Genetic and pharmacological inactivation of peptidoglycan remodeling increases antibiotic susceptibility of vancomycin-resistant Enterococcus faecium
1Department of Immunology and Microbiology, Scripps Research; La Jolla, California 92037, United States
2Cancer Center, Cold Spring Harbor Laboratory; Cold Spring Harbor, NY 11724, USA
3La Trobe Institute for Molecular Science, La Trobe University, Science Dr., Bundoora, Melbourne, VIC, 3086 Australia
4Department of Integrative Structural & Computational Biology, Scripps Research; La Jolla, California 92037, United States
5Department of Molecular Medicine, Scripps Research; La Jolla, California 92037, United States
6Department of Chemistry, Scripps Research; La Jolla, California 92037, United States
7Division of Infectious Diseases, University of Pittsburgh School of Medicine, Pittsburgh, PA 15213
*Corresponding authors: moses@cshl.edu, hhang@scripps.eduAbstract
Vancomycin-resistant Enterococcus faecium (VREfm) is a leading cause of healthcare-associated infections globally and demands new approaches for treatment. Here we show that genetic and pharmacological inactivation of a highly conserved NlpC/P60 peptidoglycan hydrolase, secreted antigen A (SagA), enhanced vancomycin susceptibility of VREfm ex vivo and in vivo. Notably, genetic deletion of sagA impaired VREfm peptidoglycan remodeling, growth and increased the activity of vancomycin. We then identified first-in-class covalent NlpC/P60 peptidoglycan hydrolase inhibitors and demonstrated that pharmacological inactivation of SagA activity also impaired peptidoglycan remodeling and increased the efficacy of vancomycin across genetically distinct VREfm clinical isolates. Our study reveals peptidoglycan hydrolases are druggable targets whose inactivation improves the efficacy of vancomycin against VREfm.
Article notes
Competing Interest Statement
The authors have declared no competing interest.
Summary of Updates:
Introduction
Increasing antimicrobial resistance (AMR) in bacterial pathogens and limited antibiotic discovery require new approaches to address this major threat to human health world-wide.1 Amongst the ESKAPE pathogens,2 vancomycin-resistant E. faecium (VREfm) infections have become more prevalent among healthcare-associated infections and can acquire resistance to last-resort antibiotics like linezolid, daptomycin, and tigecycline.3,4 The AMR crisis demands new approaches to prevent and treat VREfm infections, which are also correlated with poor patient outcomes leading to high mortality rates.5,6 While the synthesis of next-generation antibiotics provides new derivatives to address AMR7,8 and innovative approaches are being employed to discover new classes of antibiotics,1,9,10 a better understanding of E. faecium biology may provide new targets for antimicrobial development.
Enterococcus is a genus of ubiquitous Gram-positive bacteria, among which E. faecium and E. faecalis are the most prominent in humans and other mammals.11 While E. faecium can acquire antibiotic resistance and cause healthcare-associated infections,2,11,12 non-pathogenic strains of E. faecium have been reported to have beneficial effects on host physiology and been developed into probiotics.13 Our mechanistic dissection of commensal E. faecium-host interactions revealed that secreted antigen A (SagA), a highly conserved NlpC/P60 peptidoglycan hydrolase, can generate non-crosslinked muropeptides to promote host immunity.14–19 Of note, we also demonstrated sagA is essential for peptidoglycan remodeling, cell separation and growth in the commensal strain of E. faecium (Com15).19 Moreover, the commensal strains of E. faecium (Com15-ΔsagA and phage-resistant Com12 expressing catalytically inactive alleles of sagA) were more susceptible to cell-wall targeting antibiotics,19–21 suggesting SagA may be a potential antimicrobial target.
As sagA is highly conserved amongst E. faecium strains, including VREfm strains containing the vancomycin resistance vanA and vanB gene clusters,17 we investigated genetic and pharmacological inactivation of this key peptidoglycan hydrolase as a therapeutic target for treating VREfm infections. We discovered that VREfm-ΔsagA showed defective peptidoglycan remodeling, impaired cell separation and growth as well as increased susceptibility to last-line antibiotics. Notably, even though a VREfm-ΔsagA strain still encodes the vanA operon, the minimal inhibitory concentration (MIC) of vancomycin was decreased and attenuated by vancomycin treatment in a mouse model of VREfm-induced sepsis in vivo.
In parallel, we explored pharmacological inactivation of SagA activity. Bacterial essential enzymes and virulence factors can be attenuated by covalent inhibitors, including recent fluorosulfates and sulfonyl fluorides that can undergo Sulfur(VI) Fluoride Exchange (SuFEx) chemistry.22,23 Sulfonyl fluorides can covalently label a broad range of amino acid residues (Ser > Thr > Tyr > Cys > Lys > His), however their reactivity is context-dependent and largely governed by catalytic environment of the target enzyme.22–24 From a library of promiscuous SuFEx-based sulfonyl fluorides accessed through a Diversity Oriented Clicking (DOC) approach24–26, we identified β-chloro alkenyl sulfonyl fluorides as the first-in-class covalent inhibitors of the NlpC/P60 cysteine endopeptidases. The most potent β-chloro alkenyl sulfonyl fluoride SagA inhibitor impaired peptidoglycan remodeling in VREfm, reduced the MIC of vancomycin in genetically distinct VREfm strains, attenuated infection of macrophages ex vivo, and improved outcomes of VREfm-induced sepsis in vivo. Our studies demonstrate that peptidoglycan hydrolases are crucial for bacterial cell wall remodeling and are druggable targets to promote the efficacy of antibiotics against VREfm.
Results
SagA is critical for peptidoglycan remodeling and antibiotic susceptibility of VREfm
SagA is a member of the highly conserved NlpC/P60-family of cysteine endopeptidases that are important for bacterial physiology.27 We previously demonstrated SagA is crucial for commensal E. faecium (Com15 strain) peptidoglycan remodeling, cell separation and growth.19 Consistent with our previous analysis,17 phylogenetic analysis confirmed the presence of sagA using an international collection of publicly available vancomycin-susceptible E. faecium (n=164) and VREfm (n=395) isolates from 99 sequence types (Extended Data Fig.1). Our analysis also revealed the distribution of other NlpC/P60 hydrolases in E. faecium, with some being more prevalent in VREfm strains (Extended Data Fig.1), but their function was unknown. Only 4 out of 10 NlpC/P60 hydrolases were found in human clinical VREfm isolate ERV165 (sequence type 412; vanA genotype; Extended Data Fig. 2a). Using improved genetic methods for Enterococcus,28,29 we generated isogenic deletion strains of all 4 NlpC/P60 hydrolases in VREfm ERV165 strain (Extended Data Fig. 2b-e) and found that only the ERV165-ΔsagA (further referred as ΔsagA) mutant exhibited impaired growth and altered colony morphology (Fig. 1a-c and Extended Data Fig. 2f), which was rescued by sagA chromosomal complementation (ΔsagA::sagA). Transmission electron microscopy (TEM) analysis revealed that ΔsagA cells failed to properly separate, resulting in the formation of aberrant cellular clusters (Fig. 1d). Further cryo-electron tomography (cryo-ET) analysis of ΔsagA showed a decrease in cell wall thickness and a modest increase in septum thickness (Extended Data Fig. 3 and 4). The cell wall thickness was restored in the sagA-complemented strain, however no significant changes in septum thickness were observed (Extended Data Fig. 4c). Similar to Com15-ΔsagA mutant19, ΔsagA also exhibited lower amounts of non-crosslinked muropeptides and increased levels of crosslinked peptidoglycan fragments (Extended Data Fig. 5). These results demonstrate that in addition to its importance in commensal E. faecium strains, sagA is also essential for VREfm peptidoglycan remodeling, bacterial cell separation and growth, even though the majority of VREfm strains have acquired additional NlpC/P60 hydrolases (Extended Data Fig. 1).
We next evaluated the contribution of SagA to VREfm antibiotic susceptibility. Based on previous antibiotic susceptibility studies in commensal E. faecium ΔsagA mutant strains,19 we analyzed the activity of ampicillin, daptomycin and ceftriaxone in VREfm ΔsagA. Surprisingly, VREfm ΔsagA only exhibited modest inhibition of bacterial growth with ampicillin, daptomycin and ceftriaxone compared to ERV165 wild-type (Extended Data Fig. 6a-c) and no significant differences in MIC (Extended Data Table 1). However, ΔsagA, but not other NlpC/P60 hydrolase deletion strains (Extended Data Fig. 6d), showed increased susceptibility to vancomycin with 2-fold difference of MIC (Extended Data Table 2), which was abrogated in the sagA-complemented strain (Fig. 2a,b and Extended Data Table 2). Adaptive laboratory evolution experiment in the presence of sub-MIC vancomycin concentration demonstrated that neither ΔsagA, nor WT or sagA-complemented strain developed additional vancomycin resistance after two weeks (Extended Data Fig 6e). To investigate the vancomycin susceptibility of ERV165-ΔsagA, we performed whole-genome sequencing (WGS), evaluated antibiotic binding and analyzed peptidoglycan remodeling further. WGS of the ΔsagA strain confirmed the sagA-gene deletion and revealed 7 missense, 1 frame-shift and 12 silent mutations (Extended Data Table 3 and 4), none of which were associated with peptidoglycan synthesis, cell wall remodeling or vancomycin resistance (Extended Data Fig. 7a). Of note, both ΔsagA and sagA-complemented strains have L10S and L68S missense mutations found in mapZ (Midcell Anchored Protein Z) homolog that encodes a protein essential for bacterial cell division (Extended Data Table 3 and 5). However, these mutations occurred outside annotated functional domains and in a region lacking conserved sequence features. Importantly, the increased vancomycin susceptibility of ΔsagA was abrogated in the sagA-complemented strain that retained the mutations as ΔsagA (Fig. 2b and Extended Data Table 2, 5 and 6). We then employed fluorescent vancomycin (Van-BODIPY)30 and fluorescent D-amino acid (HADA)31 to evaluate antibiotic binding and peptidoglycan stem peptide remodeling, respectively. Fluorescence microscopy revealed increased Van-BODIPY and HADA staining in ΔsagA that was abrogated by sagA complementation (Fig. 2c-e). These results suggest that loss of SagA expression and defective peptidoglycan remodeling led to the improved vancomycin binding and increased antibiotic susceptibility.
To evaluate the impact of SagA on VREfm infection in vivo, we employed a mouse peritonitis infection model32. All three strains (ERV165 wild-type, ΔsagA and sagA-complemented strain) caused similar levels of weight loss as a marker of infection and exhibited high bacterial burdens in the spleen and liver (Fig. 2f,g). However, treatment with clinical doses of vancomycin33 significantly improved weight loss and cleared bacterial burden in mice infected with ΔsagA strain, but not in those infected with the wild-type VREfm or sagA-complemented strains (Fig. 2f,g). These results demonstrate that SagA not only impacts peptidoglycan remodeling and activity of cell wall-targeting antibiotics in commensal E. faecium strains, but importantly also modulates vancomycin susceptibility in VREfm ex vivo and VREfm-induced sepsis in vivo.
SuFEx-based sulfonyl fluorides covalently label and inactivate SagA NlpC/P60 hydrolase activity
To identify pharmacological inhibitors of SagA endopeptidase activity, we developed a high-throughput assay based on competitive labeling34 of the only cysteine residue (C433) in the NlpC/P60 hydrolase domain active site35 using a fluorescent tetramethyl rhodamine-iodoacetamide (TMR-IA) probe (Fig. 3a and Extended Data Fig. 8a). TMR-IA selectively labeled the recombinant SagA-NlpC/P60 hydrolase domain, but not the inactive C433A mutant or wild-type pretreated with cysteine-reactive controls N-methylmaleimide (NMM) or iodoacetamide (IA) (Extended Data Fig. 8b-d). Using this assay, we screened a library of SuFEx-based sulfonyl fluorides24 for potential covalent inhibitors of SagA. We identified 86 sulfonyl fluorides that reduced TMR-IA labeling greater than 80% (Extended Data Fig. 8e). Secondary screening of the top sulfonyl fluorides using gel-based competitive TMR-IA labeling (Extended Data Fig. 8f-i) and follow-up SagA peptidoglycan hydrolase activity assays36 (Extended Data Fig. 8a), revealed a subset of β-chloro alkenyl sulfonyl fluoride compounds (peptidoglycan hydrolase inhibitors, pghi-1 to 5) with low micromolar IC50 values in vitro (Fig. 3b,c and Extended Data Fig. 9b). Small modifications in the phenyl ring (pghi-1 to 5) did not substantially affect SagA inhibition in vitro, whereas substitution of the β-alkenyl position (pghi-6) compromised inhibition potency (Fig. 3b,c and Extended Data Fig. 9b).
Computational covalent docking suggested that these peptidoglycan hydrolase inhibitors can be accommodated in the active site, stabilized by interactions with nearby H494 and F522, after reacting with cysteine (C443) of SagA via the sulfonyl fluoride group (Fig. 3d) or β-alkenyl chloride group (Extended Data Fig. 10a). Intact protein mass spectrometry analysis of recombinant SagA treated with pghi-4 revealed a mass shift that corresponds to the mass of the SagA-pghi-4 adduct (Fig. 3e). The observed mass increase matched the calculated molecular weight loss of fluorine from pghi-4, suggesting covalent modification through the sulfonyl fluoride group and formation of a thiosulfonate adduct (Extended Data Fig. 10b-d). Indeed, treatment of the catalytically inactive C433A SagA mutant with pghi-4 did not result in a mass shift (Fig. 3e). Incubation of SagA with an inactive pghi-6 resulted in minor formation of thiosulfonate-linked SagA-pghi-6 adduct (Extended Data Fig. 10e) indicating the importance of the β-vinyl chloro group for binding to SagA. Our efforts to crystallize SagA-pghi-4 complex did not yield diffraction quality crystals (see Methods), likely due to instability of the SagA-pghi-4 adduct. Further intact protein mass spectrometry analysis of the SagA-pghi-4 adducts (peak 1, thiosulfonate, Extended Data Fig. 9b-d) identified peaks 3 and 5 with masses corresponding to thiosulfonic and sulfinic acid derivatives of SagA (Extended Data Fig. 10b-d). After 16 hours of treatment, the abundance of the SagA-pghi-4 adduct significantly decreased, while the levels of hydrolysis products increased (Extended Data Fig. 10b-d), indicating instability of the SagA-pghi-4 adduct. To explore the selectivity of identified pghi-1 to 5, we evaluated these compounds with other VREfm NlpC/P60 hydrolases in vitro. Pghi-1 to 5 also inhibited the endopeptidase activity of peptidoglycan hydrolase 2 (PGH2), which has NlpC/P60 domain with high amino acid sequence and structural homology to SagA, but not the more divergent CwlT-like peptidoglycan hydrolase 3 (PGH3) (Extended Data Fig. 11a-c). Taken together, these results suggest that the identified β-chloro alkenyl sulfonyl fluorides can bind and covalently react with the catalytic cysteine of SagA and peptidoglycan hydrolase orthologs with structurally similar NlpC/P60 hydrolase domains to inactivate their enzymatic activity.
SagA inhibitors improve antibiotic susceptibility in VREfm strains
We next evaluated the activity of pghi-1 to 6 on VREfm growth and antibiotic susceptibility. Either pghi-4 or pghi-5 alone caused a mild growth delay of VREfm (ERV165), while the other compounds did not significantly affect VREfm growth under laboratory conditions (Extended Data Fig. 12a,b). Building upon our observations of SagA modulation of VREfm antibiotic susceptibility (Figs. 1 and 2), we evaluated these compounds on VREfm growth in combination with a low dose of vancomycin (Fig. 4a and Extended Data Fig. 12c). Notably, pghi-4 and pghi-5 significantly enhanced the activity of vancomycin compared to the other β-chloro alkenyl sulfonyl fluorides and the inactive analog pghi-6 (Fig. 4a,b and Extended Data Fig. 12c). The most active compound pghi-4 lowered the MIC values of vancomycin up to 8 fold in a concentration-dependent manner by checkerboard assay analysis (Fig. 4c and Extended Data Table 7). Bacterial time-kill analysis showed that pghi-4 in combination with vancomycin significantly limited VREfm growth compared to either agent alone (Fig. 4d and Extended Data Fig. 12d). Similar to ΔsagA, pghi-4 only showed very modest to no enhancement of ampicillin, daptomycin and ceftriaxone antibiotic activity in VREfm (Extended Data Fig. 12e,f).
To characterize the mechanism of action of SagA inhibitor pghi-4, we performed imaging, peptidoglycan remodeling and chemoproteomic studies of VREfm. Differential interference contrast (DIC) microscopy revealed that pghi-4 in combination with vancomycin induced aberrant bacterial colony morphology and increased cell clustering compared to control or either agent alone (Fig. 4e,f). Cryo-ET showed dead cells with impaired peptidoglycan cleavage in VREfm treated with pghi-4 in combination with vancomycin (Extended Data Fig. 13). Both pghi-4 and vancomycin alone increased cell wall thickness, but their combination resulted in reduced thickness (Extended Data Fig. 14a,b). Interestingly, septum thickness was significantly increased under all treatment conditions (Extended Data Fig. 14a,c). These results suggest that the combination of pghi-4 and vancomycin impaired the ultrastructure of VREfm peptidoglycan. Similar to ΔsagA, pghi-4 also increased Van-BODIPY (5.9 fold, Fig. 4f,g) and HADA (Fig. 4f,h) labeling, suggesting enhanced vancomycin binding due to inhibition of peptidoglycan remodeling and the accumulation of vancomycin-target D-Ala-D-Ala-containing muropeptides37. Penicillin-BODIPY (Bocillin) labeling was only slightly increased in pghi-4 treated VREfm (1.3 fold) and in ΔsagA (2.3 folds, Extended Data Fig. 15a, b) by fluorescence microcopy as well as in-gel penicillin-binding protein 5 (PBP5) labeling (Extended Data Fig. 15c), which are consistent with our β-lactam (ampicillin) susceptibility analyses (Extended Data Fig. 6a-c, Extended Data Table 1 and Extended Data Fig. 12e,f). We observed mutations (M485A, A499T and E629V) in PBP5 of VREfm ERV165 compared to commensal E. faecium Com15 strain (Extended Data Fig. 15d), which are known to drive PBP5-dependent resistance to β-lactam antibiotics.38 Genetic deletion of sagA in VREfm ERV165 did not affect those mutations (Extended Data Fig. 15d), which may explain why β-lactam antibiotics are more active in commensal E. faecium ΔsagA strains38.
Quantitative LC-MS analysis of pghi-4-treated ERV165 did not yield similar profiles of soluble fragments from peptidoglycan compared to ΔsagA (Extended Data Fig. 16a,c). However, for these experiments ERV165 was only treated with sub-inhibitory and non-bactericidal dose of 50 µM pghi-4 to obtain sufficient material for LC-MS analysis. Moreover, the genetic inactivation of sagA may result in accumulation or depletion of soluble fragments from peptidoglycan that are not fully recapitulated with incomplete pharmacological inhibition of SagA. Nonetheless, LC-MS analysis of digested peptidoglycan fragments from pghi-4 and vancomycin co-treated ERV165 revealed higher levels of D-Ala-D-Ala-containing GlcNAc-MurNAc-pentapeptide in peptidoglycan (Extended Data Fig. 16d,e). As pghi-4 targets the active site C443 of SagA in vitro (Fig. 3d,e), we also evaluated covalent labeling of SagA and other reactive Cys-residues in ERV165 proteome by competitive chemoproteomics using an iodoacetamide-alkyne (IA-alk) probe (Extended Data Fig. 17a and Supporting Data 1). Indeed, treatment of ERV165 with pghi-4 reduced IA-alk labeling of the SagA by in-gel fluorescence labeling and western blot analysis (Extended Data Fig. 17b). The quantitative proteomic analysis of pghi-4-competitive Cys-reactive protein targets showed only a few other proteins were targeted in a dose-dependent manner (Extended Data Fig. 17c). However, none of these other Cys-reactive candidate pghi-4-target proteins have been implicated in VREfm peptidoglycan remodeling or antibiotic susceptibility. These results suggest that pharmacological inactivation of SagA by pghi-4 increases vancomycin susceptibility of VREfm by impairing peptidoglycan remodeling, which increases the levels of D-Ala-D-Ala-containing muropeptides and promotes vancomycin binding.
To characterize the scope of pghi-4 adjuvant activity with vancomycin, we evaluated additional VREfm strains including clinical isolates from patients that underwent chemotherapy or hematopoietic stem cell transplantation39. Notably, some VREfm strains (700221, DVT1574, DVT3347) showed increased susceptibility to pghi-4 and vancomycin co-treatment compared ERV165 and other VREfm strains (Fig. 4i and Extended Data Fig. 18a), even though they all harbor vanA-type resistance39, of different sequence types (Extended Data Table 8) and exhibit similar vancomycin MICs (Extended Data Table 9). Interestingly, pghi-4 adjuvant activity correlated with intracellular SagA protein expression levels in these VREfm strains (Extended Data Fig. 18b). These results suggest β-chloro-alkenyl sulfonyl fluoride SagA inhibitors can increase the vancomycin susceptibility of different VREfm clinical isolates, which correlates with their SagA protein expression levels.
Pharmacological inactivation of SagA increases vancomycin susceptibility of VREfm in vivo
To investigate the therapeutic potential of SagA inhibitor pghi-4 as an antibiotic adjuvant, we evaluated its ability to potentiate vancomycin activity using cellular and mouse models of VREfm infection. The combination of pghi-4 and vancomycin, but not either agent alone, reduced VREfm infection of murine (RAW264.7) (Fig. 4j) and human (THP-1) monocytes (Extended Data Fig. 19a) in a dose-dependent manner, that was not attributed to pghi-4 cytotoxicity (Extended Data Fig. 19b, c). We next tested therapeutic efficacy of the combination therapy in vivo. A single dose of pghi-4 in combination with vancomycin did not significantly decrease VREfm counts in mouse organs 6- or 24-hours post-infection compared to PBS treatment (Extended Data Fig. 20). However, a two-dose therapeutic regimen significantly reduced weight loss in VREfm-infected mice co-treated with vancomycin and pghi-4 (Fig. 4k). Colony forming unit (CFU) analysis of spleen and liver further demonstrated a significant reduction of VREfm burden in the vancomycin and pghi-4 co-treated group compared to PBS, while monotherapies had no significant effect (Fig. 4l). These results demonstrate that pharmacological inactivation of peptidoglycan remodeling in combination with vancomycin can attenuate VREfm infection ex vivo and VREfm-induced sepsis in vivo.
Discussion
Bacterial infections cause a significant healthcare and financial burden worldwide. While antibiotics help manage bacterial infections, many bacterial pathogens have acquired antibiotic resistance and now are difficult to treat, with 4.7 million deaths associated with AMR worldwide in 2019.40 Progress in development of new antibiotic agents and targets have resulted in novel approaches to combat antibiotic-resistant bacteria.22,41–46 Recently, combination therapeutic approaches using antibiotic adjuvants (non-antimicrobial agents enhancing antibiotic activity) have provided new entities to extend the lifespan of clinical antibiotics.47–52 For example, combination treatment with β-lactamase inhibitors (BLIs) overcame antibiotic resistance and restored β-lactam activity against bloodstream infections in hematological neutropenic patients.53 This therapeutic combination of an antibiotic adjuvant coupled with an existing antibiotic showed clinical potential and is promising for further development, however, antibiotic adjuvants for other classes of antibiotics are underdeveloped. The alarming increase of VREfm infections54,55 and evolution of antibiotic resistance in patients highlights the urgent need for new therapeutic approaches to overcome resistance and prevent adaptation in vulnerable hosts.39
While natural products and their derivatives have been reported to broadly inhibit peptidoglycan remodeling or target CHAP (cysteine, histidine-dependent amidohydrolases/peptidases) domain containing hydrolases in other Gram-positive bacterial pathogens,56,57 our studies demonstrate NlpC/p60 peptidoglycan hydrolases and their inhibitors may serve as important new antibiotic targets and agents. Here we demonstrated that SagA, a NlpC/p60 hydrolase important for peptidoglycan remodeling, modulates vancomycin susceptibility in VREfm and can be pharmacologically targeted for improved combination therapy (Extended Data Fig. 21). Unlike other NlpC/P60 hydrolases that are in VREfm strains (Extended Data Fig. 2), only deletion of sagA compromised cell growth, separation, peptidoglycan remodeling and increased VREfm susceptibility to vancomycin ex vivo and VREfm-induced sepsis in vivo, which could be rescued by sagA re-expression. Long-term serial passaging of the ΔsagA strain under sub-MIC vancomycin pressure did not substantially change vancomycin susceptibility (Extended Data Fig. 6e). While we have shown that SagA expression in E. faecium and probiotic bacterial species can promote intestinal immunity14,15 and cancer immunotherapy in vivo17, the VREfm ΔsagA mutant strain is not less pathogenic in this mouse peritonitis infection model, but is more susceptible to vancomycin in vivo (Fig. 2f,g). These results suggest SagA does not significantly contribute to VREfm-induced sepsis in vivo, but can be target for antibiotic adjuvants.
Based on these observations, we identified the first-in-class NlpC/p60 peptidoglycan hydrolase inhibitors (pghi-1 to 5) and demonstrated that these β-chloro alkenyl sulfonyl fluorides covalently label the active site of SagA and inhibit peptidoglycan hydrolase activity in vitro. We previously reported the synthesis of β-chloro alkenyl sulfonyl fluorides,24 however, their activity on NlpC/p60 peptidoglycan hydrolases was not evaluated. Notably, pghi-4 effectively inhibited peptidoglycan remodeling and increased vancomycin susceptibility in several VREfm strains ex vivo (Fig. 4i and Extended Data Fig. 18). Interestingly, pghi-4 showed 8-fold enhanced vancomycin susceptibility in VREfm ERV165 compared to the 2-fold enhanced vancomycin activity in the isogenic ΔsagA mutant strain (Figs. 2b, 4c and Extended Date Tables 2, 7). Moreover, pghi-4-treated VREfm ERV165 showed increased levels of vancomycin-target D-Ala-D-Ala-containing peptidoglycan fragments compared to the isogenic ΔsagA mutant strain (Extended Data Figs. 5 and 16). These observations suggests that β-chloro alkenyl sulfonyl fluorides, such as pghi-4, may target additional peptidoglycan remodeling enzymes in VREfm beyond SagA. In fact, we showed that β-chloro alkenyl sulfonyl fluorides, including pghi-4, can also inhibit the PGH2 NlpC/p60 endopeptidases in vitro (Extended Data Fig. 11c). Even though genetic deletion of other NlpC/P60 hydrolases (pgh2, pgh3, and pgh4) did not cause growth defect or changes in vancomycin susceptibility (Extended Data Fig. 2e and 6d), pharmacological inhibition of both SagA and PGH2 VREfm NlpC/P60 hydrolase may contribute to the overall observed activity of pghi-4 in VREfm strains. Although our quantitative competitive chemoproteomic analysis of cysteine-reactive proteins with pghi-4 did not reveal other NlpC/P60 hydrolases as potential targets in VREfm (Extended Data Fig. 17), these enzymes may not be effectively labeled by iodoacetamide reagents. It is also possible that pghi-4 may react with other nucleophilic amino acids on other proteins that were not identified in our analysis. The direct analysis of pghi-4 targets will require the generation of β-chloro alkenyl sulfonyl fluoride probes that are unfortunately not accessible by our current synthetic methods and will require the development of next-generation NlpC/P60 hydrolase inhibitor and probes. Nonetheless, our discovery and development of β-chloro alkenyl sulfonyl fluorides as covalent NlpC/P60 hydrolase inhibitors demonstrate pharmacological inhibition of peptidoglycan remodeling can improve vancomycin activity in VREfm and attenuate the infection of macrophages ex vivo (Fig. 4j) and VREfm-induced sepsis in vivo (Fig. 4k-l). The further development of more potent NlpC/p60 hydrolase inhibitors should afford new antimicrobial adjuvants to prevent and treat VREfm infections.
Methods
Chemistry
The synthesis and characterization of sulfonyl fluorides identified as SagA inhibitors in this work are reported elsewhere 24. Characterization of resynthesized pghi-4 matched the previous report (Extended Data Fig. 22-25). 1H NMR (400 MHz, CDCl3) δ 7.36 (dd, J = 8.5, 2.4 Hz, 1H), 7.17 (d, J = 2.3 Hz, 1H), 6.96 (d, J = 2.3 Hz, 1H), 6.93 (d, J = 8.6 Hz, 1H), 4.88 (tt, J = 6.3, 3.1 Hz, 1H), 3.93 (s, 3H), 2.09 – 1.98 (m, 2H), 1.97 – 1.89 (m, 3H), 1.90 – 1.80 (m, 2H), 1.67 (tdd, J = 10.6, 7.5, 4.8 Hz, 2H); 13C NMR (101 MHz, CDCl3) δ 152.4, 152.1, 150.0, 125.7, 121.8, 115.4, 115.2, 113.6, 110.7, 80.9, 56.4, 33.0, 24.3; 19F NMR (377 MHz, CDCl3) δ 65.0; LCMS (ESI+): calculated for C14H16ClFO4SNa [M+H]+: m/z = 335.05, m/z found 335.09.
Bacteria
The bacterial species used in this study are listed in Extended Data Table 10. All Enterococcus were grown aerobically at 37 °C at 200 RPM shaking in Brain Heart Infusion (BHI) broth (Fisher Scientific, 237500) with appropriate antibiotics. Following Clinical & Laboratory Standards Institute (CLSI) guidelines, vancomycin susceptibility of E. faecium was evaluated in Mueller Hinton Broth (MHB, BD 275730). E. coli was grown in Luria-Bertani (LB, BD 244610) broth.
Phylogenetic analysis
To understand the distribution of NlpC/P60 hydrolases in vancomycin-susceptible and -resistant E. faecium strains we downloaded all complete genomes of E. faecium on NCBI (n= 559 as of January 2026), consisting of genotypically vancomycin-susceptible (n=164) and -resistant E. faecium (n=395) isolates from 99 sequence types. In silico multi-locus sequence typing (MLST) was assigned using the program mlst (https://github.com/tseemann/mlst) (v2.19.0). The genome assemblies were screened for antimicrobial resistance determinants using abriTAMR58 (v1.0.18) with default settings.
All assemblies were annotated using the run_prokka function in Panaroo (v1.2.10) with clean-mode set to strict, which annotates each sample with the same gene model using Prokka (v1.14.6).59,60 The pangenome was defined using Panaroo (v1.2.10), which utilizes a pangenome graph-based approach for clustering. Core genes were defined as genes present in >99% of strains, with accessory in at least >1%. From the pangenome, functional annotation was assigned using eggNOG-mapper (v2.1.2) with default Diamond mode.61 All NlpC/P60 hydrolases were identified using the Clusters of Orthologous Genes identifier COG0791 and manually verified using CD-search (v3.2). A maximum-likelihood phylogenetic tree using the alignment of core genes (core_gene_alignment_filtered.aln) was inferred using IQ-TREE (v2.1.4)62 with a general time-reversible (GTR+G4) substitution model and 1,000 bootstrap replicates. All figures were generated in R (v.4.3.0, https://www.r-project.org/) using tidyverse (v.1.3.1), patchwork (v.1.1.1), ggtree (v.3.8.2), and ggnewscale (v.0.4.5).
Plasmids construction
1.pPK99 (pJC005.gent-ΔsagA)
Using oligos oPK996 and oPK997, the sRNA promoter driving the sagA protospacer was PCR-amplified (using pUCsRNAP plasmid as template). oPK996 contains an XhoI restriction site (indicated in lowercase), and oPK997 includes the sagA protospacer from E. faecium ERV165 (indicated in lowercase and underlined) to serve as the CRISPR-Cas12a target for counter-selection during the recombineering process. Similarly, oPK998 (which includes the identical sagA protospacer sequence in underlined lowercase and a repeat region in underlined uppercase) and oPK999 were used to amplify the upstream flanking region of sagA using E. fm ERV165 genomic DNA (gDNA) as template, while oPK1000 and oPK1001 (the latter containing an AscI site, shown in lowercase) amplified the downstream sagA flanking region using the same gDNA as template. All oligos were designed such that the three resulting PCR products contained 35–40 bp overlapping regions and were assembled using splicing by overlap extension (SOE) PCR. The resulting fragment was cloned into the pJC005.gent vector via XhoI and AscI restriction digestion followed by ligation. Positive clones were screened in E. coli NEB-5α, yielding the construct pPK99. This plasmid was then transformed into E. faecium ERV165 for generation of the clean sagA deletion mutant (sPK377), as described below.
2.pPK158 (pJC005.gentR-sagA chromosomal complementation plasmid)
For sagA chromosomal complementation in the VREfm ERV165 ΔsagA clean deletion mutant (sPK377), a neutral chromosomal locus was selected downstream of the Holliday junction resolvase gene ruvX, where no signatures of nearby gene promoters or terminators were detected (Extended Data Fig. 1b). This site contained a protospacer adjacent motif (PAM), making it suitable for CRISPR-Cas12a-based counter-selection. The sagA complementation plasmid (pPK158) was constructed by integrating the sagA gene into the selected PAM site, thereby disrupting it for CRISPR-Cas12a-based counter-selection. The construct included all essential regulatory elements of ERV165 sagA: the native sagA promoter, ribosome binding site (RBS), open reading frame (ORF), and transcriptional terminator.
Briefly, the small RNA (sRNA) promoter, derived from the pUCsRNAP template and driving the neutral locus protospacer, was PCR-amplified using oligos oPK996 and oPK1020. oPK996 includes an XhoI restriction site (in lowercase), while oPK1020 contains the neutral locus protospacer from E. faecium ERV165 (in lowercase and underlined), which served as the CRISPR-Cas12a target during recombineering. The upstream region of the neutral locus was amplified using sagA clean deletion (sPK377) gDNA as template with oligos oPK1021 and oPK1022. oPK1021 has the protospacer (underlined lower case) and a repeat region (underlined uppercase). The ERV165 sagA promoter, RBS, ORF, and both translational and transcriptional terminators were amplified using oPK1023 and oPK1034 using E. fm ERV165 gDNA as template. Finally, the downstream region of the neutral locus was amplified using sPK377 gDNA as template with oPK1025 and oPK1026, which include AscI restriction sites (in lowercase). All oligos were designed such that the three resulting PCR products contained 35–40 bp overlapping regions and were assembled using splicing by overlap extension (SOE) PCR. The resulting fragment was cloned into the pJC005.gent vector via XhoI and AscI restriction digestion followed by ligation. Positive clones were screened in E. coli NEB-5α, yielding the construct pPK158. This plasmid was then transformed into E. faecium ERV165 for generation of sagA chromosomal complementation strain in ΔsagA clean deletion mutant (sPK394), as described below.
3.pPK156 (pJC005.gent- Δpgh2)
Using oligos oPK996 and oPK1002, the sRNA promoter driving the pgh2 protospacer was PCR-amplified (using pUCsRNAP plasmid as template). oPK996 contains an XhoI restriction site (indicated in lowercase), and oPK1002 includes the pgh2 protospacer from E. faecium ERV165 (indicated in lowercase and underlined) to serve as the CRISPR-Cas12a target for counter-selection during the recombineering process. Similarly, oPK1003 (which includes the identical pgh2 protospacer sequence in underlined lowercase and a repeat region in underlined uppercase) and oPK1004 were used to amplify the upstream flanking region of pgh2 using E. fm ERV165 genomic DNA (gDNA) as template, while oPK1005 and oPK1006 (the latter containing an AscI site, shown in lowercase) amplified the downstream pgh2 flanking region using the same gDNA as template. All oligos were designed such that the three resulting PCR products contained 35–40 bp overlapping regions and were assembled using splicing by overlap extension (SOE) PCR. The resulting fragment was cloned into the pJC005.gent vector via XhoI and AscI restriction digestion followed by ligation. Positive clones were screened in E. coli NEB-5α, yielding the construct pPK156. This plasmid was then transformed into E. faecium ERV165 for generation of the clean pgh2 deletion mutant, as described below.
4.pPK107 (pJC005.gent- Δpgh3)
Using oligos oPK639, the sRNA promoter driving the pgh3 protospacer was PCR-amplified (using pUCsRNAP plasmid as template). oPK844 contains pgh3 protospacer from E. faecium ERV165 (indicated in lowercase and underlined) to serve as the CRISPR-Cas12a target for counter-selection during the recombineering process. Similarly, oPK845 (which includes the identical pgh3 protospacer sequence in underlined lowercase and a repeat region in underlined uppercase) and oPK846 were used to amplify the upstream flanking region of pgh3 using E. fm ERV165 genomic DNA (gDNA) as template, while oPK847 and oPK848 amplified the downstream pgh3 flanking region using the same gDNA as template. All oligos were designed such that the three resulting PCR products contained 35–40 bp overlapping regions and were assembled using splicing by overlap extension (SOE) PCR using oPK656 and oPK659. The resulting fragment was cloned into the pJC005.gent vector via XhoI and AscI restriction digestion followed by ligation. Positive clones were screened in E. coli NEB-5α, yielding the construct pPK107. This plasmid was then transformed into E. faecium ERV165 for generation of the clean pgh3 deletion mutant, as described below.
5.pPK154 (pJC005.gent- Δpgh4)
Using oligos oPK996 and oPK1007, the sRNA promoter driving the pgh4 protospacer was PCR-amplified (using pUCsRNAP plasmid as template). oPK996 contains an XhoI restriction site (indicated in lowercase), and oPK1007 includes the pgh4 protospacer from E. faecium ERV165 (indicated in lowercase and underlined) to serve as the CRISPR-Cas12a target for counter-selection during the recombineering process. Similarly, oPK1008 (which includes the identical pgh4 protospacer sequence in underlined lowercase and a repeat region in underlined uppercase) and oPK1009 were used to amplify the upstream flanking region of pgh4 using E. fm ERV165 genomic DNA (gDNA) as template, while oPK1010 and oPK1011 (the latter containing an AscI site, shown in lowercase) amplified the downstream pgh4 flanking region using the same gDNA as template. All oligos were designed such that the three resulting PCR products contained 35–40 bp overlapping regions and were assembled using splicing by overlap extension (SOE) PCR. The resulting fragment was cloned into the pJC005.gent vector via XhoI and AscI restriction digestion followed by ligation. Positive clones were screened in E. coli NEB-5α, yielding the construct pPK154. This plasmid was then transformed into E. faecium ERV165 for generation of the clean pgh4 deletion mutant, as described below.
Preparation of electrocompetent E. faecium ERV165 cells and electroporation
The protocol described below is a standardized procedure used in our lab. A single colony of E. faecium ERV165 and their mutant derivatives were inoculated into 4 mL of plain BHI and grown overnight. The culture was then diluted 1:50 into 200 mL GS-BHI (BHI + 4% glycine + 0.5 M sucrose) and incubated overnight (14–16 hours). From this point onward, cells were handled gently to preserve viability. Cells were harvested by centrifugation at 1,000 × g for 15 min at room temperature (RT), gently resuspended in 100 mL GS-BHI (as described below), and incubated without shaking at 37°C for 1.5 hours. Cells were re-pelleted (1,000 × g, 15 min, 4°C), gently washed with 35–40 mL ice-cold electroporation solution, and kept on ice. The harvested cells were then resuspended in 1–2 mL electroporation solution and homogenized using a sterile serological pipette. Aliquots (100 µL) were prepared in pre-chilled Eppendorf tubes and stored at –80°C. For electroporation, thawed 100 µL aliquots were mixed with 2-4 µg DNA and transferred to 0.2 cm gap cuvettes. Electroporation was carried out at 25 µF, 400 Ω, and 2.5 kV. Immediately after, 0.9 mL of 1X SBHI was added, and cells were recovered for 3 h at 37°C without shaking before plating on selective BHI agar. For recombineering, the entire transformation mixture was pelleted (7,000 × g, 2 min) and plated to maximize recovery.
Generation of the ERV165 ΔsagA clean deletion mutant (ΔsagA)
The gene-editing CRISPR-cas12a protocol from Chua et al.29 was modified to account for the robustness of the ΔsagA mutant, which made deletion challenging.
a.Generation of the ΔsagA mutant
Briefly, pPK99 plasmid was electroporated into VREfm ERV165 electrocompetent cells. A few transformants were inoculated into 5 mL of BHI broth supplemented with 250 µg/mL gentamycin and grown at 37°C with shaking for two days. Sub-culture (1:1,000 dilution) into fresh BHI broth supplemented with 250 µg/mL gentamycin and incubated under the same conditions for an additional two days. Subsequently, 1 µL of the culture was streaked onto BHI agar plates supplemented with 250 µg/mL gentamycin and 250 ng/mL anhydrotetracycline (ahTC) and incubated at 37°C for 3 to 5 days. Smaller single colonies were picked and grown in BHI broth supplemented with 250 µg/mL gentamycin and 250 ng/mL ahTC. Since DsagA mutant exhibits a sedimentation phenotype, the sedimented cells were selectively taken and re-inoculated into fresh BHI broth with the same supplements. This enrichment step was repeated twice to enhance the recovery of ΔsagA mutants.
b.Morphology-based screening and validation of ΔsagA mutants
Based on our previous studies 19, the E. faecium Com15 ΔsagA mutant is known to exhibit growth defects. Therefore, numerous sick colonies were screened but were ultimately found to be false positives. During this study, we identified an alternative screening method for ΔsagA mutants, as they exhibit a distinct colony morphology that can be visualized using light microscopy (Fig. 1d). To identify such mutants, BHI plates were continuously monitored for the appearance of smaller and sick colonies. Numerous BHI plates were examined under a light microscope using differential interference contrast (DIC) imaging with a 10× objective lens, and colonies were extensively analyzed for the characteristic ΔsagA mutant morphology. Colonies with a smaller size and a rough halo texture were selected for further screening by colony PCR. However, DNA sequencing analysis revealed a high frequency of false positives. In these cases, the sagA open reading frame (ORF) remained intact, but the protospacer was disrupted,likely due to CRISPR-Cas12a-mediated double-stranded break and repair events, leading to a phenotype that mimicked the sagA mutant. Despite these challenges, we successfully isolated a single, confirmed colony that carried a clean deletion of sagA. The whole genome sequence analysis was done to check for background mutations (Extended Data Tables 3 and 4).
Generation of ERV165 ΔsagA chromosomal complementation strain (ΔsagA::sagA)
The complementation strain was constructed using the pPK158 plasmid, following a method similar to that described above for generating the ΔsagA mutant. Whole-genome sequencing analysis confirmed that the observed growth defect of ΔsagA mutant was not due to background mutations (Extended Data Tables 5 and 6).
Generation of the ERV165 pgh2, pgh3 and pgh4 clean deletion mutants (Δpgh2, Δpgh3 Δpgh4)
Strains were constructed using the pPK156, pPK107 or pPK154 plasmids respectively, following a method similar to that described above for generating the ΔsagA mutant.
Adaptive laboratory evolution
VREfm cultures grown overnight from single colonies were subcultured in fresh BHI supplemented with vancomycin (50 µg/mL) and were grown aerobically at 37 °C at 200 RPM shaking overnight. The next day, the cultures were passaged in fresh BHI supplemented with vancomycin (50 µg/mL) to reach OD600∼0.1 and were grown aerobically at 37 °C at 200 RPM shaking overnight. The passaging was repeated for 14 days and the vancomycin susceptibility was determined following CLSI guidelines.
Recombinant protein expression and purification
Truncated SagA_NlpC/P60 proteins were expressed and purified from E. coli BL21-RIL (DE3) as previously described.16,36 pET-21a(+) plasmids containing the truncated SagA_NlpC/P60 genes with a C-terminal His6 tag were transformed into BL21-CodonPlus (DE3)-RIL E. coli (Agilent 230245) according to the manufacturer’s protocol and maintained in LB broth supplemented with 100 µg mL-1 ampicillin and 25 µg mL-1 chloramphenicol. Overnight bacterial culture was subcultured in 1 L fresh BHI supplemented with appropriate antibiotics, grown until OD600∼0.5, induced with 1 mM isopropyl-D-thiogalactopyranoside, and additionally grown for 2 hours at 37 °C and 200 RPM shaking. Bacterial cells were collected by centrifugation at 4 °C, resuspended in 20 mL lysis buffer (20 mM Tris-HCl, pH 8.0, 150 mM NaCl, 0.1% SDS, 0.025 U/mL benzonase, and 1× protease inhibitor cocktail). After 15 min of sonication followed by centrifugation at 18,000 × g for 30 min at 4°C, the supernatant containing the soluble target protein was collected and loaded on 2 mL of Ni-NTA agarose (Invitrogen) equilibrated with the binding buffer (PBS). The protein-bound resin was washed with 20 mM and 40 mM imidazole sequentially, then SagA_NlpC/P60 protein was eluted with 300 mM imidazole. Semi-purified protein was dialyzed into PBS buffer at 4°C overnight using 10K MWCO Slide-A-Lyzer MINI dialysis devices (Thermo Fisher Scientific). Protein was further purified on a ENrich™ SEC 650 column (Bio-Rad) pre-equilibrated with PBS using NGC chromatography system (Bio-Rad). Fractions containing the target protein were combined and concentrated. Protein concentration was estimated by Pierce™ BCA Protein Assay (ThermoFisher) and protein was stored at −80°C.
Fluorescence polarization (FP) cysteine activity assay
The assay is based on competitive reaction of a tested compound with broad-spectrum cysteine-reactive probe tetramethylrhodamine-5-iodoacetamide (Anaspec) and was adapted from 63. 16 μL of recombinant SagA_NlpC/P60 (0.5 μM) in PBS was added to black 384-well plate. Tested compound (50 μM) were then added to corresponding wells using Bravo instrument (Agilent). The plate was covered with aluminum seal and incubated for 1 hour at room temperature with agitation, followed by treatment with tetramethylrhodamine-5-iodoacetamide (20 nM) for 30 min at 37°C shielded from light with agitation. Fluorescence polarization was measured using EnVision plate reader 2105 (Perkin Elmer) with BODIPY TMR FP filter set and calculated as follows:
Polarization (mP) = 1000 * (S – G*P)/(S + G*P), where S and P are the measured results with the S and P emission filters respectively. G is a correction factor (effect of emission filter transmission variations, differences in the emission light paths and sample viscosity).
Inhibition (%) = 100-(mPsample-mPnegative)/(mPPositive-mPnegative), where mPsample is FP value of SagA_NlpC/P60 treated with a corresponding testing compound, mPpositive is FP value of SagA_NlpC/P60 treated with DMSO, mPnegative is FP value of SagA_NlpC/P60 (C433A, inactive mutant) treated with DMSO. Calculated inhibition > 100% was assigned as 100%.
Gel-based cysteine activity assay
The assay was adapted from 63. For gel-based SagA activity assay, the conditions were similar to FP assay, except for treatment with tetramethylrhodamine-5-iodoacetamide (50 nM). The reaction was quenched with 4 × Laemmli buffer, heat-inactivated for 5 min at 90 °C, separated by SDS-PAGE. The gel was visualized using ChemiDoc MP imaging system (Bio-Rad). Relative labeling was determined by extracting fluorescence intensity of the band and compared to DMSO control. For analysis of bacterial proteome, after treatment with pghi-4, 100 μg of precipitated secreted proteins from supernatants or bacterial proteome from cell pellets lysed in PBS were treated with N-5-hexyn-1-yl-2-iodoacetamide (IA-alk, 100 μM) for 1 hour at room temperature with agitation, followed by 1 hour treatment with click mixture: 1 mM tris(2-carboxyethyl)phosphine hydrochloride (TCEP), 1 mM CuSO4, 0.1 mM tris[(1-benzyl-1H-1,2,3-triazol-4-yl)methyl]amine (TBTA), and 100 μM rhodamine-azide. The reactions were quenched by addition 4 × Laemmli buffer and proteins were separated by SDS-PAGE. In-gel fluorescence was detected by ChemiDoc MP imaging system (Bio-Rad). Relative labeling was determined by extracting fluorescence intensity of the SagA band and compared to DMSO control using Image Lab software (Bio-Rad). Protein loading was analyzed by Coomassie blue staining for recombinant SagA and by α-SagA Western Blot for bacterial proteome.
Biotin pull-down protocol was adapted from 64. For biotin pull-down experiments from bacterial lysates, 250 μg of each total cell lysates in PBS were treated IA-alk (100 μM) for 1 hour at room temperature with agitation, followed by 1 hour treatment with click mixture: 1 mM tris(2-carboxyethyl)phosphine hydrochloride (TCEP), 1 mM CuSO4, 0.1 mM tris[(1-benzyl-1H-1,2,3-triazol-4-yl)methyl]amine (TBTA), and 100 μM biotin-azide. Proteins were precipitated with 4 × volume cold methanol. Precipitated proteins were pelleted by centrifugation (18,000 × g, 4°C, 10 min), sequentially washed with cold methanol and centrifuged 3 times, followed by drying in SpeedVac. Resulted pellets were resuspended in 100 μL 4% SDS in PBS with bath sonication. 2.5% of solution was used as input (protein loading control). Total volume of incubated with 20 μL PBS-T-washed High Capacity NeutrAvidin agarose (Pierce) (500 μL PBS-T-washed twice, 2,500 × g for 60 s) at room temperature for 1 hour with end-to-end rotation. The agarose was then washed with 500 μL PBS (1% SDS) 3 times, 500 μL 1 M Urea in PBS three times, and 500 μL PBS three times. Samples were boiled with 2 × Laemmli buffer 95°C for 5 min and analyzed by western blot.
For biotin pull-down experiments from bacterial supernatants, proteins were precipitated from supernatants with 4 × volume cold methanol. Precipitated proteins were pelleted by centrifugation (18,000 × g, 4°C, 10 min). Resulted pellets were resuspended in PBS and protein concentration was estimated by BCA assay with BCA Protein Assay Kit (Thermo). 250 μg of each total proteins were processed further as described above.
Competitive cysteine-directed chemoproteomic analysis
The protocol was adapted from 65. VREfm ERV165 were grown from overnight culture in 15 mL of fresh BHI till OD600∼0.6. Bacteria were then centrifuged and resuspended in 1 mL of BHI (OD600∼9), followed by incubation with pghi-4 (10, 25 or 50 μM) for 1 hour at 37°C and 200 RPM shaking. Bacterial pellets were collected by centrifugation (4,800 × g for 10 min), washed twice with PBS and immediately processed or stored at −80 °C. Bacterial pellets were resuspended in 360 μL PBS and lysed with 0.1 mm glass beads (BioSpec) using FastPrep system (MP Biomedicals, settings: 6 m/s, 2 cycles, 45 s for each cycle). Proteins were quantified by Pierce™ BCA Protein Assay (ThermoFisher) and normalized to 2 mg/mL. 1 mg/0.5 mL of bacterial proteome was incubated with iodoacetamide-desthiobiotin (IA-DTB, 100 μM) for 1 hour at room temperature agitating. Bacterial proteins were precipitated by cold methanol (600 μL), chloroform (200 μL), and water (100 μL), followed by vortexing and centrifugation at 16,000 × g for 10 min at 4°C. Liquids were carefully aspirated, proteins were washed with cold methanol, pelleted by centrifugation (16,000 × g for 10 min at 4°C) and air-dried for 5 min. Protein pellets were resuspended in 90 μL of buffer (9 M urea, 10 mM DTT, 50 mM triethylammonium bicarbonate (TEAB) pH 8.5), heated at 65°C for 20 min, followed by treatment with iodoacetamide (50 mM) for 30 min at 37°C. The insoluble residues were pelleted by centrifugation and clear solutions were sonicated. Samples were diluted with 300 μL TEAB buffer and trypsinized (5 μL of 0.4 μg/μL trypsin in trypsin buffer supplemented with 25 mM CaCl2) overnight at 37°C. Samples were treated with 400 µL of wash buffer (50 mM TEAB, 150 mM NaCl, 0.2% NP-40) containing 50 µL of streptavidin agarose to the peptide samples, followed by rotation at room temp for 2 hours. Suspensions were briefly centrifuged, and beads-containing suspensions were loaded on BioSpin columns. The beads were sequentially washed with 3 × 1 mL wash buffer, 3 × 1 mL PBS, 3 × 1 mL MiliQ water. Beads-bound peptides were eluted by addition of 2 × 200 µL of 80% acetonitrile (0.1% formic acid) and the eluate was concentrated by SpeedVac. Peptides were resuspended in 70 μL EPPS buffer (200 mM, pH 8.0), supplemented with 30% acetonitrile, vortexed and sonicated for 5 min. Peptides were tandem mass tag (TMT)-labeled by adding 3 μL of 10 mg/mL TMT10plex tag and incubating for 1 hour at room temperature. The reaction was quenched by sequential addition of hydroxylamine (3 μL of a 5% aqueous solution, 15 min at room temperature) and formic acid (5 μL), followed by concentration using SpeedVac. Samples were resuspended in 500 μL Velos buffer A (95% water, 5% acetonitrile, 0.1% formic acid), acidified with 20 μL of formic acid, desalted (Sep-Pak C18 Cartridge) and concentrated by SpeedVac. Desalted and concentrated samples were redissolved in 500 μL Velos buffer A and HPLC-fractionated. Fractionation and TMT LC-MS analysis was followed as previously described 65.
Data were processed as previously described65 with small modifications regarding the proteome dataset. Raw files were uploaded to the Integrated Proteomics Pipeline (IP2, version 6.0.2) available at http://ip2.scripps.edu/ip2/mainMenu.html, and MS2 and MS3 files were extracted from the raw files using RAW Converter (version 1.1.0.22, available at http://fields.scripps.edu/rawconv/) and searched using the ProLuCID algorithm using the Enterococcus faecium ERV165 UniProt database (UP000005678, released 2012-04). Cysteine residues were searched with a static modification for carboxyamidomethylation (+57.02146 Da). A dynamic modification for IA-DTB labeling (+398.25292 Da) was included with a maximum number of two differential modifications per peptide. N termini and lysine residues were also searched with a static modification corresponding to the TMT tag (+229.1629 Da). Peptides were required to be at least 6 amino acids long. ProLuCID data were filtered through DTASelect (version 2.0) to achieve a spectrum false-positive rate below 1%. We excluded nonunique peptides and required at least one tryptic cleavage site and two peptides per protein. The MS3-based peptide quantification was performed with reporter ion mass tolerance set to 20 ppm with the IP2. Pghi-4 cysteine-directed activity was calculated as competition (%) of a cysteine site relative to DMSO treatment (0%) when IA-DTB fully occupies accessible cysteines. The full list of identified proteins is in Supplementary Data 1.
Peptidoglycan isolation
The peptidoglycan isolation was followed as previously described.36 Overnight VREfm cultures were subcultured in fresh BHI supplemented with or without vancomycin, pghi-4 or in combination and were grown till OD600∼0.8-1. Bacterial cell pellets were collected by centrifugation and lysed in 0.25% SDS solution in 0.1 M Tris-HCl, pH 6.8 and boiling the suspension for 20 min at 100°C. The insoluble bacterial cell wall was collected by centrifugation, washed with distilled water to remove SDS. The cell wall was next sonicated for 30 min in distilled water and treated with benzonase followed by trypsin digestion. Then, insoluble cell wall was recovered by centrifugation (16,000 × g, 10 min, 4°C), and washed with distilled water. Next, the cell wall was treated with 1 M HCl for 4 hours at 37°C and 200 RPM shaking. The insoluble material was collected by centrifugation (16,000 × g, 10 min) and washed with distilled water until the pH was 5-6. Purified insoluble peptidoglycan was digested with mutanolysin from Streptomyces globisporus (Sigma, 10 KU/mL of mutanolysin in MiliQ H2O) in 10 mM sodium phosphate buffer, pH 4.9 for 16 hours at 37°C shaking. The mutanolysin was heat-inactivated and resulting soluble peptidoglycan was used in gel-based peptidoglycan hydrolase activity assay or LC-MS analysis.
Gel-based peptidoglycan hydrolase activity assay
The peptidoglycan hydrolase activity assay is based on in-gel analysis of peptidoglycan fragments labeled with 8-aminonaphthalene-1,3,6-trisulfonic acid (ANTS). The assay was followed as previously described36 with some modifications. 10 μM recombinant SagA_NlpC/P60 (or other NlpC/P60 peptidoglycan hydrolase) was incubated with 250 μg of the mutanolysin-digested peptidoglycan in 50 mM Bis-Tris, pH 5.5 overnight at 37 °C and 220 RPM shaking. The enzymatic activity was heat-inactivated at 100°C for 5 min. The samples were centrifuged at 16,000 x g for 5 min and supernatants were transferred and concentrated. The concentrated samples were resuspended and incubated with 5 μL of each reagent (0.2M ANTS in water supplemented with 15% acetic acid and 1M NaBH3CN in DMSO) overnight at 37 °C, 220 RPM shaking and protected from light. The reaction mixtures were diluted with 10 μL 50% glycerol (v/v) and ANTS-labeled peptidoglycan fragments were separated by Native PAGE on 4-20% Criterion TGX precast gels (Bio-Rad) ran at 100 V for 30 min, and visualized by ChemiDoc MP imaging system (Bio-Rad) and SYBR-Safe settings. Inhibition was determined by extracting fluorescence intensity of the ANTS-labelled enzymatic product (GlcNAc-MurNAc dipeptide) and compared to DMSO control.
Computational covalent docking
The docking receptor file was prepared from PDB: 6B8C.16 All crystallographic waters and alternative residue positions were removed with PyMOL (https://pymol.org/). The receptor was then protonated using Reduce66 and prepared for docking using Meeko (https://github.com/forlilab/Meeko) to assign atomtypes and Gasteiger partial charges and convert to a PDBQT file67. Gridmaps were calculated with AutoGrid4 with a box size of 16 Å, 18 Å, 27 Å (0.375 Å grid spacing) and with the box center at 101 Å, 80 Å, 140.5 Å. His 506 was also designated to be in the HID (neutral, δ-nitrogen protonated) form. The Cys 443 C⍺ and sidechain were also removed.
Covalent docking as performed using the “flexible side chain” model described by Bianco et al.68. 2D line drawings of pghi-4 adducts as either the (a) thioenol sulfonyl fluoride (Michael addition product) and (b) vinylchloride thiosulfonate (sulfonyl fluoride exchange product) were prepared in ChemDraw. The 3D conformation of the vinylchloride thiosulfonate was then prepared with MolScrubber (https://github.com/forlilab/molscrub) and checked to ensure the optimal cyclopentane conformation was generated with the ether in equatorial position. For the thioenol sulfonyl fluoride, a torsional energy scan of the phenyl-thioenol torsion was performed using Gaussian16 with 10-degree incremental steps between energy calculations. Minima were found for torsions of approximately 45° and −143° between the thioenol sulfur and the phenyl carbon adjacent to that bearing the methyl ether. Both thioenol sulfonyl fluoride geometries (with the phenyl-thioenol torsion held as non-rotatable) and the single vinylchloride thiosulfonate were prepared for docking using Meeko to add Gasteiger partial charges and convert to PDBQTs67. All three models were then docked as flexible side chains with AutoDock-GPU 69, keeping the top scoring pose for further analysis. The two docked thioenol sulfonyl fluoride models were evaluated visually, with the −143° model selected as the most reasonable docked conformation. Models were additionally corroborated using reactive docking70 to confirm the state prior to reaction could be accommodated.
Intact protein analysis by mass spectrometry
10 μM recombinant SagA_NlpC/P60 (or C433A inactive control) was incubated with 50 μM pghi-4 (or inactive pghi-6) in 50 mM Tris-Cl, pH 7.6, 150 mM NaCl buffer for 1 hour (or 16 hours) at room temperature and shaking at 600 RPM. The mixtures were analyzed by ESI-TOF mass spectrometry.
Protein crystallization
Crystallization efforts to obtain a SagA-pghi-4 complex were extensive and systematic using both co-crystallization and soaking methodologies. High-throughput crystallization screening was performed using the automated Rigaku CrystalMation system at The Scripps Research Institute with the JCSG Core Suite (QIAGEN), comprising 384 distinct conditions across four 96-well plates. Robust diffraction-quality crystals were obtained reproducibly for SagA under multiple conditions for both apo and ligand-incubated samples. For co-crystallization, the inhibitor was added at 5× and 10× molar excess to purified SagA at protein concentrations of 10, 15, and 18 mg/mL, followed by incubation for approximately 30 minutes at either room temperature or 4 °C prior to crystallization setup. In parallel, soaking experiments were performed by incubating pre-grown apo-SagA crystals in pghi-4-containing solutions for varying time intervals before flash freezing. Crystals from both co-crystallization and soaking experiments diffracted strongly to sub-2.0 Å resolution under multiple conditions, including 0.1 M sodium citrate–citric acid (pH 5.6) with ammonium or lithium sulfate, and 0.2 M magnesium acetate with 20% (w/v) PEG 3350. Data were processed and refined using standard crystallographic pipelines16. However, in all cases, molecular replacement and subsequent refinement yielded apo SagA structures, with no interpretable ligand-associated electron density observed in unbiased difference or omit maps. These results suggest that, under the crystallization conditions tested, the inhibitor either does not bind SagA with sufficient occupancy to be captured crystallographically, or is destabilized, or displaced during crystal growth or soaking. Alternatively, other factors such as transient binding, competition with crystallization components, or conformational heterogeneity cannot be excluded. Despite extensive crystallization efforts, a SagA-pghi-4 complex structure was not obtained.
Bacterial growth defect and antibiotic susceptibility assay
Starting cultures were grown from single bacterial colonies overnight at 37°C at 200 RPM shaking in Brain Heart Infusion (BHI) broth. Next, the starting culture were diluted to OD600 ∼0.1 in fresh BHI in a sterile 96-well plate and supplemented with vancomycin alone, SagA inhibitor alone or combination of vancomycin and SagA inhibitor at indicated concentrations. The OD600 values were measured at 37°C with continuous orbital shaking using BioTek Cytation 5 plate reader (Agilent). For inhibition heat-maps, inhibition was calculated as percentage of OD600 values of treatment groups over vehicle control at the single time point. Following Clinical & Laboratory Standards Institute (CLSI) guidelines, vancomycin susceptibilities of ERV165 WT and generated mitants were also tested in Mueller Hinton Broth (MHB). Minimal inhibitory concentrations (MICs) were determined as the lowest concentration that inhibited the visible growth.
LC-MS-based peptidoglycan analysis
The analysis was followed as we previously described.19,36 The mutanolysin-digested peptidoglycan was treated with sodium borohydride in 0.25 M boric acid (pH 9) for 1 hour at room temperature, quenched with orthophosphoric acid, and pH adjusted to 2-3. The samples were centrifuged at 20,000 × g for 10 minutes. Then, the reduced peptidoglycan was analyzed by 1290 Infinity II LC/MSD system (Agilent technologies) using Poroshell 120 EC-C18 column (3 × 150 mm, 2.7 μm). Samples were run at flow rate 0.5 mL/min in mobile phase (A: water, 0.1% formic acid) and an eluent (B: acetonitrile, 0.1% formic acid) using following gradient: 0-5 min: 2% B, 5-65 min: 2-10% B. All solvents were HPLC grade. The absorbance of the eluting peaks was detected at 205 nm. Masses of peaks were detected with MSD API-ES Scan mode (m/z = 200-2,500) (Extended Data Table 12). For quantification of relative abundance of muropeptides, the area under the curve of assigned individual peak from chromatograms was integrated and percentage of individual peak was calculated relative to all assigned peaks.
Western blot analysis
Western blot analysis was performed as described previously.16,19 Overnight VREfm cultures were sub-cultured in fresh BHI to OD600∼0.1 and were grown overnight at 37°C and shaking at 200 RPM. The cultures were centrifuged at 4,800 x g for 10 min, cell pellets and supernatants were separated. Cell pellets were washed with PBS and lysed in PBS with 0.1 mm glass beads (BioSpec) using FastPrep system (MP Biomedicals, settings: 6 m/s, 2 cycles, 45 s for each cycle). Proteins were quantified by Pierce™ BCA Protein Assay (ThermoFisher) and separated by SDS-PAGE on 4-20% Criterion TGX precast gels (Bio-Rad), then transferred to nitrocellulose membrane. The membrane was blocked in 0.1% TBST supplemented with 5% milk for 1 hour at room temperature with agitation, incubated with primary antibody (rabbit anti-SagA polyclonal sera, diluted 1:50000 in 0.1% TBST supplemented with 5% milk) overnight at 4°C with agitation. After 2 washes (2 min each) the membrane was incubated with secondary antibody (goat anti-rabbit antibody, HRP-conjugate, diluted 1:20,000 in 0.1% TBST supplemented with 5% milk) for 1 h at room temperature with agitation. Membranes were washed with 0.1% TBST three times (10 min each) at room temperature with agitation. Blots were developed using Clarity Western ECL substrate (Bio-Rad) and imaged using a ChemiDoc MP imaging system (Bio-Rad).
Fluorescence microscopy
VREfm were grown to OD600 ∼ 0.4 with or without pghi-4 (50 μM) in BHI and sequentially labeled with 0.5 mM HADA (Tocris Bioscience) for 30 min and 1 μg/mL Vancomycin-BODIPY (Invitrogen) for 15 min at 220 RPM shaking, 37°C and protected from light. For Bocillin staining, VREfm were incubated with Bocillin (Invitrogen) at concentration 5 μM for 30 min at 220 RPM shaking, 37°C and protected from light. The cultures were centrifuged at 4,800 × g for 10 min. The cells were washed with PBS twice and fixed with 1% formaldehyde in PBS for 10 min, followed by washing and resuspending in PBS. To image cells, an aliquot of cell suspensions was transferred to the surface of a 2% (w/v) agarose pad prepared in PBS, covered with a glass coverslip, and imaged with fluorescence microscope (Nikon Ti2-E Inverted) using DAPI and GFP filters. For single-cell fluorescence analysis, fluorescence intensities from single cells normalized to similar region of interest were extracted from images. The images were processed using Icy open source imaging software71.
Cryo-electron tomography
The protocol was adapted from 19 with some modifications. VREfm ERV165 strains or ERV165 strain treated with vancomycin (5 μg/mL), pghi-4 (50 μM) or in combination used in the cryo-ET experiments were grown overnight at 37 °C in BHI broth. Fresh cultures were prepared from a 1:50 dilution of the overnight culture and then grown at 37 °C to early log phase. The culture was centrifuged at 4,800 × g for 10 min. The pellet was resuspended with growth media supplemented with 5% glycerol to OD600 ∼ 3. Next, 5 µL of bacterial samples were deposited onto freshly glow-discharged (Pelco easiGlow; 25 s glow at 15mA) Quantifoil R2/1 copper 200 mesh grids for 1 min, back-side blotted with filter paper (Whatman Grade 1 filter paper), and frozen in liquid ethane using a gravity-driven homemade plunger apparatus (inside a 4°C cold room with a ≥95% relative humidity). The samples were frozen using a Vitrobot Mark IV (Thermo Fisher Scientific) in liquid ethane/propane mixture. The Vitrobot was set to 22°C at 90% humidity, and manually back-side blotted. The vitrified grids were later clipped with Cryo-FIB autogrids (Thermo Fisher Scientific) prior to milling.
Cryo-FIB milling was performed using an Aquilos2 dual-beam cryo-FIB/SEM instrument (Thermo Fisher Scientific). Vitrified samples were sputter-coated with metallic platinum for 15 s, followed by a 30 s coating with organometallic platinum, and then sputter-coated again with metallic platinum for 15 s to prevent drift during milling. Targets were selected and milled at an 8° angle using MAPS and AutoTEM software, respectively (Thermo Fisher Scientific). The milling template performed rough milling with a current of 0.30 nA, followed by medium milling at 0.1 nA. Thinning was conducted with a current of 50 pA. Automated milling produced lamellae with a thickness of ∼300 nm, which were then manually polished to <200 nm using a 30 pA current. A final 20 s metallic platinum coating was applied to facilitate bead-like fiducial inclusions for tilt series alignment.
Cryo-lamellae were imaged using a 300 keV Titan Krios microscope (Thermo Fisher Scientific) equipped with a field emission gun, an energy filter, and a direct electron detector (Gatan K3). An energy filter with a slit width of 20 eV was used during data acquisition. The SerialEM72 package with PACEtomo73 scripts was used to collect 35 image stacks at tilt angles ranging from +51° to −51° in 3° increments using a dose-symmetric scheme with a cumulative dose of ∼105 e⁻/Ų. Data were collected at a magnification corresponding to 2.64 Å/pixel and a nominal defocus of ∼ −5 µm.
Image stacks containing 10 frames were motion-corrected using MotionCor274, then assembled into drift-corrected stacks using IMOD75. These were aligned and reconstructed into tomograms using IMOD marker-based alignment. Tomograms were binned 4×, resulting in a final pixel size of 10.55 Å/pixel. Missing wedge artifacts were corrected using IsoNet76, a deep learning-based software. Segmentation of cell membranes was performed using MemBrain77. ColabSeg78 was used to isolate segmented membranes of interest. A MATLAB script79 was used to pick subtomograms with defined Euler angles perpendicular to the membrane, spaced ∼30 nm apart. Segmented membranes and subtomogram picks were validated in UCSF Chimera 80. Subtomogram averaging was performed using the Dynamo 81 software package. Subtomograms were extracted with a box size of 120 pixels. Five iterations of averaging with minimal translational and angular searches were conducted to generate averages. To generate density profiles, IMOD drawing tools were used to draw a line through the subtomogram average. Density values along the line were extracted, plotted and used for cell wall components assignment. Manual measurements of the cell envelope (CE) were performed on the apical and septal regions of cells using IMOD’s measurement tool. Workflow of subtomogram averaging and analysis is summarized in Extended Data Fig. 2.
Macrophage infection
The protocol was adapted from 52. Macrophages (RAW264.7, ATCC TIB-71, or THP-1, ATCC TIB-202, differentiated with phorbol 12-myristate 13-acetate) were plated in 96-well plate in the growth media (DMEM or RPMI supplemented with 4.5 g/L glucose, 2 mM L-glutamine, 10% FBS) at the cell density 3×104 cells/ well. Cells were left to adhere and grow overnight at 37°C in humidified atmosphere with 5% CO2. VREfm (ERV165 strain) were grown to OD∼0.6 in BHI by sub-culturing starting cultures grown overnight from a single colony. Bacterial cultures were centrifuged, and pellets were washed with sterile PBS twice, followed by resuspension in DMEM or RPMI with or without pghi-4 (25, 50 or 100 μM), vancomycin (100 μg/mL) or in combination. Macrophages were incubated with bacterial suspensions for 3 hours at 37°C, 5% CO2. Next, bacterial suspensions were removed, and macrophages were washed with PBS three times. Macrophages were incubated with 1% Triton X-100 (Sigma-Aldrich) in PBS for 5 mins at room temperature to lyse cells for colony-forming units (CFU) analysis. Lysates of macrophages were immediately plated on BHI agar, incubated overnight at 37°C and intracellular VREfm were enumerated.
Mouse peritonitis infection
The protocol was adapted from32. Specific pathogen-free C57BL/6 (B6,000664) mice were obtained from Scripps Rodent Breeding. Mice were fed with gamma-irradiated chow (LabDiet, 5053) and sterile drinking water ad libitum. Animal care and experiments were conducted in accordance with NIH guidelines and approved by the Institutional Animal Care and Use Committee at Scripps Research. For evaluating vancomycin susceptibility of ΔsagA in vivo, bacterial cultures of VREfm strains (WT, ΔsagA or ΔsagA::sagA) grown in BHI (OD600 ∼ 0.6) were washed with PBS and resuspended in PBS at 5×108 CFU/mL. Next, 6-8 weeks female mice were infected with 109 VREfm CFU intraperitoneally. After 30 minutes, PBS or vancomycin (Van, 100 mg/kg) was administered subcutaneously. The second dose of treatment was administered 24-hour post-infection. VREfm infection was monitored by mice weight loss compared to initial weight (before infection) over 48 hours post-infection. Mice were euthanized once the weight reached 80% of the initial weight. Spleen and liver were collected, homogenized in sterile PBS for colony-forming units (CFU) analysis. Homogenates were plated on HiCrome™ selective Enterococcus faecium agar plates (HIMEDIA 1580) with Enterococcus faecium selective supplement (FD226, HIMEDIA), incubated at 37°C overnight and viable colonies were enumerated.
For evaluating combination therapy, bacterial cultures of VREfm (ERV165 strain) grown in BHI (OD600 ∼ 0.6) were washed with PBS and resuspended in PBS (0.25% carboxymethyl cellulose, CMC) at 5×108 CFU/mL. Next, 6-8 weeks female mice were co-injected intraperitoneally with 0.2 mL of bacterial suspension (109 CFU) and treatment: PBS (0.25% CMC) alone or supplemented with vancomycin (100 mg/kg), pghi-4 (25 mg/kg) or in combination. The second dose of treatment (PBS 0.25% CMC alone or supplemented with vancomycin (100 mg/kg), pghi-4 (25 mg/kg) or in combination was administered 24-hour post-infection. VREfm infection was monitored by mice weight loss compared to initial weight (before infection) over 48 hours post-infection. Mice were euthanized once the weight reached 80% of the initial weight. Spleen and liver were collected and processed as described above.
Supporting information
Data availability
The mass spectrometry proteomics raw data have been deposited to the ProteomeXchange Consortium via the PRIDE partner repository with the dataset identifier PXD075040. All other data supporting the findings of this study are available within the article and its supplementary information files and from the corresponding author on reasonable request.
Acknowledgements
This project was funded by the National Institutes of Health (NIH) R21AT012958 grant and Scripps Research start-up funds to H.C.H. J.E.M. thanks Cold Spring Harbor Laboratory for developmental funds from the NCI Cancer Center Support Grant (5P30CA045508), the Australian Research Council (ARC) for a Future Fellowship (FT170100156), and the F.M. Kirby Foundation. We thank Juliel Espinosa for providing constructs for recombinant expression of PGH2 and PGH3. We thank Francisco Martínez-Peña, Luke Lairson, Kayla Nutsch, Caroline Stanton and Michael Bollong for assisting with automated liquid handlers and EnVision plate readers. We thank Kathryn Spenser and Scott Henderson for assistance at the Scripps Microscopy Core, K. Barry Sharpless and the Hang laboratory members for their feedback.
Contributions
K.T.F. and H.C.H. conceived the project and planned initial experiments. P.K.C. generated ERV165 isogenic deletion and their derivative strains. K.T.F. and P.K.C. characterized all generated strains. C.J.S, S.K. generated screening library of sulfonyl fluorides. J.E.M and C.J.S. both designed, and C.J.S. generated the lead sulfonyl fluorides. Z.W., J.H. resynthesized pghi-4. K.T.F. developed assays, conducted screening and identified hits. K.T.F. performed all biochemical and microbiological characterizations of identified hits. K.T.F. and Y.X. conducted competitive cysteine-directed chemoproteomics. A.H.-H., M.H. performed computational docking. K.T.F, B.S. and D.P. performed cryoET experiments. K.T.F. and S.B. attempted X-ray crystallography studies. A.M.T. performed phylogenetic analysis. D.V.T. provided VREfm clinical isolates. H.C.H, J.E.M., D.W.W, D.P., I.A.W., S.F., B.F.C. supervised experiments. K.T.F. conducted all animal studies. K.T.F., H.C.H wrote the manuscript, which was edited by all the other authors. All authors approved the manuscript before submission.
Ethics declarations
Authors declare no competing interests.