Inhibition of Staphylococcus pseudintermedius Efflux Pumps by Using Staphylococcus aureus NorA Efflux Pump Inhibitors

One promising approach in treating antibiotic-resistant bacteria is to “break” resistances connected with antibacterial efflux by co-administering efflux pump inhibitors (EPIs) with antibiotics. Here, ten compounds, previously optimized to restore the susceptibility to ciprofloxacin (CIP) of norA-overexpressing Staphylococcus aureus, were evaluated for their ability to inhibit norA-mediated efflux in Staphylococcus pseudintermedius and synergize with CIP, ethidium bromide (EtBr), gentamycin (GEN), and chlorhexidine digluconate (CHX). We focused efforts on S. pseudintermedius as a pathogenic bacterium of concern within veterinary and human medicine. By combining data from checkerboard assays and EtBr efflux inhibition experiments, the hits 2-arylquinoline 1, dihydropyridine 6, and 2-phenyl-4-carboxy-quinoline 8 were considered the best EPIs for S. pseudintermedius. Overall, most of the compounds, except for 2-arylquinoline compound 2, were able to fully restore the susceptibility of S. pseudintermedius to CIP and synergize with GEN as well, while the synergistic effect with CHX was less significant and often did not show a dose-dependent effect. These are valuable data for medicinal chemistry optimization of EPIs for S. pseudintermedius and lay the foundation for further studies on successful EPIs to treat staphylococcal infections.


Introduction
The rise of bacterial pathogens resistant to antibacterial agents poses a major threat to human and animal health worldwide. The ever-increasing demand for novel antimicrobial strategies, in contrast to the current level of investments, has led to the development of several early phase-hit compounds [1]. Among these, efflux pump inhibitors (EPIs) are antimicrobial resistance breakers targeting bacterial membrane transporters, known as efflux pumps (EPs), which are involved in the extrusion of noxious compounds, including antimicrobial agents [2]. EPIs have the potential to reverse resistance to several antibacterials [3,4], reduce biofilm formation associated with overexpressed EPs [4], and suppress the evolution of resistance [5]. Particularly, EPIs can be used to optimize the process of prevention of antibacterial efflux, thus boosting the efficacy of drugs acting on intracellular bacterial targets.
The need to address antimicrobial resistance mediated by NorA, one of the most studied EPs in Staphylococcus aureus [5][6][7], led to the identification of potent EPIs by exploitation of in silico drug repositioning approaches and medicinal chemistry campaigns [8][9][10]. We have been working for years on the design and synthesis of novel S. aureus NorA EPIs by comparing their biological activity on strains overexpressing norA or norA-deleted, resulting in the identification of several molecules able to fully restore the susceptibility to fluoroquinolone ciprofloxacin (CIP) of resistant S. aureus strains overexpressing NorA EP [8]. To establish the concentrations of each EPI to be used in synergistic studies, we determined the Minimum Inhibitory Concentration (MIC) of ten selected compounds for S. pseudintermedius ATCC 49444 (wild-type) and its efflux mutant S. pseudintermedius_P8 (overexpressing norA and harboring an 11 bp deletion in the promoter region of the norA gene). The MICs are reported in Table 1. Overall, all compounds, when used alone, had weak antimicrobial activity against S. pseudintermedius, except for the pyrazobenothiazine analog 5, which showed MIC values of 6.25 mg/L against both strains. MICs of all tested compounds exhibited the same values against both strains, suggesting that these derivatives are not S. pseudintermedius NorA substrates. On the other hand, being the susceptibility of S. pseudintermedius_P8 to these molecules affected by the increased efflux (MIC values shown in Table 1), EtBr, GEN, CIP, and CHX are considered to be substrates of S. pseudintermedius efflux system. Checkerboard assays were then performed by combining EtBr, GEN, CIP, and CHX at scalar concentrations with serial dilutions of each compound used from MIC/4 to To establish the concentrations of each EPI to be used in synergistic studies, we determined the Minimum Inhibitory Concentration (MIC) of ten selected compounds for S. pseudintermedius ATCC 49444 (wild-type) and its efflux mutant S. pseudintermedius_P8 (overexpressing norA and harboring an 11 bp deletion in the promoter region of the norA gene). The MICs are reported in Table 1. Overall, all compounds, when used alone, had weak antimicrobial activity against S. pseudintermedius, except for the pyrazobenothiazine analog 5, which showed MIC values of 6.25 mg/L against both strains. MICs of all tested compounds exhibited the same values against both strains, suggesting that these derivatives are not S. pseudintermedius NorA substrates. On the other hand, being the susceptibility of S. pseudintermedius_P8 to these molecules affected by the increased efflux (MIC values shown in Table 1), EtBr, GEN, CIP, and CHX are considered to be substrates of S. pseudintermedius efflux system. Checkerboard assays were then performed by combining EtBr, GEN, CIP, and CHX at scalar concentrations with serial dilutions of each compound used from MIC/4 to MIC/128 to prevent any synergistic effect due to an intrinsic antibacterial activity of the putative tested EPIs 1-10.
When tested against norA overexpressing strain S. pseudintermedius_P8, all compounds exhibited modest to excellent synergistic activities depending on the combined antimicro- bial. Compounds 1, 3, 5-8, and 10 were synergistic with EtBr and all the antimicrobials tested (see Table 2 for modulation factor "MF" values). Of note, since all compounds were tested at serial dilutions starting from MIC/4 to MIC/128, it is essential to consider their effective concentration when comparing data (concentrations used are shown in Table 2). Overall, 2-arylquinoline derivatives 1, 3, and 4 exhibited a significant synergistic effect with CIP and EtBr and a weaker activity when combined with GEN and CHX. This discrepancy in synergistic activity can be imputed to a reduced efflux of these two antimicrobials. Indeed, MIC values of GEN and CHX are significantly less affected (4-fold difference between the two strains) than those of CIP and EtBr (64-fold for both). Of note, at MIC/128 (1.57 mg/L), compound 1 reduced the CIP MIC by 8-fold, thus exhibiting a very potent synergistic effect. On the contrary, the 2-arylquinoline analog 2 showed a weaker (or absent) synergistic effect in accordance with data previously obtained against S. aureus strains (the synergism with CIP was considered ineffective since it was not dose-dependent). Results of the pyrazolobenzothiazine derivative 5 were affected by its low MIC values, forcing us to use it at very low concentrations in checkerboard assays. However, up to MIC/8 (0.78 mg/L), compound 5 displayed a good synergistic effect (MF = 4) with GEN, CIP, and EtBr. The nicardipine drug (compound 6) exhibited a very good profile by synergizing with GEN, CIP, and EtBr up to very low concentrations. In particular, at 1.57 mg/L (MIC/128), compound 6 still retained a MF of 4 in combination with CIP, similar to the 3-phenylquinolone analog 7 and the 2-phenyl-4-carboxy-quinoline derivative 8. Promising results were also obtained with the quinazoline derivative 9 and the benzimidazole 10 that showed a significant synergistic effect with GEN, CIP, and EtBr up to low concentrations (6.25 and 12.5 mg/L, respectively). In addition, both of them reduced the CIP MIC 4-fold at 1.57 mg/L, thus highlighting a potent synergistic activity.     As a confirmation of the synergistic effect, the minimum fractional inhibitory concentration index (FICI) obtained for each combination EPI/antimicrobial is reported in Table 3. Data supported the potent synergism observed for most of the compounds, confirming a greater effect of the EPIs in combination with the fluoroquinolone CIP and EtBr. As expected, by also analyzing the effect of compounds in combinations with antimicrobials and EtBr against the wild-type strain (ATCC 49444), it is evident a poor or absent synergistic effect with antimicrobials (Table 3). On the other hand, a modest synergism is present when most of the compounds were combined with EtBr, which is a known nonspecific substrate of bacterial EPs, thus suggesting that most of the compounds likely possess the ability to reduce the efflux deriving by the basal expression of EPs.

Comparative Inhibition of EtBr Efflux
Before testing the inhibiting activity of selected compounds on S. pseudintermedius ATCC 49444 and S. pseudintermedius_P8 EPs, the capability of the bacterial strains to accumulate EtBr was tested. Predictably, EtBr accumulation was lower inside norA-overexpressing Antibiotics 2023, 12, 806 6 of 12 S. pseudintermedius_P8 than the parent strain ( Figure 2a). Overall, the assay performed in the presence of CCCP showed a detectable accumulation for both strains. medius ATCC 49444 ( Figure 2B). As expected, S. pseudintermedius_P8 showed high EtBr efflux in the absence of EPI. The residual fluorescence of EtBr in S. pseudintermedius_P8 was indeed 58% lower than S. pseudintermedius ATCC 49444, confirming the presence of highly overexpressed EPs. Compounds 3 and 6-8 used at MIC/64 did not alter the capability of S. pseudintermedius_P8 to extrude EtBr, while compound 1 at MIC/64 showed a weak inhibitory effect ( Figure 2C). At the end of the test, all compounds used at MIC/8 increased EtBr fluorescence as follows: 7 22.5% (p = 0.078); 3 29.9% (p = 0.012); 1 32.6% (p = 0.006); 6 35.1% (p = 0.003); 8 46.1% (p = 0.001) ( Figure 2C). These findings confirm that the synergistic effect observed by checkerboard assays is due to the inhibition of drug efflux.  On the strength of the data from checkerboard assays, compounds 1, 3, and 6-8 were advanced toward fluorometry-based EtBr efflux inhibition test. The EPIs were tested at MIC/8 (compounds 1 and 6-8, 25 mg/L; compound 3, 6.25 mg/L) and MIC/64 (compounds 1 and 6-8, 3.13 mg/L; compound 3, 0.8 mg/L). The results are presented in Figure 2. The use of compounds 1, 3, and 6-8 slightly influenced the EtBr efflux activity of S. pseudintermedius ATCC 49444 ( Figure 2b). As expected, S. pseudintermedius_P8 showed high EtBr efflux in the absence of EPI. The residual fluorescence of EtBr in S. pseudintermedius_P8 was indeed 58% lower than S. pseudintermedius ATCC 49444, confirming the presence of highly overexpressed EPs. Compounds 3 and 6-8 used at MIC/64 did not alter the capability of S. pseudintermedius_P8 to extrude EtBr, while compound 1 at MIC/64 showed a weak inhibitory effect (Figure 2c). At the end of the test, all compounds used at MIC/8 increased EtBr fluorescence as follows: 7 22.5% (p = 0.078); 3 29.9% (p = 0.012); 1 32.6% (p = 0.006); 6 35.1% (p = 0.003); 8 46.1% (p = 0.001) (Figure 2c). These findings confirm that the synergistic effect observed by checkerboard assays is due to the inhibition of drug efflux.

Discussion
The overuse and misuse of antibiotics in human and veterinary healthcare and agriculture have accelerated the emergence and spread of antimicrobial-resistant bacteria worldwide, posing a threat to the effective treatment of infections. [29][30][31]. One proposed approach to treat antibiotic-resistant bacteria is to "break" resistances by co-administering appropriate non-antibiotic drugs with failing antibiotics to re-sensitizing resistant bacteria [32]. Among the antibiotic resistance breakers, EPIs can reduce resistance by blocking bacterial EPs and thus increasing intracellular antibiotic concentration.
A wide array of compounds with EPI activity have been reported to date [32,33]. Most of them were optimized to target ESKAPE pathogens [34,35] that are considered critically important owing to their role in many human infections and the frequency of antibiotic resistance worryingly limiting treatment options [36]. Particularly, research efforts were addressed to target S. aureus NorA using EPIs from natural products, de novo synthesis, and repurposing of previously-approved drugs [8][9][10]. Pursuing the development of S. aureus NorA inhibitors, we have previously identified via drug repurposing approaches and medicinal chemistry campaigns a wide array of NorA EPIs characterized by different chemical scaffolds [21][22][23]25,26,28]. Most of these compounds exhibited NorA inhibition activity by restoring at very low concentrations CIP MIC against resistant S. aureus strains and overexpressing norA while not showing any effect against S. aureus strains not expressing norA.
Here, we validated the hypothesis that EPIs specifically optimized to reverse S. aureus NorA-mediated fluoroquinolone resistance can successfully reverse resistance to CIP, GEN, and CHX associated with the overexpression of norA in S. pseudintermedius. Our study clearly shows that most of the compounds could fully restore the susceptibility of S. pseudintermedius P8 to CIP, except for compound 2 (included as inactive S. aureus NorA EPI). This is important because enrofloxacin is largely metabolized to CIP in dogs [19]. Moreover, even if to a lesser extent, all the compounds synergized with GEN as well. On the other hand, the synergistic effect with CHX was less significant and often not showing a dose-dependent effect.
Although these results suggest that S. aureus NorA EPIs can also be used as S. pseudintermedius NorA EPIs, structure-activity relationship (SAR) information acquired over the years about S. aureus NorA inhibition cannot be completely transferred for designing novel S. pseudintermedius NorA EPIs. Indeed, the introduction of a methoxy group at the C-6 position of the quinoline core (compound 3), as well as the replacement of the 2-phenyl ring with a chloro-thiophene moiety (compound 4), led to an increase of the S. aureus NorA EPI activity with respect to "naked" 2-phenylquinoline derivatives such as compound 1. Herein, instead, we observed that the activity of compound 1 was comparable to or slightly better than compounds 3 and 4, thus suggesting that SAR information should be revised. Similarly, replacing the quinoline core with a quinazoline scaffold (compound 9) led to an improvement in terms of S. aureus NorA inhibition activity that was not confirmed for S. pseudintermedius NorA EP. On the other hand, 3-phenylquinolone derivative 7, which exhibited a weaker S. aureus NorA EPI activity than 2-phenylquinoline derivatives (such as compounds 1, 3, and 4), retained a promising S. pseudintermedius NorA inhibition. Interestingly, compound 7 is also endowed with good EPI activity against nontuberculous mycobacteria resistant to clarithromycin and overexpressing EPs MAV_1406 and MAV_1695 [11]. Nicardipine (compound 6), the carboxy-quinoline derivative 8, and the benzimidazole analog 10 showed a comparable inhibition of both NorA EPs. The pyrazolobenzothiazine analog 5 was instead affected by low MIC values that compromised its use as EPI.
The results from the fluorometric tests provided further evidence for the capability of compounds 1, 3, and 6-8 to inhibit S. pseudintermedius EPs, validating their inhibition as the main mechanism involved in the synergistic effect with antimicrobials.
By combining data from checkerboard assays and EtBr efflux inhibition experiments, compounds 1, 6, and 8 could be considered the best NorA EPIs able to strongly synergize with CIP and EtBr against S. pseudintermedius P8, also showing a modest synergism with GEN and CHX. In addition, considering that their EPI activity was significantly dependent on the overexpression of norA EP (poor synergistic effect and no EtBr efflux inhibition were observed against the wild-type S. pseudintermedius strain), we are confident that these three derivatives act by inhibiting NorA efflux mechanisms. Based on the previously reported cytotoxicity evaluation of these compounds [26,27], it is interesting to note that all of them showed synergistic activity with the fluoroquinolone CIP and with the aminoglycoside GEN at concentrations significantly lower than their CC 50 values towards human cells. Special attention should be given to nicardipine (compound 6), which is a non-antibiotic approved drug for the treatment of high blood pressure and angina. At 3.13 mg/L, nicardipine was able to reduce CIP MIC by 8-fold (from 8 to 1 mg/L) against S. pseudintermedius P8 while showing a CC 50 of 188.75 and 68.73 mg/L towards HepG2 and A549 cell lines, respectively [26]. Although further studies should be performed, nicardipine (compound 6) could represent a promising candidate for pre-clinical studies in in vivo models. In parallel, results obtained for the 2-phenylquinoline derivative 1 and the 4-carboxyquinoline analog 8 give us great hope for future medicinal chemistry efforts aimed at improving S. pseudintermedius NorA EPI activity, safety, and pharmacokinetic profile.
Concerns have been raised questioning the therapeutic usefulness of EPIs at the community level against bacteria harboring mutations in antibiotic targets, such as gyrasecoding genes for fluoroquinolone resistance [37,38]. However, it was recently demonstrated that overexpressed EPs (specifically NorA in S. aureus) could create a high-resistanceevolvability bacterial niche by promoting the accumulation of antibiotic-resistance mutations or increasing the fitness benefit provided by resistance mutations [5,39,40]. Moreover, high efflux appears to be linked to the downregulation of DNA repair and mutagenesis [39]. From this perspective, EPIs might be used to prevent resistance evolution and preserve the efficacy of existing antibiotics. To further our research, future work should concentrate on using these EPI derivatives to inhibit the efflux-mediated evolvability of staphylococcal species.

Bacterial Strains and Growth Conditions
The strains used in this study were S. pseudintermedius ATCC 49444 and its derivative S. pseudintermedius_P8. This was obtained by adapting S. pseudintermedius ATCC 49444 to increasing concentrations of EtBr, a known substrate of bacterial efflux system [3,41]. More details on the growth conditions of S. pseudintermedius_P8 are given in our previous paper [4]. S. pseudintermedius_P8 overexpresses norA and harbors an 11 bp deletion in the norA promoter region. S. pseudintermedius_P8 shows higher MICs of EtBr, CIP, GEN, and CHX than the parent strain, as reported in Table 1.

MIC Determination
The lowest concentration of compounds 1-10 that inhibits visible bacterial growth (MIC) was determined in triplicate by broth microdilution according to CLSI recommendations [42]. A total of 96-well plates were inoculated with 100 µL of 2-fold serial dilutions of each compound in cation-adjusted Mueller-Hinton broth (CAMHB) to test a dose range of 100-3.125 mg/L. S. pseudintermedius ATCC 49444 and P8 colonies were resuspended in a sterile medium, and the suspension turbidity was measured spectrophotometrically at OD 600 . The bacterial suspensions were then adjusted in CAMHB to 5 × 10 5 CFU/mL. Plates were inoculated with 100 µL of bacterial suspensions and incubated at 37 • C for 20 h.

Sinergy Studies
The synergistic activity of compounds 1-10 combined with EtBr, CIP, GEN, and CHX was evaluated on S. pseudintermedius ATCC 49444 and S. pseudintermedius P8 using two-dimensional checkerboard assays using 96-well microtiter plates according to the standards [43]. Each EPI was tested in a concentration range between MIC/4 and MIC/128. For S. pseudintermedius P8, the antibiotic dilutions tested ranged from the MIC to -1 doubling dilution (1 log 2 ) below the MIC value for the original strain. For S. pseudintermedius ATCC 49444, antibiotics were tested from the MIC to MIC/8. Inoculum preparation was performed in CAMHB by colony suspension from 24 h cultures on Cation-adjusted Mueller Hinton agar. One hundred microliters of the bacterial suspension were inoculated in each well at a final concentration of 5 × 10 4 CFUs. The plates were incubated aerobically at 37 • C for 20 h. After reading well optical turbidity, the FICI was calculated for synergy interpretation as follows: The combination was considered synergistic when the FICI was ≤0.5. MF, Modulation Factor, represents the n-fold reduction of the MIC of the corresponding antimicrobial when combined with the EPI.

Fluorometric Tests
Our experimental setup is based on the one proposed by Kaatz [44]. S. pseudintermedius ATCC 49444 and P8 were grown overnight at 37 • C in static conditions in 10 mL of TSB without EtBr or supplemented with EtBr at 16 mg/L, respectively.
Preliminarily, we evaluated the capability of the strains to accumulate EtBr. Bacteria were pelleted by centrifugation at 8000 rpm for 10 min and washed twice with sterile PBS. The turbidity of the suspensions was adjusted to 0.6 OD 600 in PBS. Bacterial cells were loaded with EtBr at 10 mg/L using 20 mg/L of CCCP. The fluorescence of DNA-bound EtBr was measured at 37 • C at 60 s intervals for 30 min using a TECAN Infinite 200 PRO reader at excitation/emission 530/600 nm.
To assay the inhibitory effect of selected compounds on EtBr efflux, S. pseudintermedius ATCC 49444 and P8 were loaded with EtBr at 10 mg/L using 20 mg/L of CCCP, as described. After 20 min of incubation at room temperature in the dark, the bacterial suspensions were pelleted, and cells were resuspended in sterile PBS at 0.6 OD 600 . One hundred microliters of each suspension were added to wells in a white microtiter plate containing, in duplicate: where FI t is the fluorescence intensity at different time points, and FI t0 is the fluorescence intensity at T0.

Statistical Analyses
Statistical analyses were performed using SPSS (version 17.0). The Mann-Whitney U test was applied to compare the biological replicates between treatments in fluorescence assays. A p-value < 0.05 was assumed as significant.

Conclusions
To summarize, this work provided valuable data for the medicinal chemistry optimization of EPIs for S. pseudintermedius. It has led us to conclude that (1) a small set of 'in-house' compounds optimized to reverse CIP resistance of S. aureus overexpressing norA have EPI activity against S. pseudintermedius; (2) all compounds synergized with CIP and GEN; (3) the synergistic effect resulted dependent on overexpressed norA as observed by fluorometric assays with EtBr; (4) compounds 1, 6, and 8 emerged as able to strongly inhibit the activity of S. pseudintermedius NorA EP and completely restore the activity of CIP against the resistant strain S. pseudintermedius_P8.
These data pave the way for further studies on antimicrobial-resistant staphylococcal species other than S. aureus, which have highly virulent features for both humans and companion animals.

Data Availability Statement:
The data presented in this study are available upon request from the corresponding author.

Conflicts of Interest:
The authors declare no conflict of interest.