Mycobacterium marinum MMAR_2380, a predicted transmembrane acyltransferase, is essential for the presence of the mannose cap on lipoarabinomannan

Abstract

Lipoarabinomannan (LAM) is a major glycolipid in the mycobacterial cell envelope. LAM consists of a mannosylphosphatidylinositol (MPI) anchor, a mannan core and a branched arabinan domain. The termini of the arabinan branches can become substituted with one to three α(1→2)-linked mannosyl residues, the mannose cap, producing ManLAM. ManLAM has been associated with a range of different immunomodulatory properties of Mycobacterium tuberculosis during infection of the host. In some of these effects, the presence of the mannose cap on ManLAM appears to be crucial for its activity. So far, in the biosynthesis of the mannose cap on ManLAM, two enzymes have been reported to be involved: a mannosyltransferase that adds the first mannosyl residue of the mannose caps to the arabinan domain of LAM, and another mannosyltransferase that elongates the mannose cap up to three mannosyl residues. Here, we report that a third gene is involved, MMAR_2380, which is the Mycobacterium marinum orthologue of Rv1565c. MMAR_2380 encodes a predicted transmembrane acyltransferase. In M. marinum ΔMMAR_2380, the LAM arabinan domain is still intact, but the mutant LAM lacks the mannose cap. Additional effects of mutation of MMAR_2380 on LAM were observed: a higher degree of branching of both the arabinan domain and the mannan core, and a decreased incorporation of [1,2-(14)C]acetate into the acyl chains in mutant LAM as compared with the wild-type form. This latter effect was also observed for related lipoglycans, i.e. lipomannan (LM) and phosphatidylinositol mannosides (PIMs). Furthermore, the mutant strain showed increased aggregation in liquid cultures as compared with the wild-type strain. All phenotypic traits of M. marinum ΔMMAR_2380, the deficiency in the mannose cap on LAM and changes at the cell surface, could be reversed by complementing the mutant strain with MMAR_2380. Strikingly, membrane preparations of the mutant strain still showed enzymic activity for the arabinan mannose-capping mannosyltransferase similar to that of the wild-type strain. Although the exact function of MMAR_2380 remains unknown, we show that the protein is essential for the presence of a mannose cap on LAM.

Fig. 1.

Absence of the mannose cap on mutant LAM. (a) Cell lysates from M. marinum wild-type, ΔMMAR_2380 and complemented ΔMMAR_2380 (ΔMMAR_2380 comp.) strains and M. smegmatis wild-type (control for ‘capless’ LAM) were immunoblotted with α-AraLAM antibody F30-5, which recognizes LAM, and α-ManLAM antibody 55.92.1A1, which recognizes the mannose cap. (b) Mannooligosaccharide cap analysis of LAM. Purified and partially degraded LAM was analysed for the presence of the mannose caps by CE. Shown are the profiles of LAM purified from M. marinum E11 (trace 1) and M. marinum ΔMMAR_2380 (trace 2). A, Ara-APTS; M, Man-APTS; S, internal standard, mannoheptose-APTS; AM, Manp-α(1→5)-Ara-APTS (monomannoside cap); AMM, Manp-α(1→2)-Manp-α(1→5)-Ara-APTS (dimannoside cap).

Fig. 2.

Enzymic analysis. Mannosyltransferase capping assay with synthetic Ara₆ as acceptor and membranes from M. marinum wild-type, ΔMMAR_2380 and complemented ΔMMAR_2380 (ΔMMAR_2380 comp.), and as control, membranes from M. marinum wild-type without acceptor. Shown are means of triplicates; error bars, sd.

Fig. 3.

NMR analysis of the glyco part of LAM from M. marinum wild-type (pattern 1) and ΔMMAR_2380 (pattern 2). 2D ¹H-¹³C heteronuclear multiple quantum coherence spectroscopy (1H-¹³C HMQC) spectra of LAM in D₂O at 313 K are shown with expanded regions (δ ¹H, 4.80–5.35; δ ¹³C, 99–113). The different resonances are labelled with the abbreviated name of the corresponding glycosyl units. See Supplementary Fig. S1 for structure of ManLAM.

Fig. 4.

Reduced [1,2-¹⁴C]acetate incorporation in the polar lipids PIM₆, LM and LAM from M. marinum ΔMMAR_2380. (a) Extracted crude lipoglycans from acetate-labelled delipidated cells (25 μg) from M. marinum wild-type, ΔMMAR_2380 and complemented ΔMMAR_2380 (ΔMMAR_2380 comp.) were counted for the incorporation of [1,2-¹⁴C]acetate and analysed by SDS-PAGE/autoradiography. Shown is the average of two independent experiments. (b) [1,2-¹⁴C]Acetate-labelled M. marinum cultures were processed and polar lipids were applied (25 000 c.p.m.) to the corners of 6.6×6.6 cm pieces of aluminium-backed TLC plates and analysed using the 2D solvent system E as described in Methods. Plates were dried and autoradiograms were produced by overnight exposure of Kodak X-Omat AR film to the TLC plates to reveal [1,2-¹⁴C]acetate-labelled lipids. PI, non-mannosylated, diacylated phosphatidylinositol anchor; PIM₂, phosphatidylinositol dimannoside; PIM₆, phosphatidylinositol hexamannoside; Ac₁PIM, tri-acylated PIM; Ac₂PIM, tetra-acylated PIM. Unassigned spots in the figure are lipooligosaccharides, and in the upper-right corner are diphosphatidylglycerol, phosphatidylethanolamine and unknown phospholipids (Burguière et al., 2005). Two independently prepared lipid extracts per strain were tested.

Fig. 5.

Physical appearance of liquid-grown M. marinum wild type, ΔMMAR_2380 and complemented ΔMMAR_2380 (ΔMMAR_2380 comp.). (a) OD₆₀₀ measurements during growth in 7H9+0.05 % Tween 80 with agitation. Shown are means of three independently grown cultures per strain; error bars sd. (b) M. marinum ΔMMAR_2380 shows increased aggregation as compared with the wild-type and the complemented strains (see Supplementary Fig. S4 for colour pictures).