MbtH is a protein encoded in the mycobactin siderophore biosynthetic gene cluster of M. tuberculosis and the founding member of the MbtH-like protein family (NCBI CDD pfam 03621) 33 . Our analysis of available genome sequences of GPL producers revealed that every GPL biosynthetic gene cluster known to date contains a mbtH-like gene located upstream of NRPS-encoding genes required for D-Phe-D-alloThr-D-Ala-L-alaninol assembly (Figure 2). The MbtH-like protein orthologues encoded by these mbtH-like genes are comprised of 69–93 amino acids and have remarkable sequence identity (80-100%) (Figure 3). This sequence identity extends to the three fully conserved tryptophan residues that are a hallmark of the protein family (NCBI CDD pfam 03621) 33 (Figure 3A). The open reading frame corresponding to the mbtH-like gene of M. avium 2151 (Figure 2) has not been previously annotated; however, our genome sequence analysis revealed its presence. The MbtH-like protein encoded by this gene is shown in the protein alignment (Figure 3A). The orthologous mbtH-like genes or MbtH-like proteins in the other species shown in Figure 2 have been annotated each as mbtH or MbtH, respectively 24 46 , presumably due to their sequence relatedness with M. tuberculosis MbtH. This name assignment is misleading as these genes are not orthologues of mbtH, the gene of the mycobactin biosynthetic pathway present in many mycobacteria, including M. smegmatis, M. abscessus, and M. avium 33 35 . This name assignment leads to gene nomenclature confusion by resulting in more than one gene named mbtH in the same species. We proposed herein to name all the orthologous mbtH-like genes associated with GPL production as gplH, a name derived from

g

Our bioinformatics analysis revealed that every GPL biosynthetic gene cluster known to date contains a mbtH-like gene, gplH. The involvement of this conserved gene in GPL production remains unproven. Herein, we sought to conclusively establish whether gplH was required for GPL production. To this end, we engineered Ms ΔgplH, a mutant with an unmarked, in-frame deletion of gplH (Figure 4A), the Ms gene upstream of the NRPS-encoding gene mps1 (Figure 2), and assessed the ability of this mutant to produce GPLs as described below. Ms was selected as a representative prototype of GPL producers for the studies presented herein due to its superior experimental tractability compared with other GPL producers (e.g., MAC members).

The gplH deletion was engineered using the gplH deletion cassette-delivery suicide vector p2NIL-GOALc-ΔgplH c(~16 kb, Figure 4B) in a homologous recombination- and counter selection-based approach that replaced gplH by a 17-codon gene remnant cloned into the vector. The deletion in Ms ΔgplH encompassed 59 central amino acids of the predicted MbtH-like protein encoded by gplH (GplH, MSMEG_0399; 76 amino acids, Figure 3A). The deletion was verified by PCR using primer pairs that produced amplicons of different sizes depending on whether the genomic DNA used as PCR template was wild type (WT) or carried the gene deletion (Figure 4C). The successful engineering of Ms ΔgplH set the stage for probing the involvement of gplH in GPL production.

We investigated the effect of the gplH deletion in Ms ΔgplH on GPL production using TLC and MS analyses. In addition, a control strain for genetic complementation analysis was constructed (Ms ΔgplH + pCP0-gplH), and the ability of this strain and that of Ms WT controls to produce GPLs was investigated. Representative results from the TLC analysis are shown in Figure 5. The analysis of lipid extracts from the parental Ms WT strain, or Ms WT bearing the empty pCP0 vector, revealed the expected production of GPLs in these WT controls. Conversely, analysis of lipid extracts from Ms ΔgplH did not reveal detectable amounts of GPLs. Transformation of Ms ΔgplH with pCP0-gplH (a pCP0-based plasmid expressing gplH) rendered the strain Ms ΔgplH + pCP0-gplH, for which TLC analysis demonstrated that production of GPLs was restored to levels comparable to those seen in the WT controls. In contrast, Ms ΔgplH retained its GPL deficient phenotype after transformation with empty pCP0 vector (strain Ms ΔgplH + pCP0). The results of our complementation analysis rule out the possibility that the GPL deficient phenotype observed in Ms ΔgplH is due to a polar effect of the gplH deletion on downstream genes required for GPL production (i.e., mps1 and mps2; Figure 2).

The presence of GPLs was probed for in lipid samples from Ms WT + pCP0, Ms ΔgplH + pCP0, and Ms ΔgplH + pCP0-gplH (complemented strain) by GC-MS analysis as well. The pCP0-bearing strains, Ms WT + pCP0 and Ms ΔgplH + pCP0, rather than their respective plasmid-free parental strains, were used in these experiments so that the WT, the mutant, and the complemented strain could all be cultured under identical conditions (i.e., kanamycin-containing growth medium) for comparative analysis by GC-MS. Representative results from the GC-MS analysis are shown in Figure 6. This analysis probed for the presence of the alditol acetate derivatives of the characteristic glycosyl residues of Ms GPLs as a fingerprint indicator of the presence of GPLs in the lipid samples analyzed 47 . The GC-MS analysis of samples from Ms WT + pCP0 revealed the expected m/z peak array consistent with the characteristic presence of alditol acetate derivatives of the 2,3,4-trimethyl-rhamnose, 3,4-dimethyl-rhamnose and 6-deoxy-talose components of GPLs 7 8 47 . Conversely, these alditol acetate derivatives were not detected by GC-MS analysis of samples from Ms ΔgplH + pCP0. The samples from the complemented strain, Ms ΔgplH + pCP0-gplH, displayed an m/z peak array comparable to that of Ms WT + pCP0 and consistent with the presence of the alditol acetate derivatives originating from GPLs (not shown). Overall, the results of the GC-MS analysis and the results of the TLC analysis are in agreement with each other and, coupled with our genetic complementation-controlled analysis, conclusively demonstrate that gplH is essential for production of GPLs.

Based on the recent elucidation of the function of MbtH-like proteins as amino acid adenylation domain activators 36 37 38 39 40 , it is reasonable to propose that GplH is required for the assembly of the D-Phe-D-alloThr-D-Ala-L-alaninol core moiety of GPLs by the Mps1-Mps2 NRPS system. GplH might act as a critical activator of the amino acid adenylation activity of one or more of the four amino acid adenylation domains predicted by sequence analysis of the Mps1-Mps2 NRPS system 22 23 . Biochemical studies will be required to investigate this possibility.

We noted that Ms has two potential mbtH-like genes located outside the GPL biosynthetic gene cluster. One of these genes is the mbtH orthologue in the mycobactin biosynthetic gene cluster of Ms mentioned above 35 . The second gene, MSMEG_0016, is clustered with genes implicated in the production of the siderophore exochelin 48 49 50 . The protein products of these two Ms gplH paralogues have considerable amino acid sequence identity between themselves and with GplH and M. tuberculosis MbtH (Figure 3B). The GPL deficiency of Ms ΔgplH indicates that neither of these two Ms gplH paralogues can support the production of GPLs in Ms ΔgplH to a meaningful level under our culturing conditions.

It is worth noting that Ms mbtH and MSMEG_0016 are associated with siderophore production pathways known to be repressed during growth under iron-rich conditions 51 52 . This fact raises the possibility that neither of these genes is expressed (or they are poorly expressed) in the iron-rich standard Middlebrook media used in our studies. With this consideration in mind, we explored whether an increase in expression of Ms mbtH (encoding the paralogue with the higher homology to GplH, Figure 3) could complement the GPL deficiency of Ms ΔgplH. To this end, we evaluated GPL production in Ms ΔgplH after transformation of the mutant with pCP0-mbtHMs (expressing Ms mbtH). TLC analysis of lipid extracts from the transformant revealed the presence of GPLs, thus indicating that plasmid-directed constitutive expression of Ms mbtH complements the GPL deficient phenotype of Ms ΔgplH (Figure 5). Thus, it appears that Ms MbtH has the potential to functionally replace GplH if present in sufficient quantities. This cross-complementation phenomenon is in line with recent cell-based studies demonstrating MbtH-like protein-mediated cross-talk between NRPS systems 41 44 . Our finding is also consistent with reported in vitro enzymology indicating that, at least in some cases, the activity of amino acid adenylation domains of NRPSs can be stimulated not only by bona fide MbtH-like protein partners, but also by MbtH-like protein homologues from disparate natural product biosynthetic pathways 39 40 .

Colony morphotype, biofilm formation and sliding motility are properties that have been shown to be altered in GPL deficient mutants 18 19 20 23 . Loss of GPL also perturbs bacterial surface properties 19 32 and reduces the cell-wall permeability barrier to chenodeoxycholate uptake 19 . Interestingly, an altered profile of GPLs has been observed in drug-resistant MAC isolates 21 . This finding raises the possibility that GPL production might have an impact on antimicrobial drug susceptibility as well. We investigated whether deletion of gplH had an effect on all these properties. Ms WT + pCP0 and Ms ΔgplH + pCP0, rather than their respective plasmid-free parental strains, were used in the experiments so that the WT, the mutant, and the complemented Ms ΔgplH + pCP0-gplH strain could all be cultured under identical conditions (i.e., kanamycin-containing growth media) for comparative analysis. Representative results from these studies are shown in Figure 7.

Ms is known to develop into smooth, reddish colonies with a glossy and translucent appearance when grown on low carbon source congo red agar plates 23 . As expected, Ms WT + pCP0 displayed this characteristic morphotype in our congo red agar plate assay (Figure 7A). Ms ΔgplH + pCP0, however, had a drastically different morphotype. The mutant was characterized by rough, whitish colonies with a non-translucent and dried appearance. The strain Ms ΔgplH + pCP0-gplH had a morphotype more similar to WT than to that of the mutant, indicating partial complementation by episomal expression of gplH in the congo red agar assay. Deletion of gplH also altered the ability of Ms to form biofilms (Figure 7B). Ms WT + pCP0 formed a continuous, thin biofilm at the liquid-air interface, as expected based on previous reports 53 54 . In contrast, Ms ΔgplH + pCP0 failed to develop such a biofilm and instead grew as chunky patches on the liquid surface. The strain Ms ΔgplH + pCP0-gplH produced biofilms comparable to those seen with Ms WT + pCP0. Sliding motility was also compromised in Ms ΔgplH + pCP0 (Figure 7C). The mutant did not show sliding motility, whereas Ms WT + pCP0 was highly active in the motility assay. Ms ΔgplH + pCP0-gplH also displayed sliding motility, although the motility was somewhat reduced compared to WT. This observation indicates partial complementation by episomal expression of gplH. Overall, these results clearly indicate that deletion of gplH has a profound impact on colony morphotype, biofilm formation, and sliding motility. These mutant phenotypes have previously been associated with other GPL deficient strains and attributed to alterations of the properties of the cell surface due to lack of GPLs. Thus, it is likely that the phenotypes observed in the gplH mutant arise from its GPL deficiency.

In contrast to the drastic impact of the gplH deletion on colony morphotype, biofilm formation and sliding motility, lack of gplH had a relatively minor effect on antimicrobial drug susceptibility. When compared with Ms WT + pCP0 (control strain), Ms ΔgplH + pCP0 showed a slight, yet consistent, increase in susceptibility to only two drugs (cefuroxime and cefotaxime) from a panel of 15 drugs of different classes tested in standard disk diffusion assays.

Interestingly, these two drugs belong to the cephalosporin class, suggesting that the hypersusceptibility of the mutant is antibiotic-class dependent. Representative results illustrating the hypersusceptibility of the mutant to these cephalosporins are shown in Figure 7D.

Streptomycin susceptibility results are also shown in Figure 7D. The streptomycin susceptibility is presented as an example of those drugs to which the mutant had no meaningful difference in susceptibility relative to the WT control. The Ms ΔgplH + pCP0-gplH strain showed a drug susceptibility pattern similar to that of Ms WT + pCP0, indicating that the hypersusceptible phenotype of the mutant was complemented by episomal expression of gplH.

The molecular mechanism behind the cephalosporin hypersusceptibility arising from the lack of gplH remains obscure. It is generally believed that the permeability barrier imposed by the mycobacterial outer membrane reduces antibiotic susceptibility by decreasing compound penetration. Thus, it is tempting to hypothesize that the observed cephalosporin hypersusceptibility arises from an alteration in the permeability barrier of the outer membrane of the gplH mutant due to the lack of GPLs. The observation that lack of GPLs correlates with a reduction in the permeability barrier to chenodeoxycholate uptake 19 is in line with this hypothesis. The absence of GPLs might produce structural or fluidity changes in the membrane that lead to an increase in cephalosporin penetration. The fact that Ms ΔgplH displays only a modest increase in antibiotic susceptibility suggests, however, that the lack of GPLs in the outer membrane of the mutant does not have a profound effect on the permeability barrier that this cell envelope structure presents to drug penetration. Thus, our results support the view that GPLs are not critical contributors to the physical integrity of the permeability barrier of the mycobacterial cell envelope.

Ms strain mc²155 (ATCC 700084) and its derivatives were routinely cultured under standard conditions (37°C, 225 rpm) in Middlebrook 7H9 (Difco) supplemented with 10% ADN (5% BSA, 2% dextrose, 0.85% NaCl), 0.2% glycerol and 0.05% Tween-80 (supplemented 7H9) or in Middlebrook 7H11 (Difco) supplemented with 10% ADN (supplemented 7H11) 55 . E. coli DH5α (Invitrogen) was cultured under standard conditions in Luria-Bertani media 56 . When required, kanamycin (30 μg/ml), hygromycin (50 μg/ml), sucrose (2%) and/or X-gal (70 μg/ml) were added to the media. General recombinant DNA manipulations were carried out by standard methods and using E. coli as the primary cloning host 56 . Molecular biology reagents were obtained from Sigma, Invitrogen, New England Biolabs, Novagen, QIAGEN, or Stratagene. Oligonucleotides were purchased from Integrated DNA Technologies, Inc. PCR-generated DNA fragments used in plasmid constructions were sequenced to verify fidelity. Chromosomal DNA isolation from and plasmid electroporation into mycobacteria were carried out as reported 55 . Table 1 lists the plasmids and oligonucleotide primers used in this study.

Ms ΔgplH was engineered using the p2NIL/pGOAL19-based flexible cassette method 57 as previously reported 4 31 35 58 . A suicide delivery vector (p2NIL-GOALc-ΔgplHc, see below) carrying a gplH (MSMEG_0399) deletion cassette (ΔgplHc) was used to generate Ms ΔgplH. The vector was electroporated into Ms and transformants with a potential p2NIL-GOALc-ΔgplHc integration via a single-crossover event (blue colonies) were selected on supplemented 7H11 containing hygromycin, kanamycin, and X-gal. The selected transformants were then grown in antibiotic-free supplemented 7H9, and subsequently plated for single colonies on supplemented 7H11 containing sucrose and X-gal. White colonies that grew on the sucrose plates were re-streaked onto antibiotic-free and antibiotic-containing supplemented 7H11 plates to identify clones that lost drug resistance, a trait indicating a possible double-crossover event with consequent loss of gplH or reversion to WT. The deletion of gplH in antibiotic sensitive clones was screened for and confirmed by PCR. Towards this end, chromosomal DNA isolated from mutant candidates was used as template along with primer pairs (pepOF and pepOR, pepF and pepR) that produced diagnostic amplicons permitting differentiation between the mutant and WT genotypes.

The plasmid p2NIL-GOALc-ΔgplHc used in the construction of Ms ΔgplH carried the gplH deletion cassette ΔgplHc. The deletion cassette contained: 995-bp segment upstream of gplH + gplH’s first 13 codons (5 fragment) followed by gplH’s last 4 codons + stop codon + 1,000-bp segment downstream of gplH (3 fragment). ΔgplHc was built by the joining of the 5' fragment and the 3' fragment using splicing-by-overlap-extension (SOE) PCR 59 . Each fragment was PCR-generated from chromosomal DNA. Primer pair pepOF and pepIR and primer pair pepIF and pepOR were used to generate the 5’ and 3’ fragments, respectively. The fragments were then used as template for PCR with primers pepOF and pepOR to fuse the fragments and create ΔgplHc (2,061 bp). The PCR-generated ΔgplHc was first cloned into pCR2.1-TOPO (Invitrogen). ΔgplHc was subsequently excised from the pCR2.1-TOPO construct using KpnI and PacI, and the excerpt was ligated to p2NIL 57 linearized by KpnI-PacI digestion. The resulting p2NIL-ΔgplHc plasmid and plasmid pGOAL19 57 were digested with PacI, and the PacI cassette (GOALc, 7,939 bp) of pGOAL19 was ligated to the linearized p2NIL-ΔgplHc to create p2NIL-GOALc-ΔgplHc. To create pCP0-gplH, the plasmid used for complementation analysis, a DNA fragment (266 bp) encompassing gplH and its predicted ribosome binding site (RBS) was PCR-amplified from genomic DNA with primer pair pepF and pepR and cloned into pCR2.1-TOPO. The RBS-gplH fragment was subsequently excised from the pCR2.1-TOPO construct using PstI and HindIII and ligated to plasmid pCP0 4 linearized by PstI-HindIII digestion to create pCP0-gplH. The cloning placed gplH under the control of the hsp60 promoter of pCP0 for gene expression in mycobacteria.

GPLs were extracted and analyzed by TLC by reported methods 22 60 . Cells from cultures (5 ml, OD₆₀₀ of 1.3-1.6) grown in supplemented 7H9 as described above were collected by centrifugation (4,700 × g, 15 min), washed with cold phosphate buffered saline (PBS, 1 ml), and processed for GPL extraction. GPLs were extacted with 2:1 CHCl₃/CH₃OH (20 μl/mg wet weight) by incubation overnight at room temperature in a rocking shaker. After incubation, insoluble material was removed by centrifugation (18,400 × g, 10 min), and the CHCl₃/CH₃OH supernatant was recovered and mixed with 0.2 volumes of 0.9% NaCl. After vigorous vortexing, the mixture was centrifuged (1,150 × g, 5 min) and the organic phase (containing GPLs) was collected and evaporated to dryness. The dried lipid extracts were dissolved in 20 μl of CHCl₃/CH₃OH (2:1) and subjected to TLC using aluminum-backed, 250-μm silica gel F₂₅₄ plates developed with CHCl₃/CH₃OH (100:7). After chromatography, TLC plates were sprayed with orcinol/sulfuric acid (0.1% orcinol in 40% sulfuric acid) and glycolipids were detected by charring at 140°C.

Alditol acetate derivatives of glycosyl units from GPLs were prepared and analyzed as reported 47 61 . Briefly, lipid samples prepared by extraction as noted above were acid-hydrolyzed in 250 μl of 2 M trifluoroacetic acid for 2 hr at 120°C. After cooling down to room temperature, samples were hexane-washed (250 μl) and dried on air bath after adding 1 μg of 3,6-O-dimethyl-glucose as an internal standard. The hydrolyzed sugars were reduced overnight at room temperature by adding 250 μl of NaBD₄ (prepared at 10 mg/ml in 1 M NH₄OH in C₂H₅OH). After reduction, glacial acetic acid (20 μl) was added to remove excess NaBD₄ and the samples were dried. CH₃OH (100 μl) was added to each sample, and after resuspension the solvent was evaporated to dryness (this step was repeated twice). The samples were per-O-acetylated with 100 μl of acetic anhydride at 120°C for 2 hr. After cooling, the samples were dried on air bath and suspended in 3 ml of CHCl₃/H₂O (2:1) by vortexing. The organic layer was extracted after centrifugation (2,500 × g, 5 min, 4°C) and dried on air bath. GC-MS analysis was performed using a Varian CP-3800 gas chromatograph (Varian Inc., Palo Alto, CA) equipped with a MS-320 mass spectrometer and using helium gas. The alditol acetate derivatives were dissolved in 50 μl of CHCl₃ before injection on a DB 5 column (30 m × 0.20 mm inner diameter) with an initial oven temperature of 50°C for 1 min, followed by an increase of 30°C/min to 150°C and finally to 275°C at 5°C/min.

The assay was carried out using reported methodologies 23 . Briefly, mycobacterial cultures (5 ml, OD₆₀₀ = 1.5) were shortly vortexed with glass beads to increase homogeneity and then centrifuged (4,700 × g, 15 min) for cell collection. The collected cells were washed with PBS (5 ml) and subsequently resuspended in PBS to an OD₆₀₀ of 1. The cell suspensions were spotted (2 μl) on congo red agar plates 23 (7H9 basal medium, 1.5% agar, 100 μg/ml congo red (sodium salt of 3,3'-([1,1'-biphenyl]-4,4'-diyl)bis(4-aminonaphthalene-1-sulfonic acid), Sigma Aldrich Co.), 0.02% glucose, 30 μg/ml kanamycin). Colony morphology was examined using an Olympus SZX7 stereo microscope after plate incubation (37°C, 3 days).

The test was performed by standard methods 19 . Briefly, mycobacterial cell suspensions in PBS prepared as noted above were spotted (2 μl) on sliding motility assay plates 19 (7H9 basal medium, 0.3% agarose, 30 μg/ml kanamycin). The sliding motility plates were incubated at 37°C and the degree of spreading (diameter of growth zone) was determined at the time points indicated in results.

The liquid-air interface biofilm assay was conducted based on reported methods 53 . Overnight mycobacterial cultures (5 ml) grown in supplemented 7H9 as noted above were centrifuged (4,700 × g, 15 min) for cell collection. The cells were washed twice with 5 ml of supplemented 7H9 without Tween-80. After washing, the cells were resuspended in supplemented 7H9 without Tween-80 to a calculated OD₆₀₀ of 10. A 25 μl aliquot of each suspension was inoculated onto the surface of 2.5 ml of supplemented 7H9 without Tween-80 loaded into a well of a 12-well polystyrene plate. The plate was incubated for 4 days without shaking at 37°C before examination for biofilm formation.

Standard disk-diffusion assays were carried out as reported 58 62 . Exponentially growing cultures (OD₆₀₀ = 0.6) in supplemented 7H9 were diluted in fresh medium to an OD₆₀₀ of 0.05, and 100 μl of diluted culture were used to seed 7H11 plates (20 ml agar/plate). Antibiotic disks were placed onto the inoculated agar and the plates were incubated at 37°C for 2 days before analysis. The antibiotics tested were doxycycline, isoniazid, streptomycin, tetracycline, cefuroxime, erythromycin, ciprofloxacin, levofloxacin, ethambutol, ethionamide, rifampicin, clarithromycin, cefuroxime, cephalexin, and cefotaxime. The antibiotics were acquired from Sigma-Aldrich, Fisher Scientific, Tokyo Chemical Industry, or Calbiochem Biochemicals.