PapA3 is an acyltransferase required for polyacyltrehalose biosynthesis in Mycobacterium tuberculosis

Abstract

Mycobacterium tuberculosis possesses an unusual cell wall that is replete with virulence-enhancing lipids. One cell wall molecule unique to pathogenic M. tuberculosis is polyacyltrehalose (PAT), a pentaacylated, trehalose-based glycolipid. Little is known about the biosynthesis of PAT, although its biosynthetic gene cluster has been identified and found to resemble that of the better studied M. tuberculosis cell wall component sulfolipid-1. In this study, we sought to elucidate the function of papA3, a gene from the PAT locus encoding a putative acyltransferase. To determine whether PapA3 participates in PAT assembly, we expressed the protein heterologously and evaluated its acyltransferase activity in vitro. The purified enzyme catalyzed the sequential esterification of trehalose with two palmitoyl groups, generating a diacylated product similar to the 2,3-diacyltrehalose glycolipids of M. tuberculosis. Notably, PapA3 was selective for trehalose; no activity was observed with other structurally related disaccharides. Disruption of the papA3 gene from M. tuberculosis resulted in the loss of PAT from bacterial lipid extracts. Complementation of the mutant strain restored PAT production, demonstrating that PapA3 is essential for the biosynthesis of this glycolipid in vivo. Furthermore, we determined that the PAT biosynthetic machinery has no cross-talk with that for sulfolipid-1 despite their related structures.

FIGURE 1.

PAT and SL-1 share related structures and biosynthetic gene clusters. A, structure of PAT. B, structure of SL-1. C, genomic arrangement of the PAT and SL-1 biosynthetic gene clusters.

FIGURE 2.

PapA3 is an acyltransferase that sequentially palmitoylates trehalose in vitro. A, PapA3 was incubated with ¹⁴C-PCoA and either trehalose (Tre) or T2P. The reactions were analyzed by TLC and phosphorimaging. Two new products (1 and 2) were observed in the reaction with trehalose, but only product 2 was observed in the reaction with T2P. B, ESI-FT-ICR MS analysis of product 1 from the PapA3 reaction with trehalose. A product ion with m/z 615.32, corresponding to the m/z of a chloride adduct of synthetic T2P, was observed in the PapA3 reaction. In contrast, the control reaction lacking PapA3 showed no product at m/z 615. C, ESI-FT-ICR MS analysis of product 2 from the PapA3 reaction with Tre and T2P. An ion with m/z 853.54 was observed in both reactions but was not present in control reactions lacking PapA3.

FIGURE 3.

Linear ion trap MSⁿ of the monoacyl product ion from the reaction of PapA3 with trehalose (A) is consistent with that of synthetic T2P (B) and not that of synthetic T3P (C). MSⁿ analysis was performed in the positive ion mode using lithium-cation coordination. Shown are the MS³ spectra of the dissociation ions obtained from the cleavage of the glycosidic bond between the individual pyranose rings of trehalose.

FIGURE 4.

PAT biosynthesis requires papA3, but not stf0, in vivo. ESI-FT-ICR MS analysis of lipid extracts from wild type, ΔpapA3::papA3, and Δstf0 M. tuberculosis strains revealed the presence of characteristic PAT lipoforms that are absent from ΔpapA3 M. tuberculosis extracts.

FIGURE 5.

Proposed PAT biosynthetic pathway. PapA3 first acylates the 2-position of one of the glucose residues of trehalose with a palmitoyl group to form T2P. A mycolipenoyl group, synthesized by Pks3/4, is then transferred to the 3-position of T2P by PapA3 to generate 2,3-diacyltrehalose. 2,3-Diacyltrehalose may either be transported to the cell surface or serve as a biosynthetic intermediate that is further elaborated with mycolipenic acids to give PAT.