Characteristics of Staphylococcus aureus from subclinical bovine mastitis in Brazil

Br. vet . J. (1990) . 146, 443

CHARACTERISTICS OF STAPHYLOCOCCUS A UREUS FROM SUBCLINICAL BOVINE MASTITIS IN BRAZIL

CARLOS ALBERTO DE MAGALHAES LOPES*, GILBERTI MORENO*, PAULO ROBERTO CURIt, ARNOLD F . GOTTSCHALK4, JOSE RAFAEL MODOLO$, ANTONIA HORACIO*, ANTONIO CORREA* and CELSO PAVAN* *Department of Microbiology and Immunology, Institute of Biosciences ; f Veterinary Hospital Statistical Service, Faculty of Veterinary Medicine ; and $Department of Hygiene and Veterinary Public Health, Faculty of Veterinary Medicine ; Estadual Paulista University, UNESP, Botucatu, SP 18610, Brazil

SUMMARY A total of 127

Staphylococcus aureus strains isolated from milk samples of cows with subclinical bovine mastitis was examined for biotype, phage pattern, in-vitro antibiotic susceptibilities and ability to produce enterotoxins . The majority of the strains showed features consistent with bovine rather than human origin . All strains were sensitive to the antibiotics tested, except penicillin and streptomycin . Enterotoxigenicity was observed in 6 (4 . 7%) strains and only enterotoxins A and C were produced .

INTRODUCTION is among the most important aetiological agents of bovine mastitis, a disease widely recognized as being of great public health importance and economic significance . The findings that some strains isolated from mastitic milk are enterotoxigenic (Olson et at, 1970 ; Untermann et at., 1973 ; Niskanen & Koiranen, 1977 ; Garcia et al., 1980) and that animals can probably be a source of antibiotic resistant strains for humans (Levy, 1987) provide further justification for considering this organism significant in the epidemiology of milk-borne diseases . Since little information is available concerning the epidemiological significance of S . aureus strains from subclinical mastitis in Brazil, the purpose of this study was to examine mastitic milk for the presence of the micro-organism, as well as to carry out biotyping and phage typing of the isolates and to examine their antibiotic susceptibilities . This paper also deals with the ability of the strains to produce enterotoxins and their importance as a possible source of food intoxication . Staphylococcus aureus

MATERIALS AND METHODS Samples and isolation of S . aureus

Three hundred and three milk samples taken from individual cows giving a positive California Mastitis Test were examined for the presence of S . aureus. Eight dairy farms



4 44

BRITISH VETERINARY JOURNAL, 146, 5

located in Sao Paulo State, Brazil, were selected for this purpose . All samples were dispatched directly to the laboratory (Department of Microbiology and Immunology, Institute of Biosciences, Botucatu, UNESP) with minimal delay and stored under refrigeration until examined . All isolations were carried out by streaking 0 . 1 ml of mastitic milk onto plates with Blood Agar Base (Difco Laboratories) containing 5% defibrinated sheep blood and Mannitol Salt Agar (Difco Laboratories) plus 3% egg yolk emulsion . After incubation for 24 and 48 h, colonies were examined by Gram staining . The identification of S. aureus was done by standard bacteriological methods (Koneman et at., 1988) and for further studies, one strain from each positive sample was chosen at random . Biotyping Subdivision of the strains into biotypes was performed using the scheme of Hajek & Marsalek (1971) . All strains were tested for the following characters : fibrinolysin, pigmentation, coagulation of human and bovine plasmas, production of alpha and beta haemolysins, reduction of tellurite, clumping factor, growth type on Crystal Violet Agar and typing by adapted phages . The comparison of the characters of the strains with the physiological profiles defined for the biotypes A (human sources) and C (bovine/ovine sources) was attempted by using indices (0-10) of correlation with 10 physiological characters of each biotype . The data were statistically analysed according to the t-test for two dependent samples (Snedecor & Cochran, 1980) . Phage typing Phage typing was done using the standard method of Blair & Williams (1961) as modified by Parker (1972), with the basic sets of phages for typing human (Subcommittee, 1975) and bovine strains (Working Group on Phage Typing of Bovine Staphylococci, 1971) . The strains non-typable with the routine test dilution (RTD) were retested at 100XRTD . Enterotoxin assay The strains were tested for enterotoxins A (SEA), B (SEB), C (SEC), D (SED) and E (SEE) by the cellophane-over-agar method for enterotoxin production and the optimal sensitivity plate method for enterotoxin detection and identification (Robbins et at., 1974) . Reference enterotoxins and corresponding antisera were kindly supplied by the Professor Merlin S . Bergdoll (Food Research Institute, University of Wisconsin, Madison, USA) . Antibiotic sensitivity testing The in-vitro drug susceptibilities of the strains were analysed by determining the minimal inhibitory concentrations (MIC) of eight antibiotics according to the agar dilution technique (Washington & Sutter, 1980) . The antimicrobial agents (in powder form) and their suppliers were : ampicillin (Amp), amikacin (Ami), oxacillin (Oxa) and tetracycline (Tet) (Bristol Laboratories) ; cephalothin (Cep) (Eli Lilly & Co .) ; gentamicin (Gen) (Schering Corp .) ; penicillin (Pen) and streptomycin (Str) (Fountoura WyethBrazil) . MICs were determined in Mueller Hinton Agar (Difco Laboratories) containing serial dilutions of a drug in distilled water to give concentrations ranging from 200 to 0 . 1 ,ug/ml . The plates were inoculated with a 1 :10 dilution of a culture equivalent in



443

S. AUREUS AND BOVINE MASTITIS

turbidity to a MacFarland scale 0 . 5 BaSO4 standard (10 4 cfu/ml) using a multiloop replicator, and incubated for 18 h . The MIC was defined as the lowest antibiotic concentration showing no growth, a faint haze or a single colony at the site of inoculation .

RESULTS Strains

A total of 127 S. aureus strains either in pure culture or in numbers sufficient to be considered the cause of mastitis was isolated . Biotyping

Biotype could be defined in 53 (41 . 7%) of the isolates ; 51 (40 . 1%) strains belonged to biotype C and two (1 . 6%) to biotype A. The remaining 77 (60 . 6%) strains could not be definitely assigned (index 10) to one of the six schematic biotypes . Table I shows the degree of approximation of the strains to biotypes A and C . On the assumption that a significant approximation to one biotype was indicated by index values in the range of 6-10 (difference between biotypes 3) ; 94 (74%) strains approximated to biotype C and five (3 . 9%) to biotype A (d=3 . 60 ; SD=2 . 9 ; t=18 . 51 ; P< 0 . 001) . The 28 (22%) remaining strains were classed as an intermediate group . Table I Results of biotyping of 127 S . aureus strains from mastitic milk

Indices

1

Biotype A

2

Total

4

Biotype C 5 6

7

8

9

10*

Total

\

1 '*'_' 2 ~.

3 4 5 6 7 8 9 10*

3

~~

1

0

~~

~~

1

1

--_2 1

1~~

1 2 ~1

~~ 1

7 3

19 16 1 2

4 43 4 ~~

1 2 0

1

1

3

1

5

8

19

38

51

1 3 29 66 11 10 3 2 2 - 127

* Clear-cut definition of a biotype . Dotted lines indicate the limits for tentative assignment of biotype on the basis of a predominance of the appropriate characters . Phage typing

All but 16 (87 . 4%) strains were phage typable (Table II) ; 39 (30 . 7%) strains (25 at RTD and 14 at 100XRTD) were lysed by phages in the human basic set . Of the different phage patterns, the ones most frequently found were 6/42E/53/54/75/77/83A/85 (13 strains) and 3A/81/6/42E/53/75/77/94/95 (11 strains) . No strains were lysed by the



446

BRITISH VETERINARY JOURNAL, 146, 5

phages 52, 79, 86, 88, 90, 92D and HK2 (human set) . Seventy-two (56 . 7%) strains were lysed by phages in the bovine set (44 at RTD and 28 at 100XRTD), while 52 of them were lysed by the specific phages 78, 102, 107, 116, 117, 118 and 119 (bovine set), in single or multiple phage patterns . The phage 119 by itself was responsible for the lysis of 27 strains, the most frequent phage patterns being 119 (27 strains) and 78/102/107/117 (6 strains) .

Table II Phage typing of 127 S . aureus strains from mastitic milk Phage group

.No . of strains Human set

I II III IV

NC* Mt Mixed II-III

2 5 11

Total

Specific bovine phages

2 1 7

12

3 38

4

III-IV

IV-M I-NC III-NC 1-111-NC II-III-NC

Bovine set

2 8 2

1 10 2

1 39(30-7%)

20(15-7%)

52(40-9%)

* NC, non-classified phage group (human set) . t M, miscellaneous phage group (bovine set) .

Enterotoxin assay The production of enterotoxin was observed in six (4 . 7%) of the 127 strains, two of biotype C, one of biotype A and three non-biotypable strains . The enterotoxigenic biotype C strains produced SEC exclusively, while the biotype A strain formed only SEA . The enterotoxigenic non-biotypable strains produced SEA and SEC concomitantly . Moreover, the enterotoxigenic biotype C strains were lysed by the specific phage 119, and the biotype A and the non-biotypable strains by the phages of group III (human set) and groups III/IV (bovine set), respectively . Antibiotic sensitivity testing The results of the antibiotic sensitivity test are summarized in Table III . The antibiotics displayed potent antimicrobial activity with values of MIC 9o ranging from 2 . 3 to 0 . 1 1ug/ml . The order of the drug effectiveness was Gen > Oxa > Cep > Amp > Pen > Tet > Str .

447

S. AUREUS AND BOVINE MASTITIS

Table III In-vitro antimicrobial susceptibilities of 127 S . aureus strains from mastitic milk Antibiotic

Penicillin Ampicillin Oxacillin Cephalothin Gentamicin Amikacin Streptomycin Tetracycline

Range

_<0- 1 -< 0 . 1 _<0-1 _<0-1 _<0 . 1 .1 -
MIC (µg/ml)

-5 -- 1 -0-5 -2 -0 . 5 -1 - 10 - 20

50% *

90%

0.1 0. 1 0 . 21 0 . 14 0.1 0.1 0.1 0. 1

0 . 44 0 . 42 0 . 40 0 . 40 0.1 0 . 44 2.3 1

*MICS for 50 and 90%, of strains, respectively .

DISCUSSION It is well established that certain biotypes of S . aureus (e .g . A) and phage types are adapted to human beings and others (e .g . biotype C) to cattle . In this survey, the results of biotyping and phage typing clearly showed a better adaptation of the majority of strains to the bovine than to the human carrier . The clear-cut definition of 51 (40 . 1%) strains into biotype C, the group of 94 (74%) strains showing a predominance of characters corresponding to biotype C and the higher sensitivities of most strains to the phages of the bovine set (especially to the specific phages) give strong support to the above view, and suggest that the intervention of human carriers in the epidemiological chain of the disease, at least in the dairy farms that we examined, is of little importance . In our methodology, the technique of assigning strains to a biotype on the basis of a predominance of the appropriate characters was particularly useful in estimating the probable origin of some non-biotypable strains . In the case of strains tentatively assigned biotypes on the basis of a predominance of appropriate characteristics, those assigned to biotype C coagulated human and bovine plasmas, produced pigment and beta-haemolysin and grew in violet (negative growth type) colonies on Crystal Violet Agar, whereas those assigned to biotype A produced fibrinolysin, pigment, alpha-haemolysin and coagulated human but not bovine plasma . In relation the phage typing, our data correlate well with the results of Garcia et al. (1980) for milk samples from cows with clinical and subclinical mastitis in Spain ; their strains were more susceptible to the specific bovine phages than to those of the basic human set . However, Rahman & Baxi (1983) in India found the majority of strains (66 . 9%) typable with phages of the human set . These authors suggested the probable human origin of the strains and proposed that human carriers may be an important factor in the spread of the disease . The drug sensitivity testing showed considerable effectiveness of the antibiotics, with the exception of tetracycline and streptomycin, against the isolates . These findings are similar to those of Garcia et all and Francis & Carrol (1986) and lead us to consider that in our region, milk is not an important vehicle in the dissemination of antibiotic-resistant S. aureus.



448

BRITISH VETERINARY JOURNAL . 146, 5

The literature shows a limited occurrence of enterotoxin-positive S. aureus strains from bovine mastitis and milk specimens without a remarkable predominance of particular antigenic types of toxins (Hajek, 1978) . According to Sinell (1971), Wieneke (1974), Olson el al. (1970) and Mayer (1975), the percentage of enterotoxigenic strains from these sources ranges from 2 to 15% . Despite the public health implications of these reports, no epidemiological work has been done . Our findings of a low percentage of enterotoxigenic strains and the non-prevalence of a specific antigenic type of toxin are in accordance with the previous reports . However, SEC and SED have been reported by Garcia et all to be produced by bovine strains, whereas ours produced only SEC . In summary, our data also suggest that staphylococcal food poisoning could occur as a consequence of the use of bovine mastitic milk in our geographical area, but the probability of occurrence of such poisoning seems to be low, and is not likely to represent a serious public health hazard .

ACKNOWLEDGEMENTS The authors wish to thank Profs . Ana Maria Uthida Tanaka and Maria Aparecida de Araujo of the Department of Parasitology, Microbiology and Immunology (Faculty of Medicine, Ribeirao Preto, Sao Paulo University) for phage typing our S. aureus strains . We also thank Sonia Maria Faraldo for typing the manuscript . This work was supported by a grant in-aid from the Conselho Nacional de Desenvolvimento Cientifico e Tecnologico (CNPq)- Proc . no . 300389/86-8 .

REFERENCES BLAIR, J . E . & WILLIAMS, R. E . O . (1961) . Bulletin of the World Health Organization 24,771 . FRANCIS, P . G . & CARROL, P . J . (1986) . Veterinary Record 118, 361 . GARCIA, M . L., MORENO, B . & BERGDOLL, M . S . (1980) . Applied and Environmental Microbiology 39(3), 584 . HAJEK, V . (1978) . Applied and Environmental Microbiology 35(2), 264 . HAJEK, V . & MARSALEK, E . (1971) . Zentralblatt fur Bakteriologie, Parasitenkunde, Infektionskrankheiten and Hygiene. I Abt . Orig. Reihe A 217, 176 . KONEMAN, E . W., ALLEN, S . D ., DOWELL, JR ., V. R ., JANA, W . M ., SOMMERS, H . M . & WINN JR ., W . C . (1988) . Diagnostic Microbiology, 3rd edn . Philadelphia : J . B . Lippincott. LEVY, S . B . (1987) . Journal of Food Protection 50(7), 616 . MAYER, S . (1975) . Milchwissenschaft 30, 60A . NISKANEN, A . & KoIRANEN, L . (1977) . Journal of Food Protection 40, 543 . OLSON, J . C ., CASMAN, E . P., BAER, E . F . & STONE, J . E . (1970) . Applied Microbiology 20, 605 . PARKER, H . T. (1972). Methods in Microbiology, vol . 7B . New York: Academic Press . RAHMAN, H . & BAXI, K. K . (1983) . Indian Veterinary Journal 60, 865 . ROBBINS, R ., GOULD, S . & BERGDOLL, M . (1974) . Applied Microbiology 28, 946 . SINELL, H . J . (1971) . Alimenta 10, 13 . SNEDECOR, G . W . & COCHRAN, W . G . (1980) . Statistical Methods, 7th edn . Ames, Iowa : Iowa State University Press . SUBCOMMITTEE (1975) . International Journal of Systematic Bacteriology 25, 241 . UNTERMANN, F., KuscH, D . & LuPKE, H . (1973) . Milchwissenschaft 28, 686 . WASHINGTON, J .A . II & SUTTER, V. L . (1980) . Manual of Clinical Microbiology, 3rd edn . Washington, DC : American Society for Microbiology . WIENEKE, A . A. (1974) . Journal of Hygiene (Cambridge) 73, 255 . WORKING GROUP ON PHAGE TYPING OF BOVINE STAPHYLOCOCCI (1971) . International Journal of Systematic Bacteriology 21, 171 . (Accepted for publication 21 November 1989)