Effect of efflux pump inhibitors on antimicrobial resistance and in vivo colonization of Campylobacter jejuni J. Lin & M. Ensminger Department of Animal Science, The University of Tennessee, Knoxville, TN Summary CmeABC, a multidrug efflux pump, contributes to Campylobacter resistance to a broad spectrum of antimicrobial agents and is also essential for Campylobacter colonization by mediation of bile resistance. We hypothesize that inhibition of CmeABC will not only control antibiotic resitance but also increase the susceptibility of Campylobacter to in vivo bile salts, consequently decreasing the colonization level of Campylobacter in poultry. Using both in vitro and in vivo systems, we examined the effect of efflux pump inhibitors (EPIs) on the susceptibility of Campylobacter to various antimicrobials. Presence of the EPI resulted in 2- to 2048-fold reduction in the MICs of antimicrobials known to be substrates of CmeABC pump in all Campylobacter strains. Particularly, the MICs of selected bile salts were dramatically decreased 64- to 512-fold when the EPI was used. The intrinsic and acquired resistance of C. jejuni to macrolide was decreased significantly (32- to 64-fold reduction in the MIC of erythromycin) in the presence of the EPI. Investigation of 57 clinical Campylobacter isolates of various origins further showed that the EPI decreased the MICs of erythromycin (2- to 512-fold) in all isolates. The inhibitory effect of the EPI was does-dependents and as low as 0.5 μg/ml of the EPI resulted in decreased MIC of antimicrobials in C. jejuni. Presence of the EPI decreased the frequency of emergence of erythromycin-resistant mutants in C. jejuni (<10-11 ), which is well below normal frequency of approximate 10-8. Notably, MIC of erythromycin was also greatly decreased (> 4-fold) in CmeB mutants in the presence of EPI, suggesting the existence of other pump(s) involved in macrolide resistance in C. jejuni. Chicken colonization study demonstrated 1
that oral administration of EPI dramatically reduced the colonization of Campylobacter in the intestine. Together, these findings indicate that inhibition of CmeABC by specific EPI is a promising approach to control antibiotic resistance and colonization of Campylobacter in human and animals. Introduction Campylobacter jejuni, the leading bacterial cause of human enteritis in many industrialized countries, has become increasingly resistant to clinical antibiotics. As a general and important resistance mechanism, multidrug efflux systems contribute significantly to the intrinsic and acquired resistance to multiple antibiotics in various bacteria. Recently, a RNDtype multidrug efflux system, CmeABC, was identified in C. jejuni. CmeABC is involved in the resistance of C. jejuni to a broad spectrum of antimicrobial agents and is essential for Campylobacter colonization in animal intestine by mediating Campylobacter resistance to bile, a group of bactericidal detergents present in the intestinal tracts of animals. In addition, the CmeABC efflux pump can be dramatically induced in intestine and by bile salts, further highlighting the significance of CmeABC in Campylobacter adaptation to the intestinal environment. Thus, inhibition of the CmeABC efflux pump may not only control antibiotic resistance but also increase the susceptibility of C. jejuni to bile salts, consequently decreasing the colonization level of Campylobacter in poultry. Recently, efflux pump inhibitors (EPIs) targeting multidrug efflux systems have been developed and proven to potentiate the activity of antimicrobial agents against Gram-negative bacteria. We demonstrated in this study that EPIs dramatically increased susceptibilities of C. jejuni to a broad spectrum of antimicrobials including in vivo bile salts and reduced colonization of C. jejuni in chickens. 2
Materials & Methods C. jejuni 81-176 (a human strain) and S3B (a chicken strain) along with their isogenic CmeB mutants were used in this study. EPI MC-207,110 (Fig. 1) was purchased from SIGMA and MC-04,124 was a gift from Olga Lomovskaya (Mpex Pharmaceuticals, CA). Bacterial susceptibilities to various antimicrobial agents were determined by standard broth microdilution method in MH broth or MH broth containing EPI. Fifty-seven clinical C. jejuni isolates of various origins were also used for evaluating the effect of MC-207,110 on the strain s susceptibility to erythromycin. To determine if EPI MC-207,110 affected the frequency of emergence of resistant bacteria, in vitro selection experiments were performed by plating wildtype C. jejuni cells onto MH-agar plates containing antibiotics or antibiotics in combination with EPI MC-207,110 at 10μg/ml. To determine the effect of the EPIs on in vivo colonization of C. jejuni in chickens, the natural host and a major reservoir for C. jejuni, multiple groups of 5-dayold chickens were inoculated with S3B at a dose of 10 5 CFU/chicken, followed by oral administration of specific EPI (Table 1). Colonization was monitored by cloacal swabbing. Table 1. Experiment design of the chicken studies (experiments A and B) Expt group EPI usage/name EPI Dose (mg/kg) EPI treatment a # per Group A 1 (control) No N/A N/A 5 2 Yes/MC-207,110 25 Single treatment: 30 7 3 Yes/MC-207,110 5 min after C. jejuni 5 4 Yes/MC-004,124 100 inoculation 6 5 Yes/MC-004,124 20 5 B 1 (control) No N/A N/A 10 2 Yes/MC-207,110 75 Three treatments: 30 10 3 Yes/MC-207,110 15 min, 24 h, 48 h after 10 4 Yes/MC-207,110 3 C. jejuni inoculation 9 Results 3
Efflux pump inhibitors increased the susceptibility of Campylobacter to antimicrobials. As shown in Table 2, the presence of MC-207,110 resulted in 2- to 2048-fold reduction in the MICs of all antimicrobials in both C. jejuni strains compared to the MICs in MH broth. Presence of the EPI resulted in 32- to 64-fold reduction in the MIC of macrolide erythromycin (Ery), a major drug of choice for treating human campylobacteriosis. Notably, the MICs of commonly found bile salts for both strains decreased dramatically in the presence of the EPI (Table 2). Compared to wild-type strains, the isogenic CmeB mutants displayed smaller magnitudes of reduction in the MICs of antimicrobials in the presence of the EPI, indicating the inhibitory effect of the EPI is primarily CmeABC-dependent. The inhibitory effect of MC- 207,110 was also dose-dependent and as low as 0.5 μg/ml of the EPI resulted in a decreased MIC for erythromycin in C. jejuni. Table 2. Susceptibilities of C. jejuni 81-176 and S3B to different antimicrobials in MH broth (MH) or MH broth supplemented with 10 μg/ml of MC-207,110 (MH+MC). Antimicrobial Minimum Inhibitory Concentration (μg/ml) C. jejuni 81-176 C. jejuni S3B MH MH+MC a MH MH+MC Erythromycin 0.125 0.0039 (32) 0.5 0.0078 (64) Ciprofloxacin 0.0625 0.03125 (2) 0.125 0.125 (-) Nalidixic acid 8 2 (4) 8 2 (4) Fusidic acid 128 2 (64) 512 2 (256) Novobiocin 16 0.125 (128) 64 0.125 (512) Rifampin 128 <0.125 (>1,024) 128 0.0625 (2048) Ampicillin 1 0.5 (2) 16 4 (4) Cefotaxime 0.125 0.0625 (2) 8 2 (4) SDS 256 4 (64) 256 4 (64) Cholic acid 4,000 250 (16) 8,000 125 (64) Chenodeoxy 4,000 62.5 (64) 8,000 125 (64) Cholic acid Taurocholic acid 64,000 125 (512) 64,000 500 (128) Glycocholic acid 32,000 250 (128) 32,000 250 (128) a The numbers in parentheses indicate the fold reductions in MICs for C. jejuni strain grown in MH broth containing 10 μg/ml of EPI MC-207,110 (MH + MC) compared to growth in MH broth (MH). 4
MC-207,110 reversed acquired Ery resistance and decreased the frequency of emergence of Ery-resistant C. jejuni. Various Ery-resistant mutants were first selected in vitro using wild-type susceptible strains 81-176 and S3B. The presence of MC-207,110 not only reduced MICs of Ery in wild-type strains 81-176 and S3B but also resulted in a marked decrease in the resistance to Ery in all Ery-resistant mutants, with MIC reductions ranging from 16-fold to >64-fold, suggesting that the CmeABC is required to maintain acquired Ery resistance in the in vitro-selected isolates and inhibition CmeABC by specific EPI could substantially reverse acquired Ery-resistance. The frequency of emergence of antibiotic resistant C. jejuni was determined on MH plates containing the specific antibiotic and MC-207,110 using wild-type susceptible strains 81-176 and S3B. The frequency of emergence of Ery resistance was approximately 10-8 on plates containing up to 32X the Ery MIC. However, when plates were supplemented with 10 μg/ml of MC-207,110, no single Ery-resistant colony emerged on the plates and the frequency of resistance occurrence was less than 1.6 X10-10 in 81-176 and 5 X 10-11 in S3B. EPIs reduced in vivo colonization of C. jejuni using chicken model system. As shown in Figure 1, colonization levels of C. jejuni in EPI treatment groups were lower than that in the control group. Specifically, the control group without EPI treatment had a 60% colonization rate 2 days postinoculation. However, there was no colonization in any chicken in the group treated one time with a low-dose of MC-207,110 and a high-dose of MC-04,124 at 2 days postinoculation (Figure 1). High dose of MC-207,110 and low dose EPI MC-04,124 treatments resulted in lower colonization levels compared to the control group at 2 days postinoculation (Figure 1). Surprisingly, MC-207,110 at the high dose (25 mg/kg) had less effect than MC- 207,110 at low dose (5 mg/ml) in experiment A (Figure 1). As the study continued, the 5
differences between treatment and control groups lessened (Figure 1), most likely due to horizontal transmission of C. jejuni among chickens within a group and the single administration of EPI at day 0. There was no significant difference in colonization between control and EPI treated chickens at 7 and 9 days postinoculation. Colonization rate (%) 120 100 80 60 40 20 0 0 2 4 6 8 10 Days postinoculation control EPI 1 H EPI 1 L EPI 2 H EPI 2 L Figure 1. Effect of EPI treatment on the colonization of C. jejuni S3B in chickens. Thirty minutes following C. jejuni inoculation, MC-207,110 (EPI 1) or MC-04,124 (EPI 2) were orally administered one time at either high (H) or low (L) dose. The treatment groups were as follows: A) control, no EPI administered (open circles); B) EPI 1 at dose of 25 mg/kg (EPI 1 H, filled circles); C) EPI 1 at dose of 5 mg/kg (EPI 1 L, filled triangles); D) EPI 2 at dose of 100 mg/kg (EPI 2 H, filled squares); and E) EPI 2 at dose of 20 mg/kg (EPI 2 L, open squares). To further define the dose effect of MC-207,110 and examine the effect of multiple EPI treatments, we conducted another chicken experiment using three different EPI doses and administered EPI MC-207,110 three times for each dosage. The second chicken study (Figure 2A) also showed that MC-207,110 reduced the percentage of chickens colonized by C. jejuni S3B when compared to the control group (90%). Similar to the unexpected dose response with MC-207,110 in the first chicken study, the highest dose of MC-207,110 resulted in highest colonization rate (70%) among three EPI-treatment groups at 2 days postinoculation (Figure 2A). Although MC-207,110 was administered for three consecutive days, inhibition of C. jejuni colonization by MC-207,110 lessened throughout the study, and by day 9 postinoculation C. jejuni S3B colonized 80 90 % of chickens for all groups. Shedding levels of chickens 6
colonized with Campylobacter were also evaluated and no significant difference was observed among four groups (Figure 2B). Colonization rate (%) 100 90 80 70 60 50 40 30 20 10 (A) control H M L 0 0 2 4 6 8 10 Days Postinoculation Shedding level (log 10 CFU/g feces) 7 6 5 4 3 2 1 0 (B) Control H M L 2 4 7 9 Days Postinoculation Figure 2. Effect of multiple EPI treatments on the colonization of C. jejuni S3B in chickens. Five-day-old chickens in each group were inoculated with 10 5 CFU of C. jejuni S3B via oral gavage. Each group of chickens received either no EPI (control), high dose (H; 75 mg/kg), medium dose (M; 15 mg/kg), or low dose (L; 3 mg/kg) of EPI MC-207,110 following C. jejuni inoculation. The EPI was orally administered into chickens three times at 30 min, 24 hr and 48 hr postinoculation. A) Percentage of chickens colonized by Campylobacter after inoculation and EPI treatment. B) The shedding levels of chickens colonized with Campylobacter after inoculation and EPI treatment. Each bar represents the mean CFU of the colonized chickens in each group. Standard errors are indicated by error bars. Conclusions The results of this study demonstrate the feasibility of developing a CmeABC efflux pump-based intervention strategy to combat antibiotic resistance and colonization of Campylobacter in humans and animal reservoirs. Similar to all infectious disease drug development, conversion of a promising EPI into a clinically useful therapeutic agent must address some key issues, such as toxicity, stability, bioavailability, production cost, etc. Therefore, development of clinically useful EPIs is still in the early stage. Despite the challenge of discovery of appropriate EPIs for applied use, it is promising and imperative to identify and 7
develop clinically useful EPIs to reverse antibiotic resistance and reduce Campylobacter colonization. More studies using new EPI candidate series in conjunction with pharmacokinetics/pharmacodynamic analysis are needed in the future. References Lin,J., Akiba,M., Sahin,O., et al. (2005a) CmeR Functions as a Transcriptional Repressor for the Multidrug Efflux Pump CmeABC in Campylobacter jejuni. Antimicrob.Agents Chemother. 49: 1067-1075. Lin,J., Cagliero,C., Guo,B., et al. (2005b) Bile salts modulate expression of the CmeABC multidrug efflux pump in Campylobacter jejuni. J Bacteriol. 187: 7417-7424. Lin,J., Akiba,M., Zhang,Q. (2005c) Multidrug Efflux systems in Campylobacter. In Campylobacter: Molecular and Cellular Biology. Ketley,J.M., Konkel,M.E. (eds). Wymondham, U.K.: Horizon Bioscience, pp 205-218. Lin,J., Sahin,O., Michel,L.O., et al. (2003) Critical role of multidrug efflux pump CmeABC in bile resistance and in vivo colonization of Campylobacter jejuni. Infect.Immun. 71: 4250-4259. Lin,J., Michel,L.O., and Zhang,Q. (2002) CmeABC Functions as a Multidrug Efflux System in Campylobacter jejuni. Antimicrob.Agents Chemother. 46: 2124-2131. Lomovskaya,O., Bostian,K.A. (2006) Practical applications and feasibility of efflux pump inhibitors in the clinic-a vision for applied use. Biochem.Pharmacol. 71: 910-918. 8