SPR741, an Antibiotic Adjuvant, Potentiates the In Vitro and In Vivo Activity of Rifampin against Clinically Relevant Extensively Drug-Resistant Acinetobacter baumannii

ABSTRACT Acinetobacter baumannii is responsible for 10% of all nosocomial infections and has >50% mortality rates when causing ventilator-associated pneumonia. In this proof-of-concept study, we evaluated SPR741, an antibiotic adjuvant that permeabilizes the Gram-negative membrane, in combination with rifampin against AB5075, an extensively drug-resistant (XDR) A. baumannii strain. In standard in vitro assays and in a murine pulmonary model, we found that this drug combination can significantly reduce bacterial burden and promote animal survival despite an aggressive infection.

A cinetobacter baumannii gained notoriety as the bacterial species most frequently isolated from U.S. soldiers from 2004 to 2010, with Ͼ3,500 infections associated with war wounds (1)(2)(3)(4). Presently, A. baumannii is a significant problem worldwide because of extensively drug-resistant (XDR) strains and the aggressive nature of some infections causing ventilator-associated pneumonia (5)(6)(7). Recently, the World Health Organization cited a critical need for A. baumannii research because of increased drug resistance and lack of treatments (8). To address the immediate need, research strategies such as antibiotic adjuvants, which involve the current antibiotic armamentarium, may provide a faster path to clinical application.
Antibiotic adjuvants are typically small molecules that sensitize a bacterium to clinically approved antibiotics (9). Specific examples with respect to A. baumannii include 2-aminoimidazole-based compounds, which disrupt two-component signaling (10), and anthracyclines, which potentiate activity of rifampin and linezolid (11). Here, we evaluated SPR741 (formerly NAB741), a polymyxin-B-derived molecule specifically designed to minimize the nephrotoxicity associated with this antibacterial class (12). SPR741 has reduced positive charge and lacks the highly lipophilic fatty-acid side chain present in polymyxins, which are the two structural features responsible for clinical nephrotoxicity (12,13). As a proof of concept, we tested a combination of SPR741 and rifampin because colistin and rifampin have proven to be an effective combination in both mouse models and in patients (14)(15)(16)(17)(18). Rifampin was also previously tested against AB5075, a highly virulent XDR strain, in a murine pulmonary model of infection (19), which facilitated dosing for this study.
First, the SPR741/rifampin combination was tested in vitro against A. baumannii. The MICs for SPR741 and rifampin were determined for AB5075 to be 128 g/ml and 4.0 g/ml, respectively, using standard CLSI methods in cation-adjusted Mueller-Hinton broth (CAMHB) (20). This combination was assessed with the checkerboard method to determine fractional inhibitory concentration (FIC) (21), where synergistic activity was defined by an FIC of Յ0.5 (22). The MIC of rifampin dropped from 4.0 to 0.5 g/ml in the presence of 2.0 g/ml SPR741, an 8-fold reduction, thus producing an FIC of 0.14 and indicating synergy. An isobologram was generated from these results (Fig. 1A). Next, we examined whether this synergy was applicable across the whole species of A. baumannii. We analyzed a previously described 28-strain diversity set (19) and determined MICs for all strains. A combination of 4.0 g/ml of SPR741 and 1.0 g/ml rifampin inhibited the growth of 96% of the strains, with a minimum 4-fold reduction of most MICs. AB3027 was the exception, as it is significantly resistant to rifampin (MIC Ͼ 128 g/ml; Table 1).
To further evaluate activity, time-kill assays (3 biological replicates) were performed with 2.0 g/ml SPR741 and 1.0 g/ml rifampin against AB5075 grown in CAHMB as previously described (23). Each drug used alone had little effect on growth. In contrast, the combination resulted in fewer than 10 organisms (limit of detection) on LB plates FIG 1 (A) An isobologram generated from checkerboard assays where AB5075 was grown with increasing concentrations of SPR741 and rifampin. (B) Time-kill assay of SPR741, rifampin, and combinations against XDR-A. baumannii. AB5075 was grown overnight in CAMHB, then subcultured into CAMHB for 2 h. Cultures were inoculated 1:10 into CAMHB alone or with 2.0 g/ml SPR741, 1.0 g/ml rifampin, or both SPR741 and rifampin at their respective concentrations. Time points were taken at 0, 2, 4, 6, and 24 h, and samples were plated for CFU. Only the combination of rifampin and SPR741 was statistically significant (red line), as tested by two-way ANOVA (P ϭ 0.0048). at 2, 4, and 6 h (Fig. 1B), a significant result (two-way ANOVA, P ϭ 0.0048). This result confirmed that the combination of SPR741/rifampin had a bactericidal, synergistic effect that should be further tested in vivo.
The methods for the murine pulmonary model of A. baumannii infection were previously detailed (19), and were conducted similarly for this study. In pilot experiments, we initially evaluated three doses of SPR741 (40, 60, or 80 mg/kg) once daily (QD) or twice daily (BID), with 5.0 mg/kg or 10.0 mg/kg doses of rifampin also provided QD or BID. Only the BID-treated mice survived over the course of 1 week, suggesting that QD treatment was not sufficient (data not shown). Next, two independent experiments were conducted using 10 mice per group. Four hours after A. baumannii inoculation, groups were treated with sterile saline (negative control), 5.0 mg/kg rifampin, 60 mg/kg SPR741 BID, or the combination of SPR741 40 mg/kg or 60 mg/kg SPR741 with 5.0 mg/kg rifampin BID for the next 3 days. The survival rate for 60 mg/ kg SPR741 combined with 5.0 mg/kg rifampin BID was 90% (Fig. 2A), a significant success for this aggressive infection model (Mantel-Cox test, P Ͻ 0.0027). In contrast, untreated animals or mice receiving SPR741 alone succumbed to infection ( Fig. 2A). Rifampin alone only provided 50% survival ( Fig. 2A).
Separate experiments were then conducted to evaluate bacterial burden via CFU (CFU/g of lung tissue). Mice were sacrificed on day 2 before the untreated control animals succumbed to infection, as previously described (19). These results mirrored the survival results, where the combination of SPR741/rifampin decreased bacterial burden by 6.0 log 10 CFU/g compared to the vehicle-alone control (Mann-Whitney U test, P Ͻ 0.0001) (Fig. 2B). When comparing the combination of SPR741/rifampin to rifampin treatment alone, a 2.0-log 10 reduction in burden was seen with the addition 128 Ϫ a MICs were determined separately in the presence of the combination of SPR741 at 4.0 g/ml and rifampin at 1.0 g/ml. b Ϫ, no growth; ϩ, growth.

SPR741 Potentiates Activity against XDR A. baumannii
Antimicrobial Agents and Chemotherapy of SPR741, which was also statistically significant (Mann-Whitney U test, P ϭ 0.0029) (Fig. 2B). This investigation is a promising start with regard to in vivo safety and efficacy of SPR741 combinations against Gram-negative pathogens. In pilot experiments, more mice did succumb (80% survival) with 80 mg/kg BID doses of SPR741. The reason for this is unclear, but clearance and complete animal survival are difficult to achieve in this model. AB5075 is highly aggressive and bacteria reach high numbers in lung tissue, followed by dissemination into the bloodstream and colonization of other organs, including heart, spleen, and kidneys (4,19). With regard to toxicity, previously pre-  Box-and-whisker plots of log 10 CFU/g of lung tissue on day 2 postinoculum. Mice were treated with rifampin at 5 mg/kg BID (green box), SPR741 60 mg/kg BID (yellow box), the combination of these doses (red box), or sterile saline (vehicle alone, control; black box) for 2 days. Boxes show median and interquartile ranges, while whiskers represent 95% confidence interval (CI). Groups were compared each day via the Mann-Whitney U test. ** represents P values of Ͻ0.01 (P ϭ 0.0029). These data are pooled from two biological replicates with at least 6 mice per group and 13 to 16 mice total per test condition. efficacy against other bacterial species. Furthermore, animal models mimicking other clinical indications and evaluating pharmacokinetics/pharmacodynamics (PK/PD) are also currently being pursued.