Experimental evidence of d-glutamate racemase activity in the uncultivated bacterium Candidatus Saccharimonas aalborgensis.

The Candidate Phyla Radiation (CPR) encompasses widespread uncultivated bacteria with reduced genomes and limited metabolic capacities. Most CPR bacteria lack the minimal set of enzymes required for peptidoglycan (PG) synthesis, leaving it unclear how these bacteria produce this essential envelope component. In this study, we analysed the distribution of d-amino acid racemases that produce the universal PG components d-glutamate (d-Glu) or d-alanine (d-Ala). We also examined moonlighting enzymes that synthesize d-Glu or d-Ala. Unlike other phyla in the domain Bacteria, CPR bacteria do not exhibit these moonlighting activities and have, at most, one gene encoding either a Glu or Ala racemase. One of these 'orphan' racemases is a predicted Glu racemase (MurICPR) from the CPR bacterium Candidatus Saccharimonas aalborgenesis. The expression of MurICPR restores the growth of a Salmonella d-Glu auxotroph lacking its endogenous racemase and results in the substitution of l-Ala by serine as the first residue in a fraction of the PG stem peptides. In vitro, MurICPR exclusively racemizes Glu as a substrate. Therefore, Ca. Saccharimonas aalborgensis may couple Glu racemization to serine and d-Glu incorporation into the stem peptide. Our findings provide the first insights into the synthesis of PG by an uncultivated environmental bacterium and illustrate how to experimentally test enzymatic activities from CPR bacteria related to PG metabolism.

Evolutionary analysis and structure modelling of the Rcs-repressor IgaA unveil a functional role of two cytoplasmic small ß-barrel (SBB) domains.

The Rcs sensor system, comprising the RcsB/RcsC/RcsD and RcsF proteins, is used by bacteria of the order Enterobacterales to withstand envelope damage. In non-stress conditions, Rcs is repressed by IgaA, a membrane protein with three cytoplasmic regions (cyt-1, cyt-2 and cyt-3). How the Rcs-IgaA axis evolved within Enterobacterales has not been yet explored. Here, we report phylogenetic data supporting co-evolution of IgaA with RcsC/RcsD. Functional exchange assays showed that IgaA from Shigella and Dickeya, but not from Yersinia or the endosymbionts Photorhabdus and Sodalis, repress the Rcs system of Salmonella. IgaA from Dickeya, however, repress only partially the Rcs system despite being produced at high levels in the complementation assay. The modelled structures of these IgaA variants uncovered one periplasmic and two cytoplasmic conserved ß-rich architectures forming partially closed small ß-barrel (SBB) domains. Conserved residues map in a connector linking cytoplasmic SSB-1 and SBB-2 domains (E180-R265); a region of cyt-1 facing cyt-2 (R188-E194-D309 and T191-H326); and between cyt-2 and cyt-3 (H293-E328-R686). These structures validated early in vivo studies in Salmonella that assigned a role in function to R188, T191 and G262, and in addition revealed a previously unnoticed "hybrid" SBB-2 domain to which cyt-1 and cyt-2 contribute. IgaA variants not functional or partially functional in Salmonella lack H192-P249 and R255-D313 interactions. Among these variants, only IgaA from Dickeya conserves the helix α6 in SSB-1 that is present in IgaA from Salmonella and Shigella. RcsF and RcsD, which interact directly with IgaA, failed to show structural features linked to specific IgaA variants. Altogether, our data provide new insights into IgaA by mapping residues selected differently during evolution and involved in function. Our data also infer contrasting lifestyles of Enterobacterales bacteria as source of variability in the IgaA-RcsD/IgaA-RcsF interactions.