Leon M.T. Dicks, Private Bag X1, Matieland 7602, Stellenbosch, South Africa. E-mail: email@example.com
Aims: To determine the antimicrobial activity of nisin F against Staphylococcus aureus in the respiratory tract.
Methods and Results: The respiratory tract of nonimmunosuppressed and immunosuppressed Wistar rats were colonized with 4 × 105 viable cells of S. aureus K and then treated by administering 8192 arbitrary units (AU) nisin F intranasal. Symptoms of pneumonia were detected in the trachea and lungs of immunosuppressed rats that had not been treated with nisin F. The trachea and lungs of immunosuppressed rats treated with nisin F were healthy. No significant differences were recorded in blood cell indices. The antimicrobial activity of low concentrations nisin F (80–320 AU ml−1) was slightly stimulated by lysozyme and lactoferrin.
Conclusions: Nisin F inhibited the growth of S. aureus K in the respiratory tract of immunocompromised rats. Treatment with nisin F at 8192 AU proofed safe, as the trachea, lungs, bronchi and haematology of the rats appeared normal.
Significance and Impact of the Study: Nisin F is nontoxic and may be used to control respiratory tract infections caused by S. aureus. This is, however, a preliminary study with an animal model and need to be confirmed with studies on humans.
Lysozyme and lactoferrin are the two most abundantly secreted antimicrobial proteins in airway secretions of humans (Brogan et al. 1975; Harbitz et al. 1984; Wilson 2005). These proteins are antimicrobial against S. aureus and have a synergistic effect with nisin (Nattress et al. 2001; Murdock et al. 2007). Studies on the protective effect of bacteriocins produced by lactic acid bacteria against respiratory tract infections are limited. Kruszewska et al. (2004) repressed the growth of methicillin-resistant S. aureus (MRSA) strains in the respiratory tract of mice with mersacidin, a lantibiotic produced by Bacillus sp. HIL Y-85 54728. Similar studies with nisin, a lantibiotic produced by Lactococcus lactis subsp. lactis, did not eradicate S. aureus from the nasal tract of rats (Kokai-Kun et al. 2003).
In a previous paper (De Kwaadsteniet et al. 2008), we described nisin F produced by L. lactis subsp. lactis F10 and reported on the in vitro activity against clinical strains of S. aureus. In the present study, nonimmunosuppressed and immunosuppressed rats were intranasally infected with a clinical strain of S. aureus and then treated with nisin F. Blood cell counts and the histology of lung and trachea tissue from rats in both groups were compared. The safety of nisin F was determined by administering the peptide intranasal to nonimmunosuppressed and immunosuppressed rats.
Materials and methods
Bacterial strains and culture conditions
Lactococcus lactis subsp. lactis F10 was cultured in De Man Rogosa and Sharpe (MRS) broth (Biolab, Biolab Diagnostics, Midrand, South Africa) at 30°C and Staphylococcus aureus K in brain heart infusion (BHI, Biolab) at 37°C. Biard-Parker agar base (Biolab) was used as selective medium for S. aureus K, isolated from the nasal cavity of a patient with acute sinusitis.
Nisin F preparation
Nisin F, produced by L. lactis subsp. lactis F10 (De Kwaadsteniet et al. 2008), was semi-purified by ammonium sulphate precipitation and dialysed according to the method described by Sambrook et al. (1989). The peptide was concentrated by freeze-drying and resuspended in sterile physiological saline. Antimicrobial activity was determined by using the agar-spot test method and expressed as arbitrary units (AU) per millilitre. One AU is the reciprocal of the highest serial two-fold dilution showing a clear zone of inhibition of the indicator strain (Van Reenen et al. 1998). The indicator strain was an 18-h-old culture of S. aureus K (106 CFU ml−1), embedded in BHI soft agar (1%).
Synergistic effect of nisin F with lysozyme and lactoferrin
The antimicrobial activity of nisin F, lysozyme, lactoferrin and combinations thereof were tested as follows: Nisin F was added to 100 μl BHI to yield final concentrations of 1280, 320, 160, and 80 AU ml−1, respectively (Table 1). In another set, lysozyme (Roche Diagnostic GmbH, Manheim, Germany) and lactoferrin (Sigma Diagnostics, St Louis, USA) were added to 100 μl BHI to yield a final concentration of 500 μg ml−1 (Table 1). Combinations of nisin F, lysozyme and lactoferrin were also prepared (Table 1). The suspensions were inoculated into separate wells of a sterile microtitre plate and then inoculated with S. aureus K to yield 105 viable cells. Nisin F, lysozyme and lactoferrin at the same concentrations, but not inoculated with S. aureus K served as controls. The microtitre plate was incubated at 37°C for 24 h and OD-readings taken at 595 nm (Model 680 Microplate Reader; Bio-Rad). All experiments were conducted in triplicate.
Table 1. Optical density readings (at 595 nm) of S. aureus K after incubation with nisin F, lysozyme, lactoferrin, and a combination of nisin F and lysozyme or lactoferrin, respectively
In the absence of lysozyme and lactoferrin
Lysozyme 500 μg ml−1
Lactoferrin 500 μg ml−1
Standard deviation displayed is a mean value of three repeats per experiment (n = 3). AU, antimicrobial units.
Nisin F 1280 AU ml−1
0·1 ± 0·1
0·1 ± 0
0·2 ± 0
Nisin F 320 AU ml−1
1·1 ± 0·2
0·4 ± 0·1
Nisin F 160 AU ml−1
0·9 ± 0·1
0·4 ± 0
0·5 ± 0
Nisin F 80 AU ml−1
0·6 ± 0
0·5 ± 0
1·2 ± 0
In the absence of nisin F
2·0 ± 0·1
1·7 ± 0
0·9 ± 0
Animal model and inoculation procedure
Approval for the experiments was obtained from the Animal Ethics Committee of Stellenbosch University (ethics reference number 2005B02003). Wistar male rats were divided into eight groups and housed in plastic cages in animal rooms with constant environmental conditions. Each group contained six rats. All rats received a standard rodent diet. The immunity of rats in groups 3 and 4 was suppressed by adding dexamethasone (2·5 mg l−1) to their drinking water for the first six days. An inoculum of S. aureus K was prepared by growing the cells in BHI for 18 h at 37°C. The cells were harvested (10 000 g, 10 min, 4°C), washed three times with 10 ml sterile physiological saline, resuspended in 10 ml sterile skim milk (10% w/v) and 4 ml dispensed into freeze-drying ampules. After 24 h of freeze-drying, the cells were resuspended in sterile saline to 107 CFU ml−1. On days 5, 6, 7 and 8 the rats in groups 1–4 were infected with 4 × 105 viable cells of S. aureus K (2 × 10 μl of 107 CFU ml−1 per nostril) per day. On days 9, 10, 11 and 12 rats in groups 1 and 3 were treated with 8192 AU nisin F (2 × 10 μl per nostril) and rats in groups 2 and 4 with sterile physiological saline (2 × 10 μl per nostril).
The safety of nisin F was tested by conducting a similar experiment. Rats in groups 7 and 8 received dexamethasone for the first 6 days as described before. On days 9, 10, 11 and 12 rats in groups 5 and 7 received 8912 AU nisin F and groups 6 and 8 sterile physiological saline, as described before. None of the rats in groups 5–8 were infected with S. aureus K. On day 13, all rats were euthanized by an overdose with pentobarbitone sodium (Centaur Labs, Bayer Animal Health Isando, South Africa) administered intraperitoneally.
Testing for adverse effects
The rats were weighed and their feed and water intake determined on a daily basis. Blood from rats infected with S. aureus K was plated on to Baird-Parker agar (Biolab) and the plates incubated at 37°C for 24 h.
The trachea, bronchi and lungs of three rats from each group were aseptically removed, fixed in 4% (v/v) formaldehyde (PBS) for 24 h at 25°C, embedded in paraffin, sectioned and stained with hematoxylin and eosin (H&E). The samples were processed and analysed at Pathcare Veterinary Pathologists (Pathcare, Dietrich, Voigt, Mia and Partners, Goodwood, South Africa).
Blood samples were collected from the rats directly after they were euthanized. Total blood counts were determined on fresh blood with an automated haematology counter (Beckman Coulter Hematology Analyzer LH; Beckman Coulter, Inc., CA, USA). All tests were conducted by Pathcare Veterinary Pathologists.
Nisin preparation and dosage
The antimicrobial activity of nisin F was 204 800 AU ml−1, as determined by using S. aureus K as target strain. The daily dosage with nisin F was thus 8192 AU (2 × 10 μl per nostril) per rat for both the infection and nisin F toxicity studies.
Effect of nisin F, alone and in combination with lysozyme and lactoferrin, on the growth of S. aureus K
Changes recorded in OD-readings of S. aureus K after incubation in the presence of nisin F, nisin F supplemented with lysozyme and lactoferrin, and lysozyme and lactoferrin without nisin F, are listed in Table 1. In the absence of nisin F, the growth of S. aureus K increased to OD (595 nm) = 2·0. Nisin at 1280 AU ml−1 prevented the growth of S. aureus K (OD595nm = 0·1). However, low concentrations of nisin F (80–320 AU ml−1) stimulated the growth of S. aureus K slightly (OD595nm = 0·6–1·1). No significant differences in the growth of S. aureus K were recorded when the cells were treated with nisin F (1280 AU ml−1) and the same level of nisin F in combination with either lysozyme or lactoferrin (OD595nm = 0·1–0·2). However, at lower concentrations of nisin F (80 AU ml−1) a slight increase in growth (OD595nm = 1·2) was observed when cells were treated with lactoferrin, but not when treated with lysozyme. This could have been an experimental error, as lactoferrin combined with a higher concentration of nisin F (160 AU ml−1) had the same antimicrobial effect as lysozyme combined with the same concentration nisin F (OD595nm = 0·4–0·5). Nisin F at 320 AU ml−1 combined with lysozyme completely inhibited the growth of S. aureus K (OD595nm = 0).
Testing for adverse effects
No significant differences were recorded in weight gain, and water and feed intake for rats in groups 1–4 (not shown). Bacterial growth was recorded in blood sampled from one of the rats in group 4 (immunosuppressed, not treated with nisin F and infected with S. aureus K). Small (1–5 mm) black colonies, surrounded by clear zones, characteristic of S. aureus, were visible on Baird-Parker agar plates. The lungs of the same infected animal had clearly visible necrotic lesions (not shown). No bacteremia or necrotic lesions were observed on lungs of rats in groups 1, 2 and 3. No significant differences were recorded in weight gain, and water and feed intake for rats in the nisin toxicity trial (groups 5–8).
No significant histopathology were recorded in lung tissue collected from rats in group 1 (non-immunosuppressed, infected with S. aureus K and treated with nisin F), group 2 (non-immunosuppressed, infected with S. aureus K and treated with sterile physiological saline), and group 3 (immunosuppressed, infected with S. aureus K, and treated with nisin F) (Fig. 1a). Severe symptoms of pneumonia were detected in lung tissue sampled from rats in group 4 (immunosuppressed, infected with S. aureus K, and treated with sterile physiological saline). The alveoli of the lungs from rats in group 4 were clearly obliterated due to the infiltration of macrophages, granulocytes and lymphocytes, and the proliferation of fibroblasts and alveolar epithelial cells (Fig. 1b). The trachea of the same rats showed signs of mucosal hyperplasia, with infiltration of lymphocytes and a few mast cells (Fig. 1d). No mucosal hyperplasia was observed in the trachea of rats from group 3 (immunosuppressed, infected with S. aureus K, and treated with nisin F) (Fig. 1c). No histological changes or differences were detected in the bronchi of rats in groups 1–4.
The histology of the trachea, lungs and bronchi of nonimmunosuppressed and immunosuppressed rats that have been treated with nisin F (groups 5 and 7) and sterile physiological saline (groups 6 and 8), and not infected with S. aureus K, appeared healthy (not shown).
The number of white blood cells, lymphocytes and neutrophils recorded in blood samples collected from nonimmunosuppressed and immunosuppressed rats that have been infected with S. aureus K (groups 1–4) did not differ significantly (not shown). The number of white blood cells, lymphocytes and neutrophils of the rats in the nisin F toxicity trial also did not differ between the nisin-treated rats (groups 5 and 7) and saline-treated rats (groups 6 and 8). No significant differences were detected in the red blood cell indices between the experimental and control groups for both the infection trial, i.e. groups 1 and 2 for nonimmunosuppressed rats and groups 3 and 4 for immunosuppressed rats, and for the nisin F toxicity trial, i.e. groups 5 and 6 for nonimmunosuppressed rats and groups 7 and 8 for immunosuppressed rats (not shown). The red blood cell distribution width (RDW) of rats that have not been infected with S. aureus K (groups 5–8) was slightly higher compared to rats that have been infected (groups 1–4).
Growth of S. aureus K was clearly inhibited by high levels of nisin (1280 AU ml−1), whether tested alone, or in combination with lysozyme and lactoferrin. Less growth inhibition of S. aureus K at lower concentrations of nisin F (80–320 AU ml−1) was to be expected. It is also not surprising to have recorded a slight increase in antimicrobial activity when nisin F was tested in combination with lysozyme and lactoferrin. This is indicative of synergistic antimicrobial activity, perhaps easier to observe in the presence of lower levels of nisin (80–320 AU ml−1) than an overwhelming concentration (1280 AU ml−1). Complete growth inhibition of S. aureus K was observed when 320 AU ml−1 nisin F was tested in combination with lysozyme, but not when the same experiment was conducted with 1280 AU ml−1 nisin F. The reason for this phenomenon is unclear and needs to be investigated.
Lysozyme and lactoferrin are present at relatively high concentrations in the upper respiratory tract of humans and in sputum at approximately 500 μg ml−1 (Brogan et al. 1975; Harbitz et al. 1984; Wilson 2005). Previous studies have shown a synergistic effect between nisin and lysozyme (Nattress et al. 2001) and nisin and lactoferrin (Murdock et al. 2007). Nisin prevents cell growth by binding to the cell wall precursor lipid II and destabilising the cell membrane, whereas lysozyme damages the glycosidic bond between N-acetylglucosamine and N-acetylmuramic acid residues in the peptidoglycan (Ganz 2004). Lactoferrin inhibits microbial growth by sequestering iron involved in respiration (Arnold et al. 1977).
Non-immunosuppressed rats dosed with S. aureus K did not develop symptoms of respiratory tract infections, confirming previous reports that the immune system of rats has to be compromised when studying S. aureus respiratory tract infections (Kruszewska et al. 2004). S. aureus have been associated with pneumonia, especially nosocomial pneumonia (Weber et al. 2007) and community-acquired pneumonia (Micek et al. 2007). It is thus not surprising that immunosuppressed rats infected with S. aureus K and not treated with nisin F developed severe symptoms of pneumonia, i.e. obliterated alveoli in lungs and mucosal hyperplasia in trachea. Mucosal hyperplasia may be due to increased production of mucus in the infected area (Fig. 1d). This is expected, as the up-regulation of mucin genes from by-products produced by S. aureus has been reported (Dohrman et al. 1998; Basbaum et al. 1999).
Necrotic lesions observed in the lung tissue and bacteremia detected in immunosuppressed rats infected with S. aureus K is indicative of damaged lung tissue through which the pathogen migrates into the blood stream. Immunosuppressed and infected rats treated with nisin F had no symptoms of pneumonia, mucosal hyperplasia, blood contamination or necrotic lesions, suggesting that the lantibiotic suppressed the growth of S. aureus K in vivo. Kokai-Kun et al. (2003), on the other hand, reported no inhibition of S. aureus in the nasal tract of cotton rats and concluded that nisin is either inactivated or absorbed in the nares. Another possible explanation is the formation of biofilms in the nasal tract, which could render the strains more resistant to nisin (Bendouah et al. 2006).
No histology changes were observed in the lungs, trachea and bronchi of rats in the nisin F toxicity trial (non-immunosuppressed and immunosuppressed groups). No differences in weight gain and food and water consumption were recorded. Nisin F had no effect on the white cell, lymphocyte and neutrophil production in the rat model. The red blood cell indices of rats treated with nisin F was similar to that recorded for rats treated with saline. Based on these results, nisin F is nontoxic and safe to use in the treatment of upper-respiratory infections.
As far as we could determine, this is the first successful report on a nisin variant being used to control intranasal S. aureus infections. Although murine models serve as an essential link between in vitro models and human clinical trails, the data presented in this study is preliminary and will have to be confirmed with studies on human volunteers before nisin F may be declared safe for human application.
Pathcare Veterinary Pathologists (Pathcare, Dietrich, Voigt, Mia and Partners, Goodwood, South Africa) for histology and blood cell counts and Dr Rob Smith, Department of Physiology, for advice on animal care. The research was supported by a grant from the National Research Foundation (NRF), South Africa.