HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 279, Issue 43, 44847-44857, October 22, 2004

This Article Abstract Full Text (PDF) All Versions of this Article: 279/43/44847    most recent M408648200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Gande, R. Articles by Eggeling, L. Search for Related Content PubMed PubMed Citation Articles by Gande, R. Articles by Eggeling, L. Social Bookmarking                      What's this?

Acyl-CoA Carboxylases (accD2 and accD3), Together with a Unique Polyketide Synthase (Cg-pks), Are Key to Mycolic Acid Biosynthesis in Corynebacterianeae Such as Corynebacterium glutamicum and Mycobacterium tuberculosis^*

Roland Gande, Kevin J. C. Gibson, Alistair K. Brown, Karin Krumbach, Lynn G. Dover, Hermann Sahm¶, Susumu Shioyama||, Tadao Oikawa||**, Gurdyal S. Besra, Lister Institute Jenner Research Fellow, and Lothar Eggeling

From the Institute for Biotechnology, Research Centre Juelich, D-52425 Juelich, Germany, the ^||Kansai University High Technology Research Center, Suita-shi, Osaka 564-8680, Japan, and the School of Biosciences, University of Birmingham, Edgbaston, Birmingham B15 2TT, United Kingdom

Received for publication, July 29, 2004

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
The Corynebacterianeae such as Corynebacterium glutamicum and Mycobacterium tuberculosis possess several unique and structurally diverse lipids, including the genus-specific mycolic acids. Although the function of a number of genes involved in fatty acid and mycolic acid biosynthesis is known, information relevant to the initial steps within these biosynthetic pathways is relatively sparse. Interestingly, the genomes of Corynebacterianeae possess a high number of accD genes, whose gene products resemble the -subunit of the acetyl-CoA carboxylase of Escherichia coli, providing the activated intermediate for fatty acid synthesis. We present here our studies on four putative accD genes found in C. glutamicum. Although growth of the accD4 mutant remained unchanged, growth of the accD1 mutant was strongly impaired and partially recovered by the addition of exogenous oleic acid. Overexpression of accD1 and accBC, encoding the carboxylase -subunit, resulted in an 8-fold increase in malonyl-CoA formation from acetyl-CoA in cell lysates, providing evidence that accD1 encodes a carboxyltransferase involved in the biosynthesis of malonyl-CoA. Interestingly, fatty acid profiles remained unchanged in both our accD2 and accD3 mutants, but a complete loss of mycolic acids, either as organic extractable trehalose and glucose mycolates or as cell wall-bound mycolates, was observed. These two carboxyltransferases are also retained in all Corynebacterianeae, including Mycobacterium leprae, constituting two distinct groups of orthologs. Furthermore, carboxyl fixation assays, as well as a study of a Cg-pks deletion mutant, led us to conclude that accD2 and accD3 are key to mycolic acid biosynthesis, thus providing a carboxylated intermediate during condensation of the mero-chain and -branch directed by the pks-encoded polyketide synthase. This study illustrates that the high number of accD paralogs have evolved to represent specific variations on the well known basic theme of providing carboxylated intermediates in lipid biosynthesis.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
The Corynebacterianeae represent a distinct and unusual group within Gram-positive bacteria, with the most prominent members being the human pathogens Mycobacterium tuberculosis and Mycobacterium leprae (1, 2). In addition, nonpathogenic bacteria belong to this taxon, such as Corynebacterium glutamicum, which is used in the industrial production of amino acids (3). A common feature to all these bacteria is that they possess unusual lipids, such as mycolic acids (1).

Mycolic acids are long chain -alkyl--hydroxylated fatty acids (R-CH(OH)-CH(R')-COOH), where R represents the meromycolate chain consisting in M. tuberculosis of up to 56 carbons possessing additional structural modifications, and R' represents a shorter aliphatic branch possessing 22–26 carbons). These two chains are then condensed together via a specialized Claisen condensation enzyme, followed by reduction to yield mature mycolic acids (4). In contrast, mycolic acids from Corynebacterium species, including C. glutamicum, represent the simplest form of these lipids, whereby two C₁₆ fatty acids condense together, followed by reduction to afford a C₃₂ mycolic acid. These -branched, -hydroxy fatty acids are found primarily as esters of the nonreducing arabinan terminus of arabinogalactan (5). In addition, mycolic acids can also be found as extractable "free" lipids within the cell wall, mainly linked to glucose and trehalose (6). These mycolic acids and their derivatives are believed to play a crucial role in the architecture of the cell envelope (1).

Interestingly, M. tuberculosis is characterized by an exceptionally high number of additional lipids and glycolipids, which are thought to aid in the persistence of the bacterium, thereby fueling the promise of the identification of new drug targets in the context of tuberculosis (7). The rich diversity of lipids present in M. tuberculosis is reflected at the genomic level by a panoply of genes involved in lipid biosynthesis (8), some of them constituting sets of paralogous genes. For instance, M. tuberculosis has 35 fadD genes (annotated as fatty acid-CoA ligases), 16 pks genes (annotated as polyketide synthases), and six accD genes (annotated as acyl-CoA carboxylases), and their detailed function in lipid biosynthesis and relevance for persistence are only now emerging (8, 9). Importantly, Corynebacterium species are considered the archetype within Corynebacterianeae, including M. tuberculosis, due to a low frequency of gene duplications and structural alterations giving rise to a strong conservation at the genomic level within this subgroup of the Corynebacterianeae (10). Together with their simpler chemical composition (as outlined above for mycolic acids), it can therefore be assumed that species of Corynebacterium have at their disposal just the core set of genes and reactions characteristic for the Corynebacterianeae. Indeed, C. glutamicum possesses just three fadD, one pks, and four accD genes (11). As a result, comparative studies using C. glutamicum have been employed in understanding the role of several M. tuberculosis proteins, e.g. Ppm1/D2 in lipoarabinomannan biosynthesis (12) and the "antigen 85" mycolyltransferases in cell wall mycolylation (13–16).

The group of four accD genes in C. glutamicum is interesting, and little is known about the function of these genes in Corynebacterianeae. These genes encode polypeptides with similarities to the -subunits of acetyl/propionyl-CoA carboxylases, and basically only one of them would be sufficient for the carboxylation of acetyl-CoA to provide malonyl-CoA for fatty acid biosynthesis. This suggests additional and specific carboxylations, probably involved in the synthesis of unusual lipids.

In this context, the still controversial issue of mycolic acid synthesis is very attractive to consider carboxylation reactions. Cell-free extracts of Corynebacterium have been shown to utilize [¹⁴C]palmitic acid (17, 18), with the newly synthesized mycolic acid exclusively labeled at C-1 and C-3 (19). The Claisen condensation reaction was hypothesized to involve a carboxylation step since it is inhibited by avidin, an inhibitor of biotin-dependent enzymes in extracts of Corynebacterium diphtheriae that produces the putative precursor of mycolic acids, 2-tetradecyl-3-oxo-octadecanoic acid (19, 20). In contrast, avidin has no effect on the Claisen condensation reaction in a cell-free extract of Corynebacterium matruchotii, which synthesizes mature mycolic acids (17, 18). What seems to support this last observation is the incorporation of [2,2-²H]palmitic acid in whole cells of C. matruchotii (21). During the preparation of this manuscript, a polyketide synthase from C. glutamicum (Cg-pks), the equivalent of M. tuberculosis pks13, that apparently plays a key role in mycolic acid biosynthesis was identified (22).

The study is centered on the four accD genes, with the latter located at the 3'-end of Cg-pks. As a result, we present our systematic study on lipid and mycolic acid biosynthesis based on defined C. glutamicum mutants, focusing on the relevance of the accD genes along with a phylogenomic analysis of these genes in Corynebacterium and Mycobacterium species whose genomes are established (8, 11, 23–25).

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
Bacterial Strains and Growth Conditions—Escherichia coli DH5mcr and C. glutamicum ATCC 13032 (the wild-type strain; referred to throughout as C. glutamicum) were grown in LB broth (Difco) at 37 and 30 °C, respectively. The mutants generated in this study were grown on LBHIS complex medium (5 g/liter tryptone, 5 g/liter NaCl, 2.5 g/liter yeast extract, 18.5 g/liter brain-heart infusion (Difco), and 90.1 g/liter sorbitol). Kanamycin and ampicillin were used at a concentration of 50 µg/ml. CGIII medium and CGXII medium (26) were used for C. glutamicum with 30 mg/liter protocatechuic acid being added to CGXII as a chelating agent. Samples for lipid analyses were prepared by harvesting cells at A = 10–15, followed by a saline wash and freezedrying. Growth comparisons of the four accD mutants were performed on unsupplemented CGXII and LBHIS plates and on LBHIS plates supplemented with 0.03% (w/v) sodium oleate, 0.03% (w/v) Tween 40, 0.04% (w/v) Brij 35, and 0.03% (w/v) butter hydrolysate.

Construction of Plasmids—To enable chromosomal inactivation of the four accD genes of C. glutamicum, internal fragments were amplified by PCR and blunt end-ligated using the Sure cloning kit (Amersham Biosciences) into the SmaI site of the non-replicative vector pK18mob (27). The primers used were accD1MuF10 (5'-GCA TGT GCA GGT GGC AAC GC-3'), accD1MuR11 (5'-GGT AAT CTT TGG AAC GGT TGC-3'), accD2MuF30 (5'-GTC ACG TGT ACT CCC CT-3'), accD2MuR31 (5'-CAA GCG AAT ACG AGG TC-3'), accD3MuF40 (5'-GTT GTA GGC GTC GCA GAT AC-3'), accD3MuR41 (5'-GCG TCC TCT GAA GAA GAG-3'), paccD4intfor (5'-TGG GGT TCA TCT GGG CAT CTC AC-3'), and paccD4intrev (5'-TGC CCC CAA CGT TTC CAT AAT CTC-3'). The inactivation vectors derived were termed pK18mobaccD1-int, pK18mobaccD2-int, pK18mobaccD3-int, and pK18mobaccD4-int.

The in-frame deletion of Cg-pks in C. glutamicum was achieved with pK19mobsacBpks. Cross-over PCR was used to enable one-step integration of fragments containing upstream and downstream sequences of Cg-pks into pK19mobsacB. In the first PCR round, two separate amplification products were generated using primer pair pCipks (5'-TGT TTA AGT TTA GTG GAT GGG AGT CGC CGC ATT GAT GAG ATT TC-3') and pCopks (5'-GGA ATT CGA CAG CGG AAG CTG ACG ACG-3') and primer pair pNopks (5'-GGA ATT CCG TTG GCA CTG CAC ACG GTG-3') and pNipks-2 (5'-CCC ATC CAC TAA ACT TAA ACA CTT CTG ATC CGA CGA TTG GCT CTG-3'). Both PCR products were purified and used as a template for amplification (using primers pCopks and pNopks) of a fragment devoid of Cg-pks that was treated with EcoRI and ligated with EcoRI-cleaved pK19mobsacB, yielding pK19mobsacBpks. The inserts in constructs used in this study were verified by sequencing.

Genomic Mutations—The non-replicative integration and deletion vectors were usually introduced via electroporation, but conjugation was used when electroporation failed. Conjugation was carried out using E. coli S17-1 as the donor, and its sensitivity to nalidixic acid (50 µg/ml) was used after plating for counterselection. Selection of recombinant C. glutamicum strains for integration of the four pK18mob-int vectors was performed using 15 µg/ml kanamycin. The correct integration of sequences into the chromosome and the absence of sequences, respectively, in the resulting recombinants were verified by PCR using two different primer pairs and by Southern analysis.

To achieve deletion of Cg-pks, plasmid pK19mobsacBpks was introduced into C. glutamicum by electroporation. Selection for resistance to kanamycin in at least 20 independent assays yielded regularly a number of clones indicating integration of the vector in the chromosome by homologous recombination. At least 100 clones derived from independent electroporation experiments were subjected to the subsequent selection for the second homologous recombination event. In this round, the presence of sacB (together with the addition of 10% sucrose to the medium) resulted in a positive selection of clones in which vector sequences were lost. Small colonies picked after 10 days were analyzed for deletion of Cg-pks. Of 60 colonies analyzed, 19 exhibited the loss of Cg-pks. One of these was further analyzed by Southern blot analysis, and the resulting strain, Cg-pks, was used in all subsequent studies.

Southern Blot Analysis—Genomic DNA was extracted from accD1–4 mutants and the Cg-pks mutant and cleaved with BstEII. The resulting fragments were separated on a 1% agarose gel and blotted onto a Nytran NY13N nitrocellulose membrane, with subsequent washings according to standard protocols. Detection was carried out with fragments of the accD genes as probes that were labeled with digoxigenin (DIG labeling and detection kit, Roche Applied Science). For accD1–4, the SacI-BglII, PvuII-SgrAI, ClaI-StuI, and SfuI-PvuI fragments were used, respectively. For Cg-pks analysis, the 645-bp fragment used was generated by PCR with primers pNopks and ppksACPintrev (5'-CAA CAT CGC GAG AGG AAA GG-3').

Extraction and Analysis of ¹⁴C-Labeled Lipids—LBHIS medium (5 ml) was inoculated with a single colony of the accD1–4 and Cg-pks mutants, respectively, and shaken at 120 rpm overnight at 30 °C. An aliquot (1 ml) of this culture was used to inoculate 5 ml of CGIII medium and grown again overnight. This second pre-culture was used to inoculate 5 ml of CGXII medium to give a starting absorbance of 0.2. Cells were incubated at 30 °C until A = 0.4, at which point, 5 µCi/ml [¹⁴C]acetate (62 mCi/mmol; Amersham Biosciences) was added, followed by overnight incubation with shaking at room temperature. Cells were harvested, washed, and freeze-dried.

Free lipids were extracted by two consecutive extractions with 2 ml of CHCl₃/CH₃OH/H₂O (10:10:3, v/v/v) for 3 h at 50 °C. These lipid extracts were combined with 1.75 ml of CHCl₃ and 0.75 ml of water, mixed, and centrifuged. The lower organic phase was recovered and washed twice with 2 ml of CHCl₃/CH₃OH/H₂O (3:47:48, v/v/v), and the resulting organic phase was dried and resuspended in 200 µl of CHCl₃/CH₃OH/H₂O (10:10:3, v/v/v). An aliquot (20,000 cpm) from each strain was subjected to TLC using Silica Gel 60 F₂₅₄ plates (Merck) developed in CHCl₃/CH₃OH/H₂O (60:16:2, v/v/v). Autoradiograms were produced by 2–3-day exposure to Kodak X-Omat AR film to reveal ¹⁴C-labeled lipids and compared with known standards (28).

The bound lipids from the delipidated extracts were released by the addition of 2 ml of a 5% aqueous solution of tetrabutylammonium hydroxide, followed by overnight incubation at 95 °C. After cooling, water (2 ml), CH₂Cl₂ (4 ml), and CH₃I (500 µl) were added and mixed thoroughly for 30 min. The lower organic phase was recovered following centrifugation, washed three times with water (4 ml), dried, and resuspended in diethyl ether (4 ml). After centrifugation, the clear supernatant was again dried and resuspended in CH₂Cl₂ (200 µl). An aliquot (10,000 cpm) from each strain was subjected to TLC using Silica Gel 60 F₂₅₄ plates developed in petroleum ether/acetone (95:5, v/v). Autoradiograms were produced by overnight exposure to Kodak X-Omat AR film to reveal ¹⁴C-labeled fatty acid and mycolic acid methyl esters and compared with known standards (21).

Acyl Carboxylation Assay—To assay acyl carboxylase activity, we adapted the method of Rainwater and Kolattukudy (29). Briefly, a clarified lysate of the appropriate strain was prepared by resuspending the bacterial cultures to a density of 0.5 g (wet weight)/ml in 0.1 M potassium phosphate buffer (pH 8.0) and 5 mM 2-mercaptoethanol. The cell paste was lysed by two passages through a French pressure cell (19,000 p.s.i., cell pre-chilled to 4 °C), and the lysate was clarified by centrifugation at 27,000 x g for 30 min at 4 °C. The supernatant was removed and stored on ice until assayed. Assay mixtures (100 µl) contained 100 mM potassium phosphate (pH 8.0), 5 mM ATP, 5 mM MgCl₂, 2% Me₂SO, 0.9 mM acyl substrate (acetyl-CoA, palmitoyl-CoA, or palmitic acid), 65 mg of bovine serum albumin, 0.5 µCi of NaH¹⁴CO₃ (0.1 Ci/mol; CFA421, Amersham Biosciences), and 250 µg of total protein from the lysate. All of the components of the assay excluding the enzyme extract were premixed and then mixed with the enzyme extract to initiate the reaction, which was held at 30 °C for 30 min. The reaction was quenched by the addition of 50 µl of concentrated HCl. This mixture was then held at 95 °C under a stream of air to remove non-fixed ¹⁴CO₂. After evaporation to dryness, the residue was dissolved in 1 ml of water and added to 10 ml of Ecoscint A (National Diagnostics, Inc.) prior to liquid scintillation counting.

Acetyl-CoA Carboxylation Assay—To assay for acetyl-CoA-dependent malonyl-CoA formation, cells grown on CGIII medium were washed twice with 0.9% NaCl, resuspended in 60 mM Tris-HCl (pH 7.2), and disrupted as described above. Assay mixtures (250 µl) contained 60 mM Tris-HCl (pH 7.2), 65 mM KHCO₃, 1mM ATP, 1.5 mM MgCl₂, and 135 µl of cell extract. The mixture was preincubated at 30 °C for 1 min, and the reaction was initiated by the addition of acetyl-CoA (lithium salt, Roche Applied Science) to give a final concentration of 2 mM and further incubated at 30 °C. Samples were taken at the given time intervals and mixed immediately with 5 µl of 30% perchloric acid to quench the reaction. The assay tubes were centrifuged subsequently (14,000 x g, 4 °C, 5 min), and 50 µl of the supernatant was withdrawn, neutralized with 12.5 µl of Na₂CO₃, and analyzed via HPLC.¹ Acyl-CoA synthesis was monitored by reversed-phase chromatography using a LiChrospher 1000 RP 18-EC-5µ column (125 x 4 mm; Merck) on an Agilent 1100 series HPLC apparatus. Samples (12 µl) were automatically injected and separated by an increasing gradient consisting of 50 mM sodium phosphate buffer (pH 5.0) containing 2% acetonitrile, increasing up to 26% acetonitrile in the same buffer at a flow rate of 0.3 ml/min, and monitored at 254 nm. Reactions were compared with known standards of acetyl-CoA, malonyl-CoA, and CoA (0.1–1 mM) dissolved in 200 mM sodium phosphate buffer (pH 3.0).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
accD Genes of Corynebacterianeae—From the recent Himar1-based transposon mutagenesis in M. tuberculosis (30), accD4 and accD6 were found to be essential genes, whereas accD1 and accD3 were found to be nonessential. In terms of accD1 (previously dtsR1) of C. glutamicum, an involvement in L-glutamate formation that is related to fatty acid synthesis was demonstrated (31). To address the problem regarding functional similarities of the carboxyltransferase proteins, we performed a phylogenomic analysis of the accD genes present in the genomes of the sequenced Corynebacterianeae (8, 11, 23–25, 32).

In fatty acid biosynthesis, the malonyl-CoA required for each elongation step is generated by carboxylation of acetyl-CoA. The process is catalyzed by an enzyme consisting of a biotinylated -subunit and a -subunit, the latter being the actual carboxyltransferase. In E. coli and Bacillus subtilis, each of these subunits consists of two polypeptides, whereas in all Corynebacterianeae analyzed, we have found that the - and -subunits consist of a single polypeptide. C. glutamicum has four putative carboxyltransferase -subunits (accD genes); its close relatives, Corynebacterium efficiens and C. diphtheriae, possess five; and M. tuberculosis, Mycobacterium bovis, and Mycobacterium marinum possess as many as six carboxyltransferases. The sequence identities of the polypeptides within one organism are also in part exceptional. For instance, AccD1 and AccD2 of M. tuberculosis share an identity of 49%, and even the least homologous pair, AccD1 and AccD3, still share 22% identity.

The primary structures of 35 polypeptides were analyzed using ClustalW (33), and the results are presented in Fig. 1. There are several remarkable features. First, there are two distinct groups (I and II) of closely related proteins in which each organism analyzed is represented only once with one carboxyltransferase. Group III is clearly distant from groups I and II, whereas the latter two are more closely related. The distinct clustering of these groups suggests that the members of each group are orthologs with an identical function. Second, there are three additional clusters in which M. bovis and M. tuberculosis are represented once. Also, M. marinum is present in two of these clusters. These results indicate that these carboxyltransferases serve more specialized functions within the mycobacterial branch of Corynebacterianeae, whereas those of the first and second clusters represent fundamental core functions specific for lipid biosynthesis in all bacteria. In addition, M. leprae possesses three pseudogenes sharing in part identities with the other carboxyltransferases (25). However, the fact that in M. leprae the two single carboxyltransferases fully retained are members of groups I and II strengthens the view that these members encode essential functions in lipid biosynthesis, as is the case with mycobacterial mycolic acid biosynthesis (4). Since carboxyltransferase activity requires an activated carboxyl group derived from a biotinylated -subunit, we searched the genomes for the corresponding proteins. The three Corynebacterium species and M. leprae have only one such subunit, whereas M. tuberculosis, M. bovis, and M. marinum possess three.

View larger version (28K): [in this window] [in a new window]   FIG. 1. Phylogenomic analysis of mycobacterial and corynebacterial acc genes. Groups of ortholog carboxyltransferases are highlighted in gray and labeled I–III. Cg, C. glutamicum; Ce, C. efficiens; Cd, C. diphtheriae; Mm, M. marinum; Mt, M. tuberculosis; Mb, M. bovis; Ml, M. leprae.

View larger version (30K): [in this window] [in a new window]   FIG. 2. A, schematic diagram of the construction of accD mutants from C. glutamicum. The scheme illustrates the situation for the two adjacent genes, accD1 and accD2. The wild-type (WT) situation is given, where the two internal fragments used for gene disruption are indicated as open boxes within the genes. Also shown are the three relevant BstEII restriction sites as well as the location of the accD1 and the accD2 probes, which are shown as hatched boxes above their complementary chromosomal DNA regions. Homologous recombination and selection for kanamycin resistance resulted in strain accD1 with disrupted accD1 (::accD1). Since the BstEII cut was present on the internal fragment, it was duplicated after correct integration of the vector, thus resulting in three fragments hybridizing with the accD1 probe. Note that there are two complementary regions of the internal fragment. One of the three fragments (3.7 kb) corresponds to the left BstEII fragment, one to the 4.4-kb middle BstEII fragment containing the vector (stippled boxes), and one to the 6.5-kb right BstEII fragment containing also accD2. With chromosomal DNA of strain accD2, the accD1 probe resulted in two fragments (::accD2) due to the BstEII cut within accD1. Therefore, also with the accD1 probe, an enlarged fragment with vector sequences was detected in the accD2 mutant. The sizes of the BstEII fragments are not drawn to scale. B, final confirmation of the constructed strains via Southern blot analyses using chromosomal DNA from the accD1–4 inactivation mutants and the wild-type (wt) strain. The lanes are marked accordingly, and the fragment sizes are given in kilobases on the right. For specific detection, the accD1 probe (upper left panel), the accD2 probe (upper right panel), the accD3 probe (lower left panel), and the accD4 probe (lower right panel) were used.

Phenotypic Characterization of Mutants—As shown in Fig. 3, the accD4 mutant (compare left and right panels) did not exhibit a phenotype, whereas growth of the accD1–3 mutants was strongly reduced on CGXII salt medium (upper row). The worst growth was in fact seen with the accD1 inactivation mutant, and the colonies visible in Fig. 3 are due to revertants. Also on LBHIS complex medium, growth of the three mutants was still impaired, illustrating that these mutants are not rescued by specific components supplied by yeast extract and tryptone, which are present in LBHIS medium. However, when this complex medium was supplemented with sodium oleate together with butter hydrolysate (and detergents for emulsification), growth of the accD1 mutant was markedly improved, whereas that of the accD2 and accD3 mutants was still impaired, indicating a direct relation of accD1 with fatty acid biosynthesis and confirming previous observations with a dtsR1 deletion mutant (31). The colony surface of both the accD2 and accD3 mutants was rougher than that of wild-type colonies, and these two mutants tended to clump in CGXII liquid culture (data not shown). Overall, the accD4 mutation has no apparent phenotype, whereas accD1 appears to be involved in lipid biosynthesis, as it was rescued by sodium oleate, suggesting that it may be involved in the synthesis of oleate from malonyl-CoA. The accD2 and accD3 mutants are likely to be very specific and essential carboxyltransferase -subunits, as also illustrated by the phylogenomic analysis.

View larger version (61K): [in this window] [in a new window]   FIG. 3. Growth of accD mutants of C. glutamicum. The accD mutants of C. glutamicum and the wild-type (wt) strain were grown for 3 days at 30 °C. The plates used were CGXII salt medium with glucose, LBHIS complex medium, and LBHIS complex medium supplemented with sodium oleate (300 mg/liter) and butter hydrolysate (300 mg/liter) together with the detergents Tween 40 (300 mg/liter) and Brij 35 (400 mg/liter). FA, fatty acids.

View larger version (41K): [in this window] [in a new window]   FIG. 4. Lipid analysis of accD mutants of C. glutamicum. A, free lipids were extracted, and an aliquot (20,000 cpm) from each strain was subjected to TLC analysis using Silica Gel 60 F254 plates developed in CHCl3/CH3OH/H2O (60:16:2, v/v/v). Autoradiograms were produced by 2–3-day exposure to Kodak X-Omat AR film to reveal 14C-labeled lipids and compared with known standards (trehalose monomycolates (TMM), trehalose dimycolates (TDM), and glucose monomycolates (GMM)) (28). B, the bound lipids from the delipidated extracts were released by the addition of tetrabutylammonium hydroxide, followed by overnight incubation at 95 °C. Following preparation of methyl esters, an aliquot (10,000 cpm) from each strain was subjected to TLC analysis using Silica Gel 60 F254 plates developed in petroleum ether/acetone (95:5, v/v). Autoradiograms were produced by overnight exposure to Kodak X-Omat AR film to reveal 14C-labeled fatty acid methyl esters (FAMES) and mycolic acid methyl esters (MAMES) and compared with known standards (21). wt, wild-type.

Carboxylation Reactions in the accD Mutants—To assess the validity of the assignment of the AccD family members as acyl carboxylase components, we assayed the fixation of ¹⁴C from radiolabeled bicarbonate in extracts of the C. glutamicum mutants. In these assays, we used acetyl-CoA and palmitoyl-CoA, as these were the obvious substrates for de novo fatty acid biosynthesis and corynomycolate condensation, respectively. The recent demonstration that the product of M. tuberculosis fadD32, which occupies the locus adjacent to pks13, activates fatty acids by adenylation (34) led us to incorporate palmitic acid into our experiments to determine whether, in the absence of commercially available palmitoyl-AMP, it may generate a suitable substrate for carboxylation. When assaying acyl carboxylase activity in clarified lysates of the parent strain, we detected that carbon fixation was 40-fold higher with acetyl-CoA than with palmitoyl-CoA and palmitic acid as acyl substrates, yielding specific activities of 2.45, 0.06, and 0.06 nmol/min/mg, respectively. The following reaction rates were normalized to these values. Basically, with each accD mutant and overexpressing strain, a significant influence on carbon fixation was apparent (Fig. 5). The most distinct effect was present for the accD4 mutant, where inactivation resulted in a uniform and significant decrease in ¹⁴C fixation, and overexpression of accD4 led to a uniform increase in activity, in both cases independent of the substrate used. The accD1 mutant also possessed lower levels of carboxyltransferase activity with respect to acetyl-CoA and palmitoyl-CoA effects; however, extracts prepared from the accD1-overexpressing strain possessed similar levels of activity in comparison with the wildtype extract. When analyzing extracts from accD2 and accD3, we observed substrate-specific variations in acyl carboxylase activity. With the accD2 inactivation mutant, acyl carboxylase activity with acetyl-CoA present was slightly reduced, whereas it was most pronounced with the C₁₆ substrates. A similar activity profile was observed with the extract of the accD3 inactivation mutant, with both observations being wholly consistent with roles for AccD2 and AccD3 specific to corynomycolate biosynthesis. When we assayed extracts from bacteria overexpressing accD2 or accD3, we observed clear differences in their substrate preference. In the extract enriched in AccD3, ¹⁴C fixation detected using acetyl-CoA was similar to that in the wild-type extract, whereas activity with C₁₆ substrates was increased. In contrast, overexpression of accD2 coincided with increased ¹⁴C fixation when using acetyl-CoA rather than the longer acyl chain length substrates.

View larger version (31K): [in this window] [in a new window]   FIG. 5. Relative acyl carboxylase activity associated with the C. glutamicum AccD family. The relative acyl carboxylase activity (percent of wild-type (wt) activity for each substrate) in extracts of C. glutamicum with either disruption (A) or overexpression (B) of the various accD family members was determined using acetyl-CoA (gray bars), palmitoyl-CoA (white bars), or palmitic acid (black bars). Carboxylation activity was followed by fixation of . Fixed radioactive carbon was measured by liquid scintillation counting after acidification of the reaction mixtures and evaporation to dryness.

View larger version (16K): [in this window] [in a new window]   FIG. 6. Malonyl-CoA formation with extracts of recombinant C. glutamicum. Malonyl-CoA synthesis (nanomoles/mg of protein) is shown in recombinant C. glutamicum pyc with the genotypes pJC1accBCpyc pEC7accD1 (), pJC1accBCpyc pEC7 (), pJC1pyc pEC7accD1 (), and the control (x) after incubation at 30 °C at the given time intervals. The inset shows comparable analyses using the separate accD1-overexpressing strain pVWEx2accD1 (•) compared with the empty vector pVWEx2 control ().

View larger version (12K): [in this window] [in a new window]   FIG. 7. The pks locus of mycolic acid-containing bacteria responsible for mycolic acid biosynthesis.

View larger version (54K): [in this window] [in a new window]   FIG. 8. Analysis of the Cg-pks deletion mutant of C. glutamicum. A, PCR analysis of the Cg-pks locus of the wild-type (wt) strain and the in-frame deletion mutant (Cg-pks). A standard was applied in the second lane, with the molecular masses given on the right in kilodaltons. Whereas with the primers annealing upstream and downstream of Cg-pks, the expected amplification product of 1.11 kb was obtained with chromosomal DNA from Cg-pks, no product could be formed under the conditions used with the wild-type DNA due to the length of Cg-pks. B, TLC analysis of extractable lipids from C. glutamicum (wt) and the deletion mutant (Cg-pks), illustrating the absence of the dominant lipids, trehalose dimycolates (TDM) and glucose monomycolates (GMM), upon Cg-pks deletion.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 
Our combined mutational and functional analyses on the accD genes, which we propose are key to lipid biosynthesis, revealed that growth of the accD1 mutant of C. glutamicum on minimal medium was severely restricted compared with the other mutants. This mutant was extremely unstable, and the phospholipids derived from the mutant culture are, in all likelihood, due to revertants. The recovery of growth by sodium oleate addition is strong evidence that AccD1 is a constituent of the acetyl carboxyltransferase enzyme, generating oleate from malonyl-CoA. This is further substantiated by the increased malonyl-CoA formation upon overexpression of accD1, yielding a higher specific activity of 40.7 nmol/min/mg compared with wild-type extracts (23.1 nmol/min/mg). In addition, when overexpression of accD1 and accBC was combined, an 8-fold increase in specific activity was observed, indicating that both polypeptides are components of acetyl-CoA carboxyltransferase, which is in agreement with an acyl-CoA carboxylase isolated from M. bovis (29), and a propionyl-CoA carboxylase from Mycobacterium smegmatis (35), both of which are composed of two individual polypeptides, with one being biotinylated.

Extracts of the accD4 mutant possessed lower levels of acyl carboxylase activity, and extracts from the accD4-overexpressing strain possessed increased levels of acyl carboxylase activity, suggesting that AccD4 is a carboxylase enzyme. However, the absence of a phenotype for the accD4 mutant, as well as its unaltered lipid profiles, led us to conclude that AccD4 is not directly involved in lipid biosynthesis. In sharp contrast are the accD2 and accD3 mutants; lipid analyses revealed that both proteins are essential for corynomycolate biosynthesis. We cannot be sure whether the general carboxylation assays reported here represent an indication of relative acyl carboxylase activity or if they simply reflect the half-life of the carboxylated products, which, we presume, are readily decarboxylated in a Claisen-type condensation reaction. However, the assays revealed a profound decrease in carboxylation activity for the accD2 and accD3 inactivation mutants with respect to the C₁₆ substrates. The data would appear wholly consistent with the proposed function of AccD2 and AccD3 in the later stages of corynomycolate biosynthesis. As expected, overexpression of accD3 led to a similar increase in acyl carboxylase activity with both C₁₆ substrates, whereas with acetyl-CoA, carboxylation activity was comparable with extracts from the parent strain. Intriguingly, overexpression of accD2 led to an increase in the acyl carboxylase activity with acetyl-CoA only. Maybe the peculiar substrate specificities implied by the above in vitro data indicate the formation of hetero-oligomeric assemblies whose composition defines substrate specificity. However, as both AccD2 and AccD3 are required for corynomycolate biosynthesis, we can rule out any complementation by AccD1 and may speculate that AccD2 and AccD3 operate as a heterodimer. Furthermore, our in vitro analyses of accD2 and accD3 overexpression strongly suggest that AccD3 might define the strict substrate specificity of such a complex. Sequence analysis of the AccD paralogs supports this hypothesis, as AccD2 is more closely related to AccD1 than AccD3.

The M. tuberculosis antigen 85 complexes, which possess mycolyltransferase activity, are located outside of the cytoplasm in the cell envelope and are often found in culture medium (13), as observed for C. glutamicum (14, 15). The localization of these enzymes suggests that they use mature mycolic acids once they are synthesized to transfer them to their final destination (5, 13–15). The mycolyltransferase locus is slightly different in the related Corynebacterium species (Fig. 7). CmytA in C. glutamicum is fused to a larger polypeptide with a domain of unknown function. In addition, a small open reading frame is present downstream (166 amino acids). A distinct possibility exists that cmytA and cmytB have different specificities, as they are conserved in Corynebacterianeae. Interestingly, downstream of the mycolyltransferase locus is a gene essential for M. tuberculosis (30). The encoded Pro-rich protein is predicted to be membrane-anchored, with the remaining segment directed toward the periplasm and exhibiting identity to an acetylxylan esterase from Penicillium purpurogenum possessing the characteristic serine esterase motif. A tempting hypothesis is that this particular gene may be responsible for the first steps of mature mycolic acid transfer from a putative carrier molecule, such as acyl carrier protein (ACP)/CoA, or from an intermediate polyprenol carrier (36) to possibly either trehalose or the cell wall arabinogalactan. This seems meaningful, as it is located in close conjunction with the mycolyltransferases and together with the fadD-pks-accD3 locus and warrants further investigation as a possible specific mycolyltransferase. Another structural characteristic is the gene arrangement of pks and its downstream accD3 gene. In the three Corynebacterium species analyzed, the genes are separated by <20 nucleotides; however, both genes overlap by 4 nucleotides in M. tuberculosis, M. bovis, and M. leprae. Such intimate structural organization has also been observed for other important genes in the Corynebacterianeae, such as ppm1, in which even protein fusion has occurred (12, 37, 38).

The -ketoacyl synthases catalyze the formation of new carbon–carbon bonds by condensation of a variety of acyl chain precursors with an elongation substrate, usually malonyl or methylmalonyl residues that are covalently attached in a thioester linkage to an ACP. The data obtained in this work, along with the structural characteristics of the fadD-pks-accD3 locus and recent findings on FadD32 (34) and pks13 (22), yield three distinct features that are very likely to represent the basis of discrete steps in the mechanism of mycolic acid biosynthesis. (i) FadD, recently shown to constitute a new class of fatty acyl-AMP ligases, activates long chain fatty acids as acyl adenylates, which are loaded onto multifunctional polyketide synthase for further chain extension (34); (ii) polyketide synthase possesses a -ketoacyl synthase domain with two phosphopantetheinyl-binding sites and a thioesterase domain; and (iii) the Acc proteins have carboxyltransferase activity, with that of AccD2 and AccD3 of C. glutamicum necessary for mycolic acid biosynthesis. An explanation for the two phosphopantetheinyl arms is that one site is occupied by the acyl chain forming the mero-chain and the other by the acyl chain resulting in the incoming -branch. The role of AccD3 as shown in these studies would be to generate an activated carboxylated acyl derivative, which would then react with the bound acyl chain to form a 3-oxo intermediate, which, after reduction, would form a mature mycolic acid (Fig. 9). The experimental data and the phylogenomic analysis indicate that in addition to AccD3 in C. glutamicum, a second Acc protein (AccD2) is also apparently required for mycolic acid biosynthesis. One possibility is that after the fixation of carbon by AccBC, this is not directly transferred via AccD3 (AccD4 in M. tuberculosis), but transmitted via AccD2 (AccD5 in M. tuberculosis). The carboxylating transferase might fall into a group of multienzyme complexes such as transcarboxylase from Propionibacterium shermanii (39), which is a 1.2-MDa multienzyme complex that couples two carboxylation reactions, transferring from methylmalonyl-CoA to pyruvate, yielding propionyl-CoA and oxalacetate. Cg-AccD3 also exhibits a high degree of identity to the 12 S subunit of this enzyme. Shuttling via intermediates between different catalytic subunits provides a complex and intriguing mechanism for regulation and modulation of this specificity. Clearly, further experiments and, in particular, an in vitro system for the Claisen condensation reaction involved in mycolic acid biosynthesis are required (studies in progress). The mechanism proposed (Fig. 9) for the Claisen condensation could well be a general mechanism of relevance for other lipids or antibiotics produced via polyketide synthases.

View larger version (17K): [in this window] [in a new window]   FIG. 9. Proposed mechanism for mycolic acid biosynthesis in Corynebacterianeae. Protein domains are shown as colored segments. ACP represents the N-terminal phosphopantetheinylated (Ppant; red bar) acyl carrier protein-like domain, and KAS represents the predicted -ketoacyl synthase domain. Step 1 depicts the charging of fatty acids (C16) by the fatty acid-ADP ligase FadD. Step 2 depicts de novo fatty acid synthesis viafatty-acid synthase Ia (FAS-Ia). The domains shown are acyltransferase (AT), enoyl reductase (ER), dehydratase (DH), malonyl/palmitoyltransferase (MPT), ACP, -ketoacyl reductase (KR), and -ketoacyl synthase (KAS). In step 3, transacylation reactions charge the phosphopantetheine residue of ACP and the active-site cysteine residue (Cys288; green bar) of the -ketoacyl synthase domain with the C16 fatty acid chains. The latter is likely to occur spontaneously, as in the case of M. tuberculosis KasA/B (40) and M. tuberculosis FabH (41), but the acylation of ACP may require an acyltransferase activity. In step 4, AccD2 and AccD3 and their biotinylated partner AccBC activate the ACP-bound C16 acyl group for condensation by -carboxylation. The removal of the carboxyl group (step 5), presumably now translocated into the active site of the -ketoacyl synthase domain, forms a carbanion that attacks the thioester bound acyl chain occupying Cys288, thus forming the oxomycolyl intermediate (step 6). The latter stages of synthesis of mycolic acid, e.g. reduction (step 7) and its transfer to extracytoplasmic trehalose and cell wall arabinan (summarized in step 8), are not certain. If the quaternary structure of polyketide synthase resembles that of fatty-acid synthase I, a head-to-tail homodimer, the following reaction series can be envisaged. The predicted thioesterase (TE) domain may transfer the mycolic acid moiety to the phosphopantetheine residue of the C-terminal ACP-like domain of the other monomer molecule, thus facilitating the transfer of the mycolate residue by the predicted acyltransferase domain to a suitable acceptor for export. The gene adjacent to those encoding the mycolyltransferases (FbpA proteins CmytA and CmytB, respectively), a putative membrane-anchored esterase, may act to transfer the mycolate residue to the mannosylated polyprenyl phosphate (Man-P-Pol) carrier described in mycobacteria (36) that presumably facilitates its export. However, other mechanisms are also possible. TMM, trehalose monomycolates; TDM, trehalose dimycolates; GMM, glucose monomycolates; mAGP, mycolylarabinogalactan-peptidoglycan complex.

	FOOTNOTES

^¶ Supported by the Fonds der Chemischen Industrie.

^** Supported by the Kansai University Overseas Research Program.

To whom correspondence should be addressed. Tel.: 121-415-8125; Fax: 121-414-5925; E-mail: g.besra{at}bham.ac.uk.

¹ The abbreviations used are: HPLC, high performance liquid chromatography; ACP, acyl carrier protein.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

 

1. Brennan, P. J., and Nikaido, H. (1995) Annu. Rev. Biochem. 64, 29-63[CrossRef][Medline] [Order article via Infotrieve]
2. Stackebrandt, E., Rainey, F. A., and Ward-Rainey, N. L. (1997) Int. J. Syst. Bacteriol. 47, 479-491[Abstract/Free Full Text]
3. Eggeling, L., Pfefferle, W., and Sahm, H. (2001) in Basic Biotechnology (Ratledge, C., and Kristiansen, B., eds) pp. 281-302, Cambridge University Press, Cambridge, United Kingdom
4. Kremer, L., Baulard, A. R., and Besra, G. S. (2000) in Molecular Genetics of Mycobacteria (Hatfull, G. F., and Jacobs, W. R., Jr., eds) pp. 173-190, ASM Press, Washington D. C.
5. McNeil, M. R., Daffe, M., and Brennan, P. J. (1991) J. Biol. Chem. 266, 13217-13223[Abstract/Free Full Text]
6. Minnikin, D. E. (1982) in The Biology of Mycobacteria (Ratledge, C., and Stanford, J., eds) pp. 95-184, Academic Press Ltd., London
7. Kremer, L. S., and Besra, G. S. (2002) Expert Opin. Investig. Drugs 11, 1033-1049[CrossRef][Medline] [Order article via Infotrieve]
8. Cole, S. T., Brosch, R., Parkhill, J., Garnier, T., Churcher, C., Harris D., Gordon, S. V., Eiglmeier, K., Gas, S., Barry, C. E., III, Tekaia, F., Badcock, K., Basham, D., Brown, D., Chillingworth, T., Connor R., Davies, R., Devlin, K., Feltwell, T., Gentles, S., Hamlin, N., Holroyd, S., Hornsby, T., Jagels, K., Krogh, A., McLean, J., Moule, S., Murphy, L., Oliver, K., Osborne, J., Quail, M. A., Rajandream, M.-A., Rogers, J., Rutter, S., Seeger, K., Skelton, J., Squares, R., Squares, S., Sulston, J. E., Taylor, K., Whitehead, S., and Barrell, B. G. (1998) Nature 393, 537-544[CrossRef][Medline] [Order article via Infotrieve]
9. Glickman, M. S., and Jacobs, W. R. (2001) Cell 104, 477-485[CrossRef][Medline] [Order article via Infotrieve]
10. Nakamura, Y., Nishio, Y., Ikeo, K., and Gojobori, T. (2003) Gene (Amst.) 317, 149-155[CrossRef][Medline] [Order article via Infotrieve]
11. Kalinowski, J., Bathe, B., Bartels, D., Bischoff, N., Bott, M., Burkovski, A., Dusch, N., Eggeling, L., Eikmanns, B. J., Gaigalat, L., Goesmann, A., Hartmann, M., Huthmacher, K., Kraemer, R., Linke, B., McHardy, A. C., Meyer, F., Mockel, B., Pfefferle, W., Puhler, A., Rey, D. A., Ruckert, C., Rupp, O., Sahm, H., Wendisch, V. F., Wiegrabe, I., and Tauch, A. (2003) J. Biotechnol. 104, 5-25[CrossRef][Medline] [Order article via Infotrieve]
12. Gibson, K. J., Eggeling, L., Maughan, W. N., Krumbach, K., Gurcha, S. S., Nigou, J., Puzo, G., Sahm, H., and Besra, G. S. (2003) J. Biol. Chem. 278, 40842-40850[Abstract/Free Full Text]
13. Belisle, J. T., Vissa, V. D., Sievert, T., Takayama, K., Brennan, P. J., and Besra, G. S. (1997) Science 276, 1420-1422[Abstract/Free Full Text]
14. De Sousa-D'Auria, C., Kacem, R., Puech, V., Tropis, M., Leblon, G., Houssin, C., and Daffe, M. (2003) FEMS Microbiol. Lett. 224, 35-44[CrossRef][Medline] [Order article via Infotrieve]
15. Brand, S., Niehaus, K., Puehler, A., and Kalinowski, J. (2003) Arch. Microbiol. 180, 33-44[CrossRef][Medline] [Order article via Infotrieve]
16. Kacem, R., De Sousa-D'Auria, C., Tropis, M., Chami, M., Gounon, P., Leblon, G., Houssin, C., and Daffe, M. (2004) Microbiology (Read.) 150, 73-84[Abstract/Free Full Text]
17. Shimakata, T., Iwaki, M., and Kusaka, T. (1984) Arch. Biochem. Biophys. 229, 329-339[Medline] [Order article via Infotrieve]
18. Shimakata, T., Tsubokura, A. K., and Kusaka, T. (1985) Arch. Biochem. Biophys. 238, 497-508[Medline] [Order article via Infotrieve]
19. Walker, R. W., Prome, J. C., and Lacave, C. S. (1973) Biochim. Biophys. Acta 326, 52-62[Medline] [Order article via Infotrieve]
20. Ahibo-Coffy, A., Aurelle, H., Lacave, C., Prome, J. C., Puzo, G., and Savagnac, A. (1978) Chem. Phys. Lipids 22, 185-195[Medline] [Order article via Infotrieve]
21. Lee, R. E., Armour, J. W., Takayama, K., Brennan, P. J., and Besra, G. S. (1997) Biochim. Biophys. Acta 1346, 275-284[Medline] [Order article via Infotrieve]
22. Portevin, D., De Sousa-D'Auria, C., Houssin, C., Grimaldi, C., Chami, M., Daffe, M., and Guilhot, C. (2004) Proc. Natl. Acad. Sci. U. S. A. 101, 314-319[Abstract/Free Full Text]
23. Garnier, T., Eiglmeier, K., Camus, J. C., Medina, N., Mansoor, H., Pryor, M., Duthoy, S., Grondin, S., Lacroix, C., Monsempe, C., Simon, S., Harris, B., Atkin, R., Doggett, J., Mayes, R., Keating, L., Wheeler, P. R., Parkhill, J., Barrell, B. G., Cole, S. T., Gordon, S. V., and Hewinson, R. G. (2003) Proc. Natl. Acad. Sci. U. S. A. 100, 7877-7882[Abstract/Free Full Text]
24. Cerdeno-Tarraga, A. M., Efstratiou, A., Dover, L. G., Holden, M. T., Pallen, M., Bentley, S. D., Besra, G. S., Churcher, C., James, K. D., De Zoysa, A., Chillingworth, T., Cronin, A., Dowd, L., Feltwell, T., Hamlin, N., Holroyd, S., Jagels, K., Moule, S., Quail, M. A., Rabbinowitsch, E., Rutherford, K. M., Thomson, N. R., Unwin, L., Whitehead, S., Barrell, B. G., and Parkhill, J. (2003) Nucleic Acids Res. 31, 6516-6523[Abstract/Free Full Text]
25. Cole, S. T., Eiglmeier, K., Parkhill, J., James, K. D., Thomson, N. R., Wheeler, P. R., Honore, N., Garnier, T., Churcher, C., Harris, D., Mungall, K., Basham, D., Brown, D., Chillingworth, T., Connor, R., Davies, R. M., Devlin, K., Duthoy, S., Feltwell, T., Fraser, A., Hamlin, N., Holroyd, S., Hornsby, T., Jagels, K., Lacroix, C., Maclean, J., Moule, S., Murphy, L., Oliver, K., Quail, M. A., Rajandream, M. A., Rutherford, K. M., Rutter, S., Seeger, K., Simon, S., Simmonds, M., Skelton, J., Squares, R., Squares, S., Stevens, K., Taylor, K., Whitehead, S., Woodward, J. R., and Barrell, B. G. (2001) Nature 409, 1007-1011[CrossRef][Medline] [Order article via Infotrieve]
26. Keilhauer, C., Eggeling, L., and Sahm, H. (1993) J. Bacteriol. 175, 5595-5603[Abstract/Free Full Text]
27. Schafer, A., Tauch, A., Jaeger, W., Kalinowski, J., Thierbach, G., and Puehler, A. (1994) Gene (Amst.) 145, 69-73[CrossRef][Medline] [Order article via Infotrieve]
28. Puech, V., Bayan, N., Salim, K., Leblon, G., and Daffe, M. (2000) Mol. Microbiol. 35, 1026-1041[CrossRef][Medline] [Order article via Infotrieve]
29. Rainwater, D. L., and Kolattukudy, P. E. (1982) J. Bacteriol. 151, 905-911[Abstract/Free Full Text]
30. Sassetti, C. M., Boyd, D. H., and Rubin, E. J. (2003) Mol. Microbiol. 48, 77-84[CrossRef][Medline] [Order article via Infotrieve]
31. Kimura, E., Abe, C., Kawahara, Y., Nakamatsu, T., and Tokuda, H. (1997) Biochem. Biophys. Res. Commun. 234, 157-161[CrossRef][Medline] [Order article via Infotrieve]
32. Nishio, Y., Nakamura, Y., Kawarabayasi Y., Usuda, Y., Kimura, E., Sugimoto, S., Matsui, K., Yamagishi, A., Kikuchi, H., Ikeo, K., and Gojobori, T. (2003) Genome Res. 13, 1572-1579[Abstract/