All animal investigations were performed in accordance with the Guidelines for the Care and Use of Laboratory Animals of the Ministry of Food and Drug Safety of Korea, and were approved by the Animal Care and Use Committee of the Korea Atomic Energy Research Institute (Jeongeup Si, Korea; IACUC protocol no. 2014–023).

The bacterial strains examined in the present study are listed in Table I. The streptococcal and staphylococcal strains were grown at 37°C in Todd-Hewitt broth (BD Biosciences, Franklin Lakes, NJ, USA) supplemented with 0.5% (w/v) yeast extract and Tryptic-soy broth (BD Biosciences), respectively. Purified melittin was purchased from Sigma-Aldrich (St. Louis, MO, USA). Synthetic melittin (GIGAVLKVLTTGLPALISWIKRKRQQ) was chemically synthesised by A&PEP Co., Inc. (DaeJeon, Korea).

Controlled colonies of natural honeybees (Apis mellifera L.) were maintained at room temperature at the National Academy of Agricultural Science (Suwon, Korea). In brief, a bee venom collector apparatus (Chunglin Biotech, Ansan, Korea) was placed on the hive, and the bees that landed on the apparatus were subjected to an electric shock sufficient to cause the bees to 'sting' a glass plate from which dried bee venom was harvested. The collected venom was dissolved in distilled water, centrifuged at 12,000 × g for 10 min to remove insoluble materials, and stored in a refrigerator until further use (15–17).

Bacteria were harvested at the early log phase (A₆₀₀=0.5) and suspended in phosphate-buffered saline (PBS) at ~10⁸ to 10¹⁰ CFU/ml. Subsequently, the bacterial samples were incubated with the indicated concentrations of bee venom or melittin at 25°C for 30 min, and surviving bacteria were evaluated using a plate counting method, as described previously (18). Briefly, samples were serially diluted in PBS and plated onto blood agar (Kisan Bio, Suwon, Korea). Following a 16 h incubation at 37°C, the number of surviving bacteria was counted.

To determine the minimum inhibitory concentration (MIC), the present study used a micro-dilution broth method, according to the recommendations of the National Committee for Clinical Laboratory Standards (19). In brief, the cells of the experimental bacterial strains were collected in the logarithmic phase of growth, suspended in 30 mM phosphate buffer (pH 7.0) with 60 mM NaCl, and adjusted to an A₆₀₀ of 0.3 arbitrary units (1×10⁵ cells/ml). The bee venom and the melittin samples were dissolved in 10 mM phosphate buffer (pH 6.0) with 130 mM NaCl and 0.2% (w/v) bovine serum albumin prior to serial dilution. Sample aliquots (10 µl) were mixed with the diluted bacterial suspensions (190 µl) followed by incubation for 20 h at 37°C. Bacterial growth was determined by measurement of the A₆₅₀ levels using a VICTOR™ X3 ELISA reader (PerkinElmer, Inc., Waltham MA, USA).

The cytotoxic effects of bee venom and melittin on cultured MCF7 cells were evaluated using a Cell Counting Kit-8 (CCK-8; Dojindo Molecular Technologies, Inc., Gaithersburg, MD, USA). The cells were seeded at a density of 5×10³ cells/200 µl/well into wells of 96-well round-bottomed plates and allowed to grow for 24 h at 37°C, followed by incubation with bee venom or purified synthetic melittin for 6 h at 37°C. The culture supernatants (100 µl quantities) were harvested and mixed with 10 µl aliquots of CCK-8 solution. Following 3 h incubation at 37°C, the optical densities at A₄₅₀ were measured using the VICTOR™ X3 ELISA reader (PerkinElmer, Inc.).

Mouse infection with Staphylococcus aureus was performed, as described previously (20). Bacteria of the USA300 strain (American Type Culture Collection, Manassas, VA, USA) were spectrophotometrically (OPTIZEN POP; Mecasys Co., Ltd., Daejeon, Korea) adjusted to the desired concentration prior to injection, and bacterial numbers were confirmed via serial dilution and Tryptic soy agar plating. The cultured USA300 bacteria were pelleted, washed and suspended in PBS at 0.5×10⁸ CFU/ml. Mice (7-week-old males) of the CD1 strain were obtained from Oriental Bio, Inc. (Seongnam, Korea), with 10 animals per treatment group. The mice were infected with the USA300 strain (200 µl) via intraperitoneal (i.p.) injection, followed by i.p. injection of 100 µl bee venom or purified melittin 1 h later. The infected animals were monitored every 3 h for up to 36 h. The mice were housed in controlled conditions: Temperature, 23±2°C; humidity 55±10%; light between 07:00 and 19:00. Each group was housed seperately. All animal experiments in the present study adhered to institutional guidelines upon review of the experimental protocol, and were approved by the Institutional Biosafety Committee and the Institutional Animal Care and Use Committee of Korea Atomic Energy Research Institute.

CD1 mice (7-week old; 3 mice/group) were used to examine skin infection. Following the induction of general anesthesia, the dorsal hair was electrically shaved and the skin was cleaned with 70% (v/v) ethanol. Skin infection was induced via subcutaneous inoculation of 50 µl volumes of USA300 suspension (10⁶ CFU/ml) in PBS. Subsequently, bee venom, melittin (purified or synthetic; 100 µg in 80 µl PBS), or sterile PBS was applied once daily to each surface lesion. Lesion progression was monitored at 24 h intervals for 10 days by measuring the lesion dimensions with callipers (Jeung Do B&P Co., Ltd., Seoul, Korea), and capturing images using a digital camera (WB5500; Samsung, Seoul, Korea).

Data are presented as the mean ± standard deviation. Statistical analysis was conducted using GraphPad InStat software version 5 (GraphPad Software, Inc., La Jolla, CA, USA). The statistical significance of between-group differences was evaluated using two-tailed Student's t-test. P<0.05 was considered to indicate a statistically significant difference.

The present study examined the antibacterial activities of bee venom against the Streptococcus agalactiae, Streptococcus gordonii, Streptococcus pneumonia, Streptococcus epidermidis, Streptococcus bovis and Staphylococcus aureus Gram-positive bacteria. As shown in Fig. 1, when all the bacterial strains were treated with the indicated concentrations of bee venom for 30 min, concentration-dependent death of the bacteria was evident. At venom concentrations between 1.25 and 12.5 µg/ml, bacterial viability decreased by >90%. The MIC values of the bee venom ranged between 1.56 and 12.5 µg/ml (Table II). Notably, the USA300 antibiotic-resistant Staphylococcus strain had the lowest observed MIC (1.56 µg/ml).

The present study further examined the antibacterial activities of bee venom against three MRSA clinical isolates. As shown in Fig. 2, the viabilities of all three strains decreased markedly upon treatment with bee venom for 30 min, and no bacteria survived incubation with 100 µg/ml venom. The MIC values for the three MRSA strains ranged between 0.78 and 3.13 µg/ml (Table II). Notably, the methicillin-sensitive Staphylococcus aureus strains (Mu50, ISP479C, PS735, PS736 and PS737) were less susceptible to bee venom (MIC=3.13–12.5 µg/ml), compared with the MRSA strains (Table II), suggesting that bee venom contains antimicrobial molecules, which specifically target MRSA strains.

To measure the cytotoxicity of bee venom, human epithelial cells were incubated with venom for 24 h and cell viabilities were measured using an MTT assay. As shown in Fig. 3, bee venom was not cellulotoxic at a concentration of 0.4 µg/ml. In addition, the administration of bee venom in vivo at up to 20 mg/kg i.p., caused no signs or symptoms of toxicity in the CD1 mice (data not shown).

The i.p injection of 1×10⁸ CFU of the USA300 strain into mice caused bacteraemia and mortality rates of 100% within 18 h. When the USA300-infected mice were administered with 1.25 or 2.5 mg/kg bee venom at the time of infection, no protective effect was evident (data not shown). A low dose of USA300 (1×10⁷ CFU per mouse) was injected 1 h following the administration of PBS or bee venom. Notably, all the mice died 18 h following the injection of USA300 with bee venom, whereas only five mice of the control group had died by 24 h post-infection (Fig. 4A). These data demonstrated that, although bee venom exhibited a marked antimicrobial effect in vitro, in vivo administration enhanced MRSA propagation and infection.

In addition, the present study examined the protective effect of bee venom in a staphylococcal skin infection model (Fig. 4B). When USA300 was inoculated intradermally and the areas of infected skin treated with PBS or bee venom (10 µg) once daily, the abscesses formed by USA300 were 21.3±4.8 and 18.8±6.8 mm in diameter in the PBS and bee venom groups, respectively, by day 5, and no significant difference was observed even following 10 days of venom treatment.

Bee venom is a complex mixture of proteins, peptides and low-molecular-weight materials. The principal components of the venom are phospholipase A2 (PLA2; 10–12%, w/w) and the melittin peptide (40–48%, w/w). The results of the present study confirmed and extended the previous results, demonstrating that melittin and PLA2 induced death in a broad range of bacteria, including MRSA strains. As shown in Fig. 5A, treatment of the USA300 and MRSA2 strains with PLA2 did not affect cell viability, whereas the viabilities of the MRSA strains treated with purified melittin decreased to levels comparable to those observed when bee venom was used. To examine whether melittin and PLA2 acted synergistically, two MRSA strains were treated with melittin admixed with PLA2 at various concentrations. When the USA300 and MRSA2 strains were treated with melittin alone (25 µg/ml), the total number of bacteria decreased by ~2.5–3 log CFU (Fig. 5B). However, when the cells were treated with melittin (25 µg/ml) in combination with various concentrations of PLA2, similar results were observed, indicating that PLA2 did not act synergistically with melittin to cause bacterial cell death.

Subsequently, the present study confirmed that synthetic melittin exhibited an antimicrobial activity similar to that of purified melittin. Initially, the toxicities of the two forms of melittin towards human epithelial cells were determined, as described above. As shown in Fig. 6A, synthetic melittin (99.2% pure) was ~25% less toxic than the 'purified' melittin (93% pure). However, the antibacterial activities of the two preparations against the MRSAs were comparable (Fig. 6B).

The present study also investigated whether melittin can protect against MRSA skin infections. USA300 bacteria (1×10⁷ CFU/mouse) were injected intradermally into CD1 mice, which were administered with either PBS, or purified or synthetic melittin (10 µg) 1 h post-infection. As shown in Fig. 7, abscesses in the PBS-treated group gradually increased in size to attain a diameter of 22±6.3 mm by day 5. When the infected areas were treated with purified or synthetic melittin once daily for 4 days, the diameters of the abscesses were significantly lower than those measured in the control group.

In addition, the protective effect of melittin was investigated in a model of MRSA bacteraemia (Fig. 8). When a high dose of USA300 was injected i.p., all the mice died following treatment with either PBS or 2.5 mg/kg melittin after 24 h. However, when the infected mice were injected with 5 mg/kg melittin 1 h post-infection, 50% of the mice survived >24 h.