![[Feedback]](/icons/banner/feedbackACT.gif)
![[For Subscribers]](/icons/banner/subscriberACT.gif)
![[Archive]](/icons/banner/archiveACT.gif)
![[Search]](/icons/banner/searchACT.gif)
![[Contents]](/icons/banner/tocACT.gif)
Previous Article | Next Article 
Applied and Environmental Microbiology, November 2006, p. 7422-7426, Vol. 72, No. 11
0099-2240/06/$08.00+0 doi:10.1128/AEM.01550-06
Copyright © 2006, American Society for Microbiology. All Rights Reserved.
Genetic Characterization of the Phenylacetyl-Coenzyme A Oxygenase from the Aerobic Phenylacetic Acid Degradation Pathway of Escherichia coli
Cristina Fernández,
Abel Ferrández,,
Baltasar Miñambres,
Eduardo Díaz, and
José L. García*
Department of Molecular Microbiology, Centro de Investigaciones Biológicas, Consejo Superior de Investigaciones Científicas, Madrid, Spain
Received 5 July 2006/
Accepted 11 September 2006
 |
ABSTRACT
|
|---|
We show here that the paaABCDE genes of the paa cluster responsible for phenylacetate degradation in Escherichia coli W encode a five-component oxygenase that hydroxylates phenylacetyl-coenzyme A (CoA), the first intermediate of the pathway. The primary structure of the subunits of bacterial phenylacetyl-CoA oxygenases revealed that these enzymes constitute the prototype of a new and distinct group of the large bacterial diiron multicomponent oxygenase family.
|
INTRODUCTION
|
|---|
The paa pathway is the core of the phenylacetyl-coenzyme A (PhAc-CoA) catabolon, a functional unit that integrates several peripheral catabolic pathways and that has been described in many bacteria (1, 2, 5, 12-15, 17). The gene cluster involved in phenylacetyl acid (PhAc) degradation in Escherichia coli W contains 14 genes organized in two divergent catabolic operons, paaABCDEFGHIJK and paaZ, and a regulatory operon, paaXY, in which PaaX acts as a transcriptional repressor that becomes inactivated by PhAc-CoA (2, 4, 5). A catabolic scheme where PhAc-CoA undergoes ring hydroxylation, ring opening, and further ß-oxidation-type degradation through a proposed pathway that involves CoA thioesters and converges with the classical ß-ketoadipate pathway at the ß-ketoadipyl-CoA intermediate has been suggested (9). Based on sequence homology, it had been proposed that the paaABDCE genes might encode a ring hydroxylation system (5). Moreover, it had been also shown that when these genes were cloned together with paaK, which encodes a PhAc-CoA ligase, PhAc could be transformed in vivo into 2-hydroxy-PhAc (2OHPhAc) (5), but a precise demonstration that all these genes are required for ring hydroxylation remained to be presented. It is worth noting that 2OHPhAc is not a true intermediate in the PhAc catabolic pathway, and the 1,2-dihydroxy-1,2-dihydroPhAc-CoA has been proposed as the real product of the hydroxylation reaction (9), with 2OHPhAc being the result of a dehydratation of the dihydrodiol-CoA intermediate (5, 9, 12, 17).
|
The paaABCDE gene products are involved in aromatic ring hydroxylation.
|
|---|
To determine if all the paaABCDE gene products are involved in the PhAc-CoA ring hydroxylation complex, we used a genetic approach based on two compatible plasmids, pAFK3 (and its derivatives) and pFBP (and its derivatives) (see Fig. 1 for plasmid maps). In brief, plasmid pAFK3 is a pSJ19Not derivative (5) containing the paaK gene that encodes the PhAc-CoA ligase and the paaXY genes encoding the system that represses the Pa promoter, which controls the expression of the paaABCDEFGHIJK catabolic operon. On the other hand, plasmid pFBP is a pUC18 derivative that harbors the paaABCDE genes encoding the putative PhAc-CoA hydroxylase activity under the control of the Pa promoter. When cells of Escherichia coli W14 (
paa), an E. coli W derivative strain lacking the paa genes (5), harboring plasmids pAFK3 and pFBP were cultured in M63 minimal medium (18) with glycerol as the sole carbon and energy source, plus PhAc as inducer of the Pa promoter and substrate of the hydroxylation complex, we observed that PhAc was converted to 2OHPhAc, which was secreted into the culture medium (Table 1). Interestingly, when the same experiment was performed with E. coli W14(pFBP), i.e., in the absence of plasmid pAFK3 expressing the PhAc-CoA ligase, we could not observe 2OHPhAc in the culture medium (Table 1), confirming that the synthesis of PhAc-CoA is mandatory to generate the true substrate of the hydroxylation complex encoded by all or some of the paaABCDE genes (5).
View larger version (21K):
[in this window]
[in a new window]
|
FIG. 1. Genetic constructions used in this work. (A) Maps of plasmids pAFK3 and pFBP and their derivatives. (B) Maps of plasmids pAFBCK and pAFBK2. (C) Maps of plasmids pAFE, pAAD, and pAAD-E. Arrows indicate the direction of gene transcription. Plasmid pAAD-E was constructed by XmnI (X) digestion and religation of plasmid pAAD::Tn1000-83 (5), which eliminates most of the Tn1000 transposon and part of the paaE gene. Plasmids pAAD, pCK01, pUC18, and pAFK3 (a pSJ19Not derivative) have been previously described (5). Plasmids of the pAFD(A-E) series, pFBP, and pFBE are pUC18 derivatives. Plasmids of the pAF(A-E)K series are pAFK3 derivatives. Plasmids pAFBCK and pAFBK2 are pCK01 derivatives. Abbreviations: Apr, ampicillin resistance; Cmr, chloramphenicol resistance. Plac, lac promoter. Pa, Px, and Pz are the promoters of the E. coli paa cluster.
|
|
We then determined whether all five of the paaABCDE genes were required for the PhAc-CoA ring hydroxylation activity by constructing five pFBP derivative plasmids carrying individual deletions in each gene (Fig. 1A). These plasmids were used to transform E. coli W14(pAFK3), and the resulting recombinants were tested for the production of 2OHPhAc as described above. The E. coli W14(pAFK3) cells transformed with plasmid pAFDA (paaA paaBCDE), pAFDB (paaA paaB paaCDE), pAFDC (paaAB paaC paaDE), or pAFDD (paaABC paaD paaE) were unable to convert PhAc into 2OHPhAc (Table 1). However, cells harboring plasmid pAFDE (paaABCD paaE) showed a low but detectable hydroxylase activity (Table 1). Nevertheless, to eliminate any polar effect of the mutations generated in the five pFBP derivative plasmids and, thus, to confirm that the only gene affected in the mutant plasmids was the corresponding deleted gene, we constructed five pAFK3 derivatives carrying, in addition to the paaK and paaXY genes, each one of the five paaABCDE genes expressed under control of the Plac promoter (Fig. 1A). When E. coli W14 cells were transformed with plasmid pAFDA, pAFDB, pAFDC, pAFDD, or pAFDE and the corresponding pAFK3 derivative, the resulting double transformants, with the sole exception of that harboring plasmid pAFBK, which contains the paaB gene, acquired the ability to efficiently hydroxylate PhAc-CoA with the concomitant release of 2OHPhAc into the culture medium (Table 1). A possible explanation for the lack of complementation of plasmid pAFDB (paaA paaB paaCDE) by the paaB gene expressed in trans from plasmid pAFBK could be that deletion of paaB might affect the translation of the downstream paaC gene. This assumption was based on the observation that the paaC gene has no consensus ribosome binding site (5), which might explain why the PaaC protein requires a cotranslation mechanism to be efficiently synthesized. To test this hypothesis, we constructed two similar plasmids expressing either the paaB and paaC genes from the same promoter (pAFBCK) or just the paaB gene (pAFBK2) (Fig. 1B). Whereas E. coli W14(pAFDB, pAFBCK) coexpressing the paaB and paaC genes was able to convert PhAc into 2OHPhAc, E. coli W14(pAFDB, pAFBK2) expressing only the paaB gene did not transform PhAc (Table 1), strongly suggesting that PaaC must be cotranslated with PaaB. It is worth noting here that the pAFCK plasmid that expresses the paaC gene alone is able to complement E. coli W14(pAFDC) lacking the paaC gene, because in plasmid pAFCK the PaaC protein is cotranslated with the N-terminal region of the LacZ protein.
The observation that 2OHPhAc was formed in E. coli W14 containing the cloned paaABCD genes in the absence of paaE (Table 1) prompted us to study whether PaaE is not strictly required for the hydroxylation of PhAc-CoA in our recombinant (multicopy) system or, on the contrary, whether its function could be replaced by a different E. coli gene product. Thus, when we compared the high hydroxylation activity observed in E. coli W14(pAFDE, pAFEK), a strain that expresses the cloned paaE gene from plasmid pAFEK, with the low activity detected in E. coli W14(pAFDE, pAFK3), a strain that does not contain the paaE gene (Table 1), we could conclude that PaaE was indeed essential for efficient hydroxylation of PhAc-CoA. To study whether the activity that accounts for the 2OHPhAc formation in our recombinant system could also support the growth in PhAc of an E. coli strain lacking paaE, we constructed plasmid pAAD-E, which harbors the complete set of paa genes with the sole exception of a disrupted paaE gene (paaZ paaABCD paaE paaFGHIJK paaXY) (Fig. 1C). Whereas E. coli W14 containing plasmid pAAD, which harbors a wild-type paa cluster (5), was able to grow in M63 minimal medium using PhAc (5 mM) as the sole carbon and energy source, the recombinant strain E. coli W14(pAAD-E) did not grow. To rule out the possibility that the deletion of the paaE gene in plasmid pAAD-E could cause polar effects on the expression of other paa genes, the recombinant E. coli W14(pAAD-E) strain was transformed with plasmid pAFE, a pUC18 derivative which expresses the paaE gene under control of the Plac promoter (Fig. 1C). As expected, E. coli W14(pAAD-E, pAFE) grew in M63 minimal medium using PhAc (5 mM) as the sole carbon and energy source. All these results taken together demonstrate that the paaE gene is essential for growth of E. coli with PhAc as well as for efficient hydroxylase activity when the paaABCD genes are expressed from a multicopy plasmid.
|
Evolutionary considerations on the PhAc-CoA oxygenase.
|
|---|
So far, only three ring hydroxylation oxygenases acting on aromatic CoA-acylated compounds have been described in the literature (7, 8, 19), but the PhAc-CoA oxygenase differs from these, since it is the first reported multicomponent enzyme system that hydroxylates the aromatic ring of a CoA-thioester.
Remarkably, the paaABCD gene organization is conserved in all bacteria possessing the paa-encoded PhAc degradation pathway (1, 2, 12, 14, 15) (http://www.tigr.org/tigr-scripts/CMR2/CMRHomePage.spl). However, although paaE is usually found adjacent to the paaABCD genes in the paa clusters of many bacteria, a paaE homologue is lacking in paa clusters from ancestral bacterial groups such as the Firmicutes group and the Thermus/Deinococcus group. This finding suggests that the paaE gene has been acquired during the evolution of the paa clusters by actinobacteria and gram-negative bacteria to guarantee an efficient hydroxylation of PhAc-CoA (see above).
Sequence comparisons of the paaABCDE gene products strongly suggest that the PhAc-CoA oxygenase belongs to the bacterial diiron multicomponent oxygenases (BMOs) family, which can be divided into five groups (11, 16). All these enzymes contain a heteromultimeric (
ß)2 (groups 4 and 5) or (ß
)2 (groups 1, 2, and 3) oxygenase component that is responsible for the hydroxylation activity. The evolutionary fingerprint of the different groups of BMOs is also based on the arrangement of the coding sequences within the corresponding operons (Fig. 2A) (11, 16). Sequence comparisons suggest that PaaACD might constitute the , ß, and subunits of the heteromultimeric diiron oxygenase component of the PhAc-CoA oxygenase. PaaB and PaaE can be the effector protein and the oxidoreductase, respectively, that mediate electron transfer from NAD(P)H. Although a similar five-subunit composition has been reported for the group 3 BMOs, the order of the genes encoding the five subunits of methane monooxygenases in the mmo operon (mmoXYBZC) differs from that of the paaABCDE genes (Fig. 2A).
View larger version (13K):
[in this window]
[in a new window]
|
FIG. 2. Phylogenetic trees of the subunits of BMOs. The bar represents 1 inferred amino acid substitution per 10 amino acids. Phylogenetic analyses were carried out according to the two-parameter method of Kimura (10), and a tree was reconstructed using the neighbor-joining method of the Phylip program (3). (A) Phylogenetic tree of the subunits of representative members of different groups of BMOs. A schematic representation of the gene arrangement for the different groups of BMOs is also shown. Genes encoding the , ß, and subunits of the oxygenase component are indicated with black blocks. The genes encoding the reductase component (N) are indicated with stippled blocks. The genes encoding the effector protein (R) are indicated with hatched blocks. The genes encoding the Fp and T proteins in BMOs of groups 1 and 2, respectively, are shown with empty blocks. The amino acid lengths of the subunits are indicated in brackets below the corresponding proteins. The gene clusters shown are the following (GenBank accession numbers are in parentheses): dmpKLMNOP from Pseudomonas sp. strain CF600 (M60276), touABCDEF from P. stutzeri OX1 (AY621080), mmoXYBZorfYmmoC from Methylococcus capsulatus (M90050), amoABCD from Nocardia coralline B-276 (D37875), thmADBCorfQ from Pseudonocardia sp. strain K1 (AJ296087), and paaABCDE from E. coli W (X97452). The subunits used for the construction of the phylogenetic tree are the following: DmpN (P19732) and TomA3 (AAL50373) of the phenol monooxygenase and toluene 2-monooxygenase from Pseudomonas sp. strain CF600 and Burkholderia cepacia G4, respectively; TouA (CAA06654) and TmoA (AAS66660) of the toluene/o-xylene monooxygenase and the toluene 4-monooxygenase from P. stutzeri OX1 and P. mendocina KR1, respectively; MmoX (AAB62392), of the methane monooxygenase from M. capsulatus; AmoC (BAA07114) of the alkene monooxygenase from N. coralline B-276; ThmA (CAC10506) of the tetrahydrofuran monooxygenase from Pseudonocardia sp. strain K1; and PaaA (CAA66090) of the phenylacetyl-CoA oxygenase from E. coli W. (B) Phylogenetic tree of the PaaA subunits of phenylacetyl-CoA oxygenases from E. coli (CAA66090), P. putida (AAC24334), Bordetella pertussis (NP_881296), Azoarcus evansii (AAG28968), Dechloromonas aromatica (YP_283606), Ralstonia eutropha (YP_297410), Rhodopseudomonas palustris (NP_949105), Sinorhizobium meliloti (CAC49959), Silicibacter pomeroyi (YP_283606), Bacillus halodurans (NP_241062), Geobacillus kaustophilus (YP_147897), Streptomyces coelicolor (NP_631518), Corynebacterium efficiens (NP_737277), Nocardia farcinica (YP_118371), Thermus thermophilus (YP_004583), and Deinococcus radiodurans (AAF11931).
|
|
A detailed amino acid sequence analysis of each of the five subunits of the PaaABCDE oxygenase from different bacteria (see above) revealed that the equivalent subunits from different organisms showed an extensive conservation of residues (data not shown). This observation is in agreement with the fact that PhAc-CoA oxygenases share a common ancestor, and they constitute a widespread new enzymatic system to hydroxylate PhAc-CoA in bacteria. However, the primary structure of the PaaABCDE subunits clearly differs from that of equivalent subunits of previously reported BMOs. Thus, the subunits (PaaA proteins) of PhAc-CoA oxygenases show a protein length, around 310 to 320 amino acids, that is significantly shorter than that of the subunits, around 500 amino acids, from other BMOs (Fig. 2A). Whereas subunits of most BMOs show the consensus patterns EX2AX25DEXRH and EX18GDX8SX3DEXRH (iron cluster ligands are in boldface; an X means any residue) for Fe1 and Fe2 binding (11), the PaaA subunits show different consensus patterns, EX2GX25DEXGH and DX13SYX8RX3EEXFH, respectively, although they still maintain the conserved EX2H motif. Interestingly, the C-terminal end of the subunits of most BMOs is lacking in the PaaA proteins. A phylogenetic analysis based on the amino acid sequences of the subunits from different BMOs reveals that PaaA proteins constitute a different branch of the phylogenetic tree (Fig. 2A).
The suggested translational coupling of PaaC with the PaaB protein in E. coli (see above) is in agreement with the predicted essential role of PaaB on the activity of the PaaACD oxygenase component. Unlike the oxygenase component, the oxidoreductase subunit of BMOs is closely related to isofunctional subunits of Rieske-type oxygenases (11). Thus, the oxidoreductase subunit comprises an N-terminal plant-type ferredoxin domain and a C-terminal reductase domain with flavin adenine dinucleotide- and NAD(P)-ribose binding regions, which corresponds to the molecular architecture of the class IB reductases according to the Batie classification system (11, 16). Interestingly, the predicted oxidoreductase subunit of the PhAc-CoA oxygenases (PaaE proteins) belongs to class IA reductases, since it has an N-terminal reductase domain and a C-terminal plant-type ferredoxin domain (5). Therefore, whereas BMOs reported so far have recruited class IB reductases, the PhAc-CoA multicomponent oxygenase has recruited a class IA oxidoreductase, suggesting a distinct evolutionary history for the latter.
|
Conclusions.
|
|---|
The data presented here suggest that the PhAc-CoA oxygenase constitutes the first member of a new and distinct group of BMOs. The phylogenetic tree of the subunits of PhAc-CoA oxygenases from different bacteria reveals the existence of three major branches, i.e., Deinococcus/Thermus, Firmicutes, and Proteobacteria/Actinobacteria (Fig. 2B). The same phylogenetic distribution can be extended to the other subunits of the PhAc-CoA oxygenases (data not shown). This observation, together with the facts that the paaABCDE genes are chromosomally located and have an average G+C content that corresponds with that of the host bacterium, suggest that the genes encoding the multicomponent PhAc-CoA oxygenase have been the subject of vertical rather than horizontal transmission. The hydroxylation of PhAc-CoA by a five-subunit oxygenase appears to be a general principle in bacteria that degrade PhAc and/or some other aromatics of the PhAc-CoA catabolon, since alternative PhAc degradation pathways have not been reported so far. The PhAc-CoA oxygenase constitutes, therefore, the first reported bacterial multicomponent oxygenase acting on a CoA-activated aromatic compound. Interestingly, although all BMOs described so far are monooxygenases, the proposed product of the reaction catalyzed by the PhAc-CoA oxygenase is a dihydrodiol, and therefore this enzyme could be a hydroxylating dioxygenase (9). Nevertheless, the characterization of the reaction product and whether the PhAc-CoA oxygenase is a monooxygenase or a dioxygenase will require further research.
|
ACKNOWLEDGMENTS
|
|---|
This work was supported by grants QLRT-2001-02884, GEN2001-4698-C05-02, and BIO2003-05309-C04-02.
We gratefully acknowledge the help of A. Díaz, S. Carbajo, M. Cayuela, and I. Alonso.
|
FOOTNOTES
|
|---|
* Corresponding author. Mailing address: Department of Molecular Microbiology, Centro de Invetigaciones Biológicas, Consejo Superior de Investigaciones Científicas, Ramiro de Maeztu 9, 28040 Madrid, Spain. Phone: 34-918373112. Fax: 34-915360432. E-mail: JLGARCIA{at}CIB.CSIC.ES. 
Published ahead of print on 22 September 2006.
C. Fernández and A. Ferrández contributed equally to this work.
Present address: DSM Nutritional Products Ltd., Basel, Switzerland.
Present address: Estación Agrícola Experimental, Consejo Superior de Investigaciones Científicas, Grulleros, León, Spain.
|
REFERENCES
|
|---|
- Alonso, S., D. Bartolomé-Martín, M. del Álamo, E. Díaz, J. L. García, and J. Perera. 2003. Genetic characterization of the styrene lower catabolic pathway of Pseudomonas sp. strain Y2. Gene 319:71-83.[CrossRef][Medline]
- Díaz, E., A. Ferrández, M. A. Prieto, and J. L. García. 2001. Biodegradation of aromatic compounds by Escherichia coli. Microbiol. Mol. Biol. Rev. 65:523-569.[Abstract/Free Full Text]
- Felsenstein, J. 1993. PHYLIP (phylogenetic inference package) version 3.5.1. Department of Genetics, University of Washington, Seattle, Wash.
- Ferrández, A., J. L. García, and E. Díaz. 2000. Transcriptional regulation of the divergent paa catabolic operons for phenylacetic acid degradation in Escherichia coli. J. Biol. Chem. 275:12214-12222.[Abstract/Free Full Text]
- Ferrández, A., B. Miñambres, B. García, E. R. Olivera, J. M. Luengo, J. L. García, and E. Díaz. 1998. Catabolism of phenylacetic acid in Escherichia coli. Characterization of a new aerobic hybrid pathway. J. Biol. Chem. 273:25974-25986.[Abstract/Free Full Text]
- Gibbs, H. D. 1927. The indophenol test. J. Biol. Chem. 72:649-664.[Free Full Text]
- Hartmann, S., C. Hultschig, W. Eisenreich, G. Fuchs, A. Bacher, and S. Ghisla. 1999. NIH shift in flavine dependent monooxygenation: mechanistic studies with 2-aminobenzoyl-CoA monooxygenase reductase. Proc. Natl. Acad. Sci. USA 96:7831-7836.[Abstract/Free Full Text]
- Ishiyama, D., D. Vujaklija, and J. Davies. 2004. Novel pathway of salicylate degradation by Streptomyces sp. strain WA46. Appl. Environ. Microbiol. 70:1297-1306.[Abstract/Free Full Text]
- Ismail, W., M. El-Said Mohamed, B. L. Wanner, K. A. Datsenko, W. Eisenreich, F. Rohdich, A. Bacher, and G. Fuchs. 2003. Functional genomics by NMR spectroscopy phenylacetate catabolism in Escherichia coli. Eur. J. Biochem. 270:3047-3054.[Medline]
- Kimura, M. 1980. A simple method for estimating evolutionary rates of base substitutions through comparative studies of nucleotide sequences. J. Mol. Evol. 16:111-120.[CrossRef][Medline]
- Leahy, J. G., P. J. Batchelor, and S. M. Morcomb. 2003. Evolution of the soluble diiron monooxygenases. FEMS Microbiol. Rev. 27:449-479.[CrossRef][Medline]
- Luengo, J. M., J. L. Garcia, and E. R. Olivera. 2001. The phenylacetyl-CoA catabolon: a complex catabolic unit with broad biotechnological applications. Mol. Microbiol. 39:1434-1442.[CrossRef][Medline]
- Martínez-Blanco, H., A. Reglero, L. B. Rodríguez-Aparicio, and J. M. Luengo. 1990. Purification and biochemical characterization of phenylacetyl-CoA ligase from Pseudomonas putida. A specific enzyme for the catabolism of phenylacetic acid. J. Biol. Chem. 265:7084-7090.[Abstract/Free Full Text]
- Mohamed, M. E., W. Ismail, J. Heider, and G. Fuchs. 2002. Aerobic metabolism of phenylacetic acids in Azoarcus evansii. Arch. Microbiol. 178:180-192.[CrossRef][Medline]
- Navarro-Llorens, J. M., M. A. Patrauchan, G. R. Stewart, J. E. Davies, L. E. Eltis, and W. W. Mohn. 2005. Phenylacetate catabolism in Rhodococcus sp. strain RHA1: a central pathway for degradation of aromatic compounds. J. Bacteriol. 187:4497-4504.[Abstract/Free Full Text]
- Notomista, E., A. Lahm, A. Di Donato, and A. Tramontano. 2003. Evolution of bacterial and archaeal multicomponent monooxygenases. J. Mol. Evol. 56:435-445.[CrossRef][Medline]
- Olivera, E. R., B. Miñambres, B. García, C. Muñiz, M. A. Moreno, A. Ferrández, E. Díaz, J. L. García, and J. M. Luengo. 1998. Molecular characterization of the phenylacetic acid catabolic pathway in Pseudomonas putida U: the phenylacetyl-CoA catabolon. Proc. Natl. Acad. Sci. USA 95:6419-6424.[Abstract/Free Full Text]
- Sambrook, J., and D. W. Russell. 2001. Molecular cloning: a laboratory manual, 3rd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.
- Zaar, A., J. Gescher, W. Eisenreich, A. Bacher, and G. Fuchs. 2004. New enzymes involved in aerobic benzoate metabolism in Azoarcus evansii. Mol. Microbiol. 54:223-238.[CrossRef][Medline]
Applied and Environmental Microbiology, November 2006, p. 7422-7426, Vol. 72, No. 11
0099-2240/06/$08.00+0 doi:10.1128/AEM.01550-06
Copyright © 2006, American Society for Microbiology. All Rights Reserved.
Top
Abstract
Introduction
