Escherichia coli utilizes multiple peptidoglycan recycling permeases with distinct strategies of recycling

Abstract

Bacteria produce a structural layer of peptidoglycan (PG) that enforces cell shape, resists turgor pressure, and protects the cell. As bacteria grow and divide, the existing layer of PG is remodeled and PG fragments are released. Enterics such as Escherichia coli go to great lengths to internalize and reutilize PG fragments. E. coli is estimated to break down one-third of its cell wall, yet only loses ~0 to 5% of meso-diaminopimelic acid, a PG-specific amino acid, per generation. Two transporters were identified early on to possibly be the primary permease that facilitates PG fragment recycling, i) AmpG and ii) the Opp ATP binding cassette transporter in conjunction with a PG-specific periplasmic binding protein, MppA. The contribution of each transporter to PG recycling has been debated. Here, we have found that AmpG and MppA/Opp are differentially regulated by carbon source and growth phase. In addition, MppA/Opp is uniquely capable of high-affinity scavenging of muropeptides from growth media, demonstrating that AmpG and MppA/Opp allow for different strategies of recycling PG fragments. Altogether, this work clarifies environmental contexts under which E. coli utilizes distinct permeases for PG recycling and explores how scavenging by MppA/Opp could be beneficial in mixed communities.

Keywords: AmpG; cell wall; muropeptides; peptidoglycan; peptidoglycan recycling.

Fig. 1.

Recycling of muropeptides in E. coli. Schematic of muropeptide recycling pathways found in E. coli. Muropeptide permeases, the focus of this work, internalize PG turnover products, so they can be either recycled or catabolized. PTS indicates transporters coupled to a phosphotransferase system. In the cytoplasm, muropeptides can be broken down to a tripeptide of m-Dap-D-Glu-L-Ala and the individual sugars anhMurNAc and GlcNAc. These products can either directly feed back into biosynthesis of PG precursors via recycling or be further catabolized.

Fig. 2.

Loss of MppA/OppBCDF suppresses the toxic recycling ldcA mutant in LB. (A) Model of toxic recycling when LdcA is disrupted. Toxic recycling depletes UDP-MurNAc and other substrates into tetrapeptide lipid II precursors that are unable to act as donors for PBP-catalyzed crosslinks. (B) Growth curves showing the impact of toxic peptidoglycan recycling when LdcA is absent. Performed in biological triplicate. (C) Suppressors of W3110 ΔldcA stationary phase lysis have disrupted the encoding gene for the MppA periplasmic binding protein. (D) Triplicate growth curves testing whether the loss of each transporter suppresses growth defects and/or stationary phase lysis of W3110 ΔldcA mutants. Error bars on growth curves show the SD. When absent, the error bar was smaller than the symbol at that time point.

Fig. 3.

AmpG and OppBCDF/MppA expression is impacted by carbon source and growth phase. (A) Weighted fold changes from triplicate RNA-seq in various media or during different growth stages (stat-stationary, log-logarithmic). (B and C) Triplicate growth curves to test the ability of ΔampG and ΔmppA to suppress the toxic recycling ΔldcA mutant when grown with glucose (B) and tryptone (C) as the carbon source. Error bars on growth curves show the SD. When absent, the error bar was smaller than the symbol at that time point.

Fig. 4.

AmpG recycles turned over products, whereas OppBCDF/MppA can also scavenge muropeptides. (A–F) Triplicate PG recycling assays utilizing the retention of tritiated ³H-Dap label in cellular peptidoglycan per generation. (A) Calculated loss of ³H-Dap per doubling. Sole expression of AmpG (B and C) or OppBCDF (D and E) to assess the ability of each transporter to recycle during growth on glucose (B and D) and tryptone (C and E). (F) Triplicate scavenging assays of ³H-muropeptides released from a strain unable to recycle. Error bars on assays show the SD. When absent, the error bar was smaller than the symbol at that time point.

Fig. 5.

High concentrations of oligopeptides inhibit MppA/OppBCDF PG recycling. (A) Triplicate PG recycling assays in LB with OppBCDF overexpression in the presence of OppA and DppA, absence of OppA, or absence of both. (B) OppBCDF triplicate recycling assays in EZgluc with the addition of Pro-Phe-Lys (PFK) tripeptide after 1 h timepoint. MppA was previously shown to weakly bind Pro-Phe-Lys with a K_D > 300 µM compared to binding of PG peptides at a K_D ~ 250 nM (21). Error bars on assays show the SD. When absent, the error bar was smaller than the symbol at that time point.

Fig. 6.

MppA/OppBCDF does not transport AnhMurNAc muropeptides. Triplicate muropeptide analysis of PG products accumulated in the cytoplasm. Loss of AmpD results in cytoplasmic accumulation of anhMurNAc-linked muropeptides containing tri- (A), tetra- (B), and penta- (C) peptides. Overexpression of each transporter in the ΔampD strain can assess whether each transporter is capable of importing anhMurNAc-linked muropeptides. Note that tetramuropeptides are low because LdcA is still present and trims them down to trimuropeptides. Error bars on assays show the SD. When absent, the error bar was smaller than the symbols.

Fig. 7.

Conservation of PG recycling genes in human gut microbiota. Phylogenetic tree of 60 representative species identified as gut microbes in the human microbiome project, showing distribution of homologs for PG sugar recycling genes (Gram-negative genes anmK and/or nagK, Gram-positive genes amiE and/or murQ), PG peptide recycling genes (mpl and/or ampD), and the four transporters involved in recycling (murPnagEampG, and mppA). Due to high homology between mppA and oppA, mppA homologs are split into two groups: i) confirmed/likely (filled box) if experimental evidence, striking similarity to MppA, or regulation by murein regulator, MurR, was present and ii) possible (outlined box) if the organism contained multiple oppA paralogs. See Dataset S1 for full details of homologs identified.

Fig. 8.

Peptidoglycan recycling could be beneficial in mixed communities. (A and B) Average and SD of the calculated percent loss of ³H-Dap per generation in mixed cultures of a strain able to recycle and a mutant lacking all three recycling permeases when grown in EZ-gluc (A) and LB (B) from triplicate assays. Superimposed expected bars indicate the expected averaged value taking into account the ratio between strains if recycling was performed independently by each strain. (C) Triplicate NOD1 activation assays with supernatants from mixed cultures of a strain able to recycle and a mutant lacking all three recycling permeases. The experimentally determined composition of mixed cultures assays is indicated as a ratio of % W3110:% Δpermeases, see SI Appendix, Fig. S6A. Error bars on assays show the SD.