Microbiological Characterization of VNRX-5236, a Broad-Spectrum β-Lactamase Inhibitor for Rescue of the Orally Bioavailable Cephalosporin Ceftibuten as a Carbapenem-Sparing Agent against Strains of Enterobacterales Expressing Extended-Spectrum β-Lactamases and Serine Carbapenemases

ABSTRACT There is an urgent need for oral agents to combat resistant Gram-negative pathogens. Here, we describe the characterization of VNRX-5236, a broad-spectrum boronic acid β-lactamase inhibitor (BLI), and its orally bioavailable etzadroxil prodrug, VNRX-7145. VNRX-7145 is being developed in combination with ceftibuten, an oral cephalosporin, to combat strains of Enterobacterales expressing extended-spectrum β-lactamases (ESBLs) and serine carbapenemases. VNRX-5236 is a reversible covalent inhibitor of serine β-lactamases, with inactivation efficiencies on the order of 104 M−1 · sec−1, and prolonged active site residence times (t1/2, 5 to 46 min). The spectrum of inhibition includes Ambler class A ESBLs, class C cephalosporinases, and class A and D carbapenemases (KPC and OXA-48, respectively). Rescue of ceftibuten by VNRX-5236 (fixed at 4 μg/ml) in isogenic strains of Escherichia coli expressing class A, C, or D β-lactamases demonstrated an expanded spectrum of activity relative to oral comparators, including investigational penems, sulopenem, and tebipenem. VNRX-5236 rescued ceftibuten activity in clinical isolates of Enterobacterales expressing ESBLs (MIC90, 0.25 μg/ml), KPCs (MIC90, 1 μg/ml), class C cephalosporinases (MIC90, 1 μg/ml), and OXA-48-type carbapenemases (MIC90, 1 μg/ml). Frequency of resistance studies demonstrated a low propensity for recovery of resistant variants at 4× the MIC of the ceftibuten/VNRX-5236 combination. In vivo, whereas ceftibuten alone was ineffective (50% effective dose [ED50], >128 mg/kg), ceftibuten/VNRX-7145 administered orally protected mice from lethal septicemia caused by Klebsiella pneumoniae producing KPC carbapenemase (ED50, 12.9 mg/kg). The data demonstrate potent, broad-spectrum rescue of ceftibuten activity by VNRX-5236 in clinical isolates of cephalosporin-resistant and carbapenem-resistant Enterobacterales.

VNRX-5236 exhibited potent inhibition of CTX-M-15 and SHV-5 (class A ESBLs), KPC-2 (class A carbapenemase), and P99AmpC (class C cephalosporinase), with inhibitor potency (K i ) values ranging from 0.01 to 0.11 mM (Table 2). Like other boronic acid-containing BLIs, inhibition of serine b-lactamases by VNRX-5236 is best described by a two-step model, in which a noncovalent complex forms, followed by a reversible covalent bond with the active site serine residue. Thus, second-order rate constants (k 2 /K i ) for covalent bond formation were determined for the aforementioned enzymes, indicating efficient inactivation by VNRX-5236, with k 2 /K i on the order of 10 4 to 10 5 M 21 s 21 ( Table 2). Dissociation rates (k -2 ) ranged from 2.5 to 24 Â 10 24 s 21 , which translated to appreciable active site residence times (t 1/2 ) from 5 to 46 min ( Table 2).
Selection of ceftibuten as an orally bioavailable partner antibiotic. A primary consideration for partner selection was the ability of VNRX-5236 to rescue b-lactam activity in strains of Enterobacterales expressing Ambler class A, C, and/or D b-lactamases, including serine carbapenemases. The activity of nine orally bioavailable b-lactam antibiotics was tested alone and in combination with VNRX-5236 (fixed at 4 mg/ml) against panels of Enterobacterales strains expressing class A ESBLs, class A KPC-carbapenemases, class C cephalosporinases, or class D OXA-carbapenemases. Many strains expressed multiple enzymes. These nine potential partner antibiotics were ceftibuten, cefixime, cefditoren, cefpodoxime, cefdinir, cefaclor, cephalexin, cefuroxime, and amoxicillin. The MIC 90 for all b-lactams tested alone against each panel of 25 strains was $64 mg/ml (26). As shown in Fig. 2, the combination of VNRX-5236 (fixed at 4 mg/ ml) with CTB (a third-generation, orally bioavailable cephalosporin; filled circles) resulted in the lowest MICs across all sets of strains tested (MIC 50 /MIC 90 of 0.12/1 mg/ ml). VNRX-5236 also demonstrated substantial rescue of cefixime and cefditoren with an overall MIC 50 /MIC 90 of 0.5/2 mg/ml. Amoxicillin was the weakest partner antibiotic (open circles). Coupled with its excellent oral bioavailability (;65%), long plasma halflife (;2.4 h), moderate human plasma protein binding (;60%), and extensive renal excretion (;56%), CTB was chosen as the preferred partner antibiotic (27,28). Lack of stand-alone activity of VNRX-5236 and VNRX-7145 against Grampositive and Gram-negative pathogens. VNRX-5236 and its prodrug, VNRX-7145 (VNRX-5236 etzadroxil), lack intrinsic antibacterial activity. Neither compound demonstrated stand-alone activity against four strains of Gram-positive organisms or nine strains of Gram-negative bacilli (all MICs were $128 mg/ml; see Table S1 in the supplemental material).
Impact of overexpression of key Ambler class A, B, C, and D enzymes on activity of CTB/VNRX-5236 and comparators in isogenic strains of Escherichia coli. The impact of overexpression of key Ambler class A, B, C, and D enzymes on the activity of CTB/VNRX-5236, and by inference the breadth of inhibitory activity of the BLI, was assessed in 20 engineered E. coli strains each overproducing individual serine-or metallo-b-lactamases and compared directly to other orally available b-lactam agents, including amoxicillin-clavulanic acid, CTB/clavulanic acid, sulopenem, and tebipenem. CTB/VNRX-5236, CTB/clavulanic acid, and amoxicillin-clavulanic acid were tested with inhibitor fixed at 4 mg/ml to avoid the stand-alone activity of clavulanic acid in these strains (the modal MIC for clavulanic acid alone against these 20 strains was 32 mg/ml; data not shown). Data are presented as the fold-increase in MIC relative to the vector control strain treated with b-lactam alone. Fold increases in MIC of #8, highlighted in gray in Table 3, indicate the stability of the b-lactam or b-lactam/BLI combination MIC to that enzyme.
As shown in Table 3, the CTB MIC was elevated in the presence of most class A, C, and B enzymes, with the exception of TEM-10 and GES-5, which increase the CTB MIC by only 4-to 8-fold. The CTB MIC was stable for most of the class D enzymes tested. Addition of VNRX-5236 at 4 mg/ml reduced the MIC shift to #4-fold in all strains except for the strain expressing the Ambler class B MBL, NDM-1. Addition of clavulanic acid fixed at 4 mg/ml reduced the CTB MIC in some strains expressing ESBLs but was ineffective against class A KPC carbapenemases, class C cephalosporinases, and class D carbapenemases.
The orally bioavailable b-lactam amoxicillin was hydrolyzed extensively by all enzymes tested, with the exception of the KPC-3 D179Y variant (29). As seen with CTB, clavulanic acid had variable activity against class A ESBLs and was largely ineffective against class A  KPC carbapenemases, class C cephalosporinases, and class D carbapenemases. Tebipenem was relatively resistant to hydrolysis by class A ESBLs and class C enzymes. It was, however, hydrolyzed extensively by class A, B, and D carbapenemases. Sulopenem had an enzyme hydrolysis profile similar to that of tebipenem.
Potentiation of CTB activity by VNRX-5236 in Enterobacterales. The antibacterial activity of CTB/VNRX-5236 and comparators was assessed against a diverse panel of 100 clinical isolates of Enterobacterales with defined b-lactamase subtypes ( Table 4). The panel does not reflect current epidemiological trends but instead highlights differences in coverage provided by CTB/VNRX-5236 relative to comparators. Antibacterial activity of CTB alone was included to ascertain the level of potentiation by the partnered BLI when tested at a fixed concentration of 4 mg/ml in all combinations except amoxicillin-clavulanic acid, where a fixed 2:1 ratio was tested (30,31). Isolates were subdivided by molecular characterization and included 25 strains of Enterobacterales expressing: (i) mixed class A ESBLs and OSBL (typically TEM-1), (ii) class A KPC-type carbapenemases, (iii) mixed class A ESBLs and OSBLs and class C cephalosporinases, and (iv) OXA-48/48-like carbapenemases.
The activity of CTB alone against isolates expressing Ambler class A ESBLs and OSBLs was variable, with an MIC 50 and MIC 90 of 8 and $64 mg/ml, respectively (Table 4). This result is consistent with the variable impact of class A enzyme expression on CTB activity in the isogenic strain panel (Table 3). VNRX-5236 rescued CTB activity against the clinical strains, with an MIC 50 and MIC 90 of 0.12 and 0.25 mg/ml. As these strains express enzymes that are variably susceptible to inhibition by clavulanic acid, the CTB MIC 90 was reduced from $64 with CTB alone to 0.5mg/ml for CTB/clavulanic acid. Amoxicillin-clavulanic acid had elevated MICs across these strains (MIC 90 of 32 mg/ml). The investigational oral penems demonstrated reasonably uniform activity against these strains, with MIC 90 s of 0.5 (sulopenem) and 0.06 (tebipenem) mg/ml, respectively. CTB alone demonstrated very high MIC 50 and MIC 90 values of 16 and $64 mg/ml, respectively, against strains expressing KPC-type carbapenemases. VNRX-5236 rescued CTB activity against these strains, with an MIC 50 /MIC 90 of 0.12 and 1 mg/ml, respectively, for the combination. Clavulanic acid was ineffective at rescuing CTB or amoxicillin in these strains. In addition, the MIC 50 /MIC 90 for sulopenem (32/$64 mg/ml) and tebipenem ($64/$64 mg/ml) demonstrated very poor activity of these agents against KPCexpressing isolates.
CTB alone demonstrated elevated MIC 50 /MIC 90 values of 16/$64 mg/ml against strains expressing class C cephalosporinases. These values were reduced to 0.12/1 mg/ml, respectively, in the presence of VNRX-5236. The strains had elevated CTB/clavulanic acid and AMC MICs (MIC 90 , $64 and $256 mg/ml, respectively), highlighting the weak protection afforded by clavulanic acid. The oral penems demonstrated good activity against these strains, with the MIC 90 for sulopenem and tebipenem being 2 and 0.5 mg/ml, respectively.
CTB demonstrated weak activity against strains expressing class D enzymes, with MIC 50 /MIC 90 values of 16 and $64 mg/ml, respectively. VNRX-5236 reduced the CTB MIC 50 /MIC 90 values to 0.12 and 1 mg/ml, respectively. CTB/clavulanic acid, sulopenem and tebipenem all demonstrated weak activity against these strains, with MIC 50 of 4 to 8 mg/ml and MIC 90 of 16 to 32 mg/ml. Amoxicillin-clavulanic acid did not demonstrate significant activity against these strains (MIC range, 128 to $256 mg/ml).
Assessment of bactericidal activity of CTB/VNRX-5236. To assess the ability of VNRX-5236 to restore the bactericidal activity of CTB against resistant isolates, 24-h time-kill assays were conducted. Shown in Table 5 are the modal MIC and minimum bactericidal concentration (MBC) values for antibacterial agents against these four bacterial strains. Shaded boxes represent data for the compound/pathogen combination shown in Fig. 3. VNRX-5236 restored CTB activity in all strains tested, reducing the CTBalone MIC from a range of 16 to 128 mg/ml to 0.12 to 1 mg/ml. The investigational penems demonstrated potent activity against strains expressing ESBLs or class C cephalosporinases, but as expected, their activity was compromised in the presence of carbapenemases such as KPC-2 and OXA-48.
Both CTB/VNRX-5236 and CTB/clavulanic acid demonstrated bactericidal activity against E. coli expressing CTX-M-15 (Fig. 3a). Amoxicillin-clavulanic acid and CTB alone or with clavulanic acid demonstrated initial killing, but the strains grew to control levels by 24 h (Fig. 3a). Both tebipenem and sulopenem demonstrated bactericidal activity against this strain at drug concentrations at or near their respective MICs (see Table 5). CTB/VNRX-5236 likewise demonstrated potent bactericidal activity against Enterobacter cloacae ECL01 expressing the class A ESBL TEM-72 and 2 class C enzymes (P99AmpC = ACT-C189 and ACT-1) (Fig. 3b), as did the penems and levofloxacin (Table 5). All other antibacterial agents failed in this time-kill assay, consistent with their elevated MICs. Results against the carbapenemase-expressing strains of K. pneumoniae are shown in Fig. 3c and d. Both strains had elevated MICs to all antimicrobial agents (8 to $256 mg/ml), with the exception of CTB/VNRX-5236 (0.12 to 0.25 mg/ml). CTB/VNRX-5236 demonstrated bactericidal activity against these 2 strains, with MBCs within 1 dilution of their respective MIC. Frequency of resistance (FoR). Experiments were conducted to assess the FoR for the CTB/VNRX-5236 combination at 4Â MIC. As shown in Table 6, no resistant colonies were obtained after 48 h of incubation. The FoR for the combination (defined as CFU in the presence of CTB/VNRX-5236 divided by total CFU plated) was ,4.3 Â 10 210 for all strains tested.
Impact of pH and human urine on activity of CTB/VNRX-5236. With urinary tract infections as the targeted indication for CTB/VNRX-5236, the activity of the combination over a range of relevant pH values in the presence of pooled human urine (100%) was assessed (32). Testing in urine at pH 6.0 produced MIC results equal to or within 1 to 3 doubling dilutions compared to standard broth microdilution for CTB and   (Table 7). Testing in urine at pH 7.2 produced MIC results equal to or 1 to 2 doubling dilutions lower than standard broth microdilution for CTB and CTB/ VNRX-5236.
In vivo efficacy of CTB/VNRX-5236 and CTB/VNRX-7145 in murine septicemia model. The ability of VNRX-5236 (dosed subcutaneously) or VNRX-7145 (VNRX-5236 etzadroxil dosed orally) to rescue CTB activity (dosed identically to the BLI) was assessed in a lethal murine septicemia model as described elsewhere (33). Briefly, mice were administered a lethal intraperitoneal inoculum of ceftibuten-resistant K. pneumoniae UNT-023 expressing KPC, TEM-1, and SHV-11 (CTB MIC, 64 mg/ml; CTB/VNRX-5236 MIC, 0.06 mg/ml). One hour later, antimicrobial agents were administered as a single dose by the route indicated. The experiment was conducted with groups of 5 mice and run twice in order to calculate median effective dose (ED 50 ) values with 95% confidence limits by probit analysis. As shown in Table 8, administration of CTB alone was ineffective at protecting mice from lethal disease (ED 50 , .128 mg/kg). In contrast, CTB/ VNRX-5236 etzadroxil dosed orally or CTB/VNRX-5236 dosed subcutaneously demonstrated robust activity with respective ED 50 values of 12.9 and 13.5 mg/kg. The results demonstrate potent rescue of CTB by VNRX-5236 and VNRX-7145 in this model. Hexagon, growth control; square, CTB alone at 8 mg/ml; open circle, CTB at 2 mg/ml with VNRX-5236 fixed at 4 mg/ml; Â inside a circle, CTB at 2 mg/ml with clavulanic acid fixed at 4 mg/ml; upside-down triangle, amoxicillin at 8 mg/ml with clavulanic acid at 4 mg/ml (a 2:1 ratio); triangle, tebipenem at 2 mg/ml; diamond, sulopenem at 2 mg/ml. High selectivity and lack of mammalian cytotoxicity. The selectivity and specificity of VNRX-5236 and its etzadroxil prodrug were evaluated in the DrugMatrixScreen panel of pharmacological targets at Eurofins/Panlabs (34). Binding, enzymatic, and uptake assays representing a wide range of cellular and subcellular target classes were performed. At the screening concentration of 100 mM for VNRX-5236, two modest hits were reported-56% inhibition of protein serine/threonine phosphatase, PPP3CA (calcineurin PP2B), and 77% inhibition of protein tyrosine kinase Fyn. The in vivo significance of these in vitro activities is not known. VNRX-7145 (VNRX-5236 etzadroxil) was tested at a concentration of 10 mM, with no notable off-target findings reported against any of the 129 in vitro assays. VNRX-5236 and VNRX-7145 (VNRX-5236 etzadroxil) were also found to be noncytotoxic to dividing HeLa, MRC-5, and 3T3 cell cultures up to the highest concentration tested (1 mg/ml and 128 mg/ml, respectively).

DISCUSSION
In spite of recent approvals of intravenous BL/BLI combinations for clinical use, no orally bioavailable BL/BLI combination has been approved in several decades to address the plethora of new b-lactamases. Two orally bioavailable investigational penem prodrugs have recently completed clinical phase 3 studies, with mixed results. Sulopenem etzadroxil failed to achieve noninferiority relative to ertapenem injection in patients with complicated urinary tract infections (cUTI; SURE-2 trial) or complicated intra-abdominal infections (cIAI; SURE-3 trial) (35)(36)(37). Tebipenem pivoxil hydrobromide recently demonstrated noninferiority relative to intravenous (i.v.) ertapenem in cUTI, including pyelonephritis (ADAPT-PO trial) (38). While the oral penems generally escape hydrolysis by Ambler class A ESBLs and class C cephalosporinases, they remain vulnerable to inactivation by class A and D serine carbapenemases. As such, unprotected oral penems lack utility as a step-down therapy in hospitalized patients with infections caused by carbapenemase-producing Gram-negative bacteria. Moreover, there are potential antimicrobial stewardship concerns in that widespread use of unprotected penems could further increase carbapenem resistance, which is already considered a major and growing issue worldwide (39)(40)(41)(42)(43).
Ceftibuten/VNRX-5236 demonstrates potent antibacterial activity against Enterobacterales producing Ambler class A serine b-lactamases, class C cephalosporinases, and class A and D carbapenemases (e.g., KPC and OXA-48-type,   (44,45). Such coverage is of critical importance in light of the ever-expanding contribution of these enzymes to b-lactam resistance globally (46,47). Indeed, the previous 2 decades have seen an explosive, global expansion of ESBLs in strains of Enterobacterales in both the hospital and community settings (48). Such strains are considered a "critical priority" by the World Health Organization, along with strains of Enterobacterales expressing carbapenemases (49). Since their first reporting in 1996 in North Carolina, KPC-expressing strains of Enterobacterales have become endemic in many parts of the world, including the northeastern United States and Puerto Rico, much of South America, eastern China, parts of southern Europe, and Israel (50,51). Likewise, since its discovery in Turkey in 2003, the OXA-48 carbapenemase has been identified in many parts of the world, including in the Mediterranean region and European countries (52,53). The rapid expansion of OXA-48 in strains of Enterobacterales, particularly in Europe, compromises the utility of approved oral b-lactams, as well as the investigational penems discussed above (53). The selection of ceftibuten as the b-lactam partner to VNRX-5236 was driven by multiple considerations, including (i) potency in combination with VNRX-5236 against strains of Enterobacterales, (ii) excellent oral pharmacokinetics, (iii) modest protein binding, (iv) significant excretion in urine (optimal for UTIs), and (v) carbapenem sparing. The currently approved dose of ceftibuten, 400 mg orally once daily, is insufficient to support plasma exposures needed to cover at least 90% of these pathogens (54). As such, extensive pharmacokinetic/pharmacodynamic (PK/PD) assessments are ongoing, both in vitro and in vivo, as well as a stand-alone phase 1 clinical trial, to select a suitable dose and dosing regimen of ceftibuten in the combination and confirm its tolerability in human volunteers (55). VNRX-5236 potently inhibited problematic b-lactamases found commonly in Enterobacterales, including KPC, CTX-M-15, P99AmpC, CMY-2, OXA-1, and OXA-48. Accordingly, in isogenic E. coli strains individually expressing these important b-lactamases, VNRX-5236 reduced the ceftibuten MIC to within 4-fold of the vector control in all strains, with the exception of the MBL NDM-1. This level of rescue far exceeded that observed when clavulanic acid was paired with ceftibuten or amoxicillin. While the penems were less susceptible than cephalosporins or amino penicillins to hydrolysis by ESBLs and class C enzymes, tebipenem and sulopenem MICs were nonetheless increased against most b-lactamase-producing isogenic strains compared to the vector control. Importantly, both investigational penems were hydrolyzed extensively by KPC and OXA carbapenemases.
In clinical Enterobacterales isolates, CTB/VNRX-5236 demonstrated potent, broad-spectrum activity against strains expressing any Ambler class of serine b-lactamase. Time-kill assays indicated that CTB/VNRX-5236 remained bactericidal over 24 h against representative strains from each Ambler serine b-lactamase class, while frequency of resistance studies revealed a very low propensity for selection of resistant variants across a range of b-lactamase-expressing strains of Enterobacterales. Proof of concept studies in the murine septicemia model provided strong evidence of in vivo efficacy for ceftibuten in combination with VNRX-7145 (VNRX-5236 etzadroxil) dosed orally and VNRX-5236 dosed subcutaneously. Indeed, previously reported studies also demonstrated potent inhibition of strains of Enterobacterales in the neutropenic mouse thigh and ascending urinary tract infection models (64)(65)(66).
The limitations of this study are noted. First, we have not conducted prevalencebased surveillance studies to support MIC 50 and MIC 90 determinations globally or by geography. Those studies are just now getting under way. Nevertheless, the predominant resistance mechanism in strains of Enterobacterales in the United States are ESBLs, which account for about 24% of all Enterobacterales infections (67). As these strains will almost certainly represent the MIC 90 for CTB/VNRX-5236, we anticipate the MIC 90 value for the combination to be linked to this resistance mechanism. Second, a limited number of representative b-lactamases were characterized biochemically and in bacterial cells. Our focus initially was to assess activity against the most prevalent enzymes found in strains of E. coli and K. pneumoniae. Finally, this study has not considered the role of non-b-lactamase-mediated resistance mechanisms (porin downregulation, efflux upregulation, penicillin-binding protein [PBP] target mutations) on the activity of the CTB/VNRX-5236 combination. Further expansion of this testing is ongoing. The data support further development of this first-in-class, broad-spectrum, oral BLI in combination with ceftibuten. Stock solutions (10 mg/ml) of cefixime, cefpodoxime, ceftibuten, sulopenem, tebipenem, and VNRX-5236 were prepared in dimethyl sulfoxide (DMSO). Then, 10-mg/ml stock solutions of amoxicillin, cefdinir, cefditoren, cephalexin, and clavulanic acid were prepared in 0.1 M phosphate buffer, pH 6. A 10 mg/ml stock solution of cefaclor was prepared in water, and a 5 mg/ml stock solution of levofloxacin was prepared in ethanol. These stocks were used to prepare the final test concentrations for antibacterial assays. Weights of the salts and purity of the antimicrobial agents were taken into account when preparing the stock solutions when applicable.

MATERIALS AND METHODS
For biochemical analyses, VNRX-5236 (lot no. EFM00077-053) was synthesized at Venatorx Pharmaceuticals, Inc. Cephalothin (lot no. 106M4780V) and cefotaxime (lot no. 094M4726V) were purchased from Sigma (Atlanta, GA); imipenem (lot no. R038R0) was from United States Pharmacopeia (Rockville, MD); nitrocefin (batch no. N005-09) was from TOKU-E (Bellingham, WA). Stock solutions (10 mM) for cephalothin and imipenem were prepared in phosphate-buffered saline (PBS), while cefotaxime and nitrocefin (15 mM) stocks were prepared in dimethyl sulfoxide (DMSO, D128-500; Fisher Chemical, Fairlawn, NJ). Working solutions were subsequently prepared in assay buffer, according to the final concentrations required for assay. Stock solutions (10 mM) of b-lactamase inhibitors (VNRX-5236, avibactam, tazobactam, and clavulanic acid) were prepared in DMSO and used to prepare working solutions in assay buffer according to the final assay concentrations required for the assay.
Antibacterial activity of CTB/VNRX-5236 in reference isolates and susceptibility testing methodology. The in vitro activity of CTB/VNRX-5236 was measured by broth microdilution as the activity of CTB in the presence of VNRX-5236 at a fixed concentration of 4 mg/ml. The rationale for choosing 4 mg/ml of VNRX-5236 for in vitro testing is based on in vitro broth microdilution assessments across a range of fixed concentrations of VNRX-5236 and correlation of in vitro results with antibacterial activity in the neutropenic mouse thigh infection model (64,66). Against 1,066 clinical urinary tract infection isolates resistant to both amoxicillin-clavulanic acid and levofloxacin, the MIC 90 of CTB was reduced from .32 mg/ml when tested alone to 2 mg/ml at a fixed concentration of 4 mg/ml of VNRX-5236 (J. A. Karlowsky, M. A. Hackel, and D. F. Sahm, submitted for publication; 68). The MIC distribution of CTB with VNRX-5236 fixed at 4 mg/ml against CTB nonsusceptible Enterobacterales from multiple challenge sets of clinical isolates resembles the MIC distribution of CTB against CTB-susceptible isolates. The overlapping MIC distributions for CTB against CTB-susceptible strains and CTB/VNRX-5236 against CTB-nonsusceptible strains (with VNRX-5236 fixed at 4 mg/ml) reflect the complete or nearly complete rescue of the activity of CTB by VNRX-5236 against target organisms. Importantly, humanized dosing of CTB and matching VNRX-5236 profiles in the neutropenic murine thigh infection model reduced the bacterial burden of all isolates of Enterobacterales with CTB/VNRX-5236 MIC values of 2 mg/ml or below by 0 to 1 log 10 at 24 h compared to the burden at the outset of therapy (64,66). These results are consistent with the MIC values for CTB derived in the presence of 4 mg/ml of VNRX-5236.
Expression plasmid construction. Plasmid DNA, PCR product purification, and gel extractions were performed using Wizard Plus SV miniprep and SV gel and PCR extraction kits (Promega). NdeI, BamHI, and XhoI restriction enzymes, T4 DNA ligase, and E. coli BL21(DE3) competent cells were purchased from New England Biolabs. All oligonucleotide primers for PCR amplification were purchased from Integrated DNA Technologies. All PCRs were performed with Phusion high-fidelity DNA polymerase and cloning performed in E. coli DH5a subcloning efficiency chemically competent cells (Thermo Fisher). The pET9a (StrateGene) expression clones were made using PCR amplification products from molecularly characterized clinical isolates carrying the desired b-lactamase gene and cloned NdeI to BamHI into pET9a in all cases. All b-lactamases were cloned with signal peptide encoding sequences. All transformants were verified by PCR amplification, restriction endonuclease mapping, and DNA sequencing. Confirmed expression plasmids were isolated with plasmid miniprep kits (Promega) and used to transform the expression cell line E. coli BL21(DE3) or E. coli JM109(DE3).
b-lactamase purification. For CMY-2 and OXA-1, a 50 ml preculture of E. coli BL21(DE3) cells containing the pET9a expression vector for the individual b-lactamases was grown at 37°C overnight in lysogeny broth (LB) medium supplemented with 50 mg/ml kanamycin. Common to all b-lactamase purifications, bacterial cells were lysed by three consecutive passes through a chilled French pressure cell at 18,500 lb/in 2 and clarified by centrifugation at 10,000 Â g for 30 min at 4°C. The b-lactamase activity was monitored using nitrocefin at 100 mM, and purity was examined using 10% SDS-PAGE with Coomassie brilliant blue staining. Purified proteins exhibiting .95% purity with SDS-PAGE were quantified with a Pierce BCA protein assay kit and bovine serum albumin as the standard (Thermo Fisher), concentrated to a working range of 1 to 5 mg/ml and frozen at -80°C in buffer containing 10% glycerol. Column purifications were performed using an AKTA fast protein liquid chromatograph (FPLC; GE Healthcare). The purification schemes were generally similar, with enzyme-specific differences described below.
CTX-M-15 was obtained from E. coli BL21(DE3) carrying plasmid pET-CTX-M-15, grown in 2 liters of MagicMedia autoinduction medium (Invitrogen) containing 50 mg/ml kanamycin (Sigma) for 24 h at 23°C. Cells were harvested at an optical density at 600 nm (OD 600 ) of 2.2 by centrifugation at 7,500 Â g at 4°C and resuspended in 60 ml of 10 mM HEPES-NaOH (pH 7.0) supplemented with 0.5 mM EDTA. The lysate was diluted 5-fold with cold 50 mM sodium acetate (pH 4.8) and incubated overnight at 4°C. The extract was clarified by centrifugation at 14,500 Â g at 4°C, filtered through an Amicon nitrogen concentrator with a 10-kDa cutoff filter to a volume of 50 ml, and loaded onto a HiTrap CaptoS column preequilibrated in 50 mM sodium acetate (pH 4.8). Protein was eluted by a linear gradient of 50 mM sodium acetate (pH 4.8) supplemented with 500 mM NaCl. Fractions containing active CTX-M-15 were pooled and concentrated, and the buffer was exchanged for 20 mM HEPES-NaOH (pH 7.2), 150 mM NaCl, and 10% glycerol using Amicon Ultra-15 centrifugal concentrators. CTX-M-15 was further purified using a Superdex 200 gel filtration column.
OXA-48 was obtained from E. coli BL21(DE3) carrying plasmid pET-OXA-48, grown in 2 liters of MagicMedia autoinduction medium (Invitrogen) with kanamycin as previously described for 24 h at 23°C (20). Cells were harvested at an OD 600 of 3 by centrifugation at 7,500 Â g at 4°C, resuspended in 60 ml of 20 mM triethanolamine buffer (pH 5.5), and then purified as described for CTX-M-15, with the exception that the final buffer contained 10 mM NaHCO 3 to keep the critical active site lysine residue carbamylated (69).
KPC-2 was obtained from E. coli BL21(DE3) carrying plasmid pET-KPC-2, grown in 3 liters of MagicMedia autoinduction medium (Invitrogen) with kanamycin as previously described for 24 h at 23°C (20). Cells were harvested at an OD 600 of 3.3 by centrifugation at 7,500 Â g at 4°C and resuspended in 70 ml of 20 mM MES-NaOH (pH 5.5). The extract was clarified by centrifugation at 14,500 Â g at 4°C, filtered through an Amicon nitrogen concentrator by use of 10-kDa cutoff filters to a volume of 50 ml, and loaded onto a HiTrap CaptoS column preequilibrated in 20 mM MES-NaOH (pH 5.5). Protein was eluted by a linear gradient of 20 mM MES-NaOH (pH 5.5) supplemented with 500 mM NaCl. KPC-2 active fractions were pooled and concentrated, and buffer was exchanged in 20 mM HEPES-NaOH (pH 7.3) and 150 mM NaCl using Amicon Ultra-15 centrifugal concentrators. KPC-2 was further purified using gel filtration chromatography with a Superdex 200 column.
P99AmpC was purified directly from the Enterobacter cloacae P99 clinical isolate after a sequence of the b-lactamase-encoding gene had been verified by DNA sequencing of PCR amplified product. To produce P99AmpC, Enterobacter cloacae P99 cells were grown in LB in the presence of a sub-MIC (0.016 mg/ ml) of imipenem to induce maximal expression of the enzyme. Cells were harvested at an OD 600 of 2.4 by centrifugation at 7,500 Â g at 4°C, resuspended in 50 ml of 20 mM MES-NaOH (pH 5.5), and otherwise purified in a similar manner to KPC-2 described above.
SHV-5 was obtained from E. coli BL21(DE3)/pLysS carrying plasmid pET-SHV-5, grown at 25°C in 2 liters of super broth supplemented with kanamycin and chloramphenicol to an OD 600 of 0.5, when IPTG (isopropyl-b-D-thiogalactopyranoside) was added to 0.05 mM final, and induction proceeded for 6 h. Cells were harvested by centrifugation at 5,500 g for 15 min at 4°C. The cell pellet was resuspended in 60 ml of 10 mM HEPES-NaOH (pH 7.5), and cells were lysed using a French press. The lysate was diluted 5-fold in 50 mM sodium acetate (pH 5.0) and kept at 4°C overnight. The extract was clarified by centrifugation at 14,500 Â g at 4°C, filtered through an Amicon nitrogen concentrator by use of 10-kDa cutoff filters to a volume of 50 ml, and loaded onto a HiTrap CaptoS column equilibrated with 50 mM sodium acetate (pH 5.0). SHV-5 was eluted by a linear gradient of 50 mM sodium acetate (pH 5.0) supplemented with 500 mM NaCl. Active fractions were pooled and concentrated, and buffer was exchanged in 20 mM HEPES-NaOH (pH 7.3), 150 mM NaCl, and 10% glycerol using Amicon Ultra-15 centrifugal concentrators. Finally, the SHV-5 sample was further purified by chromatography over a Superdex 200 gel filtration column.
Half maximal inhibitory concentration (IC 50 ) determinations. To determine IC 50 values, VNRX-5236 was serially diluted 3-fold in PBS in 96-well microtiter plates to obtain assay concentrations ranging from 10 mM to 0.00001 mM (1mM to 0.000001 mM with P99AmpC). An equal volume of enzyme was added (assay concentrations, 0.6 nM P99AmpC, 21 nM CTX-M-15, 3 nM KPC-2, 100 nM SHV-5), and the mixtures were preincubated for 15 min at 37°C. For AmpC, nitrocefin was then added to obtain a final concentration of 100 mM, and the absorbance at a wavelength (l) of 486 nm (representing hydrolysis of the nitrocefin b-lactam ring) was monitored immediately using a BioTek Powerwave XS2 microplate reader. In similar fashion, using UV-transparent 96-well microtiter plates, hydrolysis of imipenem (100 mM) by KPC-2 was monitored at l 300 nm, while cefotaxime (100 mM) hydrolysis by CTX-M-15 and SHV-5 was monitored at l 300 nm. Assays were performed in duplicate, and the percentage inhibition relative to reaction mixtures containing no inhibitor were determined for each concentration of inhibitor. The concentration of inhibitor required to reduce the initial rate of hydrolysis by 50% (IC 50 ) was determined based on the residual b-lactamase activity.
Measurement of the onset of inhibition (k 2 /K i ). VNRX-5236 binds to the active site of serine b-lactamases and reacts with the catalytic serine nucleophile to form a reversible covalent bond. This interaction is described by a 2-step model for reversible inhibition (1) shown in the scheme below, where E refers to enzyme and I to inhibitor, EI is the noncovalent enzyme:inhibitor complex, and EI* is the covalent enzyme:inhibitor complex. of VNRX-5236 were as above, and the reaction progress was monitored continuously by measuring the reduction in absorbance at l 260 nm. For KPC-2, imipenem (75 mM) was used as the substrate, and reactions were initiated with 3 nM enzyme. Reaction progress was monitored by measuring the reduction in absorbance at l 300 nm in a continuous fashion. The concentrations of VNRX-5236 tested were 10mM, 5 mM, 2.5 mM, 1.25 mM, 0.625mM, 0.313 mM, 0.156 mM, 0.0781mM, 0.0391mM, 0.0195 mM, 0.0098 mM, and 0 mM. Time courses were fit to equation 1 to obtain the observed rate constant, k obs , at each inhibitor concentration, and k 2 /K i was subsequently determined by fitting plots of inhibitor concentration versus k obs to equation 2, which includes a correction for substrate concentration and affinity.
Off rate (k off ; k 22 ) and half-life of active site occupancy (t 1/2 ) determination. Off rates for VNRX-5236 with the various b-lactamases were determined by jump dilution experiments performed in triplicate. The enzyme and inhibitor were incubated at room temperature for 10 min. Enzyme:inhibitor complexes were then diluted 800-fold into reaction buffer (50 mM HEPES, pH 7.0, 0.1 mg/ml bovine serum albumin [BSA]), and 20 ml of the diluted reaction was immediately added to 180 ml of 110 mM nitrocefin in a 96-well microtiter plate, and the absorbance at l 486 nm measured continuously on a BioTek Powerwave XS2 microplate reader. The resulting progress curves were fit to a single exponential, from which k off was derived. The half-life was determined from equation 3.
Determination of K i . Plots of the inhibitor concentration versus inverse the initial velocity were fit to a linear equation, from which K i was derived (equation 4). The observed k i was corrected for substrate concentration and affinity by: [S] and K m correspond to the substrate concentration and Michaelis Menten constant, respectively. Antibacterial activity. The in vitro antibacterial activity of b-lactams alone or in combination with BLIs was determined in cation-adjusted Mueller-Hinton broth (CAMHB) microdilution assays according to Clinical and Laboratory Standards Institute (CLSI) recommendations (30). Antibacterial potentiation by VNRX-5236 was assayed by fixing the BLI at 4 mg/ml, while clavulanic acid was either fixed at 4 mg/ml for testing in combination with CTB or tested at an amoxicillin:clavulanic acid ratio of 2:1 (30). Inocula for broth microdilution assays were prepared using the broth culture method. The b-lactams were 2fold serially diluted in CAMHB, with final concentrations generally ranging from 0.06 to 128 mg/ml when tested alone and from 0.016 to 32 mg/ml when tested in combination with their respective BLI. The MICs reported are modal values from 3 to 5 independent replicates performed on separate days and by separate investigators.
In vitro antibacterial time-kill assay. Bactericidal activity was assessed using time-kill assays according to standard CLSI methods (70). Time-kill assays were performed in 14 ml glass tubes or deep-well storage plates with a bacterial inoculum in CAMHB of 5 Â 10 5 CFU/ml. The tubes/plates were incubated at 37°C with shaking at 200 rpm, and aliquots were drawn at six time points (0, 2, 4, 6, 8, and 24 h) to conduct 0.5 log 10 dilutions in CAMHB in 96-well plates that were incubated overnight at 37°C and used to ascertain viable bacterial counts. The lowest quantifiable amount by this method is 2 log 10 CFU/ml.
Spontaneous frequency of resistance. Bacteria (2.3 Â 10 9 to 1.2 Â 10 10 CFU) were plated on cationadjusted Mueller-Hinton agar plates containing CTB at 4Â the MIC and VNRX-5236 fixed at 4 mg/ml. Plates were incubated for 48 h at 37°C, and surviving colonies were enumerated. Frequency of resistance (FoR) is defined as CFU in the presence of CTB/VNRX-5236 divided by CFU plated.
Mammalian cell cytotoxicity. Cytotoxicity evaluations were conducted on actively dividing cells to maximize assay sensitivity. Procedurally, 180 ml of cells were seeded in cell-culture-treated 96-well plates (Celltreat, Pepperell, MA) at a density of 2 Â 10 4 to 3.5 Â 10 4 cells/ml and incubated overnight at 37°C in a humidified, 5% CO 2 atmosphere. The next day, compound was diluted in PBS with 6% DMSO in a 0.5 log 10 dilution scheme at 10 times the final compound concentration array. Compound titrations (in 20 ml volume) were added to each well of the cell culture plates to affect a final 10-fold dilution ranging from 0.315 to 1,000 mg/ml for VNRX-5236, VNRX-7145 (VNRX-5236 etzadroxil), and ceftazidime. The cytotoxic agent chlorohexidine was assayed as a positive control at compound concentrations ranging from 0.31 to 100 mg/ml. The final DMSO concentration in all assays was 0.6%.
After an additional 72 h of incubation (when cells reached 90 to 100% confluence) Cell Titer 96 AQ solution was added at 20 ml per well, and plates were incubated an additional 2 to 4 h at 37°C and 5% CO 2 . The optical density of the wells was read on a Biotek Cytation 3 plate reader at 490 nm. Background signal (wells receiving Cell Titer reagent but containing no cells) was subtracted from all wells on the plate. The percent growth versus log drug concentration was plotted, and the concentrations at which 50% of the cells survived compared to growth control (CC 50 ) were calculated using a sigmoidal 4 parameter curve fit (GraphPad Prism). Experiments were run in triplicate with data from all three trials included in the data analysis.
Impact of pH and human urine on activity of CTB/VNRX-5236. Studies were conducted by IHMA, Schaumberg, IL. Pooled human urine (100%) was used for panel preparation (BIOIVT, product code 1801844; lot BRH1464632; expiry date, 2020-03-31) and adjusted to pH 6.0 and pH 7.2. MICs were determined by broth macrodilution methodology (30,31). Bacterial inocula were prepared in saline at approximately 1 Â 10 6 CFU/ml by dilution of a McFarland 0.5 in saline. Antibacterial panels containing 100 ml of antibacterial solution were inoculated with the adjusted inocula using 95-prong inoculator sets. Test panels were read visually after incubation following CLSI guidelines for each organism. MIC values corresponded to the first well with no visible growth.
Selectivity testing. The selectivity of VNRX-5236 was assessed in the DrugMatrixScreen at Eurofins Discovery Services. This screen encompasses 123 mammalian receptor and enzyme classes (and a single prokaryotic target, a b-lactamase) that can be used to assess the potential off-target activities of development candidates (34).
Data availability. The raw data for graphs and tables presented in this paper can be found in Tables S2 to S9.

SUPPLEMENTAL MATERIAL
Supplemental material is available online only. SUPPLEMENTAL FILE 1, PDF file, 0.3 MB.