Bacterial Mtt/Tat Pathway 179

Bacterial Twin-Arginine Signal Peptide-DependentProtein Translocation Pathway:Evolution and Mechanism

Received October 13, 1999; revised January 28, 2000; accepted January28, 2000. *For correspondence. Email wu@ibsm.cnrs-mrs.fr; Tel. +33-491164212; Fax. +33-4 91718914.

J. Mol. Microbiol. Biotechnol. (2000) 2(2): 179-189.

© 2000 Horizon Scientific Press

JMMB Minireview

Long-Fei Wu*, Bérengère Ize, Angélique Chanal,Yves Quentin and Gwennaele Fichant

Laboratoire de Chimie Bactérienne, UPR9043 CNRS,Institut de Biologie Structurale et Microbiologie, 31chemin Joseph Aiguier, 13402 Marseille cedex 20,France

Abstract

The recently identified bacterial Tat pathway is capableof exporting proteins with a peculiar twin-argininesignal peptide in folded conformation independentlyof the Sec machinery. It is structurally andmechanistically similar to the ∆pH-dependent pathwayused for importing chloroplast proteins into thethylakoid. The tat genes are not ubiquitously presentand are absent from half of the completely sequencedbacterial genomes. The presence of the tat genesseems to correlate with genome size and with thepresence of important enzymes with a twin-argininesignal peptide. A minimal Tat system requires a copyof tatA and a copy of tatC. The composition and geneorder of a tat locus are generally conserved within thesame taxonomy group but vary considerably to othergroups, which would exclude an acquisition of the Tatsystem by recent horizontal gene transfer. The tatgenes are also found in the genomes of chloroplastsand plant mitochondria but are absent from animalmitochondrial genomes. The topology of evolutiontrees suggests a bacterial origin of the Tat system. Ingeneral, the twin-arginine signal peptide is capable oftargeting any passenger protein to the Tat pathway.However, a structural signal carried by the mature partof a passenger protein can override targeting inform-ation in a signal peptide under certain circumstances.Tat systems show a substrate-Tat component spec-ificity and a species specificity. The pore size of theTat channel is estimated as being between 5 and 9 nm.Operational models of the Tat system are proposed.

Introduction

Bacteria export numerous proteins across the plasmamembrane into either the periplasmic space of Gram-negative species or directly into the growth medium in thecase of Gram-positive strains. Mechanisms for this step inprotein export have been extensively studied at themolecular level for over a decade. Until recently, it was

believed that the export of proteins across the cytoplasmicmembrane was mediated solely by the Sec machinery. Inaddition, it was generally accepted that protein unfoldingor prevention of protein folding prior to translocation iscrucial for the transport of proteins across the bacterialplasma membrane (Pugsley, 1993, Schatz andDobberstein, 1996).

Translocation of metallo-enzymes challenged thisgeneral schema since formation of metallo-centers isrelated to translocation. Such coordination defines threemechanisms used for the export of metallo-enzymes. Thedifference between these mechanisms lies in the respectivelocations of the metallo-center formation. The first one isrepresented by the biogenesis of respiratory cytochromes(Thony-Meyer, 1997). Apocytochrome precursors areexported through the Sec machinery, and the covalentattachment of heme appears to be enzyme-catalyzed andtakes place in the periplasm. The second possible pathway,implying a concomitant formation of the metallo-center withthe translocation of a precursor crossing the membrane,has not yet been demonstrated. In the third mechanism,the incorporation of a metal ion into a precursor occurs inthe cytoplasm. The metallo-enzyme is then translocatedacross the membrane in a folded conformation through aspecific pathway. Recently, such a novel protein exportsystem has been revealed, first in chloroplasts (∆pHpathway) (Settles et al., 1997) and then in Escherichia coli(Bogsch et al., 1998, Chanal et al., 1998, Santini et al.,1998, Sargent et al., 1998, Weiner et al., 1998). Thebacterial system was designated first as «membranetargeting and translocation» (Mtt) (Weiner et al., 1998) andthen as a «twin-arginine translocation» (Tat) pathway(Sargent et al., 1998). These findings provoked a generalinterest in this fascinating mechanism, and the salientfeatures of these pathways have been recently reviewedin two articles (Dalbey and Robinson, 1999, Settles andMartienssen, 1998). However, these reviews are morechloroplast-oriented. In addition, considerable progress hasbeen made since then in the study of the bacterial systemdue to the amenability of the microorganisms for geneticstudies. In this review we attempt to make criticalevaluations of the most current findings in the bacterial Tatsystem with a brief comparison to the ∆pH pathway ofchloroplasts only for the purpose of a better understandingof the general mechanism. We will focus on evolution,classification and specificity of the Tat system at bothsubstrate and species levels. We wish to provide anupdated view of this rapidly moving field and some insightinto the mechanism of the Tat pathway.

Evolution of the Tat and ∆pH pathways

Ubiquitous Presence of the Tat SystemThe first component of the Tat/∆pH pathway to be identifiedwas reported in maize (Settles et al., 1997). The Hcf106

• MALDI-TOF Mass Spectrometry in Microbiology

Edited by: M Kostrzewa, S Schubert (2016) www.caister.com/malditof

• Aspergillus and Penicillium in the Post-genomic Era

Edited by: RP Vries, IB Gelber, MR Andersen (2016) www.caister.com/aspergillus2

• The Bacteriocins: Current Knowledge and Future Prospects

Edited by: RL Dorit, SM Roy, MA Riley (2016) www.caister.com/bacteriocins

• Omics in Plant Disease Resistance

Edited by: V Bhadauria (2016) www.caister.com/opdr

• Acidophiles: Life in Extremely Acidic Environments

Edited by: R Quatrini, DB Johnson (2016) www.caister.com/acidophiles

• Climate Change and Microbial Ecology: Current Research and Future Trends

Edited by: J Marxsen (2016) www.caister.com/climate

• Biofilms in Bioremediation: Current Research and Emerging Technologies

Edited by: G Lear (2016) www.caister.com/biorem

• Microalgae: Current Research and Applications

Edited by: MN Tsaloglou (2016) www.caister.com/microalgae

• Gas Plasma Sterilization in Microbiology: Theory, Applications, Pitfalls and New Perspectives

Edited by: H Shintani, A Sakudo (2016) www.caister.com/gasplasma

• Virus Evolution: Current Research and Future Directions

Edited by: SC Weaver, M Denison, M Roossinck, et al. (2016) www.caister.com/virusevol

• Arboviruses: Molecular Biology, Evolution and Control

Edited by: N Vasilakis, DJ Gubler (2016) www.caister.com/arbo

• Shigella: Molecular and Cellular Biology

Edited by: WD Picking, WL Picking (2016) www.caister.com/shigella

• Aquatic Biofilms: Ecology, Water Quality and Wastewater Treatment

Edited by: AM Romaní, H Guasch, MD Balaguer (2016) www.caister.com/aquaticbiofilms

• Alphaviruses: Current Biology

Edited by: S Mahalingam, L Herrero, B Herring (2016) www.caister.com/alpha

• Thermophilic Microorganisms

Edited by: F Li (2015) www.caister.com/thermophile

• Flow Cytometry in Microbiology: Technology and Applications

Edited by: MG Wilkinson (2015) www.caister.com/flow

• Probiotics and Prebiotics: Current Research and Future Trends

Edited by: K Venema, AP Carmo (2015) www.caister.com/probiotics

• Epigenetics: Current Research and Emerging Trends

Edited by: BP Chadwick (2015) www.caister.com/epigenetics2015

• Corynebacterium glutamicum: From Systems Biology to Biotechnological Applications

Edited by: A Burkovski (2015) www.caister.com/cory2

• Advanced Vaccine Research Methods for the Decade of Vaccines

Edited by: F Bagnoli, R Rappuoli (2015) www.caister.com/vaccines

• Antifungals: From Genomics to Resistance and the Development of Novel Agents

Edited by: AT Coste, P Vandeputte (2015) www.caister.com/antifungals

• Bacteria-Plant Interactions: Advanced Research and Future Trends

Edited by: J Murillo, BA Vinatzer, RW Jackson, et al. (2015) www.caister.com/bacteria-plant

• Aeromonas

Edited by: J Graf (2015) www.caister.com/aeromonas

• Antibiotics: Current Innovations and Future Trends

Edited by: S Sánchez, AL Demain (2015) www.caister.com/antibiotics

• Leishmania: Current Biology and Control

Edited by: S Adak, R Datta (2015) www.caister.com/leish2

• Acanthamoeba: Biology and Pathogenesis (2nd edition)

Author: NA Khan (2015) www.caister.com/acanthamoeba2

• Microarrays: Current Technology, Innovations and Applications

Edited by: Z He (2014) www.caister.com/microarrays2

• Metagenomics of the Microbial Nitrogen Cycle: Theory, Methods and Applications

Edited by: D Marco (2014) www.caister.com/n2

Caister Academic Press is a leading academic publisher of advanced texts in microbiology, molecular biology and medical research. Full details of all our publications at caister.com

Further Reading

Order from caister.com/order

180 Wu et al.

gene encodes a receptor-like thylakoid membrane protein,and its mutation disrupts the localization of proteinstransported through the ∆pH pathway. Independently, aHcf106 counterpart mutant mttA was isolated in E. coli(Weiner et al., 1998). A point mutation in the mttA generesults in the mislocation of three molybdoenzymes thatcontain a peculiar twin-arginine signal peptide. mttAbelongs to the mttABC operon, and the encoding proteinsMttA, B and C are members of a large family of relatedsequences extending from archaea to higher eukaryotes(Weiner et al., 1998). Shortly after this report, revision ofthe DNA sequence revealed that mttA is actually two genes,tatA and tatB (Sargent et al., 1998) that encode two proteinssharing about 25% sequence identity (Chanal et al., 1998).In addition to the mttA1A2BC operon (also known as thetatABCD operon), a tatA homologous gene tatE was foundat 14 min on the chromosome of E. coli (Sargent et al.,1998, Settles et al., 1997). These genes represent all ofthe tat genes that have been identified to date.

Early scanning of data bases suggested that genesencoding Mtt/Tat components might be ubiquitously presentin the three domains of life (Settles et al., 1997, Weiner etal., 1998). Since then, several complete genomesequences have become accessible, we have thereforeundertaken a systematic analysis of the Tat system toprovide an updated view regarding the evolution andmechanism of this novel pathway.

The most widespread Tat component is TatD.Sequences related to tatD are found in all completegenomes except the genome of Archaeoglobus fulgidusthat contains two tatA and two tatC genes (Table 1) but notatD homologue. Since the depletion of tatD does not affectthe translocation of typical Tat substrates (Settles andMartienssen, 1998), whether or not TatD belongs to theTat pathway remains an unanswered question. Thereforewe will not consider TatD any further in this review.

TatC is an essential component of the Tat system, andthe depletion of tatC leads to mislocation of all the enzymesanalyzed (Bogsch et al., 1998). The tatC gene is presentin all complete bacterial genomes bigger than 2 Mb, but isabsent from the two genomes smaller than 1 Mb and isalso absent from half of the bacterial genomes between 1and 2 Mb (Table1). Interestingly tatC is the only tat genethat is found in the genomes of chloroplasts and of plantmitochondria (Table 1). However, it is absent from animalmitochondrial genomes (data not shown). Regarding thethree homologous components, TatA, TatB and TatE, atleast one copy of tatA is always found in a completebacterial genome containing tatC but absent from genomeslacking tatC. Therefore, a functional bacterial Tat systempresumably requires both TatC and TatA. In addition, thepresence of a Tat system seems to be perfectly correlatedwith the cellular location of some important enzymes inbacteria. For example, hydrogenases are enzymes ofwidespread occurrence in bacteria and occupy a centralplace in the energy metabolism of anaerobic bacteria(Adams, 1990). The complete genomes of Aquifex aeolicus,Escherichia coli, Archaeoglobus fulgidus and Helicobacterpylori encode a periplasmic or membrane-boundhydrogenase with the twin-arginine signal peptide and, inparallel, they contain the tat genes. By contrast, thehomologous hydrogenases from Methanococcusjannaschii and Methanobacterium thermoautotrophicum donot contain the twin-arginine signal peptide, and the tat

genes are consistently absent from these genomes. Takentogether, the Tat system is not ubiquitously present in allgenomes and is unlikely to be among the minimal genesrequired for life. Having a Tat system seems to be a«luxury» which a bacterium can offer to itself only if thesize of its genome allows to do it. Nevertheless, the Tatsystem could confer the bacterium an advantage in thecompetition for a biological niche. At present, ourknowledge regarding the physiological implication of Tatsystem is very limited (see below).

Genetic Structure of tat Operons and Evolution ofthe tat PathwayDespite the fact that the tat genes are absent from half ofthe complete bacterial genomes, the tat gene duplicationappears to have occurred frequently. The entire tatACoperon has been duplicated in the genome of Bacillussubtilis. But in most cases, only the tatA-like genes undergoduplications which lead to a second copy of tatA known astatA2, tatB or tatE (Table 1). The divergent evolution oftatB with an increase in its C-terminal domain resulted inthe formation of a cluster distinct from tatA both in sequenceand in function in Proteobacteria and in Actinobacteria(Table 1, Figure 1A, and see below). On the other hand,tatE was found only in the enterobacteria and might resultfrom a second duplication of tatA after the emergence ofthe ancestor of the enterobacteria. The minimalcomponents of a bacterial Tat system may include one copyof tatA and one copy of tatC in Richettsia prowazekii andin Aeropyrum pernix. The composition of a tat locus variesfrom a single gene to three genes, but it seems to be highlyconserved within a taxonomic group (Table 1). For example,the tatC gene is usually adjacent to tatA or tatB and theyform a tatAC or a tatBC operon, respectively. However,the three genes are located within a single locus in thegenomes of the beta and gamma subdivisions ofProteobacteria (Table 1). In contrast to such a groupedorganization, tatC in the genomes of R. prowazekii, A.pernix, Archaeoblobus fulgidus, Aquifex aeolicus,Chlorobium tepidum and Synechocytis is located alone ata locus separated from that of tatA. Such variation in geneorder suggests a complicated evolutionary process to whichthe Tat system was subjected and it would exclude apossible acquisition of the Tat system in various organismsby recent horizontal gene transfer.

Evolutionary trees have been computed with TatA/TatB/TatE (Figure 1A) or TatC (Figure1B). The speciesbelonging to the same group form a cluster, but the topologymay be different from that of the phylogenic tree obtainedwith 16S RNAs. In addition, the Tat components fromchloroplasts form a cluster with those from cyanobacteria,and those from mitochondria with the alpha-proteobacteria.These results suggest a bacterial origin of the Tat system.The tat genes were also found in the nuclear genome ofArabidopsis thaliana, Pisum salivum and Zea mays (Table1), suggesting gene transfer from chloroplasts to thenuclear genome. Transfer from nuclei or chloroplasts toplant mitochondria is unlikely since TatC from chloroplastsand mitochondria belong to different clusters on theevolutionary tree (Figure 1B). As mentioned above, the tatgenes are absent from the genomes of animalmitochondria. At present, we do not know if the tat geneshave been completely transferred to animal nucleargenomes or if they were lost during the evolution of animal

Bacterial Mtt/Tat Pathway 181

Table 1. Genes Encoding Tat Components and Their Genetic Structure in Various Genomes

Taxonomy Organisms (Abbreviation) Size tatC tatA tatB tatE Genetic structure

Archaea* Crenarchaeota Aeropyrum pernix (APR) 1.67 1 1 0 0 C; A

* Eurvarchaeota Archaeoglobus fulgidus (AFU) 2.18 2 2 0 0 A; A2; C; C2Methanobacterium thermoautotropicum (MTH) 1.75 0 0 0 0Methanococcus jannaschii (MJA) 1.66 0 0 0 0Pyrococcus abyssi (PAB) 1.77 0 0 0 0Pyrococcus horikoshii (PHO) 1.80 0 0 0 0

Bacteria* Aquificales Aquifex aeolicus (AAE) 1.50 1 2 0 0 C; AA2

* Firmicutes- Bacillus/Clostridium group Bacillus subtilis (BSU) 4.20 2 3 0 0 AC; A2C2; A3

Mycoplasma genitalium (MGE) 0.58 0 0 0 0Mycoplasma pneumoniae (MPN) 0.81 0 0 0 0Bacillus halodurans (BHA) 1 1Staphylococcus aureus (SAU) 1 1 AC

- Actinobacteria Mycobacterium tuberculosis (MTU) 4.40 1 1 1 0 AC; BMycobacterium bovis (MBO) 1 1 1 AC; BMycobacterium leprae (MLE) 1 1 1 AC; BMycobacterium avium (MAV) 1 1 1 AC; BRhodococcus. erythropolis (RER) 1 A

* Spirochaetales Borrelia burgdorderi (BBU) 1.44 0 0 0 0Treponema pallidum subsp.pallidum (TPA) 1.14 0 0 0 0

* Thermotogales Thermotoga maritima (TMA) 1.80 0 0 0 0

* Planctomyces/Chlamydia group Chlamydia trachomatis (CTR) 1.05 0 0 0 0Chlamydia pneumoniae (CPN) 1.23 0 0 0 0

* Proteobacteria- gamma subdivision Haemophilus influenzae (HIN) 1.83 1 1 1 0 ABC

Escherichia coli (ECO) 4.60 1 1 1 1 ABC; ESalmonella tiphy (STI) 1 1 1 1 ABC; EYersinia pestis (YPE) 1 1 1 1 ABC; EVibrio cholerae (VCH) 1 1 1 1 ABC; EPseudomonas aeruginosa (PAE) 1 1 1 0 ABCShewanella putrefaciens (SPU) 1 1 1 ABCThiobacillus ferrooxidans (TFE) 1 1 1 ABCPasteurella multocida (PMU) 1 1 1 ABCActinobacillus actinomycetemcomitans (AAC) 1 1 1 ABC?Azotobacter chroococcum (ACH) 1 1 1 ABCPseudomonas stutzeri (PST) 1

- delta/epsilon subdivisions Helicobacter pylori 26695 (HP2) 1.67 1 1 1 0 BC; AHelicobacter pylori J99 (HPJ) 1.64 1 1 1 0 BC; ACampylobacter jejuni (CJE) 1 1 1 BC; A

- beta subdivision Neisseria meningitidis (NME) 1 1 1 ABCBordetella pertussis (BPE) 1 1 1 ABCNeisseria gonorrhoace (NGO) 1 1 1 ABC

- alpha subdivision Rickettsia prowazekii (RPR) 1.10 1 1 0 0 C;ACaulobacter crescentus (CCR) 1 1 1 AB;C?

* Cyanobacteria Synechocystis PCC6803 (CYA) 3.57 1 2 0 0 C; A; A2* Deinococcus Drinococcus radiodurans (DRA) 3.28 1 2 0 0 AC;A2* Green sulfur Chlorobium tepidum (CTE) 1 2 C; A; A2

Eukaryota* Fungi Saccharomyces cerevisiae (SCE) 13 0 ? ? 0* Metazoa Caenorhabditis elegans (CEL) >100 0 0 0 0

*Nuclear genome encodingchloroplast proteins Arabidopsis thaliana (ATH) 1 1 A; C

Pisum salivum (PSA) 1 AZea mays (EMA) 3 ?

* Chloroplast genome Porphyra purpurea (PPC) 0.19 1 0 0 0Guillardia theta (GTC) 0.12 1 0 0 0Odontella sinensis (OSC) 0.12 1 0 0 0Zea mays (EMA) 0.14 0 0 0 0Epifagus virginiana (EVI) 0.07 0 0 0 0Nicotiana tabacum (NTA) 0.16 0 0 0 0Oryza sativa (OSA) 0.13 0 0 0 0Euglena gracilis (EGR) 0.14 0 0 0 0Toxoplasma gondii (TGO) 0.04 0 0 0 0

* Plant mitochondrial genomes Marchantia polymorpha (MPM) 0.19 1 0 0 0Porphyra purpurea (PPM) 0.04 1 0 0 0Chondrus crispus (CCM) 0.03 1 0 0 0Cyanidioschyzon merolae (CMM) 0.03 1 0 0 0Arabidopsis thaliana (ATM) 0.37 1 0 0 0Chlamydomonas eugametos (CEU) 0.02 0 0 0 0Chlamydomonas reinharditii (CRE) 0.02 0 0 0 0Pedimonas minor (PMI) 0.03 0 0 0 0Prototheca wickerhamii (PWM) 0.06 1Oenothera villaricae (OVM) 1Oenothera berteriana (OBM) 1

* Protist Reclinomonas americana (RAM) 0.07 1 0 0 0

Preliminary sequence data was obtained from National Center for Biotechnology Information, National Library of Medicine National Institutes of Health (http://www.ncbi.nlm.nih.gov)and from The Institute for Genomic Research (http://www.tigr.org). The size represents approximate (in mega base pairs) size of the complete genomes. Copy of the tatC, tatA,tatB or tatE genes found in a genome is presented by numbers. Genetic structure describes the loci which are composed of the tat genes as indicated. The loci are separated by ;and by ? when it is not established. A2, A3 and C2 represent the second or the third copies of tatA or tatC.

182 Wu et al.

Figure 1. (A) Tree of the TatA/B proteins. (B) Tree of the TatC proteins. The sequences have been aligned using ClustalW (Thompson et al., 1994). In the caseof TatA/B, only the N-terminal parts of the sequences have been considered. The tree was computed with the neighbor-joining method (Saitou and Nei, 1987)applied on a protein distance matrix derived from the multiple alignment with the PROTDIST method (Felsenstein, 1989). It was drawn with the PHYLO_WINtool (Galtier et al., 1996). The protein name refers to Table 1. The branch lengths are proportional to evolutionary distances. Only bootstrap scores greaterthan 60% (500 iterations) are indicated on the corresponding branch. The different taxonomic groups are indicated.

Bacterial Mtt/Tat Pathway 183

184 Wu et al.

mitochondrial genomes which are more compact andsmaller than the plant mitochondrial genomes. However,several observations would support loss of the tat genesfrom animal mitochondria. First, mitochondria are assumedto have evolved from the rickettsial group, and themitochondria of Reclinomonas americana are consideredto be the ancestor of all mitochondria (Lang et al., 1997).Both Rickettsia prowazekii and Reclinomonas americanacontain tat genes, strongly suggesting the presence of thetat genes in the ancestral mitochondrial DNA. Secondly,the tat genes are absent from the genome ofCaenorhabditis elegans, the first complete animal genome(Table 1).

Mechanism of the Tat pathway

Twin-Arginine Signal PeptideThe bacterial twin-arginine signal peptide was originallyfound in periplasmic and membrane bound hydrogenases(Voordouw, 1992, Wu and Mandrand, 1993). It was thendescribed as being conserved among various classes ofbacterial enzymes containing redox cofactors (Berks,1996), and was shown to direct these enzymes to the Sec-independent Tat pathway (Bogsch et al., 1998, Rodrigueet al., 1999, Santini et al., 1998, Sargent et al., 1998, Weineret al., 1998). The bacterial twin-arginine signal peptide

contains a characteristic, unusually long, N-terminal regionwith a consensus sequence S/T-R-R-X-phi-phi, of whichthe twin arginines are completely invariant and phi is ahydrophobic residue (Berks, 1996). The bacterial Tatpathway is structurally and mechanistically similar to the∆pH-dependent pathway used for importing chloroplastproteins into the thylakoid (Dalbey and Robinson, 1999,Settles and Martienssen, 1998, Wexler et al., 1998). Mostimportantly, the Tat signal peptides from bacteria andthylakoids appear to be interchangeable (Halbig et al.,1999a, Mori and Cline, 1998, Settles and Martienssen,1998, Wexler et al., 1998). In general, both the chloroplastand the bacterial twin-arginine signal peptides are capableof translocating a fusion protein across the membranethrough the ∆pH or Tat pathway regardless of whether thepassenger protein originally use the ∆pH/Tat or Secpathway (Table 2).

Specificity of the Twin-Arginine Signal Peptide forTargeting Proteins to the Tat/∆pH PathwaysThe twin-arginine signal peptides resemble classical signalpeptides in overall structural terms. Comparison of the twokinds of signal peptides reveals three differences: (a) theN-domain of the signals for the ∆pH-driven system or Tatpathway invariably contains a critical twin-arginine motifprior to the hydrophobic (H) domain whereas known Sec-

Table 2. Targeting of Proteins by a Native or Modified Signal Peptide to the Sec or Tat/∆pH Pathways

Proteinsa Translocation via pathways ofb Referencesc

Signal peptide Passenger ∆pH/Tat Sec

TorA - 23K +/+* − 1, 2FdnG - 23K + 2HyaA - PC + 316K - PC + 423K - PC + − 4-933K - 16K + 6, 7

N23K-H17K-C17K - PC + 7N23K-H23K-CPC - PC + − 7N23K-HPC-CPC - PC 60%+ + 3, 7N23K-HPC-CPC - 16K + 7N23K-HPC-CPC - PSI-N + 7N23K-HPC-CPC - PSII-T + 7N23K-H33K-C33K - PC + − 7N23K-HPC-CPC - 33K + + 7NPC-H23K-C23K - PC − − 7

TorA - P2 (Lep) +* −* 9TorA[10:10]- P2 (Lep) −* +* 9TorA - Lep −* +* 9TorA - Lep-inv −* +* 9GFOR - ß-Gal Targeted* 10PelB - Hyd2 − Targeted* 11PelB - HiPIP − + 12

a): Signal peptides and proteins originally targeted/translocated by the ∆pH/Tat or Sec pathways are present in bold or normal letters, respectively. Thylakoidalproteins are present in italics. The cytoplasmic located ß–galactosidase (ß–Gal) is underlined. The N, H and C represent the N-terminal, the hydrophobic andthe C-terminal domain of the signal peptide, respectively. Abbreviation: TorA: trimethylamine N-oxide reductase; FdnG: the gamma subunit of the nitrate-related formate reductase; Hya: the small subunit of the hydrogenase 1; PC: plastocyanin; 16K/17K, 23K, 33K are thylakoid proteins 16/17kDa, 23kDa and33kDa of the oxygen-evolving complex, respectively; PSI-N, PSII-T are proteins N and T from the photosynthetic system I and II, respectively. The [10:10]presents the replacement of the 10 central residues in the H-domain of the TorA signal peptide by a LAL8 hydrophobic stretch. P2 (Lep): P2 domain of theleader peptidase; Lep-inv: inverted leader peptidase; Hyd2: hydrogenase 2; HiPIP: high potential iron-sulfur protein.b): Passenger proteins were successfully (+), partially (60%+) or not (-) translocated by ∆pH/Tat or Sec pathways. * indicates that the translocation wasanalyzed in bacteria. “Targeted” means that the passenger protein seems to be targeted to, but not translocated across, the pathway indicated.c): References 1): (Sargent et al., 1998); 2): (Wexler et al., 1998); 3): (Mori and Cline, 1998); 4): (Michl et al., 1994); 5): (Henry et al., 1994); 6): (Henry et al.,1994); 7): (Henry et al., 1997); 8): (Bogsch et al., 1997); 9): (Cristobal et al., 1999); 10): (Halbig et al., 1999b); 11): (Rodrigue et al., 1999); 12): (Brüser et al.,1998).

Bacterial Mtt/Tat Pathway 185

dependent signals contain a lysine or an arginine; (b) theH-domains of ∆pH/Tat-dependent signals are generallyshorter, and (c) the twin-arginine signal peptides often haveone or more positively charged residues, known as «Sec-avoidance» signals, in the C-terminal region just upstreamof the signal peptidase cleavage site (Bogsch et al., 1997,Cristobal et al., 1999). It has been shown that the twin-arginine motif is critical for protein targeting by both bacterialTat and thylakoidal ∆pH pathways (Cristobal et al., 1999,Gross et al., 1999, Halbig et al., 1999b, Nivière et al., 1992,Wexler et al., 1998). However, the twin-arginine motif itselfis not sufficient to convert a Sec-signal into a Tat-signal(Brink et al., 1998, Brüser et al., 1998, Chaddock et al.,1995). Nevertheless, the N-domain of the ∆pH-signalpeptide added in front of the H- and C- domains of a Sec-signal is sufficient to allow import into thylakoids via the∆pH pathway both of Sec-dependent and ∆pH-dependentsubstrates (Table 2). In a certain context, the H-region andthe C-domain could also be determining factors in thechoice of the translocation pathway. In fact, an increase inthe hydrophobicity of the H-region, or loss of the Sec-avoidance signal by removing the charged residue leadsto a conversion of a twin-arginine signal peptide into a Sec-targeting signal peptide (Bogsch et al., 1997, Brink et al.,1998, Cristobal et al., 1999).

Although the twin-arginine or Sec-signal peptides arecapable of re-routing a protein from its original pathway tothe other one, it does not always lead to successfultranslocation or export (Table 2). The passenger proteincould be either stuck in the membrane (Halbig et al., 1999b,Rodrigue et al., 1999) or be inefficiently translocated(Nivière et al., 1992). In addition to the signal peptide, astructural signal residing in the mature part of a protein

seems to be equally important for targeting (Table 2). Forexample, the Sec-targeting signals in the leader peptidase(Lep) can override Tat-targeting information in a twin-arginine signal peptide that is added in front of Lep(Cristobal et al., 1999). Consistently, the efficient export ofnative Tat substrates depends both on an intact twin-arginine signal peptide and on the generation of a structuralexport signal induced by cofactor binding (Halbig et al.,1999b, Rodrigue et al., 1996, Santini et al., 1998).

Specificity of the Tat SystemIt was estimated that about two dozen proteins aretranslocated through the Tat pathway in E. coli. To date, atleast nine bacterial enzymes with a molecular mass rangingfrom 23 kDa to 116 kDa containing five types of cofactorshave been shown to use the bacterial Tat pathway (Table3). In addition, the twin-arginine signal peptide is capableof directing the translocation of other polypeptides lackingcofactor. These proteins might fold too quickly to be handledby the Sec pathway (Table 3). The translocation of theseproteins does not depend equally on a given Tatcomponent. For example, mutation of tatA, tatB or tatEleads to different consequences for the translocation ofvarious enzymes (Chanal et al., 1998, Sargent et al., 1998,Wu et al., 2000). Therefore, the homologous TatA, TatBand TatE components may handle a subset of substrateswith a greater or lesser overlap substrate specificity. Thespecificity could be determined by the fine-tuning variationin the twin-arginine signal peptides and/or by the targetinginformation carried by the mature part of these enzymes.

In addition to the substrate-Tat component specificitywithin a single bacterium, Tat systems also show species-specificity. Such specificity might first reside at the level of

Table 3. Characteristics of Proteins that Were Targeted or Translocated by the Bacterial Mtt (Tat) Pathway

Proteins Size Origin of Tat pathway Referencesg

Nickel enzymesHydrogenase 1 40.6 kDa+66.2 kDa (5 nma) Escherichia coli 1 to 3Hydrogenase 2 39.7 kDa+62.5 kDa (5 nma) Escherichia coli 1 to 5

MolybdoenzymesTMAO reductase 97 kDa (5x8x4.5 nmb) Escherichia coli 1 to 6DMSO reductase 82.6kDa + 23.6kDa Escherichia coli 1, 2, 3, and 6Formate dehydrogenase 116 kDa Escherichia coli 1 to 3Nitrate reductase 93 kDa Escherichia coli 6

Copper enzymeYack 57 kDac Escherichia coli 3

NADP-enzymeGlucose-fructose oxidoreductase 49 kDa (8.5x10x4.3 nmd) Zymomonas mobilis 7

Iron-sulfur proteinHiPIP 13.5 kDa (2.9x3.2x3.4e) Chromatium vinosum 8

OtherSufI 55 kDa Escherichia coli 3P2-domain of Leader peptidase 24 kDa Escherichia coli 9Leader peptidase 36 kDa Escherichia coli 9ß-lactamase 21 kDa Escherichia coli 1023 K protein 23 kDa Thylakoids 2β−galactosidase 116 kDa (17.5x13.5x9 nmf) Zymomonas mobilis 7

a): The homologous hydrogenase from D. gigas shows a globule structure with a diameter of about 5 nm (Volbeda et al., 1995). b): Dimensions of thestructure of the homologous TMAO reductase of S. massilia (Czjzek et al., 1998). c): Molecular weight calculated from the sequence. d): Dimensions of thetetramer enzyme (Kingston et al., 1996). e): calculated from the structure 1HIP (Carter Jr. et al., 1974). f): Dimensions of the active tetramer form (Jacobsonet al., 1994). g): References 1: (Sargent et al., 1998); 2: (Bogsch et al., 1998); 3: (Sargent et al., 1999); 4: (Chanal et al., 1998); 5: (Rodrigue et al., 1999); 6:(Weiner et al., 1998); 7: (Halbig et al., 1999b); 8: (Carter Jr. et al., 1974); 9: (Cristobal et al., 1999); 10: (Nivière et al., 1992).

186 Wu et al.

Figure 2. Operational models of the bacterial Tat system. Transmembrane segments and amphipathic domains are presented by cylinder and oval drawings,respectively. A): Gating-pushing model. The N-terminal transmembrane segments of TatA/B/E may serve as an anchor to keep them attached to the membrane.They could also participate in the formation of the Tat channel. Their amphipathic domain might maintain the permeability barrier of the membrane by sealingthe cytoplasmic end of the translocon pore prior to, and in the early stages of, translocation. Once the substrate with a twin-arginine signal peptide (RR)comes into contact with the channel, the amphipathic domains of TatA/B/E would successively open the channel and push it outward through the channel. B):Targeting-insertion-opening model. TatA/B/E might form a homo- or hetero-multimer (arbitrarily presented as a dimer here) complex by keeping thetransmembrane segments inside and the amphipathic structures at the periphery to ensure solubility in the cytoplasm. After recognition and binding of thesubstrate, the complex would then bring them directly or indirectly to TatC (arbitrarily presented as a monomer) which consists of a closed channel. Integrationof the transmembrane segments of TatA/B/E into the channel could increase the size of the pore and facilitate translocation. The subsequent scenario wouldbe the same as in (A).

Bacterial Mtt/Tat Pathway 187

the capacity of maturation of a given enzyme in the hostcells since the formation of an active metal-center is aprerequisite for targeting of the enzyme (Halbig et al.,1999b, Rodrigue et al., 1996, Santini et al., 1998).Therefore, failure in the formation of a metal-center in ahost cell would prohibit the export of the proteins throughthe Tat pathway. At the second level, the correctlysynthesized hetero-substrate may not be efficiently (Brüseret al., 1998) or may not be successfully (Halbig et al.,1999b, Wiegert et al., 1997) translocated by the host Tatsystem. Such a species specific mode of translocation wasfirst pointed out by Halbig et al. (1999b) and Wiegert etal.(1997). An active pre-glucose:fructose oxidoreductase(preGFOR) of Z. mobilis was synthesized in E. coli, but itwas not processed into a stable periplasmic protein. In thiscase either the preGFOR is not recognized by the Tatsystem of E. coli, or it is too big to cross the E. coli Tatchannel (see below). In an extreme situation, hetero-expressed Tat components might not assemble properlyin the host cells or might not interact with other host Tatcomponents. Indeed, we observed that the tatAC genesof Thermus thermophilus could not complement the E. colitatC mutant (unpublished results). Similarly, TatB ofHelicobacter pylori could not complement a tatB mutant ofE. coli, but the tatA gene of the former organism doescomplement the tatA-tatE mutant of the latter (Sargent etal., 1999). A directed evolutionary approach would providesome insight into the species-specificity of the Tat system.

Operational Model of the Bacterial Tat SystemOne of the intriguing questions regarding the mechanismof the Tat pathway is how a folded polypeptide could crossthe cytoplasmic membrane. TatC consists of sixtransmembrane segments, and its depletion completelyblocks the translocation of all enzymes analyzed.Therefore, we assume that TatC forms the core structureof the Tat channel (Figure 2). The crystal structures of theNiFe-hydrogenase from Desulfovibrio gigas and the TMAOreductase from Shewanella massilia have been resolved(Czjzek et al., 1998, Volbeda et al., 1995). Based on thesedata, we estimate the narrowest diameter for thehydrogenase 2 and the TMAO reductase of E. coli at about5 nm (Table 3). Therefore the pore size of the Tat channelof E. coli should be larger than 5 nm. On the other hand,the glucose-fructose oxidoreductase functions as atetramer with a dimension of 8.5x10x4.3 nm (Kingston etal., 1996), and it was not translocated into the periplasmof E. coli (Halbig et al., 1999b, Wiegert et al., 1997).Therefore the upper limit of the Tat channel of E. coli wouldbe less than 9 nm. Interestingly, ß-galactosidase withdimensions of roughly 17.5x13.5x9 nm seems to stick inthe Tat channel of Z. mobilis (Halbig et al., 1999b) whichwould limit the extendable size of the channel to less than13 nm. At present, we know neither if the size of a Tatchannel varies from bacterium to bacterium nor if there isa correlation between the copy number of tat genes foundin a genome and the size of the corresponding Tat channel.Nevertheless, we could estimate the opening size of theTat channel as being between 5 and 9 nm.

The Sec pathway is known to “thread” substrates inan unfolded state through the Sec channel. In E. coli, theSec channel is composed of three polypeptides, SecYEG,with a total of 15 potential transmembrane segments. Thesize of the pore might extend up to 5 nm (A.J.M Driessen,

personal communication). Similarly, the size of the aqueouspore of the Sec complex at the ER membrane is between4 and 6 nm (Hamman et al., 1997). These values are closeto the narrowest cross section of the hydrogenases andthe TMAO reductase. Therefore, the size of the pore mightnot be the only restriction parameter which prohibits theTat substrate from crossing the membrane through the Secchannel. Both the size and the form of a folded protein areessential factors to be considered in the mechanism of theTat pathway. Along the same vein, more than one copy ofTatC or other components would be required to form theTat channel with an appropriate pore size equal to or largerthan 5 nm.

TatA, TatB and TatE contain an homologous N-terminaltransmembrane segment and an adjacent amphipathicstructure followed by a C-terminal domain which variesboth in sequence and in length. The N-terminaltransmembrane segment may serve as an anchor to keepTatA/B/E attached to the membrane. In addition, thetransmembrane segment, and probably also theamphipathic domain, may participate in the formation ofthe Tat channel (Figure 2). This is supported by a reportthat TatB may be required to stabilize TatC (Sargent et al.,1999). The amphipathic domain may lie at the surface ofthe membrane or at the top of the channel to maintaincorrect orientation of the C-terminal domain that may havea receptor function. Such a reception might be performedby TatA, TatB or TatE that is either located separately fromother Tat components or is associated with the Tat channel(Figure 2A). In addition, the amphipathic domain mightmaintain the permeability barrier of the membrane bysealing the cytoplasmic end of the translocon pore priorto, and in the early stages of, translocation. Once thesubstrate comes into contact with the channel, theamphipathic domains of TatA/B/E would successively openthe channel and push it outward through the channel.Alternatively, TatA/B/E might also play a targeting role. Inthis case, they may form a homo- or hetero-multimercomplex by keeping the transmembrane segments insideand the amphipathic structures at the periphery to ensuresolubility in the cytoplasm (Figure 2B). After recognitionand binding of the substrate, the complex would then bringit directly to TatC, or first integrate into the membrane andthen bring it to TatC which forms the core structure of aclosed channel. The integration of the transmembranesegments of TatA/B/E into the channel could increase thesize of the pore and facilitate the passage of the substrate.The subsequent scenario would be the same as above. Atpresent, we do not know exactly how the Tat system isenergized. Nevertheless the proton motive force seems tobe essential for the bacterial Tat pathway (Santini et al.,1998) and the ∆pH gradient is the main energy source forthe chloroplast ∆pH pathway.

It should be noted that only the tatC gene is found inthe genomes of chloroplasts and plant mitochondriawhereas the tatA gene is present in the nuclear genome.Since the ∆pH pathway imports proteins into the thylakoidsuch a gene location might be functionally related. It wouldsupport the proposal that TatC forms the core structure ofthe channel whereas TatA acts in the cytosol or at thecytosolic side of the membrane.

In the gating-pushing model (Figure 2A) or targeting-insertion-opening model (Figure 2B), the N-terminal domainand the amphipathic region of TatB are supposed to play

188 Wu et al.

an essential role. At present this is pure speculation.However, observations indeed support the importance ofthe integrity of these two regions. Substitution of the prolineresidue, which is located at position 22 immediately afterthe transmembrane domain of TatB, by a leucine residueblocks the translocation of various enzymes (Weiner et al.,1998). However, the presence in trans of the tatB genetruncated for its last 72 residues including most of the C-terminal domain is capable of re-establishing translocation.A longer truncation of the last 82 residues would abolishtranslocation of TMAO reductase and decrease that of thehydrogenase 2 (Chanal et al., 1998, Sargent et al., 1999).Therefore, the transmembrane segment and theamphipathic domain must be physically connected andcorrectly orientated to fulfil their function, whereas the C-terminal domain is less restricted.

To compensate for the rigidity of a folded protein, theTat channel should be more flexible than the SecYEGtranslocon. It would imply a higher dependence of the Tatpathway on the fluidity of the membrane. Recently weobserved that depletion of the major zwitterionicphospholipid phosphatidylethanolamine, or of the twoanionic phospholipids phosphatidylglycerol and cardiolipin,has more deleterious effects on the translocation of aprotein via the Tat pathway than that via the Sec-pathway(Mikhaleva et al., 1999).

Despite considerable progress in our knowledgeregarding the mechanism of the Tat pathway, its elucidationis still in a state of infancy and some essential questionsremain unanswered. How are these enzymes specificallytargeted to the membrane? How do bacterial cellsdistinguish a precursor lacking a cofactor from the matureenzyme competent for translocation? How does such afolded enzyme cross the membrane? Do other Tatcomponents exist? How is the Tat machinery synthesizedand assembled? What is the efficiency of the Tat system?How is the Tat pathway energized? Seeking answers tothese questions will be some of the major research goalsin the study of protein translocation in the next decade.

Acknowledgements

We are grateful to A.J.M Driessen for the informationregarding the size of the SecYEG channel prior topublication. We acknowledge TIGR and NCBI data banksfor sequences accessibility and analysis facilities.Organizations that accomplished sequencing of thegenomes used in this study and the name of the fundingagency for each TIGR project can be found at http://www.tigr.org/tdb/mdb/mdb.html. This work was supportedby grants from the Centre National de la RechercheScientifique to UPR CNRS 9043. LFW was supported by“Quality of Life and Management of Living Resources”Grant QLK3-CT-1999.

References

Adams, M.W. 1990. The structure and mechanism of iron-hydrogenases.Biochim. Biophys. Acta 1020: 115-145.

Berks, B.C. 1996. A common export pathway for proteins binding complexredox cofactors? Mol. Microbiol. 22: 393-404.

Bogsch, E., Brink, S., and Robinson, C. 1997. Pathway specificity for adelta pH-dependent precursor thylakoid lumen protein is governed by a‘Sec-avoidance’ motif in the transfer peptide and a ‘Sec-incompatible’mature protein. EMBO J. 16: 3851-3859.

Bogsch, E., Sargent, F., Stanley, N.R., Berks, B.C., Robinson, C., and

Palmer, T. 1998. An essential component of a novel bacterial protein exportsystem with homologues in plastids and mitochondria. J. Biol. Chem.273: 18003-18006.

Brink, S., Bogsch, E.G., Edwards, W.R., Hynds, P.J., and Robinson, C.1998. Targeting of thylakoid proteins by the delta pH-driven twin-argininetranslocation pathway requires a specific signal in the hydrophobic domainin conjunction with the twin-arginine motif. FEBS Lett. 434: 425-430.

Brüser, T., Deutzmann, R., and Dahl, C. 1998. Evidence against the double-arginine motif as the only determinant for protein translocation by a novelSec-independent pathway in Escherichia coli. FEMS Microbiol. Lett. 164:329-336.

Carter Jr., C.W., Kraut, J., Freer, S.T., Nguyen Huu, X., Alden, R.A., andBartsch, R.G. 1974. Two-Angstrom crystal structure of oxidizedChromatium high potential iron protein. J. Biol. Chem. 249: 4212-4225.

Chaddock, A.M., Mant, A., Karnauchov, I., Brink, S., Herrmann, R.G.,Klösgen, R.B., and Robinson, C. 1995. A new type of signal peptide:central role of a twin-arginine motif in transfer signals for the delta pH-dependent thylakoidal protein translocase. EMBO J. 14: 2715-2722.

Chanal, A., Santini, C.-L., and Wu, L.-F. 1998. Potential receptor functionof three homologous components, TatA, TatB and TatE, of the twin-argininesignal sequence-dependent metalloenzyme translocation pathway inEscherichia coli. Mol. Microbiol. 30: 674-676.

Cristobal, S., de Gier, J.-W., Nielsen, H., and von Heijne, G. 1999.Competition between Sec- and TAT-dependent protein translocation inEscherichia coli. EMBO J. 18: 2982-2990.

Czjzek, M., Dos Santos, J.-P., Pommier, J., Giordano, G., Méjean, V., andHaser, R. 1998. Crystal structure of oxidized trimethylamine N-oxidereductase from Shewanella massilia at 2.5 angstrom resolution. J. Mol.Biol. 284: 435-447.

Dalbey, R.E., and Robinson, C. 1999. Protein translocation into and acrossthe bacterial plasma membrane and the plant thylakoid membrane. TrendsBiochem. Sci. 24: 17-22.

Felsenstein, J. 1989. PHYLIP — Phylogeny Inference Package (Version3.2). Cladistics 5: 164-166.

Galtier, N., Gouy, M., and Gautier, C. 1996. SEAVIEW and PHYLO_WIN:two graphic tools for sequence alignment and molecular phylogeny.Comput. Appl. Biosci. 12: 543-548.

Gross, R., Simon, J., and Kroger, A. 1999. The role of the twin-argininemotif in the signal peptide encoded by the hydA gene of the hydrogenasefrom Wolinella succinogenes. Arch. Microbiol. 172: 227-32.

Halbig, D., Hou, B., Freudl, R., Sprenger, G.A., and Klösgen, R.B. 1999a.Bacterial proteins carrying twin-R signal peptides are specifically targetedby the delta pH-dependent transport machinery of the thylakoid membranesystem. FEBS Lett. 447: 95-98.

Halbig, D., Wiegert, T., Blaudeck, N., Freudl, R., and Sprenger, G.A. 1999b.The efficient export of NADP-containing glucose-fructose oxidoreductaseto the periplasm of Zymomonas mobilis depends both on an intact twin-arginine motif in the signal peptide and on the generation of a structuralexport signal induced by cofactor binding. Eur. J. Biochem. 263: 543-551.

Hamman, B.D., Chen, J.C., Johnson, E.E., and Johnson, A.E. 1997. Theaqueous pore through the translocon has a diameter of 40-60 angstromduring cotranslational protein translocation at the ER membrane. Cell89: 535-544.

Henry, R., Carrigan, M., McCaffrey, M., Ma, X., and Cline, K. 1997. Targetingdeterminants and proposed evolutionary basis for the Sec and the DeltapH protein transport systems in chloroplast thylakoid membranes. J. CellBiol. 136: 823-832.

Henry, R., Kapazoglou, A., McCaffery, M., and Cline, K. 1994. Differencesbetween lumen targeting domains of chloroplast transit peptides determinepathway specificity for thylakoid transport. J. Biol. Chem. 269: 10189-10192.

Jacobson, R.H., Zhang, X.-J., DuBose, R.F., and Matthews, B.W. 1994.Three-dimensional structure of β-galactosidase from E. coli. Nature 369:761-766.

Kingston, R.L., Scopes, R.K., and Baker, E.N. 1996. The structure ofglucose-fructose oxidoreductase from Zymomonas mobilis: anosmoprotective periplasmic enzyme containing non-dissociable NADP.Structure 4: 1413-28.

Lang, B.F., Burger, G., O’Kelly, C.J., Cedergren, R., Golding, G.B., Lemieux,C., Sankoff, D., Turmel, M., and Gray, M.W. 1997. An ancestralmitochondrial DNA resembling a eubacterial genome in miniature. Nature387: 493-497.

Michl, D., Robinson, C., Shackleton, J.B., Herrmann, R.G., and Klosgen,R.B. 1994. Targeting of proteins to the thylakoids by bipartitepresequences: CFoII is imported by a novel, third pathway. EMBO. J. 13:1310-1317.

Mikhaleva, N.I., Santini, C., Giordano, G., Nesmeyanova, M.A., and Wu,L.-F. 1999. Requirement for phospholipids of the translocation of the

Bacterial Mtt/Tat Pathway 189

trimethylamine N-oxide reductase through the Tat pathway in Escherichiacoli. FEBS Lett. 463: 331-335.

Mori, H., and Cline, K. 1998. A signal peptide that directs non-Sec transportin bacteria also directs efficient and exclusive transport on the thylakoid∆pH pathway. J. Biol. Chem. 273: 11405-11408.

Nivière, V., Wong, S.L., and Voordouw, G. 1992. Site-directed mutagenesisof the hydrogenase signal peptide consensus box prevents export of a β-lactamase fusion protein. J. Gen. Microbiol. 138: 2173-2183.

Pugsley, A.G. 1993. The complete general secretory pathway in gram-negative bacteria. Microbiol. Rev. 57: 50-108.

Rodrigue, A., Boxer, D.H., Mandrand-Berthelot, M.A., and Wu, L.-F. 1996.Requirement for nickel of the transmembrane translocation of NiFe-hydrogenase 2 in Escherichia coli. FEBS Lett 392: 81-86.

Rodrigue, A., Chanal, A., Beck, K., Muller, M., and Wu, L.-F. 1999. Co-translocation of a periplasmic enzyme complex by a hitchhiker mechanismthrough the bacterial Tat pathway. J. Biol. Chem. 274: 13223-13228.

Saitou, N., and Nei, M. 1987. The neighbor-joining method: a new methodfor reconstructing phylogenetic trees. Mol. Biol. Evol. 4: 406-425.

Santini, C.L., Ize, B., Chanal, A., Müller, M., Giordano, G., and Wu, L.-F.1998. A novel Sec-independent periplasmic protein translocation pathwayin Escherichia coli. EMBO. J. 17: 101-112.

Sargent, F., Bogsch, E.G., Stanley, N.R., Wexler, M., Robinson, C., Berks,B.C., and Palmer, T. 1998. Overlapping functions of components of abacterial Sec-independent protein export pathway. EMBO.J. 17: 3640-3650.

Sargent, F., Stanley, N.R., Berks, B.C., and Palmer, T. 1999. Sec-independent Protein Translocation in Escherichia coli: A distinct and pivotalrole for the tatB protein. J. Biol. Chem. 274: 36073-36082.

Schatz, G., and Dobberstein, B. 1996. Common principles of proteintranslocation across membranes. Science 271: 1519-1526.

Settles, A.M., and Martienssen, R. 1998. Old and new pathways of proteinexport in chloroplasts and bacteria. Trends Cell Biol. 8: 494-501.

Settles, A.M., Yonetani, A., Baron, A., Bush, D.R., Cline, K., and Martienssen,R. 1997. Sec-independent protein translocation by the maize Hcf106protein. Science 278: 1467-1470.

Thompson, J.D., Higgins, D.G., and Gibson, T.J. 1994. CLUSTAL W:improving the sensitivity of progressive multiple sequence alignmentthrough sequence weighting, position-specific gap penalties and weightmatrix choice. Nucleic Acids Res. 22: 4673-4680.

Thony-Meyer, L. 1997. Biogenesis of respiratory cytochromes in bacteria.Microbiol. Mol. Biol. Rev. 61: 337-376.

Volbeda, A., Charon, M.H., Piras, C., Hatchikian, E.C., Frey, M., andFontecilla-Camps, J. C. 1995. Crystal structure of the nickel-ironhydrogenase from Desulfovibrio gigas. Nature 373: 580-587.

Voordouw, G. 1992. Evolution of hydrogenase genes. Adv. Inorg. Chem.38: 397-422.

Weiner, J.H., Bilous, P.T., Shaw, G.M., Lubitz, S.P., Frost, L., Thomas, G.H.,Cole, J., and Turner, R.J. 1998. A novel and ubiquitous system formembrane targeting and secretion of cofactor-containing proteins. Cell93: 93-101.

Wexler, M., Bogsch, E.G., Klösgen, R.B., Palmer, T., Robinson, C., andBerks, B.C. 1998. Targeting signals for a bacterial Sec-independent exportsystem direct plant thylakoid import by the delta pH pathway. FEBS Lett.431: 339-342.

Wiegert, T., Sahm, H., and Sprenger, G.A. 1997. Expression of theZymomonas mobilis gfo gene for NADP-containing glucose:fructoseoxidoreductase (GFOR) in I. Formation of enzymatically active preGFORbut lack of processing into a stable periplasmic protein. Eur. J. Biochem.244: 107-112.

Wu, L.-F., Chanal, A., and Rodrigue, A. 2000. Membrane targeting andtranslocation of bacterial hydrogenases. Arch. Microbiol. in press.

Wu, L.-F., and Mandrand, M.A. 1993. Microbial hydrogenases: primarystructure, classification, signatures and phylogeny. FEMS Microbiol. Rev.10: 243-269.

190 Wu et al.