Discovery and Evaluation of a Methylpyrazolopyrimidine Antibacterial Active against Methicillin-Resistant Staphylococcus aureus

Abstract

Methicillin-resistant Staphylococcus aureus (MRSA) is a nefarious human bacterial pathogen classified as a serious threat. MRSA strains are resistant to virtually all β-lactam antibiotics (including penicillins and cephalosporins). A common resistance mechanism to β-lactams is mediated by the function of the bla operon, which encodes a β-lactam sensor/signal transducer protein BlaR, a gene repressor BlaI, and a resistance determinant: a class A β-lactamase (BlaZ) and/or a unique penicillin-binding protein 2a (PBP2a). BlaR is responsible for sensing the presence of β-lactam antibiotics and transducing a signal to its cytoplasmic domain upon binding covalently to the β-lactam. This triggers a series of cytoplasmic events that culminate in full-blown antibiotic resistance. We have used a fluorescence-reporter assay in live S. aureus to screen two NCI compound libraries—natural product and diversity libraries—comprising 1,974 compounds for both antibacterial and antibiotic-potentiation activities. Compound 1, N⁴,N⁶-bis(4-bromophenyl)-1-methyl-1H-pyrazolo[3,4-d]pyrimidine-4,6-diamine, emerged from these assays as a potentiator of the activity of oxacillin (a second-generation penicillin), while also exhibiting antibacterial activity of its own. The compound binds to the BlaR sensor domain (31 μM) in shutting down the bla operon, and also binds to the structurally related PBP2 and PBP2a, which are both critical targets for cell wall assembly. Scanning electron microscopy documented cell wall damage caused by compound 1 in combination with oxacillin.

Graphical Abstract

INTRODUCTION

Staphylococcus aureus is a bacterial human pathogen that caused 306,600 hospitalizations and 10,200 deaths in the U.S. alone in 2019.¹ It typically accounts for 30% of deaths associated with antibiotic resistance in the U.S.² Penicillins were the agents of choice in treatment of infections caused by S. aureus but strains producing a penicillinase—broadly referred to as β-lactamase, a resistance enzyme to β-lactams—emerged shortly.³ The emergence of this resistant S. aureus prompted semisynthetic efforts to produce second-generation penicillins, including methicillin, which did not serve as substrates for S. aureus BlaZ β-lactamase. These organisms came to be known as methicillin-sensitive S. aureus (MSSA). Within 2-years resistance to MSSA strains emerged, exhibiting a breadth of resistance to the entire class of β-lactam antibiotics.⁴ These strains were named methicillin-resistant S. aureus (MRSA), which remains a global problem to this day. The antibiotic resistance profile of MRSA continues to expand, now including variants that are essentially resistant to all known antibiotics.^5–10 The present need for effective treatment options for S. aureus infections is urgent.

The cognate bla and mec operons are responsible for resistance to β-lactams in S. aureus. Indeed, many MRSA strains have both. The bla operon encodes a gene for a β-lactam sensor/signal transducer protein BlaR (MecR in the mec operon), a gene repressor BlaI (MecI in the mec operon) and an antibiotic-resistance determinant BlaZ β-lactamase (penicillin-binding protein 2a (PBP2a) in the mec operon) (Figure 1).¹¹ Whereas BlaR is more prevalent in controlling the resistance response, over 85% of MRSA strains produce both the BlaZ and PBP2a proteins.^12–15 BlaZ hydrolytically degrades β-lactam antibiotics; therefore, they cannot inhibit their target penicillin-binding proteins (PBPs), but PBP2a is simply impervious to inhibition by virtually all β-lactam antibiotics. The function of BlaR is key in the onset of the resistance mechanisms in an inducible manner. In the presence of β-lactam, the sensor domain of BlaR experiences covalent binding by the antibiotic. Through the propagation of conformational change across the membrane, the metalloprotease domain of the protein is activated to degrade the BlaI protein. Once the operon is derepressed by the proteolytic activity, the genes are transcribed, resulting in the formation of BlaR itself, BlaI, and the β-lactamase BlaZ and/or PBP2a. The effect of these events is the onset of the resistance process for β-lactam antibiotics in S. aureus (Figure 1).

Figure 1.

Depiction of the manifestation of resistance by the function of BlaR. (A) The surface sensor domain of BlaR is activated by the binding of a β-lactam (top left). This leads to a signaling cascade through the membrane, which causes the zinc metalloprotease domain of BlaR to cleave BlaI, the gene repressor. This cleavage allows for transcription of the bla operon, conferring resistance to β-lactams through the production of the β-lactamase (BlaZ) and/or PBP2a. Compound 1 is able to inhibit BlaR by binding to the sensor domain. PBP2 is an essential cell-wall transpeptidase and a target for some β-lactams. The expression of PBP2a compensates for the lost transpeptidase activity of PBP2.¹⁶ Compound 1 inhibited both PBP2 and PBP2a. (B) The bla operon was modified in plasmid pCN52-bla to create a fluorescent reporter plasmid. The gene for blaZ was replaced by the gene for the fluorescent protein Venus. The binding of a β-lactam to the sensor BlaR culminates in fluorescence by the Venus protein, indicating activation of the bla operon.

We recently reported that the BlaR protein is a druggable target.¹⁷ BlaR inhibition by a small molecule, without facilitating the conformational change, abrogates the function of the entire operon. These events would shut down the antibiotic-resistance mechanism, whereby S. aureus becomes susceptible to β-lactam antibiotics again. Herein, we report a screening that has identified a BlaR inhibitor that serves as an adjuvant in restoring activity to β-lactams against MRSA strains. The compound also exhibits antibacterial activity of its own, the mechanism of which will be detailed in this study.

RESULTS AND DISCUSSION

We described a fluorescent reporter S. aureus that was transformed with plasmid pCN52RI-Venus, which harbors a bla operon in which the antibiotic-resistance gene was replaced with the yellow fluorescent protein Venus—designated as S. aureus NRS144_{pCN52RI-Venus}.¹⁷ The functional bla operon causes the organism to fluoresce upon exposure to β-lactam antibiotics. We employed this strain to screen small-molecule inhibitors of the system using two compound libraries from the National Cancer Institute: the 1,974-member Diversity Library and the Natural Product Library. S. aureus NRS144_{pCN52RI-Venus} was placed in 96-well plates at an optical density of 0.4 in lysogeny broth (LB) medium. In the presence of 2-(2-carboxyphenyl)-benzoyl-6-aminopenicillanic acid (CBAP; 10 μg/mL), a good inducer penicillin devoid of interfering chromophore, the system was activated. Compounds were added at a final concentration of 25 μM in 2% dimethyl sulfoxide (DMSO). The bacteria were gently shaken at 37 °C for 2 h. Both fluorescence (for activation of the bla operon; λ_ex = 510 nm and λ_em = 532 nm) and OD₆₀₀ (for bacterial growth) were measured at that point (Figure 1B). We also evaluated the ratio of fluorescence to OD₆₀₀, which assesses the extent of induction corrected for bacterial growth in cases where the compound also exhibited anti-S. aureus activity. Compounds were organized into three categories: those that inhibited only fluorescence (likely BlaR inhibitors), those that inhibited both fluorescence and bacterial growth (ranked from the highest to lowest fluorescence/OD₆₀₀ ratio), and those that only inhibited growth. Compounds that were known antibiotics or that only inhibited growth were not considered further. All compounds that showed only inhibition of fluorescence and those that had a fluorescence/OD₆₀₀ ratio greater than 60% of the control moved to the next stage of screening.

This analysis led to a total of 103 compounds of interest (23 natural products and 80 diversity sets), which were assayed for their ability to potentiate the activity of oxacillin (OXA; second-generation penicillin) against one MSSA strain (ATCC29213) and two MRSA strains (N315 and USA300–0114). Strain N315 is an SCCmec type II strain (harboring both intact bla and mec operons), and USA300–0114 is an SCCmec type IV strain (carrying an intact bla operon and a truncated mec operon), prevalent in the US. Compounds that were found to potentiate the activity of OXA by ≥8-fold reduction of the MIC against one or more strains were then tested with an expanded number of S. aureus strains: COL, HIP07920, MRSA252, and HIP13170. COL is an SCCmec type I strain containing no bla operon and a truncated mec operon and produces PBP2a constitutively; HIP07920 is a vancomycin-intermediate S. aureus (VISA) SCCmec type IV strain containing an intact bla operon and a truncated mec operon; MRSA252 is an SCCmec type II strain, where the bla operon is encoded on the chromosome; and HIP13170 is a vancomycin-resistant S. aureus (VRSA) strain. Of the 103 compounds that received the second set of screenings, compound 1 stood out, which warranted its evaluation against a total of 30 additional S. aureus strains (Figure 2A and Table S1). Not only did compound 1 exhibit anti-S. aureus activity of its own—minimal inhibitory concentrations (MICs) of ≤1 to 4 μg/mL—it showed an exceptional ability to potentiate OXA (Table S1: HIP06297, 128-fold; M1712, 1,024-fold; 626, 1,024-fold; N315, 256-fold; BR 15, 128-fold; JE2, 512-fold; USA300–0114, 2,048-fold; AIS2006061, 32-fold; HIP13170, 4,096-fold). When 0.5 μg/mL compound 1 (1/4 the MIC) was used, there was no potentiation of OXA against the COL strain, as would be expected. However, when 1 μg/mL compound 1 (1/2 the MIC) was used, there was a 16-fold decrease in the MIC of OXA against the COL strain. This result is consistent with the inhibition of PBP2a, which is the primary resistance mechanism of the COL strain, by 1. In the absence of BlaR in the COL strain, it is the dual inhibition of PBP2 by OXA and of PBP2a by 1 that accounts for the 16-fold decrease in the MIC of OXA. Across the board, the largest potentiation effects were seen for the MRSA strains (Table S1), whereas the MSSA strains (the first five entries in the table; already susceptible to OXA) typically showed a mere 2-fold effect. Compound 1 showed antibacterial activity against Gram-positive Bacillus licheniformis and Bacillus subtilis (MICs of 2–4 μg/mL), but its potentiation of 2- to 4-fold of the activity of OXA in these cases was well below those for the MRSA strains. Although these bacilli harbor a bla operon but not a mec operon,^18,19 their BlaR protein shares low amino acid identity with the S. aureus counterpart (in the 20 percentiles). Furthermore, compound 1 was not active against Gram-negative strains of Table S2.

Figure 2.

Evaluation of compound 1 in vitro and in vivo. (A) The chemical structure of compound 1 and the checkerboard assay for compound 1 and OXA against strain N315 are shown. (B) Compound 1 was assessed in a time–kill assay against strain N315. OXA was evaluated at a concentration of 32 μg/mL (light blue solid line), and compound 1 at 1.2 and 2.4 μg/mL (red dashed line and red solid line, respectively). The OXA:1 combinations (in μg/mL) used in the time–kill assay were 32:1.2 (dark green solid line), 16:1.2 (dark green dashed line), 32:2.4 (light green solid line), and 16:2.4 (light green dashed line). (C) Concentrations of compound 1 were determined in plasma (red trace) and thigh tissue (black trace) of mice. Uninfected mice (n = 3 per time point) received a single subcutaneous dose of 20 mg/kg of compound 1. Terminal blood and thigh tissue were collected and analyzed by using UPLC with UV detection. Concentrations are expressed in μg/mL. (D) Neutropenic mice (n = 5 per group) were infected intramuscularly with strain N315 and given three subcutaneous doses of compound 1 (20 mg/kg), an OXA (20 mg/kg), or a combination of an OXA and 1 at 1, 9, and 17 h after infection. The infected thigh tissue was collected at 24 h, homogenized, and the homogenate was plated to determine bacterial counts. The data are shown as mean value ± s.e.m. *, p < 0.05; **, p < 0.01; and ***, p < 0.001, as determined by two-sided Tukey’s multiple comparison test (n.s. stands for no statistical significance).

Compound 1 was synthesized by a variation of a published method (see the Supporting Information).²⁰ A checkerboard assay was conducted to assess the properties of the combination of compound 1 and OXA using six S. aureus strains (N315, USA300–0114, 626, MW2, M1712, and HIP13170). The checkerboard experiment reveals if the two compounds are antagonistic (with a fractional inhibitory concentration (FIC) index of >4), indifferent (FIC index >0.5 to ≤4), or synergistic (FIC index of ≤ 0.5) with each other. Compound 1 and OXA were found to be synergistic in the six strains tested. The results for strain N315 (also known as NRS70) are shown in Figure 2A and those for the rest of the strains are presented in Figure S2. Next, we performed time–kill assays with S. aureus strains N315, USA300–0114, and HIP13170 (Figures 2B and Figure S3). For the N315 strain, time–kill curves were investigated with OXA at 32 and 16 μg/mL in combination with compound 1 at 0.6, 1.2, and 2.4 μg/mL. In the combination of OXA:1 at 16:0.6 μg/mL, bacterial growth decreased for 12 h, then increased at 24 h (Figure S3A). Increasing the concentration of compound 1 to 1.2 or 2.4 μg/mL in combination with 16 μg/mL of OXA resulted in ≥3-log₁₀ reduction in S. aureus N315 (Figure 2B), which is consistent with a bactericidal mode of action. For OXA at 32 μg/mL in combination with compound 1 at 0.6 or 1.2 μg/mL, bacterial growth decreased for 9 and 12 h, respectively, before increasing at 24 h. However, in OXA:1 at 32:2.4 μg/mL, a >3-log₁₀ reduction in growth was observed at 24 h (Figure 2B).

We evaluated the distribution of compound 1 to the thighs of mice (the site of infection in the disease model) in a pharmacokinetics (PK) study. Mice were sacrificed at 0.5, 1, 2, 4, and 8 h after a single subcutaneous dose of compound 1 at 20 mg/kg. Maximum concentrations of 2.3 and 0.9 μg/mL were achieved at 2 h in thigh tissue and plasma, respectively. The concentrations of compound 1 at 8 h were 1.4 and 0.4 μg/mL in thigh tissue and plasma, respectively (Figure 2C). The data indicated that dosing every 8 h at 20 mg/kg would achieve sufficient concentrations in the thighs for efficacy.

A mouse neutropenic thigh infection model was used to test the efficacy of compound 1. Prior to infection, mice were made neutropenic by intraperitoneal injection of cyclophosphamide. S. aureus strain N315 was injected intramuscularly into both thighs. Mice were then treated every 8 h for 24 h with the following: vehicle, compound 1 only, OXA only, or a combination of compound 1 and OXA. No statistically significant decrease in bacterial count was seen between the vehicle and compound 1 (p = 0.85). A 1.5-log₁₀ decrease in bacterial growth was seen between the vehicle group and the OXA only (p = 0.047). Bacterial growth decreased by 1.9-log₁₀ in the OXA:1 treatment group compared to the vehicle group (p = 0.006) and by 2.4-log₁₀ compared to compound 1 (p = 0.0001; Figure 2D).

Compound 1 was previously studied as a potential inhibitor of human dual-specificity tyrosine-phosphorylation-regulated kinase 1A (DYRK1A) but did not show any inhibitory activity at 10 μM.²¹ A few kinase inhibitors have shown antibacterial and antibiotic-potentiating activities.^21–24 Therefore, we explored whether compound 1 inhibits any of the three known protein kinases of S. aureus. The three S. aureus kinases are Stk1, Cap5B2, and RsbW.^22–24 While the biological functions of these protein kinases are currently being studied, it is known that all three kinase genes can be knocked out without an obvious phenotypic consequence. None of the kinases is essential for S. aureus. We utilized the strain S. aureus N315_Δstk1. While compound 1 showed antibacterial activity toward the knockout strain (MIC of 2 μg/mL), we found no difference in the potentiation of OXA by 1, compared to the wild-type N315. Next, we tested the strain S. aureus NRS144_{Δstk1/cap5B2/rsbW}—lacking all three kinase genes—transformed with plasmid pCN47-bla. The NRS144 strain was transformed by the plasmid pCN47-bla in order to generate a similar background as N315, as NRS144 does not harbor the bla and mec operons. Again, compound 1 showed antibacterial activity against this strain (MIC of 1 μg/mL), but there was little potentiation (4-fold) of OXA in combination with 1. We have not been able to observe any phenotypic consequences of treatment with compound 1 on both S. aureus protein-kinase knockout strains.

The fluorescent-reporter assay and potentiation results suggest that compound 1 targets BlaR. This assertion was confirmed by surface plasmon resonance (SPR) using the recombinant sensor domain of BlaR, which showed binding to the sensor domain with a K_D of 31 ± 2 μM (Figure 3A,B). Further confirmation came from microscale thermophoresis (MST), producing a K_D of 26 μM.

Figure 3.

Analysis of compound 1 binding to the BlaR sensor domain. (A) Raw SPR sensorgrams and (B) steady-state fit for compound 1 binding to the BlaR sensor domain. (C) The binding pose of compound 1 (in green) to the sensor domain of BlaR using induced-fit docking and energy minimization is depicted.

Repeated attempts to obtain an X-ray structure for the complex of compound 1 and the sensor domain failed to produce diffraction-quality results. To explore the mode of binding of compound 1 to the BlaR sensor domain, we resorted to molecular modeling methods. Our workflow started with induced-fit docking of compound 1 to the antibiotic-binding site of BlaR. This process involved flexible docking of the ligand and the binding-site amino acids, followed by energy minimization of the complex. The compound binds such that N⁶-bromophenyl group interacts with a small hydrophobic subsite. The N⁴-bromophenyl group plays a role in engaging the aromatic Phe421 and Trp424 side chains within the signal transduction loop (L_412–428). Compound 1 forms hydrogen bonds with Gln435, Asn436, and Thr527 (Figure 3C).

The interference with the function of BlaR could also conceivably occur through the inhibition of the cytoplasmic metalloprotease domain. The substrate for the metalloprotease domain is the gene repressor BlaI. We have previously reported a Western blot assay for the turnover of BlaI.²⁵ Membrane fractions of E. coli BL21(DE3) transformed with the pET24a-BlaR plasmid were purified to assess the turnover of BlaI in the presence and absence of compound 1. The proteolytic processing of BlaI was the same in both cases (data not shown); hence, compound 1 did not influence the protease activity.

The fact that compound 1 exhibited anti-S. aureus activity on its own was unexpected (Table S1). The sensor domain of BlaR is phylogenetically²⁶ and structurally^27,28 related to PBPs. Hence, we wondered whether the binding of compound 1 to the sensor domain of BlaR would imply similar binding to key PBPs of S. aureus. We investigated this possibility using four inducible antisense RNA strains for PBPs in the S. aureus COL background, which contains only the truncated (inactive) mec operon. The inducible interference of one specific mRNA for a given PBP at a time would reduce the copy number of that specific PBP, rendering the strain more sensitive to the compound if it targets the given PBP. In the presence of xylose as the inducer, growth was attenuated for the pbp2 and mecA strains (Figure 4A,B), in contrast to the pbp1 and pbp3 (Figure 4C,D). These results indicate that PBP2 and PBP2a are targets of compound 1. To explain it differently, when PBP2 is inhibited, PBP2a becomes essential for the growth of S. aureus. The experiments in Figure 4 merely demonstrate whether an individual PBP is targeted by compound 1 (attenuated specific PBP production weakens cell growth, resulting in increased sensitivity to the compound in these knockdown strains). Both PBP2 and PBP2a are targeted by compound 1. S. aureus cannot survive the inhibition of both PBP2 and PBP2a.

Figure 4.

Analysis of compound 1 binding to the PBPs of S. aureus. Growth curves showing hypersensitivity of S. aureus antisense strains for (A) pbp2 (PBP2), (B) mecA (PBP2a), (C) pbp1 (PBP1), and (D) pbp3 (PBP3) under a xylose-inducible promoter. Compound 1 (5 μM) was evaluated against antisense RNA strains. Bacteria were grown in the absence (open red circles) and presence (closed red circles) of xylose. Bacterial growth was measured using OD₆₀₀. (E) Scanning electron micrographs of S. aureus N315 grown in tryptic soy broth for 2.5 h at 37 °C, then treated with 5 μM compound 1 and 32 μg/mL OXA, followed by an additional hour of growth at 37 °C. Bacteria were fixed for imaging. S. aureus N315 (left to right) grown in the absence of antibiotics (first image), N315 treated with OXA only (second image), N315 treated with compound 1 only (third image), and N315 treated with compound 1 and OXA (fourth image). All images are shown at 50,000× magnification.

We cloned the genes, expressed them, and purified each of the four recombinant PBPs (PBP1, PBP2, PBP2a, and PBP3) of S. aureus. Using nanodifferential scanning fluorimetry (nanoDSF), we documented that compound 1 indeed bound to both PBP2 and PBP2a, exhibiting a shift in melting temperature of 1.4 °C in each case (Figure S4B,C). There was no shift in the melting temperature for PBP1 or PBP3 in the presence of 1 (Figure S4A,D). We attempted to obtain dissociation constants for compound 1 with PBP2 and PBP2a by using SPR, MST, and isothermal titration calorimetry (ITC). However, the solubility limits of compound 1 did not allow evaluation of the compound at a sufficient concentration range to extract the constants.

Inhibition of PBP2 and PBP2a by compound 1 would result in damage to the cell wall. Scanning electron microscopy (SEM) visualized the damage caused by 1 and the OXA treatments. MIC concentrations of OXA, 1, and their combination were used to expose S. aureus N315. Compared with the smooth, spherical shape of healthy S. aureus, the treated cells showed significant damage. The combination caused blebs on the surface of the bacteria and eventually led to cellular death, as evidenced by the collapse of the spherical shape of the bacteria (Figure 4E).

We describe in this report a strategy to shut down the underlying process responsible for resistance to β-lactam antibiotics in MRSA strains (the bla system) using a novel methylpyrazolopyrimidine compound. The mechanism of action of compound 1 was validated as binding to BlaR and related PBPs 2 and 2a. The generality of the process was demonstrated with a collection of 30 strains in both in vitro and in vivo experiments. The compound is not very water-soluble and requires optimization. As with any bioactive compound undergoing evaluation, its off-target effects need to be studied. The axiom that the low-hanging fruit in the field of antibiotics has already been reaped has held true for several decades. Ultimately, the strategies described in this report build on mechanistic knowledge of resistance mechanisms to address unmet medical needs.

MATERIALS AND METHODS

General Information.

Commercial chemicals and solvents were purchased from Millipore Sigma and Ambeed. NMR spectra were recorded at room temperature on a Bruker 400 MHz spectrometer. Bacterial strains were collected from various sources over the years, including BEI Resources. ATCC 29213 was purchased from the American Type Culture Collection.

Synthesis ofN⁴,N⁶-bis(4-bromophenyl)-1-methyl-1H-pyrazolo[3,4-d]pyrimidine-4,6-diamine (1, C₁₈H₁₄Br₂N₆; cas-101606–33-1).

The procedure described here is a variation of a literature method.²⁰ 4,6-Dichloro-1-methylpyrazolo[3,4-d]-pyrimidine (compound 2) was purchased from Ambeed and purified by sublimation at 100 °C under 2 mbar to obtain a white solid prior to use in synthesis. The purity of the compounds was ≥95%, as assessed by LC/MS. The purified 2 (0.58 g, 2.87 mmol, 1.0 equiv) and p-bromoaniline 3 (1.42 g, 8.23 mmol, 2.9 equiv) were placed in a round-bottom flask containing anhydrous ethanol (14 mL). The mixture was heated to reflux, and the crystals dissolved within 1 h. The crude product formed as a white precipitate after around 1.5 h. The suspension was refluxed for a total of 3 h, after which it was cooled to room temperature. The crude product was filtered, and the filtrate was discarded. The precipitate was transferred to a flask containing KOH (0.8 g) in ethanol (8.0 mL). The mixture was refluxed for 15 min and then allowed to cool to room temperature. Excess KOH was filtered off, and the filtrate was collected. Distilled-deionized water (10 mL) was added to the filtrate to initiate the precipitation of the product as a white precipitate. The product was filtered and washed with n-pentane, followed by recrystallization from ethanol, yielding the title compound as a white solid (400 mg, 29% yield). MS [M + H]⁺ 474.95, calculated; found: 474.9695. ¹H NMR (400 MHz, acetone) δ: 7.93 (s, 1H), 7.91–7.78 (m, 4H), 7.47 (m, 4H), 3.91 (s, 3H). ¹³C NMR (101 MHz, acetone): δ 158.99, 156.06, 155.40, 141.29, 139.88, 132.28, 132.14, 131.71, 123.80, 121.77, 115.93, 113.83, 98.28, 33.64.

Fluorescence Reporter Assay.

Plasmid pCN52RI-Venus was previously reported.¹⁷ For screening, NRS144_{pCN52RI-Venus} was grown in the presence of 50 μg/mL erythromycin. The overnight culture was diluted 150-fold in fresh LB medium supplemented with 50 μg/mL erythromycin and was allowed to grow to an OD₆₀₀ of 0.4 (approximately 2 h). A 95-μL aliquot of the growth was transferred to each well of a 96-well black-sided, clear flat-bottom microplate containing 10 μg/mL of 2-(2-carboxyphenyl)-benzoyl-6-aminopenicillanic acid (CBAP) and 25 μM of the test compound. The suspension was incubated at 37 °C for 2 h. Fluorescence and OD₆₀₀ of the samples were measured by using the Cytation 5 Microplate Reader (Agilent, Santa Clara, CA).

Minimal Inhibitory Concentration (MIC) Determination and Potentiation Assays.

The MIC determination using the microbroth-dilution method was performed by following the Clinical and Laboratory Standards Institute guidelines.²⁹ The MIC was determined using 2-fold serial dilutions of each compound in cation-adjusted Mueller-Hinton II broth (with and without 2% NaCl for S. aureus when tested with OXA) with 5 × 10⁵ CFU/mL of bacteria. All experiments were performed in triplicate by using 96-well plates. Samples were incubated at 37 °C for 18 or 24 h (for OXA). The MIC was defined as the concentration at which bacterial growth was not observed.

Checkerboard Assay.

The behavior of compound 1 with OXA was assessed against MRSA strains using a checkerboard assay, following a previously published procedure.³⁰ The fractional inhibitory concentration (FIC) index was calculated as reported earlier.³¹

Time–Kill Assay.

MRSA strains N315, USA300–0114, and HIP13170 were grown overnight on tryptic soy agar plates. Bacteria were diluted in Mueller–Hinton broth to 10⁸ CFU/mL. A 5 mL aliquot of Mueller–Hinton broth containing 10⁶ CFU/mL of each MRSA strain, along with compound 1 and OXA at various concentrations, was incubated at 37 °C. Samples were taken at 0, 3, 6, 9, 12, and 24 h. They were serially diluted and cultured on tryptic soy agar (TSA) plates for bacterial counts. This assay was performed in triplicate.

Animals.

Female, outbred 8-week-old Institute of Cancer Research (ICR) mice were purchased from Envigo and used for both the pharmacokinetic (PK) and neutropenic thigh infection studies. The average weight of the mice was 26.4 g (range 21.6–30.5 g) for the PK study and 26.5 g (range 23.6–30.3 g) for the neutropenic thigh infection model. For the PK study, mice were housed in polycarbonate shoebox cages, and in disposable cages for the infection model study. The cages contained 1/4-in. corncob bedding (The Andersons) and Alpha-Dri bedding (Shepherd Specialty Papers). Mice were exposed to a 12-h light/dark cycle, and the room was kept at 72 °F. The mice were provided sterile food (Teklad 2918 irradiated extruded rodent diet) and water ad libitum. All procedures during the PK and infection model studies were performed in compliance with, and with approval by the University of Notre Dame Institutional Animal Care and Use Committee (IACUC protocols 23–01-7605 and 23–01-7633).

Pharmacokinetic Study.

Mice (n = 15) were given a single 20 mg/kg subcutaneous dose of compound 1 (a 100 μL aliquot of a 6.0 mg/mL solution in 10% DMSO, 20% Tween-80, and 70% sterilized water). Mice were sacrificed at 0.5, 1, 2, 4, and 8 h. Terminal blood samples were collected by cardiac puncture into heparinized (sodium heparin, Sigma-Aldrich) syringes. Plasma was obtained via centrifugation of samples at 20,000g for 15 min at 4 °C. Samples were stored at −80 °C until processed. A 60-μL aliquot of plasma was mixed with 60 μL of a 10 μg/mL solution of the internal standard in acetonitrile (structure 4, obtained from the NCI). The internal standard was chosen due to its structural similarity to compound 1. The activity of compound 4 was significantly poorer compared to that of compound 1 in general. The sample was centrifuged at 17,000g for 10 min. A 60-μL aliquot of the supernatant was used for UPLC analysis with UV detection.

Thigh tissue was excised, weighed, and mixed with 100 μL of a 10 μg/mL internal standard in acetonitrile. Acetonitrile (900 μL) was added, and samples were homogenized in a Bullet Blender (Next Advance). The samples were centrifuged at 17,000g for 10 min, and the supernatant was collected. The extraction was repeated two additional times. The supernatants were combined and dried in a SpeedVac. The residue was then dissolved in 100 μL of acetonitrile and centrifuged at 17,000g for 10 min. A 60-μL aliquot was transferred to an autosampler vial for analysis.

A Waters Acquity UPLC (Waters Corp.) coupled with a diode-array detector with ultraviolet detection at 285–295 nm was used to analyze both plasma and thigh samples. A C18 Atlantis T3 column with a 2.1 mm inner diameter × 100 mm length, 3-μm particle size column was used at a flow rate of 0.4 mL/min. The following gradient was used: at 1 min, 80%A/20% B; 2.5 min, linear gradient from 80%A/20%B to 25%A/75%B; 8.5 min, to 100%B and 9 min, to 80%A/20%B. Solution A was 0.1% formic acid in water, and solution B was 0.1% formic acid in acetonitrile. The internal standard and compound 1 eluted at 3.6 and 3.8 min, respectively.

The calibration curves were prepared using known concentrations of compound 1 (ranging from 0.2 to 5 μg/mL) and an internal standard in blank control thigh tissue or control plasma. Concentrations of compound 1 were determined using the calibration curve’s linear regression parameters relative to the internal standard.

Neutropenic Thigh-Infection Studies.

Mice (n = 5 per group; a total of 30) were made neutropenic by injection of cyclophosphamide (Sigma-Aldrich) intraperitoneally at 200 mg/kg (100 μL of a 60 mg/mL solution in sterile saline) on days −4 and −1 of the experiment. S. aureus N315 was grown overnight on TSA plates and diluted to 1 × 10⁶ CFU/mL in brain-heart infusion (BHI) broth. Mice were infected with a 100 μL injection of the bacterial inoculum intramuscularly in both the right and left thighs on day 0. 1 h after infection, the mice received a subcutaneous injection of vehicle, 20 mg/kg of compound 1, 20 mg/kg of OXA (sodium oxacillin, TCI Chemicals), or a combination of OXA and compound 1. Treatments were given at 1, 9, and 17 h post-infection. Compound 1 and OXA were prepared in 10% DMSO, 20% Tween-80, and 70% sterile water. A vehicle group was included as a control. The mice were sacrificed at 24 h post-infection. Both thighs were excised aseptically, weighed, and homogenized with four volumes of phosphate-buffered saline, pH 7.5, in a Bullet Blender (Next Advance) at 4 °C, then centrifuged at 10,000g for 10 min. A 20-μL aliquot of supernatant was transferred to 180 μL of 0.1% agar saline and then serially diluted to 10⁻⁸ and plated on TSA plates in triplicate. The colony numbers were counted after a 24-h incubation at 37 °C.

Purification of the Sensor Domain of BlaR.

Competent E. coli BL21(DE3) were transformed with the previously cloned plasmid pET24a-BlaR^s. A culture of 50 mL LB containing 50 μg/mL kanamycin was grown overnight at 37 °C. A 10 mL aliquot of the overnight culture was used to inoculate 1 L of fresh LB. The culture was grown at 37 °C for 2 h, or until an OD₆₀₀ of 0.6 was reached. The culture was then induced by 0.4 mM isopropyl β-D-thiogalactopyranoside (IPTG) and allowed to grow overnight at 25 °C. Bacteria were harvested by centrifugation at 7,000g for 20 min. The pellet was resuspended in 30 mL of 25 mM TES, pH 7.4, and the bacteria were lysed by two passes through the French press. A second centrifugation step at 18,000g for 1 h was used to remove cell debris. The supernatant was loaded onto a high-S cation-exchange column equilibrated with 25 mM TES, pH 7.4. A 300 mL gradient of 25 mM TES, pH 7.4, containing 1 M NaCl was used for purification. The elution fractions were concentrated and run over an S200 size-exclusion column equilibrated with 10 mM sodium phosphate, pH 7.4, containing 150 mM NaCl. The protein was concentrated using an Amicon centrifugation filter with an MWCO of 10 -kDa. An SDS-PAGE gel was used to confirm protein purity, and the samples were aliquoted for freezing at −80 °C.

Binding of Compound 1 to BlaR Sensor Domain.

A Biacore X100 surface plasmon resonance instrument (Cytiva Life Sciences) was used. The BlaR sensor domain (1 μM in MES buffer, pH 6) was immobilized on the surface of flow cell 2 of a CM5 sensor chip using the amine-coupling kit (Cytiva), while flow cell 1 was blank-immobilized as described previously.¹⁷ Compound 1 was serially diluted (2-fold, 100 μM to 6.25 μM) in the HBS-EP running buffer, pH 7.4, containing 150 mM NaCl, 10 mM HEPES, 3 mM EDTA, 0.005% Tween-20, and 4% DMSO. Each concentration of compound 1 was injected for 180 s at a flow rate of 30 μL/min during the binding phase and allowed to dissociate for 600 s. After dissociation, a single pulse of 0.05% SDS was injected (10 μL/min for 30 s) to regenerate the chip surface and return to baseline. All SPR experiments were performed in triplicate. The Biacore X100 Evaluation software was used for data analysis. A blank injection (no compound) was subtracted from all analytical injections during the data analysis. Data were fit to the steady-state binding equation of the software:

$$R_{\mathrm{e}\mathrm{q}}=\frac{R_{\mathrm{m}\mathrm{a}\mathrm{x}}}{1+\frac{K_{\mathrm{D}}}{C}}$$

where $R_{\mathrm{e}\mathrm{q}}$ is the equilibrium SPR response, $R_{\text{max}}$ is the response at saturation of binding, $K_{\mathrm{D}}$ is the equilibrium dissociation constant, and $C$ is the compound concentration.

Fluorescent Labeling of BlaR Sensor Domain for Microscale Thermophoresis (MST).

NHS-Cy5 dye (50 μM) was mixed with an aliquot of BlaR sensor domain (10 μM, total reaction volume 100 μL). The mixture was incubated for 1 h at room temperature in the dark. The free dye was separated from the labeled BlaR sensor domain by gravity flow over an Econo-Pac column (Bio-Rad). The labeled protein was used immediately or flash-frozen and stored at −80 °C. A fixed concentration of labeled protein (10 nM) was incubated with increasing concentrations of compound 1 (0.1 μM to 100 μM). Samples were analyzed by using a Monolith NT.115pico instrument (NanoTemper).

Induced-Fit Docking with BlaR Sensor Domain.

All molecular modeling was carried out with the Schrödinger Suite 2024–4. The coordinates of the sensor domain of BlaR bound to a noncovalent ligand were obtained from the Protein Data Bank (PDB-I: 8C0S).¹⁷ The structure was prepared using the Protein Preparation Wizard at pH 7.4, with optimized hydrogen-bond assignment and water deletion.³² An energy-minimization step using the OPLS4 force field with all-atom constraints at 0.3 Å was performed to relieve clashes. Compound 1 was prepared using the LigPrep protocol, also employing the OPLS4 force field. Induced-fit docking³³ was performed using Glide,³⁴ and conformations of residues within 5 Å of the docked ligand poses were optimized. Glide redocking into the optimized protein conformation was carried out using standard precision.

Cloning, Overexpression, and Purification of BlaI.

The blaI gene from plasmid pET-24d-BlaI²⁵ was amplified using Q5 DNA polymerase (NEB) with primers NcoIBlaIF and XhoIBlaIR to insert the NcoI and XhoI restriction sites at the ends of the blaI gene (Table S3). Both the PCR product and the pET-24d plasmid (containing an N-terminal His-tag) were digested by using NcoI and XhoI restriction enzymes, purified with DNA gel extraction (Zymo Research), and ligated. E. coli DH5α was transformed with the ligation mixture and grown on LB agar plates containing 50 μg/mL kanamycin. Plasmids from multiple colonies were selected and purified for full plasmid sequencing (Plasmidsaurus). The construct was named pET-24d-BlaI-His. The plasmid was used to transform E. coli BL21(DE3) for overexpression. A 50 mL culture of LB medium containing 50 μg/mL kanamycin was inoculated and grown at 37 °C overnight, shaking at 200 rpm. A 25 mL aliquot of the overnight culture was added to 1 L of fresh LB containing 50 μg/mL kanamycin. Bacteria were grown to an OD₆₀₀ of 1.0 and induced with 0.5 mM IPTG for expression at 37 °C for 4 h. Bacteria were harvested by centrifugation at 7,000g for 15 min and frozen at −80 °C. The bacterial paste was resuspended in 50 mM HEPES, pH 7.5, 300 mM NaCl, 3 μg/mL DNaseI, and 1X protease inhibitor, then subjected to two passes through the French Press. The cell debris was collected by centrifugation at 18,000g for 45 min. The supernatant was incubated with Ni-NTA resin for 1 h at 4 °C, before washes with 10 column volumes of buffer A (50 mM HEPES, pH 7.5, 300 mM NaCl, 20 mM imidazole). The protein was eluted using a 100 mL imidazole gradient (buffer B: 50 mM HEPES, pH 7.5, 300 mM NaCl, 500 mM imidazole). An S200 size-exclusion column was used as the final purification step. After concentration using a 10-kDa MWCO Amicon centrifugation filter, the yield of homogeneous BlaI was 2.3 mg from 1 L of culture.

Cloning, Overexpression, and Purification of PBP1, PBP2, PBP2a, and PBP3.

The genes for PBP2, PBP2a, and PBP3 (excluding the sequences for the signal peptides) were cloned from S. aureus genomic DNA. The plasmid for PBP1 (pET24a-PBP1CN) was generated previously, and the purification procedure was followed as described.³⁵ The pET28a-TEV-His tag plasmid was amplified using Q5 DNA polymerase (NEB) and the following primers for each PBP: PBP2, pET28aPBP2TEVF and pET28aPBP2TEVR; PBP2a, pET28aPBP2aTEVF and pET28aPBP2aTEVR; and PBP3, pET28aPBP3TEVF and pET28aPBP3TEVR (Table S3). The genes for each PBP were amplified using Q5 DNA polymerase (NEB) and the following primers for each gene: PBP2, pET28aPBP2F and pET28aPBP2R; PBP2a, pET28aPBP2aF and pET28aPBP2aR; and PBP3, pET28aPBP3F and pET28aPBP3R (Table S3). The PCR products for both the plasmid and the gene were treated with DpnI and subjected to a DNA cleanup kit (Zymo Research). The plasmid and the gene were assembled using the NEBuilder HiFi DNA Assembly Kit (NEB). E. coli DH5α cells were transformed with the HiFi assembly mixture and plated on LB agar plates containing 50 μg/mL kanamycin. Plasmids from multiple colonies were selected and purified for full plasmid sequencing (Plasmidsaurus). The plasmids were named as follows: pET28a-PBP2(Δ57), pET28a-PBP2a(Δ24), and pET28a-PBP3(Δ58).

The newly generated plasmids were used to transform E. coli LOBSTR-RIL for overexpression. A 50 mL culture of LB media containing 50 μg/mL kanamycin was inoculated and grown at 37 °C overnight shaking at 200 rpm. A 10 mL aliquot of the overnight culture was added to 1 L of fresh LB containing 50 μg/mL kanamycin. Bacteria were grown to an OD₆₀₀ of 0.6, at which point they were induced with 0.5 mM IPTG at 16 °C for 18 h. Bacteria were harvested by centrifugation at 7,000g for 15 min. The paste was resuspended in buffer containing 10 mM Tris, pH 8.0, 300 mM NaCl, 100 μg/mL DNase I, and 1X protease inhibitor, and subjected to two passes through the French Press. The cell debris was collected by centrifugation at 18,000g for 1 h. The supernatant was incubated with Ni-NTA resin for 1 h at 4 °C, before washes with 10-column volumes of buffer A (10 mM HEPES, pH 8.0, 300 mM NaCl, and 0.05% DDM for PBP2 only). Protein was eluted using a 100 mL imidazole gradient (buffer B: 10 mM Tris, pH 8.0, 300 mM NaCl, 500 mM imidazole—supplemented with 0.05% DDM for PBP2 only). Proteins were concentrated and buffer-exchanged into buffer A using a 30-kDa MWCO Amicon centrifugation filter. The yield of each protein from 1 L of culture was as follows: PBP1, 6.5 mg; PBP2, 8.4 mg; PBP2a, 7.2 mg; and PBP3, 9.3 mg.

BlaI Cleavage Assay by BlaR.

E. coli membranes harboring the full-length BlaR were prepared using a previously published procedure.²⁵ The membrane fraction containing BlaR (50 μg) was incubated with 10 μg of purified BlaI. As a negative control, a sample of BlaR membranes was dialyzed overnight in 100 mM sodium phosphate (pH 7.5), 50 mM sodium bicarbonate, and 5 mM EDTA to generate a zinc-free BlaR sample prior to incubation with purified BlaI. Sample mixtures were incubated for 1 h at 37 °C. The reaction was stopped by the addition of SDS loading dye. For detection of BlaR and BlaI, samples were separated in 12% Tris SDS-PAGE gels. After electrophoresis, the proteins were transferred onto a nitrocellulose membrane (GenScript, catalog no. 1704085) using the Transblot Turbo Transfer system. Filter paper was soaked in anode buffer (1X Tris/CAPS buffer, 15% methanol, 75% water) and cathode buffer (1X Tris/CAPS buffer, 0.1% SDS, 90% water). The membrane was blocked for 1 h in 1X TBS-T (20 mM Tris-HCl, pH 7.5, 150 mM NaCl, 0.1% Tween-20) containing 3% bovine serum albumin and 3% nonfat milk. Next, the membrane was rinsed with 1X TBS-T and incubated for 1.5 h with the anti-His antibody (200-fold dilution). Lastly, the membrane was washed three times with 1X TBS-T buffer and incubated with Pierce ECL2 Western Blotting Substrate for 5 min. Images were acquired using a ChemiDoc MP Imaging System (Bio-Rad) under chemiluminescence detection.

Growth Curve of S. aureus COL Strains Expressing Antisense RNA.

S. aureus COL strains harboring an antisense RNA expression system interfering with mRNAs for PBP1, PBP2, PBP2a, or PBP3 were cultured on LB agar plates containing 34 μg/mL chloramphenicol at 37 °C overnight. A fresh LB culture was inoculated and grown for approximately 2 h for the bacteria to reach the logarithmic phase of growth (OD₆₀₀ = 0.5). Fresh LB media were prepared containing 34 μg/mL chloramphenicol, with or without 50 mM xylose. A 100-μL aliquot of the media was added to each well of a 96-well microplate. The concentration of bacteria was adjusted to 10⁷ CFU/mL, of which 5 μL was added to each well. Each strain was incubated with 5 μM compound 1, with or without 50 mM xylose (inducing agent). The plate was incubated at 37 °C shaking for 24 h. Absorbance measurements at 600 nm were taken each hour. GraphPad Prism 10.1.0 was used to plot Time (h) versus OD₆₀₀.

Assessment of Binding of Compound 1 to PBPs by nanoDSF.

Compound 1 was incubated with PBP1, PBP2, PBP2a, or PBP3 (1.0 mg/mL) in a HBS-EP buffer (10 mM HEPES, pH 7.4, 150 mM NaCl, 3 mM EDTA, 0.005% Tween-20) supplemented with 5% DMSO for 1 h on ice. A blank sample was prepared containing only protein without the compound. Using a Tycho NT.6 instrument (NanoTemper), samples were heated from 35 to 95 °C, and intrinsic tryptophan fluorescence emission was monitored at 330 and 350 nm. All measurements were performed in triplicate. The inflection point in the fluorescence ratio (350/330 nm) was taken as the melting temperature of the protein. GraphPad Prism 10.1.0 was used to plot the first derivative of the fluorescence ratio (350 nm/330 nm) versus temperature (in °C).

Scanning Electron Microscopy.

S. aureus N315 was cultured on an LB agar plate at 37 °C overnight. A 75 mL tryptic soy broth culture was inoculated to achieve an OD₆₀₀ = 0.1 and was grown at 37 °C shaking for 2.5 h (until OD₆₀₀ = 0.5). The culture was divided into 5 mL aliquots and treated with 5 μM compound 1; or 32 μg/mL OXA; or 5 μM compound 1 with 32 μg/mL OXA. The cultures were grown for 1 h at 37 °C with shaking. Bacteria were fixed following a previously published procedure.³⁶ Microscopy and imaging were performed on a Magellan 400 XHR instrument.

Supplementary Material

ASSOCIATED CONTENT

Supporting Information

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsinfecdis.5c00685.

Potentiation results of compound 1 for S. aureus (Table S1), potentiation results of compound 1 for Gram-negative and Gram-positive bacteria (Table S2), primers used in this study (Table S3), NMR and HPLC spectra (Figure S1), checkerboard graphs (Figure S2), time–kill assay plots (Figure S3), and nanoDSF plots (Figure S4) (PDF)

ABBREVIATIONS

MSSA

methicillin-sensitive Staphylococcus aureus

MRSA

methicillin-resistant Staphylococcus aureus

PBP

penicillin-binding protein

LB

lysogeny broth

CBAP

2-(2-Carboxyphenyl)-benzoyl-6-aminopenicillanic acid

DMSO

dimethyl sulfoxide

OXA

oxacillin

VISA

vancomycin-intermediate Staphylococcus aureus

VRSA

vancomycin-resistant Staphylococcus aureus

MIC

minimum inhibitory concentration

FIC

fractional inhibitory concentration

PK

pharmacokinetic

DYRK1A

dual-specificity tyrosine-phosphorylation-regulated kinase 1A

SPR

surface plasmon resonance

MST

microscale thermophoresis

nanoDSF

nanodifferential scanning fluorimetry

ITC

isothermal titration calorimetry

SEM

scanning electron microscopy

TSA

tryptic soy agar

ICR

Institute of Cancer Research

IACUC

Institutional Animal Care and Use Committee

BHI

brain-heart infusion

IPTG

isopropyl β-D-thiogalactopyranoside

PDB

protein data bank

TSB

tryptic soy broth