Mutant gltS alleles enable a Vibrio fischeri D-glutamate auxotroph to grow with lower requirements for exogenous D-glutamate

Abstract

D-glu is a key component of peptidoglycan (PG) and is essential for growth in most bacteria. To assess constraints on PG evolution and bacterial requirements for D-glu, we sought to artificially evolve PG biosynthesis, leading to either replacement of D-glu in the PG peptide or alternative pathways to D-glu incorporation. We previously found that suppression of D-glu auxotrophy in a murI racD mutant of Vibrio fischeri grown on lysogeny broth salts (LBS) medium was rare but could be accomplished by mutation of bsrF, with restoration of wild-type PG structure. Here, we selected nine additional prototrophic suppressors of the same murI racD mutant from 10¹⁰ colony-forming units (CFU) plated on LBS supplemented with ~2.7 mM D-gln. Each suppressor had a mutation in gltS, which encodes a putative sodium:glutamate symporter. Increased copy numbers of mutant gltS alleles enabled growth on unsupplemented LBS and resulted in PG containing D-glu. Examination of media components suggests that D-gln supplementation had inadvertently added ~14 μM D-glu, and that LBS itself contains ~1.4 μM D-glu. The mutations in gltS enabled growth with similarly low D-glu concentrations, but also increased sensitivity to homocysteic acid, suggesting more promiscuous transport. Surprisingly, we discovered that expression of mutant gltS in the auxotroph leads to incorporation of lysine into PG, in addition to canonical D-glu. When seawater is supplemented with D-glu, this V. fischeri mutant still colonized Euprymna scolopes and triggered PG-induced morphogenesis. Our results shed light on glutamate transport, highlight trade-offs in GltS structure and function, and reveal an unusual PG modification.IMPORTANCED-glu is an important building block in the peptidoglycan (PG) component of the bacterial cell wall, and its endogenous production is considered essential in most bacteria, even when grown in complex media. In Vibrio fischeri, in trans expression of mutant GltS symporters allows D-glu auxotrophic strains to grow on lysogeny broth salts (LBS) medium without exogenous D-glu, although there is a fitness trade-off of increased sensitivity to homocysteic acid. Our finding that LBS contains sufficient D-glu to support robust growth highlights the undervalued importance of D-amino acid transport and the ubiquity of D-amino acids. Moreover, the discovery of D-lysine in the PG peptide is an unusual PG modification that warrants further study.

Keywords: Aliivibrio; D-amino acids; glutamate; peptidoglycan; photobacterium; transport.

Fig 1

Growth of RMJ13 and suppressor mutants with and without D-glu or D-gln supplementation. (A) Final OD₆₀₀ of RMJ13 (murI::mini-Tn5 ∆racD) and its derivatives RMJ13M1 through RMJ13M8 (denoted here as M1 through M8) grown in LBS-Em supplemented with 400 µg/mL D-gln or D-glu. Stocks of D-gln were subsequently found to contain small amounts of D-glu, as described below. (B) D-glu requirements for gltS mutants were assessed by growing strains in LBS-Em containing 0, 1, 5, or 50 µg/mL D-glu supplementation. Cultures were grown for 24 hours before reading OD₆₀₀. Error bars indicate standard error of the mean (n = 4). Data from one representative experiment of at least three are shown.

Fig 2

Mutant gltS alleles lower D-glu requirement of D-glu auxotroph. (A) Final OD₆₀₀ readings for V. fischeri cultures grown for 24 hours in LBS-Em with or without supplementation of 50 µg/mL D-glu. Strains included RMJ13 alone or carrying vector pVSV105, pMNC15 (gltS), pMNC16 (gltS^C1092T), or pMNC17 (gltS^C490T). (B) Final OD₆₀₀ readings for V. fischeri cultures grown for 24 hours in fischeri minimal medium with varying amounts of D-glu as indicated. Error bars indicate standard error of the mean (n = 3). Data from one representative experiment of at least three are shown. * RMJ13 expressing WT GltS (carried by pMNC15) grows to a significantly lower final OD₆₀₀ than all other strains (P < 0.05, Student’s t-test).

Fig 3

Expressing mutant GltS proteins increases sensitivity to HCA. Shown are final OD₆₀₀ readings for V. fischeri cultures grown with increasing concentrations of HCA. Strains include ES114 alone, or carrying pVSV105, pMNC15 (gltS), pMNC16 (gltS^C1092T), or pMNC17 (gltS^C490T). Cultures were grown for 24 hours before reading OD₆₀₀. Error bars indicate standard error of the mean (n = 4). Data from one representative experiment of at least three are shown.

Fig 4

Muropeptide profiles of V. fischeri strains ES114, RMJ13 expressing gltS^C490T (carried on pMNC17), and RMJ13 expressing gltS^C1092T (carried on pMNC16). (A) Representative LC spectra from comparative muropeptide analysis in which the amount of purified and injected PG was normalized to the height of peak C. (B) Identification of peaks labeled in (A).

Fig 5

Diagram of predicted GltS protein structures of E. coli and V. fischeri. (Left) Structure of GltS of E. coli B strain WM335, which has two amino acid substitutions compared to wild-type. Both mutations (shown as colored dots) were found within the same strain. The diagram is based on SWISS-MODEL predictions and previous publications (42, 43, 45–48). (Right) Structure of GltS from V. fischeri, showing amino acid substitutions corresponding to mutations in RMJ13 suppressor strains. All mutations (shown as colored dots) are shown on one structure, although only one mutation exists per strain. Diagram based on SWISS-MODEL predictions (46–48).

Fig 6

Colonization of E. scolopes by V. fischeri ES114 RMJ13-derived strains in the presence and absence of D-glu supplementation. (A) Squid luminescence over time for wild-type (ES114) and RMJ13 carrying pVSV105 (vector) or pMNC17 (gltS^C490T), with and without persistent exogenous D-glu supplementation. “Apo” indicates uninoculated, aposymbiotic squid as negative controls. (B) Symbiont population levels (average CFU per light organ) over time, with the same strains, conditions, and treatments as in panel A. Error bars indicate standard error of the mean (n = 13). Data from one experiment are shown. This experiment was performed once.

Fig 7

D-glu auxotrophic cells lyse when placed in filtered seawater and without D-glu. Cells lyse within 4 hours of being placed in filtered seawater without D-glutamate. Expression of gltS^C490T (top) does not have a growth advantage over RMJ13 carrying the vector alone (bottom) in these conditions.