Molecular characterization and genetic diversity of Staphylococcus aureus isolates of dairy production farms in Rio de Janeiro, Brazil

Bianca da Silva Soares1, Cássia Couto da Motta2, Nicolle Lima Barbieri3, Dayanne Araújo de Melo4, Marisol Alvim Gomez5, Tatiani Abreu de Alencar4, Irene da Silva Coelho6, Shana de Mattos de Oliveira Coelho7, Catherine Mary Logue8, Miliane Moreira Soares de Souza9 1 Biologist, DSc. Programa de Pós-Graduação em Ciências Veterinárias (PPGCV), Departamento de Microbiologia e Imunologia (DMI), Instituto de Veterinária (IV), Universidade Federal Rural do Rio de Janeiro (UFRRJ), Campus Seropédica, RJ, Brazil 2 Veterinarian, DSc., PPGCV, DMI, IV, UFRRJ, Campus Seropédica, RJ, Brazil 3 Pharmacist, Dsc., Department of Population Health, College of Veterinary Medicine, University of Georgia, Athens, GA, United States of America 4 Veterinarian, DSc., Programa De Pós-Graduação Em Ciência, Tecnologia E Inovação Em Agropecuária (PPGCTIA), Universidade Federal Rural do Rio de Janeiro (UFRRJ), Campus Seropédica, RJ, Brazil 5 Veterinarian, MSc., PPGCV, DMI, IV, UFRRJ, Campus Seropédica, RJ, Brazil 6 Agronomist, DSc., DMI, IV, UFRRJ, Campus Seropédica, RJ, Brazil 7 Biologist, DSc., DMI, IV, UFRRJ, Campus Seropédica, RJ, Brazil 8 Food Science and Technology, DSc., Department of Population Health, College of Veterinary Medicine, University of Georgia, Athens, GA, United States of America 9 Veterinarian, DSc., DMI, IV, UFRRJ, Campus Seropédica, RJ, Brazil How to cite: Soares, B. S., Motta, C. C., Barbieri, N. L., Melo, D. A., Gomez, M. A., Alencar, T. A., Coelho, I. S., Coelho, S. M. O., Logue, C. M., & Souza, M. M. S. (2021). Molecular characterization and genetic diversity of Staphylococcus aureus isolates of dairy production farms in Rio de Janeiro, Brazil. Brazilian Journal of Veterinary Medicine, 43, e001120. https://doi.org/10.29374/2527-2179. bjvm001120


Introduction
In Dairy cattle, various species of microorganisms are implicated in infectious mastitis, especially Staphylococcus aureus, which causes contagious mastitis that could be clinical or chronic (usually subclinical) whose infection and dissemination usually occur during milking (Bardiau et al., 2014).
Mastitis caused by S. aureus is a result of the production of several virulence factors (VFs), which may contribute in different ways to the pathogenesis of the organism. Pathogenic differences of S. aureus strains may be a result of the geographical distribution, host, tissue type, and number of combination of virulence genes may influence the pathogenic potential of S. aureus strains (Bar-Gal et al., 2015).
The ability of S. aureus to produce biofilm is considered as a critical virulence factor capable of influencing the pathogenesis of mastitis. Biofilm aids in the adhesion and colonization of the organism in the mammary gland epithelium. The association between biofilm in infections and drug resistance has led to a growing interest in the characterization of the genes involved in biofilm formation. A single locus has been detected in most isolates of S. aureus of mastitic origin, indicating its potential role as a virulence factor in the pathogenesis of mastitis in ruminants (Melchior et al., 2006).
Another VF that is associated not only with adhesion but also internalization by cells is fibronectin-binding proteins (FBN) A and B that are multifunctional microbial surface components recognizing adhesive matrix molecules (MSCRAMMs) that recognize fibronectin, elastin, and fibrinogen. FBN promotes the internalization of S. aureus in epithelial and endothelial cells that are generally not phagocytic, favors the spread of bacteria from the bloodstream to internal organs, and the evasion of the immune response and the action of antibiotics (Burke et al., 2010).
VFs involved in toxins production in S. aureus play a significant role in intramammary infections pathogenicity. αand β-hemolysins are pore-forming exotoxins that can induce inflammatory changes in mammalian cells, inactivating the immune system, their direct cytotoxic effect degrades tissues, providing nutrients for bacteria and facilitating their dispersion to new sites (Haveri et al., 2007).
A high number of S. aureus genotypes from cattle herds worldwide have been studied to develop better strategies for the treatment of mastitis (Kot et al., 2016). Regulatory circuitry that controls virulence and adaptation of S. aureus to the environment is complex. This regulatory system can receive signals from the external environment to modulate biofilm formation and produce various exoproducts in a manner appropriate at the infection site (Ster et al., 2005). One of the regulatory systems, agr, encodes a constituent component of quorum sensing that is activated by bacterial density through auto-induced peptide secretion. Activation of agr follows the production of the molecule RNA III, which is an effector molecule of the agr system. RNA III follows the reduction of surface proteins, such as adhesins required for colonization. At the same time, RNA III allows the production of secreted proteins, such as nucleases and proteases, both of which participate in the release of biofilm bacteria (Novick & Geisinger, 2008;Otto et al., 2013).
Understanding the clonal relationship among strains is essential to determine the source and routes of infection, identify an outbreak, traceroute of transmission, recognize particularly virulent strains and evaluate the effectiveness of control measures (Pérez-Losada et al., 2013).
Methods commonly used to type and subtype S. aureus strain include Pulsed Field Gel Electrophoresis (PFGE) (Middleton et al., 2002), spa gene typing (Shopsin et al., 1999) and Multilocus Sequence Typing (MLST) (Smith et al., 2005). Although PFGE is considered a gold standard as a strain typing method for this genus and species, this method has greater discriminatory power in relation to other techniques (Adkins et al., 2016). The present study aimed to evaluate the genetic diversity of a collection of S. aureus isolated from bovine mastitis through the screening for VFs, agr and spa typing, PFGE and MLST.

Material and methods
Sampling A total of 120 milk samples were collected from cows presenting subclinical mastitis identified by California Mastitis Tests (CMT) from October to November 2012, from seven dairy farms located in the cities of Barra do Piraí, Vassouras, Passa Três, Paraíba do Sul, Rio das Flores and Carmo in the state of Rio de Janeiro, Brazil. Fifty-three strains of S. aureus were recovered and identified by phenotypic tests (Koneman et al., 2012) and confirmed by amplification of the coa gene (Hookey et al., 1998) and nuc (Ciftci et al., 2009) as showed in Table 1. Staphylococcus aureus standard strain ATCC 29213 was used as control. A total of 17 isolates of S. aureus were randomly selected for the present study, which were from four dairy farms initially selected, considering the virulence profile observed.

DNA extraction
Bacterial total DNA extraction was performed according to the protocol established by Tito et al. (2015).

Virulence genes
The analysis of VFs comprised the detection of the genes icaA and icaD (Vasudevan et al., 2003), implicated in biofilm production, fnbA and fnbB, that encodes fibronectin-binding proteins (El-Sayed et al., 2006) and the hemolysin genes hlA and hlB (Nilsson et al., 1999), that encodes αand β-hemolysins were performed by PCR (Polimerase Reaction Chain) as shown at Table 1. ATCC 29213 S. aureus was used as quality control.  (Shopsin et al., 2003).

4/12
Molecular characterization and genetic diversity of Staphylococcus aureus isolates of dairy production farms in Rio de Janeiro, Brazil

Molecular typing
agr system Classification of agr system groups was based on the hyper variable domain of agr locus according to Shopsin et al. (2003), as showed at Table 2. A sensu primer, pan-agr, corresponding to conserved sequences of the agrB gene, was used in all reactions, combined with four anti-sensu primers, each one specific for the amplification of a single agr group based on the agr locus polymorphism. spa typing spa typing was performed according to Shopsin et al. (1999), as showed at table 2, and ATCC 29213 S. aureus was used as quality control. PCR products were purified using Exo-SAP-IT® (USB Corporation, Cleveland, Ohio) as recommended by the manufacturer and then sequenced. The sequences generated were edited using the Bioedit program (Hall, 1999) and Mega version 7.0 (Kumar et al., 2016), and later were analyzed using the DNAGear program (Faroq et al., 2012) for spa type designation.
Pulsed Field Gel Electrophoresis (PFGE) PFGE typing was performed according to protocol established by the CDC (Centers for Disease Control and Prevention, 2013). Salmonella Braenderup H9812 was used as standard control. Gel images were imported into BioNumerics (Applied Maths®) for analysis. Macrorestriction patterns were compared using the BioNumerics Fingerprinting software (Version 6.5, Applied Math, Austin, TX). The similarity index of the isolates was calculated using the Dice correlation coefficient option of the software with a position tolerance of 1% and an optimization of 0.5%. The unweighted-pair group method using average linkages (UPGMA) was used to construct a dendrogram.
Multilocus sequence Typing-MLST MLST typing was performed according to protocol described by Enright et al. (2000), considering the amplification of housekeeping genes (arcC, aroE, glpF, gmk, pta, tpi e yqiL), as showed at Table 2. ATCC 29213 S. aureus was used as quality control. PCR products were purified using ExoSAP-IT (USB Corporation, Cleveland, Ohio) as recommended by the manufacturer and then sequenced. Sequences were edited using the program BioEdit (Hall, 1999) and Mega version 7.0 (Aanensen & Spratt, 2005) and the allele and sequence types (STs) were determined using the MLST website (Trust Pharmacy, 2020).
Considering the presence of fibronectin-binding proteins genes (fbnA and fbnB) the majority of the isolates presenting one or both of them (profiles 1, 2, 3, 4, 5, 6 and 7) as showed at Table 3. In this study, 41% (7/17) of the strains were positive for fbnA gene and 47% (8/17) were positive for fbnB gene.
The detection of the genes responsible for the production of hemolysins α and β, hlA and hlB respectively, demonstrated that only one profile (profile 5) did not present this capacity, as showed at Table 3. 94% (16/17) of the strains were positive for the hlA gene and 70.5% (12/17) were positive for the hlB gene.

Molecular typing of S. aureus
The agr typing classified the S. aureus isolates predominantly as agr type-II, 58.8% (10/17). It should be noted that 41.2% (7/17) of the strains were classified as non-typeable.
Through PFGE technique it was possible to observe high genetic heterogeneity among the strains examined (Figure 1). All patterns were relatively unique to the strains examined. Strains 208 and 225 appear to be closely related as they differ in only one fragment. In addition, both strains share similar virulence profiles (Table 3) and were recovered from the same farm (farm C).

7/12
Molecular characterization and genetic diversity of Staphylococcus aureus isolates of dairy production farms in Rio de Janeiro, Brazil They also share ST 126 and spa type -spa 605 and were both classified as agr type-II but were isolated from different animals. Analyses of MLST typing demonstrated five different ST/CC types (ST/CC126, ST5, ST/CC97, ST3087 and ST747) ( Table 3). Most of the strains, 64.7% (11/17), were identified as belonging to ST/CC 126 and this was distributed on three different farm properties.

Analyses of VFs detection of S. aureus
Most S. aureus strains have the genetic capacity to produce biofilm. Our results corroborate with Marques et al. (2017) that showed a high abundance of genes icaA and icaD in S. aureus isolates, producing biofilm in samples from bovine intramammary infections in Brazil. Acosta et al. (2018) reported a high frequency of fnbA (94%), fnbB (81.3%), hlA (88%) and hlB (88.3%) genes in S. aureus from milk samples, and these genes were more abundant in samples from bovine milk than goat milk in a recent study developed in cow and goat herds raised in three different geographical regions of the state of Pernambuco, Brazil, and these results corroborate with our data.
The importance of monitoring biofilm production (icaA and icaD), fibronectin-binding proteins (FBN) A and B (fnbA and fnbB) and αand β-hemolysins (hlA and hlB) genes in S. aureus isolated from dairy environment is reinforced due to its high frequency amongst the isolates investigated (Table 3). It supports the idea that these VFs are strictly related to severe mastitis cases.

Analyses of molecular typing of S. aureus
Agr typing of S. aureus isolates showed a predominance of agr type-II and non-typeable. Similar results were reported by Fabres-Klein et al. (2015) who also detected the prevalence of agr type-II suggesting that this type is better adapted to the dairy environment, and by Marques et al. (2017) that also noted that not all S. aureus strains examined could be classified, when analyzing S. aureus from bovine mastitis in Brazil.
spa type 605 has been reported in several studies of bovine mastitis (Guinane et al., 2010;Sakwinska et al. 2011). This spa type has also been reported as an ordinary strain associated with bovine mastitis in the state of Rio de Janeiro and also appears to be linked with ST126 (Aires- de-Sousa et al., 2007;Rabello et al., 2007).
Considering PFGE analyses, the similarity between the strains 208 and 225 could be occurred by a small-scale mutation or by the presence of mobile genetic elements. The fact that they shared several characteristics and only differed at the animal origin suggests that there may be similar strains circulating and causing mastitis in this particular herd. Previous studies conducted by our research group on distinct farms of the present study showed similar results, high genetic heterogeneity, and the absence of clonally related strains (Marques et al., 2013). Rabello et al. (2007) also reported similar results when using the same technique to evaluate S. aureus strains from Rio de Janeiro state. A possible explanation for the high genetic heterogeneity of S. aureus strains in properties in the state of Rio de Janeiro can be explained by the fact that the production is carried out on small properties with distinct characteristics for milking and breed. This high heterogeneity would also imply the difficulty of obtaining effective prophylactic measures to control mastitis because of the diversity of strains implicated.
ST 126 has been reported in several studies associated with bovine mastitis outbreaks (Guinane et al., 2010;Sakwinska et al., 2011) and has also been reported as implicated in bovine mastitis in the state of Rio de Janeiro (Aires- de-Sousa et al., 2007;Rabello et al., 2007). A study by Smyth et al. (2009) showed that ST126 is usually associated with agr type-II and spa type t605. A similar result was observed in our study that presenting a prevalence of spa type t605. However, some strains were also identified as t359 and t23, suggesting that spa type and ST are not exclusively linked. This can be explained by using a single locus method such as spa typing for macroepidemiologic investigations, which can distort the underlying clonal relationships .
ST 97 has also been detected in previous studies of bovine mastitis in the state of Rio de Janeiro, this ST appears to be widely distributed in several countries around the world, in addition to being related to LA-MRSA strains and has also been implicated as a causative agent of bovine mastitis (Meemken et al., 2010;Smith et al., 2005).

Comparative analysis of the genetic profiles generated by different molecular typing techniques
The clustering of strains, according to the virulence genes distribution, yielded 11 distinct profiles. Most strains shared the same virulence profile but not the same type through other typing and subtyping techniques, as already reported by Melles et al. (2004) in a study evaluating the population dynamics and expansion of pathogenic clones of S. aureus in healthy humans. Their study suggested that not all clones of S. aureus share the same VFs because the strains may be under different environmental pressure that results in strains from some clonal lineages being more virulent than others.
Regarding spa-type analysis of the strains examined in this study, seven different spa types were detected among the 17 strains examined. Genotypic characterization based on the investigation of a locus of the highly variable tandem repeat region of the spa gene is another successful tool for the genotypic characterization of S. aureus. spa gene typing also considers the recombination of X region events. Since recombination occurs more frequently within this region, this typing method typically provides more types (spa) than MLST. Although spa typing and MLST are usually concordant and allow a similar classification for distinct genotypes, there are also several cases of sorting errors by the spa typing technique, which are most likely caused by recombination events. Thus, spa typing is a useful method to obtain information about the genotype of the strains but should not replace the MLST technique .
Regarding the detection of STs by MLST technique, five different sequence types (ST) or clonal complexes (CC) were detected. According to MLST technique, a clone is defined based on the sequence type, so an isolate is defined as a clone if it shares the same ST. This definition of clones using the MLST technique is very useful for investigating the evolutionary history of bacteria. However, it may lead to misconception since the term clone is defined as bacterial isolates with indistinguishable genotypes. However, in the case of bacteria, it is not possible to define clones based only on this definition, since high rates of mutation and recombination, as well as the acquisition or loss of mobile genetic elements, can lead (depending on the species) to a bacterial genome that is highly variable. Permanent diversification of the ancestral genome results in an increasingly diverse set of genotypes. The MLST method, however, provides an efficient method to identify genetic lineages that share the same ancestor, due to MLST method being based on the amplification of multiple internal fragments in the DNA sequence of relatively conserved housekeeping genes (with approximately 450-500 bp for each gene) that are present in all strains of the same species. Genetic relationships between the strains is determined by the analysis of these housekeeping genes sequences, which are compared by analysis of nucleotide substitution in the sequence. Based on this definition, clones (MLST) include isolates of the same sequence type but are not necessarily genetically identical (Ranjbar et al., 2014;Spratt, 2004).
PFGE analysis generated quite distinct profiles demonstrating the heterogeneity of the strains studied. Compared with MLST results, this technique allows detection of more frequent genomic alterations. Thus, using PFGE it is possible to demonstrate the diversity of S. aureus strains sharing the same sequence type and thus show that clones based on MLST definition are not necessarily genetically identical (Murchan et al., 2003).
PFGE and MLST techniques are considered as some of the best methods for molecular typing of S. aureus but did not have a reasonable correlation when applied individually. However, when combined, it is possible to obtain an accurate overview of the population of S. aureus present in dairy herds. In view of the results generated from this study, it can be concluded that the choice of the typing method and its application depends on the type of epidemiological study that is to be carried out. However, it should be emphasized that in order to carry out the study, the application of multiple typing techniques that can be combined provides higher power to the analysis and the research question.
Although there is much still that remains to be investigated and understand regarding recombination events in S. aureus it seems clear that it occurs frequently enough to be detectable but not often enough to destroy vertical or clonal signals. Understanding how the mix of horizontal and vertical processes of clonal evolution contributes to the structuring and evolution of this bacterial species is a goal to be achieved, and techniques such as Whole Genome Sequencing (WGS) can clarify these aspects.
However, for these studies to be reliably validated, epidemiological and clinical biases need to be eliminated. Currently, most of the phylogenomic studies of S. aureus consist of sample collection and analysis related to clonal complexes of medical interest, which hinders an evolutionary view of these transformations (Planet et al., 2016). Regarding isolates from animal production, especially those of bovine origin, these studies are still emerging and point to a highly diversified phylogenetic structure. Possibly, due to the intense selection pressure exerted by the use of antimicrobials and sanitizing agents in the production environment. Also, the great diversity of bacterial species in a constant competition that induces elaboration of strategies for survival and adaptation to the host, fomites, and the environment as a whole.

Conclusions
A high genetic variety of S. aureus strains associated of the presence of different virulence factors justified the absence of clonal strains at the properties evaluated. Besides that, the definition of S. aureus clonal strains is not fully understood since there are a limited number of studies in S. aureus associated with bovine mastitis. Therefore, for a better understanding of phylogenetic diversity of this agent more studies are necessary to clarify the role of population dynamics and clones of S. aureus at the pathogenicity of mastitis in bovines, improving the monitoring and the control of this disease.