallographic techniques at 2.3



AA resolution, which

allowed a redeﬁnition of the residues involved in the

substrate binding sites [7,8]. The crystallographic

structure is a trimer, however, there is a report of di-

meric structure for the human enzyme [9], which may

change the subunit interface. Since the active site is

located near the interface of two subunits, changing the

putative interactions between enzyme and inhibitors

should have a bearing on structure-based inhibitor

design.

Here we report the combination of geometric docking

simulations and SAXS studies to assess the human PNP

quaternary structure in solution. The general procedure,

here described, may be used to study the spatial con-

ﬁguration of the macromolecular subdomains of other

proteins in solution.

Materials and methods

Integration of geometric docking simulation and SAXS experiments.

In order to assess the quaternary structure of PNP a scheme was used

that involved both geometric docking simulations and SAXS experi-

ments. A ﬂowchart describing the overall strategy is shown in Fig. 1.

This procedure was used to generate the dimeric models for PNP and

the trimeric structure was built using the crystallographic symmetry. In

order to speed up the geometric docking simulations, a parallel version

of the program GRAMM [10] was used to generate the dimeric models

for human PNP. Each step of the procedure is described in the

following sections.

Geometric docking simulations. In order to obtain the dimeric

structure of human PNP, docking simulation was performed using the

geometric recognition algorithm, which was developed to identify

molecular surface complementarity. The monomeric structure of hu-

man PNP (PDB access code: 1M73) [7] was docked against its own

structure. It generated a total of 100 dimers. The geometric recognition

algorithm is based on a geometrical approach and involves an auto-

mated procedure including: (i) a digital representation of the molecules

by three-dimensional discrete functions; (ii) the calculation of a cor-

relation function that assesses the degree of molecular surface overlap

and penetration upon relative shifts of the molecules in three dimen-

sions; and (iii) a scan of the relative orientations of the molecules [10].

The procedure is equivalent to a six-dimensional search but consid-

erably faster by design, and the computation time is only moderately

dependent on molecular size. This procedure has been applied to assess

protein–protein and protein–ligand interactions. The geometric rec-

ognition algorithm was implemented in the program GRAMM [10].

All geometric docking simulations were performed on a Beowulf

cluster, with 16 nodes (B16/AMD Athlon 1800+; BioComp, S

aao Jos



do Rio Preto, SP, Brazil).

SAXS studies. X-ray scattering data were collected at room tem-

perature using Cu Ka X-rays radiation generated by a Rigaku RU300

rotating anode generator operated at 50 kV and 90 mA and collimated

with a block slit system [11]. The scattering intensity was measured

using a linear position sensitive detector (CBPF-Brazil).

The SAXS measurements were performed within an angular range

deﬁned by 0:02



À1

< h < 0:450



À1

where h ¼ð4psinÞ=k,2h being the

angle between the incident and the scattered X-ray beam and k the

X-ray wavelength. The contributions to the scattering intensity from

the solvent, capillary, and air were subtracted from the total intensity.

Recombinant human PNP was expressed and puriﬁed as previously

described [12]. The SAXS measurements were carried out using human

PNP solution, which was concentrated to 12 mg ml

À1

against 10 mM

potassium phosphate buﬀer (pH 7.1). The counting time was 12 h. The

extrapolated experimental SAXS intensity function was desmeared to

suppress the inﬂuence from the slit collimation system yielding the

corrected intensities, IðhÞ.

A structural parameter related to the overall size of the macro-

molecule, the radius of gyration R

, was determined by using the

Guinier equation [13]

IðhÞ¼Ið0Þ exp

: ð1Þ

Eq. (1) applies to macromolecules in the limits of a dilute solution and

small h values. More detailed information of the molecular structure

can be obtained from the distance distribution function pðrÞ, which is

related to the SAXS desmeared (free from geometrical collimation

eﬀects) intensity IðhÞ by

Fig. 1. Flowchart describing the overall strategy to assess protein

complex conformations.

924 W.F. de Azevedo Jr. et al. / Biochemical and Biophysical Research Communications 309 (2003) 923–928

pðrÞ¼

IðhÞðhrÞ sinðhrÞdh: ð2Þ

The pðrÞ function is proportional to the number of pairs of electrons

separated by the distance r, which is encountered by combinations

between all the elements of the macromolecule. The radius of gyration

of macromolecules in solution is usually determined by applying Eq.

(1). The distance distribution function, p

exp

ðrÞ, has been determined by

indirect Fourier transformation using the ITP program [11]. This

program was also used to determine the intensity IðhÞ, free from

smearing collimation eﬀects. The theoretical function p

theo

ðrÞ was cal-

culated using the program MULTIBODY [11], modiﬁed in order to

make molecular model building easier [4]. The program MULTI-

BODY calculates the resulting function pðrÞ of the complete set of

atomic coordinates of each structural model for the macromolecule. In

the present study we calculated the p

theo

ðrÞ for monomer, for the

dimers, generated by geometric docking, and for the trimer, obtained

by application of crystallographic rotations.

Correlation between geometric docking simulations and SAXS

experiments. To assess the correlation between theoretical and exper-

imental distance distribution function p

exp

ðrÞ and p

theo

ðrÞ, respectively,

we have calculated the linear correlation coeﬃcient (CC), which is

deﬁned as follows [14]:

CC ¼

exp;i

ðrÞj

À jp

exp

ðrÞj



Âjp

theo;i

ðrÞj

À jp

theo

ðrÞj

hi

exp;i

ðrÞj

À jp

exp

ðrÞj



theo;i

ðrÞj

À jp

theo

ðrÞj





;

ð3Þ

where

exp

ðrÞj

is the mean of the jp

exp;i

ðrÞj

, jp

theo

ðrÞj

is the mean of

theo;i

ðrÞj

, and sums are made over all available pðr Þ. When a corre-

lation is known to be signiﬁcant, CC is one conventional way of

summarizing its strength. The complex, which generates the highest

correlation coeﬃcient, is considered the right macromolecular con-

formation. In the present work, we calculated the CC for all dimeric

models obtained from the geometric docking simulations and for the

monomeric and trimeric structures.

Results and discussion

Guinier plot (log IðhÞ versus h

) of the desmeared

scattering function is displayed in Fig. 2. The slope of

linear portion of this plot was determined to obtain the

radius of gyration of human PNP. The R

value was

29.8



AA.

PNPs from most mammalian and some of the bac-

terial sources appear to be trimeri c although dimeric

quaternary structures have been proposed for the

human enzyme [9]. Analysis of the crystallographic

structures of human PNP indicates a trimeric structure

(PDB access codes: 1ULA, 1ULB, 1M73, and 1PWY)

[6–8,15]. However, in a number of instances the qua-

ternary structure observed in the crystalline state is not

conserved in solution [3]. Furthermore, in the case of

human PNP the low pH used in the crystallization

condition [16] may generate diﬀerences in the spatial

conﬁguration of the macromolecular subdomains ob-

served in the crystal when compared to the structure in

physiological pH. In addition, since the active site of the

PNP is located near the interface of two subunits within

the trimer, the precise information about the biological

unit in solution is of capital importance to guide the

structure-based design of inhibitors because its target is

a structure as close as possible to the structure found in

the physiological conditions, where the drug will act. Up

to now all structure-based designs of PNP inhibitors

have used the low-resolution structures of human PNP

(PDB access codes: 1ULA and 1ULB) and consider the

trimer as the target for molecular modeling studies.

Three families of theoretical models, based on the

high-resolution crystallographic structure (PDB access

code: 1M73) [7], were used to determine the theoretical

distance distribution function, p

theo

ðrÞ, using the pro-

gram MULTIBODY and then compared with the ex-

perimental function pðrÞ determined using the ITP

program from SAXS data. Figs. 3A–C show structural

models and the experimental distribution function

against the theoretical distribution function for

the monomer, dimer, and trimer, of human PNP,

respectively.

The atomic coordinates for monomeric structure

were obtained from the asymmetric unit content of the

crystallographic structure of human PNP solved at 2.3



resolution [7]. Previous statistical analysis of low-reso-

lution docking indicated that gross structural features of

protein–protein interactions could be identiﬁed for a

signiﬁcant percentage of protein complexes [1]. There-

fore, the low-resolution protocol of the GRAMM pro-

gram [10] was used to generate the dimeric models. A

total of 100 models for the dimeric structure were built,

only the complex, which generated the highest correla-

tion coeﬃcient between theoretical and experimental

distance distribution function is shown in Fig. 3B. The

trimeric structure was built applying two successive

rotations of 120° along z-axis on the atomic coordinat es

of the monomer. The radii of gyration for the structural

models are 18.7, 26.5, and 28.7



AA, for the monomer,

dimer, and trimer, respectively.

Fig. 2. Guinier plot of the SAXS intensities, IðhÞ, for human PNP. The

straight line was obtained by least-squares ﬁtting in the region

< 1:8:10

À3



W.F. de Azevedo Jr. et al. / Biochemical and Biophysical Research Communications 309 (2003) 923–928 925

The value CC lies betw een )1 and 1. It has a value of

1, when the data points lie on a perfect straight line with

positive slope. If the data points lie on a perfect straight

line with negative slope, then CC has the value )1 [14].

The correlation coeﬃcient between theoretical and ex-

perimental distance distribution functions ranges from

0.591 to 0.995, an d the highest correlati on coeﬃcient

was obtained for the trimeric structure, which also

presented the radius of gyration closer to the experi-

mental radius of gyration.

The contact area at interface between each subunit in

the PNP trimer is 1124



, which indicates that the

subunits are strongly bound to each other. Fig. 4 shows

the electrostatic potential surface at subunit interface of

the trimeric structure generat ed using GRASP [17].

Analysis of the electrostatic potential surface at the

Fig. 3. Proposed structural models for human PNP and the corresponding theoretical (continuous line) and experimental (dotted line) distance

distribution functions, pðrÞ, for: (A) monomer, (B) dimer, and (C) trimer. The model ﬁgures were generated by MOLSCRIPT [27] and Raster3D [28].

926 W.F. de Azevedo Jr. et al. / Biochemical and Biophysical Research Communications 309 (2003) 923–928

subunit interface indicates good shape complementarity

and some charge complementarity; however, most of the

contacts are hydrophobic and involve residues Tyr88,

Phe141, Phe159, Phe200, and Leu209.

The trimeric PNP struc ture has been extensively used

for structure-based studies of PNP inhibitors [5,6,18–

25]. However, the quaternary structure of human PNP

in solution and in physiological pH has not been pre-

viously investigated using low-resolution methods, such

as SAXS. The present analysis of the SAX S experiments

integrated with geomet ric docking simulation strongly

indicates that human PNP is a trimer in solution, the

agreement found between the experimental and theo-

retical pðrÞ functions for the trimer suggests that struc-

ture in solution adopts approximately the same

conformation identiﬁed in the high-resolution crystal-

lographic structure (PDB access code: 1M73) [7]. The

radius of gyration determined for the trimeric structure

is slightly smaller than that determined from the Guinier

plot (log IðhÞ versus h

) of the desmeared scattering

function. The possible reasons for this discrepancy may

be the cryogenic conditions used to solve the high-res-

olution structure of human PNP and the absence of

solvents in the theoretical model.

The integration of a high-eﬃcient algorithm for

geometric docking with SAXS experiments allowed the

investigation of the possible quaternary structures not

observed in the crystalline state, such as the putative

PNP dimeric structure [9]. The pr ocedure adopted to

analyze the interaction between PNP subunits can be

used for other protein complexes. The main applications

of the present methodology are: (1) analysis of interac-

tions between biological macromolecules using struc-

tural models obtained from crystallography or NMR,

(2) validation of structural models obtained from

molecular modeling [26] of complexes of biological

macromolecules, and (3) analysis of complexes of

biological macromolecules in conditions closer to the

biological environment.

Geometric docking simulations may be omitted from

the strategy if the atomic coordinates for the complexes

are available. We are applying the procedure, here de-

scribed, to assess the quaternary structure of a number

of protein complexes, such as hemoglobins, PNPs, and

crotoxin.

Acknowledgments

This work was supported by grants from FAPESP (SMOLBNet,

Proc. Num. 01/07532-0), CNPq, CAPES and Instituto do Millenium

(CNPq-MCT). W.F.A. (CNPq, 300851/98-7), M.S.P. (CNPq, 500079/

90-0), and L.A.B. (CNPq, 520182/99-5) are researchers for the Bra-

zilian Council for Scientiﬁc and Technological Development.

References

[1] A. Tovchigrechko, C.A. Wells, I.A. Vakser, Docking of protein

models, Protein Sci. 11 (2002) 1888–1896.

[2] N. Jing, C. Marchand, J. Liu, R. Mitra, M.E. Hogan, Y.

Pommier, Mechanism of inhibition of HIV-1 integrase by G-

tetrad-forming oligonucleotides in vitro, J. Biol. Chem. 275 (2000)

21460–21467.

[3] D.I. Svergun, C. Barberato, M.H.J. Koch, L. Fetler, P. Vachette,

Large diﬀerences are observed between the crystal and solution

quaternary structures of allosteric aspartate transcarbamylase in

the R state, Proteins Struct. Funct. Genet. 27 (1997) 110–117.

[4] J.R. Olivieri, A.F. Craievich, The subdomain structure of human

serum albumin in solution under diﬀerent pH conditions studied

by small angle X-ray scattering, Eur. Biophys. J. 24 (1995) 77–

84.

[5] J.A. Montgomery, Purine nucleoside phosphorylase: a target for

drug design, Med. Res. Rev. 13 (1993) 209–228.

[6] S.E. Ealick, Y.S. Babu, C.E. Bugg, M.D. Erion, W.C. Guida, J.A.

Montgomery, J.A. Secrist III, Application of crystallographic and

modeling methods in the design of purine nucleoside phosphor-

ylase inhibitors, Proc. Natl. Acad. Sci. USA 91 (1991) 11540–

11544.

Fig. 4. Electrostatic potential surface at subunit interface of human PNP, calculated with GRASP [17] and shown from )10 to +10 kT. Uncharged

regions are white.

W.F. de Azevedo Jr. et al. / Biochemical and Biophysical Research Communications 309 (2003) 923–928 927

[7] W.F. de Azevedo Jr., F. Canduri, D.M. Santos, R.G. Silva, J.S.

Oliveira, L.P.S. Carvalho, L.A. Basso, M.A. Mendes, M.S. Palma,

D.S. Santos, Crystal structure of human purine nucleoside

phosphorylase at 2.3



AA resolution, Biochem. Biophys. Res. Com-

mun. 308 (2003) 545–552.

[8] D.M. Santos, F. Canduri, J.H. Pereira, M.V.B. Dias, R.G. Silva,

M.A. Mendes, M.S. Palma, L.A. Basso, W.F. de Azevedo, D.S.

Santos, Crystal structure of human purine nucleoside phosphor-

ylase complexed with acyclovir, Biochem. Biophys. Res. Com-

mun. 308 (2003) 553–559.

[9] A.S. Lewis, B.A. Lowy, Human erythrocytes purine nucleoside

phosphorylase: molecular weight and physical properties, J. Biol.

Chem. 254 (1979) 9927–9932.

[10] E. Katchalski-Katzir, I. Shariv, M. Eisenstein, A.A. Friesem, C.

Aﬂalo, I.A. Vakser, Molecular surface recognition: determination

of geometric ﬁt between proteins and their ligands by correlation

techniques, Proc. Natl. Acad. Sci. USA 89 (1992) 2195–2199.

[11] O. Glatter, in: O. Glatter, O. Kratky (Eds.), Small Angle X-Ray

Scattering, Academic Press, London, 1982.

[12] R.G. Silva, L.P. Carvalho, J.S. Oliveira, C.A. Pinto, M.A.

Mendes, M.S. Palma, L.A. Basso, D.S. Santos, Cloning, overex-

pression, and puriﬁcation of functional human purine nucleoside

phosphorylase, Protein Expr. Purif. 27 (2003) 158–164.

[13] A. Guinier, G. Fournet, Small-Angle Scattering of X-rays, Wiley,

New York, 1955.

[14] W.H. Press, S.A. Teukolsky, W.T. Vetterling, B.P. Flannery,

Numerical Recipes in FORTRAN. The Art of Scientiﬁc Com-

puting, second ed., Cambridge University Press, New York, 1992.

[15] S.E. Ealick, S.A. Rule, D.C. Carter, T.J. Greenhough, V. Babu,

W.J. Cook, J. Habash, J.R. Helliwell, J.D. Stoeckler, R.E. Parks

Jr., F. Chen, C.E. Bugg, Three-dimensional structure of human

erythrocytic purine nucleoside phosphorylase at 3.2



AA resolution,

J. Biol. Chem. 265 (1990) 1812–1820.

[16] W.J. Cook, S.E. Ealick, C.E. Bugg, J.D. Stoeckler, R.E. Parks Jr.,

Crystallization and preliminary X-ray investigation of human

erythrocytic purine nucleoside phosphorylase, J. Biol. Chem. 256

(1981) 4079–4080.

[17] A. Nicholls, K. Sharp, B. Honig, Protein folding and association:

insights from the interfacial and thermodynamic properties of

hydrocarbons, Proteins Struct. Funct. Genet. 11 (1991) 281–296.

[18] P.W.K. Woo, C.R. Kostlan, J.C. Sircar, M.K. Dong, R.B.

Gilbertsen, Inhibitors of human purine nucleoside phosphorylase.

Synthesis and biological activities of 8-amino-3-benzylhypoxan-

thine and related analogues, J. Med. Chem. 35 (1992) 1451–1457.

[19] J.-W. Chern, H.-Y. Lee, C.-S. Chen, Nucleosides. 5. Synthesis of

guanine and formycin B derivatives as potential inhibitors of

purine nucleoside phosphorylase, J. Med. Chem. 36 (1993) 1024–

1031.

[20] M.D. Erion, S. Niwas, J.D. Rose, S. Ananthan, M. Allen, J.A.

Secrist III, Y.S. Babu, C.E. Bugg, W.C. Guida, S.E. Ealick, J.A.

Montgomery, Structure-based design of inhibitors of purine

nucleoside phosphorylase. 3. 9-Arylmethyl derivatives of 9-

deazaguanine substituted on the methylene group, J. Med. Chem.

36 (1993) 3771–3783.

[21] J.A. Secrist III, S. Niwas, J.D. Rose, Y.S. Babu, C.E. Bugg, M.D.

Erion, W.C. Guida, S.E. Ealick, J.A. Montgomery, Structure-

based design of inhibitors of purine nucleoside phosphorylase. 2.

9-Alicyclic and 9-heteroalicyclic derivatives of 9-deazaguanine, J.

Med. Chem. 36 (1993) 1847–1854.

[22] W.C. Guida, R.D. Elliot, H.J. Thomas, J.A. Secrist III, Y.S.

Babu, C.E. Bugg, M.D. Erion, S.E. Ealick, J.A. Montgomery,

Structure-based design of inhibitors of purine nucleoside phos-

phorylase. 4. A study of phosphate mimics, J. Med. Chem. 37

(1994) 1109–1114.

[23] S. Niwas, P. Chand, V.P. Pathak, J.A. Montgomery, Structure-

based design of inhibitors of purine nucleoside phosphorylase. 5.

9-Deazahypoxanthines, J. Med. Chem. 37 (1994) 2477–2480.

[24] P.E. Morris, A.J. Elliott, S.P. Walton, C.H. Williams, J.A.

Montgomery, Synthesis and biological activity of a novel class of

purine nucleoside phosphorylase inhibitors, Nucleosides Nucleo-

tides Nucleic Acids 19 (2000) 379–404.

[25] A. Andricopulo, R.A. Yunes, Structure–activity relationships for

a collection of structurally diverse inhibitors of purine nucleoside

phosphorylase, Chem. Pharm. Bull. 49 (2001) 10–17.

[26] A. Sali, T.L. Blundell, Comparative protein modelling by satis-

faction of spatial restraints, J. Mol. Biol. 234 (1993) 779–815.

[27] P.J. Kraulis, MOLSCRIPT: a program to produce both detailed

and schematic plots of proteins, J. Appl. Cryst. 24 (1991) 946–950.

[28] E.A. Merritt, D.J. Bacon, Raster3D: photorealistic molecular

graphics, Methods Enzimol. 277 (1997) 505–524.

928 W.F. de Azevedo Jr. et al. / Biochemical and Biophysical Research Communications 309 (2003) 923–928

ANEXO II

Selection of an Escherichia coli host that expresses mutant forms

of Mycobacterium tuberculosis 2-trans enoyl-ACP(CoA) reductase

and 3-ketoacyl-ACP(CoA) reductase enzymes

Simone S. Poletto,

Isabel O. da Fonseca,

Luiz P.S. de Carvalho,

Luiz A. Basso,

and Di



ogenes S. Santos

Departamento de Biologia Molecular e Biotecnologia, Universidade Federal do Rio Grande do Sul, Avenida Bento Gonc

alves, 9500,

Porto Alegre, RS 91501-970, Brazil

Faculdade de Farmacia Pontif



ıcia, Universidade Catolica do Rio Grande do Sul, Porto Alegre, RS 90619-900, Brazil

Received 14 August 2003, and in revised form 17 October 2003

Abstract

Tuberculosis (TB) still remains a worldwide health concern. Eﬀorts to understand the complex biology of Mycobacterium

tuberculosis, the causative agent of TB, are important for new antitubercular drug development. Despite the completion of the

genome sequence and the development of new genetic tools to manipulate this organism, the availability of suﬃcient amounts

of mycobacterial proteins still remains an essential and laborious step to study the biochemical features of this pathogen. The

T7-RNA polymerase-based pET system has been largely employed to express mycobacterial proteins in Escherichia coli, but it

presents some limitations. To overcome problems with unstable expression of an M. tuberculosis inhA-encoded enoyl reductase

mutant protein and lack of expression of two mabA-encoded ketoacyl reductase mutants, a sub-population of E. coli

BL21(DE3) host cells was selected from a small-opaque colony. This empirically selected host, named BL21(DE3)NH, allowed

stable expression of these mutant proteins. Although the mechanism that led the BL21(DE3)NH host to express the re-

combinant mutant proteins remains unknown, the persistent phenotype points to a stable genetic switch. This genetic alteration

resulted in a tight control of the highly processive T7 RNA polymerase. Moreover, the absolute requirement for IPTG to

obtain protein expression in the BL21(DE3)NH host cells suggests that no inherent defect in the transcriptional activity of the

T7 promoter is present. Empirical host selection requires no further genetic manipulation of recombinant plasmids and may

represent a means of obtaining tailor-made E. coli strains that overcome toxic eﬀects associated with heterologous protein

expression.

Keywords: Protein expression; Escherichia coli; Host selection; Mycobacteria

Tuberculosis (TB), caused by Mycobacterium tuber-

culosis, remains one of the deadliest diseases in the

world. It is estimated that 8.2 million new TB cases

occurred worldwide in the year 2000, 1.8 million deaths

occurred in the same year, and more than 95% of those

were in developing countries [1]. Possible factors un-

derlying the resurgence of TB include the HIV epidemic,

increase in the homeless population, and decline in

health care structures and national surveillance [2].

Another contributing factor is the evolution of multi-

drug-resistant strains of M. tuber culosis (MDR-TB),

deﬁned as resistant to at least isoniazid and rifampicin,

which are the most eﬀective ﬁrst-line drugs [3]. Treat-

ment of MDR-TB requires the administration of sec-

ond-line drugs (amikacin, kanamycin, capreomycin,

cycloserine, para-aminosalicylic acid, ethionamide, and

ﬂuoroquinolones) that are more toxic and less eﬀective

and are given for at least three times as long and at 100

times the cost of basic short-course chemotherapy regi-

mens [4]. Hence, new antimycobacterial agents are need-

ed to improve the treatment of MDR-TB, to shorten the

treatment course to improve patient compliance, and to

Corresponding author. Fax: +55-51-33167309.

E-mail addresses: [email protected] (L.A. Basso),

[email protected] (D.S. Santos).

doi:10.1016/j.pep.2003.10.009

Protein Expression and Puriﬁcation 34 (2004) 118–125

www.elsevier.com/locate/yprep

provide for more eﬀective treatment of latent tubercu-

losis infection [5].

The study of biology, pharmacology, and host–

pathogen interactions is important for the development

of new drugs as well as vaccines against M. tuber culosis.

The completion of the genome sequence of M. tuber-

culosis [6] and availability of genetic tools to manipulate

this organism [7] have accelerated these studies. Despite

these advances, the availability of suﬃcient amounts of

proteins of M. tuberculosis still remains an essential and

laborious step to study the biochemical features of this

pathogen. This fact is due to (i) the slow-growth of the

bacillus whose generation time is 24 h; (ii) since aerosols

containing bacteria are the way of spread of the disease,

a P3 facility is required to grow the bacillus culture; (iii)

and the highly lipophilic envelope of mycoba cteria that

represents an obstacle to protein extraction and puriﬁ-

cation. On the other hand, the improvements in the re-

combinant DNA technology, the advances in molecular

genetic studies in mycobacteria and, as previously

pointed out, the availability of the complete genome

sequence have facilitated the cloning and he terologous

expression of mycobacterial proteins. Escherichia coli

and Mycobacterium smegmatis (a fast-growing bacte-

rium that belongs to the same genus of M. tuberculosis)

are the ﬁrst choices for hosts [8]. E. coli, in particular,

remains one of the most attractive systems for heterol-

ogous protein expression due to its ability to grow

rapidly and at high density on inexpensive substrates, its

well-characterized genetics, and the avail ability of an

increasingly large number of cloning vectors and mutant

host strains [9]. The pET system (Novagen) developed

by Studier and Moﬀat [10] has been one of the most

widely used systems for expression of mycobacterial

proteins in E. coli [11–15]. In this system, target genes

are positioned downstream of the bacteriophage T7 late

promoter on medium copy number plasmids. The highly

processive T7 RNA polymerase gene (DE3 lysogen) of

production hosts is placed under control of the IPTG-

inducible lacUV5 promoter. However, expression of

membrane proteins and proteins toxic to the host cells

[16], formation of inclusion bodi es [17] or complete

absence of expression are amongst the diﬃculties en-

countered with mycobacterial proteins.

The inhA-encoded 2-trans enoyl-ACP (CoA) reduc-

tase (ENR) and mabA-encoded 3-ketoacyl-ACP (CoA)

reductase (KAR) are enzymatic components of the

Type II fatty acid synthase (FAS-II) of mycobacteria

and are thus involved in mycolic acid biosynthesis [18].

The mycobacterial NADH-dependent ENR is the tar-

get for isoniazid [11,19], the most prescribed drug to

treat tuberculosis, making it an excellent target for TB

drug development. Another reductive step of the

mycobacterial FAS-II system is catalyzed by the

NADPH-dependent KAR enzyme [20]. The FAS-II

system enzymes have been shown to be essential for

mycobacterial growth and conﬁgured into high

throughput screening for identiﬁcation of new drugs

against TB [21] . The amino acid residues tyrosine-158

and lysine-165 have been implicated in the catalytic

mechanism of M. tuberculosis ENR [11,22,23]. More-

over, the side chain of Y158 of M. tuberculosis ENR

has been shown to unde rgo a rotation upon binding of

triclosan, an e nzyme inhibi tor, to form a hydrogen

bond with the hydroxyl group of this inhibitor [18].

The mycobacterial KAR and ENR enzymes are

members of the short-chain dehydrogenase/reductase

(SDR) family and thereby display the amino acid sig-

nature of this family, S–(X)

–Y–(X)

–K [24]. As part

of our eﬀorts to develop novel antitubercular drugs

based on rational design, we are studying the catalytic

mechanism of M. tuberculosis ENR and KAR en-

zymes, identifying amino acid residues involved in ei-

ther catalysis or substrate binding, and trying to solve

their three-dimensional structures. Accordingly, seven

mutant proteins were produced for the putative cata-

lytic amino acids (ENR: Y158F, K165A, K165Q;

KAR: S140T, S140A, Y153F, and K157A), cloned into

the pET system, and transformed into E. coli

BL21(DE3) host strain. However, the Y158F ENR

mutant failed to show stable protein expression in this

system, and no expression could be obtained for the

Y153F and K157A KAR mutant enzymes.

Miroux and Walker [16] have selected mutant hosts

derived from E. coli BL21(DE3) that allowed protein

expression at higher levels than the parent host. The

mutant host C41(DE3) has been employed to express E.

coli inner membrane enzyme acyl acyl carrier protein

synthase [25] and M. tuberculosis acyl carrier protein

(AcpM) and malonyl-CoA:AcpM transacylase [15].

Building on the strategy adopted by Miroux and Walker

[16], we selected a sub-population of E. coli BL21(DE3),

named BL21(DE3)NH, that was able to express

theY158F ENR, Y153F KAR, and K157A KAR mu-

tant proteins, which could not be expressed in the E. coli

BL21(DE3) parent strain. Here, we report site-directed

mutagenesis to obtain theY158F ENR, Y153F KAR,

and K157A KAR mutants, and selection of a host cell

that was able to express these mutant proteins in soluble

form. Scanning electron microscopy has been performed

to evaluate changes in BL21(DE3)NH morphology as

compared to the BL21(DE3) parent strain.

Abbreviations used: E. coli, Escherichia coli; ENR, enoyl reduc-

tase; InhA, enoyl-ACP (CoA) reductase from Mycobacterium tuber-

culosis; KAR, 3-ketoacyl-ACP (CoA) reductase; LB, Luria–Bertani

medium; SDS–PAGE, sodium dodecyl sulfate–polyacrylamide gel

electrophoresis; IPTG, isopropyl-b-

-thiogalactopyranoside; MabA,

3-ketoacyl-ACP (CoA) reductase from Mycobacterium tuberculosis;

PCR, polymerase chain reaction; bp, base pairs; WT, wild type; EtdBr,

ethidium bromide; SEM, scanning electron microscopy; OsO

, osmium

tetraoxide; expression plasmid names consist of the name of the parent

plasmid vector followed by the recombinant DNA insert they encode.

S.S. Poletto et al. / Protein Expression and Puriﬁcation 34 (2004) 118–125 119

Materials and methods

Materials

The pET-23d(+)::inhA recombinant plasmid was a

gift from Dr. John S. Blanchard at Albert Einstein

College of Medicine (AECOM, Bronx, NY) and pET-

3d::mabA recombinant plasmid was a gift from Dr.

William R. Jacobs Jr. at AECOM (Bronx, NY). pET-

23d(+), pET- 23a(+), BL21(DE 3) E. coli strain, and

Perfect Protein Marker were purchased from Novagen;

Luria–Bertani medium (LB), oligonucleotides, T4 DNA

ligase, IPTG, DH10B cells, and DNA and protein mo-

lecular weight standards were from Gibco-BRL; car-

benicillin was from Sigma; BamHI, HindIII, NcoI, and

NdeI restriction enzymes were from Boehringer–Mann-

heim and New England Biolabs; Pfu DNA polymerase

was from Stratagene; Agarose was from SEA KEM

GTG; Zero Blunt PCR Cloning Kit pCR-Blunt and

Top10 One Shot were from Invitrogen; Quantum Prep

Plasmid Miniprep kit, QIAEX II Agarose Gel Extrac-

tion kit, and QIAprep Spin Miniprep kit were from

Qiagen; Thermo Sequenase radiolabeled terminator cy-

cle sequencing kit was from Amersham Life Science; and

SDS–polyacrylamide gels and Tris–glycine running

buﬀer were from Bio-Rad.

Cloning of wild-type mabA gene into pET23a(+) vector

The synthetic oligonucleotide primers used in this

study are listed in Table 1. The primers mabA5, con-

taining a NdeI restriction site (bold), and mabA3, con-

taining a HindIII restriction site (bold), were used to

amplify the mabA gene from the pET-3d:: mabA re-

combinant plasmid using Pfu DNA polymerase at

standard conditions. The PCR product was puriﬁed

from agarose gel using QIAEX II Kit and inserted into a

pCR-blunt vector. E. coli Top 10 One-Shot competent

cells were transformed with the recombinant plasmid.

Plasmid DNA was extracted using Quantum Prep

Plasmid Miniprep Kit (Qiagen). The insert was removed

by NdeI and HindIII double digestion and ligated into a

pET23a(+) digested with the same restriction enzymes.

The integrity of the mabA gene was conﬁrmed by DNA

sequencing.

Site-directed mutagenesis of inhA an d mabA genes

Site-directed mutations were introduced in the inhA

and mabA genes using the method described by Ho et al.

[26]. A list of the primers used for all ampliﬁcation steps

is given in Table 1. It should be noted that only the

primers to obtain theY158F ENR, Y153F KAR, and

K157A KAR proteins by site-directed mutagenesis are

described here, since these were the mutants that showed

unstable or no protein expression. The experimental

protocol described here is exempliﬁed by Y158F inhA

mutant gene production. First, two independent PCRs

using Pfu DNA polymerase were performed using

primers P1 and Y158b

—reaction 1, and JSB02 and

Y158a

—reaction 2, and the plasmid pET23d (+)::inhAas

template. The PCR program was as follows: one step of

98 °C for 3 min, 35 cycles at 98 °C for 1 min, 65 °C for

1 min, and 72 °C for 2 min, followed by a ﬁnal extension

step at 72 °C for 2 min, using a DNA Thermal Cycler

(PTC200, MJ, USA). The PCR products of these reac-

tions were puriﬁed by electrophoresis on 1% agarose gel

and DNA was extracted using the QIAEX II Kit. A

second PCR was performed using the products of the

reactions 1 and 2 as templates and the primers P1 and

JSB02. The overlap PCR conditions were identical to

those used in the ﬁrst round of ampliﬁcation. The PCR-

ampliﬁed band was puriﬁed and cloned as described in

the previous section, yielding pET23d(+)::inhAY158F

recombinant plasmid. The primers used to make the

mabA mutants are listed in Table 1 and followed the

same protocol just described, yielding pET23a(+)::ma-

bAY153F and pET23a(+)::mabAK157A recombinant

plasmids. The sequence of the DNA inserts in all re-

combinant plasmids was conﬁrmed by the dideoxynu-

cleotide chain termination method using the Thermo

Sequenase radiolabeled terminator cycle sequencing kit

(Amersham Biosciences).

Protein expression of WT and ENR mutants in

BL21(DE3)

To assess the expression of these constructs,

BL21(DE3) electrocompetent cells were transformed

with the following recombinant plasmids: pET23-

d(+)::inhA and pET23d(+)::inhAY158F and grown in

LB medium containing carbenicillin (50 lgml

À1

)at

various times after induction with IPTG (1 mM) at

37 °C. Control experiments were performed under the

same experimental conditions, except that E. coli

Table 1

Primers used for ampliﬁcation and mutagenesis of inhA and mabA

genes from M. tuberculosis H37Rv

Primers Restriction

sites

Sequence

P1 NcoI5

-att gaa cca tgg cag gac tgc tgc acg-3

JSB02 BamHI 5

-gcggatccgctagagcaattgggtgtgcgc-3

Y158a — 5

-atg ccg gcc ttc aac tgg atg-3

Y158b — 5

-cat cca gtt gaa ggc cgg ca t-3

mabA5 NdeI5

-att cat atg act gcc aca gcc act gaa gg-3

mabA3 HindIII 5

-t aag ctt tca gtg gcc cat acc cat gcc-3

Y153F5 — 5

-cag gcc aac ttc gca gcc tcc-3

Y153F3 — 5

-gga ggc tgc gaa gtt ggc ctg-3

K157A5 — 5

-gca gcc tcc gcg gcc gga gtg-3

K157A3 — 5

-cac tcc ggc cgc gga ggc tgc-3

The sequences underlined correspond to degenerate codons and the

sequences in bold correspond to restriction sites.

120 S.S. Poletto et al. / Protein Expression and Puriﬁcation 34 (2004) 118–125

BL21(DE3) host cells harbored the expression vector

lacking the target gene. Cells were harvested (14,000g

for 10 min) at 4, 8, 12, 16, 20, and 24 h after induction,

lysed by sonication, and cell debris was removed by

centrifugation (20,000g for 20 min). The resulting solu-

ble protein content was analyzed by 12% SDS–PAGE.

Stable host production

Escherichia coli BL21(DE3) electrocompetent cells

were transformed with pET23d(+)::inhAY158F, plated

on LB solid medium containing carbenicillin

(50 lgml

À1

), and grown overnight at 37 °C, yielding sub-

populations of large-transluc ent and small-opaque col-

onies. These two types of colonies were examined for

their ability to express Y158F ENR protein in liquid

culture, using the protocol of IPTG induction as previ-

ously described for wild-type ENR. The degree of pro-

tein expression, as monitored by SDS–PAGE, was

dependent on colony morphology, with smaller colonies

expressing the desired protein, whereas larger colonies

did not. Since sub-culturin g may eventually result in loss

of phenotype, E. coli BL21(DE 3) host cells wer e trans-

formed with pET23d(+)::inhAY158F recombinant plas-

mid and plated on LB–carbenicillin overnight at 37 °C. A

single small colony was used to inoculate 5 ml LB–car-

benicillin medium, grown to midlog phase (0.4–0.6

600

)at37° C, and induced with 1 mM IPTG for 3 h.

At midlog phase of growth, an aliquot was withdrawn to

prepare a glycerol culture stock. The remaining of cells

were harvested by centrifugation at 14,000g for 10 min.

Pelleted material was solubilized in 400 llof20mM

Pipes buﬀer, pH 7.3, disrupted by sonication, and the

resulting crude extract was centrifuged at 14,000g for

10 min. Expression of the recombinant protein in soluble

form was conﬁrmed by 12% SDS–PAGE. The glycerol

stock of this recombinant protein expressing host was

cured of pET-23d(+)::inhAY158F by growth in LB li-

quid medium in the absence of carbenicillin for 15 days.

Each day, a portion of the culture (5 ll) was used to in-

oculate 5 ml of LB. After 15 days, small-colony derived

BL21(DE3) cells lacking the recombinant plasmid arose,

which was veriﬁed both by its inability to grow on LB

solid medium containing carbenicillin and no recovery of

recombinant plasmid from the cells. To conﬁrm the

ability of the new host, henceforth named

BL21(DE3)NH (NH standing for New Host), to confer

stable overexpression of Y158F ENR protei n,

BL21(DE3)NH electrocompetent cells were transformed

with either pET-23d(+) plasmid as control or pET-

23d(+)::inhAY158F recombinant plasmid, plated on LB-

agar plate containing carbenicillin (50 lgml

À1

), and

grown overnight at 37 °C. Seven single colonies were

separately used to inoculate 5 ml LB–carbenicillin liquid

medium, grown to an OD

600

of 0.4–0.6, induced with

1 mM IPTG, and grown for 3 h at 37 °C after induction.

The cells were harvested by centrifugation at 14,000g for

10 min, pelleted material was solubilized in 400 llof

20 mM Pipes buﬀer, pH 7.3, and disrupted by sonication.

Soluble proteins were analyzed by SDS–PAGE.

Protein expression of mycobacterial KAR mutants in

BL21(DE3)NH host cells

To address the ability of this new host to express pro-

teins that, in our hands, have failed to do so in the

BL21(DE3) parent host, BL21(DE 3), and BL21(DE3)

NH electrocompetent cells were transformed with the

following plasmids: pET-23a(+), pET-23a(+)::mabAY

153F and pET-23a(+)::mabAK157A, plated on LB–car-

benicillin medium, and incubated overnight at 37 °C.

Single colonies were used to inoculate 5 ml LB medium

containing carbenicillin (50 lgml

À1

) and 1 mM of IPTG

was added to the liquid cultures as they reached an OD

600

between 0.4 and 0.6. The cultures were grown at 37 °C for

8 h as determined for the wild-type KAR protein in

BL21(DE3) (data not shown). Cells were harvested by

centrifugation (14,000g for 10 min) and pellets were sol-

ubilized in 400 ll of 50 mM Tris–HCl, pH 7.8, buﬀer,

disrupted by sonication, and clariﬁed by centrifugation

(14,000g for 10 min). The soluble fraction of the cellular

extracts was analyzed by SDS–PAGE.

Scanning electronic microscopy

The procedure for scanning electronic microscopy of

E. coli was carried out as described elsewhere [27,28].

Slides of E. coli BL21(DE3) and BL21(DE3)NH were

obtained by cutting single colonies (1 mm

and 0.5 mm of

thickness) from the LB agar plates, followed by a 2-h

immersion in Karmovisk solution at room temperature to

chemically stabilize the sample and an additional ﬁxation

with 1% OsO

. Samples were then washed 10 times with

cacodylate buﬀer solution. For dehydration, samples

were washed with 30–100% ethanol and, subsequently,

with aqueous carbonic acid. The specimens were coated

with gold by the sputtering process to obtain a good sec-

ondary electron production for image formation. Mi-

crographs were taken using Zeiss DMS 940A and

operated in electron emission mode at 20 kV accelerating

voltage. Samples were observed in 3000Â ampliﬁcation at

12 mm distance.

Results and discussion

Site-directed mutagenesis of both inhAandmabA

genes was performed by the overlap extension method

[26] and mutations were conﬁrmed by DNA sequencing.

The inhA mutant was cloned into pET-23d(+) and

mabA mutants were inserted into pET-23a(+) expres-

sion plasmid.

S.S. Poletto et al. / Protein Expression and Puriﬁcation 34 (2004) 118–125 121

Overexpression of mycobacterial K165A, K165Q,

and Y158F ENR mutant protei ns is shown in Fig. 1.

However, mycobacterial Y158F ENR mutant protein

overexpression was not stable, that is, not all colonies

picked up at random expressed the recombinant protein.

The appearance of two types of colonies with distinct

morphologies (large and small) of BL21(DE3) trans-

formed with pET-23d(+)::inhAY158F plasmid was ob-

served on LB–carbenicillin plates incubated overnight at

37 °C. The small colonies, which have diameters ap-

proximately 30% smaller than those of large colonies,

were opaque, while the large colonies were translucent.

In addition, the small-opaque colonies expressed the

desired proteins while the large-translucent colonies did

not. Variations in colony morphology of E. coli hosts

have been associated with detection of recombinant

colonies [29] and expression of the recombinant gene

[30]. Moreover, empirical selection has been employed

by Miroux and Walker [16] to tailor expression hosts to

overcome toxic eﬀects associated with protein overex-

pression. Building on Miroux an d Walker strategy, a

sub-population of BL21(DE3) cells having a small-

opaque phenotype that expressed the Y158F ENR

mutant protein was selected, isolated, and cured. The

selected host cell, named BL21(DE3)NH, transformed

with pET-23d(+)::inhAY158F and plated on LB–car-

benicillin solid medium, yielded a population of small

colonies only. To evaluate whether protein expression of

the recombinant protein requires IPTG induction, seven

isolated colonies were picked up at random, suspended

in 200 ll LB–carbenicillin medium and 50 ll of the re-

sulting suspension used to inoculate 5 ml LB–carbeni-

cillin medium, and the experimental protocol described

in Material and methods was followed. Fig. 2A shows a

representative result for a single colony, suggesting that

expression of Y158 ENR in the BL21(DE3)NH host can

be achieved only in the presence of the inducer. Fig. 2B

shows that all seven colonies express the recombinant

Y158F ENR mutant protein in soluble form upon IPTG

induction.

A protocol had been designed to optimize protein

expression of mycobacterial wild-type KAR, involving

addition of 1 mM IPTG to an LB–carbenicillin liquid

culture (0.4 < OD

600 nm

< 0.6) of BL21(DE3) and further

growth for 8 h at 37 °C. The mycobacterial Y153F and

K157A KAR mutants, however, did not show any ex-

pression under these experimental conditions (Fig. 3A).

To test the ability of the empirically selected host cells to

express these mutants, BL21(DE3)NH cells were trans-

formed with pET23a(+)::mabAY153F and pE-

T23a(+)::mabAK157A recombinant plasmids, and

pET23a(+) plasmid (control). LB–carbenicill in liquid

medium was inoculated with single colonies, and cell

growth and induction was carried out as for the

Fig. 1. SDS–PAGE (12%) of the soluble fraction of the cell extracts

expressing M. tuberculosis ENR proteins upon IPTG induction in E.

coli BL21(DE3) host cells. Lane M, Protein molecular weight stan-

dards (Gibco-BRL: insulin a and b chains, 3 kDa; bovine trypsin in-

hibitor, 6.2 kDa; lysozyme, 14.3 kDa; b-lactoglobulin, 18.4 kDa;

carbonic anhydrase, 29 kDa; and ovalbumin, 43 kDa); lane 1, wild-

type ENR; lane 2, pET-23d(+); lane 3, K165A ENR; lane 4, K165Q

ENR; and lane 5, Y158F ENR. The arrow points the band corre-

sponding to M. tuberculosis ENR proteins ($29 kDa).

Fig. 2. SDS–PAGE (12%) of the soluble fraction of the cell extracts of BL21(DE3)NH transformed with either pET-23d(+) (control) or pET-

23d(+)::inhAY158F recombinant plasmid. (A) Lane 1, 1 mM IPTG; lane 2, no IPTG; lane 3, control with 1 mM IPTG; lane 4, control in the absence

of IPTG; and lane M, protein molecular weight standards (same as Fig. 1). (B) Lane 1, control with 1 mM IPTG; lane M, protein molecular weight

standards (same as Fig. 1); and lanes 2–8, protein expression of seven colonies of BL21(DE3)NH transformed with pET-23d(+)::inhAY158F plasmid

upon 1 mM IPTG induction. The arrows point the band corresponding to mycobacterial ENR ($29 kDa).

122 S.S. Poletto et al. / Protein Expression and Puriﬁcation 34 (2004) 118–125

wild-type KAR enzyme. Cells were harvested and the

resulting pellets were suspend ed in Tris–HCl 50 mM, pH

7.8, disrupted by sonication and the solubl e fractions

were analyzed by SDS–PAGE. High level of protein

expression was observed in BL21(DE3)NH host cells

only upon IPTG induction (Fig. 3B).

The intrinsic toxicity to E. coli BL21(DE3) host cells of

pET vectors harboring no insert in the presence of IPTG

[16] has not been observed for the BL21(DE3)NH host

cells (Figs. 2 and 3B). Moreover, the BL21(DE3)NH host

phenotype was stable, since it continued to give rise to

small colonies on LB–carbenicillin agar plates upon

transformation with the recombinant plasmids tested.

Scanning electron microscopy (SEM) results show that

the E. coli BL21(DE3)NH cells have a distinct morphol-

ogy of long bacillary shape (F ig. 4A). The elongation is

evident and contrasts with the E. coli BL21(DE3) mor-

phology (Fig. 4B). The BL21(DE3)NH host cell is a

Gram-negative bacterium. Moreover, the biochemical

properties of BL21(DE3) NH host cells were determined

by standard methods for E. coli identiﬁcation: urease

production, indole production, carbohydrate fermenta-

tion, b-galactos idase activity, citrate utilization, and

growth on MacConckey agar medium. All these tests

conﬁrm that BL21(DE3)NH host cell is a strain of E. coli.

The mechanisms of translucent to opaque phenotype,

variations in colony morphology (large and small), and

alteration to the long bacillary shape of BL21(DE3)NH

cells remain unknown. It has been shown that opacity is

correlated with a robust transcription of the recombinant

gene whereas translation of recombinant RNA is not re-

quired [30]. Interestingly, uncoupling of transcription

from translation has been implicated in toxicity of re-

combinant protein expression leading to host cells pro-

ducing large amounts of the recombinant mRNA while

target proteins are maintained at rather low level s [16]. In

addition, expression of gene products that do not con-

tribute to the metabolism of bacterium, under the growth

conditions used in a particular experiment, has been

shown to result in cumulative breakdown of rRNAs and

ensuing loss of ribosomes and protein synthetic capacity

[31]. Although the mechanism that led the empirically

selected hos t to tolerate the recombinant mutant protein

expression is unknown, the BL21(DE3)NH cells derived

from small-opaque colonies transformed with re-

combinant plasmids resulted in small-opaque transfor-

mants, suggesting that the persistence of the phenotype

points to a stable genetic switch.

Leaky expression has been shown to occur in the pET

system [13,32,33]. The BL21(DE3)NH transformants

showed an absolute requirement for IPTG induction to

achieve expression of the recombinant proteins tested

(Figs. 2 and 3B). Accordingly, the genetic alteration of

the empirically selected host reported here appears to

have resulted in a tight control of the highly processive

T7 RNA polymerase that is under control of the IPTG-

inducible lacUV5 promoter. Moreover, the IPTG ab-

solute requirement for expression of the mycobacterial

ENR and KAR proteins in the BL21(DE3)NH host cell

suggests that no inherent defect in the transcriptional

activity of the T7 promoter is present.

It has been suggested that the mechanism responsible

for the diﬀerences in colony morphology is bacterial

growth rate [29]. Although we have also found this to be

generally true, any causal relationship between opacity

and growth rate seems to be unwarranted because it is

somewhat diﬃcult to establish a quantitative relationship

between them based on a qualitative observation

(opacity).

Protein production and crystallization must be opti-

mized if structural genomics will ever reach its goal of

solving the three-dimensional structure of the whole

Fig. 3. SDS–PAGE (12%) of the soluble fraction of cell extracts of either BL21(DE3) or BL21(DE3)NH transformed with either pET-23a(+)

(control), pET-23a(+)::mabAK157A, or pET-23a(+)::mabAY153F recombinant plasmids. (A) Plasmids transformed into BL21(DE3). Lane 1,

control in the absence of IPTG; lane 2, control with 1 mM IPTG; lane M, protein molecular weight standards (same as Fig. 1); lane 3, K157A KAR

in the absence of IPTG; lane 4, K157A KAR with 1 mM IPTG; lane 5, Y153F KAR in the absence of IPTG; and lane 6, Y153F KAR with 1 mM

IPTG. (B) Plasmids transformed into BL21(DE3)NH. Lane 1, control in the absence of IPTG; lane 2, control with 1 mM IPTG; lane M, protein

molecular weight standards (same as Fig. 1); lane 3, K157A KAR in the absence of IPTG; lane 4, K157A KAR with 1 mM IPTG; lane 5, Y153F

KAR in the absence of IPTG; and lane 6, Y153F KAR with 1 mM IPTG. The arrow points to the band corresponding to the expected molecular

weight of M. tuberculosis KAR proteins (25.6 kDa).

S.S. Poletto et al. / Protein Expression and Puriﬁcation 34 (2004) 118–125 123

proteome encoded by a given genome [34]. Unfortu-

nately, even when a genome can be sequenced, only up to

20% of the protein targets can produce soluble proteins

under very basic experimental conditions [35]. Thus, ex-

pression of proteins in soluble form has been identiﬁed as

an important bottleneck in eﬀorts to determine biological

activity and crystal structure of M. tuberculosis proteins

[17]. A number of strategies for optimizing the yields of

heterologous protein expression in E. coli have been de-

scribed [36,37], including the use of diﬀerent types of

promoters, placement of transcription terminators up-

stream of the promoter that drives expression of the gene

of interest, reduction of potential secondary-structure

formation at the 5

end of the transcript, use of transla-

tional enhancer, positioning of speciﬁc sequences in the 5

untranslated or 3

unstranslated regions of labile heter-

ologous mRNAs to prolong their half-lives, altering rare

codons in the target gene or coexpressing genes that en-

code rare tRNAs, mini mizing proteolysis by mutations to

avoid the E. coli ‘‘N-end rule’’ proteolytic pathway, and

altering fermentation conditions such as nutrient com-

position, temperature, and pH. Incidentally, although

wild-type KAR has nine rare codons (2Â CCC–proline,

5Â GGA–glycine, and 2Â ATA–isoleucine) and wild-

type ENR has six rare codons (2Â CCC–proline, 3Â

GGA–glycine, and 1Â AGG–arginine), they have no

adverse eﬀect on the synthesis and yield of these re-

combinant proteins, which corroborates the conclusion

arrived by others [37] that the mere presence of rare co-

dons in a gene does not necessarily dictate poor transla-

tion of that gene.

The E. coli host reported here may represent an ad-

ditional tool to obtain mycobacterial soluble proteins

using T7-RNA polymerase-based systems as has been

the case for the C41(DE3) and C43(DE3) hosts

[15,16,25] now marketed by Avidis S.A. (France). Em-

pirical host selection requires no furt her genetic ma-

nipulation of recombinant plasmids and may yield

strains that overcome toxic eﬀects associated with the

overexpression of heterologous protei ns. As poin ted out

by Miroux and Walker [16], even though the host se-

lection described here is empirical, it has the advantage

that it encompasses the entire complexity of the biology

of the expression system and provides a means of

modifying it. The expression of M. tuberculosis ENR

and KAR mutants will provide protein in quantities

necessary for determination of the mechanisms of action

of these enzymes by steady -state and pre-steady-state

kinetics as well as for crystallization trials aiming at X-

ray data collection. Enzymological and structural stud-

ies of M. tuberculosis ENR and KAR mutants should

help in the design of enzyme inhibitors to be tested as

new antitubercular agents.

Acknowledgments

Financial support for this work was provided by

Millennium Initiative Program MCT-CNPq, Ministry

of Health-Secretary of Health Policy (Brazi l) to D.S.S.

and L.A.B. D.S.S. and L.A.B. also acknowledge grants

awarded by PADCT, CNPq, an d FINEP. L.A.B.

(CNPq, 520182/99-5) is a researcher for the Brazilian

Council for Scientiﬁc and Technological Development.

We also thank Dr. Luiz Ant

onio Suita de Castro and

Embrapa Clima Temperado for the preparation of mi-

crographs and laboratory facilities.

References

[1] M.C. Raviglione, The TB epidemic from 1992 to 2002, Tubercu-

losis 83 (2003) 4–14.

Fig. 4. Scanning electron microscopy of E. coli hosts observed at

3000Â ampliﬁcation. (A) E. coli BL21(DE3)NH and (B) E. coli

BL21(DE3).

124 S.S. Poletto et al. / Protein Expression and Puriﬁcation 34 (2004) 118–125

[2] B.R. Bloom, C.J.L. Murray, Tuberculosis: commentary on a

reemergent killer, Science 257 (1992) 1055–1064.

[3] L.A. Basso, J.S. Blanchard, Resistance to antitubercular drugs,

Adv. Exp. Med. Biol. 456 (1998) 115–144.

[4] A. Plabos-M



endez, D.K. Gowda, T.R. Frieden, Controlling

multidrug-resistant tuberculosis and access to expensive drugs: a

rational framework, Bull. World Health Org. 80 (2002) 489–495.

[5] R.J. OÕBrien, P.P. Nunn, The need for new drugs against

tuberculosis, Am. J. Respir. Crit. Care Med. 162 (2001) 1055–

1958.

[6] S.T. Cole, R. Brosch, J. Parkhill, T. Garnier, C. Churcher, D.

Harris, S.V. Gordon, K. Eiglmeier, S. Gas, C.E. Barry III, F.

Tekaia, K. Badcock, D. Basham, D. Brown, T. Chillingworth, R.

Connor, R. Davies, K. Devlin, T. Feltwell, S. Gentles, N. Hamlin, S.

Holroyd, T. Hornsby, K. Jagels, A. Krogh, J. McLean, S. Moule, L.

Murphy, K. Oliver, J. Osborne, M.A. Quail, M.A. Rajandream, J.

Rogers, S. Rutter, K. Seeger, J. Skelton, S. Squares, R. Squares, J.E.

Sulston, K. Taylor, S. Whitehead, B.G. Barrell, B.G. Deciphering,

Deciphering the biology of Mycobacterium tuberculosis from the

complete genome sequence, Nature 393 (1998) 537–544.

[7] M.S. Glickman, W.R. Jacobs Jr., Microbial pathogenesis of

Mycobacterium tuberculosis: dawn of a discipline, Cell 104 (2001)

477–485.

[8] T.M. Shinnick, C.H. King, F.D. Quinn, Molecular biology,

virulence, and pathogenicity of mycobacteria, Am. J. Med. Sci.

309 (1995) 92–98.

[9] F. Baneyx, Recombinant protein expression in Escherichia coli,

Curr. Opin. Biotechnol. 10 (1999) 411–421.

[10] F.W. Studier, B.A. Moﬀat, Use of bacteriophage T7 RNA

polymerase to direct selective high-level expression of cloned

genes, J. Mol. Biol. 189 (1986) 113–130.

[11] A. Qu



emard, J.C. Sacchettini, A. Dessen, C. Vilch



eze, R. Bittman,

W.R. Jacobs Jr., J.S. Blanchard, Enzymatic characterization of

the target for isoniazid in Mycobacterium tuberculosis, Biochem-

istry 34 (1995) 8235–8241.

[12] L.A. Basso, D.S. Santos, W. Shi, R.H. Furneaux, P.C. Tyler, V.L.

Schramm, J.S. Blanchard, Purine nucleoside phosphorylase from

Mycobacterium tuberculosis analysis of inhibition by a transition-

state analogue and dissection by parts, Biochemistry 28 (2001)

8196–8203.

[13] J.S. Oliveira, C.A. Pinto, L.A. Basso, D.S. Santos, Cloning and

overexpression in soluble form of functional shikimate kinase and

5-enolpyruvylshikimate 3-phosphate synthase enzymes from My-

cobacterium tuberculosis, Protein Express. Purif. 3 (2001) 430–435.

[14] W. Shi, L.A. Basso, D.S. Santos, P.C. Tyler, R.H. Furneaux, J.S.

Blanchard, S.C. Almo, V.L. Schramm, Structures of purine

nucleoside phosphorylase from Mycobacterium tuberculosis in

complex with immucillin-H and its pieces, Biochemistry 28 (2001)

8204–8215.

[15] L. Kremer, K.M. Nampoothiri, S. Lesjean, L.G. Dover, S.

Graham, J. Betts, P.J. Brennan, D.E. Minnikin, C. Locht, G.S.

Besra, Biochemical characterization of acyl carrier protein

(AcpM) and malonyl-CoA:AcpM transacylase (mtFabD), two

major components of Mycobacterium tuberculosis fatty acid

synthase, J. Biol. Chem. 276 (2001) 27967–27974.

[16] B. Miroux, J.E. Walker, Over-production of proteins in Esche-

richia coli: mutant host that allow synthesis of some membrane

proteins and globular proteins at high levels, J. Mol. Biol. 260

(1996) 289–298.

[17] R. Vicentelli, C. Bignon, A. Gruez, S. Canaan, G. Sulzenbacher,

M. Tegoni, V. Campanacci, C. Cambillau, Medium-scale struc-

tural genomics: strategies for protein expression and crystalliza-

tion, Acc. Chem. Res. 36 (2003) 165–172.

[18] E.K. Schroeder, O.N. de Souza, D.S. Santos, J.S. Blanchard, L.A.

Basso, Drugs that inhibit mycolic acid biosynthesis in Mycobac-

terium tuberculosis, Curr. Pharm. Biotechnol. 3 (2002) 197–225.

[19] A. Banerjee, E. Dubnau, A. Qu



emard, V. Balasubramanian, K.S.

Um, T. Wilson, D. Collins, G. Lisle, W.R. Jacobs Jr., inhA, a gene

encoding a target for isoniazid and ethionamide in Mycobacterium

tuberculosis, Science 263 (1994) 227–230.

[20] A. Banerjee, M. Sugantino, J.C. Sacchettini, W.R. Jacobs Jr., The

mabA gene from the inhA operon of Mycobacterium tuberculosis

encodes a 3-ketoacyl reductase that fails to confer isoniazid

resistance, Microbiology 144 (1998) 2697–2704.

[21] P.J. Brennan, Structure, function, and biogenesis of the cell wall of

Mycobacterium tuberculosis, Tuberculosis 83 (2003) 91–97.

[22] D.A. Rozwarski, C. Vilch



eze, M. Sugantino, R. Bittman, J.C.

Sacchettini, Crystal structure of the Mycobacterium tuberculo-

sis enoyl-ACP reductase, InhA, in complex with NAD

and a

C16 fatty acyl substrate, J. Biol. Chem. 274 (1999) 15582–

15589.

[23] S. Parikh, D.P. Moynihan, G. Xiao, P.J. Tonge, Roles of tyrosine

158 and lysine 165 in the catalytic mechanism of InhA, the enoyl-

ACP reductase from Mycobacterium tuberculosis, Biochemistry 38

(1999) 13623–13634.

[24] B. Persson, Y. Kallberg, U. Oppermann, H. J

€

ornvall, Coenzyme-

based functional assignments of short-chain dehydrogenases/

reductases (SDRs) family, Chem. Biol. Interact. 143-144 (2003)

271–278.

[25] J. Shanklin, Overexpression and puriﬁcation of the Escherichia

coli inner membrane enzyme acyl acyl carrier protein synthase in

an active form, Protein Express. Purif. 18 (2000) 355–360.

[26] S.N. Ho, H.D. Hunt, R.M. Horton, J.K. Pullen, L.R. Pease, Site

directed mutagenesis by overlap extension using the polymerase

chain reaction, Gene 77 (1989) 51–59.

[27] M.A. Hayat, Basic Electron Microscopy Techniques, Nostrand

Press, New York, 1972.

[28] M.A. Hayat, Basic Techniques for Transmission Electron Mi-

croscopy, Academic Press, Orlando, 1986.

[29] R.C. Austin, D. Singh, P.C.Y. Liaw, H.J. Craig, Visual detection

method for identifying recombinant bacterial colonies, Bio.Tech-

niques 18 (1995) 380–384.

[30] S. Barik, Relationship between opacity of transformed E. coli

colonies and over-expression of the recombinant transcript,

BioTechniques 22 (1997) 112–118.

[31] H. Dong, L. Nilsson, C.G. Kurland, Gratuitous overexpression of

genes in Escherichia coli leads to growth inhibition and ribosome

destruction, J. Bacteriol. 177 (1995) 1497–1504.

[32] T.H. Grossman, E.S. Kawasaki, S.P. Punreddy, M.S. Osburne,

Spontaneous cAMP-dependent derepression of gene expression in

stationary phase plays a role in recombinant expression instability,

Gene 209 (1998) 95–103.

[33] K.C. Kelley, K.J. Huestis, D.A. Austen, C.T. Sanderson, M.A.

Donoghue, S.K. Stickel, E.S. Kawasaki, M.S. Osburne, Regula-

tion of sCD4-183 gene expression from phage-T7-based vectors in

Escherichia coli, Gene 156 (1995) 33–36.

[34] W.R.A. Taylor, A Ôperiodic tableÕ for protein structures, Nature

415 (2002) 657–660.

[35] S.A. Lesley, P. Kuhn, A. Godzik, A.M. Deacon, I. Mathews, A.

Kreusch, G. Spraggon, H.E. Klock, D. McMullan, T. Shin, J.

Vincent, A. Robb, L.S. Brinen, M.D. Miller, T.M. McPhillips,

M.A. Miller, D. Scheibe, J.M. Canaves, C. Guda, L. Jaroszewski,

T.L. Selby, M.A. Elsliger, J. Wooley, S.S. Taylor, K.O. Hodgson,

I.A. Wilson, P.G. Schultz, R.C. Stevens, Structural genomics of

the Thermotoga maritima proteome implemented in a high-

throughput structure determination pipeline, Proc. Natl. Acad.

Sci. USA 99 (2002) 11664–11669.

[36] G. Hannig, S.C. Makrides, Strategies for optimizing heterolo-

gous protein expression in E. coli, Trends Biotechnol. 16 (1998)

54–60.

[37] S.C. Makrides, Strategies for achieving high-level expression of

genes in Escherichia coli, Microbiol. Rev. 60 (1996) 512–538.

S.S. Poletto et al. / Protein Expression and Puriﬁcation 34 (2004) 118–125 125

ANEXO III

Protein Expression and PuriWcation 40 (2005) 23–30

www.elsevier.com/locate/yprep

doi:10.1016/j.pep.2004.06.040

DAHP synthase from Mycobacterium tuberculosis H37Rv: cloning,

expression, and puriW cation of functional enzyme

Caroline Rizzi

, Jeverson Frazzon

, Fernanda Ely

, Patrícia G. Weber

Isabel O. da Fonseca

, Michelle Gallas

, Jaim S. Oliveira

, Maria A. Mendes

Bibiana M. de Souza

, Mário S. Palma

, Diógenes S. Santos

d,¤

, Luiz A. Basso

a,¤

Departamento de Biologia Molecular e Biotecnologia, Universidade Federal do Rio Grande do Sul, Porto Alegre, RS 91501-970, Brazil

Departamento de Ciência dos Alimentos, Universidade Federal do Rio Grande do Sul, Porto Alegre, RS 91501-970, Brazil

Departamento de Biologia/CEIS, Universidade do Estado de São Paulo, Rio Claro, SP 13506-900, Brazil

Centro de Pesquisa e Desenvolvimento em Biologia Molecular e Funcional, PontíWcia Universidade Católica do Rio Grande do Sul,

Porto Alegre, RS 90619-900, Brazil

Received 13 April 2004, and in revised form 18 June 2004

Available online 8 December 2004

Abstract

Tuberculosis (TB), caused by Mycobacterium tuberculosis, remains the leading cause of mortality due to a bacterial pathogen.

According to the 2004 Global TB Control Report of the World Health Organization, there are 300,000 new cases per year of multi-

drug resistant strains (MDR-TB), deWned as resistant to isoniazid and rifampicin, and 79% of MDR-TB cases are now “super

strains,” resistant to at least three of the four main drugs used to treat TB. Thus there is a need for the development of eVective new

agents to treat TB. The shikimate pathway is an attractive target for the development of antimycobacterial agents because it has been

shown to be essential for the viability of M. tuberculosis, but absent from mammals. The M. tuberculosis aroG-encoded 3-deoxy-

D-

arabino-heptulosonate 7-phosphate synthase (mtDAHPS) catalyzes the Wrst committed step in this pathway. Here we describe the

PCR ampliWcation, cloning, and sequencing of aroG structural gene from M. tuberculosis H37Rv. The expression of recombinant

mtDAHPS protein in the soluble form was obtained in Escherichia coli Rosetta-gami (DE3) host cells without IPTG induction. An

approximately threefold puriWcation protocol yielded homogeneous enzyme with a speciWc activity value of 0.47 U mg

¡1

under the

experimental conditions used. Gel Wltration chromatography results demonstrate that recombinant mtDAHPS is a pentamer in solu-

tion. The availability of homogeneous mtDAHPS will allow structural and kinetics studies to be performed aiming at antitubercular

agents development.

Keywords: Mycobacterium tuberculosis; Shikimate pathway; DAHP synthase; Protein expression

Tuberculosis (TB)

remains the leading cause of mor-

tality due to a bacterial pathogen, Mycobacterium tuber-

culosis. The interruption of centuries of decline in case

rates of TB occurred, in most cases, in the late 1980s and

involved the USA and some European countries due to

increased poverty in urban settings and the immigration

Corresponding authors. Fax: +55 51 3166234.

E-mail addresses: diogenes@pucrs.br (D.S. Santos), labasso@dna.cbiot.ufrgs.br (L.A. Basso).

Abbreviations used: TB, tuberculosis; MDR-TB, multidrug-resistant; PEP, phosphoenolpyruvate; E4P, D-erythrose-4-phosphate; DAHPS, 3-de-

oxy-

D-arabino-heptulosonate 7-phosphate synthase; DMSO, dimethyl sulfoxide; LB, Luria–Bertani; SDS–PAGE, sodium dodecyl sulfate–polyacryl-

amide gel electrophoresis; IPTG, isopropyl -

D-thiogalactoside; -ME, -mercaptoethanol; DAHPS(Phe), phenylalanine-regulated DAHPS;

DAHPS(Try), tyrosine-regulated DAHPS; DAHPS(Trp), tryptophan-regulated DAHPS.

24 C. Rizzi et al. / Protein Expression and PuriWcation 40 (2005) 23–30

from TB high-burden countries [1]. Thus, no sustainable

control of TB epidemics can be reached in any country

without properly addressing the global epidemic. It is

estimated that 8.2 million new TB cases occurred world-

wide in the year 2000, with approximately 1.8 million

deaths in the same year, and more than 95% of those

were in developing countries [2]. Approximately, 2 bil-

lion individuals are believed to harbor latent TB based

on tuberculin skin test surveys [3], which represents a

considerable reservoir of bacilli. Possible factors under-

lying the resurgence of TB worldwide include the HIV

epidemic, increase in the homeless population, and

decline in health care structures and national surveil-

lance [4]. Another contributing factor is the evolution of

multi-drug resistant strains (MDR-TB), deWned as resis-

tant to isoniazid and rifampicin, which are the most

eVective Wrst-line drugs [5]. According to the 2004

Global TB Control Report of the World Health Organi-

zation, there are 300,000 new cases per year of MDR-TB

worldwide, and 79% of MDR-TB cases are now “super

strains,” resistant to at least three of the four main drugs

used to treat TB [6]. The factors that most inXuence the

emergence of drug-resistant strains include inappropri-

ate treatment regimens, and patient noncompliance in

completing the prescribed courses of therapy due to the

lengthy standard “short-course” treatment or when the

side eVects become unbearable [7]. Hence, faster acting

and eVective new drugs to better combat TB, including

MDR-TB, are needed.

The shikimate pathway is an attractive target for the

development of herbicides and antimicrobial agents

because it is essential in algae, higher plants, bacteria,

and fungi, but absent from mammals [8]. In mycobacte-

ria, the shikimate pathway leads to the biosynthesis of

chorismic acid, which is a precursor for the synthesis of

aromatic amino acids, naphthoquinones, menaquinones,

and mycobactins [9]. The salicylate-derived mycobactin

siderophores have been shown to be essential for M.

tuberculosis growth in macrophages [10]. More recently,

the shikimate pathway has been shown by disruption of

aroK gene, which codes for the shikimate kinase enzyme,

to be essential for the viability of M. tuberculosis [11].

The absence from the human host and essentiality of

mycobacterial shikimate pathway indicate that any of its

enzymes are promising targets for the development of

potentially non-toxic antimycobacterial agents.

Homologues to enzymes in the shikimate pathway

have been identiWed in the genome sequence of M. tuber-

culosis [12]. The Wrst committed step in the shikimate

pathway is catalyzed by 3-deoxy-

D-arabino-heptuloson-

ate 7-phosphate (DAHP) synthase (DAHPS; EC

4.1.2.15). DAHPS catalyzes the stereospeciWc condensa-

tion of phosphoenolpyruvate (PEP) and

D-erythrose

4-phosphate (E4P), forming DAHP and inorganic phos-

phate [13]. Based on phylogenetic analysis, DAHPS has

been divided into two classes, class I and class II [14].

Escherichia coli expresses three DAHPS isoenzymes that

are representative of class II and require divalent metal

for activity [15], which play a role in catalysis and/or

structural integrity [16]. Each isoenzyme is speciWcally

inhibited by one of the three aromatic amino acids [8].

DAHPS(Phe), a homotetramer encoded by the aroG

gene, is feedback inhibited by phenylalanine; the aroH-

encoded DAHPS(Trp) and aroF-encoded DAHPS(Tyr)

are homodimers feedback inhibited by, respectively,

tryptophan and tyrosine. In M. tuberculosis genome,

however, only the aroG (Rv2178c) encoded DAHPS iso-

enzyme (mtDAHPS) has been proposed to be present by

sequence homology.

To determine the mechanism of action of mtDAHPS

by steady-state and pre-steady-state kinetics as well as

for X-ray crystal structure determination aiming at the

rational design of antimycobacterial agents, expression

of aroG encoded mtDAHPS in functional form and in

large quantity are needed. Accordingly, we here

describe the PCR ampliWcation, cloning, sequencing,

expression in E. coli Rosetta-gami (DE3) cells, puriWca-

tion to homogeneity, oligomeric state determination,

and assay of mtDAHP enzyme activity. Measurements

of enzyme activity conWrm the correct assignment to

the structural gene encoding mtDAHPS in M. tubercu-

losis. The availability of mtDAHPS will allow enzyme

kinetics and structural studies to be undertaken to pro-

vide a framework on which to base the design of new

agents with antitubercular activity with, hopefully, low

toxicity.

Materials and methods

PCR ampliWcation and cloning of M. tuberculosis aroG

gene

The design of synthetic oligonucleotide primers used

for PCR ampliWcation of aroG gene (5Ј-ggacatatgaactgg

accgtcgacatac-3Ј and 5Ј-cggatcctcagtcccgcagcatctccgc-3Ј)

was based on the complete genome sequence of M.

tuberculosis H37Rv [12]. These primers were comple-

mentary to, respectively, the amino-terminal coding and

carboxy-terminal noncoding strands of aroG gene con-

taining 5ЈNdeI and 3ЈBamHI restrictions sites, which are

in bold. This pair of primers was used to amplify the M.

tuberculosis aroG gene (1389 bp) from genomic DNA

using standard PCR conditions and the enzyme Pfu

DNA polymerase (Stratagene), which is a thermostable

polymerase that exhibits low error rate, thus lowering

the likelihood of introducing unwanted mutations. PCR

ampliWcation required the presence of 10% of dimethyl

sulfoxide (DMSO) in the reaction mixture. The PCR

product was puriWed by electrophoresis on low melting

agarose, digested with NdeI and BamHI (Boehringer–

Mannheim), and cloned into pET-23a(+) (Novagen)

C. Rizzi et al. / Protein Expression and PuriWcation 40 (2005) 23–30 25

expression vector, which had previously been digested

with the same restriction enzymes. To both conWrm the

identity of the cloned gene and ensure that no mutations

were introduced by the PCR ampliWcation step, the

DNA sequence of the ampliWed M. tuberculosis aroG

structural gene was determined by dideoxy-chain termi-

nation method [17], using the Thermo Sequenase radio-

labeled terminator cycle sequencing kit (Amersham

Biosciences).

Expression of mtDAHPS

The recombinant plasmid pET-23a(+)::aroG was

transformed into electrocompetent E. coli Rosetta-gami

(DE3) cells (Novagen), and selected on LB agar plates

containing 50 gmL

¡1

carbenicillin, 15 gmL

¡1

kana-

mycin, 34 gmL

¡1

chloramphenicol, and 12.5 gmL

¡1

tetracycline. Single colonies were used to inoculate

500 mL LB medium, containing the same antibiotics and

concentrations of LB solid medium, and grown at 37 °C

and 180 rpm for 24 h, without addition of isopropyl -

D-

thiogalactopyranoside (IPTG). Cells were harvested by

centrifugation at 48,000g for 20 min at 4 °C, and stored

at ¡20 °C. For protein expression analysis, 10 mg of

stored cells was resuspended in 500 L BuVer A (50 mM

Tris–HCl, pH 7.8), disrupted by soniWcation, and cell

debris was removed by centrifugation. Both soluble and

insoluble fractions were analyzed by SDS–PAGE 12%

[18]. Control experiments were performed under the

same experimental conditions except that E. coli host

cells were transformed with the expression vector lack-

ing the target gene.

PuriWcation of recombinant mtDAHPS

Approximately, 36 g of cells was collected by centri-

fugation (48,000g for 20 min) from 6 L of LB medium.

All subsequent steps were performed on ice or at 4 °C.

Frozen cells (36 g) were thawed and resuspended in

BuVer A (4 mL of buVer per gram of cell paste) contain-

ing 1 mM -mercaptoethanol (-ME) (Sigma) and

0.2 mg mL

¡1

of lysozyme, and the mixture was stirred

for 30 min. Cells were disrupted by sonication, and cell

debris was removed by centrifugation (48,000g for

30 min). The supernatant was incubated with 1% w/v of

streptomycin sulfate for 15 min, and centrifuged

(48,000g for 30 min). Solid ammonium sulfate was

added to the supernatant fraction to a concentration of

25% saturation, incubated for 30 min, and centrifuged

as above. The resultant pellet was resuspended in 70 mL

BuVer A containing 1 mM -ME and dialyzed against

three changes of 2 L of the same buVer using a dialysis

tubing with molecular weight cut-oV of 12,000–4000 Da.

The sample was clariWed by centrifugation and loaded

on a Q-Sepharose Fast Flow (2.6 cm £ 8.2 cm) anion

exchange column (Amersham Biosciences) previously

equilibrated with BuVer A and fractionated using a

600 mL 0.0–0.6 M NaCl linear gradient. The fractions

containing mtDAHPS (0.35–0.38 M NaCl) were pooled,

concentrated to 8.0 mL using an Amicon ultraWltration

cell (MW 30,000 Da), and loaded on a Sephacryl S-200

HR (2.6 cm £ 60 cm) gel Wltration column (Amersham

Biosciences) at 0.5 mL min

¡1

. The protein was eluted

with BuVer B (BuVer A containing 200 mM NaCl) at the

same Xowrate. The active fractions were loaded on a

Mono Q HR 16/10 anion exchange column (Amersham

Biosciences) equilibrated with BuVer A and eluted with

400 mL linear 0.0–0.6 M NaCl gradient. The active frac-

tions, which exhibited a single band on SDS–PAGE,

were pooled, quickly frozen in liquid nitrogen, and

stored at ¡80 °C.

Determination of protein concentration

Protein concentrations were determined using the

Bio-Rad Laboratories protein assay kit (Bradford

method) [19] and bovine serum albumin as standard.

Determination of mtDAHPS molecular mass

The molecular mass of native mtDAHPS homoge-

neous protein was determined by gel Wltration chroma-

tography using a Sephacryl S-200 (HR 10/30)

(Amersham Biosciences) equilibrated with BuVer B at a

Xowrate of 0.4 mL min

¡1

. Protein molecular mass stan-

dards were from Gel Filtration LMW and HMW Cali-

bration Kit from Amersham Biosciences. Protein elution

was monitored at 280 nm.

mtDAHPS enzyme assay

Enzyme activity of recombinant mtDAPHS protein

was assayed in the forward direction by a continuous

spectrophotometric method described by Shoner and

Hermann [20], monitoring the decrease in phosphoenol-

pyruvate (PEP) concentration at 232 nm

( D 2.8 £ 10

¡1

) on a Multi-Spec 1501 photodi-

ode array spectrophotometer (Shimadzu). All reactions

were carried out at 25 °C and initiated with addition of

enzyme to a reaction mixture containing: 50mM Tris–

HCl, pH 7.0, 400 M E4P (Sigma), 1 mM -ME, and

200 M PEP (Acrós Organics) in a total volume of

500 L. One unit of enzyme activity (U) is deWned as the

amount of enzyme catalyzing the conversion of 1 mol

PEP/min at 25 °C.

N-terminal amino acid sequencing

The N-terminal amino acid residues of homogeneous

recombinant mtDAHPS were identiWed by automated

Edman degradation sequencing using a PPSQ 21A gas-

phase sequencer (Shimadzu).

26 C. Rizzi et al. / Protein Expression and PuriWcation 40 (2005) 23–30

Mass spectrometry analysis

The homogeneity of recombinant protein preparation

was assessed by mass spectrometry (MS), employing

some adaptations made to the system described by

Chassaigne and Lobinski [21]. Samples were analyzed on

a triple quadrupole mass spectrometer, model QUAT-

TRO II, equipped with standard electrospray (ESI)

probe (Micromass, Altrinchan), adjusted to ca.

250 Lmin

¡1

. The source temperature (80 °C) and needle

voltage (3.6kV) were maintained constant throughout

the experimental data collection, applying a drying gas

Xow (nitrogen) of 200 Lh

¡1

and a nebulizer gas Xow of

20 Lh

¡1

. The mass spectrometer was calibrated with

intact horse heart myoglobin and its typical cone-volt-

age induced fragments. The subunit molecular mass of

recombinant protein mtDAHPS was determined by ESI-

MS, adjusting the mass spectrometer to give a peak with

at half-height of 1 mass unit, and the cone sample to

skimmer lens voltage controlling the ion transfer to mass

analyzer was set to 38 V. About 50 pmol (10 L) of each

sample was injected into the electrospray transport sol-

vent. The ESI spectrum was obtained in the multi-chan-

nel acquisition mode, scanning from 500 to 1800 m/z at a

scan time of 7 s. The mass spectrometer is equipped with

MassLynx and Transform software for data acquisition

and spectra handling.

Results and discussion

The PCR ampliWcation of aroG structural gene from

M. tuberculosis H37Rv genomic DNA required the pres-

ence of 10% DMSO in the reaction mixture (data not

shown). DMSO is a cosolvent that improves GC-rich

DNA denaturation and helps to overcome the diYculties

of polymerase extension through secondary structures,

altering the structural conformation of DNA templates

[22]. This result is consistent with the 65.6% G + C con-

tent of M. tuberculosis H37Rv genome [12]. PCR frag-

ment was inserted into pET23a(+) expression vector [23]

between NdeI and BamHI restriction sites. DNA

sequencing of the entire aroG structural gene by the dide-

oxy chain termination method both conWrmed the iden-

tify of the cloned PCR product and showed that no

mutations were introduced by the DNA ampliWcation

step.

Recombinant plasmids were introduced into E. coli

BL21(DE3) host cells by electroporation. Unfortunately,

recombinant mtDAHPS remained in the insoluble frac-

tion (Fig. 1). Since one of the goals of the present work

was to conWrm the correct assignment to the structural

gene encoding mtDAHPS, eVorts were made to express

recombinant M. tuberculosis DAHPS in its soluble,

active form avoiding unfolding and refolding protocols

because they cannot guarantee that they will yield large

amounts of biologically active product [24]. In addition,

a number of protocols were tested to obtain mtDAHPS

in the soluble fraction to no avail, including buVer addi-

tives (urea, deoxycholic acid, Triton X-100, and high

NaCl concentrations) and reduced cultivation tempera-

ture (20, 25, and 30 °C). In practice, it is usually worth-

while to test several diVerent vector/host combinations

to obtain the best possible yield of protein in its desired

form. Accordingly, a number of commercially available

strains of E. coli host cells were tested in an attempt to

produce mtDAHPS in the soluble fraction. Analysis of

the relationship between codon preference and expres-

sion level led to the classiWcation of E. coli genes into

three main classes [25]. Class II genes, which correspond

to genes highly and continuously expressed during expo-

nential growth that is likely to resemble the tRNA popu-

lation available for recombinant protein expression,

have a number of avoided codons with frequencies of

less than 6%. InsuYcient tRNA pools can lead to prema-

ture translational termination, translation frameshifting

or amino acid misincorporation that might result in

expression of nonproperly folded recombinant protein

[26]. Rare codons near the N-terminus of a coding

sequence can have a severe eVect on heterologous

expression in E. coli [27]. Four rare codons for heterolo-

gous gene expression in E. coli are present near the N-

terminus of M. tuberculosis aroG structural gene

(1 £ AUA for isoleucine, 3 £ CCC for proline). To test

whether these rare codons may have any eVect on aroG

expression, E. coli Rosetta (DE3) strain harboring

tRNA genes for AGG, AGA, AUA, CUA, CCC, and

Fig. 1. SDS–PAGE (12%) of the soluble and insoluble fractions of the

cell extracts of either BL21(DE3) or Rosetta-gami (DE3) host cells

transformed with either pET-23a(+) (control) or pET-23a(+)::aroG.

Expression conditions were 24 h at 37 °C without IPTG addition. Lane

1: insoluble fraction of BL21 (DE3) transformed with pET-23(+); lane

2: insoluble fraction of BL21(DE3) transformed with pET-

23(+)::aroG; lane 3: soluble fraction of BL21(DE3) transformed with

pET-23(+); lane 4: soluble fraction of BL21(DE3) transformed with

pET-23(+)::aroG; lane 5: MW marker “High range” (Gibco-BRL);

lane 6: insoluble fraction of Rosetta-gami (DE3) transformed with

pET-23(+); lane 7: insoluble fraction of Rosetta-gami (DE3) trans-

formed with pET-23(+)::aroG; lane 8: soluble fraction of Rosetta-gami

(DE3) transformed with pET-23(+); and lane 9: soluble fraction of

Rosetta-gami (DE3) transformed with pET-23(+)::aroG. Molecular

mass of mtDAHPS is approximately 50.6 kDa.

C. Rizzi et al. / Protein Expression and PuriWcation 40 (2005) 23–30 27

GGA rare codons on a chloramphenicol-resistant plas-

mid [28] was transformed with pET-23a(+)::aroG recom-

binant plasmid. Disappointingly, recombinant

mtDAHPS remained in the insoluble fraction (data not

shown), thereby discarding any eVect of the mycobacte-

rial aroG rare codons on recombinant protein expres-

sion. Although E. coli DAHPS has no disulWde bridges

despite possessing three cysteine residues, there is no

experimental evidence for the absence of disulWde

bridges in mtDAHPS, which possesses Wve cysteine resi-

dues (Cys 87, Cys 231, Cys 365, Cys 420, and Cys 440),

and a less reducing cytoplasmatic environment could

improve mtDAHPS solubility. The Origami E. coli host

strains (Novagen) have mutations in both the thiore-

doxin reductase (trxB) and glutathione reductase (gor)

genes, which greatly enhances disulWde bond formation

in the cytoplasm [29,30]. Unfortunately, none of the pro-

tocols tested yielded soluble mtDAHPS. The Rosetta-

gami (DE3) E. coli host strain (Novagen) combines the

features of Rosetta and Origami strains. SDS–PAGE

analysis showed that expression of recombinant

mtDAHPS protein in its soluble form with the expected

molecular mass (»51 kDa) could be achieved using the

Rosetta-gami (DE3) cells grown at 37 °C for 24 h with

no IPTG induction (Fig. 1). The underlying reason for

this result is unclear; however, it underscores the need

for optimization of vector/host combinations to achieve

soluble recombinant protein expression before attempt-

ing any unfolding/refolding protocols.

It should be pointed out that a screening of experi-

mental conditions was carried out to obtain high yield of

recombinant protein expression, including temperature

of growth, culture aeration, medium type, hours of

growth after IPTG induction, and hours of growth in the

absence of IPTG. The best results were obtained from

Rosetta-gami (DE3) E. coli cells grown for 24 h at 37 °C

in LB medium without IPTG induction as described

above. In the pET vector system (Novagen), target genes

are positioned downstream of the bacteriophage T7 late

promoter. Typically, production hosts contain a pro-

phage (DE3) encoding the highly processive T7 RNA

polymerase under control of the IPTG-inducible lacUV5

promoter that would ensure tight control of recombi-

nant gene basal expression [31,32]. In agreement with the

results presented here, leaky expression has been shown

to occur in the pET system [33–37]. It has been proposed

that leaky protein expression is a property of the lac-

controlled system as cells approach stationary phase in

complex medium and that cyclic AMP, acetate, and low

pH are required to achieve high-level expression in the

absence of IPTG induction, which may be part of a gen-

eral cellular response to nutrient limitation [38].

Enzyme activity measurements demonstrated that

there was a 92-fold increase in speciWc activity for

mtDAHPS when Rosetta-gami (DE3) E. coli harboring

either pET-23a(+)::aroG or pET-23a(+) crude extracts

were compared (Table 1), indicating that mtDAHPS was

expressed in its soluble and functional form. The puriW-

cation protocol of recombinant mtDAHPS, protein

determination, enzyme assay, and SDS–PAGE analysis

were as described in Materials and methods. Recombi-

nant mtDAHPS enzyme was puriWed approximately 3-

fold (Table 2) to electrophoretic homogeneity (Fig. 2).

The puriWcation protocol yielded approximately 5 mg of

homogeneous protein. A signiWcant loss in protein yield

occurred in the 25% ammonium sulfate precipitation

step (Table 2) because some mtDAHPS remained in the

supernatant. Protein precipitations by higher ammo-

nium sulfate concentrations were also carried out, yield-

ing larger amounts of recombinant mtDAHPS in the

pellet (data not shown). However, a number of contami-

nants co-precipitated with mtDAHPS. In particular, a

contaminant that co-eluted in subsequent chromato-

graphic steps when larger than 25% ammonium sulfate

concentrations were used. Accordingly, the 25% ammo-

nium sulfate precipitation step was deemed more appro-

priate for the puriWcation protocol because a signiWcant

amount of contaminants remained in the supernatant,

while mtDAHPS with a lower protein-contaminating

background remained in the pellet thus making the

Table 1

Measurements of recombinant DAHPS enzyme activity

Crude cell extract in 50 mM Tris–HCl, pH 7.0.

UmL

¡1

/mg mL

¡1

Cell extract

SpeciWc activity

(SA, U mg

¡1

) AS cloned/SA control

Control 0.0018 1

DAHPS 0.1651 92

Table 2

PuriWcation of M. tuberculosis 3-deoxy-

D-arabino-heptulosonate 7-phosphate synthase expressed in E. coli Rosetta-gami(DE3) transformed with

pET-23a(+)::aroG

Typical puriWcation protocol starting from 36 g wet weight cells obtained from 6 L of culture.

UmL

¡1

/mg mL

¡1

PuriWcation step Total protein (mg) Total enzyme activity (U) SpeciWc activity

(U mg

¡1

)PuriWcation fold Yield (%)

Crude extract 2442.38 403.23 0.17 1.00 100

Ammonium sulfate 136.78 44.11 0.32 1.95 11

Q-Sepharose Fast Flow 21.42 7.00 0.33 1.98 2

Sephacryl S-200 HR 17.83 4.48 0.25 1.52 1

Mono Q HR 16/10 4.77 2.24 0.47 2.84 0.6

28 C. Rizzi et al. / Protein Expression and PuriWcation 40 (2005) 23–30

subsequent puriWcation steps less demanding. The

samples of each puriWcation step were assayed for

DAHPS enzyme activity and compared to control exper-

iments to demonstrate that the observed values are

actual velocities of mtDAHPS activity. Enzyme activity

of homogeneous mtDAHPS was linearly dependent on

sample volume added to the reaction mixture (Fig. 3),

thereby showing that the initial velocity is proportional

to total enzyme concentration and that true initial veloc-

ities are being measured. E. coli DAHPS(Phe) has been

shown to be sensitive to oxidation, leading to inactiva-

tion [39]. Moreover, higher mtDAHPS enzyme activity

values could be observed in the presence of -ME (data

not shown). Accordingly, -ME was used in all steps of

the mtDAHPS puriWcation protocol. The mtDAHPS

speciWc activity has been found to be stable for at least

two months when stored at ¡80 °C.

The subunit molecular mass of active mtDAHPS was

determined to be 50510.38 Da by electrospray ionization

mass spectrometry (ESI-MS), consistent with the post-

translational removal of the N-terminal methionine resi-

due from the full length gene product (predicted mass:

50641.51 Da). The ESI-MS result also revealed no peak

at the expected mass for the three isoforms of E. coli

DAHPS enzymes (38804.03, 38735.18, and 38009.53 Da),

thus providing evidence for both the identity and purity

of the recombinant protein. The Wrst 11 N-terminal

amino acid residues of the recombinant protein were

identiWed as NWTVDIPIDQL by the Edman degrada-

tion chemistry protocol. This result unambiguously iden-

tiWes the homogeneous recombinant protein as

mtDAHPS and conWrms removal of the N-terminal

methionine residue from it. A common type of co-/post-

translational modiWcation of proteins synthesized in

prokaryotic cells is modiWcation at their N-termini.

Methionine aminopeptidase catalyzed cleavage of initia-

tor methionine is usually directed by the penultimate

amino acid residues with the smallest side chain radii of

gyration (glycine, alanine, serine, threonine, proline,

valine, and cysteine) [40]. The N-terminal methionine was

removed from the E. coli expressed recombinant

mtDAHPS enzyme, consistent with the Wnding that some

middle-sized penultimate amino acid residues (Asn, Asp,

Leu, and Ile) undergo N-terminal processing [41].

A value of 253 § 25 kDa was determined for the

molecular mass of native mtDAHPS homogeneous pro-

tein by analytical gel Wltration chromatography (data

not shown), suggesting that mtDAHPS is a pentamer in

solution. Whereas E. coli DAHPS(Phe) is a tetramer [42]

and E. coli DAHPS(Trp) is a dimer [43]. Interestingly,

more recently, recombinant DAHPS from Pyrococcus

furiosus has been shown to be a dimer in solution and

not to be inhibited by phenylalanine, tyrosine, or trypto-

phan [44].

DAHPS in most, not all, microorganisms is the target

for pathway regulation by negative feedback inhibition,

which controls carbon Xow into the shikimate pathway.

The most intensively investigated microorganism

DAHPS has been the E. coli enzyme, which possesses

three isoenzymes, each speciWcally regulated by one of

three aromatic amino acid end products, either Phe, Tyr,

or Trp [45]. The three isoforms have a common require-

ment for a metal cofactor, which can be similarly satis-

Wed by a range of divalent metal ions [46]. DAHPS

enzymes from a number of microorganisms have been

studied, such as Corynebacterium glutamicum [47], Ther-

motoga maritima [48], Bacillus subtilis [49], Saccharomy-

ces cerevisiae [50], and P. furiosus [44]. However, to the

best of our knowledge, this is the Wrst report on cloning,

expression, and puriWcation of functional DAHPS from

M. tuberculosis.

Fig. 2. SDS–PAGE analysis of pooled fractions from the puriWcation

steps of mtDAHPS. Lane 1, MW marker “High range” (Gibco-BRL);

lane 2, crude extract; lane 3, ammonium sulfate precipitation; lane 4,

Q-Sepharose Fast Flow ion exchange; lane 5, S-200 gel Wltration; and

lane 6, Mono Q ion exchange.

Fig. 3. Linear dependence of mtDAHPS activity on homogeneous pro-

tein volume. The rates of enzyme activity were performed in the for-

ward direction by continuously monitoring the decrease of

phosphoenolpyruvic acid at 232 nm.

C. Rizzi et al. / Protein Expression and PuriWcation 40 (2005) 23–30 29

Homogeneous mtDAHPS protein will provide pro-

tein in quantities necessary for studies on the enzyme

mechanism of action by steady-state and pre-steady-

state kinetics, its metal requirement, if any, and feedback

inhibition by Phe, Tyr, and Trp. The expression and

puriWcation of mtDAHPS reported here will also provide

protein for crystallization trials aiming at three-dimen-

sional structure determination by X-ray diVraction. The

three-dimensional structures of four forms of the Phe-

regulated isoenzyme of E. coli DAHPS have been solved

by X-ray crystallography [42,51–53], which should facili-

tate screening of experimental conditions to obtain crys-

tals of mtDAHPS in complex with its substrates,

possible metal cofactor and feedback inhibitors, if any.

Expression of functional proteins in soluble form has

been identiWed as an important bottleneck in eVorts to

determine biological activity and crystal structure of M.

tuberculosis proteins [54]. We hope that the results

reported here will contribute to eVorts towards the struc-

ture determination of potential targets in M. tuberculo-

sis. The enzymological and structural studies on

mtDAHPS should help in the design of enzyme inhibi-

tors to be tested as antimycobacterial agents.

Acknowledgments

Financial support for this work was provided by Mil-

lennium Initiative Program MCT-CNPq, Ministry of

Health-Department of Science and Technology-UNE-

SCO (Brazil) to D.S.S. and L.A.B. D.S.S. and L.A.B. also

acknowledge grants awarded by PADCT, CNPq, and

FINEP. L.A.B. (CNPq, 520182/99-5), D.S.S. (CNPq,

304051/1975-06), and M.S.P. (CNPq, 300337/2003-50)

are researchers awardees from the National Council for

ScientiWc and Technological Development of Brazil.

M.A.M. is a post-doctoral fellow from FAPESP (01/

05060-4).

References

[1] M.C. Raviglione, The TB epidemic from 1992 to 2002, Tuberculo-

sis 83 (2003) 4–14.

[2] E.L Corbett., C.J. Watt, N. Walker, D. Maher, B.G. Williams, M.C.

Raviglione, C. Dye, The growing burden of tuberculosis: global

trends and interactions with the HIV epidemic, Arch. Intern. Med.

1639 (2003) 1009–1021.

[3] C. Dye, S. Scheele, P. Dolin, V. Pathania, M.C. Raviglione, Global

burden of tuberculosis: estimated incidence, prevalence, and mor-

tality by country, JAMA 282 (1999) 677–686.

[4] B.R. Bloom, C.J.L. Murray, Tuberculosis: commentary on a

reemergent killer, Science 257 (1998) 1055–1064.

[5] L.A. Basso, J.S. Blanchard, Resistance to antitubercular drugs,

Adv. Exp. Med. Biol. 456 (1998) 115–144.

[6] World Health Organization, Anti-tuberculosis drug resistance in

the world: Third Global Report, 2004.

[7] K. Duncan, Progress in TB drug development and what is still

needed, Tuberculosis 83 (2003) 201–207.

[8] R. Bentley, The shikimate pathway—a metabolic tree with many

branches, Crit. Rev. Biochem. Mol. Biol. 25 (1990) 307–384.

[9] C. Ratledge, Nutrition, growth and metabolism, in: C. Ratledge,

J.L. Stanford (Eds.), The Biology of the Mycobacteria, vol. 1, Aca-

demic Press, London, 1982, pp. 185–271.

[10] J.J. de Voos, K. Rutter, B.G. Schroder, H. Su, Y. Zhu, C.E. Barry

III, The salicylate-derived mycobactin siderophores of Mycobac-

terium tuberculosis are essential for growth in macrophages, Proc.

Natl. Acad. Sci. USA 97 (2000) 1252–1257.

[11] T. Parish, N.G. Stoker, The common aromatic amino acid biosyn-

thesis pathway is essential in Mycobacterium tuberculosis, Micro-

biology 148 (2002) 3069–3077.

[12] S.T. Cole, R. Brosch, J. Parkhill, T. Garnier, C. Churcher, D. Har-

ris, S.V. Gordon, K. Eiglmeier, S. Gas, C.E. Barry III, F. Tekaia, K.

Badcock, D. Basham, D. Brown, T. Chillingworth, R. Connor, R.

Davies, K. Devlin, T. Feltwell, S. Gentles, N. Hamlin, S. Holroyd,

T. Hornsby, K. Jagels, B.G. Barrell, Deciphering the biology of

Mycobacterium tuberculosis from the complete genome sequence,

Nature 393 (1998) 537–544.

[13] P.R. Srinivasan, D.B. Sprinson, 2-Keto-3-deoxy-

D-arabo-heptonic

acid 7-phosphate synthase, J. Biol. Chem. 234 (1959) 716–722.

[14] M.R. Birck, R.W. Woodard, Aquifex aeolicus 3-deoxy-

D-manno-2-

octulosonic acid 8-phosphate synthase: a new class of KDO 8-P

synthase?, J. Mol. Evol. 52 (2001) 205–214.

[15] C.M. Stephens, R. Bauerle, Analysis of the metal requirement of 3-

deoxy-

D-arabino-heptulosonate-7-phosphate synthase from Esch-

erichia coli, J. Biol. Chem. 266 (1991) 20810–20817.

[16] P.A. Jordan, S. Bohle, C. Ramilo, J.N.S. Evans, New insights into

the metal center of 3-deoxy-

D-arabino-heptulosonate 7-phosphate

synthase, Biochemistry 40 (2001) 8387–8396.

[17] F. Sanger, S. Nicklen, R. Coulson, DNA sequencing with chain-ter-

minating inhibitors, Pro. Natl. Acad. Sci. USA 74 (1977) 5436–5467.

[18] U.K. Laemmli, Cleavage of structural proteins during the assem-

bly of the head of bacteriophage T4, Nature 227 (1970) 680–685.

[19] M.M. Bradford, R.A. McRorie, W.L. Williams, A rapid and sensi-

tive method for the quantiWcation of microgram quantities of pro-

tein utilizing the principle of protein-dye binding, Anal. Biochem.

72 (1976) 248–254.

[20] R. Shoner, K.M. Hermann, 3-Deoxy-

D-arabino-heptulosonate 7-

phosphate synthase, J. Biol. Chem. 252 (1976) 5440–5447.

[21] H. Chassaigne, R. Lobinski, Characterization of horse kidney

metallothionein isoforms by electrospray MS and reversed-phase

HPLC-electrospray MS, Analyst 123 (1998) 2125–2130.

[22] D. Pomp, J.F. Medrano, Organic solvents as facilitators of poly-

merase chain reaction, Biotechniques 10 (1991) 58–59.

[23] F.W. Studier, A.H. Rosenberg, J.J. Dunn, J.W. DubendorV, Use of

T7 RNA polymerase to direct expression of cloned genes, Meth-

ods Enzymol. 185 (1990) 60–89.

[24] C.H. Schein, Production of soluble recombinant proteins in bacte-

ria, Bio/technology 7 (1989) 1141–1149.

[25] A. Hénaut, A. Danchin, Analysis and predictions from Escherichia

coli sequences, or E. coli in silico, in: F.C. Neidhardt (Ed.), Esche-

chichia coli and Salmonella: Cellular and Molecular Biology, ASM

Press, Washington, DC, 1996, pp. 2047–2066.

[26] C. Kurland, J. Gallant, Error of heterologous protein expression,

Curr. Opin. Biotechnol. 7 (1996) 489–493.

[27] Y. Nakamura, T. Gojobori, T. Ikemura, Codon usage tabulated

from international DNA sequence databases: status for the year

2000, Nucleic Acids Res. 28 (2000) 292.

[28] R. Novy, D. Drott, K. Yaeger, R. Mierendorf, Overcoming the

codons bias of Escherichia coli for enhanced protein expression,

Innovations 12 (2001) 1–3.

[29] W.A. Prinz, F. Aslund, A. Holmgren, J. Beckwith, The role of the

thioredoxin and glutaredoxin pathways in redox-facilitate

attempts at crystallizing and elucidating the protein disulWde

bonds in the Escherichia coli cytoplasm, J. Biol. Chem. 272 (1997)

15661–15667.

30 C. Rizzi et al. / Protein Expression and PuriWcation 40 (2005) 23–30

[30] F. Baneyx, Recombinant protein expression in Escherichia coli,

Curr. Opin. Biotechnol. 10 (1999) 411–421.

[31] F.W. Studier, B.A. MoVatt, Use of bacteriophage T7 RNA poly-

merase to direct selective high-level expression of cloned genes, J.

Mol. Biol. 189 (1986) 113–130.

[32] J.W. DubendorV, F.W. Studier, Controlling basal expression in an

inducible T7 expression system by blocking the target T7 pro-

moter with lac repressor, J. Mol. Biol. 219 (1991) 45–59.

[33] L.A. Basso, D.S. Santos, W. Shi, R.H. Furneaux, P.C. Tyler, V.L.

Schramm, J.S. Blanchard, Purine nucleoside phosphorylase from

Mycobacterium tuberculosis. Analysis of inhibition by a transition-

state analogue and dissection by parts, Biochemistry 40 (2001)

8196–8203.

[34] J.S. Oliveira, C.A. Pinto, L.A. Basso, D.S. Santos, Cloning and

overexpression in soluble form of functional shikimate kinase and

5-enoylpyruvylshikimate 3-phosphate synthase enzymes from

Mycobacterium tuberculosis, Protein Expr. Purif. 3 (2001) 430–435.

[35] M.L.B. Magalhães, C.P. Pereira, L.A. Basso, D.S. Santos, Cloning

and expression of functional shikimate dehydrogenase (EC

1.1.1.25) from Mycobacterium tuberculosis H37Rv, Protein Expr.

Purif. 26 (2002) 59–64.

[36] R.G. Silva, L.P.S. Carvalho, J.S. Oliveira, C.A. Pinto, M.A. Men-

des, M.S. Palma, L.A. Basso, D.S. Santos, Cloning, overexpression

and puriWcation of functional human purine nucleoside phosphor-

ylase, Protein Expr. Purif. 27 (2003) 158–164.

[37] K.C. Kelley, K.J. Huestis, D.A. Austen, C.T. Sanderson, M.A.

Donoghue, S.K. Stikel, E.S. Kawasaki, M.S. Osburne, Regulation

of sCD4-183 gene expression from phage T7 based vectors in

Escherichia coli, Gene 156 (1995) 33–36.

[38] T.H. Grossman, E.S. Kawaski, S.R. Punreddy, M.S. Osburne,

Spontaneous cAMP-dependent derepression of gene expression in

stationary phase plays a role in recombinant expression instabil-

ity, Gene 209 (1998) 95–103.

[39] R. Bauerle, O.K. Park, Metal-catalyzed oxidation of phenylalanine-

sentitive 3-deoxy-

D-arabino-heptulosonate-7-phosphate synthase

from Escherichia coli: inactivation and destabilization by oxidation

of active-site cysteines, J. Bacteriol. 181 (1999) 1636–1642.

[40] W.T. Lowther, B.W. Matthews, Structure and function of the methio-

nine aminopeptidases, Biochim. Biophys. Acta 1477 (2000) 157–167.

[41] P.-H. Hirel, J.-M. Schmitter, P. Dessen, G. Fayat, S. Blanquet,

Extent of N-terminal methionine excision from Escherichia coli

proteins is governed by the side-chain length of the penultimate

amino acid, Proc. Natl. Acad. Sci. USA 86 (1989) 8247–8251.

[42] I.A. Shumilin, R.H. Kretsinger, R. Bauerle, Crystal structure of phen-

ylalanine-regulated 3-deoxy-

D-arabino-heptulosonate-7-phosphate

synthase from Escherichia coli, Struct. Folding Des. 7 (1999) 865–875.

[43] J.M. Ray, R. Bauerle, PuriWcation and properties of trypto-

phan-sentitive 3-deoxy-

D-arabino-heptulosonate-7-phosphate

synthase from Escherichia coli, J. Bacteriol. 173 (1991) 1894–

1901.

[44] L.R. SchoWeld, M.L. Patchett, E.J. Parker, Expression, puriWca-

tion, and characterization of 3-deoxy-

D-arabino-heptulosonate 7-

phosphate synthase from Pyrococcus furiosus, Protein Expr. Purif.

34 (2004) 17–27.

[45] K.M. Herrmann, The shikimate pathway as an entry to aromatic

secondary metabolism, Plant Physiol. 107 (1995) 7–12.

[46] C.M. Stephens, R. Bauerle, Analysis of the metal requirement of 3-

deoxy-

D-arabino-heptulosonate-7-phosphate synthase from Esch-

erichia coli, J. Biol. Chem. 266 (1991) 20810–20817.

[47] H.F. Liao, L.L. Lin, H.R. Chin, W.H. Hsu, Serine 187 is a crucial

residue for allosteric regulation of Corynebacterium glutamicum 3-

deoxy-

D-arabino-heptulosonate-7-phosphate, FEMS Microbiol.

Lett. 184 (2001) 59–64.

[48] J. Wu, D.L. Howe, R.W. Woodard, Thermotoga maritima 3-deoxy-

D-arabino-heptulosonate-7-phosphate synthase: the ancestral

eubacterial DAHP synthase?, J. Biol. Chem. 278 (2003) 27525–

27531.

[49] R.A. Jensen, E.W. Nester, Regulatory enzymes of aromatic amino

acid biosynthesis in Bacillus subtilis. I. PuriWcation and properties

of 3-deoxy-

D-arabino-heptulosonate 7-phosphate synthetase, J.

Biol. Chem. 241 (1966) 3365–3372.

[50] G. Paravicini, T. Schmidheini, G. Braus, PuriWcation and proper-

ties of 3-deoxy-

D-arabino-heptulosonate 7-phosphate synthase

(phenylalanine-inhibitable) of Saccharomyces cerevisiae, Eur. J.

Biochem. 186 (1989) 361–366.

[51] T. Wagner, I.A. Shumilin, R. Bauerle, R.H. Kretsinger, Structure

of 3-deoxy-

D-arabino-heptulosonate-7-phosphate synthase from

Escherichia coli: comparison of the Mn

2-phosphoglycolate and

the Pb

2-phosphoenolpyruvate complexes and implications for

catalysis, J. Mol. Biol. 301 (2000) 389–399.

[52] I.A. Shumilin, C. Zhao, R. Bauerle, R.H. Kretsinger, Allosteric

inhibition of 3-deoxy-

D-arabino-heptulosonate-7-phosphate syn-

thase alters the coordination of both substrates, J. Mol. Biol. 320

(2002) 1147–1156.

[53] I.A. Shumilin, R. Bauerle, R.H. Kretsinger, The high-resolution

structure of 3-deoxy-

D-arabino-heptulosonate-7-phosphate syn-

thase reveals a twist in plane of bound phosphoenolpyruvate, Bio-

chemistry 42 (2003) 3766–3776.

[54] R. Vicentelli, C. Bignon, A. Gruez, S. Canaan, G. Sulzenbacher, M.

Tegoni, V. Campanacci, C. Cambillau, Medium-scale structural

genomics: strategies for protein expression and crystallization,

Acc. Chem. Res. 36 (2003) 165–172.

ANEXO IV

ANEXO V

Protein Expression and PuriWcation 47 (2006) 614–620

www.elsevier.com/locate/yprep

doi:10.1016/j.pep.2006.02.012

Expression, puriWcation, and circular dichroism

analysis of human CDK9

Andreia Machado Leopoldino

, Fernanda Canduri

, Hamilton Cabral

, Magno Junqueira

Alessandra Bernadete Trovó de Marqui

, Luciano H. Apponi

, Isabel Osório da Fonseca

Gilberto Barbosa Domont

, Diógenes S. Santos

, Sandro Valentini

Gustavo Orlando Bonilla-Rodriguez

, Marcelo Andrés Fossey

Walter Filgueira de Azevedo Jr.

, Eloiza Helena Tajara

a,¤

Faculdade de Medicina de São José do Rio Preto, SJRP, SP, Brazil

Universidade Federal do Mato Grosso do Sul, Campo Grande, MS, Brazil

Universidade Estadual Paulista, SJRP, SP, Brazil

Rede Proteômica-Rio, Universidade Federal do Rio de Janeiro, RJ, Brazil

Faculdade de Ciências Farmacêuticas, UNESP, Araraquara, SP, Brazil

Centro de Pesquisas em Biologia Molecular e Funcional/PUCRS, Porto Alegre, RS, Brazil

Faculdade de Biociências-PUCRS, Porto Alegre, RS, Brazil

Received 3 November 2005, and in revised form 2 February 2006

Available online 10 March 2006

Abstract

The human cyclin-dependent kinase 9 (CDK9) protein was expressed in E. coli BL21 using the pET23a vector at 30 °C. Several milli-

grams of protein were puriWed from soluble fraction using ionic exchange and ATP-aYnity chromatography. The structural quality of

recombinant CDK9 and the estimation of its secondary structure were obtained by circular dichroism. Structural models of CDK9 pre-

sented 26% of helices in agreement with the spectra by circular dichroism analysis. This is the Wrst report on human CDK9 expression in

Escherichia coli and structure analysis and provides the Wrst step for the development of CDK9 inhibitors.

Keywords: CDK9; Cancer; AIDS; Structure; Dichroism analysis; Molecular modeling; Expression

The cyclin-dependent kinase 9 (CDK9)

is a cdc2-related

kinase protein ubiquitously expressed in most cells [1] and

involved in many physiological processes, including cell

diVerentiation and apoptosis [2,3]. In contrast to other

CDKs with cell cycle regulatory functions, this serine–thre-

onine kinase exhibits protein levels unchanged in human

cells entering and progressing through the cell cycle [4].

Originally named PITALRE for its PSTAIRE-like

sequence [5,6], CDK9 is involved in transcriptional events

[7] and is regulated by the regulatory subunit, cyclin T1,

cyclin T2 or cyclin K [8,9]. Recently, a novel isoform of

CDK9 of 55 kDa [10], which originates from an alternative

upstream promoter was identiWed.

CDK9/cyclin T belongs to the multiprotein complex

P-TEFb, an elongation factor for RNA polymerase II-

directed transcription, and is responsible for the phosphor-

ylation of the C-terminal domain of the largest subunit of

RNA polymerase II [11]. It has been found that CDK9/

cyclin T1 is also required for the viral transactivator called

Tat to stimulate the processivity of RNA pol II in HIV,

Corresponding author. Fax: +55 17 3201 5790.

E-mail address: tajara@famerp.br (E.H. Tajara).

Abbreviations used: CDK9, cyclin-dependent kinase 9; LB, Luria–Ber-

tani medium; MALDI-TOF, matrix-assisted desorption ionization time-

of-Xight; CD, circular dichroism.

A.M. Leopoldino et al. / Protein Expression and PuriWcation 47 (2006) 614–620 615

suggesting a possible involvement of this kinase in AIDS

(for references, see [12]). Ammosova et al. [13] observed

dephosphorylation of CDK9 prior to its association with

HIV-1 transcription initiation complex, which might be

important for regulating HIV-1 transcription.

Recent data have indicated that Cdk9 is a critical deter-

minant of cardiac hypertrophy, in vitro and in vivo. Sano

and Schneider [14] observed that trophic signals for

increasing cardiac mass activated Cdk9, suggesting a role in

heart disease.

The CDK inhibitor Seliciclib (CYC202, R-roscovitine)

competes for the ATP binding site on the kinase. It has

greatest activity against CDK2/cyclin E, CDK7/cyclin H,

and CDK9/cyclin T. Seliciclib induces apoptosis in tumor

cell lines, reduces tumor growth in xenografts in nude mice

and is currently in phase II clinical trials for lung cancer

and B cell malignancies [15].

The functional aspects of human CDK9 have been stud-

ied by many groups but no conformational data is known.

Since the CDK9 protein is a potential therapeutic target in

AIDS, heart disease, and cancer, data on its 3D structure

becomes an essential step in the design of new and speciWc

inhibitors which could improve the treatment of patients.

In this work, we expressed and puriWed recombinant

human CDK9 in Escherichia coli (BL21) suitable for future

structural analysis. The secondary structural data obtained

using circular dichroism was in agreement with the confor-

mation obtained by homology modeling.

Materials and methods

PCR ampliWcation and cloning of human cdk9 cDNA

For RT-PCR, the RNA total was isolated from human

tissue using Trizol reagent (Invitrogen). To obtain the full-

length cdk9 cDNA, RT-PCR was carried out using the oli-

gonucleotides primers forward CgCATATggCAAAG

CAGTACGACTCGGTG and reverse gCggATCCTCAg

AAgACgCgCTCAAACTCC designed based on the cDNA

sequence from GenBank Accession No. NM_001261. The

5Ј NdeI and 3Ј BamHI restriction sites are shown in italic.

The start and stop codons are shown in bold and were

designed to be complementary to 24 and 22 bases of the 5Ј

and 3Ј ends of the CDK9 cDNA, respectively. The cdk9

cDNA (1119 bp) was ampliWed using standard PCR condi-

tions with Pfu DNA polymerase (Stratagene). To introduce

the 5Ј NdeI and 3Ј BamHI sites the ampliWed fragments

were digested with NdeI and BamHI, puriWed using the

Qiagen Gel PuriWcation, and inserted into a pET23a(+)

expression vector (Novagen), previously digested with the

same enzymes. The correct ORF was conWrmed by restric-

tion analysis and sequencing.

Expression of recombinant human CDK9

E. coli

BL21(DE3)pLys cells were transformed with the

pET23a(+)::cdk9 and recombinant colonies were selected

on LB agar plates containing carbenicillin (50 gmL

¡1

) and

chloramphenicol (50gmL

¡1

). Following transformation,

several recombinant colonies were used to inoculate 5 mL

LB medium containing carbenicillin (50 gmL

¡1

) and

chloramphenicol (50 gmL

¡1

). To determine the best con-

dition of human CDK9 protein expression, E. coli

BL21(DE3)pLys cells harboring pET23a(+)::cdk9 were

grown in the presence of IPTG (Wnal concentration of 1,

0.5, and 0.1 mM) at four diVerent temperatures (37, 30, 28,

and 20 °C) at 180 rpm after OD

600

measurements (Abs 0.4–

0.6) [16,17]. One sample was removed at each hour after

induction. The cells were harvested by centrifugation,

stored at ¡80 °C or resuspended in hypotonic lysis buVer

(10 mM Tris–HCl, pH 7.4, 25 mM NaCl, and 1 mM

EDTA), and disrupted by sonication (three short bursts,

about 10 s each, allowing the bacterial suspension to cool

on ice between each burst). The proteins and cell debris

were separated by centrifugation at 20,800g for 30 min at

4 °C. The soluble and insoluble fractions were analyzed by

SDS–PAGE.

PuriWcation of recombinant human CDK9

Preculture (20 mL) of one single colony E. coli (DE3)pLys

containing the recombinant plasmid was diluted in 500 mL

of Luria–Bertani medium (LB) supplemented with appropri-

ate antibiotics [carbenicillin (50gmL

¡1

) and chlorampheni-

col (50gmL

¡1

)]. The culture was conducted at 30 °C at

180 rpm in a shaking incubator until the cells reached mid-

log growth (OD

600

measurements of 0.4–0.6). At this point,

the expression of the target protein was induced by adding

IPTG (0.1 mM) and continued incubation at 30°C for 3 h.

The cells were harvested by centrifugation and resuspended

in 30 mL lysis buVer (10 mM Tris–HCl, pH 7.4, 25mM NaCl,

1 mM EDTA, and 1 mM PMSF). After sonication (three

short bursts, about 30 s each, allowing the bacterial suspen-

sion to cool on ice between each burst), the lysate was clari-

Wed by centrifugation for 1 h at 20,800g and 4 °C. The

supernatant was loaded over a DEAE Sepharose column

(2.5 £ 9 cm) pre-equilibrated with lysis buVer. The absorbed

proteins were eluted with a linear gradient (25–500 mM

NaCl, elution volume 60mL) by the use of a peristaltic pump

at 60 mL/h. After, the fractions were visualized in silver-

stained SDS gels, the selected CDK9 fractions were pooled

and dialyzed against a buVer containing 10 mM Hepes, pH

7.4, 25mM NaCl, and 1mM EDTA. The dialyzed CDK9

were loaded onto an ATP aYnity column (Sigma; 1 £ 2.5 cm;

2 mL) pre-equilibrated with buVer A [10mM Hepes, pH 7.4,

25 mM NaCl, 1 mM EDTA, 10% glycerol (v/v), and 0.5 mM

dithiothreitol]. After washing, bound proteins were eluted

with a 50mL linear salt gradient (25–500 mM NaCl in buVer

A). Fractions containing CDK9 were pooled and concen-

trated (up to approximately 6 mg/mL) and dialyzed using an

Amicon ultraWltration cell (MWC 10,000 Da) against 10 mM

Hepes, pH 7.4, and 1 mM EDTA. The protein content was

analyzed by SDS–PAGE and visualized using the Silver

Staining kit, Protein (GE HealthCare).

616 A.M. Leopoldino et al. / Protein Expression and PuriWcation 47 (2006) 614–620

Protein determination

The protein concentration was carried out as proposed

by Bradford [18] using bovine serum albumin as standard,

and after puriWcation by direct absorbance readings at

280 nm.

Bidimensional electrophoresis

Bidimensional electrophoresis (2DE) [19] was performed

with the IPGphor IEF system and Ruby electrophoresis

unit (GE HealthCare). After sonication, the protein sample

(100 g) was diluted in buVer containing 8 M urea, 2%

Chaps (3-[(3-cholamidopropyl)dimethylammonio]-1-pro-

panesulfonate), 0.3% DTT (dithiothreitol), 0.5% IPG buVer,

pH 3–10, bromophenol blue trace and loaded into immobi-

lized pH gradient (IPG) gel strips. Isoelectric focusing was

run at 20 °C with 8000 V for until 16,500 Vh. The IPG strips

were subsequently either stored at ¡80 °C or equilibrated

for 15 min each at room temperature in SDS equilibration

buVer (50 mM Tris–HCl, pH 8.8, 6 M urea, 2% w/v SDS,

30% glycerol, and few grains of bromophenol blue) with 1%

w/v DTT. The strips were again incubated in SDS equili-

bration buVer with 2.5% w/v iodoacetamide instead of

DTT. The strips were sealed in 12.5% SDS–PAGE and elec-

trophoresis was conducted at 30 mA/gel for 5 h. The gels

were Wxed overnight in 50% ethanol and 10% acetic acid,

washed in destain solution (50% ethanol and 5% acetic

acid) for 3 min and stained in 0.05% Coomassie blue solu-

tion in 40% methanol and 10% acetic acid for 90 min. The

gels were soaked in destain solution for 15, 45, 120, and

120 min, respectively, and preserved in 5% acetic acid. The

gels were incubated in 10% methanol for 48 h and storage

in 50% glycerol until analysis. The 2D images were scanned

by using the ImageScanner (GE HealthCare) and analyzed

by using IMAGEMASTER 2D PLATINUM software

(GE HealthCare). The intensity of each spot was Wrst pro-

cessed by background subtraction and then normalized

between gels as a proportion of the total intensity protein

from the gel. 2DE was chosen to CDK9 validation because

it separated the proteins by charge and molecular weight

(two biochemical characteristics). Therefore, it is more

eYcient for proteomics analysis and mass spectrometry

than SDS–PAGE.

MS and protein identiWcation

The matrix-assisted desorption ionization time-of-Xight

(MALDI-TOF) mass spectrometry (MS) was used to con-

Wrm the identity of the recombinant CDK9. The protein

spot of interest identiWed by 2DE absent in control, was cut

from the gel for subsequent tryptic digestion. For MALDI-

TOF MS analysis, extracted peptides were resuspended

with 0.1% triXuoroacetic acid in 50% acetonitrile, mixed

with 1 l of matrix (10 mg/ml -cyano-4-hydroxycinnamic

acid) and 1 l of this mixture was pipetted in the MALDI

plate and allowed to dry. Peptide maps were obtained using

a Voyager–DE PRO MALDI-TOF mass spectrometer

(Applied Biosystem) in positive ion reXectron mode. Pep-

tide fragment mass values were searched against the data-

base by using the MS-Fit program (http://

prospector.ucsf.edu) with the minimum parameters: Swiss-

Prot database, match with three peptides, mass tolerance of

50 ppm, and 15% coverage percentage.

Circular dichroism

Circular dichroism (CD) spectra in the UV–visible were

recorded on a Jasco J710 dichrograph operating at room

temperature, range 250–190 nm, interfaced with a PC. The

spectrum was analyzed through the self-consistent method

published by Sreerama and Wood [20], and an interactive

graphic program for calculation the secondary structures

by Deléage and Geourjon [21]. Measurements in the far UV

were carried out on 25 m cuvettes and the proteins were

dissolved in 10 mM Hepes, pH 7.4, and 1 mM EDTA.

Molecular modeling

Homology modeling is usually the method of choice

when there is a clear relationship of homology between the

sequence of a target protein and at least one known struc-

ture. Model building of the CDK9 were carried out using

the program Parmodel [22], which is a web server for auto-

mated modeling and protein structural assessment. Par-

model runs a parallelized version of MODELLER [23].

MODELLER is an implementation of an automated

approach to comparative modeling by satisfaction of spa-

tial restraints. The modeling procedure begins with align-

ment of the sequence to be modeled (target) with the

sequence of related known three-dimensional structures

(templates). This alignment is usually the input to the

program. The output is a three-dimensional model for the

target sequence containing all mainchain and sidechain

non-hydrogen atoms [23].

Two models were built, the Wrst in the absence of

ligands, and the second in the presence of ATP. The models

of CDK9 were based on the atomic coordinates of CDK2

apoenzyme (PDB code: 1HCL) [24] and CDK2:ATP (PDB

code: 1HCK) [25].

Several slightly diVerent models can be calculated by

varying the initial structure. A total of 1000 models for each

modeling were generated by the program Parmodel [22],

the Wnal models were selected based on stereochemical

quality. All optimization process was performed on a Beo-

wulf cluster with 16 nodes (BioComp, AMD Athlon XP

2100+).

Results and discussion

CDK9 is a serine–threonine kinase involved in transcrip-

tion and responsible for the activation of RNA polymerase

II. Due to its role in these processes, it is a potential thera-

peutic target in human disease. Therefore, data on its 3D

A.M. Leopoldino et al. / Protein Expression and PuriWcation 47 (2006) 614–620 617

structure are essential to the development of new and spe-

ciWc inhibitors.

Studying protein 3D structure represents a challenge in

protein biochemistry. In vitro studies such as crystallization

are reliant on the solubilization, and on appropriate quan-

tity and quality of proteins. Here, we present data that led

to the successful expression and puriWcation of human

CDK9.

Overproduction of human CDK9

CDK9 protein has been produced using pET vector

system and E. coli BL21(DE)pLys (Fig. 1, lane 4). The

expression of CDK9 at 37 °C produced insoluble pro-

teins that were resuspended in denaturing buVer con-

taining 8 M urea. However, no CD spectrum was

detected when an aliquot of dialyzed CDK9 (refolding)

was analyzed by circular dichroism at a protein concen-

tration of 13 mg/mL. Others studies [16,17] reported sim-

ilar results indicating that refolding is incomplete. The

incomplete refolding of CDK9 may be due to exposition

to urea for a long period (up to 7 days) before diluting

out the denaturant.

An optimization of the expression in diVerent tempera-

tures was used to yield soluble proteins. The best condition

was by culturing the selected colonies with the plasmid

pET23a(+)::cdk9 in LB medium at 30 °C and 0.1 mM IPTG

(Fig. 1). The expression at 37 °C resulted in high amounts of

precipitated proteins; at lower temperature (28 and 20 °C),

the expression resulted in low amounts of protein (Fig. 2).

In respect to IPTG, we found that cultures with lower

IPTG concentrations produced appropriate amounts of

soluble protein. Almost 40 mg mL

¡1

of total protein was

synthesized and the soluble proteins were recovered after

centrifugation (Table 1).

By using direct absorbance readings at 280 nm, we esti-

mated that, on average, about 10 mg of pure CDK9 was

puriWed from a 500 mL culture of transformed bacteria.

Validation and puriWcation of human CDK9

The proteins before and after induction by IPTG were

analyzed by 2DE for validation. A spot with pI and molec-

ular weight close to the theoretical values calculated for

CDK9 (pI 8.97 and 42 kDa) was present in 2DE gel after

induction but absent in control. Protein spot was excised,

in-gel digested with trypsin, and identiWed by MALDI mass

spectrometry. The processed peptide mass Wngerprint data

analyzed by the MS-Fit search identiWed the human CDK9

as the Wrst hit with coverage of 82% (data not shown).

Recombinant protein was puriWed by ionic exchange

(DEAE Sepharose) and ATP-aYnity chromatography

(Table 1; Fig. 1, lane 3). The resins and elution conditions

were chosen based on our experience acquired during

CDK2 puriWcation [27]. The CDK2 puriWcation protocol

was optimized for CDK9. The fractions of puriWed CDK9

were pooled and concentrated for future analysis and

Fig. 1. PuriWcation of the recombinant human CDK9 protein by ion

exchange and ATP-aYnity chromatography. Lane: 1, crude extract, no

CDK9 expression (control); 2, ion exchange chromatography of total sol-

uble protein; 3, ATP-aYnity chromatography of human CDK9; 4, crude

extract with recombinant human CDK9; 5, LMW protein marker (GE

HealthCare). Silver staining. The resulting CDK9 is almost 99% pure.

Fig. 2. SDS–PAGE of the crude extract containing recombinant human

CDK9. Lanes: 1–3, crude extract from expression at 20, 37, and 28 °C,

respectively. Coomassie blue staining.

Table 1

The optimal puriWcation

The purities were estimated from silver-stained SDS gels and the

quantiWcation was estimated using direct absorbance readings at 280 nm.

Approximate yield (%) Approximate purity

(%)

Cell lysate 50–60 20

DEAE column 70 50

ATP-aYnity column 90 100

618 A.M. Leopoldino et al. / Protein Expression and PuriWcation 47 (2006) 614–620

maintained at 4 °C in the presence of proteases inhibitors to

avoid degradation.

Structural analysis of CDK9

Circular dichroism is an important technique to esti-

mate secondary structure content, to validate protein

quality, and for monitoring conformational changes of

proteins due to drug binding. The CDK9 deconvoluted

CD spectrum showed that it contains 30–34% helix, 33–

36% beta, and 21–29% of coil (Fig. 3). Such CD data

permitted to obtain information about the CDK9 struc-

tures of the regions not modeled. The results suggest

that human recombinant CDK9 maintains secondary

structures and is viable to functional and structural

studies.

Fig. 4 shows the sequence alignment of CDK9 and

CDK2. The alignment presents 38% of identity, considering

the region from residues 17 to 333. The overall stereochem-

ical quality of the Wnal CDK9 apoenzyme and CDK9:ATP

models and average G-factor were calculated by the PRO-

CHECK software [26]. Analysis of the Ramachandran

plots indicates that 89.5% of the residues are in most

favored regions, and 8.4% of the residues in additional

allowed regions for the CDK9 apoenzyme whereas 87% of

the residues are in most favored regions, and 9.9% in addi-

tional allowed regions for CDK9:ATP model. The G-factor

values were 0.14 and 0.38 for CDK9 apoenzyme and

CDK9:ATP models, respectively. Ideally, scores should be

above ¡0.5.

The models for CDK9 are folded into the typical bilo-

bal structure, with the smaller N-terminal lobe consisting

predominantly of -sheet structure and the larger C-ter-

minal lobe consisting primarily of -helices. The N-termi-

nal lobe of CDK has a sheet of Wve antiparallel -strands

(1–5) and a single large helix (1). The C-terminal lobe

contains a pseudo-4-helical bundle (2, 3, 4, 6), a small -

ribbon (6–8), and two additional helices (5, 7). The

ATP molecule is found in the cleft between the two lobes.

The core (the -sheet and the helical bundle) of the CDK9

structure is very similar to that of CDK2 [25,27,28], as

shown in Figs. 5A–C.

The diVerences between the structures of CDK2 and

CDK9 are in the N-terminal (residues 1–17) and C-termi-

nal regions (residues 333–372); the CDK9 sequence is

Fig. 3. Circular dichroism spectrum of human CDK9. The protein confor-

ation consists of 30–34% helices, 33–36% beta-strands, and 21–29% coils.

Fig. 4. Sequence alignment of CDK2 with CDK9 (38% identity). The alignment was performed with the program MULTIALIGN [30].

A.M. Leopoldino et al. / Protein Expression and PuriWcation 47 (2006) 614–620 619

longer than CDK2. Therefore, the N- and C-terminal were

not modeled, due to the absence of models. These regions

can provide speciWcity to CDK9, such as the diVerences in

the loops 52–55, 88–95, 182–185, and 317-322, which can be

observed in the alignment of their primary sequences

(Fig. 4). Considering the modeled region, CDK9 shows

26% of helices. Such result is consistent with CD data of the

entire CDK9 protein.

Pinhero et al. [29] established a protocol for the expres-

sion and puriWcation of the complex His

-CDK9/CycT1

using recombinant baculovirus in Spodoptera frugiperda

insect cells. On an average, from 1 L culture of infected Sf9

cells, these authors could purify about 1.2 mg of pure His

CDK9/CycT1. Compared to this study, our protocol in E.

coli presents some advantages in respect to yield of puriWed

protein, fusion tags, expression system, and operating costs/

complexity.

In summary, ours results show that recombinant

human CDK9 can be cloned and expressed in E. coli and

puriWed from soluble fraction in two chromatographic

steps. Mass spectrometry identiWcation validated the

expression of CDK9. Moreover, expression at a lower

temperature using 0.1 mM IPTG resulted in a secondary

structured recombinant protein as shown by CD analysis

validating its use in future structural analysis. This is the

Wrst report on human CDK9 expression in E. coli and

structure analysis and provides the Wrst step for the devel-

opment of new inhibitors for the treatment of AIDS,

heart disease, and cancer.

Acknowledgments

This work was supported by grants from FAPESP

(Proc.01/07532-0, 02/04383-7, and 02/09388-7), FAPERJ,

CNPq, CAPES.

References

[1] X. Grana, A. De Luca, N. Sang, Y. Fu, P.P. Claudio, J. Rosenblatt,

D.O. Morgan, A. Giordano, PITALRE, a nuclear CDC2-related pro-

tein kinase that phosphorylates the retinoblastoma protein in vitro,

Proc. Natl. Acad. Sci. USA 91 (1994) 3834–3838.

[2] G. De Falco, Giordano A, CDK9: from basal transcription to cancer

and AIDS, Cancer Biol. Ther. 1 (2002) 342–347.

[3] X. Yang, M.O. Gold, D.N. Tang, D.E. Lewis, E. Aguilar-Cordova,

A.P. Rice, C.H. Herrmann, TAK, an HIV Tat-associated kinase, is a

member of the cyclin-dependent family of protein kinases and is

induced by activation of peripheral blood lymphocytes and diVerenti-

ation of promonocytic cell lines, Proc. Natl. Acad. Sci. USA 94 (1997)

12331–12336.

[4] J. Garriga, S. Bhattacharya, J. Calbo, R.M. Marshall, M. Truongcao,

D.S. Haines, X. Grana, CDK9 is constitutively expressed throughout

the cell cycle, and its steady-state expression is independent of SKP2,

Mol. Cell. Biol. 23 (2003) 5165–5173.

[5] J. Pines, The cell cycle kinases, Semin. Cancer Biol. 5 (1994) 305–313.

[6] P.D. JeVrey, A.A. Russo, K. Polyak, E. Gibbs, J. Hurwitz, J. Mass-

agué, N.P. Pavletich, Mechanism of CDK activation revealed by

the structure of a cyclinA-CDK2 complex, Nature 376 (1995)

313–320.

[7] S.H. Chao, D.H. Price, Flavopiridol inactivates P-TEFb and blocks

most RNA polymerase II transcription in vivo, J. Biol. Chem. 276

(2001) 31793–31799.

[8] J. Peng, N.F. Marshall, D.H. Price, IdentiWcation of a cyclin subunit

required for the function of Drosophila P-TEFb, J. Biol. Chem. 273

(1998) 13855–13860.

[9] T.J. Fu, J. Peng, G. Lee, D.H. Price, O. Flores, Cyclin K functions as a

CDK9 regulatory subunit and participates in RNA polymerase II

transcription, J. Biol. Chem. 274 (1999) 34527–34530.

[10] S.M. Shore, S.A. Byers, W. Maury, D.H. Price, IdentiWcation of a

novel isoform of Cdk9, Gene 307 (2003) 175–182.

[11] Y. Zhu, T. Pe’ery, J. Peng, Y. Ramanathan, N. Marshall, T. Marshall,

B. Amendt, M.B. Mathews, D.H. Price, Transcription elongation fac-

tor P-TEFb is required for HIV-1 tat transactivation in vitro, Genes

Dev. 11 (1997) 2622–2632.

[12] J. Garriga, X. Grana, Cellular control of gene expression by T-type

cyclin/CDK9 complexes, Gene 337 (2004) 15–23.

[13] T. Ammosova, K. Washington, Z. Debebe, J. Brady, S. Nekhai,

Dephosphorylation of CDK9 by protein phosphatase 2A and protein

Fig. 5. Ribbon diagram of (A) CDK2 structure; (B) CDK9 model; and (C) CDK9:ATP model. Figure was generated by MolMol [31].

620 A.M. Leopoldino et al. / Protein Expression and PuriWcation 47 (2006) 614–620

phosphatase-1 in Tat-activated HIV-1 transcription, Retrovirology 2

(2005) 47.

[14] M. Sano, M.D. Schneider, Cyclins that don’t cycle—cyclin T/cyclin-

dependent kinase-9 determines cardiac muscle cell size, Cell Cycle 2

(2003) 99–104.

[15] D.E. MacCallum, J. Melville, S. Frame, K. Watt, S. Anderson, A. Gia-

nella-Borradori, D.P. Lane, S.R. Green, Seliciclib (CYC202, R-Rosco-

vitine) induces cell death in multiple myeloma cells by inhibition of

RNA polymerase II-dependent transcription and down-regulation of

Mcl-1, Cancer Res. 65 (2005) 5399–5407.

[16] J.J. Chalmers, E. Kim, J.N. Telford, E.Y. Wong, W.C. Tacon, M.L.

Shuler, D.B. Wilson, EVects of temperature on Escherichia coli over-

producing beta-lactamase or human epidermal growth factor, Appl.

Environ. Microbiol. 56 (1990) 104–111.

[17] A. Lombardi, M. Sperandei, C. Cantale, P. Giacomini, P. GaleY,

Functional expression of a single-chain antibody speciWc for the

HER2 human oncogene in a bacterial reducing environment, Protein

Expr. Purif. 44 (2005) 10–15.

[18] M.M. Bradford, A rapid and sensitive method for the quantitation of

microgram quantities of protein utilizing the principle of protein-dye

binding, Anal. Biochem. 72 (1976) 248–254.

[19] A. Gorg, W. Postel, A. Domscheit, S. Gunther, Two-dimensional elec-

trophoresis with immobilized pH gradients of leaf proteins from bar-

ley (Hordeum vulgare): method, reproducibility and genetic aspects,

Electrophoresis 9 (1988) 681–692.

[20] N. Sreerama, R.W. Woody, Protein secondary structure from circular

dichroism spectroscopy. Combining variable selection principle and

cluster analysis with neural network, ridge regression and self-consis-

tent methods, J. Mol. Biol. 242 (1994) 497–507.

[21] G. Deleage, C. Geourjon, An interactive graphic program for

calculating the secondary structure content of proteins from cir-

cular dichroism spectrum, Comput. Appl. Biosci. 9 (1993)

197–199.

[22] H.B. Uchoa, G.E. Jorge, N.J. da Silveira, J.C. Câmera, F. Canduri,

W.F. De Azevedo, Parmodel: a web server for automated compara-

tive modeling of proteins, Biochem. Biophys. Res. Commun. 325

(2004) 1481–1486.

[23] A. Sali, T.L. Blundell, Comparative protein modelling by satisfaction

of spatial restraints, J. Mol. Biol. 234 (1993) 779–815.

[24] U. Schulze-Gahmen, J. Brandsen, H.D. Jones, D.O. Morgan, L. Mei-

jer, J. Vesely, S.-H. Kim, Multiple modes of ligand recognition: crystal

structures of cyclin-dependent protein kinase2 in complex with ATP

and two inhibitors, olomoucine and isopentenyladenine, Proteins 22

(1995) 378–391.

[25] U. Schulze-Gahmen, H.L. De Bondt, S.-H. Kim, High-resolution crys-

tal structures of human cyclin-dependent kinase 2 with and without

ATP: bound waters and natural ligand as guides for inhibitor design,

J. Med. Chem. 39 (1996) 4540–4546.

[26] R.A. Laskowski, M.W. MacArthur, D.S. Moss, J.M. Thornton, PRO-

CHECK: a program to check the stereochemical quality of protein

structures, J. Appl. Cryst. 26 (1993) 283–291.

[27] W.F.JR. De Azevedo, S. Leclerc, L. Meijer, L. Havlicek, M. Strnad, S.-

H. Kim, Inhibition of cyclin-dependent kinases by purine analogues:

crystal structure of human CDK2 complexed with roscovitine, Eur. J.

Biochem. 243 (1997) 518–526.

[28] F. Canduri, H.B. Uchoa, W.F. De Azevedo Jr., Molecular models of

cyclin-dependent kinase 1 complexed with inhibitors, Biochem. Bio-

phys. Res. Commun. 324 (2004) 661–666.

[29] R. Pinhero, P. Liaw, K. Yankulov, A uniform procedure for the puriW-

cation of CDK7/CycH/MAT1, CDK8/CycC and CDK9/CycT1, Biol.

Proced. Online 6 (2004) 163–172.

[30] F. Corpet, Multiple sequence alignment with hierarchical clustering,

Nucleic Acids Res. 16 (1988) 10881–10890.

[31] R. Koradi, M. Billeter, K. Wüthrich, MOLMOL: a program for dis-

play and analysis of macromolecular structures, J. Mol. Graph. 14

(1996) 51–55.

Livros Grátis
( http://www.livrosgratis.com.br )
 
Milhares de Livros para Download:
 
Baixar livros de Administração
Baixar livros de Agronomia
Baixar livros de Arquitetura
Baixar livros de Artes
Baixar livros de Astronomia
Baixar livros de Biologia Geral
Baixar livros de Ciência da Computação
Baixar livros de Ciência da Informação
Baixar livros de Ciência Política
Baixar livros de Ciências da Saúde
Baixar livros de Comunicação
Baixar livros do Conselho Nacional de Educação - CNE
Baixar livros de Defesa civil
Baixar livros de Direito
Baixar livros de Direitos humanos
Baixar livros de Economia
Baixar livros de Economia Doméstica
Baixar livros de Educação
Baixar livros de Educação - Trânsito
Baixar livros de Educação Física
Baixar livros de Engenharia Aeroespacial
Baixar livros de Farmácia
Baixar livros de Filosofia
Baixar livros de Física
Baixar livros de Geociências
Baixar livros de Geografia
Baixar livros de História
Baixar livros de Línguas

Baixar livros de Literatura
Baixar livros de Literatura de Cordel
Baixar livros de Literatura Infantil
Baixar livros de Matemática
Baixar livros de Medicina
Baixar livros de Medicina Veterinária
Baixar livros de Meio Ambiente
Baixar livros de Meteorologia
Baixar Monografias e TCC
Baixar livros Multidisciplinar
Baixar livros de Música
Baixar livros de Psicologia
Baixar livros de Química
Baixar livros de Saúde Coletiva
Baixar livros de Serviço Social
Baixar livros de Sociologia
Baixar livros de Teologia
Baixar livros de Trabalho
Baixar livros de Turismo
 
 