Secrets of getting started: regulation of the first committed step of peptidoglycan synthesis by protein phosphorylation in Enterococcus and other Gram-positive bacteria

Abstract

Regulation of the first committed step of peptidoglycan precursor synthesis by MurA-enzyme homologs has recently taken center stage in many different bacteria. In different low-GC Gram-positive bacteria, regulation of this step has been shown to be regulated by phosphorylation of homologs of the IreB/ReoM regulatory protein by PASTA-domain Ser/Thr-protein kinases. In this issue, Mascari, Little, and Kristich determine this regulatory pathway and its links to resistance to cephalosporin β-lactam antibiotics in the major human pathogen, Enterococcus faecalis (Efa). Unbiased genetic selections identified MurAA (MurA-family homolog) as the downstream target of IreB regulation in the absence of the IreK Ser/Thr-protein kinase. Physiological and biochemical approaches, including determination of MICs to ceftriaxone, Western blotting of MurAA cellular amounts, isotope incorporation into peptidoglycan sacculi, and thermal-shift binding assays of purified proteins, demonstrated that unphosphorylated IreB, together with proteins MurAB (MurZ-family homolog), and ReoY(Efa) negatively regulate MurAA stability and cellular amount by the ClpCP protease. Importantly, this paper supports the idea that ceftriaxone stimulates phosphorylation of IreB, which leads to increased cellular MurAA amount and precursor pathway flux required for E. faecalis cephalosporin resistance. Overall, findings in this paper significantly contribute to understanding variations of this central regulatory pathway in other low-GC Gram-positive bacteria.

There has been a recent burst of progress in understanding how protein phosphorylation by PASTA-domain Ser/Thr-protein kinases regulate the first committed step of peptidoglycan synthesis in different low-GC Gram-positive bacteria, including major pathogens (Kelliher et al., 2021; Rismondo, Bender, & Halbedel, 2017; Sun, Hürlimann, & Garner, 2023; Tsui et al., 2023; Wamp et al., 2022; Wamp et al., 2020). This first step is catalyzed by homologs of the MurA UDP-N-acetylglucosamine 1-carboxyvinyltransferase that converts PEP and UDP-GlcNAc to Pi and UDP-N-acetyl-3-O-(1-carboxyvinyl)-alpha-D-glucosamine (Brown, Vivas, Walsh, & Kolter, 1995; Du et al., 2000). MurA enzymes are inactivated by covalent binding of the antibiotic fosfomycin, which is a PEP analog (Kahan, Kahan, Cassidy, & Kropp, 1974; Marquardt et al., 1994; Skarzynski et al., 1996). In low-GC Gram-positive bacteria, there are two MurA homologs, often designated as the MurA-family and MurZ-family. Except for S. pneumoniae, the MurA-family enzyme plays the predominant role in peptidoglycan synthesis (see (Tsui et al., 2023)); however, the roles of the second homolog are not fully understood.

A series of concurrent, recent papers have used genetic and biochemical approaches to demonstrate that phosphorylation of homologs of a single protein target, designated as IreB, ReoM, or YrzL in different bacteria and called IreB/ReoM here, regulates PG synthesis through the stability of MurA-family homologs in Listeria monocytogenes (Kelliher et al., 2021; Wamp et al., 2022; Wamp et al., 2020), Bacillus subtilis (Sun et al., 2023; Wamp et al., 2020), and Staphylococcus aureus (Kelliher et al., 2021), or the enzymatic activities of the MurZ-and MurA-family homologs in Streptococcus pneumoniae (Tsui et al., 2023). Unphosphorylated IreB/ReoM acts as a negative regulator of protein stability or enzymatic activity, whereas phosphorylation of IreB/ReoM by PASTA-containing Ser/Thr protein kinases prevents this negative regulation. Two evolutionarily distinct mechanisms underlie this negative regulation (Fig. 1). In L. monocytogenes, B. subtilis, and S. aureus, unphosphorylated ReoM is thought to interact with the MurA-family homolog, and this interaction directs the MurA-family homolog for degradation by the ClpCP protease (Kelliher et al., 2021; Sun et al., 2023; Wamp et al., 2022; Wamp et al., 2020). In these species, the MurZ-family homolog and another protein called ReoY (YpiB in B. subtilis) are also required for this degradation, possibly by acting as protease adaptors (Sun et al., 2023; Wamp et al., 2022; Wamp et al., 2020). Thus, lack of phosphorylation of ReoM leads to a decrease in cellular MurA amount and a decrease in PG synthesis. By contrast, in S. pneumoniae the MurZ-family homolog is predominant physiologically and enzymatically, and the MurA-family homolog is also enzymatically active (Du et al., 2000; Tsui et al., 2023). Genetic evidence (Tsui et al., 2023) and biochemical assays (Joseph, M, 2023, unpublished data) support a model where interaction with unphosphorylated IreB decreases the enzymatic activity, but not the cellular amounts, of S. pneumoniae MurZ and MurA. Consistent with this different mechanism of negative regulation, a ReoY homolog is absent in S. pneumoniae and deletion of clpP or clpC does not change MurZ or MurA cellular amounts (Tsui et al., 2023). Nevertheless, in both mechanisms, the PASTA-domain Ser/Thr-protein kinases and their cognate Ser/Thr-protein phosphatases alter the flux through the PG precursor synthesis pathway by changing the phosphorylation level of the IreB/ReoM protein (Kelliher et al., 2021; Sun et al., 2023; Wamp et al., 2022; Wamp et al., 2020).

Figure 1.

Regulatory pathway that leads to destabilization of MurAA by binding to unphosphorylated IreB in E. faecalis proposed in (Mascari et al., 2023). Cell wall stresses signal the IreK Ser/Thr-protein kinase to transfer a phosphoryl group from ATP to the IreB regulatory protein. Phosphorylated IreB~P does not bind to MurAB or MurAA, whereas unphosphorylated IreB dimer binds to both MurAB and MurAA, possibly sequentially. The MurAB-IreB-MurAA complex together with ReoY(Efa) acts as an adaptor for MurAA degradation by the ClpCP protease. Decreased cellular MurAA amount lessens PG precursor synthesis, cephalosporin resistance, and growth during other stress conditions. An analogous regulatory pathway has been described in L. monocytogenes (Kelliher et al., 2021; Wamp et al., 2022; Wamp et al., 2020), B. subtilis (Sun et al., 2023), and S. aureus (Kelliher et al., 2021). In S. pneumoniae (Spn; middle of figure), binding of unphosphorylated IreB to MurZ and MurA inhibits their enzymatic activity directly without leading to MurZ or MurA degradation by ClpCP (Tsui et al., 2023). See text for additional details.

A new paper in this issue by Mascari, Little, and Kristich provides important new insights into this mechanism and its control of resistance to cephalosporin family β-lactam antibiotics in Enterococcus faecalis (Mascari, Little, & Kristich, 2023). This paper builds on earlier work from the Kristich laboratory that linked absence of the E. faecalis IreK Ser/Thr-protein kinase or IreB regulatory protein to decreased or increased cephalosporin resistance, respectively (Hall et al., 2017; Kristich, Wells, & Dunny, 2007). Consistent with this relationship, deletion of the cognate IreP Ser/Thr-protein phosphatase also increased cephalosporin resistance (Kristich, Little, Hall, & Hoff, 2011). Other interesting links were the findings that IreK Ser/Thr-protein kinase activity increased in exponentially growing E. faecalis cells stressed by exposure to a cephalosporin and that IreK Ser/Thr-protein kinase activity was inactivated in stationary phase (Labbe & Kristich, 2017). It was also shown that MurAA (the MurA-family homolog in E. faecalis) catalytic activity, but not MurAB (the MurZ-family homolog), was required for intrinsic cephalosporin resistance (Vesic & Kristich, 2012). Together, these results suggested that unphosphorylated IreB was a negative regulator of a downstream factor that contributed to cephalosporin resistance (Hall, Tschannen, Worthey, & Kristich, 2013; Kristich et al., 2007). These findings also provided a framework for interpretation of IreB/ReoM function in the other bacteria mentioned above. This new paper by Mascari, Little, and Kristich demonstrates that this downstream factor is MurAA, which is destabilized by unphosphorylated IreB (Fig. 1) (Mascari et al., 2023).

MurAA was implicated as a target of IreB regulation by two independent forward genetic selections (Mascari et al., 2023). One selection was for increased cephalosporin (in this case, ceftriaxone) resistance in a mutant containing a catalytically inactive IreK Ser/Thr-protein kinase. Although this selection hypothetically could turn up components in many kinase substrates, only amino acid changes at one position in MurAA (A115T and A115V) were recovered. The other selection relied on the fact that overexpression of IreB is lethal in a ΔireK Ser/Thr-protein kinase mutant. This selection was highly informative and turned up several suppressor mutations with amino acid changes in a region of MurAA away from the catalytic site (e.g., N188K). In addition, a mutation (clpP(A46E)) that was predicted to inactivate the ClpP protease was recovered. The murAA(N188K) and clpP(A46E) mutations from the second selection increased cephalosporin resistance in the wild-type ireK⁺ or ΔireK genetic backgrounds. Importantly, PG substrate flux, determined by rate of [¹⁴C]-GlcNAc incorporation into purified PG sacculi, decreased in the absence of the IreK Ser/Thr-protein kinase compared to wild-type in E. faecalis (Mascari et al., 2023), which likely lacks PG turnover like S. pneumoniae (Boersma et al., 2015). This decrease in PG substrate flux was counteracted in ΔireK murAA(A115T) and ΔireK murAA(N188K) mutants. Together, these results support that MurAA function is a downstream target of protein phosphorylation by the IreK Ser/Thr-protein kinase, and they reinforce that MurAA function is positively correlated with cephalosporin resistance (Vesic & Kristich, 2012).

The involvement of the IreK Ser/Thr-protein kinase, IreP Ser/Thr-protein phosphatase, IreB regulatory protein, and ClpP protease from these new and past results (Kelliher et al., 2021; Sun et al., 2023; Wamp et al., 2022; Wamp et al., 2020) suggested the hypothesis that MurAA amount was negatively regulated by unphosphorylated IreB via the ClpCP protease pathway. Indeed, western blotting showed increased relative cellular amounts of MurAA in ΔireB, ΔireP, ΔclpP, and ΔclpC mutants in exponentially growing cells. Interesting, cellular MurAA amount is strongly dependent on growth phase and was shown to markedly decrease in stationary phase (Mascari et al., 2023). MurAA amount also increased in ΔireB, ΔireP, ΔclpP, and ΔclpC mutants in stationary phase. Likewise, the newly isolated murAA(N188K), murAA(A115T), and clpP(A46E) mutants showed increased relative amounts of MurAA. Somewhat surprising, the relative cellular amount of MurAA was unchanged in the ΔireK Ser/Thr-protein kinase mutant, perhaps reflecting a complication caused by lack of phosphorylation of other IreK kinase targets (Mascari et al., 2023).

Notably, relative cellular MurAA amount also increased in cells treated with a sublethal concentration of ceftriaxone (Mascari et al., 2023), which activated protein phosphorylation by the IreK Ser/Thr-protein kinase (Labbe & Kristich, 2017; Minton, Djorić, Little, & Kristich, 2022). The ceftriaxone-induced increase in MurAA amount was blocked by ΔireK or ΔireB mutations. Furthermore, protein stability determinations after addition of the translation inhibitor chloramphenicol showed that ΔclpP or ΔireB stabilized MurAA. Last, absence of MurAB or the ReoY homolog (OG1RF_11272) increased the relative amount of MurAA in wild-type E. faecalis and increased cephalosporin resistance of a ΔireK mutant (Mascari et al., 2023). By contrast, the relative cellular amount of MurAB was not proteolytically regulated under any condition tested. However, MurAB catalytic activity was not required for its role in facilitating degradation of MurAA by ClpCP (Mascari et al., 2023). Thus, Mascari, Little, and Kristich establish that unphosphorylated IreB, MurAB, and ReoY(Efa) negatively regulate cellular MurAA amount through ClpCP-mediated proteolysis. This mechanism is analogous to the one reported recently in L. monocytogenes (Kelliher et al., 2021; Wamp et al., 2022; Wamp et al., 2020), B. subtilis (Sun et al., 2023; Wamp et al., 2020), and S. aureus (Kelliher et al., 2021), but different from that of S. pneumoniae (Tsui et al., 2023), which is an ovoid-shaped bacterium, like E. faecalis (Mascari et al., 2023).

Mascari, Little, and Kristich go on to provide other new results about the mechanism of phosphorylation-mediated regulation of MurAA stability in E. faecalis and other Gram-positive bacteria. Thermal-denaturation profiling of purified proteins demonstrated that unphosphorylated IreB directly interacted with MurAA only when bound to both of its substrates, UDP-GlcNAc and PEP (or its analog fosfomycin) (Mascari et al., 2023). This striking finding suggests that MurAA substrate amounts could indirectly influence MurAA stability and amount. In contrast, they report for the first time that IreB also binds directly to MurAB, but in the absence of substrates. The amino acid change in E. faecalis MurAA(N188K) is in a domain that was identified in L. monocytogenes MurA (Wamp et al., 2022) and S. pneumoniae MurZ (Tsui et al., 2023) as being required for interaction with unphosphorylated IreB. In agreement with this idea, purified MurAA(N188K) did not bind to unphosphorylated IreB. Conversely, biochemically phosphorylated IreB (IreB~P) did not bind to MurAA. Given that unphosphorylated IreB is a dimer (Hall et al., 2017), these binding experiments suggest that one subunit of the IreB₂ dimer may bind to MurAA, while the other binds to MurAB. In E. faecalis, this complex would target MurAA for degradation by ClpCP (Mascari et al., 2023), whereas in S. pneumoniae, this complex would simultaneously inhibit the enzymatic activities of MurZ and MurA (Tsui et al., 2023).

Throughout these studies, interesting correlations emerged between MurAA amount and response to the cephalosporin, ceftriaxone. Treatment with ceftriaxone increased cellular MurAA amount in E. faecalis, and the greater MurAA amount in mutants defective in its proteolytic degradation increased ceftriaxone resistance (Mascari et al., 2023). However, although there was a correlation, there was not a strict linear relationship between cellular MurAA amount and ceftriaxone resistance. In addition, ΔireK mutants lacking the Ser/Thr-protein kinase showed growth defects to cell stresses, such as the presence of bile, higher temperature, and hydrogen peroxide addition (Mascari et al., 2023). Growth of ΔireK mutants was partially or fully restored by ΔireB, ΔclpC, or murAA(N188K) mutations, which increased cellular MurAA amount. These results indicated that increased amounts of MurAA helped the ΔireK mutant deal with these stresses. However, increased MurAA amount did not lead to greater stress resistance in wild-type cells and was sometimes detrimental, affirming the importance maintaining normal MurAA amounts. Consistent with this idea, previous studies showed that overexpression of MurA in L. monocytogenes in high salt (Wamp et al., 2022; Wamp et al., 2020) or MurZ in S. pneumoniae (Tsui et al., 2023) inhibited growth and resulted in aberrant cell morphologies. The mechanisms underlying the link between increased MurAA amount that increases PG precursor flux and cephalosporin resistance in E. faecalis reported in this paper (Mascari et al., 2023) and previously in L. monocytogenes (Wamp et al., 2022; Wamp et al., 2020) remains to be fully determined. It likely reflects the catalytic requirements and expression levels of additional Class B penicillin-binding proteins (PBPs) with low cephalosporin reactivity that are present in E. faecalis (Djoric, Little, & Kristich, 2020) and L. monocytogenes (Wamp et al., 2022), but absent in other bacteria, such as S. pneumoniae (Lamanna et al., 2022).

Importantly, this paper raises many fundamental questions for future studies about the physiology of Ser/Thr-protein kinase phosphorylation of IreB/ReoM to regulate MurA enzyme amount or activity in different Gram-positive bacteria. One question concerns the setpoint of IreB phosphorylation in exponentially growing and stationary cells in different media. In E. faecalis, only about 20% of IreB is phosphorylated in exponentially growing cells, and this percentage drops to nearly zero in stationary phase cells (Labbe & Kristich, 2017). One implication of this relatively low level of IreB phosphorylation in unstressed, growing cells is that there is considerable capacity to increase the amount of MurAA through increased activity of the IreK Ser/Thr-protein kinase, which occurs in response to cephalosporins. On the other hand, there may be differences in this IreB/ReoM phosphorylation setpoint in other bacteria. L. monocytogenes ΔreoM and S. pneumoniae ΔireB mutations fully suppress the requirement for their essential Ser/Thr-protein kinases in the strains tested during exponential growth (Tsui et al., 2023; Wamp et al., 2022). These results indicate that maintaining protein phosphorylation of ReoM or IreB to maintain MurA stability or MurZ enzymatic activity, respectively, is a primary function of Ser/Thr-protein phosphorylation in unstressed cells of these Gram-positive bacteria. How PG and Lipid II sensing feed into setting the activities of PASTA-domain Ser/Thr-protein kinases in different exponentially growing and stressed Gram-positive bacteria is an important, unanswered question.

Several other questions are raised by this paper. The relative increase in MurAA amount in mutants defective in the proteolytic regulation is modest in E. faecalis (or in B. subtilis (Sun et al., 2023)) compared to large increases of MurA reported for L. monocytogenes (Rismondo et al., 2017; Wamp et al., 2022; Wamp et al., 2020). It is possible that the proteolytic regulation of MurAA amount is just more limited in E. faecalis and B. subtilis. Alternatively, binding of unphosphorylated IreB may also inhibit the enzymatic activities of E. faecalis MurAA and MurAB directly, which is the sole mechanism found in S. pneumoniae (Tsui et al., 2023) (Joseph M, 2023, unpublished data). In this regard, Enterococcus MurAA and MurAB are evolutionarily closely related to pneumococcal MurA and MurZ, respectively (Tsui et al., 2023). This paper also contributes new information about the long-standing and largely unanswered question about why there are two MurA homologs in low-GC Gram-positive bacteria (Du et al., 2000; Mascari, Djorić, Little, & Kristich, 2022; Tsui et al., 2023; J. Zhou et al., 2022). The paper shows that MurAB binds to unphosphorylated IreB and that even catalytically inactive MurAB can participate with ReoY(Efa) in delivering MurAA to the ClpCP protease for degradation. The mechanism of this delivery remains unknown, but MurAB binding may precede MurAA binding to IreB and thereby provide a failsafe mechanism against cells having an insufficiency of both MurAB and MurAA for PG precursor synthesis (Mascari et al., 2023). Other questions raised by some of the phenotypes in this paper are whether IreB binds to and regulates other proteins and whether there is regulation of the cognate IreP Ser/Thr-protein phosphatase.

Finally, a genetic selection in this paper for increased ceftriaxone resistance in a catalytic mutant of the IreK Ser/Thr-protein kinase recovered the murAA(A115T) mutation (Mascari et al., 2023). The A115T amino acid change is near the active site away from the domain of MurAA involved in IreB binding. The A115T amino acid change increases the cellular amount of MurAA(A115T) compared to wild-type, but does not protect MurAA(A115T) from proteolysis. One possibility is that the A115T affects feedback inhibition of MurAA(A115T) by UDP-MurNAc, which is the produced by MurB in the next step of the pathway (Mizyed, Oddone, Byczynski, Hughes, & Berti, 2005; Y. Zhou et al., 2023). Together, this paper by Mascari, Little, and Kristich (Mascari et al., 2023), along with recent parallel studies in other Gram-positive bacteria (Kelliher et al., 2021; Rismondo et al., 2017; Sun et al., 2023; Tsui et al., 2023; Wamp et al., 2022; Wamp et al., 2020) and by different mechanisms in Gram-negative bacteria (Hummels et al., 2023) and Mycobacterium tuberculosis (Boutte et al., 2016) affirm the central role played by metabolic pathway regulation in PG synthesis and its integration with the synthesis of other macromolecules.