Emergence of Methicillin-Resistant Staphylococcus aureus ST239/241 SCCmec-III Mercury in Eastern Algeria

In this paper, we investigate the epidemiology of infections-associated Staphylococcus aureus (S. aureus) from the Medical Intensive Care Unit (MICU) at University Hospital Center of Constantine (UHCC) in Algeria, with a special emphasis on methicillin-resistant strains (MRSA) revealed by cefoxitin disks (30 μg), then confirmed by penicillin-binding protein (PBP2a) agglutination and real-time polymerase chain reaction (RT-PCR) targeting mecA and mecC genes. Staphylococcal Cassette Chromosome mec (SCCmec type), staphylococcal protein A (spa-type), multilocus sequence type (MLST), Panton–Valentine Leucocidin (PVL), and toxic shock syndrome toxin-1 (TSST-1) were further investigated in all isolates, and whole genome sequencing was performed for a selected subset of three hospital-acquired MRSA (HA-MRSA) isolates. A measurement of 80% out of the 50 S. aureus isolates were identified as HA-MRSA harbouring the mecA gene, and 72.5% of them were multidrug resistant (MDR). Twelve STs, four different SCCmec cassettes, fourteen spa types, ten isolates Panton–Valentine Leukocidin (PVL)-positive, and three isolates TSST-1 were identified. Interestingly, there was a high prevalence (n = 29; 72.5%) of a worrisome emerging clone: the HA-MRSA ST239/241 SCCmec-III mercury with PVL negative, resistant to β-lactams, aminoglycosides, quinolones, and tetracyclines. Other clones of HA-MRSA isolates were also identified, including PVL-positive ST80 SCCmec-IV/SCCmec-unknown (22.5%), ST34 SCCmec-V with TSST-1 positive (2.5%), and PVL-negative ST72 SCCmec-II (2.5%). Genome analysis enables us to describe the first detection of both PVL-negative HA-MRSA ST239/241 SCCmec-III mercury carrying ccrC, as well as SCCmec-V cassette, which dramatically changes the epidemiology of S. aureus infections in one of the hospitals in eastern Algeria.


Introduction
Staphylococcus aureus (S. aureus) is one of the leading bacteria that cause skin and soft tissues infections, bacteraemia, endocarditis, osteomyelitis, pneumonia, urinary tract infections, toxic shock syndrome and bloodstreams infections [1,2]. Methicillin-resistant S. aureus (MRSA) isolates have been involved in healthcare-associated infections (HA-MRSA) since the 1960s and community-associated infections (CA-MRSA) since the end of the 1980s [3]. Nowadays, treatment of S. aureus infections has become more challenging due to (i) the widespread and rapid dissemination of MRSA isolates in both community and hospital settings and (ii) the increasing prevalence of highly multi-drug resistant (MDR) MRSA, particularly in hospitalised patients [4,5]. Methicillin resistance is mediated by the expression of an altered penicillin-binding protein (PBP2 or PBP2a) encoded by methicillin-resistant gene variants (i.e., mecA, mecC) carried on the Staphylococcal Cassette Chromosome mec (SCCmec) [6][7][8]. Thus far, fourteen SCCmec cassettes (SCCmec-I to XIV) have been described on the basis of their structural organisation, their genetic content, and the combination of the mec and ccr (cassette chromosome recombinase) gene complexes [9][10][11][12]. Five of these mobile elements (SCCmec-I, II, III, IV and V) are the most-reported cassettes in MRSA isolates in the world [13]. SCCmec types I, II, and III are more frequently described in hospital-acquired (HA)-MRSA isolates, whereas community-acquired (CA)-MRSA isolates are more characterised by SCCmec types IV or V [14,15]. Interestingly, it has been reported that S. aureus isolates acquired SCCmec cassettes from coagulase-negative staphylococci (CoNS), especially Staphylococcus sciuri (S. sciuri) [16]. These CoNS harbour a high variety of SCCmec cassettes and can serve as a potential reservoir for SCCmec elements [17,18].
To elucidate the distribution and the epidemiology of S. aureus, the genotyping approach is predominantly performed based on staphylococcal protein A (spa) typing and multilocus sequence typing (MLST) [19]. However, molecular characterisation of SCCmec cassettes is also used for typing S. aureus isolates and for differentiating HA-MRSA from CA-MRSA isolates and their respective epidemiological origins [20]. In addition to antimicrobial resistance elements, the pathogenicity of S. aureus isolates is also related to the expression of various virulence genes encoding Panton Valentine Leukocidin (PVL) and toxic shock syndrome toxin-1 (TSST-1), among others. These genes can also serve as additional genetic markers in MRSA isolates characterisation [21,22].
Compared to other regions of the world, epidemiological studies published on MRSA isolates in North Africa are quite scarce [23] in spite of some investigations reporting the spread of MRSA isolates in Algerian hospitals. With the exception of the study by Basset et al. in 2015 describing the detection of SCCmec-III mercury-positive MRSA isolates identified as the typically nosocomial Brazilian clone ST239 at Bologhine Ibn Ziri University Hospital in Center of Algeria [24], no epidemiological studies have been carried out on HA-MRSA isolates harbouring this genetic element elsewhere in the country (especially in the east). Moreover, the work of Ouchenane et al. on MRSA at the Military Hospital of Constantine (from 2005 to 2007) did not present any complementary molecular data to really demonstrate the carriage of SCCmec-V [25]. Thus, we aim in this study to conduct an exhaustive epidemiological investigation on S. aureus isolates collected over two years from patients hospitalised in the Medical Intensive Care Unit (MICU) at University Hospital Center of Constantine (UHCC), Algeria (Eastern Algeria), through their prevalence, determination, and molecular characterisation with whole genome sequencing to investigate the clonality, the virulomes, and the resistomes of HA-MRSA.

Clinical Isolates
As presented in Table 1, 50 clinical isolates of S. aureus were successfully collected from patients with various infections. Blood from septicaemia cases was the main sample (44%; n = 22), followed by endotracheal intubation (40%; n = 20), central venous catheter, chest drain and urine (4%; n = 2), and, finally, pleural and urinary catheter (2%; n = 1 for each one). A measurement of 54% (n = 27) of the specimens was collected from women and the remaining 38% (n = 19) from men. All patients were between the ages of 2 and 84, and they were classified into nine age groups, with the age group 41-50 years old (16%; n = 8) being the most infected by S. aureus.
Seven resistance phenotypes were observed in MRSA isolates, with the phenotype penicillin-kanamycin-tobramycin-gentamicin-ciprofloxacin-ofloxacin-tetracycline being the most prevalent (55%; n = 22; p = 0.001) ( Figure 1). The statistical analysis could not prove any significant impact or correlation between the nature of samples, the sex/age of the patients, and the prevalence of MDR strains.

Molecular Characterisation
As shown in Figure 1, none of the 40 HA-MRSA clinical isolates harboured the mecC gene, and the mecA gene was detected in all of them. SCCmec typing revealed that both SCCmec-III and I were absent in all isolates; however, 72.5% (n = 29) of them were confirmed to harbour the SCCmec-III mercury cassette. The presence of this cassette type was significantly higher (p < 0.001), followed by SCCmec-IV cassette (20%; n = 8), and the remaining isolates carried the types II and V and an unknown SCCmec cassette (2.5%; n = 1 for each one).
Seven resistance phenotypes were observed in MRSA isolates, with the phenotype penicillin-kanamycin-tobramycin-gentamicin-ciprofloxacin-ofloxacin-tetracycline being the most prevalent (55%; n = 22; p = 0.001) ( Figure 1). The statistical analysis could not prove any significant impact or correlation between the nature of samples, the sex/age of the patients, and the prevalence of MDR strains.  Interestingly, co-resistance to β-lactams, aminoglycosides, fluoroquinolones, and tetracyclines was more established among the S. aureus isolates harbouring the SCCmec-III mercury cassette with a high prevalence of 75.9% (p < 0.001). All HA-MRSA strains with an MDR phenotype harboured the SCCmec-III mercury (p < 0.001). Inversely, the SCCmec-IV cassette was only identified among strains resistant to β-lactams, to kanamycin, and to erythromycin.
The only TSST-1 gene detected in this study was found in a SCCmec-V HA-MRSA strain of ST34 with t166 ( Figure 1).

Genomic Analysis of SCCmec Cassettes from Sequenced Genomes
To investigate the genetic structure of the detected SCCmec cassettes of our isolates, three of them (SAUR390, SAUR678, and SAUR1404, marked with a star in Figure 1) were sequenced using Illumina Miseq. The genome assembly resulted in genome sizes of 2 920 844-bp, 2 903 748-bp, and 2 853 533-bp assembled into 124 contigs, 147 contigs, and 177 contigs for SAUR390, SAUR678, and SAUR1404 isolates, respectively. As shown in Figure 2, sequence comparison with the SCCmec cassettes between our isolates and that of SCCmec-III mercury reveals high similarity (≥80% nucleotide identity), although some recombinations are identified. Interestingly, we can observe an integration of a complete Type I restriction-modification system (RMS) within the SCCmec of SAUR678 and SAUR1404 ( Figure 2).  Table S1.

Resistance Genes Analysis of HA-MRSA SCCmec-III Mercury
As shown in Table 3 and Supplementary Material Table S2, antibiotic resistance genes were identified across the chromosomes and the plasmids of the three sequenced genomes. These included β-lactams resistance genes mecA (methicillin) and blaZ (penicillin);  Table S1.

Discussion
In this epidemiological study, as part of the first molecular and genomic investigation of S. aureus healthcare-associated infections in MICU, our findings reveal a worrisome emergence with a high prevalence (72.5%) of HA-MRSA ST239/241-t037/138 isolates carrying PVL-negative SCCmec-III mercury element, responsible for healthcare-associated infections in the Constantine hospital (north-eastern Algeria). The high prevalence of HA-MRSA isolates we have recorded (80%) strongly corroborates earlier Algerian reports describing a prevalence of 75% in 2011 and 71.4% in 2014 at the cities of Tlemcen and Annaba, respectively [26,27]. This prevalence seems to have increased since the first study, rising from 33.2% in 2006 at one hospital in Algiers [28] to 80% in our recent investigation. Furthermore, it remains higher than those reported in European countries, including Italy (29%), Greece (27.3%), Romania (21.7%), Spain (12.5%), Germany (8.3%), France (7%), and the UK (0%) [29].
Most of our HA-MRSA isolates (42.5%) were recovered from the bloodstream of patients aged 41 to 50, most of whom were female [30], which again proves that HA-MRSA is one of the leading causes of bloodstream infections in MICU. In addition, our findings reveal significant spread diffusion of multidrug resistance among HA-MRSA isolates harbouring the SCCmec-III mercury element, which limits the options for effective antibiotic treatment of the infections due to these isolates. Inversely, previous results from a hospital setting in Algiers indicated that HA-MRSA strains harbouring the SCCmec-III mercury element are less prevalent and only few (28.6%) of them are MDR [24]. None of our MSSA isolates were MDR, which is similar to previously published results from Algeria [31]. On the other hand and in line with most previous reports in Algeria [32][33][34], our S. aureus isolates were entirely susceptible to vancomycin.
SCCmec types I, II, and III are mostly reported in HA-MRSA isolates, while SCCmec-IV and V are related to CA-MRSA [9]. The Brazilian MRSA clone (ST239, SCCmec-III, and PVL negative) has been reported in various countries [23,35]. Therefore, it seems that the prevalence of the Brazilian clone linked to the variant SCCmec-III mercury element in MICU at Constantine hospital is high (72.5%) in HA-MRSA isolates without the presence of PVL genes. In our study, which focused on the genetic characterisation of HA-MRSA isolates, the presence of S. aureus isolates harbouring different SCCmec-III mercury, spa types, and ST(s) has been attested. This is the first detection of ST239 and ST241 HA-MRSA isolates with SCCmec-III mercury and spa t037 and t138 in eastern Algeria. These ST239/ST241 SCCmec-III mercury clones are predominant (72.5%) with a highly significant prevalence of multi-resistance phenotypes (55%) (p < 0.001). As compared to reports from central Algeria, only 8.3% of HA-MRSA carried the SCCmec-III mercury element [24]. However, HA-MRSA isolates carrying SCCmec-V have been reported worldwide [36,37] and also in another hospital in the city of Constantine by no thorough molecular analysis [25,38]. Advanced analyses of the sequencing genome in our investigation improved our understanding of the structural organisation of the SCCmec cassettes. Therefore, our study has all the advantages of confirming the first detection of SCCmec-V, which is TSST-1 gene positive. We report here the detection of ST34 HA-MRSA isolates with SCCmec-V and spa t166, which had previously only been detected in MSSA strains in Africa [39,40]. The presence of this clone carrying SCCmec-V may be due to horizontal gene transfer [41].
SCCmec-IV, ST80 has been reported in both community-acquired and hospital-acquired infections in Algerian hospitals [24,34]. Patients in this MICU suffer from infections caused by 20% of HA-MRSA carrying the SCCmec type IV, and our own findings confirmed the presence of the MRSA-IV European clone.
Analysis of drug resistance and virulence genes revealed evidence of a great diversity in their combinations among our characterised strains. In general, HA-MRSA strains are commonly multidrug resistant [42], harbour SCCmec type II and III, and demonstrate multiresistance, containing additional antibiotic resistance genes, such as erythromycin and tetracycline, as well as methicillin. The SCCmec-III element confers resistance to a broad spectrum of antibiotics and heavy metals [43]. The multi-resistant nature of our MRSA clones could be explained by the presence of several resistance genes in the SCCmec cassettes, as previously described [44]. As reported by Soliman et al. (2020) [45] our strains harboured both mecA and blaZ encoding resistance to β-lactams. Several studies have mentioned a concordance between phenotypic and genotypic resistance results in MRSA strains [46,47], which is in perfect agreement with our findings. Rahimi et al. (2014) [48] reported that ermB was prevalent in just a few erythromycin-resistant staphylococci, and in the study by Abbassi et al. (2017) [49], this gene was not identified in HA-MRSA. The resistance to erythromycin in these studies was due to the carriage of ermA and ermC by MRSA strains in Iran. This aligns well with our findings, since ermB plays no role in macrolide-resistant HA-MRSA. The multi-drug resistant patterns for the HA-MRSA in our study are in keeping with what has been described in other countries in Africa. According to Bendary et al. (2020) [50], MDR HA-MRSA ST239/241 strains are less toxicogenic. This may be due to the acquisition of toxin genes on mobile elements to the detriment of extended antibiotic resistance [51]. As in our study, HA-MRSA ST239/241 strains with very low-level resistance to glycopeptides have been reported in many other countries [52,53], which proves that they are among the most effective drugs against infections due to this clone. Regarding resistance to fluoroquinolones, Japoni et al. (2011) [54] observed that SCCmec III isolates showed high rates of resistance to ciprofloxacin. This concurs well with our findings concerning the variant SCCmec-III mercury, as we detected many genes encoding resistance to these antibiotics among our isolates.
The wide collection of virulence factors harboured by S. aureus are differently present within clones [55,56] in relation to the high variety of infections for which they are responsible. PVL is a crucial cytotoxic virulence factor in serious infections (necrotising pneumonia, skin and soft tissue infections) [57,58]. It is widely described as the primary virulence factor driving the epidemic spread of the European clone [59]. This corroborates our results, as we found, for instance, ST80 to be PVL-positive and ST239/241 to be PVL-negative.
Several enterotoxin genes can be harboured by enterotoxigenic S. aureus strains [60]. Toxinogenic strains may pose a public health risk to consumers, as they can contaminate food and cause food poisoning. The TSST-1-positive MRSA strains are highly virulent and provoke a variety of health disorders, including toxic shock syndrome and various suppu-Pathogens 2021, 10, 1503 9 of 14 rative infections. In addition, the toxic shock syndrome toxin (TSST-1) and staphylococcal enterotoxins are reported to play a significant role in the proliferation of T cells, irrespective of antigenic specificity [61,62].

Conclusions
Due to the critical situation of patients admitted in to intensive care and given the high prevalence of HA-MRSA isolates described in Algeria, our study confirms that these findings in the MICU at the UHCC are linked to a new molecular and epidemiological change of HA-MRSA following the first detection of the ST239/241 SCCmec-III mercury and ST34 SCCmec-V clones. The first is expressed by high frequency and is significantly associated with multi-drug resistance to the antimicrobial agents used to treat patients in intensive care. This implies a serious threat in hospital settings, as well as in the community, requiring adequate control measures to prevent the spread, transmission dynamics, and epidemics of these clones. Proper management of antimicrobial agents and a better awareness among healthcare personnel is therefore needed to combat these nosocomial MRSA infections. In addition, other studies are needed to screen for CA-MRSA isolates based on isolation and decolonisation upon admission to evaluate the HA-MRSA-IV clone reservoir. The rapid MRSA diagnostic methods used in this investigation should be established to monitor and implement the successful treatment of these MRSA nosocomial infections.

Isolates and Identification
From January 2012 to December 2013, non-duplicate S. aureus isolates were collected from MICU UHCC in the east of Algeria. All S. aureus isolates were cultured from clinical samples, including from the bloodstream, tracheal, drain, urine, pleural and urinary catheters, in infection contexts with or without the presence of polynuclear cells in the biological samples. HA-MRSA isolates are defined as isolates recovered from hospitalised patients 48 h after admission presenting resistance to methicillin. Demographic characteristics and clinical data were collected from the patient records. All isolates were grown on Chapman and Colombia blood agar media for 24 h at 37 • C. Isolates were pre-emptively identified using conventional methods. The purity of the S. aureus isolates transported was checked using a Trypticase Soja Agar medium at the IRD, APHM, MEPHI, IHU-Méditerranée Infection (Marseille, France), where their molecular identification was confirmed using matrix-assisted laser desorption-ionization time of flight mass spectrometry Microflex (MALDI-TOF-MS) (Bruker Daltonics, Bremen, Germany) as described [63]. The pure identified S. aureus strains were conserved in a storage medium for further molecular characterisation.

Antimicrobial Susceptibility Testing
Susceptibility testing against 21 antibiotics was performed using the Kirby-Bauer disk diffusion method on Mueller-Hinton (MH) agar. The results were interpreted according to the Clinical Laboratory Standard Institute (CLSI) guidelines [64], with the exception of fusidic acid, for which we used the cut-off values of the European Committee on Antimicrobial Susceptibility Testing (EUCAST) [65]. MRSA isolates were identified using cefoxitin disks [66], in addition to (i) a growth test on MH agar supplemented with 4% NaCl and 6 µg/mL of oxacillin, incubated in ambient air for 24 h, and (ii) an MRSA-screen latex agglutination test to detect PBP2a proteins [67]. S. aureus ATCC 25,923 was used as reference control strain.

Molecular Characterisation
Genomic DNA was extracted from bacterial cultures using a commercial DNA extraction kit (QIAGEN-Germany). Real time PCR (RT-PCR) was applied to detect the presence of mecA and mecC genes and also to investigate SCCmec types I to V among the isolates. All S. aureus isolates were also characterised by standard PCR to confirm the type of SCCmec cassette. The primers and probes used in this study are presented in (Supplementary Materials Table S4). The presence of PVL and TSST-1 virulence genes was checked by RT-PCR. Spa typing analysis was performed by DNA sequencing of repeat regions of the protein A gene for all S. aureus isolates, as previously described [68]. Multi-locus sequence typing (MLST) analysis of all isolates was carried out by PCR amplification and sequencing of seven house-keeping genes including arc, aroE, glpF, gmk, pta, tpi, and yqil, as reported [69]. Allelic profiles and sequence types (STs) were assigned according to the S. aureus MLST database (Available online: http://mlst.zoo.ox.ac.uk, accessed on 1 February 2017).

Whole Genome Sequencing (WGS)
WGS was applied to three selected HA-MRSA strains (each representative of an MDR profile) to perform comparative genomics and to match their SCCmec cassettes against those reported in the literature. For this, their respective purified genomic DNAs were sequenced on the Illumina MiSeq machine (Illumina, San Diego, CA, USA) using 250-bp paired-end reads and barcodes according to the TruSeq protocol and the manufacturer's recommendations. Read quality was evaluated with the FastQC programme (Available online: http://www.bioinformatics.babraham.ac.uk/projects/fastqc/, accessed on 2 January 2018) and filtered using the Fastq-mcf programme (Ea-utils package: http://code.google.com/p/ea-utils, accessed on 2 January 2018). Genome sequences were then assembled using an A5-miseq [70] pipeline and annotated with the Prokka v1.12 programme [71]. Genes conferring antimicrobial resistance and virulence genes were investigated using the ABRICATE pipeline, which hosts different databases, including antibiotic resistance genes and virulence genes databases. The genetic environment of the SCCmec-III mercury element was investigated using the Easyfig version x2.2.3 programme (Available online: http://mjsull.github.io/Easyfig/, accessed on 15 June 2020).

Statistical Analyses
The statistical analysis was performed using the Chi-square test or Fisher's exact test as appropriate in the IBM SPSS statistics software (V. 24.0; 2016) (C D C, Atlanta, GA, USA). Results were considered to be statistically significant at a p value ≤ 0.05.

Supplementary Materials:
The following are available online at https://www.mdpi.com/article/ 10.3390/pathogens10111503/s1, Supplementary Table S1: Predicted function of genes shown in Figure 2. Supplementary Table S2: Predicted function of resistance genes presented in Table 2 and their % of identities with described genes of resistance genes. Supplementary Table S3: Predicted function of virulence genes presented in Table 3 and their % identities with described genes. Supplementary Table S4: Primers and probes used in this study for the molecular characterisation of the clinical S. aureus isolates from the MICU at the UHCC. Whole Genome Sequences of the three sequenced S. aureus isolates here are deposited at DDBJ/ENA/GenBank under the accession numbers S. aureus SAUR1404 (PKSS01000000), S. aureus SAUR678 (PKST01000000), and S. aureus SAUR390 (PKSU01000000).  Data Availability Statement: The datasets generated during and/or analyzed during the current study can be find in the main text and the Supplementary Materials.

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