In vivo efficacy of auranofin in a hamster model of Clostridioides difficile infection

Clostridioides difficile infections (CDIs) are an urgent public health threat worldwide and are a leading cause of morbidity and mortality in healthcare settings. The increasing incidence and severity of infections combined with the scarcity of effective anti-CDI agents has made treatment of CDI very challenging. Therefore, development of new, effective anticlostridial agents remains a high priority. The current study investigated the in vivo efficacy of auranofin in a CDI hamster model. All hamsters treated with auranofin (5 mg/kg) survived a lethal challenge with C. difficile. Furthermore, auranofin (5 mg/kg) was as effective as vancomycin, the drug of choice for treatment of CDIs, against relapsing CDI. Furthermore, auranofin (5 mg/kg) generated a 3.15-log10 reduction (99.97%) in C. difficile count in the cecal contents of hamsters. These results indicate that auranofin warrants further investigation as a new agent to replenish the pipeline of anti-CDI therapeutics.


Results and discussion
In vivo efficacy of auranofin in a C. difficile ileocecitis hamster model. Auranofin was previously reported to exhibit potent antibacterial and antivirulence activities against C. difficile in vitro 23 . Additionally, auranofin, at clinically achievable concentrations, was able to protect mice against C. difficile challenge 24 . These results encouraged us to investigate auranofin's efficacy in a C. difficile ileocecitis hamster model and auranofin's potential to protect hamsters from CDI recurrence. The hamster model is routinely used to evaluate therapeutics for treatment of CDI. CDI in hamsters exhibits key morphological features similar to CDI in humans such as colon enlargement, fluid accumulation and pseudomembrane formation. Additionally, dysbiosis induced by clindamycin treatment, which leads to proliferation of C. difficile, is observed both in hamsters and in humans 27,28 .
In contrast, CDI in hamsters is rapidly fatal if left untreated, a pattern that is not characteristic of human CDI. Thus, the CDI hamster model can be considered as a prevention of death model 29 .
The Golden Syrian hamster model was used to evaluate auranofin's ability to prevent ileocecitis induced by C. difficile, compared to vancomycin. The initial study investigated the activity of low doses of auranofin. Two groups of infected hamsters (n = 10) were treated with 0.125 mg/kg or 0.25 mg/kg of auranofin. One group was treated with vancomycin (positive control) and the last group received the vehicle alone (negative control). Treatments were continued for 5 days during which hamsters were observed for disease symptoms. As shown in Fig. 1A, vehicle-treated hamsters exhibited 100% mortality by day 5 of the study, in agreement with previous reports [30][31][32] . Vancomycin (20 mg/kg) protected 100% of the infected hamsters up to 5 days, in coincidence with previous reports 30,33,34 . Hamsters administered 0.125 mg/kg of auranofin exhibited 40% survival by day 5. Auranofin, at 0.25 mg/kg, was more efficacious resulting in 60% survival of infected hamsters on day 5, which was statistically significant compared to the vehicle-treated group.
We next tested the effect of higher doses of auranofin (1 mg/kg and 5 mg/kg) (Fig. 1B). After 5 days of auranofin (1 mg/kg) treatment, 60% of hamsters infected with C. difficile survived. On the other hand, administration of 5 mg/kg auranofin resulted in 100% survival of infected hamsters during the 5-day treatment period. It is worth mentioning that the results obtained in the in vivo C. difficile ileocecitis hamster model were slightly different from the results previously reported in the in vivo CDI mouse model 24 . This effect could be attributed to a difference in auranofin's pharmacokinetic profile between hamsters and mice. The rates of metabolism and excretion for auranofin may differ between hamsters and mice, which could lead to a difference in the drug's concentration at the infection site. This factor would need to be further explored in future studies. Another factor that might have contributed to the difference in results obtained between the hamster and mice studies is the overall surface area of the infection site. The site of infection is expected to be larger in hamsters compared to mice. Thus, we suspect that higher doses of auranofin were needed in hamsters to achieve a similar protective effect observed in mice.
During the experiment, the average weight of surviving hamsters in each treatment group was measured every other day ( Fig. 2A,B). In the first experiment ( Fig. 2A), hamsters in the vehicle-treated group experienced slight weight loss by day 4, but the weight loss was not statistically significant. No decrease in weight was observed for hamsters treated either with auranofin (at 0.125 mg/kg and 0.25 mg/kg) or vancomycin ( Fig. 2A). Similarly, in the second experiment, hamsters treated with either auranofin (1 mg/kg and 5 mg/kg) or vancomycin did not exhibit signs of weight loss (Fig. 2B).
In vivo efficacy of auranofin in a relapsing CDI hamster model. One of the main problems associated with CDI is the high incidence of recurrence (in 15 to 50% of cases) following initial success with antibiotic treatment 35 . Relapsing CDI occurs due to the presence of C. difficile spores that germinate in the gut into vegetative cells that colonize the intestine and subsequently produce toxins 36,37 . Recurrence of infection occurs in approximately 20% of patients 15,38,39 . Additionally, 20% of patients who experienced a relapsing episode of www.nature.com/scientificreports/ C. difficile reportedly died within 30 days of diagnosis 40 . Moreover, it was reported that up to 65% of patients successfully treated from CDI recurrence will relapse again in the future 41,42 . Consequently, relapsing CDI represents a difficult and challenging problem facing healthcare systems that requires the discovery of new, more effective agents. With this issue in mind, we sought to investigate the activity of auranofin in preventing C. difficile relapse in hamsters. We initially tested the activity of low doses of auranofin (0.125 mg/kg and 0.25 mg/kg) in a relapsing CDI hamster model (Fig. 3A). This study was followed by another relapsing CDI hamster study investigating the activity of higher auranofin doses (1 mg/kg and 5 mg/kg) (Fig. 3B). In both studies, animals were infected with C. difficile and treatments were discontinued after 5 days. Hamsters were subsequently monitored for survival and possible CDI relapse.
As depicted in Fig. 3A, vehicle-treated hamsters became moribund following C. difficile challenge resulting in 100% mortality by day 5. This result is in agreement with previous studies [30][31][32] . Following discontinuation of treatment, vancomycin protected 100% of infected animals through day 12 (90% survival). By day 19, 80% of hamsters in the vancomycin treatment group were alive, which remained unchanged until the end of the experiment. On the other hand, animals administered auranofin (0.25 mg/kg) exhibited a recurrence rate of 50%. Three hamsters died during the post-treatment period resulting in 30% survival by the end of the experiment (statistically significant protection when compared to the vehicle-treated group). Auranofin (0.125 mg/kg) was slightly less efficacious with an overall survival of 20% (50% recurrence rate, as 2 out of 4 hamsters died after the discontinuation of treatment) (Fig. 3A).
We next tested the ability of auranofin at higher doses to prevent CDI recurrence. Two doses of auranofin (1 mg/kg and 5 mg/kg) were evaluated in addition to the vehicle (negative control) and vancomycin (standardof-care antibiotic). As shown in Fig. 3B, CDI resulted in the mortality of vehicle-treated hamsters with 40%, 60%, 90% and 100% mortality observed on days 3, 4, 5 and 6, respectively. During the post-treatment stage, vancomycin protected all infected animals through day 13. Starting on day 14, a stepwise pattern of mortality was observed ultimately resulting in 60% survival (40% relapse) at the end of the experiment. This pattern is typically observed with vancomycin treatment 30,42 . Auranofin (5 mg/kg) protected all infected hamsters after the discontinuation of treatment through day 8, at which point one hamster died. By day 10, 80% of hamsters The data are presented as average weight (g) (mean ± standard deviation) for each group. A twoway ANOVA with post-hoc Dunnett's test for multiple comparisons (P < 0.05) found no significant difference between the average weight for each group after receiving treatment, as compared to that before the start of treatment (day 0). www.nature.com/scientificreports/ treated with auranofin (5 mg/kg) were alive. By day 12, 60% of hamsters were alive, which was maintained until the end of day 21. On the other hand, auranofin (1 mg/kg) was less efficacious with 40% mortality observed during the treatment stage (until day 5). Following the discontinuation of treatment, 3 hamsters succumbed to relapsing CDI, which resulted in 30% survival by the end of the experiment (statistically significant protection as compared to the vehicle-treated group) (Fig. 3B). Additionally, the average body weight results of the two experiments evaluating low and high doses of auranofin (Fig. 4A,B), found that hamsters treated with auranofin exhibited a slight loss in their average body weight through days 8-10. This was followed by an increase in body weight of hamsters until the end of the study. The initial weight loss was attributed to animals that died later in the study whereas hamsters that survived exhibited an overall increase in average body weight.
After the conclusion of each experiment, hamsters were humanely euthanized and the cecal tissues were aseptically removed, homogenized, diluted and plated to determine the C. difficile CFU count inside each hamster's cecum.
Low doses of auranofin (0.125 mg/kg and 0.25 mg/kg) were less effective in reducing the C. difficile counts inside the cecal tissues generating a 0.48-log 10 reduction (with 0.125 mg/kg dose) and 1.2-log 10 reduction (with 0.25 mg/kg dose), respectively (Fig. 5A). Statistical analysis of the data for auranofin (both at 0.125 mg/kg and 0.25 mg/kg) determined that this reduction in bacterial burden was not significant. Notably, two-thirds of hamsters in the auranofin (0.25 mg/kg) group that survived until the end of the study exhibited bacterial CFU counts in the ceca that were below the limit of detection (2.80 log 10 (CFU/mL)). One hamster in the auranofin (0.125 mg/kg) group also exhibited a CFU count that was below the limit of detection. In contrast, vancomycin significantly reduced the bacterial CFU count by 3.1-log 10 , with 7 hamsters exhibiting bacterial CFU counts in the ceca that were below the limit of detection (Fig. 5A).
In the second experiment, evaluating the activity of higher doses, auranofin (5 mg/kg) was slightly superior to vancomycin in decreasing the burden of C. difficile in the cecal tissues of infected hamsters (Fig. 5B). Auranofin (5 mg/kg) significantly reduced the C. difficile CFU count generating a 3.15-log 10 reduction. On the other hand, vancomycin (20 mg/kg) generated a 2.65-log 10 reduction. It is worth noting that 7 hamsters in the auranofin (5 mg/kg) group and 6 hamsters in the vancomycin group presented with C. difficile CFU counts that were below the limit of detection (2.80 log 10 (CFU/mL). Additionally, auranofin (1 mg/kg) significantly reduced the  www.nature.com/scientificreports/ C. difficile CFU count by 1.75-log 10 ; 2 hamsters in this group exhibited bacterial CFU counts that were below the limit of detection.
In conclusion, this study investigated the efficacy of auranofin in vivo in a CDI hamster model. Auranofin significantly protected hamsters against lethal CDI when administered at the doses of 1 mg/kg or 5 mg/kg. Furthermore, auranofin (5 mg/kg) was as effective as vancomycin in preventing CDI recurrence in hamsters. Interestingly, auranofin (5 mg/kg) was superior to vancomycin in reducing C. difficile counts present in the cecum of infected hamsters. These results indicate that auranofin merits further investigation as a supplement to the dry pipeline of anti-CDI therapeutics. Follow-up studies are warranted to investigate the efficacy of higher doses of auranofin and to evaluate the in vivo activity of auranofin in combination with other anti-CDI drugs.

Preparation of C. difficile inoculum for infection of hamsters.
C. difficile VA11 (UNT103-1) was used to infect hamsters. This bacterial strain is a toxigenic clinical isolate that was responsible for multiple CDI outbreaks in North America 43,44 . The C. difficile inoculum used for infection was prepared as described previously 43 . Briefly, resuspended bacterial plates grown onto reinforced clostridial medium + 1% oxyrase were diluted to 1 × 10 7 CFU/mL. The bacterial inoculum was diluted, plated, and counted on reinforced clostridial medium before being used to infect hamsters.
In vivo efficacy of auranofin in a CDI ileocecitis hamster model. Both studies were performed as a service provided by the University of North Texas Health Science Center (Fort Worth, TX, USA). The studies are in compliance with the Animal Research: Reporting of In Vivo Experiments (ARRIVE) guidelines. Male Golden Syrian hamsters (weighing 80-100 g) were housed in individually ventilated cages (2 per cage) and received food and water ad libitum. The CDI hamster model was performed as described previously 30,33,34,45 . All hamsters were injected with clindamycin (10 mg/kg) subcutaneously. Twenty-four hours after clindamycin pretreatment, hamsters were infected via oral gavage with 0.75 mL of the previously prepared C. difficile inoculum (~ 7.5 × 10 6 CFU/hamster). The bacterial inoculum used was re-counted after infection to confirm the infective dose.
Following infection, hamsters were randomly allocated into groups (n = 10) for treatment. Twenty-four hours post-infection, hamsters were treated with low doses of auranofin (0.125 mg/kg and 0.25 mg/kg), vancomycin (20 mg/kg), or the vehicle (10% DMSO in PBS). In a follow-up study, two groups of infected hamsters were treated with higher doses of auranofin (1 mg/kg and 5 mg/kg), one group was treated with vancomycin (20 mg/ kg), and one group was vehicle-treated. Treatments were administered orally via oral gavage and continued once daily for 5 days. Hamsters were observed throughout the duration of each experiment for signs of mortality and morbidity, the presence of diarrhea (wet tail), and overall appearance (activity, general response to handling, touch, ruffled fur). Hamsters were weighed every other day. Animals exhibiting moribund state such as prolonged periods of weight loss, prolonged lethargy (> 3 days), paralysis, skin erosions, hunched posture and distended abdomen, were euthanized 43 . In vivo efficacy of auranofin in a relapsing CDI hamster model. Hamsters in each experiment were infected, as described above. In the first study, hamsters were treated with either auranofin (0.125 mg/ kg), auranofin (0.25 mg/kg), vancomycin (20 mg/kg), or the vehicle for 5 days. In the second study, hamsters were treated with higher doses of auranofin (1 mg/kg and 5 mg/kg), vancomycin (20 mg/kg), and the vehicle for 5 days. Thereafter, treatments were discontinued and hamsters were observed for disease symptoms, recurrence of CDI and signs of mortality and morbidity (described above) until the 21 st day. Hamsters judged to be in a moribund state were euthanized. Animals that died during the observation period in each experiment were necropsied. Additionally, the contents of deceased hamsters' cecal tissues were diluted in PBS and plated anaerobically onto modified reinforced clostridial agar to obtain C. difficile CFU counts. After the end of each experiment, surviving hamsters from each experiment were humanely euthanized using CO 2 asphyxiation. The contents from each hamster's cecal tissues were diluted in PBS and plated onto modified reinforced clostridial agar to obtain CFU counts.

Data availability
Data presented in this study are available from the corresponding author upon a proper request.