Biofilm formation and prevalence of adhesion genes among Staphylococcus aureus isolates from different food sources

Abstract To assess biofilm formation ability and identify differences in the prevalence of genes involved in biofilm formation among Staphylococcus aureus strains isolated from different food samples, the ability of biofilm formation among 97 S. aureus strains was evaluated using a colorimetric microtiter plate assay. Thirteen genes encoding microbial surface components recognizing adhesive matrix molecules, and the intracellular adhesion genes were detected by PCR using specific primers. Approximately 72% of the isolates produced biofilms. Among these isolates, 54.64% were weak biofilm producers, while 14.43% and 3.09% produced moderate and strong biofilms, respectively. The icaADBC, clfA/B, cidA, and fib genes were detected in all the S. aureus strains, whereas the bap gene was not present in any of the strains. The occurrence of other adhesin genes varied greatly between biofilm‐producing and nonbiofilm‐producing strains. However, a significant difference was observed between these two groups with respect to the fnbpB, cna, ebps, and sdrC genes. No obvious evidence was found to support the link between PFGE strain typing and the capacity for biofilm formation. Considerable variation in biofilm formation ability was observed among S. aureus strains isolated from food samples. The prevalence of adhesin‐encoding genes also varied greatly within strains. This study highlights the importance of biofilm formation and the adhesins of S. aureus strains in food samples.

indicates that by adopting this lifestyle, bacteria in biofilms gain some advantages over planktonic cells. For example, biofilms can protect this microbe from the action of antibiotic drugs, proteases released by host defense cells and environmental stress factors (Singh, Ray, Das, & Sharma, 2010). This protection may contribute to the persistence of S. aureus in food processing environments, consequently increasing cross-contamination risks and subsequent economic loss due to recalls of contaminated food products (Vazquez-Sanchez et al., 2013). Therefore, an improved understanding of the development of staphylococcal biofilms at the molecular level is imperative to generate new strategies for biofilm-associated contamination.

| Bacterial strains and growth conditions
Ninety-seven S. aureus isolates from six types of marketed food or associated with food poisoning outbreaks in Hangzhou, Zhejiang Province were investigated (see Table 1). All the isolates were analyzed by cultivation on sheep blood agar and Baird-Parker agar (Merck) and identified as S. aureus by determination of specific properties. S. aureus ATCC 6538, which is used as a reference gram-positive strain in the United States and in European standard bactericidal tests, was used as a strong biofilm-forming strain in this study. Staphylococcus epidermidis ATCC 12228 was used as a negative control based on previous research (Arciola, Baldassarri, & Montanaro, 2001). Bacterial stocks of each strain were maintained at −80°C in tryptic soy broth (TSB) containing 20% glycerol (v/v). All the strains were thawed and subcultured in tryptic soy agar (TSA) for 18-24 hr prior to use.

| Biofilm formation assay
Biofilm formation was measured as previously described by Stepanovic et al. (2007), with minor modifications. Briefly, single colonies from TSA were suspended in 3 ml of TSB and incubated without shaking for 18 hr. The bacterial cultures were adjusted to match the turbidity to that of the 0. air-dried, and the stain was dissolved by adding 100 µl of 33% glacial acetic acid (v/v) per well for 10 min at room temperature.
The absorbance was read at 490 nm using an iMark microplate absorbance reader (Bio-Rad Laboratories). The experiment was performed in triplicate at least, and the absorbance of wells containing sterile TSB was used as a negative control. Considering a low cutoff (ODc) to be represented by 3 × SD above the mean values of the control wells, the strains were classified into the following categories: no biofilm production (OD ≤ ODc); weak biofilm producer (ODc < OD ≤ 2ODc); moderate biofilm producer (2ODc < OD ≤ 4ODc); and strong biofilm producer (4ODc < OD).

| Detection of adhesin genes
Template DNA was obtained from pure cultures of the strains. All strains were grown overnight in TSB. A 1-ml aliquot of each over- using UV light after electrophoresis on a 1.5% agarose gel (w/v). Pattern similarity was calculated using the Dice coefficient with 1% optimization and a band matching tolerance of 1%.

| Statistical analysis
A statistical software package (SPSS 19.0 for Windows; SPSS, Inc.) was used to perform statistical analysis. Differences in gene prevalence between biofilm-producing groups and nonbiofilm-producing groups were calculated using the chi-squared test for each gene. p-values <.05 were considered statistically significant.

| Biofilm formation analysis
A total of 72.16% of the S. aureus isolates tested were found to be adherent to varying degrees. Only three isolates (3.09%) were defined as strong biofilm producers; 14.43% of the clones were moderate producers, and more than half (54.64%) were found to be weak producers. A total of 27.84% exhibited no biofilm production ( Table 3). The S. aureus ATCC 6538 strain was found to be strongly adherent based on the OD490 values, while the S. epidermidis ATCC12228 strain was found to be nonadherent based on the OD490 values.

| PCR assay
All the primers used in the experiment exhibited specificity with a single band. We detected 13 MSCRAMMs and 5 biofilm-related genes involved in S. aureus cell attachment and multiplication. The results showed that these genes varied among the different S. aureus isolates.  Table 4, significant differences were detected between biofilm-positive and biofilmnegative isolates with respect to the fnbpB, cna, epbs, and sdrC genes.

As shown in
In addition, considering the strain population as a whole, the presence of sdrC and sdrD significantly improved biofilm formation (Table 5).
In particular, the strains with the sdrC(+)/sdrD(+) genotype exhibited strong or moderate biofilm formation more easily than the other strains. In addition, the sdrC(−)/sdrD(−)/sdrE(−) strains all exhibited no biofilm formation. Isolates that were concomitantly PCR positive for the sdrC, sdrD, and sdrE genes were all positive for biofilm formation.
With respect to the agr group, agrI was the most common type and was detected in 43 (44.33%) of all the isolates. Thirty-three (34.02%) and 20 (20.62%) isolates were positive for agrII and III, respectively, while only one strain was positive for agrIV (Table 6). Interestingly, all agrIII-positive isolates were able to form biofilms, with isolates 15-80, 15-83, and 16-22 exhibiting the highest biofilm production.

| Determination of genetic relatedness by PFGE
Pulsed-field gel electrophoresis analysis of SmaI-digested genomic DNA was performed to determine the genetic relatedness of S. aureus isolates using the CHEF Mapper system as previously described. Isolates were assigned the same pulsotype if the value of the Dice coefficient of similarity was >80% (Harastani, Araj, & Tokajian, 2014).
Clusters were designated with capital letters from A to X (Figure 1).
The pulsotype designated with the letter H was the most common, accounting for 21.92% (16/73) of the strains tested, followed by group J, which contained 12 isolates. However, the degree of biofilm formation varied greatly between these two groups.

| D ISCUSS I ON
The ability of S. aureus to produce biofilms on surfaces is believed to contribute to food poisoning (Doulgeraki, Di Ciccio, Ianieri, & Nychas, 2017 studies that showed the percentage of fnbpA and clfA in biofilm-producing strains to be significantly higher than that in nonbiofilm-producing strains (Rahimi et al., 2016). to that of the study conducted by Filipello et al. (2019), where the fnbpB gene was not detected in two isolates with strong biofilmforming ability. As in many reports, the bap gene was not detected in any isolate (Khoramian, Jabalameli, Niasari-Naslaji, Taherikalani, & Emaneini, 2015;Tang et al., 2013;Vautor, Abadie, Pont, & Thiery, 2008;Vazquez-Sanchez et al., 2013). The bap gene is present in the pathogenicity island SPIbov2, which has been identified in only a small proportion of S. aureus isolates (Vautor et al., 2008), originating only from bovine subclinical mastitis (Cucarella et al., 2001) , 1995). Our results showed that cna was present significantly more frequently in biofilm-producing strains than in nonbiofilm-producing strains. Another study reported that cna-positive isolates (20%) were identified as moderate or strong biofilm producers (Pereyra et al., 2016). In contrast, Khoramian et al. (2015) found that there was no obvious difference in the prevalence of the cna gene between these two groups, which is consistent with the findings of Tang et al.
The elastin-binding protein of S. aureus (ebps) is an adhesin that is responsible for attachment to host cells via binding to elastin (Park, Rosenbloom, Abrams, Rosenbloom, & Mecham, 1996). However, inactivation of ebps has a minimal effect on the binding of S. aureus to immobilized elastin (Roche et al., 2004). The ebps-deficient strain not only continued to form biofilms but also exhibited significantly en-  isolates from skin demonstrated that the ica operon and sdrC are highly expressed in response to biofilm formation (Shin et al., 2013).
Based on our result that 94.29% of biofilm-positive isolates carried the sdrC gene, sdrC may be an important molecule for bacterial intercellular binding and subsequent biofilm formation.
The agr sensing system has been shown to downregulate genes of cell wall-associated adherence factors, leading to decreased biofilm initiation (Moormeier & Bayles, 2017). In this study, agrI was the dominant agr type among the tested S. aureus strains (44.33%), followed by agrII (34.02%) and agrIII (20.62%), which was consistent with the results of previous studies (Bardiau, Detilleux, Farnir, Mainil, & Ote, 2014;Bar-Gal et al., 2015;Filipello et al., 2019;Khoramrooz et al., 2016;Mitra et al., 2013). Furthermore, it was interesting that all of the agrIII isolates were identified as being biofilm positive, 11 of which were strong/moderate biofilm-producing strains. This result was also similar to those of the studies conducted by Khoramrooz et al. (2016) and Rahimi et al. (2016).
Thus, it is likely that there is a significant association between ag-rIII and biofilm production in S. aureus isolates.
A second group of virulence factors that contribute to biofilm formation is PIA/PNAG, which is synthesized by icaADBC operonencoded enzymes (O'Gara, 2007). In this study, all isolates tested were found to be positive for the icaADBC genes. These findings were similar to the observations of Atshan et al. (2012) and Arciola et al. (2001), as there was no difference in the distribution of the icaA and icaD genes in the biofilm-positive and biofilm-negative strains.
However, the prevalence rates of the icaA and icaD genes vary greatly among different studies. For example, when S. aureus is exposed to different temperatures and contact surfaces for different amounts of time, distinct gene expression profiles can be observed (Atshan et al., 2013;Kroning et al., 2016;Stanley & Lazazzera, 2004

| CON CLUS IONS
In general, considerable variation in biofilm formation ability was observed among S. aureus strains isolated from food samples. The prevalence of adhesin-encoding genes also varied greatly within strains. There was no significant difference in the prevalence rate of MSCRAMM genes among the nonbiofilm-producing isolates and among those producing weak, moderate, and strong biofilms, except for fnbpB, cna, ebps, and sdrC. Our results, in combination with those of previous studies, indicate that detection of sdrC is a practical approach for the prediction of biofilm formation. Further research on a large number of isolates may be needed to verify this possibility and explore the connection between the genetic background of S. aureus and the biofilm formation ability based on microbial subtyping.

CO N FLI C T O F I NTE R E S T S
None declared.

AUTH O R CO NTR I B UTI O N S
QC designed and performed all of the experiments described above and wrote the paper. SMX analyzed the results. XQL helped with the conceiving of the study and the manuscript draft. SC and XDL performed the experiments after first revision. HQW, WZ, and ZBZ purchased materials and participated in the study's coordination.

E TH I C A L S TATEM ENT
None required.

DATA AVA I L A B I L I T Y S TAT E M E N T
All data associated with the article have been included in this manuscript.