Lactic Acid Bacteria and Bifidobacteria-Inhibited Staphylococcus epidermidis

Introduction. Lactic acid bacteria (LAB) and bifidobacteria are able to produce antimicrobial compounds to inhibit opportunistic-wounding skin pathogen. The antimicrobial compounds produced are organic acids, putative bacteriocin, hydrogen peroxide, and diacetyl. Staphylococcus epidermidis is well-known as an opportunistic wounding skin pathogen in wound infections related to implanted medical devices. Objective. To screen 87 strains of LAB and 3 strains of bifidobacteria for antimicrobial activity against Staphylococcus epidermidis. Additionally, this study sought to determine and quantify types of antimicrobial compounds produced by LAB. Materials and Methods. Inhibitory activity of LAB and bifidobacteria on S. epidermidis was assessed with the spectrophotometric method using a 96-well microplate reader. Characterization of cell-free supernatant (CFS) was done using analytical methods. Lactobacillus fermentum (collected by the Bioprocess Department at the Universiti Sains Malaysia [BD]) 1912d, Lactobacillus casei BD 1415b, Lactobacillus fermentum BD 8313a, Pediococcus pentosaceus BD 1913b, and Weissella cibaria (collected by the Food Technology Department at the Universiti Sains Malaysia [FTDC]) 8643 with high percentage of inhibition (P < 0.05), ranging from 73.7% to 88.2%, as compared to the control, were selected for subsequent analyses. Upon neutralization, the antimicrobial activity showed a drastic drop (P < 0.05) in the percentage of inhibition. Concentrations of the inhibitive metabolites were produced in varying amounts and were strain dependent. Results. Results demonstrated that lactic acid in all strains was produced in a more predominant amount than acetic acid. Protein concentration production ranged from 0.081-0.215 mg/mL. L. fermentum BD 1912d yielded as much as 0.014 mg/mL hydrogen peroxide, which was significantly higher than other strains studied. Diacetyl was produced in a higher concentration by Weissella cibaria FTDC 8643 at 2.884 ng/mL; the lowest concentration of 0.465 ng/mL was produced by Ped. pentosaceus BD 1913b. Conclusion. Antimicrobial metabolites from CFS of lactic acid bacteria were effective in repressing the growth of opportunistic wounding dermal pathogen Staphylococcus epidermidis.

  Staphylococcus epidermidis is a commensal organism having a benign relationship with the host on human skin. However, S. epidermidis is also an opportunistic pathogen and is ranked first in nosocomial¹ and implant-based infections such as peripheral or central intravenous catheters, prosthetic heart valves, and prosthetic joints.² Infections of S. epidermidis can occur at the device insertion site or in wounds. Due to their abundance, there is a high probability of device contaminations during insertion through the introduction of bacteria from the skin of the patient or from health care personnel. At least 22% of surgical site infections and cardiac device infections are caused by S. epidermidis.¹ The occurrence of infection by this opportunistic wounding dermal pathogen has been attributed to the ability of S. epidermidis to adhere to biomaterial and to wound sites, causing the wound healing process to slowdown.³ Although S. epidermidis infections rarely develop into life-threatening diseases, their ubiquity and resisitance to treatment has led to an increased burden for the public health system, costing $2 billion annually in the United States.¹ Similar to S. aureus, S. epidermidis also exhibits its virulence through biofilm formation, intercellular aggregation, protective exopolymers, toxins, and exoenzymes.¹ The formation of biofilm has complicated various antibiotic treatments (methicillin, rifamycin, gentamycin, and erythromycin) at infection sites.¹S. epidermidis contains abundant surface proteins, bifunctional adhesion and autolysin, and the biofilm associated protein (Bap), all of which are likely to contribute to the hydrophobicity of the cell surface and promote the adhesion to abiotic surfaces such as catheters.¹ A fibrinogen-binding protein of S. epidermidis is necessary and sufficient to promote S. epidermidis adhesion to fibrinogen in vitro and promotes central intravenous catheter-associated infection in vivo.¹ Due to increasing resistance of the pathogen against antibiotics, the use of lactic acid bacteria (LAB) and bifidobacteria to inhibit pathogens is seen as a natural and nontoxic alternative.   

  Lactic acid bacteria and bifidobacteria are Gram-positive bacteria which have been associated with various health effects ranging from digestive health to metabolic diseases and, recently, dermal health as well.^4,5 Most of these health benefits are related to the ability of LAB and bifidobacteria to produce a broad range of bioactive metabolites and acids in situ.⁶ Some antimicrobial metabolites exerted by LAB that are crucial for skin health include lactic and acetic acids, bacteriocin, hydrogen peroxide, diacetyl, and carbon dioxide.⁷ However, bifidobacteria do not produce hydrogen peroxide or bacteriocins.⁷    

  Organic acids (lactic and acetic acids) are the main inhibitors among the antimicrobial compounds produced via fermentation. The acidic nature provides a suitable environment to prevent and inhibit the growth of many pathogenic and spoilage microorganisms. Lactic acid inhibits the pathogens through the penetration of the undissociated form across the membrane which interferes with the metabolic functions of the pathogen. The decrease in the intracellular pH causes dissipation of the membrane and leads to membrane disruption.⁸

  Bacteriocins exhibit antimicrobial activity towards closely related Gram-positive bacteria.⁸ As reviewed by Savadogo et al,⁹ bacteriocins generally consist of antimicrobial peptides that exhibit hydrophobic or amphiphilic properties to target bacterial membrane.⁹ They form pores on the membrane of the Gram-positive pathogen which increase membrane permeability and cause rapid cell death.¹⁰ On the other hand, hydrogen peroxide performs its inhibitory effect by oxidizing lipids in the membrane layer which increases membrane permeability, leading to denaturing of basic molecular structures of cellular proteins and subsequently causes cell destruction.¹¹

  Bacterial cultures. All strains of LAB, bifidobacteria, and S.epidermidis were obtained from the Culture Collection Centre of Bioprocess Technology Division, School of Industrial Technology, Universiti Sains Malaysia (Gelugor, Penang, Malaysia). Lactic acid bacteria and bifidobacteria cultures were propagated in sterile de Man-Rogosa-Sharpe (MRS) broth (Biomark, Pune, India) while S.epidermidis was propagated in Tryptone Soy Broth (TSB) (Biomark, Pune, India) successively for 3 consecutive activation times of 24 hours each. The propagated cultures were incubated at 37°C for 24 hours¹³ using 10% (v/v) inoculum. The activated cultures were kept in the fridge and subcultured for further analysis. Stock cultures were suspended in sterile 40% (v/v) glycerol and stored at -20°C until further use. The origins of the LAB strains are listed according to category in Table 1, 2, 3 and 4.

  Preparation of standardized cell-free supernatant. To prepare the cell-free supernatant (CFS) for antimicrobial screening, standardization was carried out by adjusting the activated culture to an optical density (OD) of 1.0 using sterile MRS at 600 nm using an ultraviolet-visible (UV-vis) spectrophotometer (Hitachi U-1900, Hitachi High-Tech, Tokyo, Japan). This was followed by centrifugation at 1100 x g for 15 minutes at 4°C, after which the supernatant was collected as CFS.

  Antimicrobial screenings. Antimicrobial screenings with non-neutralized and neutralized CFS were completed using a 96-well microplate reader. Standardized S.epidermidis was prepared by adding sterile TSB to adjust the culture to an OD of 0.3 at 595 nm using an UV-vis spectrophotometer. Neutralization with 1.0 M NaOH or 1.0 M HCL to adjust the pH to 7.0 was followed by filter sterilization using a 0.22 µm syringe filter. Test samples were prepared by adding equal volume of standardized CFS and standardized pathogen. de Man-Rogosa-Sharp broth without LAB, and TSB without S.epidermidis, were used as a blank with the volume of 1:1 ratio; whereas the negative control used was MRS broth without LAB and with OD-adjusted pathogen. Absorbance was measured using a microplate reader (Dynamica, Newport Pagnell, UK). Percentage of inhibition was calculated according to the formula below: % Inhibition = ([Control OD - Sample OD]/[Control OD]) ×100%

  Determination of pH. The pH of the LAB culture before and after incubation was measured with a pH meter with automatic temperature compensation (ATC) probe (CyberScan pH 510, Eutech Instruments Pte Ltd, Singapore).

  Antimicrobial activity of protease-treated cell-free supernatant. Neutralized CFS prepared was treated with enzyme protease K (Sigma-Aldrich, St. Louis, MO) and trypsin (Sigma-Aldrich, St. Louis, MO) followed by incubation at 37°C for 1 hour¹³ and heat-inactivation at 100°C for 3 minutes. The antimicrobial effect was measured as absorbance using a microplate reader (Dynamica, Newport Pagnell, UK) at a wavelength of 595 nm.

  Antimicrobial activity of precipitated protein fraction. Protein fraction collected in neutralized CFS was precipitated using 40% (w/v) ammonium sulfate precipitation.¹⁴ Ammonium sulphate (Merck, Whitehouse Station, NJ) was added slowly and left overnight at 4°C. Pellets collected from the centrifugation (8590 x g for 1 hour at 4°C) were resuspended with 0.05 M sodium phosphate buffer, pH 7.0, filtered with a 0.22 µm cellulose acetate syringe filter to collect precipitated protein fraction. The antimicrobial activity was observed by dispensing the precipitated protein fraction and standardized S.epidermidis culture into a 96-well microplate at a volume of 1:1 ratio. The absorbance of the samples was measured at 595 nm.

  Determination of organic acids. The concentration of organic acids produced by LAB strains was determined by high performance liquid chromatography (HPLC) (Shidmazu, Kyoto, Japan).¹⁵ The HPLC system consists of Aminex HPX-87H Column (300 x 7.8 mm, Bio-Rad Laboratories, Hercules, CA) and the temperature for the column was maintained at 35°C. Samples were filtered through a 0.22 µm cellulose acetate syringe filter into HPLC vials. Organic acids were detected at the wavelength of 215 nm. Degassed mobile phase (0.004 M H2SO4) was used at a flow rate of 0.6 mL/min. Standards used for the analysis were HPLC-grade acetic and lactic acid (Sigma-Aldrich, St. Louis, MO).

  Quantification of protein. Protein concentration was quantified using the Bradford protein assay method.¹⁶ Bovine serum albumin was used to construct a standard curve. The absorbance of each sample was measured with a microplate reader at 595 nm. 

  Quantification of hydrogen peroxides. Spectrophotometric method using o-dianisidine and horseradish peroxidase was used to quantify hydrogen peroxide produced.¹⁷ The absorbance was measured at 430 nm with a microplate reader. A standard curve was established using freshly prepared hydrogen peroxide.

  Quantification of diacetyl. Colorimetric reaction was used to determine the concentrations of diacetyl using creatine and a-naphthol.^11,15 The absorbance was measured at 525 nm (Shimadzu Corp, Kyoto, Japan). Freshly prepared diacetyl (Merck, Whitehouse Station, NJ) was used to perform a standard curve. 

  Data analysis was performed using IBM SPSS Statistics for Windows, version 20.0 (IBM Corp, Armonk, NY).One-way analysis of variance (ANOVA) was performed to analyze the statistical difference between sample means. Tukey’s test was used to compare multiple means with a level of significance at a = 0.05. All data presented are mean values obtained from triplicates (n = 3), unless stated otherwise. 

  Screening of lactic acid bacteria for antimicrobial activity. The results of antimicrobial screening on 5 different categories of LAB are presented in Figures 1, 2, 3-4. In general, all 90 strains of LAB and bifidobacteria showed inhibition effect on the growth of S. epidermidis with the percentage of inhibition ranging from 2.3%-88.2%. Lactobacillus fermentum BD 8313a, Pediococcus pentosaceus BD 1913b, Lactobacillus casei BD 1415b, Weissella cibaria FTDC 8643, and Lactobacillus fermentum BD 1912d showed higher growth inhibition among the 90 strains studied (P < 0.05) and were thus selected for further evaluation.

   Effect of neutralization on antimicrobial activity. In this study, the antimicrobial activity of 5 selected LAB strains on S. epidermidis was evaluated using neutralized CFS (Figure 5). It was found that neutralized CFS exhibited less inhibition on the growth of the dermal pathogen when compared to the non-neutralized CFS (P < 0.05). A drastic decrement in the percentage of inhibition was observed to an extent of lower than 20% in comparison to the non-neutralized CFS, ranging from 13.04% to 18.53%. L. fermentum BD 8313a exhibited higher reduction (as high as 74.22%) compared to other strains studied.

  Effect of protease treatment. Current results showed that the percentage of inhibition of protease-treated CFS (Figure 5) of all strains was significantly (P < 0.05) decreased compared to the non-neutralized CFS. While comparing the percentage of inhibition of protease-treated CFS to neutralized CFS in each strain, a slight decrease in the percentage of inhibition was observed in the protease-treated CFS in Ped. pentosaceus BD 1913b, L. fermentum BD 8313a, and W. cibaria FTDC 8643; whereas L. fermentum BD 1912d and L. casei 1415b showed a slight increase in the percentage of inhibition.

  Effect of ammonium sulphate precipitation. From the precipitated protein CFS using 40% (w/v) ammonium sulphate, better inhibitions were established by 2 strains, namely L. fermentum BD 8313a and W. cibaria FTDC 8643. The bar graph in Figure 5 illustrates that W. cibaria FTDC 8643 exerted a higher percentage of inhibition, which was 11.83% higher (P < 0.05) than L. fermentum BD 8313a. Also, these 2 strains showed significant difference (P < 0.05) in pathogen growth inhibition as compared to the protease treated CFS. These 2 strains (L. fermentum BD 8313a and W. cibaria FTDC 8643) demonstrated an incremental increase in the percentage of inhibition, as much as 10.05% and 16.07%, respectively.   

 Lactic acid and acetic acid. Concentrations of organic acids, namely lactic acids and acetic acids produced by the 5 selected LAB strains, are shown in Table 5. From the data obtained, it was found that the concentrations of the acids were strain-dependent. In comparison between concentration of lactic acid and acetic acid, all strains studied showed higher concentration of lactic acid than acetic acid. L. casei BD 1415b showed a significantly higher (P < 0.05) concentration in lactic acid than other strains studied. The pH changes of MRS broth upon fermentation for 20 hours at 37°C were measured (Table 5). It was found that Ped. pentosaceus BD 1913b and L. fermentum BD 8313a showed significantly higher (P < 0.05) values in pH changes (1.37 and 1.36, respectively). 

  Protein. Protein was precipitated with 40% ammonium sulphate and its concentration was quantified by Bradford protein assay. Data obtained showed that all strains studied contained varying concentrations of protein, ranging from 0.081 mg/mL to 0.215 mg/mL. Table 6 shows that Ped. pentosaceus BD 1913b contained 0.215 mg/mL higher (P < 0.05) protein concentration when compared to the other strains studied. Hydrogen peroxide. All strains studied produced hydrogen peroxide with different concentrations (Table 6). Data obtained showed that hydrogen peroxide concentrations produced by all strains were significantly different (P < 0.05), except Ped. pentosaceus BD 1913b and L. fermentum BD 8313a. The concentration of hydrogen peroxide produced ranged from 0.004 mg/mL to 0.014 mg/mL (Table 6). The current study showed that L. fermentum BD 1912d produced a higher (P < 0.05) concentration of hydrogen peroxide.

  Diacetyl. Data showed all 5 strains of lactic acid bacteria studied produced diacetyl. Concentrations of diacetyl produced ranged from 0.465 ng/mL to 2.884 ng/mL. When compared among the strains studied, W. cibaria FTDC 8643 and L. casei BD 1415b contained significantly higher (P < 0.05) concentrations of diacetyl at 2.884 ng/mL and 2.171 ng/mL, respectively.

  S. epidermidis is an opportunistic wounding dermal pathogen which causes infections on indwelling medical devices and surgical wounds. Infections are easily caused by accidental contamination during surgery and insertion of devices, as S. epidermidis is a permanent and ubiquitous colonizer of human skin^3,18; however, complications in treatment of wounds occurred in the presence of antibiotic-resistant strains.¹

  Lactic acid bacteria and bifidobacteria are known to be probiotic agents which were traditionally used in fermented dairy products; their safety is asserted as Generally Recognized As Safe (GRAS) as determined by the U.S. Food and Drug Administration.¹⁸ They are used as biopreservatives to increase the shelf life of food products by fermentation and exert antimicrobial effects to enhance food safety. A recent review by Tan et al¹¹ showed they were effective against foodborne pathogens such as K. pneumonia, P. aeruginosa, E. coli, and S. aureus. Lactic acid bacteria and bifidobacteria were also found to be densely populated in the human intestinal tract,¹⁸ where they are effective against pathogens of the gastrointestinal tract, such as Clostridium botulinum, Bacillus cereus, and S. aureus, which cause diarrhea and other intestinal diseases.¹⁹ The production of lactic acid and acetic acid was found to have an important role in the preservation of foods and fermented product as well as inhibiting growth of dermal pathogens such as P. aeruginosa and S. aureus.¹⁵ Thus, it would be possible to inhibit non-gut pathogens such as Gram-positive opportunistic-wounding dermal pathogen, S. epidermidis.

  Lactic acid bacteria and bifidobacteria have been reported to exert antagonistic action against Gram-positive and Gram-negative pathogenic microorganisms by in situ production of antimicrobial compounds. In this study, the LAB strains performed better than bifidobacteria in inhibiting S. epidermidis. Lactic acid bacteria produced antimicrobial metabolites including lactic acid and acetic acid, bacteriocin, hydrogen peroxide, and diacetyl, which are mostly produced in the extracellular extracts.^6,8,11,15,20 However, hydrogen peroxide and bacteriocin were not found in bifidobacteria.⁷ Therefore, it can be postulated that all the metabolites produced by LAB act synergistically to be more effective than bifidobacteria, which only produced organic acid and diacetyl. Hence, in this study, extracellular extracts of LAB strains were employed as CFS. The antimicrobial activity was strain dependent. In this study, the strains Pediococcus pentosaceus BD 1913b, Lactobacillus casei BD 1415b, Lactobacillus fermentum BD 1912d, Lactobacillus fermentum BD 8313a, and Weissella cibaria FTDC 8643 from different groups of origin showed higher (P < 0.05) growth inhibition among the 90 strains studied and were selected for further evaluations.

  Lactic acid and acetic acid are the primary antimicrobial compound produced by LAB⁹ via carbohydrate catabolism.^11,16 The results obtained showed the high percentage of inhibition was mainly contributed by organic acids. This was later proven by neutralization of CFS where a drastic drop (P < 0.05) was observed in the percentage of inhibition. Hence, it can be postulated that lactic acids and acetic acids were the main contributor in the antimicrobial activity. The varying concentrations of lactic acid and acetic acid were due to different modes of fermentation applied. In this study, the 5 selected LAB strains studied were postulated to employ heterofermentation as compounds other than lactic acid, including acetic acid, hydrogen peroxide, and diacetyl, were also produced. Lactic acid is produced as the major metabolite of lactic acid bacteria. This was clearly observed from the results presented where lactic acid was produced in amounts ranging from 0.0322 mmol/mL to 0.0345 mmol/ mL. Though the concentration of acetic acid produced was low, it was reported that at a low concentration of 0.0083 mmol/mL, it was able to eliminate Pseudomonas aeruginosa from burns and soft tissue wounds within 2 weeks.¹⁵ The pH changes shown were caused by the accumulation of the lactic acids.

  Based on the high (P < 0.05) percentage of inhibition obtained, it was deduced that bacteriocin partly contributed to the inhibition. The neutralized CFS was treated with enzyme protease to determine whether the bioactive compounds produced were protein-based. Protease treatment was used to confirm if the bioactive compound in the extract was protein-based (Figure 5). Based on the results presented, Ped. pentosaceus BD 1913b, L. fermentum BD 8313a, and W. cibaria FTDC 8643 showed decrement in the growth inhibition. The authors postulated that the bioactive compounds produced were protein-based and responsible for inhibition. Ammonium sulphate precipitation at 40% (w/v) was done to concentrate the protein (bacteriocins-like peptides) presented in the CFS. The protein precipitated CFS of L. fermentum BD 8313a and W. cibaria FTDC 8643 showed significant difference (P < 0.05) in the inhibition of S. epidermidis, indicating that the bioactive compounds were protein-based. Thus, the strains were confirmed to have produced bioactive peptides with bactericidal activity. It is known that in chronic wound fluid, there is an elevated protease level that has a deleterious effects on wound healing.²¹ However, as protease activity is extremely sensitive to pH (protease are inactivated below pH 4),^21,22 based on the current findings, the use of non-neutralized CFS in wound healing could help to eliminate the effect of protease as organic acids were also produced in the extract, which lowered the pH to the range of pH 3.75 to pH 4.10 (this raw data is not shown). The authors also suspect that the protein produced might be able to promote growth of S. epidermidis as shown in the case of L. casei BD 1415b, as a lower percentage of inhibition (P < 0.05) than neutralized and protease-treated CFS was observed.

  Table 6 shows that all strains studied were capable of producing hydrogen peroxide with varying concentration ranging from 0.004-0.014 mg/mL. It has been reported that hydrogen peroxide exerted inhibitory activity on Gram-positive dermal pathogens at a concentration as low as 0.010 mg/mL.¹¹ This indicated that the hydrogen peroxide concentrations produced by several strains in this study were sufficient to exert inhibitory activity.

  Diacetyl is produced through the conversion of a-acetolactate by Pseudomonas aeruginosa-acetolactate synthases.⁸ Results obtained exhibited that different concentrations of diacetyl, ranging from 0.465 ng/mL to 2.884 ng/mL, were generated. The strain W. cibaria FTDC 8643 presented significantly (P < 0.05) higher concentration among the strains studied. It was reviewed that a minimal concentration of diacetyl (as low as 3 x 10⁶ ng/mL) was required.¹⁶ However, the concentrations of diacetyl produced were too low to be effective in inhibiting the growth of S. epidermidis. Hence, a synergistic effect with other antimicrobial metabolites suggested that diacetyl could exert inhibitory effect on the growth of S. epidermidis.

  Lactic acid bacteria can produce antimicrobial compounds to inhibit growth of S. epidermidis when lactic acid and acetic acid are the main inhibitor. Through penetration of the undissociated form of acid molecules across the pathogen’s membrane, pathogens will not be able to adhere to the surface of the biomaterials and the infection sites. Combined with a synergistic effect with other inhibitory compounds, the growth of S. epidermidis will be restrained.

The authors are from the School of Industrial Technology, Universiti Sains Malaysia, Penang, Malaysia.

Address correspondence to:
Min-Tze Liong, PhD
School of Industrial Technology
Universiti Sains Malaysia
11800 Penang
Malaysia
mintze.liong@usm.my

Disclosure: The authors disclose this work received financial support from the Science Fund Grant (305/PTEKIND/613222) provided by the Malaysian Ministry of Science, Technology and Innovation; the FRGS grant (203/PTEKIND/6711239) provided by the Malaysian Ministry of Higher Education; and the USM Universiti Sains Malaysia RU grants (1001/PKIMIA/855006, 1001/PTEKIND/815085) and USM Fellowship provided by Universiti Sains, Penang, Malaysia.