Bovine-associated non-aureus staphylococci suppress Staphylococcus aureus biofilm dispersal in vitro yet not through agr regulation

Biofilm formation is a significant virulence factor in Staphylococcus (S.) aureus strains causing subclinical mastitis in dairy cows. A role of environmental signals and communication systems in biofilm development, such as the agr system in S. aureus, is suggested. In the context of multispecies biofilm communities, the presence of non-aureus staphylococci (NAS) might influence S. aureus colonization of the bovine mammary gland, yet, such interspecies interactions have been poorly studied. We determined whether 34 S. chromogenes, 11 S. epidermidis, and 14 S. simulans isolates originating from bovine milk samples and teat apices (TA) were able to affect biofilm formation and dispersion of S. aureus, and if so, how isolate traits such as the capacity to regulate the S. aureus agr quorum sensing system are determinants in this process. The capacity of an agr-positive S. aureus strain to form biofilm was increased more in the presence of S. chromogenes than in the presence of S. simulans and S. epidermidis isolates and in the presence of NAS isolates that do not harbor biofilm related genes. On the other hand, biofilm dispersion of this particular S. aureus strain was suppressed by NAS as a group, an effect that was more pronounced by isolates from TA. Furthermore, the observed effects on biofilm formation and dispersion of the agr-positive S. aureus strain as well as of an agr-negative S. aureus strain did not depend on the capacity of NAS to repress the agr system. Supplementary Information The online version contains supplementary material available at 10.1186/s13567-021-00985-z.


Introduction
Bovine mastitis, a multifactorial disease-complex resulting from interactions among the host, environment, and typical bacteria, remains the most prevalent and challenging disease on dairy farms worldwide [1]. The nonaureus staphylococci (NAS), a heterogeneous group of staphylococcal species other than Staphylococcus aureus [2][3][4] have become the most frequently isolated bacteria from bovine milk samples as revealed by numerous studies [5][6][7][8][9]. They are commonly considered to be minor mastitis pathogens causing only a slight increase of the milk somatic cell count (SCC) [10][11][12] with no impact on milk yield. Other studies have suggested a protective effect of bovine NAS against, among other bacteria, S. aureus [13][14][15][16][17], including the repression of agr related virulence factors of S. aureus [18,19].
Biofilms are defined as a structured surface-associated community of bacterial cells (sessile) encompassed by an extracellular matrix [24]. Biofilm development is related to environmental signals and communication systems [25]. In staphylococci it relates to specific gene expression such as the intracellular adhesion locus (icaABCD) [26], the biofilm-associated protein bap [27], and the accumulation-associated protein aap [28]. In S. aureus, the switch from planktonic (free-floating forms) to sessile forms is controlled by quorum-sensing proteins encoded by the agrABCD operon, whose abundance is related to virulence and pathogenicity [29]. Therefore, high levels of agr activity is typically associated with dispersal of S. aureus biofilms, whereas the agr system of cells in a biofilm is predominantly repressed [30,31].
The thickness and the composition of biofilm might determine its functionality [32] and the persistence of bacteria in the bovine mammary gland [33]. Thicker biofilms are notoriously more difficult to eradicate as they are generally resistant to antibiotic therapy [34] and clearance by host defenses in the mammary gland [32]. Furthermore, multispecies biofilms might play a role in the context of host colonization, since the presence of specific bacteria can either promote or decrease growth of other bacteria [35,36].
Although a number of studies in literature reported on the protective effect of NAS in the bovine mammary gland, little is known in the context of biofilm formation and whether agr-mediated interactions between bovine NAS and S. aureus are important. Also, Goetz et al. [32] reported on the capacity of NAS isolates, mainly S. chromogenes and S. simulans, to reduce biofilm formation and to stimulate dispersion of pre-established biofilm of S. aureus. However, the potential mechanisms behind the biofilm-inhibition and dispersion effects of NAS have not yet been identified. Such dual-species interactions might benefit both species with respect to host colonization [37]. Moreover, a robust multispecies biofilm might contribute to the persistence of S. aureus infections [38], also in the bovine udder.
Based on the rationale that regulation of the agr system can determine biofilm production by S. aureus, we hypothesized that agr-mediated interactions between NAS and S. aureus affect S. aureus biofilm formation and dispersal in the bovine mammary gland. In order to investigate the hypothesis, we examined whether bovineassociated NAS originating from milk samples and teat apices (TA) influence biofilm formation and dispersion of S. aureus in vitro, and if so, to determine what NAS traits are involved and whether such effects are associated with the regulation of the S. aureus agr quorum sensing system.

General study design
First, the presence of genes related to biofilm formation (biofilm genotype) and the capacity to produce biofilm (biofilm phenotype) were mapped for 59 bovine NAS isolates originating from milk or TA (Additional file 1). The capacity to inhibit the growth of S. aureus 8325-4 (an agr + strain) in vitro (in vitro growth inhibition capacity) and to modulate the agr system of S. aureus 8325-4 (agr interaction capacity) for the 59 NAS isolates was known from previous work [19] (Additional file 1).
Next, it was studied whether the 59 NAS isolates were able to affect S. aureus 8325-4 biofilm formation as well as whether they were able to disperse pre-established S. aureus 8325-4 biofilm and to what extent this was related to specific NAS traits [species, origin (milk versus TA), biofilm genotype, biofilm phenotype, and in vitro growth inhibition capacity, respectively].
Last, it was studied whether the capacity of the 59 NAS isolates to repress the agr system of S. aureus was related to biofilm formation and biofilm dispersion by S. aureus strains.

Staphylococcal isolates and traits
The isolates used in this study are listed in Additional file 1. Unless otherwise stated, bacteria were grown in Trypticase Soy Broth (TSB) overnight at 37 °C.
Species identification of all NAS isolates was carried out by matrix-assisted laser desorption/ionization-time of flight (MALDI-ToF) mass spectrometry analysis in which protein fingerprints of the isolates were compared with the commercial databank of bovine reference spectra (Bruker Daltonics), microbial spectra provided by Cameron et al. [40], and additional microbial spectra of field isolates from our lab covering four additional species (S. jettensis, S. lentus, S. rostri, and S. saprophyticus).

Biofilm genotype
The bacterial DNA of the NAS was extracted [43], with some modifications to identify biofilm associated genes among NAS. In the final step, 3 µL of RNase (1 mg/ mL, Roche, Mannheim, Germany) was added. Polymerase chain reaction (PCR) assays were applied to detect the presence of the genes related to biofilm formation (aap, ica, and bap) and quorum sensing (agr) according to methodology previously described [27,28,44,45] ( Table 1). PCR products were analyzed by electrophoresis through 1.5% (wt/vol), previously stained with Ethidium Bromide (Sigma-Aldrich). Staphylococcus xylosus ATCC 29971 was used as a positive control for the bap gene and S. epidermidis ATCC 35984 for the aap and ica genes. Positive control for the agr system included S. aureus strain 8325-4. Isolates with an equal amplicon size to the positive control were considered PCR positive for the particular gene under scrutiny.
The NAS isolates that presented at least one of the four genes were recoded as positive for biofilm-related genes (BG), whereas the absence of all four genes was recoded as negative for biofilm-related genes (NBG; Additional file 1) for statistical analyses.

Biofilm phenotype
The potential of the NAS isolates to form biofilm was evaluated according to the assay described by Tremblay et al. [23], with some modifications. Briefly, overnight cultures of S. aureus strain 8325-4 [41] and strain 8325-4 Δagr [42] (Additional file 1), and the 59 NAS isolates were adjusted to OD 600 = 0.2 in TSB and then further diluted 1:100 in TSB supplemented with 0.2% glucose (TSBG). Therefore, an aliquot of 200 μL of each isolate was seeded per well in sterile 96-well polystyrene tissue culture plates at 37 °C for 24 h. After incubation, planktonic cells (supernatant) were removed and the wells were rinsed three times with distilled water. Adherent biofilms were fixed with 95% ethanol for 15 min, and stained with 0.1% (wt/vol) safranin for 20 min. After rinsing three times with distilled water, plates were allowed to dry at room temperature. The stained biofilms were dissolved in 250 μL of 33% (v/v) acetic acid, and optical densities (OD) were measured at 490 nm using Multiskan GO plate reader (Thermo Fisher Scientific, USA). The ability of the isolates to form a biofilm was classified as negative (absorbance at 490 nm, A 490 < 0.110), weak (A 490 0.110-0.500), moderate (A 490 0.500-1.500), or strong producers (A 490 > 1.500) (Additional file 1). Uninoculated wells containing TSB with glucose served as blanks. Staphylococcus epidermidis strain ATCC 35984 was used as a strong biofilm forming control strain [46], and S. epidermidis strain ATCC 12228 was used as a weak biofilm forming control strain [47].
Each NAS isolate was tested for biofilm production in triplicate on two independent days and results were averaged over the replicates.
All NAS isolates that presented any capacity to produce biofilm (weak, moderate or strong producers) were recoded as positive for biofilm production (BP) for statistical analyses. On the other hand, the absence of biofilm production by NAS was recoded as negative for biofilm production (NBP; Additional file 1).

In vitro growth inhibition capacity
The potential growth inhibition of S. aureus 8325-4 by the NAS isolates was evaluated before [19] (Additional file 1). Following the methodology described by De Vliegher et al. [13], the pattern of the in vitro growth inhibition of the S. aureus strain 8325-4 by the NAS isolates was classified as NAS exhibiting no growth inhibition, partial inhibition, or total inhibition.
The NAS isolates that presented partial or total inhibition were recoded as positive for in vitro growth inhibition of S. aureus (GI), whereas no growth inhibition was recoded as negative for in vitro growth inhibition of S. aureus (NGI; Additional file 1) for statistical analyses.

agr repression
Some NAS isolates were able to repress the agr system of S. aureus 8325-4 by downregulating the rnaIII expression  [19] (Additional file 1). As described by Canovas et al. [42], the effect of NAS isolates on S. aureus agr was classified as none, weak, moderate, or strong. All NAS isolates that showed any capacity to repress S. aureus agr (weak, moderate, or strong) were recoded as positive for agr repression (RP) for statistical analyses. Conversely, the absence of agr repression was recoded as negative for agr repression (NRP; Additional file 1). All plates were stained with 0.1% (wt/vol) safranin and the optical density measured (as abovementioned) to confirm the complete detachment of the biofilm fraction from the bottom (data not shown). Each biofilm fraction recovered from the wells was tested for biofilm formation in triplicate on two independent days and results were averaged over the replicates. A CFU/mL lower than numbers recovered from the positive control indicates less S. aureus biofilm formation, whereas the opposite is true for higher values of CFU/mL.

Staphylococcus aureus biofilm dispersion
To evaluate the potential of the NAS isolates to disperse the preformed biofilm of S. aureus, the biofilm dispersion assay was performed [32, adapted]. Briefly, S. aureus strain 8325-4 biofilms were grown as described in the Section "Biofilm phenotype", and after the 24-h incubation, planktonic cells were discarded. The pre-formed S. aureus biofilms were washed once with sterile water. After, overnight cultures of all 59 NAS were adjusted in TSBG as abovementioned (Section "Biofilm phenotype") and 200 µL of the suspensions were added directly to the S. aureus pre-formed biofilms. Plates were incubated for another 24 h at 37 °C. Finally, the planktonic cells were removed (dispersed fractions) and serially diluted. The dilutions were plated on MSA plates for CFU enumeration of S. aureus 8325-4 (the outcome variable of interest). Single-species biofilm of S. aureus 8325-4 served as positive control and uninoculated wells containing TSBG served as negative control.
All plates were stained with 0.1% (wt/vol) safranin and the optical density measured (as abovementioned) to confirm the complete detachment of the biofilm fraction from the bottom (data not shown). Each dispersion fraction recovered from the wells was tested for biofilm dispersion in triplicate on two independent days and results were averaged over the replicates. A CFU/mL lower than numbers recovered from the positive control indicates less S. aureus biofilm dispersion, whereas the opposite is true for higher values of CFU/mL.

Statistical analyses
All statistical analyses were performed using SPSS v.27.0 (IBM Corp., Armonk, NY, USA) and P ≤ 0.05 was considered significant. One-way analysis of variance (ANOVA) was used to determine whether the biofilm formation as well as biofilm dispersion of S. aureus 8325-4 (CFU/mL; normally distributed as assessed through Q-Q plots and Kolmogorov-Smirnova test of normality) differed: (1) between NAS according to their species (3 levels: S. chromogenes, S. epidermidis, and S. simulans), (2) between NAS isolates originating from the two different habitats (2 levels: milk or TA), and (3) between NAS harboring biofilm-related genes or not [2 levels: positive for at least one gene (BG) and negative (NBG); biofilm genotype), between NAS with capacity to produce biofilms themselves or not [2 levels: positive for any capacity to produce biofilm (BP) and negative (NBP); biofilm phenotype], between NAS with capacity to inhibit growth of S. aureus in vitro or not [2 levels: positive for partial or total growth inhibition of S. aureus (GI) and negative (NGI)]. In addition, one-way ANOVA was carried out to verify whether the biofilm formation and dispersion of S. aureus 8325-4 and 8325-4 Δagr (CFU/mL) was affected by the NAS capacity to suppress the agr system of S. aureus 8325-4 [2 levels: positive for any suppression of S. aureus agr (RP) and negative (NRP)].
Least significant difference (LSD) test was used to compare with the controls and post-hoc Bonferroni corrections were applied for all other comparisons.

Staphylococcus aureus biofilm formation
Biofilm formation of S. aureus 8325-4 was quantified in the presence and absence (control) of NAS ( Figure 1).
Overall, NAS as a group (comprising S. chromogenes, S. epidermidis, and S. simulans) did not influence S. aureus 8325-4 biofilm formation. Also, neither the origin of the NAS isolates (milk versus TA; Figure 1B) nor the biofilm phenotype ( Figure 1D), nor the in vitro growth inhibition capacity of NAS ( Figure 1E) were related to S. aureus biofilm formation. Still, S. chromogenes stimulated S. aureus biofilm formation (P = 0.028; Figure 1A). Also, the stimulation of the S. aureus biofilm formation was more pronounced for S. chromogenes than for S. epidermidis (P = 0.001) and S. simulans (P = 0.053; Figure 1A). Also, NAS not carrying biofilm genes stimulated S. aureus biofilm formation (P = 0.048), which was not true for NAS carrying biofilm genes although there was a strong tendency (P = 0.069; Figure 1C).
The capacity of S. aureus to form biofilm did not depend on the capacity of NAS to repress the S. aureus agr system, as evidenced for both an agr-positive and agr-negative S. aureus strain (Figure 2A).

Staphylococcus aureus biofilm dispersion
The quantitative determination (CFU/mL) of biofilm dispersion of S. aureus 8325-4 in the presence and absence (control) of NAS isolates is shown in Figure 3.
Overall, NAS as a group (comprising S. chromogenes, S. epidermidis, and S. simulans) significantly suppressed S. aureus 8325-4 biofilm dispersion, although this was more pronounced in NAS isolates originating from TA than from milk (P = 0.01; Figure 1B). Neither the NAS species ( Figure 1A), nor the biofilm genotype ( Figure 1C), nor the biofilm phenotype ( Figure 1D), nor in vitro growth inhibition capacity of NAS ( Figure 1E) were associated with S. aureus biofilm dispersion.
The ability of NAS to suppress S. aureus biofilm dispersion did not depend on their capacity to repress the agr system, as evidenced by the fact that the suppression was observed for both an agr-positive and agr-negative S. aureus strain ( Figure 2B).

Discussion
We showed that S. chromogenes stimulates S. aureus biofilm formation. The latter effect is most likely, at least partially, explained by the fact that S. chromogenes isolates were seldomly carrying biofilm genes or producing biofilms themselves, an association that was also evident from our experiments, a finding that is discussed later. On the other hand, all NAS isolates, especially those originating from TA, were able to suppress S. aureus biofilm dispersion. Interestingly, neither the biofilm formation nor the biofilm dispersion of S. aureus in the presence of bovine NAS isolates was regulated via the S. aureus agr quorum sensing system.
As was expected, the capacity of the NAS isolates to form biofilm themselves varied among the species, with S. epidermidis isolates presenting the highest ability to do so (72.7%). Conversely, a study [23] reported that S. epidermidis displayed the lowest ability to produce biofilm among the same three NAS species as used in our study, although the overall percentage of isolates forming biofilm was quite similar to ours (69.3%). The percentage of S. chromogenes (8.8%) as well as S. simulans isolates (35.7%) able to form biofilm in our study was considerably lower than the percentage observed by the other study [23], with respectively 84.7% and 84.9% of their isolates being biofilm formers. The media used for the biofilm assays (BHI vs. TSB) could be one of the reasons for the differences observed between studies. As well important, in the other study [23] all the NAS isolates originated only from bovine milk, whereas ours from both bovine milk and TA. Still, the 44.1% of S. chromogenes isolates harboring at least one of the biofilm-related genes in our study was also much lower than the 82% found by Tremblay et al. [23].
Interestingly, the biofilm phenotype observed in our isolates was not well-correlated with the identification of genes associated with biofilm formation (bap, ica, aap, and agr) in the same isolates. We demonstrated that the carriage of genes was detected in the majority of NAS originating from milk (63.2%) yet that only half of these isolates actually produced biofilm. Piessens et al. [48] also observed that only nine out of the 30 bap-and/or icaA-positive NAS isolates produced biofilm in the phenotypic assay. However, other study [23] reported that a combination of multiple genes, such as icaA-bap-aap and icaA-aap, was associated with a greater ability to form a biofilm. In our study, oddly, none of the NAS isolates possessed bap or ica genes, yet, seven out of nine aap-agr positive S. epidermidis isolates were able to produce biofilm. The failure to Toledo-Silva et al. Vet Res  (2021) 52:114   S. a. + S. c. S. a. + S. e. S. a. + S. s. S. a. ( detect bap or ica genes in the PCR analyses may occur, but does not necessarily imply the inability to form a biofilm by NAS [27,49], as other factors can mediate the biofilm-forming process. Additionally, it was reported [48] that so-called environmental NAS species (only causing IMI sporadically), such as S. epidermidis, S. sciuri, and S. xylosus, are a more significant reservoir of biofilm-related genes than so-called contagious NAS species (primarily causing IMI). Our findings regarding the effect of NAS on the S. aureus biofilm formation and dispersion were in contrast with the observations revealed in a recent study [32]. In the latter study, a reduction in biofilm formation as well as an increase in dispersion of pre-established biofilm of S. aureus by bovine NAS, especially by S. chromogenes and S. simulans isolates, was reported. Unlike our study though, their S. chromogenes isolates were essentially weak-biofilm producers, whereas ours rarely did produce biofilm. Strikingly, the S. chromogenes isolates in our collection were less likely to carry biofilm-related genes. Also, the capacity of bovine NAS isolates to stimulate S. aureus biofilm formation was more extended in NAS isolates that did not carry biofilm-related genes. Although the presence of such genes and the capacity to form biofilm were not well-correlated, we believe this finding explains, at least partially, the S. chromogenes effect and that the discordant results might be attributed to differences in intrinsic traits of the NAS isolates as well as to the combination of S. aureus and NAS species.
Unfortunately, in-depth information regarding the traits of the NAS isolates used was not informed by Goetz et al. [32].
In addition to the effect of S. chromogenes isolates on the S. aureus biofilm formation, the dispersion of the pre-established biofilm of S. aureus by the bovine NAS isolates was suppressed, even more so by TA isolates (including the isolate S. c. 29-a strong in vitro growth inhibitor). Some studies have reported advantages provided by living in a biofilm community, which may result in, for example, decreased antibiotic susceptibility and protection against immune defenses [50,51]. Furthermore, according to our previous findings [19], the same NAS isolates originating from TA, except for one S. epidermidis isolate, demonstrated in vitro S. aureus growth inhibition. However, the capacity of the bovine NAS isolates to inhibit in vitro the growth of S. aureus did not play a role in the biofilm regulation of S. aureus under the conditions tested in our study. Likewise, other study [32] observed no marked growth inhibition when S. aureus isolates were grown with the S. chromogenes and S. simulans isolates that had the strongest biofilm-inhibition activity.
Different interactions between NAS and S. aureus colonizing the bovine mammary gland have been suggested [18,19], including the competitive behavior between NAS and S. aureus in multispecies biofilm communities [32]. In a previous study using the same NAS isolates, we demonstrated that NAS have the capacity to repress  .); B originating from milk or teat apices (TA); C harboring biofilm-related genes (BG) or not (NBG); D with the capacity to produce biofilm themselves (BP) or not (NBP); E with the capacity to inhibit the growth of S. aureus (GI) or not (NGI) and compared to biofilm formed by S. aureus alone (black bars).
the agr system of S. aureus [19]. Furthermore, based on the concept that agr might coordinate both biofilm formation and dispersal [31], we verified for the first time whether bovine NAS isolates originating from milk and TA might affect biofilm formation or dispersion of S. aureus by regulation of the agr system Although it was previously reported that the repression of the agr system would be necessary to form a biofilm, whereas the reactivation of the agr system in established biofilms would trigger its dispersion [31], our NAS isolates not only suppressed biofilm dispersion of the S. aureus strain with a functional agr system (8325-4) but also from the agr mutant strain (8325-4 Δagr). Also, the effect of NAS on the biofilm dispersion of both the parent and mutant S. aureus strain (8235-4 and 8325-4 Δagr) further reinforced the hypothesis that the bovine NAS isolates can coordinate biofilm formation and dispersion of S. aureus by other mechanisms rather than regulation of agr quorum sensing system. Some authors have discussed that the presence or absence of bovine NAS isolates sharing the same niche as S. aureus may play a role in the host colonization [18], whereas others [37,52] observed a robust biofilm formation by NAS plus S. aureus in isolates from pigs, with no evident out-competition of one over the other species when growing in physical contact. According to Otto [53], in fact more than one mechanism could be involved in the biofilm formation and dispersion of staphylococci species, such as the bacterial social behavior (cooperation and/or competition), the molecular mechanism (e.g., bacterial signaling, coaggregation, or co-metabolism) [51], the immune response of the host (bovine mammary gland immune response), and both intrinsic and acquired traits of the isolates involved in the dual-species bacterial biofilms [54]. Nevertheless, as biofilms are extremely hard to eradicate by both the host and by antimicrobial therapies, these dual-species interactions would allow both species to persist in a colonizing state more robustly in one hand, but on the other hand would also provide a constant reservoir for possible S. aureus chronic infections [38]. Still, the reduction of dispersion, as was evident in our results, could potentially prevent S. aureus to spread and colonize other niches.
In conclusion, we demonstrated that S. chromogenes as well as NAS isolates in which no biofilm-related genes were demonstrated, stimulate biofilm formation of S. aureus. As well, bovine NAS isolates were effective in suppressing biofilm dispersion of S. aureus. Interestingly, the habitat of the NAS isolates seems to play an important role in the dispersion of the S. aureus biofilm, suggesting strain differences related to niche adaptation. Importantly, our findings highlight the existing interactions between bovine-related NAS and S. aureus in the biofilm communities and suggest as well that the S. aureus biofilm formation and dispersion, though not through regulation of the agr quorum sensing system, can be affected by (some) NAS (species). The mechanisms that coordinate and determine the biofilm formation and dispersion of S. aureus in the presence of NAS isolates have yet to be identified. Our findings contribute to the knowledge on biofilm formation and dispersion of S. aureus and may lead to new preventive or treatment measures, especially for chronic subclinical mastitis caused by this bacterium in dairy cows.