MmpL Genes Are Associated with Mycolic Acid Metabolism in Mycobacteria and Corynebacteria

1Mycolic acids are vital components of the waxy cell wall of Mycobacterium tuberculosis, the causative agent of tuberculosis (TB). Synthesized by the concerted and sequential action of three enzymatic units, fatty acid synthase I (FAS-I), fatty acid synthase II (FAS-II), and a polyketide synthase (Pks13), these very long-chain α-alkyl, β-hydroxy fatty acids are essential for viability and virulence. While core mycolate biosynthesis is now well studied in mycobacteria and in the related Corynebacterium glutamicum, the processing of newly synthesized mycolic acids and their subsequent transport for deposition in the cell wall remains poorly understood. An earlier report of an ABC transporter mutant affected in mycolic acid composition turned out to be an indirect effect due to altered lipoglycan biosynthesis. suggested a hypothetical pathway that involves the transfer of a mature mycolate to an isoprenoid carrier, forming MycPL (6-O-mycolyl-β-D-mannopyranosyl-mono phospho- heptaprenol), and subsequently to trehalose inside the cell to yield trehalose monomycolate. TMM is then proposed to be transported by a yet unknown mechanism outside the cell where it is used as a substrate by the mycolyltransferase enzymes of the Ag85 complex, to transfer the mycolate residue to either the arabino-galactan (AG) complex (wall bound mycolates), or to another TMM (resulting in formation of TDM). Alternatives to this hypothetical pathway include intracellular synthesis of TDM prior to transport, or the flipping of Myc-PL to the outside for subsequent use as a substrate for extracellular formation of TMM. Apart from the identification of Myc-PL as a carrier-bound intracellular intermediate, and the mycolyltransferase activity of the Ag85 proteins, not much is known about the intracellular processing and subsequent transport of mycolic acids.

2The genomes of M. tuberculosis H37Rv and other mycobacteria contain a number of genes encoding proteins that belong to a family of multidrug resistance pumps termed RND (resistance, nodulation, and division) proteins. Designated MmpL (for mycobacterial membrane protein large), these proteins typically contain 12 transmembrane (TM) domains and two non-TM loops. None of the mycobacterial MmpLs studied so far appear to play any role in drug resistance (mmpL7 conferred resistance to isoniazid when overexpressed in Mycobacterium smegmatis but the M. tuberculosis mmpL7 mutant did not display altered sensitivity to the drug; ). Instead, many mmpL genes are associated with clusters involved in the biosynthesis of cell wall-associated glycolipids like sulfolipids, polyacylated trehalose, glycopeptidolipids, lipooligosaccharides, and other complex lipids like phthiocerol dimycocerosate (PDIM). MmpLs are proposed to function as scaffolds for the biosynthetic machinery, allowing localized synthesis of a cell wall associated lipid, pairing with an ABC transporter, and in some cases an MmpS protein, to facilitate transport. Indeed, a two hybrid screen using the non-TM domains of the PDIM transporter MmpL7 revealed interactions with enzymes involved in the final stages of biosynthesis of the phthiocerol moiety. In some cases, such “mega complexes” may involve localized “flipping” of a biosynthetic intermediate for presentation to extracytoplasmic enzymes involved in the biosynthesis of the lipid, as was observed with MmpL8 and sulfolipid SL-I. Thus, it seemed likely that an MmpL protein may be involved in the translocation of a mycolate-containing glycolipid (TMM, TDM, or Myc-PL) to the outside of the bacterial cell for subsequent use as a substrate for cell wall mycolylation.

3Mycobacterium leprae, which contains a high degree of “genetic decay” but still possesses an intact cell wall containing mycolic acids, represents a valuable reference when considering candidate genes for essential cell wall-related functions like mycolic acid metabolism. Homologs of only 5 of 14 M. tuberculosis mmpL genes are present in M. leprae, of which only mmpL3 is predicted to be an essential gene based on studies in M. tuberculosis. Given the essentiality of mycolic acids for viability of mycobacteria, it was likely that mmpL3 was involved in mycolate transport in M. tuberculosis. In this study, we have investigated the role of mmpL3 in mycolic acid transport by generating and characterizing a conditional mutant of MSMEG0250, the homolog of mmpL3 in the fast growing Mycobacterium smegmatis. In parallel, we also compared and contrasted the role of mmpL genes in Corynebacterium glutamicum, a bacterium that has a cell wall similar to mycobacteria including the presence of mycolic acids, called corynomycolic acids (both belong to the suborder Corynebacterineae). A particular advantage of using corynebacteria is the ability to generate viable corynomycolate-deficient mutants and our studies demonstrated a functional redundancy of corynomycolate metabolism-associated MmpLs in this genus.

4M. tuberculosis mmpL3 (Rv0206c) was first proposed to be an essential gene by on the basis of the inability to obtain an mmpL3 knockout mutant. While the gene is not located near any known mycolic acid biosynthesis genes, the gene cluster surrounding mmpL3 shows a high level of synteny with other mycobacterial species including M. leprae (Figure 1). This includes the presence of three putative membrane associated proteins (Figure 1), and a second mmpL gene, mmpL11, situated downstream of mmpL3. While did not detect any alterations in the lipid profile of an M. tuberculosis mmpL11 mutant, the strain was attenuated in the mouse model of infection.

5The protein encoded by mmpL3 is predicted to be a membrane associated protein, and contains 12 TM domains and two non-TM loops, L1 and L2 of sizes 151 and 142 amino acids, respectively (Figure 2). We first constructed a knockout phage designed to delete MSMEG0250, the M. smegmatis homolog of M. tuberculosis mmpL3, and consistent with findings from previous studies in M. tuberculosis, we were unable to generate a null mutant. Subsequently, CESTET, a tool designed to test gene essentiality in M. smegmatis, was used to obtain a conditional mutant of MSMEG0250. The strain, ΔMSMEG0250 contained a recombinant, integrated copy of mmpL3 under the control of the inducible acetamidase promoter, while the native copy of the gene was deleted and replaced with a hygromycin resistance cassette. The viability of ΔMSMEG0250 was dependent on the addition of the inducer acetamide in the growth medium confirming that MSMEG0250 was an essential gene (Figure 3).

6Recently, spontaneous mutants resistant to BM212 and SQ109, compounds with anti-TB activities, were shown to contain mutations mapping to mmpL3 (LaRosa et al., 2012; ). Furthermore, cells exposed to SQ109 were shown to accumulate TMM suggesting that MmpL3 was involved in the efflux of TMM. To address the putative role of MmpL3 in TMM efflux, we assessed the effects of depletion of MSMEG0250 on lipid metabolism in the M. smegmatis conditional mutant, making use of a three stage extraction procedure: (1) petroleum ether extraction of the cell pellet that does not affect the integrity of the cells, but extracts the surface-exposed, noncovalently bound cell wall lipids; (2) extraction of apolar and polar lipids from the “petroleum ether stripped” pellet; and (3) extraction of AG-bound mycolic acids from the resultant delipidated cells. Cultures of the ΔMSMEG0250 conditional mutant were grown in broth in the presence or absence of acetamide, labeled with [14C]-acetate at different time points and subjected to the above three stage extraction procedure. For cultures grown in the presence of acetamide, TMM was predominantly found in the apolar lipid extract from cell pellets (Figure 4A). However, TDM was found in both the petroleum ether extract, and in the apolar lipid extract from the subsequent cell pellets. An earlier examination we had conducted of M. smegmatis cell pellets extracted with consecutive treatments of petroleum ether indicated that no further TDM was extracted from cells after the third consecutive extraction. Each petroleum ether extract of the ΔMSMEG0250 conditional mutant shown in Figure 4A represent a pool of five consecutive extractions of the same sample and thus the presence of TDM in the subsequently processed cell pellet was not due to insufficient petroleum ether extraction. The exact location of these cell-associated TDM molecules that resist petroleum ether extraction is unknown, but it is possible that they represent newly synthesized TDM that is closer to the inner membrane and thus inaccessible to petroleum ether extraction.

7In contrast to the cultures of the conditional mutant grown in the presence of acetamide, [14C]-labeled cultures grown in the absence of acetamide showed an increasing amount of TMM in cell pellets, that was accompanied by decrease in TDM levels both in the petroleum ether extracts and in the apolar fraction from cells (Figure 4A; Figure S1 available online). Conditional depletion of MSMEG0250 also resulted in a decrease in mycolylation of AG (Figure 4B). The accumulation of TMM in the cells over time following conditional depletion suggested that loss of MSMEG0250 function affected TMM transport either directly or indirectly.

8The cell walls of corynebacteria, which are similar to those of mycobacteria, contain trehalose monocorynomycolate (TMCM) and trehalose dicorynomycolate (TDCM), the equivalents of TMM and TDM, as the outer, noncovalently bounds glycolipids. As corynomycolic acid biosynthesis is a nonessential process, it is possible to generate viable mutants of corynebacteria that do not produce corynomycolates. In a parallel study on the role of MmpLs in corynomycolate transport in C. glutamicum, we conducted a BLAST search of the C. glutamicum ATCC 13032 genome, using the amino acid sequences of open reading frames from M. tuberculosis mmpL3-mmpL11 region as queries. The search revealed that there was no corynebacterial equivalent of the mycobacterial mmpL3-mmpL11 cluster. However, the genome of C. glutamicum does encode four mmpL genes, NCgl0228, NCgl0559, NCgl0887, and NCgl2769 in distinct locations on the genome. As TMCM and TDCM are the only abundant, noncovalently bound outer envelope glycolipids in C. glutamicum, it seemed likely that one or more of these mmpL genes were involved in corynomycolate transport. As NCgl0599 was associated with a cluster involved in the biosynthesis of a carotenoid pigment, it was not considered for further study.

9NCgl0228 and NCgl2769 revealed predicted topologies similar to M. tuberculosis MmpL3 (12 TM domains and two non-TM loops), while that of NCgl0887 appeared different (Figure 2). Interestingly, NCgl0887 is situated in a cluster containing a trehalose corynomycolyl transferase, and NCgl2769 is present downstream of the corynomycolate condensase gene, pks13. To probe the role of C. glutamicum mmpLs in corynomycolate transport, we generated individual in frame deletion mutants of NCgl0228, NCgl0887, and NCgl2769 in C. glutamicum ATCC 13032. Loss of corynomycolic acids from the cell envelope results in poor growth rates leading to smaller colonies. None of the three mutant strains (ΔNCgl0228, ΔNCgl0887, and ΔNCgl2769) showed any alterations in colony size or appearance (Figure 5). To investigate potential functional redundancies between the three C. glutamicum mmpL genes, double and triple mutant strains were constructed. While the colony sizes and appearance of the double mutants, ΔNCgl0228-ΔNCgl0887 and ΔNCgl0887-ΔNCgl2769, remained unaltered (Figure 5), the ΔNCgl0228-ΔNCgl2769 mutant formed smaller colonies and showed “clumpy” growth in broth. A similar growth defect was observed for the triple mutant ΔNCgl0228-ΔNCgl0887-ΔNCgl2769 (Figure 5). Thus, while single deletions of NCgl0228 or NCgl2769 did not have any effect, the concurrent loss of NCgl0228 and NCgl2769 severely affected the growth patterns of the mutant strain, indicating potential alterations in the cell wall.

10To determine whether the phenotype observed above for some of the C. glutamicum mmpL mutants were linked to defects in corynomycolate transport, petroleum ether extracts of surface-exposed lipids, and subsequent, remaining total lipids were extracted from [14C]-acetate-labeled cell pellets. TLC analysis showed that the parental strain and most mutant strains produced TMCM and TDCM, with the latter found predominantly in the petroleum ether extracts containing surface-exposed lipids (Figures 6A and 6B). Furthermore, levels of AG-bound corynomycolic acids for most mutant strains were the same as the parental, wild-type strain C. glutamicum (Figure 6C). The two exceptions were the ΔNCgl0228-ΔNCgl2769 and ΔNCgl0228-ΔNCgl0887-ΔNCgl2769 mutant strains, in which no TMCM or TDCM was detected in the petroleum ether extracts or in the cell pellets (Figures 6A and 6B). Additionally, no AG-bound corynomycolic acids were observed in the ΔNCgl0228-ΔNCgl2769 and ΔNCgl0228-ΔNCgl0887-ΔNCgl2769 strains (Figure 6C). The absence of any TMCM, TDCM, and AG-bound corynomycolates suggested either that there was a complete cessation of corynomycolic acid biosynthesis in the double and triple mutants or that corynomycolates were being transported in an alternate (carrier-bound) form that was used as substrate for TMCM, TDCM, and AG-bound corynomycolate synthesis. However, the latter seemed less likely as no intermediates were seen to accumulate in the ΔNCgl0228-ΔNCgl2769 and ΔNCgl0228-ΔNCgl0887-ΔNCgl2769 mutants. These results also indicated a functional redundancy between NCgl0228 and NCgl2769 with regards to corynomycolate metabolism. No additional alterations in lipid profiles were observed in the triple mutant suggesting that NCgl0887 did not play any apparent role (Figures 6A–6C). Complementation of the double mutant with either plasmid-borne NCgl0228 or NCgl2769 restored corynomycolate biosynthesis, though in case of the latter gene, the complementation was partial (Figures 7A and 7B and Figure S2).

11Our studies demonstrated that mmpL3 plays a role in mycolic acid transport; the intracellular accumulation of TMM in the M. smegmatis MSMEG0250 mutant following conditional depletion suggested that TMM was the likely carrier for the mycolic acid moiety. The exported TMM is presumably used as substrate by the enzymes of the Ag85-complex for mycolylation of the cell wall AG and the formation of cord factor (TDM). Our findings were complemented by the drug-to-target approach used by and, with the latter demonstrating that treatment with a compound that targeted MmpL3 resulted in a phenotype identical to what we observed with the M. smegmatis conditional mutant. Additionally, during the review of this manuscript, reported that AU1235, an adamantly urea inhibitor of M. tuberculosis, also targeted MmpL3. Notably, the authors showed a decrease in the levels of TDM and AG-bound mycolates in M. tuberculosis cells treated with AU1235, and accumulation of TMM in the inner membrane. Additionally, they were also able to reproduce the effects of AU1235 in cultures of a M. smegmatis conditional mmpL3 mutant grown under nonpermissive conditions; a result similar to that obtained for the ΔMSMEG0250 mutant described in this study.

12MmpL3 may not be solely responsible for transport; in M. smegmatis; glycopeptidolipid transport requires the concerted action of an MmpL protein, MmpS4 and a third transmembrane protein termed Gap. The region downstream of mmpL3 encodes two transmembrane proteins, Rv0204c and Rv0205, which may form part of a complex involved in the transport of TMM (though the M. leprae homolog of the latter is annotated as a possible pseudogene). Additionally, the biosynthetic clusters for PDIM also contain ABC transporters and thus the role of ABC transporters in mycolate transport (acting in conjunction with MmpL3) remains a possibility.

13A recent study by has suggested a role for the mmpL3-mmpL11 gene cluster in heme transport. The authors report that they were also able to delete a region of the M. smegmatis genome that included MSMEG0250. In our studies, we were unable to generate an individual null mutant of MSMEG0250 in M. smegmatis mc2155 despite repeated attempts (also observed for the M. tuberculosis homolog by ). We were, however, able to demonstrate the essentiality of MSMEG0250 using CESTET and show that conditional depletion caused a loss of TDM and of cell wall (AG) mycolylation, and the intracellular accumulation of TMM.

14In contrast to mycobacteria, we observed several differences in the role of mmpL genes in corynomycolate transport in C. glutamicum. First, there was no equivalent corynebacterial cluster exhibiting synteny with mycobacterial mmpL3-mmpL11 region. Instead, a functional redundancy was observed between NCgl0228 and NCgl2769 located in two distinct regions of the C. glutamicum genome. Individual deletion of either of these mmpL genes did not have any effect, but simultaneous deletion of both genes led to complete loss of TMCM, TDCM and AG-bound corynomycolates. This was in contrast to the M. smegmatis conditional mutant in which conditional depletion of MSMEG0250 led to loss of TDM and AG mycolylation, but intracellular accumulation of TMM. As mentioned above, while this result raised the possibility that corynomycolate transport may involve a carrier other than TMCM, this seemed highly unlikely, given the absence of any accumulating intermediate in the C. glutamicum ΔNCgl0228-ΔNCgl2769 and ΔNCgl0228-ΔNCgl0887-ΔNCgl2769 mutant strains. Instead, variations in the phenotypes of the mycobacterial and corynebacterial mutants may be a reflection of fundamental differences in mycolic acid biosynthesis in the two genera. Mycobacterial mycolic acid biosynthesis requires several components including a multidomain type-I fatty acid synthase (FAS), a multienzyme type-II FAS complex, modifying enzymes and Pks13. Corynomycolate biosynthesis on the other hand requires two type-I FASs and Pks13. MmpLs are also envisaged to act as scaffolds for the localized biosynthesis on the cell membrane, facilitating concerted biosynthesis and transport. In mycobacteria, due to the far larger number of components involved, mycolate biosynthesis may be delinked from transport. Instead, MmpL3 could form a complex with other proteins involved in TMM transport. As a result, loss of MmpL3 would not cause a direct cessation of mycolate biosynthesis. In contrast, in C. glutamicum, where corynomycolate biosynthesis requires fewer enzymatic components, late-stage enzymes such as Pks13 may depend on interaction with an MmpL-complex to ensure that corynomycolate biosynthesis is coupled to transport. Thus, a concurrent loss of NCgl0228 and NCgl2769 would lead to not just the loss of a transport complex, but also disrupt the late stages of corynomycolate biosynthesis. The identification of mmpL3 as an essential membrane protein-encoding gene involved in mycolate transport opens up new avenues for targeting this vital and under-exploited mycobacterial pathway for developing new anti-TB drugs, an opportunity highlighted by the identification of MmpL3 as the target of three drugs.

15One of the greatest challenges for combating TB is the recent rise of multidrug-resistant (MDR) and extensively drug-resistant (XDR) strains of M. tuberculosis. Despite being critical for viability and virulence (and the target of the hallmark anti-TB drug isoniazid), pathways involved in the biosynthesis of mycolic acids remain poorly exploited as drug targets. Additionally, components involved in the postbiosynthesis processing and transport of mycolates are not known and remain an untapped source of potential drug targets. This study identifies a gene (mmpL3) involved in mycolic acid transport in mycobacteria. Furthermore, the results obtained from the C. glutamicum mutants indicate that transport processes may be coordinated with biosynthesis via protein-protein interactions with MmpLs acting as molecular scaffolds, thus opening up new avenues for studying novel transport mechanisms for mycobacterial lipids.

16The M. smegmatis conditional mutant ΔMSMEG0250 was generated using CESTET. Briefly, a merodiploid was first generated by introducing an integrating vector pMV306-mmpL3 by electroporation into M. smegmatis mc2155 (pMV306-mmpl3 consists of mmpL3 cloned downstream of the acetamidase promoter in the integrative vector pMV306). The merodiploid strain mc2155::pMV306mmpl3 was then subjected to specialized transduction as previously described using phΔMSMEG0250, a temperature sensitive, recombinant phage designed to replace MSMEG0250 with a hygromycin resistance marker. Transductants were selected at the nonpermissive temperature of 37°C on selective plates containing 150 μg/ml hygromycin. After confirmation of gene replacement by Southern blot, one such transductant was named ΔMSMEG0250 and was selected for further analysis. Conditional depletion of MSMEG0250 in ΔMSMEG0250 to visualize [14C]-labeled mycolic acids and other lipids was carried out as described before.

17To generate specific deletions of NCgl0228, NCgl0887, or NCgl2769 in the chromosome of C. glutamicum, the nonreplicable vector pK19mobsacB was used, containing inserts synthesized by GeneArt. The inserts contained 12 nucleotides (nt) of the 3′ end of the respective gene together with 300 bp genomic upstream sequences, and 36 nt of the 5′ end together with 300 bp genomic downstream sequences. The respective vector constructed was used to transform C. glutamicum ATCC 13032 to kanamycin-resistance (Kanr) indicating chromosomal integration. Sucrose-resistant (Sucr) clones were selected in a second round of positive selection, indicating loss of the vector-encoded sacB function. After the second recombination event about half of the recombinants contained the desired deletion, as verified by PCR, and one strain each was termed C. glutamicum ΔNCgl0228, C. glutamicum ΔNCgl0887, and C. glutamicum ΔNCgl2769, respectively. The double-mutant C. glutamicum ΔNCgl0887ΔNCgl2769 was generated using C. glutamicum ΔNCgl0887 and using the appropriate pK19mobsacB construct. In-frame deletion of NCgl0228 in a single- or double-mutant background was not possible despite repeated attempts. As an alternative, we used pK19mobsacBNCgl0228, containing an internal fragment of NCgl0228 amplified by primer pairs 5′-CATAGAATTCGTGGCTGTGCTCATTGCGTTGAC-3′ and 5′-GTACGTCGACCTCTGCCATCAAATCAGCCGACTG-3′. This vector was used to disrupt NCgl0228 in C. glutamicum ΔNCgl0887, in C. glutamicum ΔNCgl2769 and in C. glutamicum ΔNCgl0887-ΔNCgl2769 to generate the double-mutants C. glutamicum ΔNCgl0228-ΔNCgl0887 and ΔNCgl0228-ΔNCgl2769, and the triple mutant C. glutamicum ΔNCgl0228-ΔNCgl0887-ΔNCgl2769. Growth of all mutants was characterized on the complex medium brain heart infusion broth at 30°C, with 25 μg ml-1 kanamycin when appropriate. Extraction of [14C]-labeled of lipids and corynomycolic acids from all strains was done as described previously.

18NCgl0228 was amplified using primer pairs F0228 (5′-CTCATTTGTCGACAAGGAGATATAGGTGGCGAAATTGCTATTCAGG-3′) and R0228 (5′-GGTGGGATCCCTAACGTGCAGCCTGCTTCTCC-3′); and NCgl2769 using the primer pair F2769 (5′-ACAATTGTCGACAAGGAGATATAGGTGTTTTCTAAATGGGGCCAC-3′) and R2769 (5′-TCCTCGCGGATCCTTAATCTAGATCCTCAAGCCTGC-3′), using C. glutamicum ATCC 13032 chromosomal DNA as template. PCR products were cloned in pVWEx2 to yield pVWEx2-NCgl0228 and pVWEx2-NCgl2769, respectively. The inserts in the plasmids were verified by sequencing, and subsequently used to transform C. glutamicum ΔNCgl0228ΔNCgl0887 to tetracycline resistance (5 μg ml-1) using transformation protocols described by.