Introduction

Infectious diseases caused by antimicrobial resistant (AMR) bacteria are a serious global concern, as they can make treatment with antimicrobials less effective or even ineffective1,2,3. Among Gram-positive AMR bacteria, methicillin-resistant Staphylococcus aureus (MRSA) is recognized as a major AMR bacterium. MRSA has been widespread not only in hospitals but also in communities4,5. The glycopeptides, arbekacin, daptomycin, linezolid and tedizolid are believed to be effective against MRSA infections, but strains resistant to these antimicrobials has already been reported6,7,8,9,10. In addition, vancomycin-resistant enterococci (VREs) are known to cause serious opportunistic infections in elderly and immunocompromised individuals11. Since therapeutic options for infections caused by multidrug-resistant (MDR) or extensively drug-resistant (XDR) bacteria are limited, the development of new classes of antimicrobial agents is needed. Bacteriocins, which are antimicrobial peptides or proteins produced by bacteria, are known as novel antimicrobial agent candidates12,13,14. It is generally known that bacteriocins are strongly effective against the same species or genus12. Considering the effectiveness against MRSA, those produced by Staphylococcus species may serve as promising therapeutic candidates. Several Staphylococcus spp. strains are known to produce bacteriocins such as epidermin, Pep5, and nukacin ISK-115. Some bacteriocins, such as Pep5, epidermin, and nukacin ISK-1, have anti-MRSA activity, and their efficacy has been evaluated in several studies16,17.

Staphylococcus capitis is a coagulase-negative staphylococci that primarily resides on human skin, particularly on the scalp, and is considered a low-pathogenicity commensal bacterium. S. capitis has been reported to produce several bacteriocins, such as capidermicin, epidermin, gallidermin and nisin J18,19,20. In this study, we screened bacteriocin-producing strains from S. capitis clinical isolates and identified one S. capitis strain that showed strong antimicrobial activity against Gram-positive bacteria. We also evaluated the inhibitory effects of this strain on MRSA, VRE, and other Gram-positive pathogens.

Results

Identification and antimicrobial activity of the bacteriocin-producing S. capitis strain HBC3

We performed direct assays to identify potent bacteriocin-producing S. capitis strains using the S. aureus MRSA strain MW2 as an indicator. Among 18 S. capitis strains, we found that one strain, HBC3, exhibited antimicrobial activity against MW2 (Table 1). Next, we investigated the antibacterial activity of HBC3 against several bacterial species. S. capitis HBC3 exhibited antibacterial activity against Enterococcus faecium, Enterococcus faecalis, Staphylococcus epidermidis, Streptococcus pyogenes, Listeria monocytogenes, Bacillus coagulans, Clostridium difficile, Corynebacterium accolens, and Corynebacterium propinquum whereas HBC3 showed no activity against Klebsiella pneumoniae, Escherichia coli or Acinetobacter baumannii (Table 1). In addition, a direct assay was performed with multiple indicator strains, including S. aureus (n = 61), S. epidermidis (n = 20), E. faecalis (n = 12), and E. faecium (n = 18). The mean zones of inhibition for S. aureus, S. epidermidis, E. faecalis, and E. faecium were 25.66 mm, 18.96 mm, 28.53 mm, and 30.53 mm, respectively (Fig. 1A). S. epidermidis was significantly less susceptible than were S. aureus, E. faecalis, and E. faecium (P < 0.0001). MRSA strains were significantly more susceptible than methicillin-susceptible Staphylococcus aureus (MSSA) strains were (P < 0.001), but there was no significant difference in susceptibility between vancomycin-resistant and vancomycin-susceptible E. faecalis and E. faecium strains (Fig. 1B).

Table 1 Antibacterial activities of HBC3.
Full size table
Fig. 1
Fig. 1
Full size image

Antibacterial activity of S. capitis HBC3 against 4 bacterial species. (A) A soft agar overlay assay was performed using S. capitis HBC3 as a bacteriocin-producing strain and various strains, including Staphylococcus aureus (n = 61), Staphylococcus epidermidis (n = 20), Enterococcus faecalis (n = 12), and Enterococcus faecium (n = 18), as indicators. (B) The inhibitory zones of the antibiotic-susceptible and antibiotic-resistant strains of each species were compared. MSSA, methicillin-susceptible S. aureus; MRSA, methicillin-resistant S. aureus; VSE, vancomycin-susceptible enterococci; VRE, vancomycin-resistant enterococci. One-way analysis of variance was performed to determine statistically significant differences among species, and Student’s t test was performed to determine statistically significant differences between drug-resistant and drug-susceptible strains. Ns, not significant; *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Identification of bacteriocin gene clusters in HBC3

To identify the bacteriocin produced by the S. capitis strain HBC3, we performed a genomic analysis. Using BAGEL4 and antiSMASH, we identified two bacteriocin gene clusters, the capidermicin operon and the tcl operon, which are responsible for the production of micrococcin P1 (MP1) in several Staphylococcus species21 on a plasmid named pHBC3_1. The pHBC3_1 plasmid has a length of 32,690 bp and 38 putative open reading frames (ORFs), including the genes for MP1 synthesis and its immunity, capidermicin synthesis and its immunity, replication-associated factors, transposases, and hypothetical proteins (Fig. 2A; Table 2). The HBC3 strain without pHBC3_1 (designated HBC3 p(-)) showed no antimicrobial activity against MW2, VRE1 or B. coagulans (Fig. 2B).

Fig. 2
Fig. 2Fig. 2Fig. 2
Full size image

Gene map of the pHBC3_1 plasmid in HBC3. (A) Gene map of pHBC3_1 and comparison with 8 other publicly available plasmids. The outermost circle displays the gene map of pHBC3_1, with genes presented as arrows and orientations referring to the direction of gene transcription. A comparison of the pHBC3_1 plasmid with 8 other plasmids was performed with BRIG using pHBC3_1 as the reference. The two innermost circles represent the GC content (black) and CG skew (purple/green). The color key provides the names of the 11 plasmids, which are arranged from the inner to the outer colored ring. (B and C) The bacteriocin-coding regions of MP1 in HBC3 were compared with the plasmids of S. hominis S34-1 (GenBank: CP040733.1) and the plasmid of S. capitis CIT060 (GenBank: MN234131.1). (D) Antibacterial activity of HBC3 with or without the pHBC3_1 plasmid. A direct assay was performed to evaluate the antibacterial activity against 3 bacterial species.

Table 2 The genes encoded in the micrococcin P1 and capidermicin-hourboring plasmid pHBC3_1.
Full size table

The MP1 gene cluster was located at nucleotide positions 373 to 11,357 on pHBC3_1. This region showed homology with the tcl region of S. hominis S34-1, a strain known to produce MP122(Fig. 2C). The amino acid sequence of the tclE gene, encoding the MP1 leader peptide, was completely identical to that of S34-1. Most of the other genes also presented identical amino acid sequences, but ORF3 and tclI presented one amino acid substitution, tclL presented two amino acid substitutions, and tclP presented a frameshift mutation in S34-1. The gene encoding the capidermicin peptide was found at nucleotide positions 15,537 to 15,689. Compared with the capidermicin region of the S. capitis CIT060 strain, a known capidermicin producer20, the genomic region between ORF13 and the front part of ORF20 on pHBC3_1 corresponded with the reported CIT060 sequence (Fig. 2D). The amino acid sequence of the capidermicin gene was completely the same as that in a previous report20. The genes encoding the putative immunity proteins ydbS and ydbT were also identified between nucleotide positions 16,318 and 18,210, and compared with the ydbS of CIT060, only ydbS had a single amino acid substitution.

A comparison of pHBC3_1 with 8 publicly available plasmids with the highest identity, as analyzed by nucleotide BLAST, revealed that the pSc1516939_1 plasmid, which is hosted by S. capitis subsp. urealyticus strain Sc151693923, had a nearly identical nucleotide sequence to that of pHBC3_1 across the entire plasmid (percent identity of nucleotide sequence: 99.85%, query coverage: 100%) (Fig. 2A). In addition, the plasmids of the S. hominis S34-1 strain, and five S. aureus strains showed sequence homology with the MP1 biosynthetic locus on pHBC3_1, and one S. capitis-derived plasmid showed homology with regions on pHBC3_1, excluding the MP1 and capidermicin loci.

Purification and identification of capidermicin and MP1

To identify the bacteriocins encoded by pHBC3_1, bacteriocins were purified from HBC3 culture supernatants by two methods: the Macro-Prep resin method16 and the butanol extraction method24. Following purification of the bacteriocin by HPLC, the exact molecular mass of the fraction with antibacterial activity was measured using Electrospray ionization mass spectrometry (ESI-MS). By using the Macro Prep resin, the mass of the fraction with antibacterial activity corresponded to the predicted mass of capidermicin (5464.390 Da) (Fig. 3A and B). However, MP1 was not detected by this method. By using the butanol extraction method, one fraction with antibacterial activity matched the predicted mass of MP1 (1144.360 Da) (Fig. 3C and D). The peaks corresponding to capidermicin and MP1 were not observed in the samples obtained from the culture supernatants of plasmid-cured strains (Fig. 3A and C).

Fig. 3
Fig. 3
Full size image

Purification and identification of capidermicin and MP1. (A, C) Purification of bacteriocin from the culture supernatant of HBC3 (red) and its plasmid-deleted strain (blue) was performed by reverse-phase HPLC. Capidermicin and MP1 were isolated with a linear gradient from 0% to 100% acetonitrile containing 0.1% trifluoroacetic acid at a flow rate of 1 mL/min. (B, D) Molecular mass determination of MP1 and capidermicin was performed by ESI‒MS.

Antimicrobial spectrum of capidermicin and MP1

To investigate the antimicrobial spectrum of MP1 and capidermicin, we evaluated their antibacterial activity with a spot-on-lawn assay using purified capidermicin and MP1 against several bacterial species. MP1 showed strong antibacterial activity against all of the tested Gram-positive bacteria, while capidermicin had strong antimicrobial activity against B. coagulans JCM2257 and weak activity against E. faecium VRE1 and L. monocytogenes 698, whereas no activity was observed in S. aureus MW2, S. epidermidis KSE36 and S. pyogenes SF340 (Fig. 4A).

Fig. 4
Fig. 4
Full size image

Antibacterial activity of purified MP1 and capidermicin against. (A) The spot-on-lawn method was used to evaluate the susceptibility of the strains to 6 bacterial species. A 2µL spot of each bacteriocin solution at 60µM was spotted. (B) The susceptibility to 3 bacterial species was evaluated using the MIC assay.

Minimum inhibitory concentrations (MICs) of capidermicin and MP1

To evaluate the antibacterial activity of capidermicin and MP1, the minimum inhibitory concentration (MIC) was determined using the broth microdilution method (Fig. 4B). The MICs of capidermicin were 0.5 µM (2.7 µg/mL) against B. coagulans JCM2257 and greater than 16 µM (86.96 µg/mL) against S. aureus MW2 and E. faecium VRE1. In contrast, the MICs of MP1 were 2 µM (2.29 µg/mL) against MW2 and JCM2257 and 4 µM (4.58 µg/mL) against VRE1.

MP1 and capidermicin inhibited bacterial growth at sub-MIC concentrations

The antibacterial activity of HPLC-purified MP1 and capidermicin was evaluated against B. coagulans JCM2257. MP1 or capidermicin alone inhibited the growth of JCM2257 in a concentration-dependent manner (Fig. 5A, B). An additive inhibitory effect on bacterial growth was observed when MP1 at 1/2 MIC was combined with capidermicin at either 1/2 MIC or 1/4 MIC (Fig. 5C). A similar additive effect was also detected when 1/4 the MIC of MP1 was combined with 1/2 the MIC of capidermicin, with a fractional inhibitory concentration index (FICI) of 0.75 (Fig. 5D).

Fig. 5
Fig. 5
Full size image

Effects of MP1 and capidermicin on the growth of B. coagulans JCM2257. Various concentrations of MP1 and capidermicin were incubated with 1 × 10⁵ MW2 or JCM2257 cells. Bacterial growth was monitored by measuring the optical density at 660 nm (OD₆₆₀) at 10-minute intervals over 24 h using a microtiter plate reader. Each experiment was performed in triplicate, and the data are presented as the mean values.

Coculture of S. capitis with MW2, VRE1 and JCM2257

To investigate the inhibitory effect of HBC3 on the growth of competing bacteria, coculture assays were performed using S. aureus MW2, E. faecium VRE1, and B. coagulans JCM2257. Coculturing with the HBC3 WT strain completely suppressed the proportions of MW2, VRE1, and JCM2257 (Fig. 6), whereas coculturing with HBC3 p(-) resulted in markedly increased colonization rates of 96.8% for MW2, 17.16% for VRE1, and 16.03% for JCM2257.

Fig. 6
Fig. 6
Full size image

Coculture assay of S. capitis with MW2, VRE1 and JCM2257. Coculture assays were performed according to the methods described in the Materials and methods. The proportions of MW2, VRE1 and JCM2257 cocultured with HBC with or without the pHBC3_1 plasmid were calculated. The statistical significance of the differences was determined by Student’s t test. ***, P < 0.001; ****, P < 0.0001.

Discussion

In this study, we identified a Staphylococcus capitis strain, HBC3, that produces two antimicrobial peptides, micrococcin P1 (MP1) and capidermicin. HBC3 showed strong antimicrobial activity against several Gram-positive bacteria, including S. aureus, E. faecium, S. pyogenes and C. difficile (Table 1). Interestingly, S. aureus was more susceptible to HBC3 than S. epidermidis was, and MRSA was more susceptible than MSSA (Fig. 1). Since we found that the plasmid-deleted strain of HBC3 showed no antibacterial activity against these bacteria, the production of MP1 and capidermicin reflected the results of the direct assay. Additionally, capidermicin showed no antibacterial activity against Staphylococci, suggesting the differences in susceptibility between S. aureus and S. epidermidis as well as between MSSA and MRSA strains are attributable to MP1. MP1 is a 26-membered ring-based thiopeptide found in several genera, such as Staphylococcus, Micrococcus and Macrococcus25; however, this is the first isolation of MP1-producing S. capitis. It is also reported that MP1 exhibit synergistic effects with existing antibiotics such as rifampicin have also been reported26. The mechanism of action of MP1 has been reported to involve the inhibition of protein synthesis by the targeting of the 50 S ribosomal subunit25,27,28. Liu et al. reported that, owing to the hydrophobicity of MP1, MP1 is incorporated into membrane vesicles, resulting in the cytotoxicity of membrane vesicles29. MP1 integrates into and then penetrates through the membrane, resulting in the inhibition of the 50 S ribosomal subunit. Reifsteck F et al. reported the analysis of hydrophobicity-hydrophilicity among various staphylococcal strains and reported that most staphylococci, including S. aureus, were hydrophobic, but the hydrophobicity was influenced by several factors, such as the presence of capsules and cell surface proteins30. Rawlinson LB et al. reported that S. aureus isolates tend to be more hydrophobic than S. epidermidis isolates are, although several individual S. aureus and S. epidermidis clinical isolates have similar surface hydrophobicities31. Therefore, it is speculated that the different susceptibilities among staphylococci may be caused by variations in the hydrophobicity of the bacterial membrane. Capidermicin was first identified from the S. capitis CIT060 strain20, and capidermicin is a class II leaderless bacteriocin predicted to be a cationic peptide with a theoretical net charge of + 5.34 (pI = 10.22), consisting of four α-helices20. The antibacterial mechanism of capidermicin has not been elucidated in detail, but it is thought that capidermicin can permeabilize cell membranes or forming pores resulting in cell death because of the structural similarity with LnqQ and A5332,33. Capidermicin has shown antibacterial activity against several staphylococcal species, including S. aureus and Listeria monocytogenes20,34. In this study, we detected no or weak antibacterial activity against S. aureus. The different results from those of previous reports may be due to the use of different strains or MIC evaluation methods (in a previous report, the MIC was determined after 16 h of incubation, but we evaluated the MIC after 24 h of incubation). In contrast, we observed clear antibacterial activity against L. monocytogenes and B. coagulans, which are Gram-positive rods. Therefore, we concluded that capidermicin exhibited more effective antibacterial activity against Gram-positive rods than cocci. The difference in the properties of these two bacteriocins results in a wide range of antimicrobial activities, which allows this bacterium to competitively exist in the indigenous flora of the skin and other parts of the body.

We thought about the possibility that cross-resistance with antimicrobial agents that share the same target site as bacteriocins. While MP1 targets the ribosome and several macrolide resistance factors exert their effects through ribosomal modification or protection35, we verified whether these factors affect MP1 susceptibility in S. aureus strains. The antimicrobial activity of HBC3 against S. aureus was thought to be due to MP1, thus the halo sizes were compared between ermA, tet(M), msrA-carrying strains and other strains in 61 S. aureus strains used in Fig. 1A. None of these resistance genes involved in MP1 resistance (Fig. S1). As for capidermicin cross resistance, daptomycin, a clinically used antimicrobial agent against MRSA, also disrupts bacterial membrane as capidermicin. The emergence of S. aureus daptomycin-nonsusceptible (DAP-NS) strains has been reported36, and its major mechanism of daptomycin resistance is the mutations in mprF gene, encoding membrane protein that mediates lysinylation of phosphatidylglycerol and weaken cell surface negative charge37. Capidermicin is also a cationic peptide as daptomycin, and we revealed that S. aureus ΔmprF strain exhibits increased susceptibility to capidermicin (Fig. S2). Thus, it is predicted that cross-resistance to daptomycin and capidermicin may occur. To evaluate whether the immunity factors of MP1 and capidermicin are involved in antimicrobial resistance, we determined the MICs of erythromycin, tetracycline, gentamicin, chloramphenicol, daptomycin and vancomycin in HBC3 and its plasmid-deleted strain. There were no differences between two strains (Table S2), suggesting that the immunity factors do not involve in antimicrobial resistance.

Genomic analysis revealed that HBC3 possesses MP1 and capidermicin gene clusters on the plasmid pHBC3_1 (Fig. 2A). The nucleotide sequence of pHBC3_1 is almost the same as that of the plasmid pSc1516939_1 of the S. capitis subsp. urealyticus strain Sc1516939 (identity: 99.85%)23. However, the report of pSc1516939_1 primarily focused on determining the complete genomes of S. capitis strains. Therefore, this is the first report to examine the antimicrobial activity and bacteriocin production of an S. capitis strain that carries this plasmid. Interestingly, the MP1 and capidermicin gene clusters were located on a single plasmid. Previously, the MP1 or capidermicin gene cluster was separately identified in individual plasmids. In several MP1-encoding plasmids, IS6 family transposases and their inserted sequences were found at the flanking region of the MP1 gene cluster. This IS6 sequence is also observed in pHBC3_1. Insertion sequences, as mobile elements, play important roles in shaping host genomes38. ISs in ISfinder (https://www-is.biotoul.fr/index.php) are grouped into many families, including IS6, and are further divided into subgroups39. IS6 is known to be involved in generating clusters of antibiotic resistance genes40. Therefore, the presence of IS6 family insertion sequences within pHBC3_1 suggests that horizontal gene transfer and recombination events may have facilitated the assembly of this composite plasmid. This unique genetic configuration likely endows HBC3 with increased ecological competitiveness by enabling the simultaneous production of two potent bacteriocins. Gene cluster analysis revealed that the MP1 gene cluster is quite similar to that of S. hominis strain S34-1, a previously characterized MP1-producing strain21. Additionally, the capidermicin gene cluster of HBC3 was almost the same as that of the known capidermicin producer strain CIT06020. Therefore, it is possible that pHBC3_1 is generated by the fusion of two plasmids similar to these plasmids. The region outside the bacteriocin gene clusters contains several genes such as plasmid replication initiator repA and tyrosine-tRNA ligase (Fig. 2A). The nucleotide sequence of a plasmid from S. capitis in NCBI database only showed homology to this region (pSKB0123-1). Therefore, it is predicted that this plasmid could spread to only S. capitis.

Since the deletion of pHBC3_1 from the HBC3 strain resulted in no antibacterial activity (Fig. 2B) and the two purified bacteriocins completely matched those previously identified by ESI-MS analysis (Fig. 3B, D), the antibacterial activity of HBC3 is clearly due to the production of these two bacteriocins. A spot-on-lawn assay using purified MP1 and capidermicin revealed that, compared with capidermicin, MP1 exhibited broader activity against Gram-positive bacteria, whereas capidermicin displayed a narrower spectrum. (Fig. 4). Notably, against B. coagulans JCM2257, capidermicin showed the strongest activity among all tested strains, while MP1 comparatively less effective. These results indicate that the two peptides have distinct target preferences and may function complementarily. To further investigate this, a checkerboard assay was performed using purified MP1 and capidermicin against Bacillus coagulans JCM2257 to evaluate their combined effects. Growth inhibition was observed when 0.5 µM MP1 (1/4 MIC) and 0.25 µM capidermicin (1/2 MIC) were combined (Fig. 5D). On the basis of these concentrations, the fractional inhibitory concentration index (FICI) was calculated to be 0.75, indicating an additive effect. These findings support the idea that MP1 and capidermicin can act cooperatively to inhibit the growth of B. coagulans. A coculture assay revealed that HBC3 potently inhibited the growth of S. aureus MW2, E. faecium VRE1 and B. coagulans JCM2257 in the same environment (Fig. 6). This elimination effect was significantly lower in the HBC3 p(-) mutant, suggesting that the two peptides encoded by the plasmid have inhibitory effects. On the basis of the antibacterial characteristics of the two bacteriocins, MP1 is involved mainly in eliminating S. aureus and E. faecium, whereas capidermicin is involved mainly in eliminating B. coagulans. These results indicate that the production of two different bacteriocins helps HBC3 to inhibit other competing Gram-positive bacteria in the same niche and promote its flourishing.

The results of this study suggest that the coexistence of two structurally and functionally distinct bacteriocins in a single S. capitis strain strongly inhibits the competing bacteria. Bacteriocin-producing strains are known to be candidates for future probiotic-based or decolonization therapies41,42,43. However, when bacterial strains are applied to probiotics, attention should be given to the pathogenicity and drug resistance of the bacteria. The S. capitis lineage called NRCS-A clone (NICU-Related Clone of S. capitis – type A) is known to show high virulence and methicillin resistance and has been repeatedly isolated from neonatal intensive care units (NICUs) worldwide44. However, HBC3 did not belong to this pathogenic lineage and did not possess any detectable antimicrobial resistance genes (data not shown). Future investigations should assess their in vivo efficacy in inhibiting the colonization of pathogenic or multidrug-resistant bacteria, toxicity toward host cells, pharmacokinetic properties, and potential synergy with conventional antibiotics45. Moreover, their effects on the composition and stability of the human microbiota should be evaluated to determine their clinical applicability. In addition, since the antibacterial effect of HBC3 against S. aureus was greater than that against S. epidermidis and Corynebacterium accolens, which are commensal bacteria in human skin and the nasal cavity, the application of probiotics using HBC3 or MP1 is expected to selectively target S. aureus in the human microbiota.

In conclusion, we characterized an S. capitis strain that shows strong antimicrobial activity against diverse Gram-positive bacteria due to the production of two plasmid-encoded antimicrobial peptides, MP1 and capidermicin. This study is expected to contribute to their potential application as novel therapeutics targeting pathogenic or AMR bacteria.

Methods

Bacterial strains and growth conditions

The bacterial strains used in this study are shown in Table S1. Staphylococcus strains were cultured at 37 °C in trypticase soy broth (TSB) (Becton, Dickinson and Company [BD], Franklin Lakes, NJ, USA) with shaking. The Streptococcus mutans was grown in TSB at 37 °C with static incubation in a 5% CO2 incubator.

The S. aureus and S. epidermidis clinical isolates used in this study were previously isolated17,46. Corynebacterium spp. were cultured in TSB supplemented with 0.05% Tween 80 at 37 °C with shaking. Clostridium difficile was cultured anaerobically at 37 °C in BHI medium (BD). Cutibacterium acnes was grown anaerobically on GAM agar (Nissui Pharmaceutical, Tokyo, Japan) at 37 °C. For anaerobic conditions, we used a GasPack system (Mitsubishi Gas Chemical Company, Inc., Tokyo, Japan). Listeria monocytogenes was grown at 37 °C in BHI medium aerobically.

Isolation of S. capitis and Corynebacterium strains from the nasal cavity

The S. capitis strains, and Corynebacterium strains used in this study were isolated from outpatients at the dental department of Hiroshima University Hospital. All methods were performed in accordance with relevant guidelines and regulations, including the Declaration of Helsinki. The study was approved by the Hiroshima University Hospital Ethics Committee (Review Board approval number: E-2525). Informed consent was obtained from all study participants or their legal guardians prior to sample collection47.

Nasal swabs were plated on 5% sheep blood agar plates (BD) and incubated for 2 days at 37 °C aerobically. The colonies were replated on tryptic soy agar (TSA), and the species were identified using matrix-assisted laser desorption ionization time‒of-flight mass spectrometry (MALDI-TOF MS) performed with a BD™ Bruker MALDI Biotyper™ Sirius system (Becton, Dickinson and Company, Franklin Lakes, NJ, USA) using MBT Compass 4.1, with the MBT Compass library Ver.9.0.0 (8468 MSPs) (Bruker Daltonik GmbH, Bremen, Germany) database used as a reference.

Screening of bacteriocin-producing strains of coagulase-negative Staphylococcus isolates

Overnight cultures (3 µL) of 18 Staphylococcus capitis strains isolated from the human nasal cavity were spotted onto TSA plates and incubated at 37 °C for 24 h. Subsequently, 3.5 mL of prewarmed half-strength TSB soft agar (0.75%) containing S. aureus MW2 (10⁷ CFU/mL) was overlaid onto the plates and incubated at 37 °C for an additional 20 h. Antibacterial activity was assessed by the presence or absence of a clear inhibition zone (halo) surrounding the spotted colonies.

Determination of the complete genome sequence of the bacteriocin-producing strain HBC3

To obtain the complete genome sequence of the Staphylococcus capitis strain HBC3, genomic DNA was extracted following a protocol based on a previously described method46. Briefly, cells cultured overnight in 5 mL of tryptic soy broth (TSB) were harvested by centrifugation and resuspended in 0.5 mL of CS buffer (100 mM Tris-HCl [pH 7.5], 150 mM NaCl, and 10 mM EDTA) containing lysostaphin at a final concentration of 70 µg/mL (Sigma‒Aldrich, St. Louis, MO, USA). After incubation at 37 °C for 2 h, 0.5 mL of saturated phenol was added, and the sample was mixed thoroughly and centrifuged to collect the aqueous phase. Subsequently, 0.5 mL of a phenol: chloroform (1:1) mixture was added, followed by mixing and centrifugation. The DNA was precipitated by adding an equal volume of ethanol, and the resulting DNA pellet was dissolved in distilled water. Whole-genome sequencing (WGS) was performed using both the Illumina HiSeq X FIVE platform (Illumina, San Diego, CA, USA) and the GridION platform (Oxford Nanopore Technologies, Oxford, UK). A hybrid assembly combining Illumina short reads and Nanopore long reads was conducted using Unicycler v0.5.048, resulting in the complete sequences of the HBC3 chromosome and all associated plasmids.

Identification of the bacteriocin gene cluster and analysis of plasmids in the HBC3 genome sequence

The bacteriocin synthesis gene clusters in the HBC3 genome sequence were searched using the website versions BAGEL4 (http://bagel4.molgenrug.nl)49 and AntiSMASH bacterial version (https://antismash.secondarymetabolites.org)50. The plasmids that carried bacteriocin genes were subjected to nucleotide BLAST to identify gene functions and to compare them with publicly available plasmids. Genomic information on pSc1516939_1 (CP145210.1), S34-1 plasmid unnamed 1 (CP040733.1), 357 plasmid unnamed 1 (CP077879.1), 358 plasmid unnamed 1 (CP077876.1), 356 plasmid unnamed 1 (CP077881.1), pD4-19 (ON936820.1), 359 plasmid unnamed 1 (CP077934.1), and pSKB0123-1 (AP031468.1) was selected and compared with HBC3-plasmid pHBC3_1 using BLAST Ring Image Generator (BRIG)51. For the comparison of bacteriocin gene clusters, genomic information on the S34-1 tcl operon (CP040733.1) and capidermicin operon of CIT060 (MN234131.1) was obtained, and the comparisons among the three clusters were visualized using Easyfig52.

Purification and identification of bacteriocins produced by HBC3

We purified two bacteriocins produced by the HBC3 strain by two different methods. To purify capidermicin, we used a modified method described previously16. A 400 ml overnight culture of HBC3 was centrifuged at 5500 rpm for 20 min. Macro-Prep cation exchange resin (1.5 ml) (Bio-Rad, USA) was added to the supernatant, and the mixture was stirred for 24 h. The resin was collected in an open column and washed three times with 10 ml of 25 mM ammonium acetate (pH 7.5). To elute the bacteriocin, the resin was treated with 1000 µl of 5% acetic acid. This elution process was repeated ten times. Each fraction was completely evaporated and dissolved in 30 µl of distilled water. The antibacterial activity of each solution was tested against S. aureus MW2 and B. coagulans JCM2257. In detail, 3.5 ml of prewarmed TSB soft agar (1%) containing the indicator strains was poured onto TSA plates. After 5 min, 3 µl of each solution was spotted onto the plates. The plates were incubated overnight at 37 °C, and growth inhibition was verified. Samples exhibiting antibacterial activity were subjected to high-performance liquid chromatography (HPLC) using an Octadecyl C18 column. The column was equilibrated with 0.1% TFA in water before sample injection. Elution was performed with a gradient of 0–50% acetonitrile from 0 to 20 min and 50–100% acetonitrile from 20 to 30 min at a flow rate of 1.0 mL/min.

To purify the thiopeptide micrococcin P1 (MP1), butanol extraction was performed as described previously24. The butanol extracts were applied to an HPLC system with a gradient of 35–65% acetonitrile from 0 to 20 min, 65–95% acetonitrile from 20 to 28 min, and 95% acetonitrile from 28 to 30 min at a flow rate of 1.0 ml/min, as described elsewhere22. Purified samples with antibacterial effects were subjected to ESI‒MS analysis using an LTQ Orbitrap XL (Thermo Fisher Scientific, USA). The mass spectra of the samples were obtained by positive mode electrospray ionization (ESI+, [M+ n4-H]4+) and analyzed using Xcalibur QUAL Browser software (Thermo Fisher Scientific, Bremen, Germany). PEAKS X (Bioinformatics Solutions Inc., Waterloo, ON, Canada) was used for MS/MS analysis to determine the peptide identities.

Isolation of the plasmid-cured strain

The bacteriocin-synthesizing gene-carrying plasmid was generated from the HBC3 strain using a modified method described previously53. HBC3 was grown aerobically at 37 °C. When the OD660 reached 0.4, ethidium bromide was added to a final concentration of 3.6 µg/ml, followed by incubation at 42 °C for 16 h. The appropriate dilutions of the cultures were plated on TSA. After the plates were incubated at 37 °C for 24 h, 200 colonies were picked and replated on normal TSA and TSA overlaid with TSB-Tween 80 soft agar (3.5 ml, 0.8% agar, 0.1% Tween 80) containing 108 C. propinquum HBC1 cells, which presented increased susceptibility to S. capitis HBC3. After incubation for 24 h at 37 °C, a colony showing no inhibition zone was isolated as a plasmid-cured strain. Deletion of the plasmid was verified by PCR using specific primers (Forward: agtaaagttatcagtgct and Reverse: ttatttgatacctactgc) designed for the detection of the capidermicin gene.

Direct assay

An overnight culture of S. capitis (3 µL) was spotted on TSA and incubated at 37 °C for 24 h. Then, 3.5 mL of prewarmed half-strength TSB soft agar (0.75%) containing the indicator strain (107 cells/mL) was poured over the TSA plate and incubated at 37 °C for 24 h.

After incubation for 24 h, the diameter of the halo (halo size) was measured. The exact halo size was calculated from the mean value of three independent halos.

Determination of the MIC

The MIC was measured by the microdilution method as described previously16. A total of 1 × 105 S. capitis HBC3, S. aureus MW2, B. coagulans JCM2257, and E. faecium VRE1 cells were inoculated into 100 µL of TSB (for capidermicin and micrococcin P1) or Mueller Hinton Broth (for erythromycin, tetracycline, gentamicin, chloramphenicol, daptomycin and vancomycin). For daptomycin, 50 µg/mL Ca++ was added to the medium. Erythromycin, chloramphenicol and tetracycline were purchased from Wako Pure Chemical Corporation, Osaka, Japan. Gentamicin was purchased from Nacalai Tesque, Kyoto, Japan. Daptomycin was purchased from Tokyo Chemical Industry Co., Ltd., Tokyo, Japan. Vancomycin was purchased from Sigma-Aldrich, St. Louis, MO, USA. The MICs were determined after incubation for 24 h at 37 °C. The experiment was carried out in triplicate.

Spot on lawn assay

A total of 3.5 ml of prewarmed semiconcentrated TSB soft agar (1%) containing bacterial cells (107 cells) was poured onto TSA plates. Two microliters of capidermicin or micrococcin P1 was then spotted onto the plate; after incubation at 37 °C for 16 h, the growth inhibition zones were measured using ImageJ analysis.

Growth curve measurement assay

The growth-inhibitory effects of purified MP1 and capidermicin were evaluated in a 96-well microtiter plate against B. coagulans JCM2257. Bacterial cultures (1 × 10⁵ cells) were inoculated into 96-well microtiter plates with a Nunclon Delta surface (Thermo Fisher Scientific, MA, USA) containing MP1 or capidermicin at final concentrations of 1/2 MIC or 1/4 MIC. The optical density at 660 nm (OD₆₆₀) was measured at 10-minute intervals over 24 h using a microplate reader (SpectraMax iD3, Molecular Devices, San Jose, CA, USA). All the experiments were performed in triplicate. To assess the combined effects of MP1 and capidermicin, a checkerboard assay was performed against only B. coagulans JCM2257 in the same microtiter plate format. MP1 and capidermicin were combined in a two-dimensional dilution matrix, with serial twofold dilutions of each compound array along the X- and Y-axes. Bacterial growth under each combination was monitored, and interactions were evaluated on the basis of the fractional inhibitory concentration index (FICI).

Coculture

We performed a coculture assay by modifying a previous method17. Overnight cultures of S. capitis HBC3 wild-type (WT), HBC3 plasmid-cured mutant, S. aureus MW2 and E. faecium VRE1 strains were adjusted to an OD660 of 1.0, and the bacterial culture was diluted 10-fold with TSB. HBC3 WT was mixed with each indicator strain as follows: 167 µL of HBC3 and 33 µL of MW2; 100 µL each of HBC3 and VRE1; and 18 µL of HBC3 with 182 µL of B. coagulans. Subsequently, 20 µL of each mixture was spotted onto TSA plates and incubated aerobically at 37 °C for 24 h. The bacterial colonies growing on agar plates were scraped and suspended in 5 mL of TSB. All suspensions were adjusted to an OD of 0.8 and appropriately diluted. One hundred microliters of each dilution was plated onto normal TSA plates and incubated overnight at 37 °C. The number of colonies was counted to quantify the number of surviving bacterial cells under each condition. The assay was performed three times independently.

Statistical analysis

Student’s t test was used for comparison in Fig. 6, one-way ANOVA was used for comparison in Fig. 1B and S2, and Kruskal-Wallis test was used for comparison of Fig. S1 via GraphPad Prism (GraphPad Software, San Diego, CA, USA).