Evaluation of In Vitro Efficiency of Ciclopirox Against Yersinia pestis and Francisella tularensis

TL;DR

This study shows that the antifungal drug ciclopirox can inhibit the growth of dangerous bacteria like Yersinia pestis and Francisella tularensis in the lab.

Contribution

The study demonstrates the in vitro antibacterial efficacy of ciclopirox against Tier-1 pathogens Y. pestis and F. tularensis.

Findings

Ciclopirox effectively inhibits the in vitro growth of fully virulent and avirulent strains of Y. pestis and F. tularensis.

Prolonged exposure to ciclopirox does not lead to increased resistance in these bacterial strains.

Ciclopirox may serve as a treatment alternative or combination therapy for these bacterial threats.

Abstract

Yersinia pestis and Francisella tularensis are Tier-1 pathogens with high interest for biodefense and public health. Evaluating the antibacterial activity of repurposed drugs against these high-priority pathogens is a key element in the ongoing effort to develop diversified antimicrobial strategies. Drug repurposing offers a cost-effective and time-efficient approach to address antibiotic resistance by identifying new applications for existing therapeutics. In this study, we demonstrate in vitro antibacterial effect of the antifungal agent ciclopirox and offer this drug as a potential antibacterial treatment. Ciclopirox in vitro activity was previously reported against various Gram-negative bacteria, including resistant strains, primarily through iron chelation that disrupts key metabolic pathways and virulence mechanisms. Additionally, it exhibits antibiofilm activity and can potentiate the efficacy of certain antibiotics. Our findings reveal that ciclopirox effectively inhibits the in vitro growth of fully virulent strains of Y. pestis and F. tularensis, as well as avirulent isolates, including avirulent mutants that their wild-type susceptibility was reduced through selection to MIC levels defining them as “nonsusceptible” to ciprofloxacin (Y. pestis Kim53Δ70Δ10 and F. tularensis LVS) and doxycycline (LVS), or resistant to doxycycline (Kim53Δ70Δ10) according to CLSI interpretive criteria. Additionally, prolonged exposure of Y. pestis and F. tularensis to sub-MIC and MIC concentrations of ciclopirox did not lead to an increase in observed MIC during the study period. These results highlight ciclopirox as a potential candidate for treatment alternative, combined with other antibiotic substances or repurposed drugs against these bacterial threats.

Figures

Full text

1. Introduction

Antibiotic resistance causes a serious and growing threat to global healthcare systems. This challenge is rapidly escalating into a crisis, largely fueled by the overuse and misuse of antibacterial drugs (such as treating viral infections, widespread use in agriculture, animal husbandry, etc.). At the same time, there is limited incentive for pharmaceutical companies to invest in developing new antibiotics, as these are often considered less profitable compared to other treatments (such as long-term or chronic disease drugs). As a result, over the past decade, only a small number of new antibiotic compounds have been approved by regulatory agencies such as the Food and Drug Administration (FDA) and European Medicines Agency (EMAs) [1,2].

The CDCs (Centers for Disease Control and Prevention) have categorized several bacteria as posing alarming, significant, and threats to healthcare systems. Many of these already lead to significant burdens (economic and clinical) on healthcare systems around the world as well as on patients and their families. Furthermore, antibiotics resistance in these bacteria can be readily acquired synthetically in the laboratory [3].

Drug repurposing has emerged in recent years as a highly studied therapeutic approach, serving as an effective strategy for re-evaluating clinically approved and safe drugs for the treatment of new medical conditions. This approach is characterized by lower development costs, shortened development times, and improved safety profiles compared to de novo drug development [4]. Specifically, there is extensive scientific literature on drug repurposing aimed at combating bacteria in general, and antibiotic-resistant bacteria in particular.

Ciclopirox is an antifungal drug approved for topical use in humans for many years. As evidenced by the literature, this drug is a suitable candidate for evaluation as a repurposed drug to combat antibiotic-resistant bacteria. Ciclopirox exhibits direct antibacterial activity against clinical isolates of the Gram-negative bacteria A. baumannii, E. coli, K. pneumoniae, and P. aeruginosa, with sensitivity in the range of MIC = 5–15 µg/mL (regardless of whether the bacterium is susceptible or resistant to other antibiotics). Its mechanism of action (specifically in E. coli) involves inhibiting LPS synthesis and galactose metabolism (two pathways critical for virulence) [5,6]. A broader direct antibacterial mechanism of ciclopirox, which appears to be broad-spectrum (antifungal, antibacterial, and antiviral), involves iron chelation [7,8]. Furthermore, iron chelation may have indirect implications, as iron is a critical factor for bacterial virulence. The drug may also be effective against biofilms [9] and can reduce the expression and activity of virulence factors [10]. Synergism of ciclopirox with polymyxin B has been observed against Gram-negative bacteria, including resistant strains (E. coli, A. baumannii, and P. aeruginosa) [11].

Although ciclopirox is approved only for topical treatment, clinical trials have been conducted over the past decade evaluating its efficacy (repurposing for cancer treatment) via oral administration (www.clinicaltrials.gov—[NCT00990587](NCT00990587), NCT05647343). Additionally, intravenous administration of the prodrug of ciclopirox (fosciclopirox) appeared to be well tolerated in clinical trials (www.clinicaltrials.gov—[NCT03348514](NCT03348514), NCT04956042) [12]. The favorable safety data, coupled with the lack of evidence for the development of resistance in fungi [13,14,15] and to date, in bacteria [7], may serve as criteria for developing treatments for resistant bacteria.

In the study presented here, we demonstrate that ciclopirox exhibits in vitro antibacterial activity against Tier-1 Gram-negative pathogens Y. pestis (virulent strain Kim53) and F. tularensis (virulent strain Schu S4), as well as against the respective attenuated Y. pestis derivatives Kim53Δ70Δ10, Kim53Δ70Δpgm, and Kim53Δ10Δpgm, and the F. tularensis live Vaccine Strain (LVS). Susceptibility to ciclopirox was not affected by mutations that reduced the wild-type sensitivity of Kim53Δ70Δ10 or LVS to first choice treatment drugs, namely doxycycline or ciprofloxacin to levels categorizing them as “nonsusceptible” or resistant according to CLSI interpretive guidelines [16]. These findings warrant further investigation into the therapeutic potential of this compound against high-consequence pathogen.

2. Results

2.1. Inhibition of Bacterial Growth

Determination of the ciclopirox MICs against strains of F. tularensis and Y. pestis was conducted by microdilution assays according to CLSI guidelines [16] (Figure 1). The MIC obtained at 48 h of incubation for F. tularensis Schu S4 was in the range of 3–6 µg/mL (44% MIC = 3 µg/mL, 56% MIC = 6 µg/mL [n = 16]), and for LVS in the range of 1.5–6 µg/mL (50% MIC = 1.5, 43% MIC = 3 µg/mL, 7% MIC = 6 µg/mL [n = 30]). These results indicate that the LVS strain is highly similar to the virulent Schu S4 strain in respect to ciclopirox sensitivity, as their MIC ranges differ by about one 2-fold dilution (Figure 1a,b respectively). Similar MIC results were obtained for the virulent and avirulent derivatives of Y. pestis. The MIC obtained after 24 h for the virulent strain Kim53 was 12 µg/mL [n = 6], while for the avirulent derivative Kim53Δ70Δ10 the MIC was in the range of 6–12 µg/mL (60% 6 µg/mL, 40% 12 µg/mL [n = 20]) (Figure 1c,d respectively). MICs of 12 µg/mL were obtained also for the Kim53Δ70Δpgm and Kim53Δ10Δpgm derivatives, indicating that the absence of the pgm locus or the virulence plasmids pCD1 and pPCP1 (all required virulence factors in vivo), do not lead to a measurable effect on MICs by iron depravation of ciclopirox in vitro.

2.2. Minimal Bactericidal Concentration

Minimal Bactericidal Concentration (MBC) of ciclopirox to F. tularensis (LVS) was determined by plating a standard concentration of bacterial culture (ca. 5 × 10^4^ cfu/well) into microplate wells containing a series of twofold dilutions of ciclopirox in HLMHI and incubating at 37 °C under conditions identical to the microdilution MIC assay. The CFU counts of the F. tularensis culture was measured at time zero (start of the experiment), and after 48 h of incubation. Bacterial survival in the wells was assessed by preparing serial dilutions and plating for viable plate counts on mCHA plates. The MBC value was defined as the minimal concentration at which a 3-log reduction (99.9% kill or ≤0.1% survival) in the number of surviving bacteria occurs relative to time zero. The percentage of LVS surviving bacteria is presented in Figure 2a. The ciclopirox concentration at which a 3-log reduction was observed at the time of MIC determination (48 h, MIC = 3 µg/mL) relative to the initial bacterial count (at time zero) was MBC = 24 µg/mL. At this concentration, 0.01% survived, and this is the first concentration where less than 0.1% survived (>3 log reduction) [17]. The MBC/MIC ratio for this bacterium at 48 h is eight. This value is >3 MIC multiplications, thus indicating bacteriostatic activity of ciclopirox against F. tularensis, although only by a factor of two above the bactericidal “cutoff” definition. However, it should be noted that for slow-growing organisms, further incubation should be allowed (as doubling time and killing may take longer), and indeed, when checked after 56, 64 and 72 h, the MBC/MIC ratio drops to the bactericidal definition range of MBC/MIC = 4 (see time-kill curves in Figure 3a).

In a similar assay the MBC of ciclopirox was examined against Y. pestis Kim53Δ70Δ10 in CAMHB at 28 °C. The MIC at 24 h was 12 µg/mL. Viable plate counts revealed that only at a concentration of 384 µg/mL there was a reduction in more than 3 logs, to less than 0.1% relative to the initial inoculum at time zero (in this case, a reduction to 0.02%) (Figure 2b). Therefore, this drug MBC/MIC = 32 was clearly found to show bacteriostatic activity for Kim53Δ70Δ10 bacteria.

2.3. Time Kill Curves

The time-kill (TKC) measurements were conducted in multiple replicates of the microdilution MIC assay, by sampling two replicates at each indicated time point for CFU counts in serial 10-fold dilutions. The TKC for LVS (Figure 3a) showed that after 48 h (MIC = 3 µg/mL), a ciclopirox concentration of 24 µg/mL was the minimal concentration to produce a reduction in more than 3 log units, again defining this concentration as the MBC as shown above (see Figure 2a). Furthermore, as incubation progressed towards 72 h, a reduction of over three orders of magnitude was observed also at lower concentrations of 12 µg/mL and even 6 µg/mL (thus lowering the MBC_72h_/MIC ratio to 2). This phenomenon supports the conclusion that ciclopirox’s mode of action against the F. tularensis bacterium can be considered as bactericidal.

For Y. pestis Kim53Δ70Δ10 (Figure 3b), the ciclopirox MIC was 12 µg/mL, which remained consistent throughout the incubation. Ciclopirox concentrations below the MIC (0, 3, 6 µg/mL) allowed for bacterial proliferation. However, even at concentrations significantly above the MIC (24–192 µg/mL), the reduction in bacterial count was limited to less than two orders of magnitude. A substantial reduction of approximately four orders of magnitude, defining the MBC, was only observed at 384 µg/mL. With an MBC/MIC ratio not lower than 32, ciclopirox acts as a bacteriostatic growth inhibitor against Kim53Δ70Δ10.

2.4. Susceptibility to Ciclopirox in Ciprofloxacin and Doxycycline Reduced Susceptibility Isolates

Antibacterial activity of ciclopirox was investigated in avirulent isolates of LVS and Kim53Δ70Δ10 that were selected for reduced susceptibility to either ciprofloxacin or doxycycline. The term reduced susceptibility is used to describe the range of non-WT MIC range for F. tularensis that can only be defined as susceptible (S) to ciprofloxacin and doxycycline when MICs are ≤0.5 µg/mL and ≤4 µg/mL, respectively, while higher MICs suggestive of a “nonsusceptible” category cannot be used to define intermediate (I) or resistant (R) categories (CLSI, M45). Similarly, Y. pestis is defined as susceptible to ciprofloxacin for MICs ≤0.25 µg/mL, while higher MIC values cannot be used to interpret the “nonsusceptible” strain as intermediate or resistant. However, Y. pestis has defined categories for tetracyclines, with MIC interpretive criteria of ≤4, 8 and >16 µg/mL for S, I, and R categories respectively.

As shown in Table 1, ciclopirox’s MICs remained within the 6–12 µg/mL range across a broad spectrum of Y. pestis isolates exhibiting reduced susceptibility to ciprofloxacin (MIC range 0.008–16 µg/mL) or resistance to doxycycline (up to 32 µg/mL) as compared to the parental strain (Kim53Δ70Δ10). These findings align with previous research by Carlson-Banning et al. [7], who demonstrated that ciclopirox’s MIC remained within a narrow range of 5–15 µg/mL against thirty clinical E. coli isolates, half of which were highly resistant to ciprofloxacin (MIC = 50–500 µg/mL). Similar consistency was observed in A. baumanii, K. pneumoniae, and P. aeruginosa (8–9 resistant clinical isolates of each) [7].

Similarly, the MICs of ciclopirox against F. tularensis LVS isolates with reduced susceptibility to ciprofloxacin remained within a range of 1.5–3 µg/mL but demonstrated a up to one doubling dilution difference in the MIC range of the isolates with reduced susceptibility to doxycycline (Table 2). As these deviations did not occur in all replicates, they do not seem to represent a mechanistic relevance to ciclopirox susceptibility.

2.5. The Proposed Mechanism of Antibacterial Effect of Ciclopirox Is Iron Chelation

Other studies indicate that ciclopirox’s primary antibacterial (and antifungal) mechanism of action involves iron chelation [7,8]. Unlike other chelators that bind extracellular iron (e.g., deferoxamine) or reduce ferric to ferrous iron (e.g., triapine), ciclopirox is thought to bind intracellular iron [18], thereby inhibiting iron-dependent enzymes by blocking iron incorporation [19].

To elucidate whether the antibacterial mechanism of action of ciclopirox works similarly in the Tier-1 bacterial agents, we investigated whether increasing iron concentrations in the medium would inhibit or abolish ciclopirox’s growth inhibitory activity against Y. pestis (Kim53, Kim53Δ70Δ10, Kim53Δ70Δpgm and Kim53Δ10Δpgm) and F. tularensis (Schu S4 and LVS). For this purpose, these bacteria were incubated under standard microdilution conditions in the appropriate medium for ciclopirox MIC determination, while replica of the same MIC assays was incubated with various concentrations of FeSO_4_ to determine iron’s effect on the MICs. As shown in Table 3, adding increasing concentrations of FeSO_4_ led to a dramatic, dose-dependent increase in ciclopirox’s MIC values (16–≥64 fold compared to ciclopirox alone without FeSO_4_).

The above results indicate that the addition of iron (as FeSO_4_) titrates the antibacterial iron chelating activity of ciclopirox, while the addition of iron itself had little to no effect on bacterial growth rate in the growth control wells (Figure 4), indicating that the observed phenomenon is due to the drug’s effect being suppressed by addition of iron.

Cho and Kim [20] demonstrated that while selective pressure rapidly induces resistance of E. coli ATCC 25922 to azidothymidine (<2 days), the same pressure does not result in a significant increase in MIC was observed during the study period (growth for 25 days). These findings are consistent with the lack of evidence in the literature for the development of resistance to ciclopirox. Accordingly, we investigated whether this observation also applies to Y. pestis and F. tularensis. For this purpose, we exposed these strains to selective pressure of either ciclopirox or ciprofloxacin (Figure 5). Our results show that while exposure to gradually increasing concentrations of ciprofloxacin in liquid medium led to the emergence of resistance (manifested by a rapid and marked increase in ciprofloxacin MIC values for both strains), the selective pressure did not result in any increase in ciclopirox MIC during our study (cumulative exposures of 24 and 36 days for Y. pestis and LVS respectively). The lack of rise in MIC within the limited exposure cycles probably represents multiple intracellular iron-dependent targets and processes that cannot be mutually mutated to gain resistance to iron deprivation (Figure 5).

3. Discussion

In this study, we demonstrated that ciclopirox exhibits an in vitro antibacterial effect against both virulent and non-virulent strains of Y. pestis and F. tularensis. Our data suggest that despite the physiological differences between strains, accounting to the growth rates according to standardized CLSI-based guidelines (MIC determination at 24 h for Y. pestis and 48 h for F. tularensis), the drug’s mechanism of action (primarily iron chelation of Fe^3+^ and other polyvalent metal ions) targets fundamental metabolic processes that are essential for both pathogens. The antibacterial effect is more potent against F. tularensis LVS (MIC = 1.5–3 µg/mL) and Schu S4 (MIC = 3–6 µg/mL), compared to Y. pestis Kim53 (MIC = 12 µg/mL). Additionally, MBC and TKC results showed that ciclopirox is on the border definition between bactericidal and bacteriostatic activity against F. tularensis (MBC/MIC ratio = 4–8) but definitely bacteriostatic against Y. pestis (MBC/MIC = 32). These results indicate that repurposing of ciclopirox as treatment alternative would probably have to rely on a combination with other antibiotics or repurposed drugs.

Over forty years ago, Sakurai and colleagues [18] showed in fungi that the drug binds irreversibly to intracellular components (intracellular concentrations are 200-fold higher than extracellular levels, with over 97% of the drug bound to cellular organelles and components, while only a minor fraction remains free in the cytoplasm, all without intracellular metabolism or degradation). Therefore, it is unlikely that the drug serves as a substrate for efflux pumps and would thus not be rapidly expelled from the cell. Moreover, the major mechanism of action of ciclopirox is iron and possibly other polyvalent metal ions chelation, which disrupts multiple iron- and metal-dependent processes such as metalloenzymes involved in energy metabolism (e.g., catalase, peroxidases, and cytochromes), DNA repair and replication, oxidative stress pathways (disruption of oxidative stress defense, accumulation of ROS, impaired ATP production and failure of essential biosynthetic pathways). Unlike classical antibiotics for which development of resistance is usually evident within few years of clinical use, the co-disruption of multiple processes suggests a very low likelihood of resistance development or cross-resistance with classical bacterial resistance mechanisms. These factors support our findings that the MIC of ciclopirox does not change also in the avirulent Y. pestis and F. tularensis strains selected for reduced susceptibility or resistance to conventional antibiotics (ciprofloxacin and doxycycline). Furthermore, although exposure was limited to 12 cycles, we were unable to induce any change in ciclopirox MICs through the selective pressure in both of these bacteria (unlike ciprofloxacin). This observation aligns with similar studies showing that selective pressure did not result in ciclopirox resistance in various Gram-negative bacteria, in contrast to the resistance that develops under selection with antibiotics or other drugs such as ciprofloxacin [7] or azidothymidine [20]. However, further research should include comprehensive genomic and transcriptomic studies over extended passage periods, which are essential to fully determine the long-term resistance profile potential of ciclopirox.

Our results support iron chelation as a principal antibacterial mechanism of ciclopirox, consistent with the evidence summarized in the review by Shen and Huang [21]. Specifically, supplementation of the growth medium with iron (FeSO_4_) fully reversed the antibacterial effect of ciclopirox against both F. tularensis and Y. pestis. This result implies a broad mechanism of action that may extend to other antibiotic-resistant bacterial species. Because iron availability is a pivotal factor during infection, strategies that modulate the host–pathogen competition for this metal have the potential to favor the host. To this end, siderophores may be chemically modified to serve as iron-chelating agents, or existing iron chelators can be further optimized to potentiate immune-mediated restriction of bacterial growth. Additionally, siderophores have recently gained attention as targeted delivery platforms for directing antibiotics into pathogenic bacteria [22]. In the context of virulent Y. pestis, it is noteworthy that the drug’s effect on iron availability may independently help control the pathogen, as upregulation of iron storage has been associated with reduced virulence [23].

Importantly, beyond its direct antibacterial effects, ciclopirox has immunomodulatory properties [15], including inhibition of NLRP3 inflammasome activation [24] and antioxidant activity [25]. Furthermore, ciclopirox inhibits the mammalian target of rapamycin (mTOR) protein [26], thereby potentially activating autophagy (including in immune cells), which may contribute to host defense against pathogens [27]. These properties offer a distinct advantage in combating resistant bacteria, as they do not rely solely on a direct drug-specific effect but rather engage multiple endogenous host mechanisms. Consequently, ciclopirox may provide additional in vivo benefits beyond its direct antibacterial effect and iron chelation, particularly in bacterial infections accompanied by severe inflammation, such as those caused by the pathogens tested in this study [14]. However, the relatively high MBC values observed, particularly for Y. pestis (Kim53), suggest that ciclopirox may face pharmacokinetic challenges if utilized as a systemic monotherapy, as limited data from primary phase I paper by Minden et al. [28] reported for oral ciclopirox given once daily at ~80 mg/m^2^, PK values of Cmax ~220 ng/mL, plasma AUC ~750 ng h/mL and rapid elimination t½ of ~2.7 h.

To address these hurdles, the clinical presentation of pneumonic plague and respiratory tularemia provides a strong rationale for investigating localized, inhalation-based delivery. Such aerosolized administration could potentially achieve high local concentrations within the pulmonary environment that significantly exceed systemic plasma thresholds, thereby bypassing the limitations associated with systemic administration. Additionally, the use of fosciclopirox, the phosphoryl-oxymethyl-ester prodrug of ciclopirox-presents a viable solution for systemic delivery, as its maximum tolerated dose (MTD) in mice and humans is much higher than that of ciclopirox. Upon intravenous administration, fosciclopirox undergoes rapid and complete conversion to the active metabolite, ciclopirox.

Given the observed MIC and MBC ranges, ciclopirox may be most strategically applied as a synergistic adjuvant or potentiator within a multi-drug framework. By disrupting iron and other polyvalent metal ions homeostasis and compromising bacterial metabolic pathways, ciclopirox may sensitize these pathogens to frontline antibiotics, or other repurposed drugs. This potentiation could potentially lower the required therapeutic doses of traditional agents and mitigate the emergence of further resistance. Furthermore, the clinical presentation of pneumonic plague and respiratory tularemia provides a rationale for investigating alternative delivery routes. Aerosolized or inhalation-based administration of ciclopirox could potentially achieve high local concentrations within the pulmonary environment, reaching levels that significantly exceed systemic plasma thresholds. Such localized delivery strategies may bypass the pharmacokinetic limitations associated with systemic administration, offering a more feasible therapeutic pathway for treating primary respiratory infections caused by these Tier-1 pathogens. Future studies focusing on in vivo combination efficacy and pulmonary distribution are warranted to further define the clinical potential of ciclopirox in these scenarios.

Despite the antibacterial activity observed in this study, several limitations must be acknowledged. First, while we have validated the efficacy of ciclopirox against both attenuated and virulent strains, our experiments were conducted in vitro. This does not necessarily reflect the complex host–pathogen interactions or host-mediated iron sequestration that may influence drug performance. While the in vitro results provide a critical proof-of-concept, the absence of in vivo validation means that key parameters—including systemic toxicity, pharmacokinetics, and the ability of the drug to reach therapeutic concentrations within infected tissues—remain to be determined.

In conclusion, the results presented in this study demonstrate as proof of concept that ciclopirox has an antibacterial effect against Y. pestis and F. tularensis, including vitulent strains and avirulent strains possessing non-WT-reduced susceptibility to ciprofloxacin and doxycycline, through a mechanism of action involving primarily iron chelation. These findings indicate that ciclopirox could potentially serve as a combined antibacterial treatment or adjuvant for infections caused by first-choice drugs resistant Y. pestis and F. tularensis strains.

4. Materials and Methods

The primary objective of this study was to identify and evaluate potential therapeutic options against Yersinia pestis and Francisella tularensis by assessing the in vitro efficacy of ciclopirox. We explicitly state that no experimental procedures were performed with the intent to enhance the pathogenicity, transmissibility, or resistance of the strains used in this study. All methodologies were strictly limited to standard antimicrobial susceptibility testing and the evaluation of drug efficacy to address the public health need for novel treatment strategies against these agents.

Tested strains—all experiments with virulent strains Yersinia pestis, Kimberley53 (termed in short Kim53) and Francisella tularensis Schu S4 were conducted in a biosafety level 3 (BSL3) facility in accordance with the biosafety guidelines of the Israel Institute for Biological Research (IIBR). Three avirulent derivatives of Kim53 were used: Kim53Δ70Δ10 (Kim53pPCP1^−^pCD1^−^), Kim53Δ70Δpgm (Kim53pCD1^−^pgm^−^) and Kim53Δ10Δpgm (Kim53pPCP1^−^pgm^−^). All derivatives kept the wild-type susceptibility to doxycycline and ciprofloxacin of the parental Kim53 virulent strain. Selected mutants of the plasmids-cured Kim53Δ70Δ10 with reduced susceptibility to ciprofloxacin (MIC within CLSI “nonsusceptible” category) were YP-1, YP-2, YP-3, YP4, YP-5, YP6, and YP-7 [29], and mutants with MICs within the CLSI resistant category to doxycycline were YP-8 and YP-9 [30]. Francisella tularensis LVS (Live Vaccine Strain), is an avirulent tularemia strain. LVS was similarly mutagenized to achieve derivatives LVS-2 and LVS-3 with MICs categorizing them as nonsusceptible to ciprofloxacin, and LVS-4 with reduced susceptibility to doxycycline (yet categorized as susceptible according to CLSI guidelines for potential agents of bioterrorism) [16]. The isolation was performed in compliance with Israeli law for working with select agents, approved by the Institutional “Recombinant DNA Experimental Usage Committee” and performed according to specific Institutional “Biosafety committee” guidelines for containment and working procedures. Experiments using the virulent strain were performed using biosafety level 3 (BSL-3) containment and procedures. Experiments with the non-virulent strains were performed using biosafety level 2 (BSL-2) containment and procedures.

Reagents and drugs—Ciclopirox olamine (Sigma-Aldrich, Rehovot, Israel 7670603, Cat., 1134030), Doxycycline hyclate, Ciprofloxacine (Ciprofloxacin IV Altan Pharmaceuticals S.A. 28230 Las Rozas, Madrid, Spain 2 mg/mL), FeSO_4_ (Sigma-Aldrich, Rehovot, Israel 7670603, Cat. 935689), PBS (Dulbecco’s Phosphate Buffered Saline, Biological Industries Kibbutz Beit, HAEMEK 25115, Israel, Cat. 02-023-1A).

Growth Media—Modified Cystine Heart Agar (mCHA) [17] plates were used for isolation and for cfu counts of F. tularensis (Schu S4 and LVS). Hematin liquid Mueller Hinton (cations adjusted) IsoVitaleX (HLMHI) broth was used for culturing and antibacterial susceptibility testing of F. tularensis strains. The medium is the reference medium for microdilution tests for F. tularensis (CLSI, M45) [16], with addition of 6 µM Hematin (Sigma-Aldrich, 3281), which enhances the growth of F. tularensis in the liquid culture without affecting the susceptibility to ciprofloxacin or doxycycline. Brain Heart Infusion Agar (BHIA, BD Difco 241830, BD DifcoBecton Drive, Franklin Lakes, NJ 07417-1880, USA) plates were used for isolation and cfu counts of Y. pestis. Liquid Medium-Mueller-Hinton-cation-adjusted broth (CAMHB) (BD Difco, 212322) is the reference medium for determining commonly isolated bacterial susceptibility (CLSI, M07) [17]. CAMHB was used for Y. pestis in susceptibility tests (CLSI, M45) [16] and for exposure assays with increasing concentrations of ciprofloxacin or ciclopirox. Liquid Brain Heart Infusion (L-BHI, Difco, 237500) broth was used for culturing Y. pestis. Tryptic Soy Broth (Becton, Dickinson Company 1 Becton Drive Franklin Lakes, NJ 07417-1880 USA Cat. 236950) + 0.1 g/100 mL Cysteine (Sigma-Aldrich, C6852) (TSB-C) for culturing LVS. This medium was used for monitoring MIC changes under selective pressure from exposure of LVS to increasing concentrations of ciprofloxacin or ciclopirox.

Microdilution MIC (Minimal Inhibitory Concentration) assay—The MIC was determined using a microdilution assay to establish the minimal concentration of the tested compound that inhibits the growth of a standard bacterial inoculum (CLSI, M07 and M45) [16]. The test was performed by preparing colony suspensions at a concentration of 1–2 × 10^8^ cfu/mL by adjusting suspensions turbidity, measured with a Novaspec plus spectrophotometer (Amersham Scientific, Amersham Scientific Little Chalfont, Buckinghamshire, HP7 9NA, UK) to O.D.660nm = 0.075 for F. tularensis in HLMHI medium, or O.D.660nm = 0.31–0.34 for Y. pestis in CAMHB medium. The calibrated suspension was diluted 1:100 in the appropriate medium and inoculated (50 µL) into flat-bottom 96-well microplates (TPP, Cat. No. 92096) containing serial dilutions of the tested antibacterial compounds (50 µL), resulting in a final well volume of 100 µL and an initial bacterial concentration of approximately 5 × 10^4^–1 × 10^5^ cfu/well (or 5 × 10^5^–1 × 10^6^ cfu/mL) in the respective test medium. All tests were done in duplicates and repeated in at least three independent days. Dilutions for viable cfu counts were performed to verify the inoculum concentration on the appropriate agar medium for each bacterium. Turbidity monitoring was carried out using a Tecan Infinite plate reader (Tecan Austria GmbH, Untersbergstr. 1A, A-5082 Grödig, Austria), with hourly measurements of at O.D.630 for an incubation period of 24–48 h at 28 °C for Y. pestis or 48–72 h at 37 °C for F. tularensis. At the end of incubation, the threshold for growth/inhibition in the wells was defined as 10% relative to the growth control wells without the tested compound. The MIC was established both by the turbidity (O.D.630 < 10% of growth control) and visually (un-aided eye). MIC values were determined after incubation of 24 h for Y. pestis and 48 h for F. tularensis, in accordance with CLSI M45 guidelines. Media and antibiotic dilutions were verified using quality control strain E. coli ATCC 25922.

Minimum Bactericidal Concentration (MBC)—characterization experiments to determine the MBC in the presence of ciclopirox were conducted in triplicates using the same conditions used for the microdilution MIC tests.

The MIC was established by monitoring turbidity both visually (un-aided eye) and using a plate reader (O.D. ≤ 10% of growth control). At the time of MIC determination, all wells from the MIC value and higher (showing no visible growth) were sampled and serially diluted in PBS and plated on agar plates with the appropriate medium for CFU counts. The MBC was defined as the minimal concentration at which a reduction of 99.9% (<0.1% survival) in the number of surviving bacteria occurs relative to cfu in time zero. The MIC and MBC values were used to assess the bacterial susceptibility and to characterize whether the activity of ciclopirox is bactericidal (MIC/MBC ratio ≤ 4) or bacteriostatic (MIC/MBC ratio ≥ 8).

Time-Kill Curve (TKC)—Characterization of the TKCs for ciclopirox was determined using the microdilution test method, in which the MIC for Y. pestis and F. tularensis was assessed at 24 and 48 h respectively by monitoring turbidity visually and with a plate reader, accompanied by replicate test plates incubated under the appropriate temperature conditions. At predefined time points, duplicate wells were sampled, pooled for serial tenfold dilutions in PBS and plated on suitable agar plates, for CFU enumeration and quantification of surviving percentage of bacterial counts relative to the initial inoculum at time zero.

Iron Effect—The effect of iron chelation on bacterial sensitivity was determined under the same MIC test conditions, using CAMHB medium for Y. pestis and HLMHI medium for F. tularensis. The tested bacteria were suspended to a turbidity of O.D.660 = 0.31–0.34 for Y. pestis and O.D.660 = 0.073–0.078 for F. tularensis, diluted 1:100 in the appropriate medium and inoculated into wells containing medium with various concentrations of FeSO_4_ solution at each tested ciclopirox concentration. The Y. pestis (Kim53 and plasmids/pgm-cured derivatives) test plates were incubated at 28 °C and the F. tularensis (LVS Schu S4 and) plates at 37 °C, in a Tecan infinite plate reader, with continuous hourly monitoring of culture turbidity at O.D.630 nm for 24–72 h, to assess the effect of iron presence on bacterial sensitivity (MIC values) to ciclopirox.

Selective pressure—for preliminary screening of resistance isolates selection was done by plating Y. pestis (Kim53) on BHIA at 28 °C for 48 h and F. tularensis LVS on mCHA at 37 °C for 72 h. Colonies were suspended in CAMHB (Kim53 Δ70Δ10) or TSB-C (LVS) to 1 O.D.660nm. The following two drug series were prepared: ciprofloxacin at 0.5×, 1×, 2×, and 4× MICs; ciclopirox at 0.25×, 0.5×, 1×, and 2× MICs. Bacterial suspensions were diluted to 0.001 O.D.660nm in 5 mL of indicated medium and incubated shaking at 180 rpm for 48 h at 28 °C (Kim53 Δ70Δ10) or 72 h at 37 °C (LVS). After incubation, turbidity was measured. Culture with highest concentration was used to inoculate fresh medium with a new dilution series, where the lowest concentration matched 0.25× the previous highest growth concentration. This selection process was repeated for up to 12 passages or until reduced sensitivity was achieved for both ciclopirox and ciprofloxacin. At each cycle passage, cultures were plated (BHIA for Kim53Δ70Δ10; mCHA for LVS) to verify identity and MICs determination, isolates with reduced MICs were kept for future molecular and omics-based analysis.