Regulation of Peptidoglycan Synthesis in Escherichia coli – RegOPepS

Peptidoglycan (PG) is the major constituent of the bacterial wall and the target of a large number of antibiotics, particularly the ß-lactams. These antibiotics inhibit the last cross-linking step of PG polymerization by irreversibly binding to the penicillin-binding proteins (PBPs). Until recently, PBPs were the only enzymes known to catalyze PG cross-linking. However, my team has shown that this reaction is also catalyzed by a second family of enzymes, the L,D-transpeptidases (LDTs). PBPs are potentially inhibited by all classes of ß-lactams whereas LDTs are only inhibited by carbapenems. Formation of 3?3 cross-links by LDTs is predominant in mycobacteria including Mycobacterium tuberculosis. In other species, LDTs are unessential and their contribution to PG cross-linking during exponential growth is negligible. However, the proportion of 3?3 cross-links increases during the stationary phase suggesting a role of LDTs in PG maintenance in the absence of de novo synthesis of PG precursors. In Escherichia coli, I have recently shown in in vitro-selected mutants that LDTs mediate high-level resistance to most ß-lactams by fully replacing the PBPs. The bypass mechanism requires increased production of the guanosine pentaphosphate and tetraphosphate [(p)ppGpp] alarmone, which was originally identified by its capacity to trigger the stringent response. These results have shown that E. coli has the capacity to polymerize PG by two alternate cross-linking pathways whose relative contribution is controlled by the (p)ppGpp alarmone. The general objectives of the project are the identification of the enzymes involved in the LDT pathway and of the role of the (p)ppGpp alarmone in PG synthesis. Our first research hypothesis is that PG polymerization mediated by PBPs and LDTs involves different sets of proteins. By constructing strains that conditionally rely on PBPs or LDTs for PG synthesis, we will investigate three types of enzymes. (i) Since most models of PG polymerization predict that cleavage of PG cross-links is required for insertion of new subunits into the PG polymer we will identify the endopeptidases responsible for cleavage of 3?3 crosslinks. This will be based on inactivation of candidate genes and a Tn-Seq genome-wide approach. (ii) Some PBPs (Class B enzymes) contain both a transpeptidase domain and a so-called morphogenesis domain, which is poorly characterized at the functional level. We will investigate if these morphogenesis domains cooperate with the LDTs for PG synthesis. (iii) PG polymerization requires transpeptidases to form the cross-links and glycosyltransferases for elongation of glycan strands. We seek to identify the glycosyltransferases that cooperate with the LDTs for PG polymerization. Our second research hypothesis is that the (p)ppGpp alarmone is a key mediator controlling the adaptation of PG metabolism to starvation and other stresses. This may involve both a negative control of the polymerization pathway involving classical PBPs and an activation of the LDT pathway to maintain the integrity of the cell wall. We will identify the key targets of (p)ppGpp gene regulation that are directly involved in PG synthesis. We will also determine the mechanism of regulation of these genes. The scientific impact of the project concerns the regulation of PG synthesis, a topic which has been neglected. For the first time, the adaptation of PG metabolism to starvation will be globally assessed with a focus on increased synthesis of 3?3 cross-links by L,D-transpeptidases and on the factors that are essential for their function. The existence of two alternative cross-linking pathways and the possibility to control their relative contributions to PG polymerization offer a unique opportunity to look in depth into the redundancy of protein functions involved in PG synthesis. The foreseen outcomes include the identification of new targets for drug development.