The microorganism attracts special attention due to its resistance to a broad spectrum of antibiotics, and the most important are methicillin-resistant S. aureus strains (MRSA).

Treatment of staphylococcal infections is hampered not only by antibiotic resistance, but also by plenty of virulence factors, among which is the ability of staphylococci to form biofilms. Like biofilms of other bacteria, the S. aureus biofilm consists of two main components: water (about 97%) and organic matter represented by extracellular polymers and colonies of microorganisms. Extracellular polymers constitute 50–90% of all organic mass of the biofilm and include a combination of various compounds, such as extracellular DNA (eDNA), proteins, and polysaccharides. The remainder of biofilm is represented by bacterial cells [2, 3].

The biofilm ensures firm attachment of bacteria to various surfaces, including tissues of the body and medical equipment [2]. Furthermore, biofilms considerably increase resistance of bacteria to the immune system factors and antimicrobial drugs: the minimum inhibitory concentrations (MICs) of antibiotics necessary for disruption of bacteria in biofilms can 1000-fold exceed the concentrations that are enough to destroy planktonic cells [3, 4].

Bacteriophages increasingly considered as a promising remedy to treat bacterial infections caused by antibioticresistant strains. Only virulent bacteriophages that realize the lytic cycle only are used in medical practice. Such phages capable of infecting and destroying the S. aureus cells include members of the families Herelleviridae and Rountreeviridae. These demonstrate high efficacy in both in vitro experiments and in vivo models, are successfully used for therapy, which ensures reduction of bacterial load and improvement of clinical outcomes [5].

In this context, of special interest are the studies focused on assessing the effects of phages on biofilms. It has been shown that some S. aureus phages can effectively reduce the biofilm biomass [6, 7]. Nevertheless, a number of studies show that biofilm generation can be stimulated under exposure to phages, which can be associated with the features of interplay between phages and bacterial cells [8, 9]. Such effects are likely to depend on a number of factors: bacterial strain itself, bacteriophage used and its concentration, physiological state of cells, as well as morphological and structural characteristics of the biofilm.

The study aimed to assess the effects of the lytic bacteriophages vB_SauM-515A1 (family Herelleviridae) and vB_SauP-436A (family Rountreeviridae) on biofilms formed by the S. aureus clinical isolates.

METHODS

Bacterial strains and bacteriophages

In this study, we used 20 S. aureus strains, most different based on the origin and isolation locus, from the collection of the Lopukhin Federal Research and Clinical Center of PhysicalChemical Medicine of the Federal Medical Biological Agency. The strains were collected in 2015–2020 from hospitals in various regions of Russia.

Bacterial culture was grown in the liquid LB (lysogeny broth) medium (Oxoid; UK) or Miller LB agar (Oxoid) at 37 °C for 18 h. Cells grown in the liquid medium were used to produce overnight cultures, which were later used for inoculation in both experiments on determining antibiotic resistance and all the tests related to assessment of biofilm formation and the effects of antibacterial agents on biofilm formation. Bacteria were cultured in LB agar in order to maintain the bacterial culture, as well as to enumerate cells in the experiments on assessing the effects of bacteriophages on the biofilms formed.

Sensitivity to oxacillin (beta-lactam antibiotic), vancomycin (glycopeptide), gentamicin (aminoglycoside), tetracycline (tetracycline antibiotic), levofloxacin (fluoroquinolone), and erythromycin (macrolide) (all drugs manufactured by SigmaAldrich, USA) was determined by the serial dilution method in accordance with the EUCAST guidelines (v.14.0) [10]. The strains showing multidrug resistance (MDR) were determined as resistant to three or more antibiotics of different classes.

The study involved lytic bacteriophages vB_SauM-515A1 (genus Kayvirus, family Herelleviridae) and vB_SauP-436A (genus Rosenblumvirus, family Rountreeviridae) used in the form of sterile filtrates of phage lysates in the LB medium. Bacteriophages were earlier isolated from the commercially available combination phage product “Staphylococcal Bacteriophage” of the series P332 (Microgen; Russia) in the S. aureus SA515 (ST8 (ST, sequence type)) and SA436 (ST1) strains and characterized in detail [11]. Phages were grown in appropriate host strains.

Molecular genetic characteristics of bacterial strains

DNA was extracted using the DNA-Express kit (Lytech; Russia) following the manufacturer's instructions. DNA samples were stored at –20°C.

Strain typing was performed by multilocus sequence typing (MLST) in accordance with the standard scheme [12]. The icaA, icaB, icaC, and icaD genes responsible for biofilm formation were detected by polymerase chain reaction (PCR) using the previously reported primers [13]. The PCR reaction mixture (25 µL) contained 66 mM Tris-HCl (рН = 9), 16.6 mM (NH₄)₂SO₄, 2.5 mM MgCl₂, 250 µM of each dNTP, 1 unit of Taq DNA polymerase (Lytech; Russia) and 10 pmol of each appropriate primer. Amplification was performed using the DNA Engine Tetrad 2 kit (MJ Research; USA) in accordance with the previously proposed regimes [13]. The amplification products were assessed by horizontal 2% agarose gel electrophoresis with ethidium bromide visualization.

Assessment of biofilm formation

Assessment was performed by the previously reported method [14] with certain modifications. For that the suspension of bacterial cells being in their exponential growth phase (optical density (OD) at 620 nm was 0.12) was inoculated into the wells of the uncoated 96-well flat bottom ventilated plate (Thermo Scientific; USA) containing TSBg (tryptic soy broth supplemented with 1% glucose) (Himedia; India) to the final concentration of 10⁴ cells per well and incubated at 37 °C for 24 h without shaking. The final volume in each well was 200 µL. The sterile medium was used as a negative control. After incubation the wells were carefully triple washed with the sterile phosphate buffer (PBS, pH = 7.4) to remove planktonic cells, then stained with the 0.1% crystal violet (CV) aqueous-alcohol solution (Sigma-Aldrich; USA) for 30 min at room temperature. After incubation the dye was triple washed with the sterile PBS. To perform further analysis, the dye bound in each well was eluted by adding 200 µL of 96% ethanol, and optical density of the solution was measured using the Microplate Reader Flex-A (Allsheng; China) at 570 nm. All the experiments were conducted in three biological replicates.

The ability of the bacterium to form biofilms was determined in accordance with the earlier proposed criteria: OD ≤ ODc — the strain does not produce biofilms; ODc < OD ≤ 2 × ODc — the strain is a weak biofilm producer; 2 × ODc < OD ≤ 4 × ODc — the strain is a moderate biofilm producer; 4 × ODc < OD — the strain is a strong biofilm producer, where ODc — is the average OD of the negative control + 3 × SD (standard deviation). The sterile medium was used as a negative control [15].

Sensitivity of planktonic forms of strains to bacteriophages

The strains’ sensitivity to bacteriophages was determined when assessing the efficiency of plating (EOP), as previously reported [11]. EOP is a ratio of the bacteriophage titer in the test strain to the bacteriophage titer in the host strain (S. aureus SA515 for phage vB_SauM-515A1; S. aureus SA436 for phage vB_SauP436A) expressed as a percentage. Bacteriophage titer in the test strain was determined by the Gracia titration method, as previously reported [16]. For that aliquots (5 µL) of the 10-fold serial dilutions of each bacteriophage product (stock of 2 × 10⁹ plaque-forming units/mL or PFU/mL) were applied to the surface of the plates with the semi-solid LB agar (0.6% agar) containing 0.1 mL of the test strain overnight culture (10⁶ CFU/mL) and incubated at 37 °С for 24 h. Concentration of phage particles in PFU/mL was estimated for each strain. The efficacy of plating was assessed in three replicates.

Assessment of the impact of bacteriophages on the biofilms formed

The experiments involved the 24-h biofilms produced and washed in accordance with the above method. After washing, phage lysate in TSBg was added to the well to the final concentration of 10⁸ PFU/mL, the volume in each well was 200 µL. Sterile PBS was added to the control samples instead of phage lysate. Incubation was carried out for 24 h at 37 °C. Then CV staining was performed as described above, with subsequent OD measurement at 570 nm. All the experiments were conducted in three biological replicates.

In preliminary experiments, we determined the number of cells in the biofilm by adding 200 µL of sterile PBS to the prewashed biofilm and destroying the biofilm by active pipetting. Serial dilutions of the cell suspension were sown on the LB agar. Colonies were enumerated after the 24-h incubation at 37 °C.

Statistical analysis

Statistical analysis was performed using the GraphPad Prism v. 8.0.1 software (GraphPad Software Inc.; USA) based on the t-test data. During the analysis we compared optical density values obtained after the 24-h incubation of the biofilms treated and not treated with the bacteriophage. The differences were considered significant at p < 0.05. The Shapiro–Wilk test was used to confirm normal distribution of data, the data on the samples were considered to be almost normally distributed at p > 0.05.

Fisher’s exact test was used to reveal significant correlations based on two nominal traits in small samples. The correlations were considered significant at p < 0.05.

RESULTS

Characteristics of strains and their ability to form biofilms

According to the MLST results, strains of the collection belonged to eight sequence types, among which the most abundant was ST8 (4/20, 20%) (table). ST1, ST121, ST5, and ST25 accounted for three strains each (15%). As for sensitivity to antibiotics, a large number of strains belonged to MRSA (9/20, 45%). Furthermore, seven strains (35%) were characterized by MDR.

All the tested strains were able to form biofilms, and more than a half were strong biofilm producers (13/20, 65%). In the remaining cases (7/20, 35%), the strains were moderate biofilm producers. CFU enumeration showed that the number of cells in the 24-h biofilms reached 10⁸ CFU/mL for all the studied strains. The amplification results demonstrated that all the isolates contained a complete set of the icaADBC operon genes.

Effects of bacteriophages of various taxonomic groups on the S. aureus planktonic cells and biofilms

Based on the results of assessing the effects of bacteriophages vB_SauM-515A1 and vB_SauP-436A on the planktonic cells of strains of the collection it was determined that 14 strains (70%) were sensitive to phage vB_SauM-515A1, while 10 strains (50%) were sensitive to phage vB_SauP-436A. Three strains of the collection (15%) turned out to be resistant to both phages. The phage vB_SauM-515A1 efficiency of plating in sensitive strains varied between 90 and 500%, while that of phage vB_ SauP-436A varied between 75 and 400%.

The effect of bacteriophages on biofilm was assessed only for strains sensitive to the corresponding bacteriophage, based on data from experiments with planktonic cells. Biofilms were treated with the bacteriophage titer of 10⁸ PFU/mL, which, based on CFU enumeration in the untreated biofilms, corresponds to the multiplicity of infection (MOI) value of 1.

According to the experimental results (figure), in the majority of cases (9/14, 64.3%), the 24-h treatment with the vB_SauM515A1 bacteriophage did not cause reduction of the biofilm biomass but, on the contrary, stimulated biomass production. As for remaining five strains (5/14, 35.7%), the exposure to the vB_SauM-515A1 bacteriophage had no significant effect on the biofilm. The exposure to bacteriophage vB_SauP-436A, on the contrary, significantly reduced the biofilm biomass in four strains (4/10, 40%). This bacteriophage had no effect on the biofilms formed by other strains.

It should be noted that statistical analysis has revealed no significant correlations between the fact that the strain belongs to MRSA or certain sequence type and the ability of bacteriophage to stimulate biofilm formation (in the case of vB_SauM-515A1) or destroy biofilms (in the case of vB_SauP-436A). These effects also did not depend on the baseline strains’ ability to form biofilms.

DISCUSSION

In this study, the strains were considered isolated from the heterogeneous clinical material that belonged to epidemiologically significant sequence types (table). Among them the most abundant was ST8, one of the most common sequence types in Russia and all over the world among hospital strains [17, 18]. Strains of this sequence type include the pandemic clones, such as USA300, causing multiple infection outbreaks and often belonging to MRSA [19]. Along with ST8, the ST1, ST5, and ST121 strains were identified, which, according to the literature data, are characterized by antibiotic resistance and are capable of causing severe infections [20–22].

The findings showed that all the studied samples could form biofilms, regardless of the origin and sequence type. Furthermore, all strains were characterized by the presence of the icaADBC operon responsible for biosynthesis of polysaccharide intercellular adhesin, the major and best studied component of the S. aureus clinical strains’ biofilm matrix, in the genome [23, 24]. It should also be noted that the majority of strains turned out to be strong biofilm producers. In combination with antibiotic resistance, this once again emphasizes severity of the problem of treating the infections caused by these strains.

Members of the families Herelleviridae and Rountreeviridae were used to assess the efficacy of bacteriophages. Phages were selected based on the differences in their morphological (myoviruses and podoviruses), microbiological (host spectrum, parameters of infection based on the single-step growth curve), and genetic (genome size and the number of genes encoded) characteristics [11]. According to the findings, vB_SauM-515A1 (genus Kayvirus, family Herelleviridae) showed higher efficacy against planktonic cells compared to vB_SauP-436A (genus Rosenblumvirus, family Rountreeviridae), which is consistent with the earlier reported data on these bacteriophages [11]. As for other lytic bacteriophages of the family Herelleviridae (earlier referred to as Myoviridae), it has been also shown that their lytic spectrum varies between 85.3 and 99.2% depending on the collection, while in members of the family Rountreeviridae (earlier referred to as Podoviridae) this indicator is 64–68% [11, 25, 26].

As for biofilms, it was shown that bacteriophage vB_SauM-515A1 stimulated the biofilm biomass increase, while phage vB_SauP-436A reduced its quantity (figure). It was earlier reported that myovirus phiIPLA-RODI belonging to the same genus, as vB_SauM515A1, could stimulate biofilm formation in S. aureus [9]. The authors explain the findings by the increased eDNA content in the matrix, which, in turn, is associated with high lytic activity of bacteriophages of this family. The increased eDNA content contributes to the biofilm structural integrity and stability, as well as modulates formation of amyloid fibers essential for the biofilm architecture maintenance [27]. In another study, the researchers have shown that the use of phages of the genera Kayvirus и Rosenblumvirus (separately or in combination) in initial phases of biofilm formation or in mature biofilms does not result in the decrease in biofilm quantity [8]. The authors observed such an effect when using phages at both low (0.1) and high (10) MOI values. Furthermore, the increase in biofilm quantity was observed after the 24-h incubation, except for the case of treating the mature biofilm with the mixture of phages at MOI 10.

It should be noted that in the studies discussed the authors confined themselves to testing one and two bacterial isolates, which limits comprehensive comparison of the results. Nevertheless, the available data suggest the trend towards stimulation of biofilm formation in the S. aureus strains shown by phages of the genus Kayvirus. According to our data, reduction of the biomass quantity or the lack of significant effect was observed, when treating biofilms with bacteriophage of the genus Rosenblumvirus. Perhaps, vB_SauP-436A causes less intense cell lysis and lower eDNA release, which hampers stimulation of biofilm formation. In this regard, the combination use of bacteriophages, either in the form of the cocktail comprising several phages, or in combination with the agents destroying the matrix, such as depolymerases, seems to be a promising approach to combating biofilms [28].

It should also be noted that in this study phage efficacy was assessed by staining the biofilm biomass consisting of matrix and cells (both live and dead) with crystal violet. This method makes it impossible to assess the share of viable cells after treatment with antibacterial agents, so it is necessary to conduct further research for better understanding of the effects of phages on biofilms.

CONCLUSIONS

The study emphasizes complexity and ambiguity of the effects of bacteriophages on the S. aureus biofilm, especially in the case of clinical isolates showing high antibiotic resistance. Despite the ability of phages of the families vB_SauM-515A1 and vB_SauP-436A to inhibit growth of planktonic cells, their efficacy against biofilms of the S. aureus strains turned out to be low. The findings emphasized the importance of selecting phages based on their specific characteristics and efficacy against biofilms, as well as possible need for combination approaches involving bacteriophages and the agents destroying the biofilm matrix. At the same time, it should be considered that the results of this study have been obtained in the in vitro experiments by the limited number of methods. Further research is required to translate the findings into clinical practice, including the experiments involving in vivo models, which will make it possible to more accurately assess the efficacy and safety of the proposed approaches under conditions close to reality.