Affinity of Staphylococcus aureus for prostheses colonization compared to other bacteria. An in vitro study

Staphylococcus aureus biofilms have been recognized as a leading cause of multiple infections, including implantassociated infections and chronic wounds. We evaluated the colonization capacity of two distinct textured prostheses by different bacterial strains. Staphylococcus aureus, Staphylococcus epidermidis, Escherichia coli, Proteus mirabilis and Enterococcus faecalis were evaluated. Initially, the hydrophobicity and biofilm formation capacity were determined. Subsequently, 20 fragments of vascular prosthesis and 20 silicone prostheses were embedded in Research, Society and Development, v. 10, n. 5, e15310514701, 2021 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v10i5.14701 2 suspensions with the microorganisms and incubated. The prostheses were then sown in culture medium and incubated for 48 hours. Petri dishes were photographed and analyzed by fractal dimension. The Kruskal-Wallis test and the Dunn test were applied for the analysis of biofilm formation. To compare the mean intensity for the type of bacteria and the type of prosthesis, a general linear model was applied. Staphylococcus aureus was the bacterium with the highest colonization density in both prostheses (p = 0.0001). E. coli showed strong adherence in the biofilm formation capacity test (p = 0.0001), however, it did not colonize either prosthesis. We demonstrated that Staphylococcus aureus has a greater affinity for vascular and silicone prostheses than other bacteria.

suspensions with the microorganisms and incubated. The prostheses were then sown in culture medium and incubated for 48 hours. Petri dishes were photographed and analyzed by fractal dimension. The Kruskal-Wallis test and the Dunn test were applied for the analysis of biofilm formation. To compare the mean intensity for the type of bacteria and the type of prosthesis, a general linear model was applied. Staphylococcus aureus was the bacterium with the highest colonization density in both prostheses (p = 0.0001). E. coli showed strong adherence in the biofilm formation capacity test (p = 0.0001), however, it did not colonize either prosthesis. We demonstrated that Staphylococcus aureus has a greater affinity for vascular and silicone prostheses than other bacteria. Keywords: Biofilms; Prostheses and implants; Polytetrafluoroethylene; Vascular surgical procedures; Breast implant.

Introduction
According to the National Institutes of Health (NIH), over 80% of all microbial infections are related to biofilm formation. These types of infections are difficult to diagnose and treat. Biofilms are also able to colonize medical devices such as catheters and implants (Rabin et al., 2014).
In studies carried out with patients with different types of wounds associated or not with surgery, the most common bacterial species detected were Staphylococcus aureus (in most cases), followed by Proteus mirabilis, Escherichia coli, Enterococos faecalis, Pseudomonas aeruginosa, Klebsiella ssp and Corynebacterium spp (Bessa et al., 2015;Mengesha et al., 2014).
The most common microorganisms found on the skin are S. aureus and Staphylococcus epidermidis (Igari et al, 2014). Contamination of vascular or silicone prostheses often occurs in the preoperative period, when the prosthesis contact the Research, Society and Development, v. 10, n. 5, e15310514701, 2021 (CC BY 4.0) | ISSN 2525-3409 | DOI: http://dx.doi.org/10.33448/rsd-v10i5.14701 3 skin during surgery, so these are the bacteria that most contaminate these two types of prostheses (Ghiselli et al., 2001;Alp et al., 2014).
S. aureus is a major cause of nosocomial and community-acquired infections, and infections caused by this bacterium represent a high cost to the healthcare system (Lister & Horswill, 2014). Interest in infections caused by the S. aureus biofilm has been associated with the rapid increase in the use of implants and medical prostheses and the concomitant increase in infections related to these devices (Bhattacharya et al., 2015).
The surface of an implant is rich in proteins such as fibronectin, which is present at the surgical wound site. These proteins are recognized by microbial surface components that recognize matrix adhesion molecules, providing a niche for bacteria to form biofilms. The adhesion of bacteria can be influenced by the material that prosthesis is made and by its texture (Bhattacharya et al., 2015).
The hypothesis of this study is that the prevalence of S. aureus as a causal agent of biofilms or infections in prostheses and implants is due to its greater ability to adhere to medical devices when compared to other bacteria. So, we evaluate the in vitro colonization capacity of different bacterial strains on two prostheses with different textures (silicone prostheses and vascular prostheses that are two medical devices widely used in clinical practice) to assess whether S. aureus has greater prosthesis colonization capacity when compared to other bacteria that commonly colonize prostheses and wounds and whether the texture of the prostheses can influence this colonization.

Methods
This is a prospective, quantitative, and experimental study (Pereira et al., 2018).
About 50µL of bacterial frozen stocks samples (10 8 CFU/mL) were inoculated in BHI (brain heart infusion) and incubated at 37°C for 24 hours.

Determination of cellular hydrophobicity
The hydrophobicity of the bacterial cell surface was assessed by measuring the microbial adhesion to solvents (MATS) (Locatelli et al., 2004). Bacterial cultures were adjusted to an optical density (OD) 600 nm of 1.0 in 0.9% saline. A 0.4 mL volume of xylene was added to 3.6 mL of bacterial suspension. The mixture was vortexed for 5 minutes, and then the tubes were conditioned at room temperature for 20 minutes. The aqueous phase was removed, and the OD 600 nm was measured by a spectrophotometer (Digital Visible Spectrophotometer -Q898DRM, QUIMIS Scientific Apparatus, Diadema, São Paulo, Brazil). In this method, a greater interaction of microorganisms with a nonpolar liquid (p-xylene) represents a greater cellular hydrophobicity. Bacteria were considered hydrophobic when OD was less than 0.8 and hydrophilic when OD was equal to or greater than 0.8 (Locatelli et al., 2004).

Prostheses colonization test
To calculate the minimum sample size, the "pwr" package in R was used (Champely, 2018). The test power was 80%, significance level was 5%, number of groups to be compared was 10 (2 prostheses x 5 bacteria), and effect size (Cohen's d) was 0.50. So, it was concluded that at least 4 of each prosthesis per group were necessary. After incubation, the prostheses were seeded by rolling in Petri dishes (15x150 mm) containing 40 mL of agar (Electrolyte Deficient Cystine Lactose Agar -CLEDfor evaluation of P. mirabilis and blood agar for evaluation of other bacteria) and were incubated at 37°C for 48 hours.

Statistical analysis
The homogeneity of the variances was verified by the Levene test and the normality by the Kolmogorov-Smirnov test.
The analysis of biofilm formation did not present a normal distribution or homogeneity of variances, so the Kruskal-Wallis test was applied and the Dunn test to make the comparison between the groups.
To compare the mean intensity for the type of bacteria and the type of prosthesis, a general linear model was applied.
As the intensity did not present a normal distribution or homogeneity of variances, the Kruskal-Wallis test was also applied to compare the positions of each group, regardless of the type of prosthesis.
Differences were considered statistically significant when p <0.05. The tests were performed with SPSS v. 23.0.

Determination of cellular hydrophobicity
S. aureus, E. coli and E. faecalis were found to be hydrophobic, and S. epidermidis and P. mirabilis were found to be hydrophilic (Table 1).

Biofilm formation analysis
There was a significant difference in the analysis of biofilm formation between the bacterial groups (p = 0.0001) ( Figure 2). Fonte: Autores (2021).
Bacteria were considered non-adherent when OD was less than 0.049, weak adherent when OD was between 0.049 and 0.098, moderately adherent when OD was between 0.098 and 0.196 and strong adherent when OD was greater than 0.196.

Prostheses colonization test
There was a difference in intensity only for the types of bacteria (p = 0.0001), but there was no difference for the type of prosthesis (p = 0.496) or for the interaction prosthesis x bacteria (p = 0.586) (Figure 3).

Discussion
In this study, S. aureus was the bacterium that colonized both types of prostheses evaluated the most, although in the biofilm formation test, it was "weakly adherent". In contrast, E. coli was "strongly adherent" to the biofilm formation test but did not colonize either type of prosthesis evaluated. This corroborates our hypothesis that S. aureus is the most frequent cause of infection because it has a higher affinity for the materials from which medical devices are made.
Silicone breast implant infection occurs in 7 to 24% of implant breast reconstructions and the most common pathogens are S. epidermidis and S. aureus (Arad et al., 2013), and we corroborate these data.
ePTFE is chemically inert, hydrophobic and mechanically durable (Campbell et al. 1976). S. aureus and coagulasenegative Staphylococcus are responsible for 70 to 90% of vascular prosthesis infections (Turtiainen et al., 2014;Van De Vyver et al., 2017). Apparently, the microporous structure and synthetic composition of ePTFE grafts provide niches for bacterial accumulation and interfere with the ability of leukocytes to fight infections (Nadzam et al., 2000). In this study, the bacteria that most colonized the ePTFE prostheses were S. aureus followed by E. faecalis. These two bacteria have a hydrophobic character, which may justify having colonized more ePTFE prostheses. However, E. coli also presented a hydrophobic character but did not colonize these prostheses.
Interstitial porosity, surface tension and electronegativity of medical devices have been shown to attract and maintain bacterial adherence (Chang et al., 2003). This may explain the differences we found between the biofilm formation test, where E. coli was "strongly adherent" and S. epidermidis was "nonadherent", and the prosthesis colonization test, where E. coli did not colonize both prostheses, but S. epidermidis colonized the two types of prostheses in greater concentration. The microplates that were used for the biofilm formation test were made with polystyrene, a material very distinct from silicone or ePTFE, which may have led to this difference in results.
Although E. coli is a biofilm-forming bacterium (Henriques et al., 2013;Ziuzina et al., 2015), in this study, E. coli did not colonize silicone or ePTFE prostheses. Biofilm formation capacity seems to be associated with specific genes expressed in different strains of E. coli (Tiba et al., 2009). Additionally, most studies have evaluated the ability of biofilm formation by E.
coli in tissues (Tiba et al., 2009;Henriques et al., 2013;Ziuzina et al., 2015). Some studies have evaluated catheters associated with urinary tract infection and E. coli is the bacterium that colonizes these devices the most (Maharjan et al., 2018). These data may justify the fact that we did not observe colonization by E. coli in the prostheses; the difference may be due to the bacterial strain used in our study, which would make a difference in its ability to form biofilm, or because E. coli causes nosocomial infection by colonizing more tissue and only certain medical devices. In vivo studies will be needed to clarify this point.

Conclusion
With the data from this study, we demonstrated that S. aureus showed a higher affinity for colonization of silicone and ePTFE prostheses than the other bacteria tested and that E. coli, although showed strong adherence in the biofilm formation capacity test, did not colonize either prosthesis in the in vitro evaluation. This difference observed for E. coli between the biofilm formation capacity test and the adhesion in the evaluated prostheses shows that the bacterial strain and the material that prosthesis is made can influence the bacterial adhesion to them.