Role of LmeA, a Mycobacterial Periplasmic Protein, in Maintaining the Mannosyltransferase MptA and Its Product Lipomannan under Stress

Abstract

The mycobacterial cell envelope has a diderm structure, composed of an outer mycomembrane, an arabinogalactan-peptidoglycan cell wall, a periplasm, and an inner membrane. Lipomannan (LM) and lipoarabinomannan (LAM) are structural and immunomodulatory components of this cell envelope. LM/LAM biosynthesis involves a number of mannosyltransferases and acyltransferases, and MptA is an α1,6-mannosyltransferase involved in the final extension of the mannan chain. Recently, we reported the periplasmic protein LmeA being involved in the maturation of the mannan backbone in Mycobacterium smegmatis Here, we examined the role of LmeA under stress conditions. We found that lmeA transcription was upregulated under two stress conditions: stationary growth phase and nutrient starvation. Under both conditions, LAM was decreased, but LM was relatively stable, suggesting that maintaining the cellular level of LM under stress is important. Surprisingly, the protein levels of MptA were decreased in an lmeA deletion (ΔlmeA) mutant under both stress conditions. The transcript levels of mptA in the ΔlmeA mutant were similar to or even higher than those in the wild type, indicating that the decrease of MptA protein was a posttranscriptional event. The ΔlmeA mutant was unable to maintain the cellular level of LM under stress, consistent with the decrease in MptA. Even during active growth, overexpression of LmeA led the cells to produce more LM and become more resistant to several antibiotics. Altogether, our study reveals the roles of LmeA in the homeostasis of the MptA mannosyltransferase, particularly under stress conditions, ensuring the stable expression of LM and the maintenance of cell envelope integrity.IMPORTANCE Mycobacteria differentially regulate the cellular amounts of lipoglycans in response to environmental changes, but the molecular mechanisms of this regulation remain unknown. Here, we demonstrate that cellular lipoarabinomannan (LAM) levels rapidly decline under two stress conditions, stationary growth phase and nutrient starvation, while the levels of another related lipoglycan, lipomannan (LM), stay relatively constant. The persistence of LM under stress correlated with the maintenance of two key mannosyltransferases, MptA and MptC, in the LM biosynthetic pathway. We further showed that the stress exposures lead to the upregulation of lmeA gene expression and that the periplasmic protein LmeA plays a key role in maintaining the enzyme MptA and its product LM under stress conditions. These findings reveal new aspects of how lipoglycan biosynthesis is regulated under stress conditions in mycobacteria.

Keywords: Mycobacterium; biosynthesis; glycolipids; mannose; stress response.

FIG 1

M. tuberculosis LmeA is the functional ortholog of M. smegmatis LmeA. The ΔlmeA M. smegmatis strain was complemented with the M. tuberculosis LmeA ortholog (rv0817c) expressed at the mycobacteriophage L5 attB site. (A) SDS-PAGE analysis of LM and LAM. Two independent clones of the ΔlmeA::Mtb_lmeA strain were able to restore mature LM formation (arrow). (B) HPTLC analysis of PIMs. PIM profiles were not affected by complementation of Mtb_lmeA-HA.

FIG 2

Expression of the lmeA transcript and cellular levels of LM/LAM during stationary growth phase. (A) Relative expression of the lmeA transcript normalized to that of a housekeeping gene, gyrB. A significant increase in lmeA expression is noted by 72 h of growth in comparison to that at log phase (18 h). **, P < 0.05, by ANOVA. (B) SDS-PAGE analysis of LM/LAM during stationary phase. A representative image of a biological triplicate is shown. LAM decreased significantly more in the stationary phase than LM did. (C and D) Dot plots showing intensities of LM (C) and LAM (D) from biological replicates. Signal intensities were quantified using ImageQuant and normalized to the intensity of the WT at log phase. The averages are shown as horizontal bars. **, P < 0.01; *, P < 0.05, by ANOVA. ΔlmeA::C, ΔlmeA::P_nativelmeA-HA strain (also see the text).

FIG 3

Protein levels of MptA decrease significantly in the ΔlmeA mutant during stationary phase. (A and B) Western blot analysis of MptA (A) and MptC (B) over a 72-h time course of growth. MptA shows a significant decrease in the ΔlmeA mutant. MptC remains fairly constant between the WT, ΔlmeA mutant, and ΔlmeA::P_nativelmeA-HA strain. For SDS-PAGE, the same concentration of protein was loaded into each lane. (C) Relative expression levels of the mptA transcript were normalized to that of gyrB. A dramatic decrease in mRNA transcript levels is noted between log phase and 48 h (stationary phase). ***, P < 0.001; **, P < 0.05, by ANOVA. At each time point, there were no significant differences among the WT, ΔlmeA mutant, and ΔlmeA::P_nativelmeA-HA strain. ΔlmeA::C, ΔlmeA::P_nativelmeA-HA strain.

FIG 4

Effect of starvation on cellular levels of LM/LAM and their biosynthetic enzymes. (A to C) SDS-PAGE analysis of LM/LAM. LAM decreases during starvation, while LM remains constant. A representative image of biological triplicate is shown. LM (B) and LAM (C) from a biological triplicate were quantified, and changes relative to WT levels in log phase are shown in dot plots. Averages are shown as horizontal bars. (D) Western blot analysis of MptA in log phase and after starvation. MptA is decreased in the ΔlmeA mutant in comparison to levels in the WT during starvation. (E) Western blot of MptC in log phase and after starvation. No significant decrease was noted among the WT, ΔlmeA mutant, and ΔlmeA::P_nativelmeA-HA strain before and after starvation. Equal amounts of protein were loaded for Western blot analysis. (F) Transcript levels of mptA relative to levels in the WT in log phase, normalized to gyrB levels. The levels of the mptA transcript decreased from log phase to starvation, but deletion of the lmeA gene had no apparent impact. **, P < 0.01, by ANOVA. ΔlmeA::C, ΔlmeA::P_nativelmeA-HA strain.

FIG 5

Impact of overexpression of LmeA on LM/LAM biosynthesis. (A) Anti-HA Western blot to detect LmeA-HA expressed in the ΔlmeA::P_nativelmeA-HA strain (indicated in the figure as ΔlmeA::C) and LmeA-HA overexpression strain (LmeA OE). LmeA-HA was ∼10-fold more intense in LmeA OE than in the ΔlmeA::P_nativelmeA-HA strain. (B) Western blot analysis of MptA in the actively growing WT, ΔlmeA, ΔlmeA::P_nativelmeA-HA, and LmeA-HA OE strains. The same amount of protein was loaded into each lane. A representative image of a biological triplicate is shown. ΔlmeA::C, the ΔlmeA::P_nativelmeA-HA strain. (C) Dot plot showing relative abundances of MptA in a biological triplicate. Intensity was normalized to that of the WT, and the average is shown as a horizontal bar. *, P < 0.05; **, P < 0.01, by ANOVA. (D) Cellular levels of LM and LAM in the actively growing WT, ΔlmeA, ΔlmeA::P_nativelmeA-HA, and LmeA OE strains. A representative image of three biological replicates is shown. (E) Dot plot of cellular LM/LAM levels. The average from biological replicates is shown as a horizontal bar. **, P < 0.01, by ANOVA.

FIG 6

Proposed model of how LmeA functions during stress conditions. Under active growth conditions, LmeA aids MptA in its function, potentially through either maintaining MptA or facilitating the trafficking of LM/LAM or both. Upon stress exposure, such as starvation or stationary phase, LmeA prevents the decline of the cellular MptA level. IM, inner membrane; PG, peptidoglycan; AG, arabinogalactan; OM, outer membrane.