Acta vet. scand. 1999, 40. 315-321.

Identification of Coagulase-Positive Staphylococci Isolated from Bovine Milk

By A. Capurro' , C. Concha', L. Nilsson! and K. Ostensson'

I Departrnent ofObstetrics and Gynecology, Faculty ofVeterinary Medicine, Swedish University of Agricultural Sciences, and 2Departrnent of Mastitis, National Veterinary Institute, Uppsala, Sweden.

Capurro A, Concha C, Nilsson L, Ostensson K: Identification of coagulase-positive staphylococci isolated from bovine milk. Acta vet. scand. 1999,40,315-321. -A to• tal of 414 coagulase-positive staphylococcal strains obtained at the mastitis laboratory, National Veterinary Institute, Uppsala, Sweden, were studied. One hundred and seventy seven strains were used for a frequency study. Ninety-seven per cent were identified as Staphylococcus aureus, 2% as Staphylococcus intermedius and 1% as Staphylococcus hyicus. Two hundred and thirty seven strains with atypical hemolysis reactions on bo• vine blood agar were randomly selected, with the aim to increase the number of S. inter• medius and S. hyicus strains available for testing. Eight different characteristics, includ• ing physiological, enzymatical and biochemical properties, were used to identify the coagulase-positive Staphylococcus species. The results ofthis study suggest that the fol• lowing tests should be included for correct identification of the 3 different species ofco• agulase-positive staphylococci: P agar supplemented with acriflavin, fl-galactosidase and hemolytic reaction on chocolate agar. These 3 tests are simple and quick to perform and enable accurate for easy differentiation ofthe 3 coagulase-positive Staphylococcus species.

cow; mastitis; differentiation.

Introduction In the last edition ofthe Bergey's Manual ofde• use ofone single test, the coagulase test (Evans terminative bacteriology (1986) three species 1965). However, sin ce it has been shown that of coagulase-positive staphylococci (CPS) are different strains ofS. aureus can have different listed; S. aureus (Buchanan & Gibons 1974), S. coagulase reaction (Watts et al. 1984; Fox et al. intermedius (Hajek 1976) and S. hyicus subsp . 1996) and that S. hyicus subsp. hyicus is coagu• hyicus (Devriese et al. 1978). CPS are often iso• lase variable (Devriese et al. 1978), additional lated from bovine milk samples, and in routine tests have to be used for an accurate differenti• laboratory diagnostics, generally, all strains of ation. CPS have been considered as S. aureus. How• Studies, using complementary tests to the coag• ever, it is today generally considered that CPS ulase test, have been carried out during recent in milk include not only S. aureus, but also S. years to determine a good combination oftests intermedius and S. hyicus (Harmon et al. 1993, for differentiation of the 3 CPS at specie level National Mastitis Council. 1987). Historically, (Botha & Brand 1987, Roberson et al. 1992). species in the genera Staphylococcus have been Several studies in different regions and coun• divided into coagulase-positive and coagulase• tries (Rampone et al. 1993, Calzolari et al. negative staphylococci, differentiated by the 1995, Roberson et al. 1996) have reported dif-

Acta vet. scand. vol. 40 no. 4, 1999 316 A. Capurro et al. ferent frequencies for S. aureus, S. interm edius, sified as CPS. Isolates which were identified as and S. hyicus, among CPS from bovine milk, CPS were additionally tested. identified by the use of a combination of tests. In all studies there were a high predominance of Storage ofisolates S. aureus. Each of the isolates was harvested in a tube The main objectives of the present report were (Nalgene cryogenics vials, Nalge Company, to study the relative occurrence ofS. aureus, S. Rochester, NY, USA) containing I mlofTryp• intermedius, and S. hyicus among CPS, isolated ticase soy broth (TSB) +15% glycerol, (Becton• from milk samples collected from bovine mas• Dickinson, San Jose, CA, USA) for prolonged titis cases in Sweden and to establish a relevant storage at -31°C. After storage each isolate was and accurate typing scheme for identification of subcultured twice, before it was tested. the 3 CPS species . Differentiation ofstaphylococcal species Materials and methods Differentiation of staphylococcal species was Bacterial strains carried out using the following tests in the given The CPS strains were obtained from bovine order. milk samples from mastitis cases, sent to a mas• In the test ofability to grow on P agar (Phillips titis routine laboratory (National Veterinary In• & Nash 1985), the medium was supplemented stitute, Uppsala, Sweden). The total number of with 7 J-lg ofacriflavin per ml according to Har• CPS strains included in this study was 414, mon et al. (199 I). The inoculation was per• among which 327 were from subclinical and 87 formed according to Roberson et al. (1992). from clinical mastitis . The samples were classi• Growth indicated S. aureus. In the acetoin test fied in material A and B according to the he• the medium was prepared according to Rober• molysis reaction on bovine blood agar of each son et al. (1992). S. aureus is in general positive strain. The terminology by Elek & Levy (1950) in this test. Anaerobic fermentation ofmannitol was used. Material A, used for frequency study, was performed as described by the Subcommit• consisted ofthe 177 strains irrespective oftype tee on Taxonomy ofStaphylococci and Micro• of hemolysis reaction presented. In this mate• cocci (1965). S. aureus is considered to be pos• rial all CPS isolated from milk samples during itive. Culture on purple agar (Difco, Detroit, IL, a specific period oftime were selected, with the USA) with I% maltose was prepared and inter• restriction ofjust one strain per herd. Material preted according to Quinn et al. (1994). S. au• B consisted of237 strains, without the bovine S. reus and S. intermedius show positive reactions. aureus strains typical double hemolysis zones Thej3-galactosidase test (Robertson et al. 1992) on bovine blood agar. The strains were CPS was performed using a commercially available randomly selected from different herds, with substrate tablet (Rosco, Taastrup, Denmark) the aim to increase the number ofstrains other (Mac Faddin 1980), according to instructions than S. aureus, available for testing . by the company. S. intermedius is considered to be positive. Identification ofisolates Hemolytic reaction on chocolate agar was per• Gram-positive strains of non-motile cocci formed as described by Liimmler (1991). S. hy• which produced catalase (National Mastitis icus has a positive reaction. Synergistic hemol• Council 1987) and coagulase (National Mas• ysis (CAMP-like) test was performed as titis Council 1987, Quinn et al. 1994) were c1as- described by Hebert & Hancock (1985). S. hyi-

Act a vel. scand . vol. 40 no. 4. 1999 Identification ofcoagulase-positive staphylococci 317

Table I . Number and frequency of coagulase-positive staphylococcal species in material A (all coagulase• positive strains) and B (strains without the bovine coagulase-positive S. aureus typical double hemolysis zone), divided into clinical and subclinical cases ofmastitis.

Material A Material B Species Clinical Subclinical Total % Clinical Subclinical Total %

S. aureus 36 137 173 97 48 182 230 97 S. intermedius I 2 3 2 I 4 5 2 S. hyicus 0 I I I II 2 I

Total 37 140 177 100 50 187 237 100 cus has a positive reaction. Aerobic fermenta• stable DNA:RNA hybrid. The labelled DNA: tion ofmannitol was tested in mannitol salt agar RNA hybrids are measured in a Gen-Probe lu• (Acumedia, Ljusne, Sweden) and interpreted minometer. according to Botha & Brand (1987). S. aureus and S. intermedius show positive reactions. Results Table 1 shows the number and frequencies of Control strains the different species obtained among the CPS. Control strains used in all tests were obtained In samples from clinical and subclinical mas• from a collection at the Department ofMastitis, titis cases and in both materials high predomi• National Veterinary Institute, Uppsala, Swe• nance ofS. aureus was seen. There were no dif• den. The control strains were: Staphylococcus ferences in the frequencies of the 3 species aureus Newman A275, Staphylococcus aureus between material A and B. Oxford 209, Staphylococcus intermedius Ll3, A simplified key (Fig. 1) was compiled for the Staphylococcus hyicus 8610193, and Staphylo• identification ofthe different CPS species. coccus hyicus 8612948. Table 2 shows the proportion of the tested strains positive in the 8 different tests. Four Accuprobe'' method hundred and three strains'that were positive on Ten, 8 and 3 strains ofS. aureus, S. intermedius P agar with acriflavin, and positive to the aero• and S. hyicus, respectively,were tested with Ac• bic and anaerobic mannitol fermentation test, cuprobe" Staphylococcus aureus culture iden• among which 98% and 80% were positive to ac• tification test (Gen-probe, San Diego, CA, etoin test and purple agar test, respectively, USA), specific for S. aureus of human origin. were determined as S. aureus. All these S. au• The test was performed according to Freney et reus strains were negative in the other tests al. (1993). This is a nucleic acid hybridization used. test which is based on the ability of comple• The 8 strains with positive reaction in the aero• mentary nucleic acid strands to specifically bic mannitol fermentation test andfJ-galactosi• align and associate to form stable double• dase test, were considered as S. intermedius. stranded complexes. A chemiluminescent la• Among these, 2 and 6 strains were also positive belled, single-stranded DNA probe, comple• to the acetoin test and purple agar test, respec• mentary to the ribosomal RNA of the target tively. All S. intermedius strains were negative organism, combines with the RNA to form a to the test on P agar with acriflavin, anaerobic

Act a vet. scand . vol. 40 no. 4, 1999 318 A. Capurro et al.

Gram-positive cocci I Catalase test (+) ---JI_ (_) Staphylococcus Streptococcus Micrococcus I Coagulase test

S.aureus Coagulase-negative spp. S. intermedius Micrococcus S. hyicus I Acetoin test P-agar test Anaerobic fermentation of mannitol test

(+) -----1------(_) S.aureus S. intermedius S. hyicus I Aerobic fermentation of mannitol test jI-galactosidase test

(+) 1 - (_) S. intermedius S. hyicus

Synergistic hemolysis (CAMP-like) test Hemolytic reaction on chocolate agar test

(+) S. hyicus

Figure I . A simplified key for differentiating coagulase-positive staphylococci. mannitol fermentation test, and the CAMP-like test and gave a hemolytic reaction on chocolate test and showed no hemolytic reaction on choc• agar and were determined as S. hyicus. These olate agar. strains were negative in all the other tests. Only 3 strains were positive in the CAMP-like The results ofAccuprobe'" test are shown in Ta-

Acta vet. scand. vol. 40 no. 4. 1999 Identification ofcoagulase-positive staphylococci 319

TabIe 2. Results of analyses of coagulase-positive staphylococci by8 different tests forcharacterization of S aureus, S. intermedius andS. hyicus. Thefigures aregiven as percentages of positive strains.

Test Species P agar Anaerobic Aerobic ft-galactosidase Acetoin Purple CAMP-like Hemolysis fermentation fermentation agar on chocolate ofmannitol ofmannitol agar

S.aureus 100 100 100 o 98 80* o o S intermedius o o 100 100 25 75** o o Shyicus o o o o o o 100 100

*A +++ positive test. **A ++positive test.

TabIe 3. Percentages of positive reactions among Calzolari et al. (1995) are not really compar• 21 coagulase-positive staphylococcal strains tested able. However, the frequencies of the different bytheAccuprobef test. species reported in these studies, as well as in Number of Accuprobe'" the present, are rather similar. Species strains positive Roberson et al. (1996) found, compared with stested strains (%) our findings, higher frequency of S. hyicus Staphylococcus aureus 10 100 (17.7%) and lower frequency ofS. intermedius Staphylococcus intermedius 8 o (0.2%). The differences might be due to the fact Staphylococcus hyicus 3 o that Roberson et al. (1996) had 41% of primip• arous cows in their trial. According to Nicker• son et al. (1995), primiparous cows have higher ble 3. All S. aureus strains were positiveand the prevalenceof S. hyicus. Furthermore, Roberson other CPS were negative. et al. (1996) used a different chromogeniccom• pounds in thejJ-galactosidase test. Discussion In the present study different tests for identifi• The 3 species were, more or less, equally dis• cation and differentiation of the 3 CPS species tributed among clinical and sub-clinical cases were evaluated, in order to find accurate but (Table 1). The frequencies of the different CPS also simple, inexpensive and quick tests. The species were similar in material A and B but tests should havea high accuracy in identifying slightly different to what has been reported in each species and identificationshould be based other studies. El-Sukhon, (1995) found higher on positive, rather than negative test results. frequency of S. hyicus (3.8%). This may be ex• The tests in the present study (Table 2) were plained by differences in the characterization of evaluated according to the these criteria. variousspecies. It is also knownthat in the pop• According to the results of the present work a ulation of S. hyicus subsp. hyicus some strains combination of the following tests is suggested havean irregular,often negative,but sometimes to accurately differentiate the 3 CPS species delayed positive coagulase reaction (Devriese from each other: P agar supplemented with et al. 1978,Sperber 1979). acriflavin test, jJ-galactosidase test and hemo• According to differences between the materials lytic reaction on chocolate agar. These tests are and tests used,the results from studies by Botha meant to supplement rather than replace the co• & Brand (1987), Rampone et al. (1993), and agulase test. They are rather quick, easily per-

Acta vel. scand. vol. 40 no. 4, 1999 320 A. Capurro et al.

formed and interpreted, and relatively inexpen• ogenic and non-pathogenic staphylococci . 1. sive. Pathol. Bacteriol. 1950, 62, 54 I-554. To identify a strain on the basis ofa small num• El-Sukhon SN: Characterization of staphylococci isolated from mastitic cows in Jordania. Bull ber of biochemical characteristics may be con• Anim. Health Prod. Afr, 1995,43, 23I-235. sidered a problem because it increases the risk Evans JB : Current views and problems relating to the ofmisidentification. Moreover, regarding S. au• taxonomy of the Micrococcaceae. Int. Bull. bact. reus and S. intermedius, there are strains that Nomencl. Taxon. 1965, 15, 111-112. Fox LK, Besser TE, Jackson SM: Evaluation ofa co• have weak utilisation, or fail to utilise certain agulase-negative variant Staphylococcus aureus substrates in these 3 tests. Therefore we suggest as a cause of intramammary infections in a herd that anaerobic fermentation of mannitol also of dairy cattle. 1. Am. Vet. Med. Assoc. 1996, should be used. To distinguish between S. inter• 209, 1143-1146. Freney J. Meugnier H, Bes M, Fleurette J: Identifica• medius and coagulase-positive S. hyicus with tion of Staphylococcus aureus using a DNA atypical or weak reactions, we recommend probe: Accuprobe'". Ann. Bioi Clin. 1993, 51, aerobic fermentation ofmannitol to be used. 637-639. The Accuprobe" test (Table 3), specific for hu• Hajek II:' Staphylococcus intermedius. a new species man S. aureus, worked well for the lOS. aureus isolated from animals. Int. 1. Syst. Bacteriol. bovine strains tested. This seems to be a simple 1976,26.401-408. Harmon RJ. Eberhart RJ. Jasper DE: Microbiologi• and secure method but needs further evaluation cal procedures for the diagnosis ofbovine udder for CPS species isolated from bovine milk. infection. Arlington VA, National Mastitis Coun• cil Inc, 1993. Acknowledgements Harmon RJ. Langlois BE, Akers K: A simple medium The laboratory work was carried out at the mastitis for the verification of identity ofStaphylo coccus laboratory, Department of Mastitis, National Veteri• aureus of bovine origin. 1. Dairy Sci 1991, 74. nary Institute, Uppsala, Sweden. Dr. Capurro was (Suppl I), 202. awarded a scholarship by the Swedish Foundation for Hebert GA. Hancock AH: Synergistic hemolysis ex• International Co-operat ion in Research and Higher hibited by species of staphylococci. 1. Clin. Mi• Education (STINT). crobiol. 1985,22.409-415. Liimmler C: Characterization ofStaphylococcus hyi• cus with the ATB 32 Staph System and with Con• References ventional Tests. 1. Clin. Microbiol. 1991, 29. Botha FS, Brand PAJ: A simplified key for identifica• 1221-1224. tion ofcoagulase-positive staphylococc i isolated Mac Faddin JF: In: Biochemical tests for identifica• from bovine milk. S. Afr. 1. Dairy Sci. 1987, I, tion of medical bacteria. 2nd edn. Williams & 39-44. Wilkins, Baltimore, Maryland, USA, 1980, pp Buchanan RE, Gibbons NE: Bergey's manual of de• 120-128. terminative bacteriology. 8th ed. The Williams & National Mastitis Council: Staphylococci. In: Labor• Wilkens Company. Baltimore. 1974. atory and field handbook on bovine mastitis. Ar• Calzolari A, Bettera S, Frigeiro C. Rampone H, Gi• lington, VA, 1987, p. 80. raudo J: Taxonomic classification and antibiotic Nickerson SC, Owens WE, Boddie RL: Mastitis in resistance of staphylococci isolated from bovine Dairy Heifers: Initial studies on prevalence and milk. Proceedings of the 3ed international mastitis control. 1. Dairy Sci. 1995, 78, 1607-1618. seminar, Tel Aviv, Israel, 1995, session 2, 78-79. Phil/ips E, Nash P: Culture media. In: Lennette, Bal• Devriese LA, Hajek V. Oeding P. Mey er SA, Schleifer lows, Hausler Jr, Shadomy (eds.): Manual ofclin• KH: Staphylococcus hyicus (Sompo/insky 1953) ical microbiology, 4th ed. American Society for comb. nov. and Staphylococcus hyicus subsp. Microbiology, Washington, DC 1985, 1051• chromogenes subsp. nov. Int. 1. Syst. Bacteriol. 1092. 1978,28, 482-490 . Quinn PJ. Carter ME, Markey B, Carter GR: Staph• Elek SD, Levy E: Distribution ofhaemolysins in path- ylococcus species. In: Clinical Veterinary Micro-

Acta vet. scand . vol. 40 no. 4, 1999 Identification ofcoagulase-positive staphylococci 321

biology, Quinn, Carter, Markey and Carter (eds.) Sammanfattning Wolfe publishing, London, 1994, p. 118. Identifiering av koagulaspositiva stafylokocker isole• Rampone H. Bogni C. Giraudo J. Calzolari A: Iden• rade fran bovin mjolk. tification of staphylococci from milk in Argen• tina. Zbl. Bakt. 1993,279.537-543. En undersokning har genomforts dar sammanlagt Roberson JR , Fox LK. Hancock DD, Besser TE: 414 koagulaspositiva stafylokockstammar studerats. Evaluation of methods for differentiation of Co• Stammama var isolerade fran rnjolk fran fall av mas• agulase-Positive staphylococci. 1. C1in. Micro• tit i samband med rutinverksamheten vid Avdeln• bioI. 1992,30.3217-3219. ingen for mast it, Statens Veterinarmedicinska An• Roberson JR. Fox LK, Hancock DD, Gay JM, Besser stall , Uppsala, Sverige. Av dessa 414 stammar ingick TE: Prevalence of coagulase-positive staphylo• 177 stammar i en frekvensstudie, varvid 97% av cocci, other than Staphylococcus aureus, in bo• stammama identifierades som Staphylococcus au• vine mastitis. Am. 1.Vet. Res. 1996, 57, 54-58. reus, 2% som Staphylococcus intermedius och 1% Sperber WH: The identification of staphylococci in som Staphylococcus hyicus. Ytterligare 237 stammar clinical and food microbiology laboratories. Crit. med atypisk hemolysreaktion pAnotblodagar valdes Rev. Clin . Lab. Sci., 1979, 7, 121-184 . ut i syfte att oka antalet Staphylococcus intermedius• Anonymous: Subcommittee on Taxonomy of'Staphy• och Staphylococcus hyicus-stammar tillgangliga for lococci and Micrococci: Recommendations. Int. testning. Alta olika karaktarisktika, inkluderande fy• Bull. Bacteriol. Nomencl. Taxon. 1965, 15. 109• siologiska, enzymatiska och biokemiska egenskaper, 110. anvandes fOratt identifiera de koagulaspositiva stafy• Watts JF, Pankey W. Nickerson SC : Evaluation of lokockartema. Baserat pA resultatet av denna studie Staph- Evaluation Ident and STAPHase Systems foreslas att foljande tester kan anvandas fOren saker for Identification of Staphylococci from Bovine identifiering av de 3 olika artema av koagulaspositiva Intramammary Infections. 1. C1in. Microbiol. stafylokocker: P agar med acriflavin, p-galactosidas 1984, 20. 448-452. och hemolytisk reaktion pa chokladagar. Dessa 3 tester ar okomplicerade och snabba att genomfora vilket mojliggor en enkel, snabb och korrekt diffe• rentiering av de 3 koagulaspositiva stafylokockar• tema.

(Received May 25, 1998; accepted August 2, 1999).

Reprints may be obtained from : C. Concha, Department of Mastitis, National Veterinary Institute, S-751 89 Uppsala, Sweden. E-mail : [email protected], tel: +4618674260, fax: +4618309402.

Acta vel. scand , vol. 40 no. 4. 1999