In vitro concentration-dependent inhibition of early biofilm formation by Staphylococcus aureus isolated from dairy cattle using a bacteriophage cocktail Authors: Edyta Kaczorek-Łukowska, Paweł Foksiński, Natalia Szyryńska, Agnieszka Maszewska, Marta Krzyżaniak Published: 2026 ABSTRACT AbstractBovine mastitis remains one of the most economically important diseases in dairy cattle, with Staphylococcus aureus being a major etiological agent, particularly in chronic and subclinical infections. The ability of S. aureus to form biofilms significantly contributes to antimicrobial tolerance and treatment failure, highlighting the need for alternative or adjunctive therapeutic approaches. Bacteriophage therapy has re-emerged as a promising strategy; however, data on its anti-biofilm efficacy against mastitis-associated S. aureus isolates are limited. In this study, we evaluated the anti-biofilm activity of a newly developed anti–Staphylococcus aureus bacteriophage cocktail against clinical isolates obtained from subclinical bovine mastitis. Twenty-eight non-duplicate field isolates of Staphylococcus aureus obtained from dairy cattle with subclinical mastitis on different farms in north-eastern Poland were examined using a MIC-like assay, followed by biofilm inhibition experiments conducted under high bacterial inoculum conditions. Biofilm formation was assessed after 24 and 48 h using crystal violet staining and confocal laser scanning microscopy with LIVE/DEAD™ fluorescence staining. Ultrastructural changes were analyzed by scanning electron microscopy. MIC-like values did not correspond to concentrations effective against biofilm formation. Biofilm biomass and viability were reduced in a concentration- and time-dependent manner, with the most pronounced effects observed at higher bacteriophage concentrations. Microscopic analyses confirmed biofilm disruption and bacteriophage-induced cellular damage. These findings demonstrate the potential of bacteriophage cocktails as anti-biofilm agents and support further investigation of bacteriophage-based strategies targeting early biofilm development in veterinary staphylococcal infections. This in vitro study was designed to evaluate the concentration-dependent effects of bacteriophages on early biofilm development rather than clinical treatment efficacy.Keywords: Biofilm, Bacteriophages, Staphylococcus aureus, Cattle, Mastitis Introduction Mastitis remains one of the most important diseases affecting dairy cattle worldwide, posing a significant challenge to both milk producers and the veterinary sector. The disease negatively impacts animal welfare and represents a major source of economic loss in dairy herds due to treatment costs, reduced milk yield, decreased milk quality, and premature culling of affected cows [1–3]. Bacterial pathogens are the primary etiological agents of bovine mastitis and consequently antimicrobial therapy remains the most commonly applied treatment approach [4, 5]. Despite its widespread use, antimicrobial treatment frequently fails to achieve complete bacteriological cure. Treatment failure is frequently associated with inappropriate antimicrobial selection, initiation of empirical therapy without prior bacteriological culture and antimicrobial susceptibility testing, insufficient treatment duration, or premature cessation of therapy once clinical signs have resolved. Moreover, the global increase in antimicrobial resistance (AMR) further limits the effectiveness of conventional therapeutic strategies [6]. Incomplete pathogen elimination may lead to the development of chronic mastitis, a long-lasting and often recurrent inflammatory condition characterized by persistently elevated somatic cell counts. Chronic mastitis frequently presents in a subclinical form, which complicates detection and control at the herd level [7]. During the establishment and persistence of chronic intramammary infections, bacteria activate a range of virulence mechanisms that promote survival within the mammary gland. Among the bacterial pathogens associated with bovine mastitis, Staphylococcus aureus is recognized as one of the most important etiological agents, particularly in chronic and subclinical infections [8]. This pathogen is of major relevance in both veterinary and human medicine and is capable of causing infections in companion animals as well as in livestock. In dairy cattle, S. aureus intramammary infections are notoriously difficult to eliminate and are characterized by high persistence within herds and a poor response to antimicrobial therapy [9]. Recent assessments by the European Food Safety Authority (EFSA) have further identified S. aureus as one of the most relevant antimicrobial-resistant bacterial species affecting cattle in the European Union, underscoring its clinical and epidemiological significance [10]. One of the key virulence mechanisms contributing to the persistence of S. aureus infections is biofilm formation. Biofilms represent a critical interface between bacterial virulence and antimicrobial tolerance and substantially contribute to reduced treatment efficacy [11]. It is well established that bacteria embedded within a biofilm structure may exhibit up to a 1,000-fold increased tolerance to antimicrobial agents compared with their planktonic counterparts [11]. Consequently, even antimicrobial therapy selected on the basis of in vitro susceptibility testing may fail to produce the expected clinical and bacteriological outcomes. Currently, there are no distinct therapeutic guidelines addressing the treatment of biofilm-associated intramammary infections, despite increasing evidence that a growing number of mastitis-associated pathogens, including S. aureus, are capable of forming biofilms. This highlights the urgent need for further research into biofilm-targeted therapeutic strategies in the context of bovine mastitis. As mentioned above, antibiotic therapy remains the treatment of choice for bovine mastitis and thus constitutes one of the largest sources of antimicrobial use in the dairy industry [12]. Unfortunately, this contributes to the spread of antimicrobial resistance, which poses a serious threat to both human and animal health. It is estimated that by 2050, AMR will be directly responsible for approximately 1.9 million deaths annually, with the total number of deaths associated with AMR reaching up to 8.2 million per year worldwide in humans [13]. While comparable estimates for animals are currently limited, One Health assessments highlight significant impacts on morbidity, mortality, and production losses in animals due to resistant infections and treatment failures in livestock and food-producing animals. This is even more important given that more than 70% of all antibiotics used globally are used in livestock farming [14]. It is important to emphasize that animal and human health are closely interconnected under the One Health concept, and preventive as well as corrective actions should be implemented simultaneously [15]. These projections indicate that without the introduction of additional effective preventive and therapeutic measures, antimicrobial resistance will remain one of the major global public health threats. Consequently, research groups worldwide are intensively exploring alternative or adjunctive therapeutic approaches to antibiotic therapy [16]. One such approach is bacteriophage therapy. Bacteriophages are viruses that infect bacteria and lack the ability to infect eukaryotic cells, which represents a major advantage of this strategy. Their primary mode of action is lytic activity, typically restricted to specific bacterial strains, resulting in high specificity. Bacteriophages are ubiquitous throughout the biosphere, and their abundance is estimated to be approximately ten-fold higher than that of bacteria [7, 17, 18]. Although bacteriophages were first discovered as early as 1915 by William Twort, and their bactericidal properties were demonstrated in 1917 by Félix d’Hérelle, their use was largely abandoned in many regions, including Western Europe, following the introduction of antibiotics [17, 19, 20]. Currently, due to the growing AMR crisis, bacteriophages are once again being considered as an alternative or supportive therapy, not only in human medicine but increasingly also in veterinary applications [7, 17, 21]. Taken together, the widespread ability of S. aureus to form biofilms, combined with the limited effectiveness of antibiotic therapy in bovine mastitis, highlights a critical therapeutic gap. Although bacteriophage therapy has gained renewed attention, data regarding its concentration-dependent antibiofilm activity against bovine mastitis-associated S. aureus isolates remain limited. Therefore, the aim of this study was to evaluate the efficacy of a newly developed bacteriophage cocktail against biofilm formation by S. aureus isolates obtained from subclinical bovine mastitis using crystal violet assays, fluorescence microscopy, and scanning electron microscopy under different incubation times and phage concentrations. We hypothesized that higher bacteriophage concentrations would result in greater inhibition of biofilm development. Materials and methods Strains Twenty-eight non-duplicate Staphylococcus aureus strains isolated from milk samples collected from dairy cattle with subclinical mastitis on different farms located in north-eastern Poland (Warmia and Mazury region) between 2016 and 2019 were included in this study. Only one isolate per farm was selected to avoid epidemiologically related duplicates. Subclinical mastitis cases were defined as the absence of visible clinical signs with a somatic cell count exceeding 400,000 cells/mL. The strains had been previously characterised regarding virulence genes and antimicrobial resistance profiles by Kaczorek-Łukowska et al. (2022). Bacteriophage cocktail In this experiment, an anti-Staphylococcus aureus bacteriophage cocktail with a titer of 3.33 × 108 PFU/mL was used. The cocktail was developed within the framework of the project POIR.01.01.01-00-0016/16 by Proteon Pharmaceuticals and was previously characterized in the study by [7]. To prepare phage cocktail three anti-Staphylococcus aureus bacteriophages were propagated separately, and following downstream processing, the individual phages were combined at equal titers. The cocktail was subsequently evaluated by phage enumeration using the double agar overlay plaque assay and was further assessed for microbiological sterility. Minimum inhibitory concentration-like values Minimum inhibitory concentrations (MICs) were established using the broth microdilution technique following the guidelines of the Clinical and Laboratory Standards Institute (CLSI). Assays were conducted in 96-well microtiter plates containing cation-adjusted Mueller–Hinton broth (CAMHB). Each bacterial strain was inoculated to obtain a final density of 5 × 105 CFU/mL, in accordance with CLSI recommendations [22]. The anti-Staphylococcus aureus bacteriophage cocktail (stock concentration 3.33 × 108 PFU/mL) was tested across a series of two-fold dilutions ranging from 50% to 0.78% of the stock preparation. Plates were incubated at 35 ± 2 °C for 16–20 h. The MIC-like values was defined as the lowest bacteriophage concentration at which visible bacterial growth was inhibited, as assessed by visual inspection without optical aids. Biofilm-forming assay using crystal violet staining This experiment was carried out using flat-bottomed 96-well microtiter plates and the crystal violet assay as described by [23], with minor modifications. The concentration of bacterial cells in the analysed samples was 5.6 × 107 cells/mL. The bacterial growth medium used in this study was tryptic soy broth (TSB; Oxoid, Basingstoke, UK) supplemented with 1% (w/v) dehydrated glucose (Pol-Aura, Poland). In the test groups, bacterial suspensions were mixed with the bacteriophage cocktail (stock concentration 3.33 × 108 PFU/mL) at final proportions corresponding to 50%, 25%, 12.5%, and 6.25% of the tested preparation, resulting in final phage concentrations in the wells of 1.67 × 108, 8.33 × 107, 4.16 × 107, and 2.08 × 107 PFU/mL, respectively. In the control group, bacterial suspensions were mixed with TSB medium and sterile phosphate-buffered saline (PBS) instead of the bacteriophage preparation. Wells containing TSB medium without bacterial inoculation served as the negative control. Following incubation under aerobic conditions at 37 °C for 24 and 48 h to allow biofilm development, the wells were gently washed and stained with a 0.1% (w/v) aqueous crystal violet solution (Merck, Germany). After excess stain was removed, the crystal violet retained by the biofilm matrix was solubilized using 200 μL of an 80:20 (v/v) mixture of ethyl alcohol and acetone. The optical density of the resulting solution was subsequently determined at a wavelength of 595 nm using an ELISA plate reader (Sunrise Absorbance Reader, Tecan, Austria). Each S. aureus strain was analysed in eight technical replicates per plate, and the entire experiment was performed in triplicate. Confocal laser scanning microscopy (CLSM) Biofilm formation by S. aureus was evaluated after 24 h and 48 h of incubation using Live/Dead fluorescence staining by quantifying live cells, dead cells, and the total number of bacterial cells. For this experiment, two strains (ID 556 and 397) were selected because they exhibited a weaker response to lower bacteriophage concentrations in the initial screening assays. These strains were therefore considered more challenging models for evaluating the antibiofilm efficacy of the bacteriophage cocktail under microscopy-based conditions. This experiment was conducted using glass-bottomed 24-well plates. S. aureus suspensions were standardized to a concentration of 4.5 × 108 cells/mL, and 25 µL of the standardized suspension was added to 975 µL of tryptic soy broth (TSB; Oxoid, Basingstoke, UK) supplemented with 1% glucose, resulting in a final bacterial concentration of 1.125 × 107 cells/mL per well. To assess biofilm formation under different experimental conditions, cultures were incubated under aerobic conditions at 37 °C for 24 and 48 h. In the test groups, bacterial suspensions were mixed with an anti-Staphylococcus aureus bacteriophage cocktail (stock concentration 3.33 × 108 PFU/mL) at final proportions corresponding to 50%, 25%, 12.5%, and 6.25% of the tested preparation, yielding final bacteriophage concentrations in the wells of 1.67 × 108, 8.33 × 107, 4.16 × 107, and 2.08 × 107 PFU/mL, respectively. In the control groups, bacterial suspensions were mixed with TSB medium and sterile phosphate-buffered saline (PBS) instead of the bacteriophage cocktail. Additionally, a positive control consisted of bacterial suspensions mixed with PBS and TSB supplemented with 1% glucose at a 50:50 (v/v) ratio, without bacteriophage addition. Biofilm formation was evaluated after each incubation period to determine the time- and concentration-dependent effects of the bacteriophage cocktail on S. aureus biofilm development. Fluorescence-based assessment of S. aureus biofilms was performed using the FilmTracer™ LIVE/DEAD™ Biofilm Viability Kit (Invitrogen, Carlsbad, CA, USA) in accordance with the manufacturer’s instructions. Following the staining procedure, each well was gently rinsed three times with ultrapure water to remove non-adherent cells and excess staining solution. Subsequently, the biofilms were visualized using a Zeiss LSM 900 confocal laser scanning microscope (Carl Zeiss, Oberkochen, Germany) equipped with an Airyscan 2 detector and four laser lines (405, 488, 561, and 640 nm), enabling high-resolution fluorescence imaging (approximately 120–140 nm lateral resolution). Image acquisition was carried out using Zeiss ZEN 3.7 software (Carl Zeiss, Oberkochen, Germany), and image processing and quantitative analyses were performed using ImageJ software (National Institutes of Health, USA). Scanning electron microscopy (SEM) For this experiment, two strains (ID 556 and 397) were selected because they exhibited a weaker response to lower bacteriophage concentrations in the initial screening assays. These strains were therefore considered more challenging models for evaluating the antibiofilm efficacy of the bacteriophage cocktail under microscopy-based conditions. The experiment was performed using glass coverslips in 6-well plates. The number of bacterial cells was standardized to 4.5 × 108 cells/mL, then 100 µl of this suspension was added to 3900 µl of TSB broth (Oxoid, Basingstoke, Great Britain) supplemented with 1% glucose and with the addition of the bacteriophage cocktail at a concentration of 1.67 × 108 PFU/mL in each well. Each assay was performed with a control sample (untreated strain in TSB broth with 1% glucose with PBS) and a negative sample (pure broth). The samples were incubated at 37 °C under aerobic conditions for 24 h. Then, the bacterial biofilms on coverslips were immersion-fixed in a mixture of 2.5% glutaraldehyde and 1% paraformaldehyde in 0.2 M phosphate buffer (pH 7.4) for 2 h at 4 °C. They were washed three times in 0.2 M phosphate buffer (pH 7.4), dehydrated in ethanol (30, 50, 70, 90, 95, 100%, 15 min each), dried in a critical point dryer (Autosamdri®−931, Tousimis, Maryland, USA) and chromium-coated (10 nm) using a sputter coater (Quorum Q150T ES, Quorum Technologies, Laughton, UK). Thereafter, the biofilms were imaged under a field-emission scanning electron microscope (Gemini 450, Carl Zeiss, Oberkochen, Germany) at an acceleration voltage of 1.3 kV using an Everhart–Thornley SE detector. Rationale for assay-specific bacterial inoculum selection Different bacterial inocula were intentionally selected according to the objective and technical requirements of each assay. In the MIC-like assay, the bacterial density (5 × 105 CFU/mL) was selected in accordance with CLSI recommendations for planktonic susceptibility testing. In contrast, for the crystal violet biofilm assay, a higher inoculum was used to ensure robust and reproducible biofilm formation under conditions reflecting a high bacterial burden and to allow detection of concentration-dependent trends in biomass reduction. For microscopy-based analyses (CLSM and SEM), the final bacterial concentration was adjusted to 1.125 × 107 cells/mL, which provided improved visualization of biofilm architecture and more reliable image acquisition. In particular, these conditions allowed clearer fluorescence-based assessment in CLSM and reduced technical artefacts during SEM sample preparation and imaging. Since no standardized guidelines currently define bacterial inoculum levels for phage-based biofilm assays, the inocula used in this study were selected according to the biological endpoint and technical performance of each method. Statistical analysis Data are presented as mean values ± standard deviation (SD). In the crystal violet assay, each strain was analyzed using eight technical replicates within a single experiment, and the entire experiment was independently repeated three times on separate occasions (biological replicates). For statistical analysis, technical replicates were averaged within each independent experiment, and the resulting three biological replicate means were used for final comparisons. The subgroup analysis of the five strong biofilm-producing strains was based on the same pooled dataset and therefore also represents mean values derived from three independent biological experiments. For fluorescence-based biofilm analyses, two selected strains were examined in three independent biological experiments, and the mean values from these replicates were used for statistical comparisons. Scanning electron microscopy (SEM) analyses were performed on the same two selected strains as representative qualitative imaging experiments. Statistical comparisons between groups were performed using the Kruskal–Walli’s test followed by Dunn’s post hoc test with Bonferroni correction. All statistical analyses were performed using GraphPad Prism (GraphPad Software, San Diego, CA, USA). Results Minimum Inhibitory concentrations- like value The MIC-like value analysis demonstrated that the majority of the tested S. aureus strains (25/28, 89.3%) exhibited MIC-like values below 0.78% of the bacteriophage cocktail stock concentration, corresponding to < 2.6 × 106 PFU/mL. In contrast, three strains (10.7%) showed an MIC-like value of 0.78% of the stock preparation, corresponding to 2.6 × 106 PFU/mL. Biofilm-forming assay using crystal violet staining As shown in Fig. 1, treatment with the bacteriophage cocktail resulted in a statistically significant reduction of S. aureus biofilm biomass; however, the magnitude of this effect depended on both the incubation time and the applied bacteriophage concentration. After 24 h of incubation, analysis including all tested strains (n = 28) demonstrated a statistically significant reduction in biofilm biomass at all tested treatment levels compared with the untreated control group. The most pronounced inhibition was observed at the highest phage doses. Biofilm biomass was reduced by approximately 85% and 83% following treatment with the 50% and 25% cocktail concentrations, respectively. Lower doses also produced significant reductions, corresponding to approximately 65% and 73% decreases in biomass at the 12.5% and 6.25% treatment levels, respectively (Fig. 1A). Biofilm formation by Staphylococcus aureus strains after 24 h (A, B) and 48 h (C, D) of incubation in the presence of selected concentrations (6.25–50%) of the bacteriophage cocktail. C indicates the untreated control. Panels A and C present results for all tested strains (n = 28), whereas panels B and D show the subgroup analysis of five strains (n = 5) exhibiting the lowest susceptibility to the lowest bacteriophage concentrations. Data are presented as mean OD values ± SD from three independent biological experiments. Within each experiment, eight technical replicates were performed for each strain and averaged prior to statistical analysis. Asterisks indicate statistically significant differences compared with the control (*p < 0.05, **p < 0.01, ***p < 0.001) When the analysis was restricted to the five strong biofilm-producing strains after 24 h, only the two highest treatment levels remained significantly effective. Exposure to the 50% and 25% bacteriophage concentrations reduced biofilm biomass by approximately 87% and 84%, respectively. In contrast, the lower concentrations (12.5% and 6.25%) resulted in more limited reductions of 33% and 50%, respectively, and did not differ significantly from the untreated control group, indicating a concentration-dependent loss of efficacy in strains with stronger biofilm-forming capacity (Fig. 1B). After 48 h of incubation, a similar concentration-dependent pattern was observed. In the analysis including all tested strains (n = 28), biofilm biomass was significantly reduced at all bacteriophage treatment levels compared with the untreated control group. The strongest inhibition was again noted at the highest concentrations, with reductions of approximately 82% and 69% following treatment with the 50% and 25% bacteriophage cocktail, respectively. Lower concentrations (12.5% and 6.25%) also remained significantly effective, reducing biofilm biomass by approximately 55% and 56%, respectively, although these values indicated partial biofilm recovery over prolonged incubation (Fig. 1C). The analysis of the five strong biofilm-producing strains after 48 h revealed that only the highest bacteriophage concentration (50%) maintained a statistically significant inhibitory effect, corresponding to an approximately 81% reduction in biofilm biomass. Lower concentrations (25%, 12.5%, and 6.25%) resulted in reductions of 49%, 30%, and 16%, respectively, but were not sufficiently effective to achieve statistically significant differences compared with the untreated control group, confirming reduced susceptibility among strains with stronger biofilm-forming capacity (Fig. 1D). CLSM analysis of biofilm formation After 24 h of incubation, statistically significant differences in biofilm composition were observed between experimental groups; however, the effects were dependent on the bacteriophage concentration (Fig. 2A-C). The 6.25% bacteriophage concentration did not result in statistically significant differences compared with the control for any of the analyzed parameters. Analysis of live bacterial cells revealed a statistically significant reduction only at the 50% bacteriophage concentration compared with the control. Additionally, significant differences were observed between the 6.25% concentration and higher bacteriophage concentrations (25% and 50%), with both higher concentrations leading to a significant reduction in the number of live bacterial cells relative to the 6.25% treatment (Fig. 2A). For dead bacterial cells, statistically significant reductions compared with the control were observed at bacteriophage concentrations of 12.5% and 25%. No statistically significant difference was detected between the control and the 50% concentration (Fig. 2B). Importantly, the lower dead-cell signal observed at selected higher bacteriophage concentrations should be interpreted together with the reduction in total biofilm biomass, indicating fewer cells remaining within the biofilm structure rather than reduced phage activity. Evaluation of the total number of bacterial cells demonstrated a statistically significant reduction in biofilm biomass at bacteriophage concentrations of 25% and 50% compared with the control. Furthermore, the 50% concentration differed significantly from the 6.25% concentration, indicating a concentration-dependent effect on overall biofilm cell numbers, as shown in Fig. 2C. Confocal laser scanning microscopy (CLSM) analysis of biofilm formation by Staphylococcus aureus (n = 2) strains after 24 h (A–C) and 48 h (D–F) of incubation in the presence of selected concentrations of the bacteriophage cocktail. Panels A–C show biofilm development after 24 h exposure to phages, whereas panels D–F present biofilms formed after 48 h. The untreated control consisted of TSB supplemented with glucose and PBS. Biofilms were stained using the Live/Dead viability assay. Green fluorescence represents viable bacterial cells, red fluorescence indicates dead or membrane-compromised cells, and blue fluorescence corresponds to the total bacterial cell population. Data are expressed as mean fluorescence-derived values ± SD from three independent biological experiments. Asterisks indicate statistically significant differences compared with the control (*p < 0.05, **p < 0.01, ***p < 0.001) After 48 h of incubation, the bacteriophage cocktail exhibited a pronounced concentration-dependent effect on biofilm formation (Fig. 2D-F). As observed at 24 h, the 6.25% concentration did not differ significantly from the control for any of the analyzed parameters. Analysis of live bacterial cells showed a statistically significant reduction at bacteriophage concentrations of 25% and 50% compared with both the control and lower concentrations (6.25% and 12.5%). The 12.5% concentration maintained a relatively high number of live cells and differed significantly from the higher bacteriophage concentrations, indicating limited inhibitory activity at this concentration (Fig. 2D). In the case of dead bacterial cells, statistically significant differences were observed between the control and higher bacteriophage concentrations (25% and 50%), with both concentrations resulting in markedly lower numbers of dead cells. Additionally, the 12.5% concentration differed significantly from 25% and 50%, further supporting a distinct response at higher bacteriophage doses (Fig. 2E). Similarly to the 24 h time point, this reduction in dead-cell signal at higher concentrations should be interpreted together with the decrease in total biofilm biomass, suggesting substantial biofilm disruption and fewer retained cells within the analyzed structure rather than reduced bactericidal activity. Assessment of the total number of bacterial cells revealed a statistically significant reduction in biofilm biomass at bacteriophage concentrations of 25% and 50% compared with the control. Moreover, the 12.5% concentration differed significantly from the higher bacteriophage concentrations, confirming a clear concentration-dependent inhibition of mature biofilm development after prolonged incubation (Fig. 2F). Overall, bacteriophage concentrations of 25% and 50% consistently reduced biofilm formation at both 24 h and 48 h, with the strongest and most reproducible effects observed at the 50% concentration. Lower concentrations (6.25% and 12.5%) were insufficient to significantly inhibit biofilm development, particularly after prolonged incubation. Reductions in both live- and dead-cell fluorescence signals at higher concentrations were associated with decreased total biofilm biomass, supporting the conclusion that the bacteriophage cocktail primarily disrupted biofilm structure and reduced bacterial retention within the biofilm matrix. These results indicate a clear concentration-dependent anti-biofilm activity of the bacteriophage cocktail against S. aureus, which becomes more pronounced during biofilm maturation. Representative CLSM images illustrating these effects are shown in Figs. 3 and 4. Representative confocal laser scanning microscopy images of Staphylococcus aureus biofilms formed after 24 h and stained with the LIVE/DEAD™ Biofilm Viability Kit. Green fluorescence indicates viable bacterial cells, whereas red fluorescence indicates membrane-compromised/dead cells; merged fluorescence channels are shown. Images are presented as follows: A untreated control (without bacteriophages), B 6.25%, C 12.5%, D 25%, and E 50% bacteriophage cocktail. Scale bars = 20 µm Representative confocal laser scanning microscopy images of Staphylococcus aureus biofilms formed after 48 h and stained with the LIVE/DEAD™ Biofilm Viability Kit. Green fluorescence indicates viable bacterial cells, whereas red fluorescence indicates membrane-compromised/dead cells; merged fluorescence channels are shown. Images are presented as follows: A untreated control (without bacteriophages), B 6.25%, C 12.5%, D 25%, and E 50% bacteriophage cocktail. Scale bars = 20 µm SEM analysis Two S. aureus strains and a 50% concentration of the bacteriophage cocktail were selected for ultrastructural analysis using scanning electron microscopy. After 24 h of incubation, control samples exhibited dense, well-organized biofilm structures composed of tightly packed coccoid cells embedded within an extensive extracellular matrix. Numerous multilayered bacterial aggregates and microcolonies were visible, indicating mature biofilm formation as shown in Fig. 5A, B. In contrast, bacteriophage-treated samples showed a profound disruption of biofilm architecture. The biofilm structure was largely absent, and only sparse, irregularly distributed bacterial cells were observed. Many cells exhibited morphological alterations, including deformed shapes and disrupted cell surfaces, consistent with bacteriophage-induced lysis. Residual extracellular matrix fragments and scattered cellular debris were occasionally detected, suggesting biofilm destabilization rather than simple growth inhibition (Fig. 5C, D). Scanning electron microscopy (SEM) analysis of biofilm structure formed by Staphylococcus aureus strains (n = 2) after 24 h of incubation. Biofilms were developed under control conditions (TSB supplemented with glucose and PBS) and in the presence of 50% bacteriophage cocktail. Control samples showed dense multilayer biofilm structures, whereas bacteriophage-treated samples exhibited reduced surface coverage and disrupted bacterial aggregates Discussion While the National Institutes of Health (NIH) estimate that approximately 65% of microbial infections and up to 80% of chronic infections in humans are associated with biofilm formation, comparable quantitative epidemiological data are lacking in veterinary medicine; nevertheless, available evidence suggests that a similar biofilm-associated pattern likely occurs in animals [24]. Biofilm-forming bacteria exhibit marked tolerance to multiple antibiotics and evade host immune responses, frequently leading to therapeutic failure [25]. In view of the substantial challenges associated with the treatment and control of biofilm-related infections, there is a clear unmet need not only for the development of non-antibiotic or adjunctive therapeutic strategies aimed at preventing or disrupting biofilm formation, but also for a shift in current diagnostic and therapeutic recommendations. Existing guidelines for susceptibility testing and antimicrobial therapy are largely based on planktonic bacteria or on concentrations that do not reflect the high cell density and structural complexity of biofilms. Once established, biofilms significantly increase tolerance to antimicrobial agents and facilitate intercellular communication. This enables the activation and dissemination of collective defence mechanisms, which can make even targeted therapies ineffective. According to the European Medicines Agency [17], bacteria may develop resistance to bacteriophages if phages are used incorrectly or at insufficient doses. Therefore, both conventional and emerging alternative therapeutic approaches should explicitly account for the biofilm mode of growth to improve clinical efficacy [26, 27]. As mentioned earlier, bacteriophages have been known for many years; however, in many countries they were largely abandoned in favor of antibiotics. After decades of excessive use of these active substances and the emergence of a global problem—antimicrobial resistance affecting both humans and animals—there has been a renewed interest in these viruses due to their therapeutic potential against bacterial pathogens. Currently, an increasing number of research groups are working on optimizing bacteriophage-based therapies for both human and veterinary medicine [28]. In the present study, we investigated the effects of a newly developed anti–Staphylococcus aureus bacteriophage cocktail on biofilm formation by clinical field isolates. Initially, a MIC-like value was determined in accordance with CLSI recommendations [22]. Subsequently, selected phage concentrations were evaluated for their ability to inhibit biofilm formation after 24 and 48 h of incubation. Our results demonstrated that MIC-like values obtained under planktonic conditions were poor predictors of concentrations effective against biofilm formation. This observation is biologically expected, as bacteria growing within biofilms exhibit substantially greater tolerance than their planktonic counterparts due to high cell density, extracellular matrix protection, and metabolic heterogeneity. These findings highlight the limitations of standard MIC testing when evaluating therapies intended for biofilm-associated infections and support the need for biofilm-oriented parameters such as minimum biofilm inhibitory concentration (MBIC). This discrepancy may be attributed, at least in part, to differences in bacterial density. In the biofilm assays, a substantially higher bacterial inoculum was used compared with MIC-like value testing. However, these higher bacterial concentrations more closely reflect natural infection conditions and were intentionally applied to assess the performance of the bacteriophage cocktail under particularly challenging conditions. Unlike antimicrobial susceptibility testing, there are currently no standardized guidelines defining bacterial inoculum levels for biofilm-related assays. In our experiments, a bacterial concentration of 5.6 × 107 cells/mL of field S. aureus isolates (n = 28) was used, while bacteriophages were tested at concentrations of 1.67 × 108, 8.33 × 107, 4.16 × 107, and 2.08 × 107 PFU/mL. These concentrations were markedly higher than the MIC-like value (< 2.6 × 106 PFU/mL). Crystal violet staining showed that after 24 h, the two highest phage concentrations were most effective at reducing biofilm formation. After 48 h, however, only the highest concentration maintained significant activity. Although statistically significant inhibition of biofilm formation was observed for all phage concentrations tested when the entire strain collection was considered, limiting the collection to five strong biofilm-producers (17.9% (5/28) of the tested pool) showed a markedly increased tolerance to bacteriophages, leading to a reduction in the number of concentrations showing significant effects. Similar results were obtained in the fluorescence-based analyses, where the most pronounced effects were observed at the two highest bacteriophage concentrations. Furthermore, scanning electron microscopy confirmed the high efficacy of the highest concentration in preventing biofilm formation by the tested S. aureus strain. These observations are consistent with previous reports showing that low phage concentrations were insufficient to achieve a significant reduction in bacterial counts [29]. The authors suggested that the low abundance of Salmonella in the chicken gut, combined with the complexity of the intestinal environment—including physicochemical factors such as viscosity—was not conducive to efficient phage amplification. Biofilm-associated growth of S. aureus may substantially reduce susceptibility to both conventional antimicrobials and bacteriophages through several complementary mechanisms. In the case of antibiotics, the extracellular biofilm matrix can limit antimicrobial diffusion, while high cell density, altered gene expression, metabolic heterogeneity, and the presence of slow-growing or persister-like cells further decrease treatment efficacy. In addition, biofilm conditions may promote stress-adaptive responses and facilitate the survival of subpopulations with reduced susceptibility. Similar limitations may also apply to bacteriophage therapy. Because successful phage infection depends on access to bacterial surface receptors, the biofilm matrix may physically hinder phage penetration and reduce contact between phage particles and bacterial cells [30]. In S. aureus, wall teichoic acids (WTA) are considered major phage receptors and are also closely linked to susceptibility to cell wall-targeting antibiotics. Therefore, loss of WTA or alterations in its biosynthesis or glycosylation pattern may simultaneously affect both antibiotic efficacy and phage adsorption efficiency. Moreover, recent studies have shown that the agr quorum-sensing system regulates the α/β-GlcNAc glycosylation pattern of WTA by suppressing tarM expression during stationary phase, thereby influencing phage susceptibility. Since quorum sensing plays a central role in biofilm maturation and regulation, these findings suggest that biofilm-associated quorum-sensing activity may additionally modulate bacteriophage sensitivity through changes in WTA structure [31–33]. Likewise, mutations affecting cell wall architecture, such as those involving femA, may also reduce phage susceptibility. femA is involved in pentaglycine cross-bridge formation within peptidoglycan, and its disruption has been associated with altered cell wall organization, reduced cross-linking, and localized thickening, which may hinder phage penetration and subsequent infection [34, 35]. In addition, these bacteria possess a wide range of virulence factors and a remarkable ability to acquire new resistance mechanisms, enabling rapid adaptation to hostile environmental conditions such as antimicrobial exposure, immune pressure, and nutrient limitation. This high adaptive potential further contributes to persistence of infection, treatment failure, and the difficulty of developing consistently effective therapeutic strategies against this pathogen. Therefore, in biofilm-forming S. aureus, reduced susceptibility to phages may result not only from restricted phage diffusion within the biofilm matrix, but also from structural adaptations of the bacterial cell envelope that impair phage–host interactions. Taken together, these mechanisms further support the need to evaluate phage-based therapies under biofilm-relevant conditions rather than relying exclusively on planktonic susceptibility testing. Although increasing phage titers in the applied preparations improved antibacterial efficacy, complete bacterial eradication was not achieved. According to the European Medicines Agency [17], bacteria may acquire resistance to bacteriophages if they are used incorrectly or at too low a dose. This is a significant risk that must be considered when evaluating the safety and effectiveness of phage-based therapeutic products. Accordingly, the developed anti-Staphylococcus aureus phage cocktail was intentionally formulated to achieve the highest possible efficacy while limiting the emergence of bacterial resistance—one of the most common challenges associated with antibiotic therapies—as previously described by [7]. In S. aureus, phage resistance is most commonly associated with mutations affecting cell surface structures involved in phage adsorption, particularly wall teichoic acids, which serve as primary phage receptors [36, 37]. However, resistance may also arise from mutations in genes indirectly involved in cell wall integrity, such as femA, leading to altered peptidoglycan architecture and impaired phage attachment [38]. Consequently, our findings support the use of sufficiently high bacteriophage cocktail concentrations—within established safety limits for the host—in potential clinical applications, in order to minimize bacterial survival and reduce the likelihood of resistance development. This consideration is particularly important in the context of commercial applications in animal populations, such as dairy cattle, where an individualized therapeutic approach is neither economically nor logistically feasible. In such settings, the application of suboptimal bacteriophage concentrations may not only fail to eliminate target bacteria but may also induce stress responses that activate multiple bacterial defense strategies. Under stress conditions, S. aureus can produce and release masking molecules, including lipoproteins capable of binding phage receptors, either actively or as a consequence of bacterial lysis [39]. Additionally, the formation of outer membrane vesicles has been shown to act as a phage decoy mechanism, adsorbing and inactivating phage particles before they reach the bacterial cell surface [40]. Stress exposure may further promote increased production of extracellular matrix components—composed primarily of polysaccharides, proteins, lipids, and extracellular DNA—leading to biofilm formation that physically limits phage access to their receptors. Collectively, these mechanisms can substantially reduce phage efficacy and facilitate bacterial survival, thereby increasing the likelihood of selecting for phage-resistant S. aureus populations when insufficient phage concentrations are applied [39]. Biofilms represent not only a physical barrier limiting bacteriophage penetration but also a dynamic selective environment that promotes the emergence of phage-resistant subpopulations. High cell density, metabolic heterogeneity, and stress-induced defense mechanisms collectively create conditions favorable for the selection of resistant phenotypes, particularly when bacteriophages are applied at sublethal concentrations. Although these experiments were conducted under in vitro conditions, we believe they provide novel and valuable insights into the antibiofilm potential of the bacteriophage cocktail against Staphylococcus aureus isolates associated with bovine mastitis. Nevertheless, the present findings also raise several important questions that should be addressed in future studies. Further investigations should include models that better reflect mammary gland-like conditions, evaluation of the efficacy of the bacteriophage cocktail against mature pre-established biofilms, assessment of repeated-dose treatment regimens, and monitoring of the potential emergence of bacteriophage resistance among tested isolates. It should also be acknowledged that different bacterial inocula were used across the applied assays, resulting in varying effective multiplicities of infection (MOI), which may limit direct quantitative comparison between methods. However, each assay was intentionally optimized for a distinct experimental endpoint, including planktonic growth inhibition, quantitative biofilm biomass assessment, or microscopy-based structural evaluation. Another limitation of the present study is that microscopy-based analyses were restricted to two selected isolates with lower susceptibility to bacteriophage treatment. Although this approach provided a stringent model for efficacy testing, the observed responses cannot be considered fully representative of all tested strains. Therefore, while our study has inherent limitations, it represents a strong and informative foundation for more advanced analyses and future translational research. Although the present findings are promising, translation into clinical practice requires consideration of several factors, including phage stability in milk, intramammary delivery efficiency, interactions with host immunity, strain variability among herds, repeated dosing strategies, and regulatory approval. Moreover, efficacy against mature biofilms and mixed-species intramammary infections should be evaluated under conditions more closely resembling the in vivo mammary environment. Based on our observations, the use of sufficiently high bacteriophage concentrations may be preferable in therapeutic settings, provided that patient safety is ensured through appropriate toxicological evaluation, including cytotoxicity and genotoxicity assessments. Such an approach may help maximize antibacterial efficacy and potentially reduce the risk of selecting phage-resistant bacterial subpopulations, which is particularly relevant in clinical applications at the herd level where transmission dynamics may facilitate dissemination of resistant strains. Conclusion Importantly, the present study primarily addresses the ability of the bacteriophage cocktail to interfere with early stages of biofilm formation rather than to eradicate fully mature biofilms. This distinction is clinically relevant, as early intervention during subclinical mastitis may prevent the establishment of highly structured and highly tolerant biofilms. Accordingly, the observed concentration-dependent effects should be interpreted mainly in the context of biofilm prevention or early disruption rather than complete eradication of established biofilms. From a practical perspective, repeated or prolonged exposure of bacterial populations to suboptimal bacteriophage concentrations may increase selective pressure and favor the emergence of phage-resistant variants. Therefore, the definition of minimal effective phage concentrations and optimized dosing regimens is essential for herd-level applications. Analogous to antibiotic therapy, repeated administration of bacteriophages may be required to maintain effective antibacterial pressure over time, and this aspect should be specifically addressed in future studies. Although all experiments were conducted under in vitro conditions, the results demonstrate pronounced, concentration-dependent anti-biofilm activity of the bacteriophage cocktail against clinical S. aureus isolates. These findings support the potential of bacteriophage-based interventions as preventive or adjunctive strategies for controlling biofilm-associated subclinical mastitis in dairy cattle and highlight the need for further studies evaluating repeated dosing and efficacy against mature biofilms under conditions more closely resembling the in vivo mammary environment. Acknowledgements Not applicable. Authors’ contributions E.K.: Conceptualization; Methodology; Investigation; Formal analysis; Data curation; Validation; Writing—original draft; Writing—review and editing; Project administration. P.F.: Investigation; Methodology. A.M.: Phage preparation; Writing—review and editing. M.K.: Phage preparation; Writing—review and editing. N.S.: Methodology; Validation of microscopic methods; Writing—review and editing. All authors have read and approved the final manuscript. Funding Funded by the Minister of Science under “the Regional Initiative of Excellence Program”. Data availability The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request. Declarations Ethics approval and consent to participate Milk samples were collected as part of routine veterinary diagnostics/herd health monitoring (no experimental procedures were performed on animals). According to the Polish law Act of 15 January 2015 on the protection of animals used for scientific or educational purposes, ethical approval was not required. Samples were residual diagnostic material and were analyzed anonymously; no identifiable farm-level data were collected. Consent for publication Not applicable. Competing interests The authors declare no competing interests. Footnotes Publisher’s Note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.