OUP user menu

Staphylococcus aureus strains isolated from bovine mastitis: virulence, antibody production and protection from challenge in a mouse model

G. Leitner, O. Krifucks, A. Glickman, A. Younis, A. Saran
DOI: http://dx.doi.org/10.1016/S0928-8244(02)00458-3 99-106 First published online: 1 March 2003


Septic arthritis in mice was used as a model to evaluate the virulence of Staphylococcus aureus and coagulase-negative staphylococci (CNS) isolated from cases of bovine mastitis. In addition, the model was used to evaluate the cross protection elicited by heterologous antibodies. Mice were intramuscularly inoculated with serial bacterial doses of different strains of S. aureus or CNS, for virulence determination; they were monitored for arthritis, gangrene or death up to 20 days. Antibody response, cross reactivity and resistance to challenge were tested by subcutaneous inoculation with a low dose of one of the S. aureus or CNS strains followed by challenge with two S. aureus strains. S. aureusα-hemolysin isolate was the most virulent, followed by α+β-hemolysin and β-hemolysin isolates. The least virulent isolates were the non-hemolytic S. aureus strains but even they were more virulent than the CNS strains tested. Antibodies against three different S. aureus antigens were detected by the ELISA in all mice that were inoculated with the S. aureus strains but not in any of those with the CNS strains. Immunoblot test against various S. aureus strains as antigens showed high cross-reactivity among the S. aureus strains but only a slight similarity, restricted to the bands above 36 kDa, with the CNS sera. Low-dose inoculation of α or α+β strains before challenge with homologous and heterologous strains protected the mice, whereas the two β strains provided only partial protection. The inoculations of non-hemolytic S. aureus or the CNS strains did not elicit any protection. Our findings demonstrate that pre-exposure of mice to a low dose of certain S. aureus strains could provide protection and that the antibodies produced could have an important protective role.

  • Staphylococcus aureus
  • Mastitis
  • Mice model

1 Introduction

Staphylococcus aureus is a ubiquitous pathogen that causes a variety of infections in man and animals. A high degree of natural immunity to S. aureus has been detected in infected and non-infected human and animal hosts [14]. Despite this natural immunity and the existence of experimental vaccines, S. aureus infection continues to present a difficult problem and is the main etiological cause of mastitis in cattle, sheep and goats. The disease causes huge economic losses worldwide, estimated at $2 billion annually (or $200/cow) in the US alone [5]. Commercially available S. aureus vaccines against bovine mastitis have shown limited efficacy under field conditions [611]. In dairy cows, S. aureus and coagulase-negative staphylococci (CNS) cause contagious mastitis that is mainly subclinical and chronic but may appear as a peracute event in mammary quarters. In general the former leads to the more severe damage to the mammary gland, with an enhanced inflammation response. This suggests that after S. aureus and/or CNS penetrate(s) into the gland, the bacterial exosecretions are potentially responsible for the severity of the disease. S. aureus can produce and secrete proteins that exhibit superantigen (SAg) activity; they belong to a family of pyrogenic toxins and induce massive aberrant oligoclonal activation of T lymphocytes. The various strains of S. aureus can produce a number of potential virulence factors such as hemolysins (α, β, γ or δ), coagulase, leukocidin, enterotoxins (A–E), toxic shock syndrome toxin 1 (TSST-1), etc. [12,13]. The degree or severity of subclinical mastitis caused by S. aureus, which is partially related to these factors, can also vary among the S. aureus strains. In an attempt to test the role of S. aureus and CNS exosecretions in the bovine udder, cows were inoculated intracisternally with various strains of sterile bacterial exosecretions and were monitored for 96 h post-inoculation (PI) [14]. This study showed an association between bacterial exosecretions and the severity of the cows’ response as measured by the induction of inflammation in the inoculated quarters. Mouse models, if suitable, are less costly than cows as well as being a prerequisite for experiments preceding trials in humans. Previous studies [15,16] used mouse as an animal model in order to understand which components of S. aureus are of important in the development and persistence of septic arthritis. Although the virulence of the bacteria in mice is not necessarily similar to that in other species, we attempted to evaluate the virulence of S. aureus and CNS by means of this model, which we also used in an attempt to evaluate the degree of cross protection among isolates. All strains were field isolates from bovine mastitis in Israel and at the time of isolation cows had subclinical mastitis, however, some of the cows presented acute inflammatory processes before examination that later became subclinical [17,18]. The present study addressed the questions of bacterial virulence, antibody production and protection of mice after pre-exposure to a low bacterial dose of live homologous and heterologous S. aureus strains.

2 Materials and methods

2.1 Mice

Female Swiss mice, 6–8 weeks of age, were maintained in the animal facility of the Kimron Veterinary Institute. Three mice were housed in each cage under standard conditions of light (12/12 h light/dark) and temperature (22°C) and were fed standard laboratory chow and water ad libitum. At the end of the experiments, mice were euthanized with CO2

2.2 Bacterial strains

Seven strains of S. aureus, two strains of Staphylococcus chromogenes and one each of Staphylococcus intermedius and Staphylococcus xylosus were used. All strains were isolated from cases of chronic bovine mastitis [17,18]. Each strain was tested for coagulase production {(tube test) Anilab, Tal-Shachar, Israel}[19], subjected to the slide latex agglutination test (BACTI Staph, Remel, Santa Fe Drive, Lenexa, KS, USA), identified by means of the ID-32-API STAPH test (BioMerieux S.A, France), and phage typed (coagulase positive) with phages issued by the International Reference Laboratory, Colindale, UK, by the method of Blair and Williams [20] as modified by Samra et al. [21]. The strains did not produce any of the enterotoxins for which we tested (A–E) or TSST-1 as assayed by means of the microslide-immuno-diffusion-gel technique [22] at the Department of Food Hygiene, Ministry of Public Health, Haifa, Israel. The production of hemolysins (α or β) was confirmed by cultivating the bacteria in Columbia broth (Difco, Detroit, MI, USA) at 37°C for 10 h, centrifuging at 3000×g and testing the supernatant against rabbit or sheep erythrocytes as described by Nilsson et al. [15] For challenge and for electrophoresis each bacterium was cultivated in nutrient broth (Difco, Detroit, MI, USA) at 37°C for 24 h and harvested by centrifuging at 3000×g 15 min at 4°C. The pellet was resuspended in non-pyrogenic PBS (pH 7.6; 0.01 M) and washed twice, and a portion was stored at −20°C for the electrophoresis tests. For challenge, cells were resuspended to McFarland equivalence turbidity standard 4 (1.2×109 cells ml−1) (Remel Santa Fe Dr., Lenexa, KS, USA). When a higher concentration was required (>1.2× 109 cells ml−1), the bacteria were repelleted and resuspended in one-quarter of the non-pyrogenic PBS. The concentration was adjusted to the appropriate level by addition of non-pyrogenic PBS, and 0.25 ml of the bacterial suspension was inoculated into the left hind leg of each mouse. The exact bacterial concentration injected to the mouse was determined by serial dilutions (×2) on blood-agar plates.

2.3 Clinical evaluation

Mice were inoculated intramuscularly (IM) with various bacteria and examined individually thereafter. The injected limb was visually inspected daily for visible erythrema gangrene. Macroscopic inspections during 20 days of observation revealed either: (0) normal appearance; (1) morbidity — erythrema gangrene; (2) mortality. Virulence was calculated as the combined percentage of mortality and morbidity among all the mice inoculated and expressed as the number of CFUs of the bacteria that resulted in death (mortality — which occurred within the first 5 days post-challenge (PC)) or erythrema gangrene (morbidity) in 50–70% of the inoculated mice.

2.4 Experimental design

Two sets of experiments were performed. The first set determined the virulence of the isolates to mice. Mice were inoculated IM: with one of the S. aureus strains, BS1, AU2873, AU3133, ZU452, FR2449, GL3497 or NR2602; with S. chromogenes (m10); with S. intermedius (M2). For each strain, nine mice were used receiving 5×106, 5×107 or 5×108 CFU. The mice were inspected for arthritis, gangrene or mortality for up to 20 days. Isolates whose virulence was not determined: mice did not show any symptoms of arthritis even with the highest dosage of CFU and were tested again as mentioned above but the dosage of bacteria was increased to 5×109 CFU/mouse or decreased to 5×105 CFU/mouse.

The second set of experiments was conducted to study antibody responses, the cross reactivity among them, and resistance to challenges with various S. aureus strains. Two independent trials were conducted.

Trial 1 included five groups, each of six mice. The mice in each group were inoculated subcutaneously (SC) with 1×105 CFU/mouse of: one of the S. aureus strains, BS1, AU2837 or AU3133; S. chromogenes (m3); or S. xylosus (743)) and with the non-pyrogenic PBS as negative control. Antibody response was determined 40 days PI by ELISA.

Trial 2, included 10 groups, each of eight mice. The mice in each group were inoculated SC with 1×105 CFU/mouse of: one of S. aureus strains, BS1, AU2837, AU3133, ZU452, FR2449, GL3497 or NR2602; or one of the CNSs, S. chromogenes (m10) or S. intermedius (M2) and with the non-pyrogenic PBS as negative control. Antibody response was tested twice, at 15 and 25 days PI. At 30 days PI the mice in each group were divided into two subgroups, each of which was challenged with one S. aureus isolate, either BS1 or NR2602. The mice were inoculated IM with 5×108 CFU/mouse of BS1 or with 1×107 CFU/mouse of NR2602, and were monitored for mortality and arthritis for up to 20 days PC.

2.5 Electrophoresis and immunoblot

One-dimension sodium dodecyl sulfate-polyacrylamide gel electrophoresis (1-D SDS–PAGE) was performed according to Laemmli [23]. Bacterial cells were disrupted with glass beads in a homogenizer (Braun Melsungen AG, Germany) for 10–15 min. The glass beads and the remaining bacteria were removed by centrifugation (1000×g, 15 min). The supernatant was filtered through 0.2-µm filters. Protein concentrations of the disrupted bacteria were determined with the Bio-Rad protein assay [24]. The samples (antigens) were adjusted to a final protein concentration of 1 mg ml−1 and 33 µl of each sample was mixed with 25 µl of 4× NuPAGE sample buffer, 32 µl of ultrapure water and 10 µl of reducing agent (0.5 M DTT). The sample mixtures were heated to 70°C for 10 min and loaded into the gels at 30 µl per lane. The gels were 7% [NuPAGE Tris acetate (+TA buffer)] and 10% (Bis–Tris Gel with w/MOPS) (NOVEX, San Diego, CA, USA); they were stained with coloidal blue. Molecular-mass markers for these gels were: See Blue Pre-Stained, 191–14 kDa for the 10% gel and 210–41 kDa for the 7% gel). For the immunoblot assay, a nitrocellulose membrane (0.2 µm) was blocked with 3% casein and incubated with 1:50 to 1:100 dilutions of serum from mice immunized with three different S. aureus strains. The blot was developed with goat anti-mouse IgG (H+L) alkaline peroxidase conjugate (1:1000), with a substrate of 3,3-diaminobenzidine tetrahydrochloride (ICN Pharmaceuticals, Inc.). Molecular-mass markers for these gels were: NOVEX Marker 12 Standard Bands (200–2.5 kDa.

2.6 Immunological assay

ELISA: 96-well immunoplates (Nunc, Kamstrup, Denmark) were coated overnight at 4°C with 200 µl/well of various strains: BS1, AU2837, AU3133, of agitation-killed S. aureus, suspended in carbonate-bicarbonate buffer (pH 9.6) at 5×108 CFU ml−1. When an assay was performed, 200 µl of 3% casein were added to each well and incubated for 40 min at room temperature. The plates were washed three times for 5 min with PBS (pH 7.6) containing 0.5% Tween 20. They were then dried and 100 µl of tested serum, at 1:400 dilutions was added, and the plates were incubated for 1 h at 37°C. This procedure was carried out in duplicate. The plates were washed and affinity-purified goat anti-mice (H+L) peroxidase conjugate (Kirkegaard and Perry Laboratories, Gaithersburg, MD, USA) at 1:1500 dilution was added to each well and the material was incubated for 1 h at 37°C. The bound antibodies were detected by adding 100 µl of 2.2′-azino-di(3-ethyl-benzthiazolin sulfonate) peroxidase substrate (Kirkegaard and Perry Laboratories, Gaithersburg, MD, USA) after washing. The respective plates were read in a microplate autoreader (Dynatech MR5000, Guernsey, UK) at 410 nm. To each plate, positive and negative serum standards were added, prepared from hyperimmune challenged and unimmunized non-infected mice, respectively. The antibody level was (the OD of the tested serum divided by the OD of the negative serum) minus 1, i.e., (OD tested serum/OD negative serum) −1

3 Results

The virulence ratings of the isolates (Table 1) were determined by inoculation with a range of concentrations of: the S. aureus strains, BS1, AU2873, AU3133, ZU452, FR2449, GL3497 and NR2602; S. chromogenes (m10); and S. intermedius (M2). The most virulent isolate was the α-hemolysin, S. aureus NR2602 (5×106 CFU/mouse). Moderate virulence was shown by the one of α+β-hemolysins, isolate FR2449 (5×107 CFU/mouse), followed by the other α+β-hemolysin, isolate BS1 and by one of the β-hemolysin isolates, S. aureus AU2873 (1×108 CFU/mouse). The least virulent isolates were the β-hemolysin, S. aureus ZU452 and the non-hemolysin isolate, AU3133 (>5×109 CFU/mouse). The non-hemolysin isolate GL3497 did not cause any mortality but the morbidity was high in all three levels of inoculation therefore, virulence was calculated as 1×108. Neither of the two CNSs, S. chromogenes (m10) or S. intermedius (M2), elicited any symptoms, even when the highest challenge was used.

View this table:
Table 1

Virulence {number of CFUs of bacteria that induced mice death (mortality) or erythrema gangrene (morbidity) in 50% of the mice} derived from inoculation with several concentrations of various strains of S. aureus (BS1, AU2873, AU3133, ZU452, FR2449, GL3497 and NR2602), S. chromogenes (m10) or S. intermedius (M2)

BacteriumCFU/miceMortality (%)Morbidity (%)Mortality and morbidity (%)Virulence
S. chromogenes (m10)5×109>5×109
S. intermedius (M2)5×109>5×109
  • Percentage of mice mortality during 20 days PI.

  • Percentage of mice with arthritis and gangrene during 20 days PI.

In the 1-D SDS–PAGE tests, a high degree of similarity was found among the various S. aureus strains, in their protein bands above 42 kDa, whereas the CNS strains tested showed substantial differences in this region (Fig. 1). The 42-kDa bands identified as protein A and tested by anti-protein-A antibody (data not shown), appears only in the S. aureus strains. Among the S. aureus strains, differences were found in the range of 40–28 kDa (Fig. 2), while no differences could be observed below 28 kDa (data not shown).

Figure 1

1-D SDS–gels 10% (Bis–Tris gel with w/MOPS) electrophoresis of disrupted S. aureus and CNS strains. Molecular-mass markers are See Blue Pre-Stained (191–14 kDa).

Figure 2

1-D SDS–gels 7% [NuPAGE Tris acetate (+TA buffer)] electrophoresis of disrupted S. aureus and CNS strains. Gels express the protein patterns between 55 and 36 kDa. Molecular-mass markers are See Blue Pre-Stained (210–41 kDa).

Table 2 presents summaries of the mortality and morbidity after inoculation, of mice pre-exposed to low concentrations of: the S. aureus strains, BS1, AU2873, AU3133, ZU452, FR2449, GL3497 and NR2602; S. chromogenes (m10); or S. intermedius (M2). Mice that had been pre-exposed to the α strain NR2602 or to the α+β strain BS1, were protected when inoculated with homologous and heterologous strains, while pre-exposure to the α+β strain FR2449 resulted in only a partial protection against inoculation with the NR2602 strain. No protection could be shown against the BS1 strain. Among the mice pre-exposed to the two β strains, AU2873 elicited full protection against BS1 while ZU452 failed to do so. The pre-exposure to the non-hemolytic S. aureus strains, AU3133 and GL3497, to the S. chromogenes (m10) or to S. intermedius (M2) did not confer any protection relative to control when the mice were inoculated with NR2602 and only poor protection when they were inoculated with BS1 strains. Moreover, the two CNS strains used: S. chromogenes (m10) and S. intermedius (M2) when challenged with NR2602, caused an increase in mortality above control (100% vs. 25% respectively). Table 3 summarizes the results of ELISA tests of antibody levels against antigens of three S. aureus strains: NR2602/4, AU2873 and GL3497. In both trials (1 and 2), antibodies were detected in all the mice that were inoculated with the various S. aureus strains while such antibodies could not be detected in any of the mice, which were inoculated with CNS. In the comparisons among the antibody levels elicited by the various antigens (cross reactivity), only strain AU3133 showed higher titer with the homologous strain than with the two heterologous strains. In all the other six S. aureus strains tested, antibody levels of homologous and heterologous strains showed no significant differences (Table 3).

View this table:
Table 2

Mortality and morbidity of mice pre-exposed to low concentrations of various strains of S. aureus (BS1, AU2873, AU3133, ZU452, FR2449, GL3497 and NR2602), S. chromogenes (m10) or S. intermedius (M2) and then inoculated with different concentration of the bacteria

BacteriumHemolysisBS1 (5× 108 CFU)NR2602 (1× 107 CFU)
Mortality (%)Morbidity (%)Mortality (%)Morbidity (%)
S. chromogenes (m10)0100100
S. intermedius (M2)050100
  • Percentage of mice mortality during 20 days PI.

  • Percentage of mice with arthritis and gangrene during 20 days PI.

  • Control: mice which were not pre-exposed to low concentrations of any bacteria.

View this table:
Table 3

Antibody titers of mice 40 (experiment 1) and 15 and 25 (experiment 2) days after SC inoculation with a small number (5×105 CFU/mouse) of one of seven different strains of S. aureus (BS1, AU2873, AU3133, ZU452, FR2449, GL3497 and NR2602), two strains of S. chromogenes (m3 and m10), S. xylosus (743) or S. intermedius (M2)

1S. chromogenes (m3)40000
1S. xylosus (743)400.060.090.08
2S. chromogenes (m10)150.100
2S. intermedius, (M2)15000
  • ELISA tested sera against each of the S. aureus strains: BS1, AU2873 and AU3133.

  • Days post inoculation.

  • Antibodies titer within row with underline represents homologous strain.

Immunoblot tests of sera of all S. aureus strains, with different S. aureus strains used as antigens, revealed high cross-reactivity while at the same time they also showed some specific protein bands, the nature of those however are not known at this stage. Fig. 3 presents one such immunoblot with serum against strain AU2873; sera from the mice inoculated with the different CNS strains show some similarity in the immunoblot to the S. aureus strain, above the 36-kDa marker.

Figure 3

1-D SDS–gels 10% (Bis–Tris gel with w/MOPS) immunoblot assay. Electrophoresis of disrupted S. aureus and CNS strains. Molecular-mass markers are NOVEX Marker 12 Standard Bands (200–2.5 kDa).

4 Discussion

The present study describes the use of a mouse model in an attempt to elucidate S. aureus virulence, antibody production and protection after exposure to a low bacterial dose of live homologous and heterologous S. aureus strains. This model may assist to find S. aureus strains suitable for the development of vaccines against bovine mastitis caused by S. aureus, as well as more information for the understanding of the mammary gland immunity complex. α-Toxin [25] or α- and γ-toxin [15] play major roles in the pathogenesis of septic arthritis. α-Toxin has been found to contribute to arthritic events by promoting adherence of neutrophils to endothelial cells [26] and macrophages, so producing proinflammatory mediators such as TNF-α. In the present study, the field strains of S. aureus producing the α-toxin were found to be 1.5–2 log more virulent than the α- and β-toxin and those higher than β-toxin or the non-hemolytic strains. These results support the putative role of α-toxin as the major toxin in the pathogenesis of septic arthritis, although the higher virulently of the NR2602 (α-toxins) then BS1 and FR2449 (α- and β-toxin) is not yet clear. It is possible that the level of α-toxin produced by NR2602 was higher however, the exact levels of each toxin were not titered. It has been proposed [15] that β-toxin might have some protective properties. However, in the present study, β-toxin strains were found to be moderately virulent compared with the non-hemolytic strains of S. aureus and the CNS strains tested. The non-hemolytic S. aureus strains were more virulent than the S. chromogenes and S. intermedius strains. It seems, therefore, that although hemolysin plays a major role in the development of erythrema gangrene and mortality in mice, other, even more dominant factors may influence the pathogenesis of the disease.

Mice, which were pre-exposed to the α-toxin NR2602, or the α+β-toxin strains, BS1 and FR2449 were protected from challenge with the S. aureusα-toxin NR2602 strain. In contrast, pre-exposure to the β-toxin or the non-hemolytic S. aureus strains did not provide any protection. The β-toxin strain AU2873 did not provide any protection against challenge with the α-toxin strain NR2602 but did provide protection against the α+β BS1. These results suggest that β-toxin may have an independent pathogenesis of septic arthritis. Pre-exposure of mice to β-toxin provided a protection against β-toxin but not against the α-toxin. Interestingly, the mice pre-exposed to the two CNSs, S. chromogenes (m10) and S. intermedius (M2) exhibited a higher protection rate against the α+β BS1 strain than against the α NR2602 one. α-Toxin and probably also β-toxin play a major role in the pathogenesis of septic arthritis. However, other S. aureus SAg toxins such as TSST-1 also contribute to the arthritogenicity [27,28]. It has been suggested that S. aureus arthritis is a T-cell-dependent and SAg-mediated disease [29]. The role of B cells in this disease has been studied in B-cell-deficient mice [30] and no impact on the progress of symptoms was found. However, in the present study, our aim was to study the possibility of using a pre-exposure to low concentrations of various S. aureus strains, to prevent the development of symptoms of septic arthritis as a result of a subsequent higher exposure. It has been reported [31] that vaccination of mice with recombinant staphylococcal enterotoxin A provided protection. Indeed, our present results indicate that B cells were involved in the production of antibodies against the various bacteria used and that those may have had an important role in the protections process.

In the mammary gland, S. aureus and CNS are the predominant pathogenic bacteria that cause chronic infection, and the disease is generally more severe when caused by S. aureus than when caused by CNS strains. Other studies [32,33] have demonstrated the role of SAgs as one of the virulence factors that modulate the immune response and promote survival of the bacterium. Therefore, the mechanisms by which SAgs cause septic arthritis in mice and the severity of mastitis in cows appear to be the same. If so, protecting mice from septic arthritis and elucidating the mechanism involved may contribute to the development of anti-S. aureus immunization in cows. One may question the relevance of the mouse model to the problem of bovine mastitis caused by S. aureus strains. Therefore, a composition comprising three strains of S. aureus derived from cows affected with mastitis was evaluated by means of this model. The said composition was found to be highly efficacious in protecting mice against both homologous and heterologous strains. This composition, with an adjuvant was subsequently used as a vaccine, administered to cows, which were then challenged with a virulent strain of S. aureus, and it was found to confer significant immunity on the cows [34]. It seems, therefore, that this mouse model, in spite the differences between the clinical symptoms in mice and cows, can serve as a useful tool for the evaluation of the virulence of S. aureus isolates derived from wounds or mastitis cases. Moreover, it can be used as a tool for evaluating the degree of immunity conferred by potential vaccines designed against S. aureus strains.


  1. [1].
  2. [2].
  3. [3].
  4. [4].
  5. [5].
  6. [6].
  7. [7].
  8. [8].
  9. [9].
  10. [10].
  11. [11].
  12. [12].
  13. [13].
  14. [14].
  15. [15].
  16. [16].
  17. [17].
  18. [18].
  19. [19].
  20. [20].
  21. [21].
  22. [22].
  23. [23].
  24. [24].
  25. [25].
  26. [26].
  27. [27].
  28. [28].
  29. [29].
  30. [30].
  31. [31].
  32. [32].
  33. [33].
  34. [34].
View Abstract