Peptidoglycan Remodeling by an L,D-Transpeptidase, LdtD during Cold Shock in Escherichia coli

Abstract

Peptidoglycan (PG) is a unique and essential component of the bacterial cell envelope. It is made up of several linear glycan polymers cross-linked through covalently attached stem peptides making it a fortified mesh-like sacculus around the bacterial cytosolic membrane. In most bacteria, including Escherichia coli, the stem peptide is made up of l-alanine (l-Ala¹), d-glutamate (d-Glu²), meso-diaminopimelic acid (mDAP³), d-alanine (d-Ala⁴), and d-Ala⁵ with cross-links occurring either between d-ala⁴ and mDAP³ or between two mDAP³ residues. Of these, the cross-links of the 4-3 (d-Ala⁴-mDAP³) type are the most predominant and are formed by penicillin-binding D,D-transpeptidases, whereas the formation of less frequent 3-3 linkages (mDAP³-mDAP³) is catalyzed by L,D-transpeptidases. In this study, we found that the frequency of the 3-3 cross-linkages increased upon cold shock in exponentially growing E. coli and that the increase was mediated by an L,D-transpeptidase, LdtD. We found that a cold-inducible RNA helicase DeaD enhanced the cellular LdtD level by facilitating its translation resulting in an increased abundance of 3-3 cross-linkages during cold shock. However, DeaD was also required for optimal expression of LdtD during growth at ambient temperature. Overall, our study finds that E. coli undergoes PG remodeling during cold shock by altering the frequency of 3-3 cross-linkages, implying a role for these modifications in conferring fitness and survival advantage to bacteria growing in diverse environmental conditions. IMPORTANCE Most bacteria are surrounded by a protective exoskeleton called peptidoglycan (PG), an extensively cross-linked mesh-like macromolecule. In bacteria, such as Escherichia coli, the cross-links in the PG are of two types: a major fraction is of 4-3 type whereas a minor fraction is of 3-3 type. Here, we showed that E. coli exposed to cold shock had elevated levels of 3-3 cross-links due to the upregulation of an enzyme, LdtD, that catalyzed their formation. We showed that a cold-inducible RNA helicase DeaD enhanced the cellular LdtD level by facilitating its translation, resulting in increased 3-3 cross-links during cold shock. Our results suggest that PG remodeling contributes to the survival and fitness of bacteria growing in conditions of cold stress.

Keywords: DeaD; L,D-transpeptidases; LdtD; NlpI; cold shock; peptidoglycan.

FIG 1

Schematic depiction of peptidoglycan sacculus of E. coli. The figure represents a Gram-negative bacterial cell in which the OM was partially peeled off. PG is a highly cross-linked sac-like structure situated between the OM and IM in the periplasmic space. PG is made up of linear glycan strands of NAG (light blue) and NAM (dark blue) bonded through β,1-4 linkages. Each NAM residue is attached to a stem peptide, which is usually a tetrapeptide made up of l-Ala¹, d-Glu², mDAP³, and d-Ala⁴. Glycan strands are interlinked to each other via stem peptides cross-bridged either between d-Ala⁴ and mDAP³ (4-3; depicted as ‘a’) or between two mDAP³ residues (3-3; depicted as ‘b’). Cross-linking is indicated by a black line between the stem peptides.

FIG 2

Overexpression of LdtD or LdtE alleviated ΔnlpI phenotypes. (A) Growth of ΔnlpI mutant carrying the empty vector (ASKA plasmid; P_T5-lac::) or its ldtD or ldtE derivatives at 37°C on indicated plates with 50 μM IPTG. (B) Growth of ΔnlpI mutant carrying the empty vector (pTrc99a; P_trc::) or its derivatives harboring ldtD or ldtE on LB or LBON plates supplemented with 50 μM IPTG. (C) Growth curve of ΔnlpI mutant carrying the empty vector (pTrc99a) or its derivatives carrying ldtD or ldtE in LBON broth supplemented with 50 μM IPTG at 42°C. (D) Differential interference contrast (DIC) microscopic images of ΔnlpI cells with pTrc99a or its derivatives were grown in LBON containing 50 μM IPTG at 42°C. The scale bar represents 5 μm.

FIG 3

Deletion of nlpI is polar on the expression of downstream gene deaD. Western blots showing levels of (A) LdtD-Flag or (B) LdtE-Flag in the WT and ΔnlpI strain backgrounds. (C) Western blot showing an LdtD-Flag in the ΔnlpI mutant with plasmid-borne nlpI (pTrc99-nlpI) (D) LdtE-Flag levels in indicated strains. (E) Schematic representation indicating promoter sites of deaD (Yrbnp1 and p2) within the nlpI gene. Block arrows indicate the direction of transcription. (F) Western blot showing DeaD-HA levels in the WT and ΔnlpI mutant. (G) LdtD-Flag levels of the WT and ΔnlpI mutant with empty vector (P_T5-lac::) or vector carrying deaD grown in the presence of 50 μM IPTG. Indicated strains are grown in LB at 37°C until OD₆₀₀ of 1.0, harvested, and analyzed via Western blotting. NlpI levels were also measured in all the strains used. MepS are used as a positive control to confirm the NlpI complementation because it is known to accumulate in the absence of NlpI (23). FtsZ was used as a loading control. Bar diagrams indicate the relative quantification of respective protein levels from three replicates; *, P < 0.05; **, P < 0.005; ***, P < 0.001; ns (not significant); n = 3.

FIG 4

DeaD modulated LdtD levels and 3-3 cross-linkages in the PG sacculi. Western blots showing levels of (A) LdtD-Flag or (B) LdtE-Flag in the WT and ΔdeaD strain backgrounds. (C) Western blot of LdtD-Flag in the WT and ΔdeaD carrying the empty vector (pCA24N) or vector carrying the deaD gene (pCA24N-deaD) grown in the presence of 50 μM IPTG. FtsZ and NlpI were used as controls. Bar diagrams represent the respective quantified protein levels. *, P < 0.05; **, P < 0.005; ns (not significant); n = 3. (D) HPLC chromatogram profile of the WT and ΔdeaD mutant carrying the empty vector (pCA24N) and that of ΔdeaD mutant with pCA24N-deaD. Strains were grown in LB supplemented with 50 μM IPTG and 15 μg/mL of Cm until an OD₆₀₀ of 1.0. PG analysis was done as described in Materials and Methods. Peak “a” was identified by mass-spectrometric analysis and its structure is depicted. The graph represents the average peak-area% of Tri-Tetra muropeptide species. ***, P < 0.001; ns (not significant); n = 3.

FIG 5

DeaD is required for efficient translation of LdtD. (A) Western blot showing LdtD-His levels in ΔldtD or ΔldtDΔdeaD strains carrying pTrc99a-ldtD–His plasmid. Strains were grown in LB supplemented with IPTG (0, 10, or 100 μM) until an OD₆₀₀ of 1.0, and normalized cell extracts were subjected to Western blotting. (B) Western blot showing posttranslational stability of LdtD-Flag. In rapidly growing WT or ΔdeaD cells at an OD₆₀₀ 0.6, 300 μg/mL spectinomycin was added to block the protein synthesis. Fractions were collected at 0, 30, 60, and 120 min post-antibiotic treatment, and normalized cell extracts were analyzed through Western blotting, as described in Materials and Methods. FtsZ was used as a loading control. *, P < 0.05; **, P < 0.005; ***, P < 0.001; ns (not significant); n = 3.

FIG 6

Effect of cold shock on LdtD levels and composition of PG. (A) Western blot indicating LdtD-Flag and DeaD-HA levels of WT during cold shock. WT strain carrying both LdtD-Flag and DeaD-HA constructs was subjected to cold shock (as described in Materials and Methods) and fractions were collected at indicated intervals until an OD₆₀₀ of 1.0 and normalized cell extracts were analyzed (lanes1 to 5). Lane 6 has cell extract of the WT continuously incubated at 37°C. The relative fold change of LdtD-Flag levels between lanes 5 and 6 is shown in the bar diagram. DeaD-HA served as a positive control, whereas FtsZ was used as a loading control. NlpI levels remained unaltered during cold shock. (B) Western blot indicating LdtD-Flag levels in a deaD deletion mutant. ΔdeaD mutant carrying LdtD-Flag subjected to cold shock was analyzed as described above (lanes 1 to 5). Lanes 6 and 7 had extracts of the WT and ΔdeaD grown continuously at 37°C. The values obtained from lanes 5 and 7 were used for comparison. (C) HPLC chromatograms of WT and ΔdeaD strains subjected to cold shock. Cells were grown, PG sacculi were isolated, and digested with mutanolysin and soluble muropeptides were resolved using HPLC. Peaks a, b, c, and d were analyzed further as described in Materials and Methods. Peaks a, c, and d showed a significant increase during cold shock, whereas peak b is unaltered. Bar diagrams represent the average peak area percentage of respective muropeptide species in the chromatogram. The molecular structures of the analyzed peaks are shown. **, P < 0.005; ***, P < 0.001; ns (not significant); n = 3.