Metabolic tagging reveals surface-associated lipoproteins in mycobacteria

Abstract

Mycobacteria such as the causative agent of tuberculosis, Mycobacterium tuberculosis, encode over 100 bioinformatically predicted lipoproteins. Despite the importance of these posttranslationally modified proteins for mycobacterial survival, many remain experimentally unconfirmed. Here we characterized metabolic incorporation of diverse fatty acid analogues as a facile method of adding chemical groups that enable downstream applications such as detection, crosslinking and enrichment, of not only lipid-modified proteins, but also their protein interactors. Having shown that incorporation is an active process dependent on the lipoprotein biosynthesis pathway, we discovered that lipid-modified proteins are also located at the mycobacterial cell surface even though mycobacteria do not encode known lipoprotein transporters. These data have implications for uncovering a novel transport pathway and the roles of lipoproteins at the interface with the host environment. Our findings and the tools we developed will enable the further study of pathways related to lipoprotein function and metabolism in mycobacteria and other bacteria in which lipoproteins remain poorly understood.

Keywords: fatty acids; lipoproteins; mycobacteria; surface proteins.

Figure 1.. Fatty acids are incorporated into cell envelope proteins via the lipoprotein processing pathway.

A) Following export, primarily by the Sec secretion system, proteins are targeted for lipoprotein processing via a lipobox motif (dark grey fill) containing a conserved cysteine. In the first committed step, Lgt modifies the Cys sidechain with a diacylglyceryl group via a thioether linkage (blue). LspA then cleaves the signal peptide, leaving the Cys as the N-terminal residue (green). Lnt then esterifies the amino terminus with a fatty acid (red), yielding the final N-terminally triacylated protein product. The lipobox motif and phosphatidyl glycerol (for Lgt) and phosphatidyl ethanolamine (for Lnt) donors are presumed based on the canonical bacterial lipoprotein biosynthesis pathway. Fatty acid chains (16:0; 19:0) are shown as reported for PG and presumed for PE. One of two possible configurations for sn-1/sn-2 is displayed and tuberculostearic acid is depicted as the most likely assignment for 19:0. (Signal peptide, chemical structures, and proteins are not to scale relative to the plasma membrane.) B) Structures of modified fatty acids used in this study with commercial names and abbreviations (bold).

Figure 2.. Msm and Mtb incorporate the modified fatty acid BC12 into proteins.

A) Msm or Mtb was incubated with varying concentrations of BODIPY-C12 (BC12) for ~1 doubling time (2 h for Msm, 20 h for Mtb). Five μM BC12 was used for all subsequent experiments. Heat-killed (HK) cells served as a negative control and total lysates were analyzed by SDS-PAGE. B) Same as in A) except −/+ 10% OADC was supplemented during BC12 incubation.All samples in each panel were run on the same gel; lanes scanned in separate detection channels appear slightly separated. GroEL immunoblot was used as a loading control. Data are representative of at least n = 3 independent experiments.

Figure 3.. Modified fatty acids added to proteins are labile to base hydrolysis with some dependence on the Lgt MSMEG_3222 for incorporation.

Msm (A-D) wild type (“WT”), (C) ^ΔMSMEG_3222 null (“Δlgt”) or ΔMSMEG_3222::MSMEG_3222 complement (“C”) strains were incubated with 5 μM BC12, 20 μM azFA, or 20 μM palmitic acid as a negative control (“FA”) for 2 h before harvesting. Total lysates of azFA-treated cells were subjected to CuAAC with alk-AZ488 or alk-TB. Lysates were then treated with (A) 0.05 mg/mL proteinase K or (B) 0.1 M sodium hydroxide for 1h before inactivation or quenching and protein precipitation. All samples in each panel were run on the same gel; some lanes were scanned in separate detection channels. GroEL immunoblot was used as a loading control. Data are representative of n = 1–3 independent experiments.

Figure 4.. Enrichment of azFA-labeled proteins confirms modification of the lipoprotein LprG.

Msm expressing LprG-3xFLAG (MSMEG_3070–3xFLAG) or harboring empty vector control (EV) was cultured with azFA for 2 h. Total lysates were subjected to CuAAC with alk-TB followed by avidin enrichment. (A) Fluorescence scan for tetramethylrhodamine (TAMRA) and (B) anti-FLAG immunoblot. (I = input, O = output, ctrl = untreated total lysate from Msm LprG-3xFLAG as anti-FLAG positive control. Asterisk indicates non-specific cross-reacting protein.)

Figure 5.. Click coupling to detergent extracts and whole-cells reveals surface-accessible lipoproteins.

A) Msm was cultured without (−) or with (+) Tween 80, incubated with 20 μM azFA for 2 h. Cells were then extracted with 1% Tween 80. The resulting supernatant extract (S) and the total lysate of the remaining cell pellet (P) were subjected to CuAAC. Protein load for −/+ Tween pellet fractions were normalized approximately for the degree of observed labeling. B) Msm in standard +Tween medium was incubated with μM for 2 h. CuAAC to alk-AZ488 was then performed on whole cells and total lysates. Protein load for lysate and whole-cell samples were normalized approximately for the degree of observed labeling. (C, D) Same as in B) except azFA was incubated with Msm for the specified times prior to CuAAC to alk-AZ488 on whole cells. E) Same as in B) using Msm wild-type, ΔMSMEG_3222 (Δlgt), and complement ΔMSMEG_3222::MSMEG_3222 (C) strains.

Figure 6.. Incorporation of a photoclick-FA in Msm yields UV-dependent adducts.

Msm was incubated with palmitic acid (FA), azFA, alkFA, or pcFA for 30 min. and then UV irradiated for a total of 30 min. Total lysates were subjected to CuAAC with either alk-AZ488 or az-110 as indicated. GroEL immunoblot served as a loading control. Data are representative of n = 3 independent experiments.