Comparison of genotyping methods and toxin gene profiles of Staphylococcus aureus isolates from clinical specimens

Abstract Staphylococcus aureus is a frequent cause of infections worldwide. Methicillin-resistant S. aureus (MRSA) is one of the main causes of Gram-positive infections, and methicillin-susceptible strains (MSSA) primarily colonize and infect community hosts. Multiple virulence factors are involved, with toxins playing a significant role in several diseases. In this study, we assess the prevalence of toxin genes in 89 S. aureus clinical isolates (31 MRSA and 58 MSSA). We evaluated the discriminatory power of the association of internal transcribed spacer-PCR (ITS-PCR) and 3’- end coa gene ( coa-PCR) when compared with other more commonly used and costly techniques. The isolates showed a high level of genetic diversity, and toxins were found in all the isolates. While most toxin classes displayed no statistically significant correlations and were equally distributed in isolates regardless of their resistance status, classic enterotoxins ( sea-see) showed a positive correlation with MSSA isolates. The combination of coa-PCR with ITS-PCR showed a discriminatory index of 0.84, discriminating 22 genotypes that agree with previously determined data by PFGE and MLST. This association between the two PCR-based methods suggests that they can be useful for an initial molecular epidemiological investigation of S. aureus in hospitals, providing significant information while requiring fewer resources.


Introduction
Methicillin-resistant Staphylococcus aureus (MRSA) and methicillin-sensitive S. aureus (MSSA) are major pathogens associated with serious infections in both hospitals and communities worldwide (Tong et al., 2015;Arias et al., 2018;Junie et al., 2018;Lakhundi and Zhang, 2018).In Latin American countries like Brazil, Bolivia, Chile, and others, more than 50% of S. aureus isolates are already categorized as MRSA and can be considered resistant to most β-lactams (Lee et al., 2018).
Exotoxins produced by S. aureus can cause diarrhea that is either related or not to the use of antibiotics, gastroenteritis, or food poisoning (Ortega et al., 2010;Pinchuk et al., 2010).This pathogen is also connected to severe illnesses like Toxic Shock Syndrome (TSS) and Scalded Skin Syndrome (SSS), which are regarded as superantigens and caused by toxin production (Ladhani et al., 1999;Dinges et al., 2000).The group of superantigens includes staphylococcal enterotoxins (SE), toxic shock syndrome toxin (TSST), and exfoliative toxins (ET) (Portillo et al., 2013).
For epidemiological purposes, it is crucial to identify the distribution of clinical isolates, and genotyping has emerged as a key tool in medical investigations to identify strain origin, clonal relatedness, and outbreak epidemiology (Olive and Bean, 1999;Arias et al., 2018).Genotyping methods often involve applying different molecular techniques based on PCR, sequencing, or genomic macrorestriction (Rodríguez-Noriega et al., 2010).
By combining these methods, strains can be classified into different lineages, or clones.Some clones of S. aureus are known as epidemic clones, meaning they are descendants of the same ancestor and are widely distributed among different countries.Some of these lineages, such as the Brazilian Epidemic clone (BEC), the Pediatric clone (PC), and the Cordobes/Chilean clone, have already been identified as native to Latin America and are useful for describing the various genetic backgrounds of S. aureus (Zurita et al., 2016).
Since few data have been published in Brazil, particularly in the Northeast region, our goal was to investigate the exotoxin gene profile for MRSA and MSSA isolates from patients admitted to hospitals in Recife city, Pernambuco state, Brazil.Additionally, we aimed to evaluate the ability of coagulase gene typing (coa-PCR) and ribosomal 16S-23S internal transcribed spacer (ITS-PCR) to distinguish isolates and clones from clinical onsets in comparison to techniques more frequently used for molecular epidemiology studies of S. aureus, such as and 20 µM of each oligonucleotide.Amplification consisted of denaturation 95 ºC for 2 min, then 30 cycles of 95 ºC for 1 min, 60 ºC for 1 min and 72 ºC for 2 min.Products were separated by electrophoresis through 1.5% agarose gel.
Toxin profiling was done by converting the toxin gene data into a binary matrix.For each isolate, a concatenation of these data yielded a binary profile resembling a barcode sequence.Table S2 has a detailed listing of these data.

coa-PCR
The coagulase gene typing was performed as previously described (Aarestrup et al., 1995).PCR assays were prepared to a final volume of 50 µL, containing 40 ng of chromosomal DNA, 0.8 mM of deoxynucleotide triphosphates, 1X PCR buffer, 2.5 mM MgCl 2 , 1 U Taq DNA polymerase (Promega, Madison, WI, USA), and 20 µM of each oligonucleotide.Amplification was made up of an initial denaturation of 95 ºC for 2 min, then 30 cycles of 95 ºC for 1 min, annealing at 55ºC for 1 min and extension at 72 ºC for 1 min.Products were submitted to electrophoresis through 1.5% agarose gel.S. aureus strain ATCC 25923 was used as a positive control.

ITS-PCR
Amplification of the 16S-23S intergenic spacer region (ITS-PCR) employed a single pair of primers, as previously described (Jensen et al., 1993).PCR assays were prepared to a final volume of 50 µL, containing 40 ng chromosomal DNA, 1 mM of deoxynucleotide triphosphates, 1X PCR buffer, 3 mM MgCl 2 , 1.5 U Taq DNA polymerase (Promega, Madison, WI, USA) and 20 µM of each oligonucleotide (Jensen et al., 1993).Amplification was made up of 95 ºC for 5 min, then 30 cycles of 95 ºC for 1 min, 55 ºC for 1 min and 72 ºC for 2 min.Electrophoresis was performed using a 2% agarose gel.For quality assurance, S. aureus strains ATCC 33591 and ATCC 25923 were employed.

Sequencing
A random sample of purified PCR products from toxigenic genes, ITS-PCR and coa-PCR (Purelink PCR purification kit, Invitrogen, Carlsbad, CA, USA) were chosen for DNA sequencing using the Big Dye Terminator Kit v3.1 and an ABI 3730xl DNA analyzer (Applied Biosystems, Foster City, CA, USA) in order to confirm only the specificity of the amplified genes.The PureLink PCR purification kit was provided by Invitrogen, Carlsbad, California, USA (Applied Biosystems, Foster City, CA, USA).The nucleotide sequences obtained were compared with the S. aureus sequence database in the GenBank through BLAST (http://www.ncbi.nlm.nil.gov).

Discriminatory analysis
In order to compare the discriminatory index of typing methods, we used the formula described by Hunter and Gaston (1988): DI = 1 -[1 / N(N -1)] Σs nj(nj -1), where N is the total number of isolates in this population, s is the total number of different types, and nj is the number of isolates representing each type.This formula is based on the probability that two unrelated strains taken from the population sample will be placed in different types of groups.

Statistical analysis
We investigated any connections between the presence of toxins and the genotypes identified by Coagulase/ITS-PCR, as well as between the presence of toxins and the resistance status (either MRSA or MSSA).We used the Jamovi software's Pearson's correlation test, and only results with a p value of 0.05 or lower were considered statistically significant (Jamovi, 2023).
Statistically, however, the only relevant correlation happened between the MSSA group and classical enterotoxins.19 MSSA isolates carried a sum of 21 genes, while MRSA isolates carried only one (sea, in isolate Sa86 -toxin profile 33) (Table 2).
All strains, regardless of susceptibility, carried at least two egc-related genes.A complete set of the egc, consisting of seg, sei, sem, sen, and seo, was found in 63/89 (71%), of which 22 are MRSA and 41 MSSA.
Isolates were subdivided into 45 genetic profiles based on their toxigenic content, and these genotypes were called toxin profiles (Table S2).The 31 MRSA isolates exhibited 16 profiles, while the 58 MSSA isolates were associated with 39, with both groups sharing 10 of these toxin profiles.The toxin profile 22 -representing a complete egc in addition to genes sej, sek, and sel -was the most prevalent genotype, occurring in 10 isolates (10/89 -16%; 5 MRSA and 5 MSSA).For both groups, toxin profile 22 was also the most frequent, followed by 17, which had four representatives in each group and corresponds to a complete egc in addition to gene sek.
The association between coa-PCR and ITS-PCR (C/R) analysis revealed 22 genotypes (Table 3).The discriminatory index obtained by this combination was 0.84, similar to the index obtained by MLST (0.86) and spa-typing (0.89) but lower than the index obtained by PFGE (0.99).The 22 genotypes and the toxins under investigation did not show any statistically significant correlations.
Genotype C/R-16 consists of only one isolate, a MSSA with ST5 and in PFGE cluster F, and therefore, also related to PC/USA800 (Table 3).Genotypes C/R-9 and C/R-16 differ only in their ITS-PCR profiling.While R1 (ITS-PCR profiling for genotype C/R-9) has 5 bands, R9 (ITS-PCR profiling for genotype C/R-16) has 4, and 3 of these bands are exactly at the same height.We can therefore suggest that both of these genotypes are related.

Discussion
Staphylococcus aureus is responsible for a broad spectrum of diseases in humans due to its ability to express several virulence factors, including enterotoxins, toxic shock syndrome toxin, and exfoliative toxins.Among them, SEs are the major cause of staphylococcal food poisoning (Argudin et al., 2010).Ferry et al. (2005) described that S. aureus strains that cause sepsis, with or without shock, harbor at least one superantigen-encoding gene.Several diseases, including infectious endocarditis and food poisoning, have already been linked to egc (Johler et al., 2015;Stach et al., 2016).
Different rates of toxigenic genes have been found in S. aureus clinical isolates from multiple countries, according to a number of investigations (Song et al., 2016;De Carvalho et al., 2019).In our study, an elevated frequency of S. aureus clinical isolates comprising toxigenic genes was observed, especially those with a complete egc.All isolates related to the USA800/PC clone contained a complete egc, as well as other toxins, except for two isolates (Sa1 and Sa76) which exhibited only part of this cluster.According to Monecke et al. (2011), CC5 isolates carry the enterotoxin gene cluster, although partial deletions have been observed.
Only a few studies have examined the frequency of toxigenic genes in S. aureus in Brazil, particularly in the North and Northeast regions and especially in clinical specimens (Vasconcelos et al., 2011).In one study, 14% of MRSA strains from a university general hospital in Recife during 2002-2003 were related to the USA800 clone and harbored egc.Additionally, approximately 70% of MRSA strains were related to BEC, and none of them had toxigenic genes (De Miranda et al., 2007).
Only the classic staphylococcal enterotoxins (sea-see) were statistically associated with the resistance status of the isolates, and MSSA had the highest frequency of these superantigens.According to earlier research, MRSA makes up no more than 10% of colonizing strains, which suggests that MSSA is very common in colonizing and/or communityacquired infections (Mehraj et al., 2016).The predominance of these genes in MSSA isolates raises concerns for communityassociated infections because the classical enterotoxins are strongly associated with food poisoning (seb), lethal sepsis (sec), and infective endocarditis (sec) (Salgado-Pabón et al., 2013;Ahmad-Mansour et al., 2021).
The eta gene was found in only one isolate, and the low rates of isolates carrying the eta and etb genes responsible for SSS are in accordance with other studies, which have also demonstrated the low frequency of these genes in S. aureus isolates (Becker et al., 1998(Becker et al., , 2003)).
The isolates from our study were from different clinical specimens.It is interesting to note that one isolate (MSSA Sa08, tst positive) was obtained from vaginal secretion.TSS was initially associated with the use of superabsorbent tampons in women with S. aureus tst producers on vaginal secretion; however, later, cases of non-menstrual TSS in the community and hospitals became prevalent (Fitzgerald et al., 2001;Durand et al., 2006).
In the current study, we found significant genetic diversity among S. aureus isolates as well as a high frequency of toxigenic genes.Molecular typing techniques can be used to understand this diversity and how these strains are related.In some circumstances, where speed is required to identify a local outbreak and design containment plans, PCR-based genotyping approaches, such as coagulase gene typing and ITS-PCR, are fast and offer significant discriminatory power.According to Hunter and Gaston (1988), a discriminatory index greater than 0.90 can be interpreted as reliable and is thus desirable.However, even though PFGE has shown greater discriminatory power (DI 0.99), the combination of PCRbased typing methods proved to be a useful and inexpensive procedure for conducting epidemiological surveys of S. aureus on a local or regional scale, even with a DI of 0.84.
The analysis of PCR-coa identified five different amplicons.The 3´end coding region of the coa gene contains a series of repeating 81bp DNA sequences that differ in the number of tandem repeats.Since this region exhibits polymorphism, it is useful as a typing method (Aarestrup et al., 1995).
We chose ITS-PCR as one of our genotyping methods in this study because of its practicality, cost-effectiveness, and alignment with the specific research objectives (Liu et al., 2008).This technique has several advantages, such as simplicity, speed, and the ability to discern between closely related strains based on variations within the ITS region (Boom et al., 1990;Gray et al., 2014).However, it is important to recognize that there are other genotyping techniques, such as ribotyping, that provide extensive information about the ribosomal RNA sequence and rely on the sequencing or hybridization of ribosomal RNA segments (Bouchet et al., 2008).
On the other hand, ITS-PCR focuses specifically on the amplification and analysis of a specific genomic region, the ITS region of the ribosomal RNA operon.This targeted approach allows efficient screening of large collections of strains and can provide valuable information about the genetic diversity and lineage of strains based on variations within the ITS region (Ryberg et al., 2011;Lian and Zhao, 2016).By employing ITS-PCR in conjunction with other molecular typing techniques such as coagulase typing, PFGE, spa typing and SCCmec genotyping, we were able to obtain comprehensive information on the molecular epidemiology and clonal relationships between the S. aureus strains under investigation.
In this study, C/R-9 was the most frequent genotype observed among isolates.All 26 isolates in C/R-9 were associated with Clonal Complex 5, grouping all USA800/ PC and USA100 clones.Thus, the results indicate that the C/R combination was able to distinguish USA800/PC and USA100 clones from other strains.Only one isolate related to the USA800 clone (Sa28 ST5/ newly described t10548, cluster F) exhibited a different genotype, C/R-16, but this genotype appears to be related to C/R-9.
The BEC represents a multidrug-resistant lineage described in Brazil in 1992, linked to hospital-acquired infections (HA-MRSA), which was widely distributed in Brazil and later in other countries.However, in the first ten years of the 2000s, there was an increase in "imported" clones, such as the Pediatric clone (PC), the New York/Japan Clone, and other less common lineages (Andrade et al., 2020).
Since then, BEC prevalence appears to be decreasing, while infections by community-acquired (CA-MRSA) clones, such as PC and OSPC, have been steadily increasing.Complete BEC substitution has already been reported in some hospitals (Chamon et al., 2017;Monteiro et al., 2019), and CA-MRSA clones appear to be becoming more common across Brazil (Romero and Cunha, 2021).Historically, there have been distinctions between the epidemiologies of HA-MRSA and CA-MRSA, with CA-MRSA typically being linked to more virulent infections and infecting patients who have few or no risk factors.However, these distinctions are becoming less clear today (Lee et al., 2018).
All isolates within ST1 (Cluster E, CC1) were grouped into the genotype C/R-6 association.Coagulotype and ITS-PCR analyses were capable of distinguishing these isolates from others also clustered into PFGE cluster E and related to USA400 that exhibited ST669 and ST97, both STs from CC97.The genotypes C/R-11 and C/R-12 were able to group all isolates related to clone USA1100, except for isolate Sa65 (C/R-3).
We observed a more robust correlation between coa-PCR/ITS-PCR and PFGE/MLST patterns in MRSA isolates.The C/R association allowed us to observe the clonal spread of MRSA and MSSA within the main hospital analyzed (hospital 1).Patients from these isolates were dispersed throughout various hospital wings.Additionally, we discovered closely related isolates between hospital 1's isolates and all four of hospital 2's isolates.
No one genotyping technique that is now available is thought to be best for epidemiological studies.Every circumstance is unique, so it is important to assess the benefits and drawbacks of each technique both separately and collectively in order to choose the best methodology based on the targets and objectives outlined in each study.
Through combining coagulotype and ITS-PCR analysis, which showed a relationship with PFGE genotype and MLST as well as a minor correlation with spa typing, we found a high genetic diversity among the isolates in our study and observed clonal spread of MRSA and MSSA in hospital settings.It is important to emphasize that this specific association between techniques may be practical, quick, and affordable for initial epidemiological investigations in hospitals or local outbreaks, thus becoming an interesting strategy for countries and institutions with fewer resources.
We emphasize the need for further studies for epidemiological surveillance of MRSA and MSSA due to the change in S. aureus epidemiology and the growing threat of this pathogen to hospital and community environments.

Table 3 .egc=
Molecular and toxigenic profiles of 89 Staphylococcus aureus isolates from hospitals in Recife, Brazil.Enterotoxin gene cluster (seg, sei, sem, sen and seo); g ST2381, g ST2382 and g ST2383= new sequence types described; h t10548 and h t10550= new spa types describe;

Table 1 -
Oligonucleotides used for the detection of exotoxin genes.

Table 2 -
Pearson's correlation between toxin gene counts and isolates' resistance status (MRSA or MSSA).