Phenotypical and Genotypical Comparison of Clostridium difficile Isolated from Clinical Samples: Homebrew DNA Fingerprinting versus Antibiotic Susceptibility Testing (AST) and Clostridial Toxin Genes

Background Clostridium (Clostridioides) difficile is recognized as the major cause of healthcare antibiotic-associated diarrhea. We surveyed a molecular epidemiological correlation between the clinical isolates from two general hospitals in Iran through clustering toxigenic types and antibiotic susceptibility testing (AST) accuracy. Methods Study population included 460 diarrhoeic specimens from inpatients with a history of antibiotic therapy. All samples underwent enriched anaerobic culture, confirmed by detection of gluD gene with PCR. Toxin status and AST were assessed by the disk diffusion method (DDM) and minimal inhibitory concentrations (MICs) of metronidazole, vancomycin, and rifampin. C. difficile outbreak was analyzed through conventional PCR by tracing toxin genes and Homebrew pulsed-field gel electrophoresis (PFGE) for characterizing isolates within our healthcare systems. Results A total of 29 C. difficile strains were isolated by enriched anaerobic culture from the clinical samples. Among them, 22 (4.8%) toxigenic profiles yielded toxins A and B (tcdA, tcdB) and binary toxins (cdtA, cdtB). The minimum inhibitory concentration (MIC) was 18.1% and 9% for vancomycin and metronidazole, and all isolates were susceptible to rifampicin and its minimum inhibitory concentration was at <0.003 μg/mL. The most dominant toxigenic and antibiotic-resistant “pulsotype F” was detected through PFGE combined with multiple Clostridial toxigenic pattern and AST. Conclusions DNA fingerprinting studies represent a powerful tool in surveying hypervirulent C. difficile strains in clinical settings. Resistance to vancomycin and metronidazole, as first-line antibiotics, necessitate accomplishment of proper control strategies and also prescription of tigecycline as a more appropriate option.


Introduction
Clostridium difficile (new taxonomic name: Clostridioides difficile) is a Gram-positive, spore-forming, obligate anaerobe and recognized as the most common cause of nosocomial and gastrointestinal infections such as mild diarrhea, severe pseudomembranous colitis, and toxic megacolon. e pathogenicity of this bacterium is related to the toxin production of A and/or B and binary toxins which are encoded by tcdA, tcdB, and cdtA and cdtB genes [1,2].
C. difficile infection (CDI) is initiated following antimicrobial consumption and eventuates in disruption of the normal colon microflora. Antibiotic therapy may also cause C. difficile antibiotic resistance in patients suffering from CDI and is a source of high morbidity and mortality worldwide.
Epidemiological studies of C. difficile in European countries have shown an increase in infection rate over time. CDI incidence in the United States in 2011 was more than 400,000 cases and resulted in 29,000 deaths in patients mostly aged above 65 years. CDI mortality rate before the year 2000 was low with a rate of less than 2%; however, it has since increased to 16.7% [3,4]. e severity of CDI in the world is influenced by overuse of antibiotics (especially fluoroquinolones, macrolides, and β-lactams), prolongation of hospitalization, and increase in the aging population [5,6]. e easy transmission of C. difficile through the oral-fecal route results in high persistence in the environment and is a major hospital problem. Early detection of CDI is important to prevent the transmission of the organism in clinical settings, as well as in managing the prescription of antibiotics [7].
Although the first-line treatments for CDI are metronidazole and vancomycin, fidaxomicin is also considered a complementary therapy. However, several cases of resistance to metronidazole and vancomycin have been reported around the world [8,9]. Determination of antimicrobial resistance patterns is critical for both patient treatment and epidemiological studies.
For instance, reports of outbreaks from Canada, the United States, Asia, and the UK confirmed that fluoroquinolone resistance was related to the emergence of C. difficile NAP1/027/BI [5,10].
According to studies from Iran by Goudarzi et al., the resistance to metronidazole and vancomycin was 5.3% and 8%, respectively [11]. However, Shoaei et al. reported 100% of isolates to be susceptible to metronidazole and 11.7% of isolates to be resistant to rifampicin in 2019 [10]. Unlike the several epidemiological investigations reported from Europe, North America, and Australia, there are limited studies in Middle East Asia [12].
Metronidazole (nonsevere CDI) and vancomycin (severe CDI) are the first-line treatments of CDI. Although vancomycin was more effective than metronidazole for chemotherapy of severe and mild/moderate CDI, several cases of resistance to metronidazole and vancomycin have been reported from around the world [8,9,13].
Some antibiotics that are mostly related to C. difficileassociated disease (CDAD) are clindamycin, ampicillin, and cephalosporin, which may contribute as important risk factors for the progress of CDAD [12,14,15]. erefore, the emergence of metronidazole-nonsusceptible C. difficile is a serious concern in clinical settings [16].
Furthermore, multidrug-resistant (MDR) strains have emerged owing to the uncontrolled usage of antibiotics. Hence, disclosure of susceptibility profile is a strategy toward lowering the increasing antibiotic resistance trend.
In the present study, molecular epidemiology of C. difficile infection in the two general hospitals of Tehran was characterized by pulsed-field gel electrophoresis (PFGE), possession of A and B toxin and binary toxin genes, and also antimicrobial susceptibility pattern.

Study Population.
A total of 460 clinical stool samples were collected during the years 2017 to 2019. Diarrhoeic stool samples belonged to adult patients with a history of antibiotic therapy from 2 to 8 weeks, followed by symptomatic antibiotic-associated diarrhea [17]. e study population was hospitalized in several wards, e.g., the intensive care unit (ICU), bone marrow transplantation, gastroenterology, cardiac surgery, renal disorder, and pulmonary disease ( Figure 1).

Enriched Toxigenic Culture and Identification C. difficile.
e fecal specimens were transported at room temperature and cultured anaerobically within 8 hr of collection or stored at 4°C for no more than 48 hr. Toxigenic culture was performed after isolation of C. difficile [18]. One portion of each sample was cultured regularly, and the rest was exposed to alcoholic shock for 1 hr before being cultured to inhibit the growth of other bacteria in feces [19].
Treated and untreated samples were inoculated onto the cycloserine-cefoxitin fructose agar, enriched with vitamin K 1 solution (1 mg/mL and hemin solution 5 mg/mL), placed in anaerobic jars (Merck) with a Gas Pack Anaerocult ® A (Merck, Germany), and incubated at 37°C for 2-5 days [20].
To optimize the growth of C. difficile, suspicious colonies were subcultured under anaerobic condition into BHI agar supplemented with 5% (v/v) sheep blood and incubated at 37°C for 24 hr. BHI agar was used for investigating colony characteristics (flat, horse barn odor, and Gram staining) and DNA extraction. Molecular identification of C. difficile was performed by conventional PCR of specific gene glutamate dehydrogenase (gluD). Confirmed colonies were preserved at − 80°C for long-term storage. (Qiagen, Germany), according to the manufacturer's instructions.

Detection of C. difficile Toxin Genes.
To detect enterotoxin (tcdA) and cytotoxin (tcdB) and binary toxin (cdtA, cdtB) genes, endpoint PCR was performed on DNA extracted from C. difficile isolates. e primers are shown in Table 1. PCR amplification was conducted as described in a previous study. In brief, thermocycler condition covered denaturation at 94°C for 10 min, 30 cycles of 94°C for 50 s, 52°C for 50 s, and 72°C for 50 s, with a final extension at 73°C for 10 min [21][22][23].

Antibiotic Disks.
Disk diffusion was performed with the following material: antibiogram disks obtained from Merk, Germany, and ROSCO, Denmark. Minimum inhibitory concentration (MIC) test strips were bought from Liofilechem, Italy. MIC of vancomycin was determined by the agar dilution method and MIC for metronidazole was determined by the test strips (Liofilechem, Italy) method as recommended by the EUCAST breakpoints on Brucella blood agar supplemented with 5% sterile sheep blood, 5 μl/mL hemin, and 1 μl/mL vitamin K 1 , after 24 hr of incubation at 37°C in the anaerobic jar (Gas Pack Anaerocult ® A Merk, Germany) [24].
MIC values were tested using the following MIC ranges: vancomycin >2 mg/L and metronidazole >2 mg/L, based on the EUCAST breakpoint [25]. An agar plate without any antimicrobial agent was permanently incubated as growth control. An isolated C. difficile colony was tested for susceptibility to vancomycin and metronidazole by the agar dilution method [26]. Double dilution of each antibiotic was conducted in 1280 μg/mL of stock solution, and it was added to enriched Brucella agar with 5 μl/mL hemin, 1 μl/mL vitamin K 1 , and 5% (v/v) sheep blood.
Plates with double concentration of antibiotics were prepared, namely, 0.25-16 μg/mL for vancomycin, 0.0002-32 μg/mL for rifampicin, and 0.12-64 μg/mL for metronidazole. e suspension equivalent of C. difficile 0.5 McFarland standard was prepared in Brucella broth. e MIC results were read after 48 hr of incubation at 37°C under anaerobic condition (Gas Pack Anaerocult ® A Merk, Germany).

Clinical Data of Patients.
During this study, we screened 460 patients who were suspicious of CDI. Twenty-nine (6.3%) stool samples were positive in culture and confirmed as C. difficile through endpoint PCR for gluD. Intended patients were hospitalized in the gastroenterology (n � 11), intensive care unit (ICU) (n � 9), bone marrow transplantation (n � 3), pulmonary disease (n � 3), cardiac surgery (n � 2), and renal disorder (n � 1) wards ( Figure 1). Of 29 positive cultures, 16 patients were female (55.1%) and 13 were male (44.8%) and the average age was 54.3 years. Analysis of patient history demonstrated that the medium hospitalization duration was 17.1 days and 60% of patients used at least 3 antibiotics before sampling. e frequent antibiotics administered were meropenem (79.3%) and vancomycin (48.2%) ( Table 2).

Antibiotic Susceptibility Tests.
We used the agar dilution method to assess the minimum inhibitory concentration of vancomycin, metronidazole, and rifampicin in toxigenic isolates. e MIC50 and MIC90 results for the three antibiotics are demonstrated in Table 4. Metronidazole and vancomycin resistance was shown in 9% and 18.1% of isolates, respectively, while all isolates were susceptible to rifampicin and the minimum inhibitory concentration was at <0.003 μg/mL.

Multidrug Resistance (MDR). MDR indicates resistance
to one agent in three or more antibiotic classes. High-level resistance to ciprofloxacin was detected in most of the C. difficile isolates. Two (5.5%) isolates were MDR and exhibited resistance to vancomycin, metronidazole, and ciprofloxacin.

PFGE.
A dendrogram, produced from PFGE data, demonstrated 22 toxigenic isolates divided into 11 clusters and 13 subtypes (based on a similarity value of 0.80) (Figure 2). e most dominant type was pulsotype F which was identified in 3 (13.6%) isolates from gastroenterology and ICU wards. Pulsotype F was toxigenic with (tcdA+/ tcdB+) and binary toxin (cdtA and cdtB) genes. Twenty (tcdA+/tcdB+) isolates had 10 different pulsotypes and 12 subtypes. ese pulsotypes were identified in gastroenterology and ICU wards in both hospitals. All of the isolates that distinguished into 4 pulsotypes were screened in the Van: vancomycin, Mtz: metronidazole, and Rif: rifampicin.
ICU ward. Two (tcdA− /tcdB+) isolates showed the same pulsotype which belonged to the ICU ward. Both types C1 and K2 showed concurrent resistance to metronidazole and vancomycin; these types were isolated from gastroenterology and ICU wards.

Statistical Analysis.
e results were analyzed through one-way analysis of variance (ANOVA) and pairwise two-

Discussion
In the last decade, with increasing nosocomial diarrhea among people in North America and Europe, CDI has become a major problem [30]. However, the epidemiology of CDI is less known in Asia in general and the Middle East, in particular [31]. In this study, 460 suspicious patients were evaluated for C. difficile infection, antibiotic resistance pattern, and molecular characteristics. PFGE was performed to demonstrate the epidemiological characteristics of C. difficile isolates in our local health systems.
CDI prevalence in our study was 4.8% (22/460), comparable to the studies from the United States and Europe [4]. e prevalence of CDI in Kuwait and Qatar was reported to be 7.2% and 7.9%, respectively [32,33]. In a survey performed in Saudi Arabia, the incidence of CDI was 1.7 per 10,000 patients [34]. e annual CDI prevalence in Iran in the years 2017 and 2019 was 18.1% and 11.4%, respectively [10,33].
e risk factors associated with CDI include old age (≥65 years), antibiotic consumption, hospitalization, and exposure to healthcare systems [35]. Our sample population had been exposed to antibiotics for 2 months prior to the study (Table 2), and the mean duration of hospitalization was 17.1 days. Analysis of the patients' history demonstrated that  Canadian Journal of Infectious Diseases and Medical Microbiology beta-lactams were the most common antibiotics before the occurrence of CDI. Our study reported antibiotics therapy panels including beta-lactams, fluoroquinolones, and lincosamides.
Although metronidazole and vancomycin are the current choices for treatment of mild-to-moderate CDI and severe infection, susceptibility to these antibiotics has been gradually decreasing [13,36,37]. In a study conducted in Israel, the susceptibility to metronidazole and vancomycin among ribotype 027 was 44.6% and 87.7%, respectively [38]. A study of antimicrobial resistance among toxigenic C. difficile isolates in Iran in 2013 showed resistance to metronidazole and vancomycin to be 5.3% and 8%, respectively [11]. Also, current studies in Iran showed a susceptibility decrease to all antibiotics [10].
In our study, the resistant phenotype was observed in 5.5% isolates. e MIC90 for metronidazole was 1 mg/L. However, 77.2% of isolates were inhibited in <1 mg/L concentration of metronidazole and 2 isolates were resistant to >256 mg/L of metronidazole. According to data from the present study, up to 81% of isolates were inhibited with 1 mg/L of vancomycin. However, 4 isolates were resistant to vancomycin (MIC was 4 mg/L for two isolates and 8 mg/L for two isolates). e percentage of MDR C. difficile ranges from 2.5% to 66% in various countries. Noticeably, resistance to vancomycin and metronidazole is a great concern that necessitates a proper consumption route.
Previous studies from Iran's neighboring countries report low resistance to metronidazole and vancomycin as assessed by disk diffusion assay and MIC. In addition, in East Asian and European countries, the rate of resistance to these antibiotics has been low (0-6.3%) as confirmed by various methods. Owing to the high-level metronidazole resistance, its prescription and consumption should be confined.
Based on disk diffusion assays, all isolates were susceptible to tigecycline. e majority of isolates were susceptible to commonly prescribed agents based on both the antibiotic susceptibility test and MIC results.
In the present study, the MIC50 for vancomycin was 1 mg/L and MIC90 was 8 mg/L, breakpoint to vancomycin was MIC >2 mg/L, and 4 isolates were vancomycin-resistant. e MIC50 and MIC90 of metronidazole were 0.5 mg/L that was significantly lower than the susceptibility category breakpoint of ≥32 mg/L. Only two isolates were resistant to metronidazole with MIC ≥265 mg/L. Toxigenic and drug-resistant C. difficile has been reported in various regions of the world. Accordingly, an urgent antibiotic susceptibility test report is essential alongside pathogenicity assessment to avoid the selection of nonsusceptible isolates. erefore, based on previous research studies on susceptibility to metronidazole and vancomycin, a subinhibitory concentration of these antibiotics can promote the production of biofilms and the resistance to metronidazole and vancomycin in C. difficile isolates. In case of failure of antibiotic therapy, tigecycline has been proved highly effective [39,40].
Furthermore, resistance to metronidazole and vancomycin may be due to overuse of these antibiotics in patients. According to our study, 16.6% and 15.5% of patients had a history of usage metronidazole and vancomycin, respectively, and generally, 60% of the patients used at least three prior antibiotic therapies.
Our dendrogram analysis showed that PFGE type F was the most common pulsotype identified (13.6%). All the patients harbouring pulsotype F were positive for binary toxins (cdtA and cdtB) and also tcdA and tcdB genes with a high genetic correlation. ese patients were hospitalized in two different wards in the same hospital.
In the present study, genetic diversity among 22 toxigenic C. difficile strains was high and isolates had a low genetic correlation with each other. In addition, both pulsotypes C and K (4 isolates) were vancomycin-resistant types, but they had a low genetic correlation. Isolates in pulsotypes C and K were detected in different wards in a hospital, namely, gastroenterology, ICU, and BMT wards. A− B+ toxigenic genotype was observed in 2 isolates, belonging to pulsotype A, and these were obtained from the ICU. is pulsotype was completely susceptible to metronidazole, vancomycin, and rifampicin. It is noteworthy that pulsotypes with A− B+ toxin gene were different in our study from that of Goudarzi et al. [11].

Conclusion
Our study of adult inpatients covered antibiogram pattern and showed low correlation genetic diversity in the C. difficile toxin profile. Our findings highlight the necessity for continuous monitoring of the clinical history of the inpatients and antibiotic treatment procedures. It is noteworthy that our analysis was limited by the lack of strain diversity and could be improved by including more hospitals. Furthermore, the assumption of clonal transmission between present pulsotypes proved false. Finally, high susceptibility to tigecycline could prove useful for CDI treatment and must be investigated as an alternate therapy.

Conflicts of Interest
e authors declare that they have no conflicts of interest.