Methicillin-Resistant Staphylococcus aureus from Diabetic Foot Infections in a Tunisian Hospital with the First Detection of MSSA CC398-t571

This study sought to analyze the antimicrobial resistant phenotypes and genotypes as well as the virulence content of S. aureus isolates recovered from patients with diabetic foot infections (DFIs) in a Tunisian hospital. Eighty-three clinical samples of 64 patients were analyzed, and bacterial isolates were identified by MALDI-TOF. The antimicrobial resistance phenotypes were determined by the Kirby–Bauer disk diffusion susceptibility test. Resistance and virulence genes, agr profile, spa and SCCmec types were determined by PCR and sequencing. S. aureus was detected in 14 of the 64 patients (21.9%), and 15 S. aureus isolates were recovered. Six out of the fifteen S. aureus isolates were methicillin-resistant (MRSA, mecA-positive) (40%). The isolates harbored the following resistance genes (number of isolates): blaZ (12), erm(B) (2), erm(A) (1), msrA (2), tet(M) (2), tet(K) (3), tet(L) (1), aac(6′)-aph(2″) (2), ant(4″) (1) and fexA (1). The lukS/F-PV and tst genes were detected in three isolates. Twelve different spa-types were identified and assigned to seven clonal complexes with the predominance of agr-type III. Furthermore, the SCCmec types III, IV and V were found among the MRSA isolates. Moreover, one MSSA CC398-t571-agr-III isolate was found; it was susceptible to all antimicrobial agents and lacked luk-S/F-PV, tst, eta and etb genes. This is the first report on the prevalence and molecular characterization of S. aureus from DFIs and also the first detection of the MSSA-CC398-t571 clone in human infections in Tunisia. Our findings indicated a high prevalence S. aureus in DFIs with genetic diversity among the MSSA and MRSA isolates.


Introduction
Staphylococcus aureus (S. aureus) strains have become a leading cause of hospitalassociated and community-associated infections worldwide. Two mechanisms confer resistance to β-lactams in staphylococci with the most common being the production of β-lactamase, encoded by blaZ, which produces the hydrolysis of the β-lactam ring, rendering the β-lactam inactive. More than 90% of staphylococcal isolates now produce penicillinases [1]. The second mechanism is due to an altered penicillin-binding protein PBP2a, encoded by mecA, which is carried in a variable mobile element, namely, the staphylococcal chromosome cassette mec (SCCmec) [2]. This mechanism leads to resistance to a semi-synthetic penicillinase-resistant β-lactam called methicillin. Furthermore, the term methicillin resistance manifests as resistance to virtually all β-lactams with the exception of the latest generation of cephalosporin β-lactams [2]. MRSA strains can also acquire additional resistance to several commonly used non-β-lactam antimicrobials (e.g., aminoglycosides, macrolides, fluoroquinolones and tetracycline) and are currently considered as the first class of multidrug-resistant (MDR) pathogens [3].
Diabetic foot infection (DFI), defined as soft tissue or bone infection below the malleoli, is the most common diabetic complication that often leads to hospitalization and non-traumatic lower extremity amputation [4][5][6]. Many studies have shown that DFI is polymicrobial [7][8][9]. Particularly, S. aureus is the bacteria implicated the most [10]. In addition to its ability to acquire antimicrobial resistance (AMR) to many clinically important drugs [11], this microbe plays a significant role in DFIs by causing infections ranging from superficial to severe and potentially fatal systemic infections [12].
The overuse of antibiotics is one of the most serious issues with DFI treatment. The prescription of unsuitable antimicrobial treatment has an impact on the microbiota and encourages the selection and growth of MDR bacteria. Thus, the global emergence of MRSA has considerably restricted the available therapeutic options for staphylococcal infections [13].
The epidemiology of AMR in Tunisia has been very dynamic in recent years, and the available data give insights into the alarming situation, especially in hospital settings [14]. In contrast, only three earlier studies have been conducted on DFI. These studies were limited to the bacteriological profile of DFI patients and showed controversial results. Enterobacteriaceae were the main bacteria causing the infection in diabetics in two of them [8,9], while S. aureus was the most frequent pathogen isolated in the remaining one [15]. Thus, there are currently no data on the molecular characterization of the bacterial strains involved, risk factors or treatment of multi-drug resistance organisms in patients with DFI in Tunisia. The aim of the present study was to evaluate the prevalence of S. aureus isolated from diabetic patients admitted for infected foot ulcers in the multidisciplinary diabetic foot center of the International Hospital, Carthagene in Tunisia during the COVID-19 pandemic and to investigate their genetic relatedness, antibiotic resistance pattern and virulence characteristics.

Patient Characteristics and MRSA Prevalence in Ulcer Samples
As shown in Table 1, 64 patients were included in this study (48 men and 16 women with a mean age of 62.28 years); 6 of them had diabetes type 1 (4 male and 2 female), 57 diabetes type 2 (43 male and 14 female) and 1 male had diabetes secondary to acute pancreatitis. Since the diabetic foot center is international, the patients were from different African countries (Tunisia, Algeria, Libya, Chad and Guinea). Accordingly, the distribution of the nationality among the patients with foot ulcer infection was as follows: Tunisia (51.6%), Libya (39%), Algeria (6.3%), Chad (1.6%) and Guinea (1.6%). S. aureus was detected in 21.9% of all patients with DFI analyzed in this study (14/64). One isolate per positive sample was included, except in one patient in which two different isolates were recovered and both of them were included making a collection of fifteen S. aureus isolates. Six of these fifteen S. aureus isolates were MRSA (cefoxitin-resistant and mecA positive) (40%), and the remaining nine isolates were methicillin-susceptible S. aureus (MSSA). One of the patients carried one MSSA and one MRSA isolate. Thus, the overall prevalence of MRSA in the ulcers was 9.4% and reflected a proportion of 42.9% of the participants who had S. aureus foot ulcer infections.

Antimicrobial Resistance Pattern of the S. aureus Isolates
The resistance profiles of MRSA and MSSA isolates to the antimicrobial agents tested are presented in Table 2. All fifteen isolates showed resistance to at least one antibiotic. Multidrug resistance was found in 53.4% of isolates. All six MRSA isolates showed resistance to fusidic acid, four isolates to tetracycline and three isolates to tobramycin and gentamicin. Two MRSA isolates were resistant to ciprofloxacin, levofloxacin, trimethoprimsulfamethoxazole and erythromycin, and one to clindamycin, minocycline, mupirocin and rifampicin. All the MRSA isolates were susceptible to tigecycline, vancomycin, teicoplanin and linezolid.
Among the nine MSSA isolates, the blaZ gene was found in seven isolates (77.8%). The presence of erm(B) alone or in association with erm(A) was detected in two erythromycinresistant isolates. The msrA gene was identified in two isolates and the fexA gene in one isolate. Two strains had no resistance genes (Table 3).

Molecular Typing of Isolates
Twelve different spa types were identified among the fifteen S. aureus isolates. The spa type t127 was detected in three isolates, while others were detected only once: (t311, t037, t15077, t688, t084, t188, t355, t091, t012, t223 and t571). One of the fifteen isolates could not match any known spa sequence. For one patient, two S. aureus isolates were identified corresponding to two different spa types (t311 and t571). The isolate belonging to spa-type t571 was assigned to the clonal complex CC398.
Among the six isolates that carried the mecA gene, two of them harbored SCCmec type V, one isolate SCCmec type IVb and another SCCmec type III, and the remaining two isolates were not typable. The characterization of the agr system showed a predominance of agr group III (11 isolates, 73.3%). The agr group IV, II and I were detected in four isolates.

Virulence Profile
The lukF/lukS-PV genes encoding for Panton-Valentine leukocidin (PVL) were detected in two isolates (13.33%) typed as t127-MRSA and t355-MSSA. The tst and eta genes were found, each in one isolate. None of the strains carried genes encoding the ETB toxin.
In addition, all MRSA and seven MSSA isolates carried the scn gene of the IEC system, and they were ascribed to different IEC types (A, B, C, D, E and G). In addition, two MSSA isolates lacked the scn gene (IEC-negative). Thirteen isolates (87%) contained an IECconverting βC-Φs, as demonstrated by the presence of scn. The predominant IEC variant was type D (sea, sak and scn) found in four isolates (30.7%). Variant E (sak and scn), G (sep, sak and scn), C (chp and scn), A (sea, sak, chp and scn) and B (sak, chp, and scn) were present in three, two, two, one and one isolates, respectively (Table 3).

Discussion
Methicillin-resistant S. aureus is a dominant hospital pathogen in Tunisia and worldwide [16]. Although antibiotic resistance of healthcare-associated staphylococci is well documented in Tunisia, no detailed information is available on antibiotic resistance and the molecular characterization of S. aureus isolated from diabetic ulcers. The current study fills this knowledge gap by analyzing the prevalence and molecular characteristics of S. aureus isolates in the International Tunisian Hospital, Carthagene. Of note, the patients involved were from different countries (Tunisia, Algeria, Libya, Chad and Guinea), thus reflecting the characteristics of S. aureus associated with DFIs not only in Tunisia but also on a wider geographical scale.
Studies have shown that prior use of antibacterial agents, hospitalization, MRSA nasal carriage and chronic wounds are risk factors for MRSA acquisition in patients with DFIs [17].
In our study, the prevalence of S. aureus detected in DFIs was 21.9%. Three previous retrospective studies from different Tunisian hospitals showed that S. aureus was isolated in 9%, 17% and 31% of DFIs [8,9,15]. This variation in percentages might be due to a difference in the geographical areas, the method applied to obtain cultural samples and study periods, especially the case of our study which took place during the COVID-19 pandemic and partly explaining the decrease in the number of consulting patients.
In our study, multi-resistance was found in 53.4% of isolates. The MDR phenotype is considered a common trait of MRSA isolates from various origins with resistance to several clinically relevant antimicrobial agents typically due to the acquisition of various mobile genetic elements (plasmids and transposons) causing treatment failure and significant associated human health burdens and healthcare costs [26,27]. In Tunisia, the high proportion of isolates showing this resistance phenotype may be related to the abuse of antibiotics with a frequent practice of self-medication. In this study, vancomycin, tigecycline, teicoplanin and linezolid were effective against all the S. aureus isolates. Fortunately, these antibiotics remain the best option to treat MRSA-associated infections, thus appropriate use of these antibiotics is highly recommended to avoid the selection of resistant strains.
Overall, high genetic diversity was found among the S. aureus isolates demonstrated by thirteen spa types and four agr groups with the predominance of agr-type III and spa-type t127 (CC1). These results are consistent with those of a recent review which highlighted that S. aureus strains isolated from diabetic foot ulcers in different countries are genetically diverse [28].
The spa type t127 was the predominant (three isolates, 20%). It was previously reported that t127 is associated with serious human infections in the United States and Germany [29][30][31]. In addition to a clinical origin, the spa type t127 has recently been reported in processed foods in China and in animals, indicating the risk of MRSA transfer from food and animal origins to humans or vice versa [32,33].
Interestingly, an MSSA isolate belonging to the spa type t571 (CC398 lineage) was found in one of the patients in our study. To our knowledge, this is the first report of MSSA-CC398 in human infections in Tunisia. Even though livestock-associated (LA)-MRSA-CC398 is closely related to food-producing animals [34], this MSSA-CC398-t571 subclade seems to be livestock-independent and has been detected in human invasive infections in different countries [35][36][37]. Similarly, a national French study showed a consistent and significant association between MSSA-CC398 and diabetic foot osteomyelitis [38].
The Panton-Valentine leukocidin is the most studied toxin produced by S. aureus [39]. In our study, only two S. aureus isolates contained the genes of PVL (13.3%). It has been suggested that PVL-positive strains are less frequently detected among DFIs as this gene is mostly prevalent in community species [20]. However, some studies reported higher PVL gene rates reaching 14.1% and 57% in Algeria and India, respectively [24,40].
The tst gene, encoding toxic shock syndrome toxin-1, was found in one isolate. Other studies reported different rates of tst positive strains in cases of diabetic foot ulcers ranging from 13% to 19% [40][41][42], whereas no strains of S. aureus from DFIs were positive to tst in a previous African study [43].
Among our fifteen S. aureus isolates, one of them (6.7%) harbored the eta gene and none harbored the etb gene. This result was similar to a previous report from France showing that 3% of S. aureus strains from diabetic foot ulcers were eta-positive, but no strain harbored the etb gene [41]. Another European study noted that 13% and 17% of the strains from DFIs harbored eta and etb genes, respectively [42]. However, these virulence factors were not detected in cases of DFIs previously reported in Algeria [24] or in diabetic foot osteomyelitis in France [38]. S. aureus is an extremely versatile pathogen in humans with different virulence phenotypes, suggesting that the virulence determinants did not spread homogeneously among various genetic backgrounds.

Bacteria Collection and Identification
Between September 2019 and October 2020, a total of 83 samples (tissue biopsy and/or deep swab and/or aspiration) were analyzed from 64 patients who were admitted for DFI at the diabetology department of the International Hospital Carthagene of Tunisia; this hospital has a capacity of 300 beds and 55 intensive care beds. Inclusion criteria were diabetic patients with any type of diabetes and aged ≥18. Patients who received antibiotic therapy within 3 months before the consultation, pregnant patients and those with a mental disorder that precluded understanding the scope of the study were excluded from the work.
Samples (aspirations (n = 5), deep swabs (n = 42), and tissue biopsies (n = 36)) were collected after wound debridement and cleansing with sterile physiological saline (as part of part of the routine clinical work of the hospital). Swabs were taken from open wounds by sterile cotton swabs from the base of the ulcer wound and aspirations were taken for closed lesions (abscesses and other fluctuant infected tissues) by needle aspirates after cleaning with polyvidone-iodin solution. In operated patients, intra-operative samples were obtained by infected soft tissues biopsies in the operating room. The samples were transported in sterile tubes without transport medium and were processed immediately upon arrival for bacteria recovery in the clinical laboratory as part of the routine diagnosis at the hospital.
The samples were inoculated on blood agar (Oxoid, UK, CM0271) and incubated in stove at 37 • C for 24 h. Colonies suspected to be Staphylococcus were subcultured on mannitol salt agar selective medium (Oxoid, UK, CM0085) for specific detection of S. aureus. All isolates were identified by MALDI-TOF-MS system using the standard extraction protocol recommended by Bruker (Bruker, Bremen, Germany). The identification of the S. aureus colonies was also confirmed by PCR of the gene nuc [44].

Screening of Methicillin-Resistant S. aureus Isolates (MRSA)
Isolates resistant to cefoxitin (FOX, 30 µg) on MH agar according to CLSI recommendations were confirmed for the presence of the mecA gene by PCR technique as described previously [45,46]. mecA-positive isolates were considered as MRSA isolates.

Molecular Typing of Isolates
All S. aureus isolates were characterized by amplification and sequencing of the polymorphic region of the staphylococcal protein A-encoding gene (spa) [47]. The obtained sequences were analyzed using Ridom Spa-type software version 1.5.21 (Ridom GmbH, Würzburg, Germany) to determine the spa type. In addition, a specific PCR was carried out to identify the CC398 lineage, targeting the CC398-specific variant of sau1-hsdS1 [48]. The clonal complex of the remaining isolates was assigned, when possible, according to the spa-type.
MRSA isolates were subjected to SCCmec typing by PCR strategy to determine the mec gene complex and the ccr gene complex as described by Zhang et al. [49]. The identification of the agr allele group (I-IV) was also determined by multiplex PCR as described earlier [50].

Occurrence of Virulence and Immune Evasion Cluster (IEC) Genes
All S. aureus isolates were screened using PCR for the following staphylococcal virulence genes: Panton-Valentine leukocidin (lukS/F-PV), toxic shock syndrome toxin (tst) and the exfoliative toxins (eta and etb) as previously described [51]. The immune evasion cluster (IEC) genes (scn, chp, sak, sea and sep) were examined by PCR and, based on the genes obtained, the isolates were classified into seven IEC types [51,52]. The scn gene (encoding the staphylococcal complement inhibitor) was used as a marker of the IEC system. Positive controls from the collection of the University of La Rioja were included in all PCR assays.

Conclusions
This is the first study to report the prevalence rate, the antimicrobial resistance profile, virulence genes and molecular typing of S. aureus isolates obtained from diabetic foot wounds in Tunisia. Our results indicate a high prevalence of S. aureus in DFIs with genetic diversity among the MSSA and MRSA isolates. A high number of MDR isolates harbored various AMR and virulence genes.
This study elucidates the recent regional epidemiological data on S. aureus implicated in DFIs which will be relevant for better guidelines for antibiotic use in clinical settings. These findings highlight the need for further studies focusing on the molecular surveillance of AMR for optimal management of DFI.