The Mycobacterium avium Complex gtfTB Gene Encodes a Glucosyltransferase Required for the Biosynthesis of Serovar 8-Specific Glycopeptidolipid

Abstract

Mycobacterium avium complex (MAC) is one of the most common opportunistic pathogens widely distributed in the natural environment. The 28 serovars of MAC are defined by variable oligosaccharide portions of glycopeptidolipids (GPLs) that are abundant on the surface of the cell envelope. These GPLs are also known to contribute to the virulence of MAC. Serovar 8 is one of the dominant serovars isolated from AIDS patients, but the biosynthesis of serovar 8-specific GPL remains unknown. To clarify this, we compared gene clusters involved in the biosynthesis of several serovar-specific GPLs and identified the genomic region predicted to be responsible for GPL biosynthesis in a serovar 8 strain. Sequencing of this region revealed the presence of four open reading frames, three unnamed genes and gtfTB, the function of which has not been elucidated. The simultaneous expression of gtfTB and two downstream genes in a recombinant Mycobacterium smegmatis strain genetically modified to produce serovar 1-specific GPL resulted in the appearance of 4,6-O-(1-carboxyethylidene)-3-O-methyl-glucose, which is unique to serovar 8-specific GPL, suggesting that these three genes participate in its biosynthesis. Furthermore, functional analyses of gtfTB indicated that it encodes a glucosyltransferase that transfers a glucose residue via 1→3 linkage to a rhamnose residue of serovar 1-specific GPL, which is critical to the formation of the oligosaccharide portion of serovar 8-specific GPL. Our findings might provide a clue to understanding the biosynthetic regulation that modulates the biological functions of GPLs in MAC.

Mycobacteria are pathogens that cause diseases such as tuberculosis and leprosy. In addition, nontuberculous mycobacteria, which are widely distributed in the natural environment, cause opportunistic pulmonary infections resembling tuberculosis. These mycobacteria are distinguished by a multilayered cell envelope consisting of peptidoglycan, mycolyl arabinogalactan, and surface glycolipids (9, 13). The surface glycolipids are abundant and structurally different, and they may act as a barrier to immune responses (9, 13). Glycopeptidolipids (GPLs) are major glycolipid components present on the surface of several species of nontuberculous mycobacteria (40). All of these GPLs have a conserved core structure that is composed of a fatty acyl tetrapeptide glycosylated with 6-deoxytalose (6-d-Tal) and O-methyl-rhamnose (O-Me-Rha) and are termed non-serovar-specific GPLs (nsGPLs) (2, 4, 14). On the other hand, the GPLs of Mycobacterium avium complex (MAC), nontuberculous mycobacteria consisting principally of two species, M. avium and M. intracellulare, have various haptenic oligosaccharides linked to the 6-d-Tal residue of nsGPLs, resulting in serovar-specific GPLs (ssGPLs) (2, 4, 40). The oligosaccharide portions of ssGPLs define MAC serovars that are classified into 28 types. The serovar 1-specific GPL, with Rha linked to the 6-d-Tal residue, is the basic oligosaccharide unit of all ssGPLs (11). The Rha residue of serovar 1-specific GPL is further extended by various glycosylation steps, such as rhamnosylation, fucosylation, and glucosylation (11). These glycosylation steps generate structural diversity in GPLs of MAC (11). However, because of their complexity, most of the biosynthetic pathways for ssGPLs have not been fully determined. We recently showed that the biosynthesis of nsGPLs was regulated by a combination of glycosyltransferases (31). Therefore, each glycosyltransferase might mediate a specific step in the biosynthesis of ssGPLs.

In terms of biological activity, it has been reported that the properties of ssGPLs are notably different from each other and that some of the properties play a role in affecting host responses to MAC infections (3, 5, 21, 27, 37, 38). Moreover, epidemiological studies have shown that serovars 1, 4, and 8 are distributed predominantly in North America and are also frequently isolated from AIDS patients (24, 39, 41). However, in contrast to other ssGPLs, the serovar 8-specific GPL is reported to be able to induce altered immune responses (3, 21). The biosynthetic pathway for serovar 8-specific GPL, particularly its oligosaccharide portion that includes a unique 4,6-O-(1-carboxyethylidene)-3-O-methyl-glucose (Glc) residue (7, 8) that may determine the specificity of serovar 8, remains unknown (Table 1). In this study, we investigated the genomic region assumed to be associated with the biosynthesis of GPL in MAC serovar 8 strain and identified the genes involved in the glycosylation pathway leading to the formation of serovar 8-specific GPL.

TABLE 1.

Oligosaccharide structures of serovar 1- and 8-specific GPLs

Serovar	Oligosaccharide	Reference(s)
1	α-l-Rha-(1→2)-l-6-d-Tal	17
8	4,6-O-(1-carboxyethylidene)-3-O-methyl-β-d-Glc-(1→3)-α-l-Rha-(1→2)-l-6-d-Tal	7, 8

MATERIALS AND METHODS

Bacterial strains, culture conditions, and DNA manipulation.

Table 2 shows the bacterial strains and vectors used in this study. MAC strains were grown in Middlebrook 7H9 broth (Difco) with 0.05% Tween 80 supplemented with 10% Middlebrook ADC enrichment (BBL). Recombinant M. smegmatis strains used for GPL production were cultured in Luria-Bertani broth with 0.2% Tween 80. Isolation of DNA and transformation of M. smegmatis strains were performed as previously described (32). The genomic regions of MAC strains were amplified by a two-step PCR using TaKaRa LA Taq with GC buffer and the following program: denaturation at 98°C for 20 s and annealing-extension at 68°C for an appropriate time depending on the length of the targeted region. Escherichia coli strain DH5α was used for routine manipulation and propagation of plasmid DNA. When necessary, antibiotics were added as follows: kanamycin, 50 μg/ml for E. coli and 25 μg/ml for M. smegmatis; and hygromycin B, 150 μg/ml for E. coli and 75 μg/ml for M. smegmatis. Oligonucleotide primers used in this study are listed in Table 3.

TABLE 2.

Bacterial strains and vectors used in this study

Strain or vector	Characteristics	Source or reference
Bacteria
E. coli DH5α	Cloning host	TaKaRa
M. smegmatis mc2155	Expression host	35
M. intracellulare ATCC 35771	MAC serovar 8 strain	29
M. avium JATA51-01	Source of rtfA	17
Vectors
pYM301	Source of pYM301a	30
pYM301a	Site-specific integrating mycobacterial vector carrying an hsp60 promoter cassette and AflII site	This study
pMV261	E. coli-Mycobacterium shuttle vector carrying an hsp60 promoter cassette	36
pMV-rtfA	pMV261 with rtfA	This study
pYM-gtfTB	pYM301a with gtfTB	This study
pYM-gtfTB-orf3-orf4	pYM301a with gtfTB, ORF3, and ORF4	This study

TABLE 3.

Oligonucleotide primers used in this study

Primer	Sequencea	Restriction site
RTFA-S	5′-CGGGATCCCATGAAATTTGCTGTGGCAAG-3′	BamHI
RTFA-A	5′-AACTGCAGCTCAGCGACTTCGCTGCGCTTC-3′	PstI
GTFTB-S	5′-AACTGCAGAAATGACCGCCACAACCAGGGC-3′	PstI
GTFTB-A	5′-GGAATTCTCAGGCGCTCAGTGGCTCGTC-3′	EcoRI
ORF4-A	5′-GGAATTCCTAGGGCGCCAATTCGATGAG-3′	EcoRI
GTFB-U4	5′-GGAATTCGGTCGACTCGACGAAGCCGAC-3′	EcoRI
DRRC-A	5′-GGAATTCTGCAGGCGGGGCGACTCCTGCT-3′	EcoRI

^aUnderlining indicates restriction sites.

Construction of expression vectors.

The rtfA gene was amplified from genomic DNA of M. avium strain JATA51-01 using primers RTFA-S and RTFA-A. The PCR products were digested with each restriction enzyme and cloned into the BamHI-PstI site of pMV261 to obtain pMV-rtfA. To use the site-specific integrating mycobacterial vector more conveniently, we constructed pYM301a containing an AflII site in pYM301. The region encompassing gtfTB, ORF3, and ORF4 was amplified from genomic DNA of MAC serovar 8 strain ATCC 35771 using primers GTFTB-S and ORF4-A. In addition, gtfTB was amplified using primers GTFTB-S and GTFTB-A. The PCR products were digested with each restriction enzyme and cloned into the PstI-EcoRI site of pYM301a to obtain pYM-gtfTB-orf3-orf4 and pYM-gtfTB (Table 2).

Isolation and purification of GPLs.

Harvested bacterial cells were allowed to stand in CHCl₃-CH₃OH (2:1, vol/vol) for several hours at room temperature. After water was added, total-lipid extracts were obtained from the organic phase and evaporated to dryness. Total-lipid extracts were subjected to mild alkaline hydrolysis as previously described (32, 33) to obtain crude GPL extracts. For analytical thin-layer chromatography (TLC), crude GPLs obtained from the same wet weight of harvested bacterial cells were spotted on Silica Gel 60 plates (Merck) using CHCl₃-CH₃OH-H₂O (30:8:1, vol/vol/vol) as the solvent and were visualized by spraying the plates with 10% H₂SO₄ and charring. Purified GPLs were prepared from crude GPLs by preparative TLC on the same plates, and each GPL was extracted from the corresponding band. Perdeuteriomethylation was carried out as previously described (6, 12, 17).

GC-MS and MALDI-TOF MS analysis.

Crude and purified GPLs were hydrolyzed in 2 M trifluoroacetic acid (2 h, 120°C), and the released sugars were reduced with NaBD₄ and then acetylated with pyridine-acetic anhydride (1:1, vol/vol) at room temperature overnight. The resulting alditol acetates were separated and analyzed by gas chromatography-mass spectrometry (GC-MS) with a TRACE DSQ (Thermo Electron) equipped with an SP-2380 column (Supelco) using helium gas. The following program was used: temperature increased from 52 to 172°C at a rate of 40°C/min and then increased from 172 to 250°C at a rate of 3°C/min. To determine the total mass of the purified GPLs, matrix-assisted laser desorption ionization—time of flight (MALDI-TOF) mass spectra (in the positive mode) were obtained with a QSTAR XL (Applied Biosystems) using a pulse laser with emission at 337 nm. Samples mixed with 2,5-dihydroxybenzoic acid as the matrix were analyzed in the reflectron mode with an accelerating voltage of 20 kV and with operation in positive ion mode.

Nucleotide sequence accession number.

The 4.6-kb genomic region amplified from MAC serovar 8 strain ATCC 35771 using primers GTFB-U4 and DRRC-A has been deposited in the DDBJ nucleotide sequence database under accession number AB437139.

RESULTS

Isolation and sequencing of the 4.6-kb genomic region responsible for GPL biosynthesis in MAC serovar 8.

Lacking information on the genes responsible for biosynthesis of serovar 8-specific GPL, we compared and analyzed the genomic regions likely to be responsible for GPL biosynthesis in several MAC serovars (16, 28). Most of these regions have high homology to each other, while the segment between the gtfB and drrC genes was found to vary in the strains. Therefore, we assumed that this segment contains genes involved in the formation of the unique Glc residue in serovar 8-specific GPL. To clone the gtfB-drrC region by using PCR, we designed various primers containing sequences derived from other MAC strains. By examining combinations of several pairs of primers, a 4.6-kb fragment was amplified from genomic DNA of a MAC serovar 8 strain when primers GTFB-U4 and DRRC-A were used (Fig. 1). Sequencing of this 4.6-kb fragment revealed four complete open reading frames (Fig. 1). The deduced amino acid sequences encoded by ORF1, ORF2, ORF3, and ORF4 were found to be identical to the amino acid sequences of four functionally undefined proteins from M. avium strain 104, MAV_3253, MAV_3255, MAV_3256, and MAV_3257, respectively (GenBank accession no. NC_008595.1). M. avium strain A5 also possessed a genomic region harboring ORF2, ORF3, and ORF4 (GenBank accession no. AY130970.1). These four open reading frames are predicted to encode the following proteins: ORF1, a putative glycosyltransferase similar to GtfD, which has been identified as a fucosyltransferase involved in the biosynthesis of serovar 2-specific GPL (73% identity) (30); ORF2, a putative glycosyltransferase, designated GtfTB, showing high homology to Rv1516c of M. tuberculosis (61% identity) (28); ORF3, a putative polysaccharide pyruvyltransferase similar to MSMEG_4736 and MSMEG_4737 of M. smegmatis (61 and 58% identity, respectively) (GenBank accession no. NC_008596.1); and ORF4, a putative O-methyltransferase similar to MSMEG_4739 of M. smegmatis (55% identity) (GenBank accession no. NC_008596.1).

FIG. 1.

Organization of the 4.6-kb genomic region isolated from MAC serovar 8 strain. Filled triangles indicate the primers used for PCR amplification.

Identification of the genes required for synthesis of the sugar residue unique to serovar 8-specific GPL.

Based on the deduced functions of the genes in the 4.6-kb fragment, we focused on gtfTB (ORF2), ORF3, and ORF4 and characterized them by performing expression analyses. Because the serovar 8-specific GPL has a structure in which the Rha residue of serovar 1-specific GPL is further glycosylated (Table 1), it was necessary to prepare a strain producing serovar 1-specific GPL that could be the substrate for the enzymes participating in the biosynthesis of serovar 8-specific GPL. For this, as previously demonstrated, we created a recombinant M. smegmatis strain, designated MS-S1, by introducing the plasmid vector pMV-rtfA having the M. avium rtfA gene, which converts nsGPLs to serovar 1-specific GPL (30). We then introduced the integrative expression vector pYM-gtfTB possessing gtfTB into MS-S1 and assessed GPL profiles by performing a TLC analysis (Fig. 2). By comparison with the profile of MS-S1/pYM301a (vector control) (Fig. 2, lane A), two new spots, designated spots GPL-SG-U and -D, were observed in MS-S1/pYM-gtfTB (Fig. 2, lane B), indicating that serovar 1-specific GPL was converted to structurally different compounds by expression of gtfTB. Moreover, when the expression vector pYM-gtfTB-orf3-orf4 containing gtfTB, ORF3, and ORF4 was introduced into MS-S1, another new spot, designated GPL-S8, appeared (Fig. 2, lane C), implying that the structure of GPL-SG-U and -D was further modified by the products of ORF3 and ORF4. To confirm that these compounds contain the sugar residues associated with serovar 8-specific GPL, we performed a GC-MS analysis of the monosaccharides released from crude GPL extracts of each recombinant strain and the MAC serovar 8 strain (Fig. 3). The results showed that there was an excess of Glc, together with Rha, 6-d-Tal, 3,4-di-O-methyl-Rha, and 2,3,4-tri-O-methyl-Rha, in the profile of MS-S1/pYM-gtfTB compared with other profiles, as well as minor Glc peaks presumably derived from traces of trehalose-containing glycolipids (Fig. 3B). This indicates that the gtfTB gene mediates the transfer of a Glc residue to serovar 1-specific GPL. In contrast, the profile of MS-S1/pYM-gtfTB-orf3-orf4 revealed the presence of 4,6-O-(1-carboxyethylidene)-3-O-methyl-Glc, which was also detected in the MAC serovar 8 strain (Fig. 3C and D), demonstrating that the three genes are associated with the formation of the unique sugar residue of serovar 8-specific GPL.

FIG. 2.

TLC of crude GPL extracts from recombinant M. smegmatis strains MS-S1/pYM301a (A), MS-S1/pYM-gtfTB (B), and MS-S1/pYM-gtfTB-orf3-orf4 (C). GPL extracts were prepared from the total-lipid fraction, and this was followed by mild alkaline hydrolysis. Samples were spotted and developed using CHCl₃-CH₃OH-H₂O (30:8:1, vol/vol/vol).

FIG. 3.

GC-MS of alditol acetate derivatives from crude GPL extracts of recombinant strains M. smegmatis MS-S1/pYM301a (A), MS-S1/pYM-gtfTB (B), and MS-S1/pYM-gtfTB-orf3-orf4 (C) and a MAC serovar 8 strain (D). GPL extracts were prepared from the total-lipid fraction, and this was followed by mild alkaline hydrolysis. Asterisks indicate noncarbohydrates. Me, methyl.

Functional characterization of gtfTB.

Expression analysis showed that serovar 1-specific GPL was converted to new compounds containing Glc when the gtfTB gene was expressed (Fig. 2, lane B, and Fig. 3B). Although these results suggested that the product of gtfTB participates in the formation of a Glc residue, it is not clear whether gtfTB encodes the glycosyltransferase that transfers Glc via 1→3 linkage to the Rha residue of serovar 1-specific GPL, whose linkage was previously detected in serovar 8-specific GPL (7, 8). To elucidate the function of gtfTB, we determined the linkage of sugar moieties of GPL-SG-U and -D, which were produced by recombinant strain MS-S1/pYM-gtfTB (Fig. 2, lane B). After extraction of the products from the corresponding bands on the TLC plate, purified GPL-SG-U and -D were subjected to perdeuteriomethylation followed by GC-MS. The differences in the TLC profiles of GPL-SG-U and -D might have been due to the presence or absence of fatty acid methylation, which is often observed in M. smegmatis GPLs (23, 31), whereas the GC-MS profiles and fragmentation ions for GPL-SG-U and -D were identical, demonstrating that GPL-SG-U and -D had the same sugar moieties and linkages. Therefore, the profiles of GPL-SG-U shown here are representative of GPL-SG-U and -D. The GC-MS profile of GPL-SG-U contained four peaks corresponding to 6-d-Tal, Rha, Glc, and 2,3,4-tri-O-methyl-Rha (data not shown). The characteristic spectra for Glc, Rha, and 6-d-Tal are shown in Fig. 4. The spectrum of Glc had fragment ions at m/z 121, 167, and 168, which represent the presence of deuteriomethyl groups at positions C-2, C-3, and C-4 (Fig. 4A). In contrast, fragment ions at m/z 121, 134, 193, and 240 were detected for Rha, indicating that a deuteriomethyl group was introduced at positions C-2 and C-4 of Rha, in which position C-3 was acetylated (Fig. 4B). In addition, detection of fragment ions at m/z 134, 181, and 193 (Fig. 4C) revealed that there was deuteriomethylation at positions C-3 and C-4 in 6-d-Tal. These results demonstrated that position C-1 of Glc is linked to position C-3 of Rha but not to position C-2 of 6-d-Tal, because it has been determined previously that position C-1 of Rha is linked to position C-2 of 6-d-Tal in the oligosaccharide of serovar 1-specific GPL (17). Accordingly, the oligosaccharide structures of GPL-SG-U and -D were determined to have Glc-(1→3)-Rha-(1→2)-6-d-Tal at D-allo-Thr, demonstrating that gtfTB encodes the glucosyltransferase that transfers a Glc residue via 1→3 linkage to the Rha residue of serovar 1-specific GPL.

FIG. 4.

GC-MS spectra and fragment ion assignments for Glc (A), Rha (B), and 6-d-Tal (C), which were derived from alditol acetates of sugars released from deuteriomethylated GPL-SG-U. Ac, acetate; D, deuterium.

Structural assignment of GPL-S8 synthesized by expression of gtfTB, ORF3, and ORF4.

GC-MS of the crude GPL extract from MS-S1/pYM-gtfTB-orf3-orf4 revealed the presence of 4,6-O-(1-carboxyethylidene)-3-O-methyl-Glc (Fig. 3C). To confirm that this structural component was derived from GPL-S8, we performed GC-MS and MALDI-TOF MS analyses of purified GPL-S8. The results showed that GPL-S8 contained a 4,6-O-(1-carboxyethylidene)-3-O-methyl-Glc residue and two main pseudomolecular ions (m/z 1,565.9 and 1,579.8 [M + Na]⁺) (data not shown). Consequently, as shown in Fig. 5, these results were consistent with the proposed structure for GPL-S8-1 and -2 containing 4,6-O-(1-carboxyethylidene)-3-O-methyl-Glc, with differences in pseudomolecular ions due to fatty acid methylation.

FIG. 5.

Proposed structure and biosynthetic genes of GPL-S8 (serovar 8-specific GPL). Me, methyl.

DISCUSSION

Structural diversity of the ssGPLs, notably in their sugar residues, defines 28 serovars of MAC. Although these ssGPLs are known to contribute to the virulence of MAC, the mechanisms of their biosynthetic regulation are largely unknown. In this study, we clarified the biosynthetic pathway for serovar 8-specific GPL, specifically the glycosylation step in which a Glc residue is transferred to the Rha residue of serovar 1-specific GPL.

To isolate the genomic region associated with the biosynthesis of serovar 8-specific GPL, we compared the GPL biosynthetic gene clusters in several MAC strains and found significant differences in the gtfB-drrC region. The segment flanking the 3′ end of the gtfB-drrC region includes several genes responsible for the serovar 1-specific GPL whose structure is found in all ssGPLs. On the other hand, it is experimentally clarified that the gtfB-drrC regions of serovar 2-, 7-, and 16-specific GPL-producing strains contain the genes involved in the formation of the specific sugar residues that are transferred to the Rha residue of serovar 1-specific GPL (18, 19, 30). Thus, this region could play an important role in generating the structural diversity of ssGPLs. As shown in this study, the specific functions for formation of sugar moieties of serovar 8-specific GPL were due to the genes present in the gtfB-drrC region, suggesting that focusing on this region might provide clues for elucidating the characteristics of other ssGPLs whose biosynthesis is still not known.

It has been reported previously that the gtfTB gene in M. avium strains 104 and A5 was not likely to be associated with GPL biosynthesis because its ancestral homologue, Rv1516c (61% identity with the GtfTB gene), was the gene of M. tuberculosis, which produces no GPLs (28). Thus, it was interesting that gtfTB encodes a glycosyltransferase that does participate in GPL biosynthesis in which a Glc residue is transferred to serovar 1-specific GPL, yielding the serovar 8-specific GPL. M. avium strains 104 and A5 synthesize serovar 1-specific GPL as a final product and intermediate, respectively, while it has been recognized that neither of these strains produces serovar 8-specific GPL in spite of the presence of gtfTB in the GPL biosynthetic gene cluster (28). These observations raised the possibility that the transcription of gtfTB is inefficient in both strains due to the upstream sequences. Actually, in M. avium strain 104, a transposase sequence was observed upstream of gtfTB, indicating that this strain might be deficient in glucosylation, and consequently a serovar 1-specific GPL-producing strain is obtained (28). On the other hand, it has been shown that the biosynthetic gene cluster for serovar 7-specific GPL in M. intracellulare strain ATCC 35847 contains a putative glycosyltransferase gene which encodes amino acid sequences that are similar to the amino acid sequences encoded by gtfTB (59% identity) (18). Structural analysis of sugar moieties in serovar 7-specific GPL indicated that this GtfTB homologue may serve as a glycosyltransferase during formation of the terminal amidohexose residue that structurally resembles Glc (18).

The deduced amino acid sequences encoded by ORF3 and ORF4 showed that these genes putatively encode polysaccharide pyruvyltransferase and O-methyltransferase, respectively. Expression of ORF3 and ORF4 together with gtfTB led to structural alterations in which Glc was modified with both 4,6-O-(1-carboxyethylidene) and 3-O-methyl groups. Based on these observations, it is strongly suggested that ORF3 is associated with the formation of the 4,6-O-(1-carboxyethylidene) group that is synonymous with the cyclic pyruvate ketal and that ORF4 is associated with the 3-O-methylation of the Glc residue (Fig. 5). In mycobacteria, homologues of ORF3 and ORF4 were found only in M. smegmatis, as MSMEG_4736 (for ORF3), MSMEG_4737 (for ORF3), and MSMEG_4739 (for ORF4). M. smegmatis also produces glycolipids containing 4,6-O-(1-carboxyethylidene)-3-O-methyl-Glc as a sugar moiety (25, 34), which suggests that both homologues participate in the synthesis of these glycolipids. Sugar residues with a 4,6-O-(1-carboxyethylidene) group substitution have been found in carbohydrates such as extracellular polysaccharide and N-linked glycan, which are produced by some bacteria and yeasts (1, 15, 20, 22, 26). It has been shown that an increase in 4,6-O-(1-carboxyethylidene)-containing sugar residues leads to enhanced viscosity of extracellular polysaccharide from Xanthomonas sp., which alters the cell surface properties related to cellular attachment and protection from environmental stress (10). Accordingly, in terms of the properties of serovar 8-specific GPL, the presence of the 4,6-O-(1-carboxyethylidene) group might influence the pathogenicity of MAC serovar 8.

With regard to the antibody reactivity, it is unclear whether serovar 8-specific antibodies react with GPL-S8 because there are minor structural differences in the methylated positions of fatty acids and the terminal Rha residue linked to the tetrapeptide between GPL-S8 and serovar 8-specific GPL of MAC. Evaluation of the antibody response to GPL-S8 using serovar 8-specific antibodies would facilitate understanding the immunoreactivity mediated by ssGPLs.

In this study, we proved that gtfTB and adjacent genes in the GPL biosynthetic gene cluster in MAC serovar 8 strain are responsible for the formation of a unique glucose residue in serovar 8-specific GPL (Fig. 5). In particular, gtfTB encodes the glucosyltransferase that plays a critical role in the pathway leading from serovar 1-specific GPL to serovar 8-specific GPL. Through further study, including generation of gtfTB knockout mutants of MAC serovar 8 strains, results relevant to the biosynthesis of serovar 8-specific GPL might help clarify the biological function of ssGPLs and their role in the host-pathogen relationships of MAC.