Screening, synthesis, in vitro potency, and structure-activity relationship

Screening a small and focused library of anti-infectives yielded a single series of very potent antibiotic macrocycles of the rifampicin class. After synthetic modifications, several analogs assayed inhibited growth of S. aureus both in a standard broth assay and an intracellular killing assay at sub-micromolar to nanomolar minimum inhibitory concentration (MIC) concentrations. We believe that the intracellular assay provides us with a measure of penetration, and thus, compounds to push forward into animals. The most potent and instructive compounds are listed in Table 1.

Table 1 Cellular activity of rifampicin-based antibiotics

Synthesizing and assaying rifampicin analogs, we were able to remove the labile imine functionality and replace it with a stable benzo-oxazine ring system (rigidifying the molecule) and introduce the basic amine found in many potent antibiotics [6]. Although rifampicin 1 is potent in the broth assay, it lacks activity in the intracellular assay (Table 1). Our analogs (compounds 2, 3, 4, and 5) gain potency with the introduction of the fused 5 ring system and the alkylation of the basic amine in the broth assay. As the primary amine protons are replace with methyl groups, a log fold increase in potency is realized in the broth assay. However, the only compounds that gain potency in the intracellular killing assay are 4 and 5, both tertiary amines, as expected.

Compound 6 shows the fine selectivity of the bacteria when a small change in the structure is introduced. By shifting the tertiary amine tail by one carbon position, it reduces the intracellular potency to rifampicin levels and causes a log reduction in the broth assay.

Taking rifampicin, compounds 4 and 5 into Gram-negative Salmonella strains AR-0031 and ATCC 14028, we see activity in both the broth and intracellular assays (Table 1). Although, not as potent as observed against S. aureus, they have an appreciable activity over and above rifampicin. Given these results, we believe that compounds 4 and 5 can have broader antibacterial properties.

In vivo efficacy [7,8,9,10]Median S. aureus NRS384 kidney burden in mice treated with rifampicin in combination with vancomycin

In this experiment, mice infected with S. aureus MSRA strain NRS384 were treated with vancomycin alone (control) or vancomycin in combination with 0.002 mg/kg to 25 mg/kg rifampicin.

As shown in Fig. 1, intravenous (IV) infection with S. aureus MRSA strain NRS384 results in high bacterial burden in the kidneys. Vancomycin treatment alone results in a ~ 2–3 log reduction in S. aureus kidney burden. Combination treatment of rifampicin with vancomycin further reduces kidney burden. Rifampicin was effective in combination with vancomycin at 25 mg/kg and perhaps 0.25 mg/kg, but efficacy was reduced to vancomycin alone levels at lower doses.

Fig. 1

Mice infected with S. aureus MRSA strain NRS384 then treated with vancomycin (110 mg twice daily) and rifampicin. ***p < 0.001, *p < 0.05, ns = not significant (t-test). Figure adapted from PCT WO 2022204499

Median S. aureus NRS384 kidney burden in mice treated with rifamycin analogs in combination with vancomycin

In this experiment, mice infected with S. aureus MSRA strain NRS384 were treated with vancomycin alone (control) or vancomycin in combination with 0.01–0.25 mg/kg of compounds 4 and 5.

As shown in Fig. 2, IV infection with S. aureus MRSA strain NRS384 results in high bacterial burden in the kidneys. Vancomycin treatment alone resulted in a ~ 2–3 log reduction in S. aureus kidney burden. Combination treatment of compounds 4 and 5 with vancomycin further reduced kidney burden.

Fig. 2

Mice infected with S. aureus MRSA strain NRS384 then treated with vancomycin (110 mg twice daily) and compounds 4 and 5. p < 0.0001 (two-way ANOVA). Figure adapted from PCT WO 2022204499

Compounds 4 and 5 in combination with vancomycin are effective at reducing S. aureus kidney burden at lower doses than rifampicin in combination with vancomycin. For example, compound 5 significantly reduced S. aureus kidney burden to below the limit of detection even at the lowest tested dose of 0.01 mg/kg.

Median S. aureus NRS384 kidney burden in mice treated with rifampicin or compound 5 with or without combination with vancomycin

In this experiment, mice infected with S. aureus MSRA strain NRS384 were treated with 25 mg/kg rifampicin (control) or 0.002–0.25 mg/kg compound 5 in combination with or without vancomycin.

As shown in Fig. 3, IV infection with S. aureus MRSA strain NRS384 results in high bacterial burden in the kidneys. Vancomycin treatment alone resulted in a ~ 2–3 log reduction in S. aureus kidney burden. Rifampicin reduced kidney burden at 25 mg/kg when combined with vancomycin but was ineffective as a monotherapy. Compound 5 significantly reduced S. aureus kidney burden by ~3 logs at 0.25 mg/kg when administered as a monotherapy.

Fig. 3

Mice infected with S. aureus MRSA strain NRS384 and treated with or without vancomycin (110 mg twice daily) and compound 5. ns = not significant, **p < 0.005 (two-way ANOVA). Figure adapted from PCT WO 2022204499

As shown in Fig. 3, compound 5 is highly effective at low doses, including 0.05 and 0.002 mg/kg, at reducing the bacterial load to below the limit of detection when given in combination with vancomycin. Compound 5 in combination with vancomycin is more effective than either monotherapy, i.e., than compound 5 alone or vancomycin alone, in reducing the bacterial burden.

As shown in Fig. 3, the efficacy of compound 5 is enhanced by its use in combination with vancomycin. The efficacy of vancomycin is enhanced by its use in combination with compound 5.

Median S. aureus N315 kidney burden in mice treated with rifampicin or compounds 4 and 5 in combination with vancomycin

Similar results were observed using another MRSA strain. In this experiment, mice infected with S. aureus MSRA strain N315 were treated with 25 mg/kg rifampicin (control), or 0.01–0.75 mg/kg compound 4, or 0.002–0.25 mg/kg compound 5 in combination with vancomycin.

As shown in Fig. 4, IV infection with S. aureus MRSA strain N315 results in high bacterial burden in the kidneys. Vancomycin treatment alone resulted in a ~ 4 log reduction in S. aureus kidney burden. Combination treatment of 25 mg/kg rifampicin or 0.01–0.75 mg/kg compound 4 or 0.002–0.25 mg/kg compound 5 with vancomycin further reduced kidney burden to or near the limit of detection for the experiment.

Fig. 4

Mice infected with S. aureus MRSA strain NRS315 and treated with vancomycin (110 mg twice daily) and compounds 4 and 5. ns = not significant,**p < 0.005 (two-way ANOVA). Figure adapted from PCT WO 2022204499

Both leads, compounds 4 and 5, were efficacious in combination with vancomycin to reduce the bacterial load to below the limit of detection at a dose of 0.01 mg/kg and 0.002 mg/kg, respectively, although the scatter in the vancomycin control group limited the statistical analysis.

ADME

Compounds 4 and 5 were evaluated in standard assays for metabolic stability in rodent S9 microsomes, human plasma protein binding, permeability, and transporter activity, along with human CYP inhibition and induction (Tables 2–4). As shown in Table 2, plasma protein binding is in the range of 80-98% with compound 4 having the least binding. These values compare well to the approved antibiotics Rifampicin or Rifabutin. Table 2 also shows the rodent S9 microsomal stability for the compounds. Compounds 4 and 5 have half-lives greater than 120 min, which compares well to the approved antibiotics. Both compounds 4 and 5 exhibit no in vitro clearance in rat and mouse over 2 h which corresponds to low hepatic extraction. In Table 3 the permeability and transporter activity of compounds 4 and 5 are listed. Both leads have very good CACO-2 permeability ratios that rests between the values determined for rifabutin and rifampicin. As for the MDCK-MDR1 assay, again efflux ratios reside between the values determined for rifabutin and rifampicin. These values indicate that the compounds are effluxed by transporters. Finally, Table 4 shows the results for the CYP inhibition and induction activity for compounds 4 and 5. Both leads inhibit all CYP isozymes tested, except for compound 4, at uM values. Inhibition of CYP isozymes can potentially cause drug-to-drug interactions if the antibiotic is dosed high enough. This contrasts with both rifampicin and rifabutin that mainly are devoid of that inhibition. Table 4 also contains CYP induction activity for compounds 4 and 5. Neither lead compound induces the CYP isozymes tested. However, rifabutin and rifampicin are strong inducers of these 3 CYP isozymes. The hERG potassium current was estimated to be >3 uM. The positive control, Cisapride, inhibited hERG current at 90 nM by 87.6%. It would not be expected that either lead would cause long QT syndrome.

Table 2 Human plasma protein binding and rodent S9 microsomal stability for compounds 4 and 5Table 3 Permeability and transporter activity for compounds 4 and 5Table 4 CYP inhibition and induction activity for compounds 4 and 5Toxicity

To investigate the toxicity of these antibiotic leads, a 2 week IV once-a-day (QD) dosing study was performed in Sprague Dawley rats to compare compounds 4 and 5 (see Table 5 for the study design). For compound 4, approximately dose-proportional with the trend of less than dose proportional increase in Cmax and AUCtau was observed at low dose. The AUC or Cmax of compound 4 was not impacted when coadministered with vancomycin. For compound 5, greater than dose proportional increase in Cmax between 0.1 and 10 mg/kg and AUCtau between 1 and 10 mg/kg groups was observed. The AUC or Cmax of compound 5 was not impacted when coadministered with vancomycin, except for Day 1 when a slightly greater Cmax and AUC were observed on Day 1 in the animals coadministered compound 5 and vancomycin compared to the animals administered compound 5 only. No apparent accumulation was observed following 14 daily doses. No sex difference in exposure was observed. No apparent accumulation was observed following 14 daily doses. There were no body weight, body weight gains, or food consumption changes that were considered related to rifampicin, vancomycin, compounds 4 or 5 given alone, or in combination with vancomycin.

Table 5 Rat toxicity study design for once-a-day (QD) dosing of compounds 4 and 5

At the end of the treatment period, hematology changes were limited to minimal increases in reticulocyte counts (compound 4 at ≥0.1 mg/kg/day and compound 5 at 10 mg/kg/day). Changes observed in males given the combination of compounds 4 or 5 with vancomycin consisted of minimal decreases in red cell mass parameters (red blood cells, hemoglobin, and hematocrit), mild decreases in neutrophils counts and minimal increase in platelet counts. These changes were also noted in males given vancomycin alone and were therefore considered vancomycin-related. Compound 4 related changes in coagulation parameters was limited to minimal increases in fibrinogen in males given the combination, and were also noted in males given vancomycin alone, as such, the changes were attributed to vancomycin. Clinical chemistry changes consisted of minimal increases in phosphorus (compound 4 at ≥0.1 mg/kg/day), minimal decrease in albumin concentration (compound 4 at 10 mg/kg/day and the combination of compounds 4 or 5 with vancomycin), minimal increase in glucose concentration (compound 4 at 1 mg/kg/day and the combination of compound 4 with vancomycin) and minimal increase in urea nitrogen (combination of compound 4 with vancomycin). For compound 4, at the end of the recovery period, additional changes observed consisted of minimal to mild decreases in globulin in males at ≥0.1 mg/kg/day and in males given the combination. In addition, there was an apparent minimal increase in alkaline phosphatase in males at ≥1 mg/kg/day and increase in glucose at ≥0.1 mg/kg/day. Additional changes observed in males given the combination included a minimal decrease in calcium after the 14-day recovery. Changes in coagulation parameters were no longer observed following the 14-day recovery period. For compound 5, at the end of the recovery period all of the hematology changes were no longer observed with the exception of persistent increase in reticulocytes and red blood cell distribution width in males given 10 mg/kg/day and there were apparent minimal increases in platelet counts in males at ≥1 mg/kg/day and in males given the combination. There was an apparent minimal prolongation of activated partial thromboplastin time in males previously given 10 mg/kg/day following the 14-day recovery period. Decrease in globulin/increase in albumin/globulin ratio in males given 1 mg/kg/day, were still noted at the end of 14-day recovery period. A minimal decrease in alanine aminotransferase activity and mild increase in triglyceride concentrations were observed in males given the combination following the 14-day recovery period.

None of the compound 4 or 5 related findings were considered adverse. Overall, compounds 4 and 5 were well-tolerated. The no-observed-adverse-effect level (NOAEL) was considered to be 10 mg/kg/day, the highest dose administered for compound 4 and 5.