INTRODUCTION
Staphylococcus aureus can establish infections in humans in a wide range of metabolic niches due to several signal transduction pathways, as well as virulence genes encoded in its genome. Significant advances have been made in the study of bacterial virulence factors and their functions in human disease. However, we have only just begun to understand the metabolic pathways required for bacterial proliferation in the host and their contribution to antibiotic resistance.
The
blaZ-encoded β-lactamase hydrolyses the β-lactam ring in penicillin, conferring penicillin resistance in
S. aureus (
1). Methicillin resistance is mediated by alternative penicillin-binding protein 2a (PBP2a), encoded by
mecA (
2) located on a mobile genetic element, the staphylococcal chromosome cassette (SCC
mec) (
3,
4). PBP2a has low affinity for β-lactam antibiotics and thus is able to cross-link peptidoglycan (PG) strands even in the presence of β-lactam antibiotics and in this manner confers resistance (
2). Methicillin-resistant
S. aureus (MRSA) strains are resistant to methicillin, as well as to all the other β-lactam antibiotics (
1,
2), consequently making infections with MRSA difficult to treat.
β-Lactam resistance in
S. aureus is typically expressed heterogeneously within a given population (
5). The majority of cells within a heterogeneous population exhibit susceptible or borderline susceptible resistance to β-lactams. A subpopulation of approximately 0.1% can survive antibiotic treatment and, upon reexposure to the antibiotic, a homogeneously resistant population emerges (
5). The mechanisms underpinning this switch from heterogeneous resistance (HeR) to homogenous resistance (HoR) are associated with accessory mutations outside
mecA (
6). High-level β-lactam resistance is accompanied by significant energy demands that impose a fitness cost on the cell (
7,
8).
The activity of the tricarboxylic acid (TCA) cycle is an important source for the generation of NADH, and therefore membrane potential, during aerobic respiration. In addition, the TCA cycle generates metabolic intermediates that are then used in various other pathways in the cell. The genes encoding enzymes for the TCA cycle are repressed when preferred nutrients, such as glucose, remain available in the surrounding medium (
9). Once post-exponential growth is reached and glucose is depleted from the medium, TCA cycle gene expression is derepressed in
S. aureus (
10–12). The activity of the TCA cycle has previously been linked to β-lactam resistance in
S. epidermidis, where a dysfunctional TCA cycle is common among clinical isolates and is associated with alterations in the cell envelope and increased tolerance to β-lactams (
13). In
S. aureus, TCA cycle activity regulates ATP levels, which controls tolerance to several antibiotics, including β-lactams (
12). Increased TCA cycle activity to fuel cell wall biosynthesis has also been shown to accompany mutations that enable the transition from the HeR to HoR phenotypes (
14,
15). Furthermore, disruption of the TCA cycle via mutations in
acn and
citZ has been reported to block the production of HoR mutants (
14,
15).
Post-translational modification (PTM) of proteins is one of the most effective mechanisms in diversifying protein function and regulation (
16). PTMs can change the charge and structure of a protein, thus affecting activity, as well as the ability to interact with other proteins/binding partners (
17,
18). Lysine is a basic residue that is critical for protein structure and function (
19). The side chain of lysine in particular can be modified by a variety of PTMs, including phosphorylation (
20), succinylation (
21–23), ubiquitination (
24), methylation (
25), acetylation (
16,
26), and lipoylation (
27). Relatively little is known about PTMs in
S. aureus metabolism and antibiotic resistance, and advances in our understanding of these systems will generate new insights into fundamental cellular processes and virulence mechanisms and potentially identify new therapeutic targets.
In the present study, we report that mutations in the succinyl coenzyme A (succinyl-CoA) synthetase genes, sucC and sucD, lead to increased susceptibility to β-lactam antibiotics in MRSA strain JE2. The impact of these mutations on growth was measured and compared to other TCA cycle mutants. The relative intracellular concentrations of metabolites from the pyruvate node of glycolysis and the TCA cycle were measured, and PG architecture and autolytic activity compared in the sucC mutant and wild-type JE2. We describe the first profile of the lysine succinylome in MRSA and the impact of the sucC mutation on the global succinylome. Our data reveal that increased accumulation of succinyl-CoA from the TCA cycle increases susceptibility to β-lactam antibiotics and reduces autolytic activity via perturbation of global lysine succinylation in MRSA.
DISCUSSION
The TCA cycle is centrally involved in the production of biosynthetic precursors, reducing potential and energy. Here, we report that mutations in sucC and sucD genes encoding the α and β subunits of succinyl-CoA synthetase, which catalyzes the conversion of succinyl-CoA to succinate, significantly increased susceptibility to β-lactams. The sucC and sucD mutants grew as smaller, less-pigmented colonies on MHA and exhibited impaired growth in MHB. Genetically blocking the production of succinyl-CoA in the sucC mutant by mutating sucA or sucB reversed the growth and β-lactam susceptibility phenotypes. In contrast, mutation of sdhA in the sucC mutant had no phenotypic impact. Succinyl-CoA levels were significantly increased in the sucC mutant and were restored to wild-type levels by psucCD complementation or mutation of sucA.
The accumulation of succinyl-CoA in the
sucC mutant perturbed global protein succinylation, which is an important PTM previously described in several pathogens (
21,
22,
43), including
S. epidermidis (
23). Although several PBPs, including
mecA-encoded PBP2a, were among the proteins with the highest number of succinyl-lysines, PG architecture and cross-linking were unchanged in the
sucC mutant, even under oxacillin stress when
mecA expression is increased. The absence of structural changes in PG also indicates that the accumulation of succinyl-CoA does not impact β-lactam resistance via altered biosynthesis of lysine, which is an important component of the cell wall (
53). Succinyl-CoA is used in the biosynthesis of lysine from aspartate and in
Corynebacterium glutamicum disruption of the
sucCD locus, and the resulting accumulation of succinyl-CoA was accompanied by overproduction of lysine (
54). It is difficult to envisage how increased lysine accumulation would reduce β-lactam resistance, and in any event the unchanged PG structure in the
sucC mutant does not implicate lysine biosynthesis in this phenotype. Succinyl-CoA synthetase activity generates GTP, which is a substrate for RelA, RelP, and RelQ enzymes that produce the stringent response alarmone (p)ppGpp. (p)ppGpp plays a central role in the control of MRSA β-lactam resistance (
31–34). However, the ability of the
sucC mutant to produce stable HoRs with mutations in
relA or
relQ suggests that intracellular GTP is not limited in the
sucC and
sucD mutants or associated with reduced β-lactam resistance. Metabolomic analysis further revealed significantly reduced levels of acetyl-CoA in the
sucC mutant, raising the additional possibility that the acetylome may also have been perturbed. In this context, changes in PG acetylation have also been implicated in autolysis and antibiotic resistance (
55), and future comparison of the acetylome and succinylome in
sucCD mutants to identify proteins that are modified by both PTMs may provide further insights into how these mutations impact resistance.
The reduced pigmentation of the
sucCD mutants may be a consequence of altered production of the
S. aureus carotenoid staphyloxanthin, composed of a glucose residue esterified with a 30-carbon carboxylic acid chain and a 15-carbon fatty acid (
56). In
S. aureus most fatty acids are odd-numbered branched-chain fatty acids (
57). The β-oxidation of odd-numbered fatty acids generates acetyl-CoA and propionyl-CoA, the latter of which can be converted to succinyl-CoA. Interestingly, staphyloxanthin-derived lipids interact with flotillin to form functional membrane microdomains required for oligomerization and activity of PBP2a (
58). In contrast to the reduced pigmentation, the proteomic analysis revealed increased levels of staphyloxanthin biosynthetic enzymes in the
sucC mutant (MassIVE ID
MSV000086976 and MassIVE ID
MSV000086971), perhaps reflecting efforts by the
sucC mutant to compensate for reduced pigmentation.
The discovery that Atl was the most succinylated protein in the MRSA proteome and that 12 of the 82 succinyl-lysines were significantly more succinylated in the
sucC mutant suggested a potential connection to reduced β-lactam resistance. Fisher and Mobashery recently proposed a model in which the bactericidal activity of β-lactams is the result of deregulated Atl activity at the cell division septum (
59). Our data showing that autolytic activity was significantly reduced in the
sucC mutant and that the oxacillin MIC of the
atl transposon mutant NE460 was unchanged are not consistent with this possibility. Furthermore, previous work in our laboratory linked increased autolytic activity with increased β-lactam resistance. Specifically,
atl transcription was activated in a HoR mutant of USA300 LAC (oxacillin MIC > 256 µg/ml), which exhibited significantly increased autolytic activity, and growth of USA300 LAC in sub-MIC oxacillin was also associated with significantly increased autolysis (
60). Thus, while it seems clear that autolysis and β-lactam susceptibility are interconnected, the precise mechanistic interactions between these two phenotypes needs to be elucidated further.
Processing of the Atl proprotein, produces a signal peptide, a propeptide, a
N-acetylmuramoyl-
l-alanine amidase (AM) enzyme, and a C-terminally located endo-β-
N-acetylglucosaminidase enzyme (GL) (
61). The region between the AM and GL catalytic domains contains three repeat regions (R1 to R3) with GW-dipeptide motifs required to target Atl proprotein to the equatorial ring on the cell surface during cell division (
62). None of these 12 lysine residues exhibiting increased succinylation are in the AM catalytic domain, 5 are in the GL catalytic domain, 2 are in the propeptide region, and the remaining 5 are in the R1 and R2 regions. The single lysine residue (K
200) of Sle1 that is more succinylated in the
sucC mutant is located in a LysM cell wall hydrolase domain and may be important for activity of the enzyme.
Susceptibility to the LtaS inhibitor Congo red was unchanged in the
sucC mutant, indicating that increased β-lactam susceptibility may not be associated with impaired expression or stability of WTA or LTA. Similarly, susceptibility to the alanylation inhibitor DCS was the same in the wild-type and
sucC mutant. Given that
d-alanine is an important component of PG, this observation is consistent with the absence of any changes in PG structure in the
sucC mutant. Furthermore, because
d-alanine is an important component of WTA and LTA, unchanged susceptibility to DCS does not point to roles for these cell envelope glycopolymers in
sucC-dependent β-lactam susceptibility. Nevertheless, given that almost 58% of all quantifiable succinylated peptides were significantly changed in the
sucC mutant, we propose a model in which perturbation of the succinylome likely modulates the activity of multiple enzymes, including Atl and Sle1, that collectively control growth and interconnected cell envelope characteristics such as autolysis and susceptibility to β-lactam antibiotics (
Fig. 8). The U.S. Food and Drug Administration-approved anticancer drug streptozotocin specifically targets succinyl-CoA synthetase in human cells to limit proliferation (
63) and is used primarily to treat tumors that cannot be surgically removed. The findings described here may open the door to the possibility of sensitizing MRSA to β-lactam antibiotics using compounds that specifically target succinyl-CoA synthetase or protein succinylation generally within the cell.
ACKNOWLEDGMENTS
C.C., C.F., M.S.Z., L.A.G., and J.P.O. were supported by grants from the Health Research Board (HRA-POR-2015-1158 and ILP-POR-2019-102) (
www.hrb.ie), the Irish Research Council (GOIPG/2014/763 and GOIPG/2016/36) (
www.research.ie), and Science Foundation Ireland (19/FFP/6441) (
www.sfi.ie). E.B. and F.C. were supported by a grant from Svenska Forskningsrådet Formas. G.C., H.M.O., T.L.F., and J.A. were supported by the IARPA FunGCAT program, NIGMS grant GM103493, and Department of Energy contract DE-AC05-76RLO 1830. K.W.B., P.D.F., V.C.T., D.S., J.N.A., and F.R. were supported by National Institute of Allergy and Infectious Diseases (NIAID) grant P01-AI83211. K.W.B. and J.N.A. were supported by NIAID grant R01-AI125589. D.S. and V.C.T. were supported by grant NIAID R01AI125588.
The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.
Claire Fingleton: conceptualization, formal analysis, investigation, methodology, writing - original draft, writing—review and editing; Chris Campbell: conceptualization, formal analysis, investigation, methodology, writing—original draft, writing—review and editing; Merve S. Zeden: data curation, formal analysis, investigation, methodology, writing—original draft, writing—review and editing; Emilio Bueno: formal analysis, investigation, methodology, writing—review and editing; Laura A. Gallagher: formal analysis, investigation, methodology; Dhananjay Shinde: investigation, methodology; Jongsam Ahm: investigation, methodology; Heather M. Olson: investigation, methodology; Thomas L. Fillmore: investigation, methodology; Joshua N. Adkins: funding acquisition, supervision, formal analysis, writing—review and editing; Fareha Razvi: formal analysis, writing—review and editing; Kenneth W. Bayles: funding acquisition, supervision, writing—review and editing; Paul D. Fey: funding acquisition, supervision, writing—review and editing; Vinai C. Thomas: formal analysis, funding acquisition, supervision, investigation, methodology, writing—review and editing; Felipe Cava: funding acquisition, formal analysis, supervision, writing—review and editing; Geremy C. Clair: conceptualization, formal analysis, investigation, methodology, supervision, writing—review and editing; and James P. O’Gara: conceptualization, formal analysis, funding acquisition, project administration, supervision, writing—original draft, writing—review and editing.