Metabolites of bovine-associated non-aureus staphylococci influence expression of Staphylococcus aureus agr-related genes in vitro

Communications via quorum sensing (QS) between non-aureus staphylococci (NAS) and Staphylococcus (S.) aureus in the bovine mammary gland remains largely unexplored. We determined whether 34 S. chromogenes, 11 S. epidermidis, and 14 S. simulans isolates originating from bovine milk samples and teat apices were able to regulate the QS of S. aureus, and if so, how in vitro growth inhibition of S. aureus by NAS, or NAS metabolites, or NAS cells themselves play a role in this process. In co-culture with S. aureus we observed that these 3 NAS species in general downregulated the expression of rnaIII, the effector molecule of the QS system, but this effect was more pronounced in S. chromogenes and S. simulans isolates than in S. epidermidis isolates. In vitro growth inhibition of S. aureus by NAS resulted in a small underestimation of the downregulating effect of NAS on rnaIII expression of S. aureus. Additionally, the culture supernatant of these NAS isolates and supernatant treated with proteinase K expressed greater regulatory activity over S. aureus virulence genes rnaIII, hla, and spa than washed NAS cells suspended in sterile water. These microbial interactions may influence S. aureus virulence and pathogenesis within the host. Isolation and identification of NAS metabolites affecting the QS system of S. aureus might help to develop alternative strategies for treatment and control of S. aureus mastitis. Supplementary Information The online version contains supplementary material available at 10.1186/s13567-021-00933-x.


Introduction
Bovine mastitis, an inflammation of the bovine mammary gland, can be caused by more than 100 bacterial species and subspecies [1]. The disease remains a major challenge to the dairy industry due to reduced milk production and quality, and substantial treatment costs [2]. One of the most common causative agents is Staphylococcus aureus [3][4][5]. Staphylococcus aureus remains extremely important for dairy herds because of its pathogenicity, contagiousness, capability to persist in the mammary gland, colonization of skin or mucosal epithelia, and poor cure rates when causing intramammary infections with the currently available therapies [3,5].
The collective behaviors of bacteria are largely controlled by cell-cell communication or quorum sensing (QS), in response to changes in the population density and species composition of the adjacent community [6]. Quorum sensing via the accessory gene regulator (agr) system plays a significant role in the pathogenesis of staphylococci, especially S. aureus [7]. The agr system is composed of a 2-component signal transduction complex which in response to a secreted auto-inducing peptide (AIP) stimulates the expression of a regulatory RNA designated rnaIII, the effector molecule of the agr system [8]. The accumulation of AIP to a critical threshold drives the transcription of rnaIII which, as a reciprocal effect, results in upregulated expression of exoproteins, including hla encoding α-hemolysin, and downregulation of surface-associated proteins, such as protein A encoded by the spa gene [9,10]. In addition to its regulation of the genes encoding individual virulence factors, QS coordinates activities including biofilm development, bioluminescence, bacterial competence, and virulence [11]. Therefore, QS has been the focus of many studies taking into account the potential advantages of using certain compounds for combating pathogens [12,13].
The coagulase-negative staphylococci, now more often called the non-aureus staphylococci (NAS), have become the most common bacteria isolated from bovine milk samples [14][15][16]. The role of NAS for bovine udder health is under scrutiny, with recent work discussing a wide diversity between species and even strains in epidemiology, ecology, virulence, and host-interaction [17][18][19][20][21]. Previous studies reported NAS as a relevant cause of bovine mastitis [17,22,23], while others reported only a slight increase of the milk somatic cell count (SCC) as an indicator of mammary gland inflammation [24,25] and no impact on milk yield, despite the elevated SCC [26]. A protective effect of NAS intramammary infections or NAS teat apex colonization against intramammary infections caused by major pathogens has also been reported [27][28][29]. Despite the literature supporting the predominance of NAS in the bovine mammary gland, the variation in the findings of research studies with regard to the role of the different bovine-associated NAS species originating from different habitats (e.g. milk, teat apex) needs further clarification.
agr-mediated interactions between NAS and S. aureus colonizing the same host niche have been recently suggested [30]. Some studies reported on the presence of potential NAS secreted compounds that might be responsible for the suppression of the S. aureus QS system, negatively affecting the ability of S. aureus to produce toxins [12,31,32]. However, to date, the interaction between S. aureus and NAS isolated from the bovine mammary gland or from bovine teat apices remains widely unstudied, except for one recent study [33]. Better insights in the interplay of NAS-S. aureus might be essential to understand the bacterial colonization process in the bovine mammary gland and to target S. aureus mastitis.
Based on the evidence that bacterial communications play an important role in niche generation and competition, we hypothesize that interactions of bovine NAS with S. aureus regulate the virulence of the latter in the bovine mammary gland. In order to investigate the hypothesis, we first investigated if bovine NAS isolates affect the agr QS of S. aureus when sharing the same niche, accounting for potential in vitro growth inhibition of S. aureus by NAS. Second, we examined whether metabolites produced and secreted by bovine NAS or NAS cells themselves influence the expression of S. aureus virulence factors controlled by the agr quorum sensing system.

General study design
First, the in vitro growth inhibition of S. aureus by NAS was evaluated in order to reveal potential bactericidal effects of bovine NAS isolates (n = 59) belonging to three different species [S. chromogenes (n = 34), S. epidermidis (n = 11), and S. simulans (n = 14)] and originating from two different habitats, milk (45 isolates) and teat apices (14 isolates) of primiparous and multiparous cows (in vitro growth inhibition assay). Next, it was studied whether NAS can act as inhibitors of S. aureus agr by measuring the β-Galactosidase activity of rnaIII::lacZ (β-Galactosidase liquid assay), taking into account the in vitro growth inhibition results. Last, it was examined whether active substances produced by NAS or NAS cells themselves influence the expression of S. aureus QSrelated genes (β-Galactosidase plate assay).

Bacterial isolates
Non-aureus staphylococci isolates were obtained from our repository. The isolates were previously assigned to a species by matrix-assisted laser desorption/ionizationtime of flight (MALDI-ToF) mass spectrometry analysis. In short, protein fingerprints of the isolates were compared with the commercial databank of bovine reference spectra (Bruker Daltonics), microbial spectra provided by Cameron et al. [34], and additional microbial spectra of field isolates from our lab covering 4 additional species (S. jettensis, S. lentus, S. rostri, and S. saprophyticus). The 59 NAS isolates were selected representing the three most prevalent species in milk samples and on teat apices of dairy cows and heifers [16,35]: 45 from milk-S. chromogenes (n = 28), S. epidermidis (n = 7), and S. simulans (n = 10), and 14 from teat apices (TA)-S. chromogenes (n = 6), S. epidermidis (n = 4), and S. simulans (n = 4) (see Additional file 1). The S. aureus isolates used in this study and their references are listed in Additional file 1 as well.
Unless otherwise stated, bacteria were grown in Tryptic Soy Broth (TSB) for 16-24 h at 37 °C.

In vitro growth inhibition assay
To evaluate potential growth inhibition of S. aureus by NAS, the cross-streaking method was used [27].
First, bacterial cultures of the 59 NAS isolates were adjusted to 0.5 McFarland turbidity standard, and inoculated as a center-streak (width of 5 mm) on Columbia Sheep Blood Agar Petri-dishes for 24 h at 37 °C. The agars were circularly loosened and then turned upside down and 0.5 McFarland standard suspensions of the S. aureus strain 8325-4 (see Additional file 1) was swabbed to achieve full coverage. The presence of a growth inhibition zone (measured in mm) within and adjacent to the center-streak was investigated after another 24 h of incubation for each NAS-S. aureus combination. As part of the set of 59 NAS isolates, the isolate SC29-"S. chromogenes TA", a characterized inhibitor-S. chromogenes C2 in De Vliegher et al. [27]-was used as a positive control. A zone of total growth inhibition of S. aureus strain 8325-4 (TGI) was declared when no colonies were observed, whereas a zone of partial growth inhibition zone (PGI) was declared when smaller and/or less colonies of S. aureus strain 8325-4 were observed. If the same size/numbers of colonies of S. aureus strain 8325-4 were present as on the positive control plate (S. aureus 8325-4 alone), the zone was declared as no growth inhibition (NGI).
The experiments were performed in triplicate on three different days and results were averaged over the replicates. The distribution (%) of NAS inhibiting the growth of S. aureus 8325-4 was compared among the three different NAS species (S. chromogenes, S. epidermidis, S. simulans) and between the two different NAS habitats (milk and TA).

β-Galactosidase liquid assay
This assay aimed to verify S. aureus agr inhibitory activity by bovine NAS when sharing the same niche environment. The overnight cultures of the SH101F7 S. aureus rnaIII reporter isolate [12] (see Additional file 1) and the 59 abovementioned NAS isolates were diluted 100 × in 15 mL of TSB, and allowed to reach an OD 600 of 0.5. Each isolate was adjusted to an OD 600 of 0.1 in TSB, and added to the same tube at a ratio of 1:1. From each culture, 1 mL was taken hourly (for 4 h) and centrifuged for 3 min at 7.2 g and 4 °C. The supernatants were removed after centrifugation, and the pellets recovered in 1 mL of TRIS 50 mM, pH 8 and 3 µL of lysostaphin. The mix was incubated at 37 °C for 30 min to allow cell lysis. Z-buffer (400 µL) was then added to each sample which was further incubated for 5 min at 28 °C. Lastly, 100 µL of ONPG (ortho-Nitrophenyl-βgalactosidase) (4 mg/mL) was added to the mix, and the time necessary for the solution to turn yellow was controlled.
Each assay was performed in triplicate on three different days. The absorbance at 420 and 550 nm for each sample was measured. The activity was calculated in Miller units as described by Miller [36]. It was analysed whether the rnaIII expression of S. aureus differed between the three different NAS species (S. chromogenes, S. epidermidis, S. simulans), between NAS originating from the two different habitats (milk and TA), and between NAS isolates based on their in vitro growth inhibition of the abovementioned S. aureus strain 8325-4 (no growth inhibition, partial inhibition and total inhibition as determined with the in vitro growth inhibition assay). Staphylococcus schleiferi 2898 served as positive control [12] and rnaIII expression of S. aureus in the absence of NAS served as negative control.

β-Galactosidase plate assay
To determine whether active substances produced and secreted by NAS and NAS cells themselves have an effect on the expression of S. aureus virulence factors controlled by the agr quorum sensing system (QS) the β-Galactosidase plate assay was performed [37].
First, supernatants and bacterial cell suspensions were obtained from cultures of the 59 NAS isolates. Briefly, 16 h cultures of NAS isolates in TSB were centrifuged at 3000 g for 1 h [38] and the supernatants were carefully collected and filtered through 0.4 μm pore size filters. The pellets were washed twice with sterile water and suspended in 10 mL of sterile water, in order to obtain fresh NAS cell suspensions (CS). For each isolate, part of the filtered culture supernatant was left untreated (SP) and the other part was treated with proteinase K (SPK) at a concentration of 50 μg mL −1 for 1 h at 37 °C in a reaction mixture containing 0.5% SDS, 0.01 M Tris and 0.005 M EDTA. Following treatment, the enzyme was inactivated by adding phenylmethylsulfonyl fluoride. As an internal control, the supernatant of S. aureus 8325-4 (AIP-I; Additional file 1) was used to induce agr.
Then, three reporter strains of S. aureus carrying lacZ fused to central virulence genes (hla, rnaIII, and spa) encoding resistance to erythromycin (see Additional file 1) were grown for 16 h, and 2 mL of 10 3 -diluted cultures were placed in Greiner plates to which 50 mL of tryptone soy agar (TSA) (~40 °C) containing 150 µg/ mL of 5-bromo-4-chloro-3-indolyl-D-galactopyranoside (X-Gal). In order to maintain selective pressure for the plasmids, 5 µg/mL of erythromycin was added to the agar. The plates were subsequently left to solidify, and wells were shaped manually with a sterile sharp iron drill (4 mm). Aliquots of 20 μL of SP, SPK, CS, and controls were added to plates containing the different reporter strains. The incubation time varied between the different reporter strains. Therefore, the plates were incubated at 37 °C for 9 to 36 h until a blue color appeared on the plates ( Figure 1). The presence of a halo zone around the well indicated the regulation of the virulence genes (rnaIII, hla and spa) with the degree of the effect depending on the diameter of the halo zone (measured in mm) [12] and stratified as exhibiting no effect (≤ 10 mm), a slight effect (11-15 mm), a moderate effect (16-20 mm), or a severe effect (≥ 25 mm) on gene expression.
All experiments were performed in triplicate on three different days and results were averaged over replicates. The distribution (%) of NAS isolates affecting S. aureus virulence gene expression was compared between the three different NAS species (S. chromogenes, S. epidermidis, S. simulans), between the NAS belonging to the two different habitats (milk and TA) and between the three different culture preparations (SP, SPK and CS). The SP, SPK and CS of S. schleiferi 2898 [12] were used as positive controls, whereas H 2 0 was used as negative control.

Statistical analyses β-Galactosidase liquid
The expression of the rnaIII gene of S. aureus (outcome variable) during co-culture with different NAS species (predictor variable of main interest) was studied using linear mixed models (PROC MIXED; SAS version 9.4; SAS Institute Inc., Cary, NC, USA) taking into account the repeated measurements. A natural logarithmic transformation of rnaIII expression values (LnRnaIII) was performed to obtain a normal distribution.
First, a linear mixed model was fit with staphylococcal species [5 levels: S. aureus only (negative control), S. aureus + S. schleiferi (positive control), S. aureus + S. chromogenes, S. aureus + S. epidermidis, and S. aureus + S. simulans] as a categorical predictor variable of main interest to compare NAS with the positive and negative controls, respectively. Time of measurement (4 levels: 1 h, 2 h, 3 h, and 4 h) and the interaction term between staphylococcal species and time of measurement were included as additional categorical predictor variables. Isolate was included as random effect to account for the three replicates per isolate and replicate was included as repeated effect to account for the 4 repeated measurements per replicate.
Second, a linear mixed model was fit with NAS species [3 levels: S. aureus + S. chromogenes, S. aureus + S. epidermidis, and S. aureus + S. simulans], the habitat of the NAS isolates (2 levels: milk and TA) and the in vitro growth inhibition of S. aureus by NAS (3 levels: total, partial, or no growth inhibition; see before) as categorical predictor variables and all possible two-way interactions. Isolate was included as random effect to account for the three replicates per isolate and replicate was included as repeated effect to account for the 4 repeated measurements per replicate. Non-significant variables (p > 0.05) were omitted from the full model , or (C) the spa::lacZ (PC203; Ery r ) reporter strain of Staphylococcus aureus were exposed to 20 μL of either Supernatant (SP), Supernatant + Proteinase K (SPK), or Cell suspension (CS) obtained from overnight cultures of Staphylococcus chromogenes, Staphylococcus epidermidis, and Staphylococcus simulans, respectively. Staphylococcus schleiferi (strain 2898) [12] and H 2 O were used as positive control (P) and negative control (N), respectively. Halos around the wells appeared between 12 and 36 h of incubation at 37 °C, and the diameter (measured in mm) was classified as NE: no effect (≤ 10 mm), SLE: slight effect (11-15 mm), ME: moderate effect (16-20 mm), and SE: severe effect (≥ 25 mm) on gene expression (upregulation for rnaIII and hla, and downregulation of spa). This figure is representative for all β-Galactosidase plate assays. using a backward stepwise approach. The goodness-offit measures included −2 × log-likelihood, the Akaike information criterion, and the Bayesian information criterion. Residuals were evaluated graphically and plotted against the predicted values. A Bonferroni's correction was used to correct for multiple comparisons. Significance was assessed at p ≤ 0.05.

β-Galactosidase plate assay
The Fisher's exact test was used to determine whether the distribution of effects (no effect, slight effect, moderate effect, or severe effect on the expression of rnaIII, hla, and spa genes) differed (1) between NAS isolates belonging to the three different species (3 levels: S. chromogenes, S. epidermidis, and S. simulans), or (2) between NAS isolates originating from the two different habitats (2 levels: milk or teat apices), or (3) between the different culture preparations (3 levels: SP, SPK, CS). Statistical analyses were performed using SPSS v.26.0 (IBM Corp., Armonk, NY, USA) and p ≤ 0.05 was considered significant.

In vitro growth inhibition
The patterns of the in vitro growth inhibition of S. aureus 8325-4 by the NAS isolates are presented in Table 1. Fifty NAS isolates out of 59 (84.7%) were able to inhibit the growth of S. aureus at least partially. Total growth inhibition was observed only by one S. simulans isolate from a TA and by the positive control strain S. chromogenes SC29-TA, also part of the set of 59 NAS isolates.

β-Galactosidase liquid assay Comparison with the controls
The rnaIII expression increased significantly over time [p = 0.0453; Least Square Means (LSM) increased from 40.95 to 58.95 back-transformed Miller units] and differed between staphylococcal species (p < 0.0001); yet the evolution of the expression over time significantly differed between staphylococcal species (p < 0.0001; Table 2 and Figure 2A

Association with NAS species, NAS habitat and NAS in vitro growth inhibition
The rnaIII expression increased significantly over time  Table 3 and Figure 2B). Neither the NAS habitat nor the NAS in vitro growth inhibition of S. aureus 8325-4 influenced the rnaIII expression significantly. However, forcing NAS in vitro growth inhibition into this final model in order to better understand whether the differences in effect between species was due to in vitro growth inhibition of S. aureus in the co-culture (Table 3, Figure 2B), the NAS species effect became slightly stronger (i.e. LSM values became smaller, indicating the reduction in rnaIII expression was more pronounced).

Discussion
We demonstrated that bovine-associated NAS downregulate the S. aureus rnaIII gene of the strain when sharing the same niche in vitro, an effect that is much more pronounced in S. chromogenes and S. simulans than it is in S. epidermidis. The difference in downregulation between S. chromogenes and S. simulans on the one hand and S. epidermidis on the other hand, is even slightly more pronounced when the in vitro growth inhibition on S. aureus is taken into account. As well, we demonstrated that rnaIII regulation observed in the liquid assay was comparable with the results from the plate assay, with less effect Table 2 Final linear mixed effect model for the β-galactosidase liquid assay for rnaIII gene expression including the positive and negative controls 1 Estimate. 2 Standard error. 3 Least square means after back transformation. 4 Overall P value for fixed effects. 5 Staphylococcus aureus rnaIII::lacZ reporter strain SH101F7 was used as negative control when growing alone and as positive control when in co-culture with Staphylococcus schleiferi strain 2898. 6 The interaction term is visualized in Figure 2A. Staphylococcal species x Time of measurement 6 < 0.0001 of S. epidermidis, and no differences in effect according to the origin of the NAS isolates (milk versus teat apices). The plate assay also showed that substances produced by the NAS can regulate rnaIII, hla, and spa expression. The effects of the washed NAS cells themselves were less pronounced. This is the first study that investigated whether the in vitro growth-inhibitory effect of S. aureus by bovine NAS also affects the activation of S. aureus agr system. Previously it has been suggested that bacteriocins produced by NAS, mainly originating from teat apices, are responsible for growth inhibition of S. aureus [16,27,[39][40][41]. Some studies also reported the role of bacteriocins as signaling peptides in quorum sensing and bacterial cross talk within microbial communities [42], but none described this mechanism specifically for the NAS species. In our study, total growth inhibition of S. aureus by two bovine NAS isolates belonging to S. chromogenes (isolate SC29-likely bacteriocin producer, Additional file 1) and S. simulans and both originating from teat apices was confirmed, next to partial S. aureus inhibition exerted by most of the other isolates. The in vitro growth inhibition pattern, however, did not have a major impact on the growth of S. aureus when sharing the niche in the short span (4 h) of the experiments. As shown, in a separate small experiment with a subset of NAS isolates, the number of S. aureus cells actually remained unchanged within the time frame of the beta-galactosidase liquid assay (4 h) (see Additional file 2). In addition, when the statistical model was corrected by including the effect of growth inhibition, only a slightly change in the LSM of the species effects was observed. Actually, correcting  [27]. Each bar represents the least square means of three biological replicates and the error bars represent the standard deviation.
for the inhibitory effect of NAS showed that the downregulation of rnaIII would be slightly more pronounced when such effect would not be present. These results suggest that some NAS isolates can affect the activation of S. aureus agr system via a mechanism that does not involve growth inhibition.
Despite the well documented literature suggesting the effect of NAS on the regulation of the agr system [12,[30][31][32], our understanding of the crosstalk between bovinerelated NAS and S. aureus is still limited [33,43]. In that respect, the use of the laboratory strain 8325-4 has been widely accepted in the animal research field since it was first used [44]. Since, the strain has served as a model for global virulence regulation in S. aureus [12,30,33]. Here, we present data confirming the capacity of different NAS to suppress S. aureus rnaIII expression, a finding that aligns with the results from Mahmmod et al. [33], in which NAS originating from milk and teat apices from dairy cows had the ability of cross-interfering with the S. aureus agr quorum sensing system.
Although the majority of the NAS isolates were able to regulate agr in our study, it was more pronounced in S. chromogenes and S. simulans isolates. Nevertheless, the degree of agr regulation displayed by the NAS isolates (no effect to severe effect) was not the same, suggesting that the ecological niche the isolates originate from might also be important [12]. However, we report that the agr activity of S. aureus was equally regulated by isolates originating from milk and TA, which is slightly different from what was reported earlier [33] where NAS isolates Table 3 Final linear mixed regression model for the β-galactosidase liquid assay for rnaIII gene expression and final model corrected for in vitro growth inhibition of Staphylococcus aureus by bovine non-aureus staphylococci (NAS) 1 Estimate. 2 Standard error. 3 Least square means after back transformation. 4 Overall P value for fixed effects. 5 Non-significant. 6 See Materials and methods. 7 The interaction term is visualized in Figure 2B. 8  originating from TA numerically appear to be more likely to regulate S. aureus agr. In our study, differences observed within and between species may be related to the different traits of the NAS isolates. Therefore, further investigations at the strain level must be carried out for a better understanding of their behavior. It has been demonstrated that the signaling molecules produced by NAS resemble the AIPs of the S. aureus agr quorum sensing system [12,[30][31][32]. Others suggested that AIP-like molecules present in the culture supernatant of NAS might be responsible for downregulating rnaIII and hla expression and upregulating expression of spa [33]. We not only studied the effect of NAS supernatant (SP) on the regulation of rnaIII, hla, and spa but also of supernatant treated with proteinase K (SPK), and cell suspension (CS). Our finding that SP obtained from different NAS regulates the agr system of S. aureus aligns with previous studies [12,30,33]. Canovas et al. [12] reported a considerable suppression across all four agr groups of S. aureus conferred by the AIP-containing supernatant of the S. schleiferi strain 2898 we also used as positive control. They also reported that the pure synthetic S. schleiferi AIP was able to completely abolish agr induction of an S. aureus reporter strain. Later, Peng et al. [30] showed that NAS isolated from pigs regulate the agr system by competing with the S. aureus AIPs for AgrC binding, generally resulting in suppression of the S. aureus agr. Although Mahmmod et al. [33] did not identify the exact mechanism involved in the cross-talk between S. aureus and NAS originating from milk and TA, they suggested that AIP-like molecules are most likely responsible for the agr-suppression caused by their NAS isolates.
The overall regulatory effect of proteinase K treated supernatant (SPK) was not different from the nontreated supernatants (SP). Notwithstanding that we did not isolate and identify which compounds are responsible for such regulatory mechanism, we speculate that NAS produce AIP-like molecules, which are resistant to proteinase K [45], that inhibit the S. aureus

Table 4 Downregulation of the rnaIII gene activity of Staphylococcus aureus by three different culture preparations obtained from bovine non-aureus staphylococci (NAS) isolates from milk and teat apices (TA) [no. (%)]
1 Culture preparations = SP-Supernatant; SPK-Supernatant + Proteinase K; CS-Cell suspension. 2 Number of isolates. 3 Regulation effect was rated according to the size of inhibition zone around the well (measured in mm) [12] and was classified as exhibiting no effect (≤ 10 mm), a slight effect (11-15 mm  agr system. Preliminary results of additional tests performed in our lab seem to substantiate that nature (data not shown). These peptides are synthesized in the bacterial cell and transported across the cell membrane by specialized membrane transport proteins [8]. agr activity in the bacterial cells will be activated during the late log-phase bacterial growth, when the concentration of the AIP in the medium is high [46]. As the AIP molecule is secreted extracellularly with the higher concentration obtained at late log-phase bacterial growth, we speculate that the concentration of the AIP molecules must be greater in SP and SPK than in the CS, which could partially explain our results. This hypothesis is supported when we consider the results of the plate assay. The use of lacZ as a gene reporter in the S. aureus strains (Additional file 1) claims to measure the rnaIII, hla, and spa promoter activity under selective pressure of erythromycin [47], a condition to guarantee plasmid stability. Therefore, we consider that growth and production of metabolites by NAS cells were probably reduced in erythromycin-susceptible isolates, affecting the capacity of CS to regulate the virulence genes of S. aureus, and potentially (partially) explaining the reduced CS effects. This data shown that the agr regulatory effect is probably not related to the use of cells (CS), but with metabolites present in the supernatant (SP and SPK); however, further investigations have to be carried out to confirm such hypothesis. Previous studies identified S. epidermidis AIP as a potent suppressor of the S. aureus agr system, which might explain the predominance of S. epidermidis on the human skin [31]. In contrast, we observed a less pronounced regulatory effect of the S. epidermidis isolates against S. aureus growth as well as on the regulation of the expression of the S. aureus agr system compared with S. chromogenes and S. simulans. Our data corroborate with the findings of Mahmmod et al. [33], which also reported a variable effect of S. epidermidis isolated   Table 6 Upregulation of the spa gene activity of Staphylococcus aureus by three different culture preparations obtained from bovine non-aureus staphylococci (NAS) isolates from milk and teat apices (TA) 1 Culture preparations = SP-Supernatant; SPK-Supernatant + Proteinase K; CS-Cell suspension. 2 Number of isolates. 3 Regulation effect was rated according to the size of inhibition zone around the well (measured in mm) [12] and was classified as exhibiting no effect (≤ 10 mm), a slight effect (11-15