BOVINE COAGULASE · NEGATIVE STAPHYLOCOCCI : BIOCHEMISTRY AND POLYMERASE CHAIN REACTION

Hum mel, R., G . L e h man n : Bovine coagulase-negative Staphylococci: Biochemistry and Polymerase Chain Reaction. Acta vet. Brno, 63,1994: 133-139. Staphylococcus (S.) simulans, S. chromogenes, and S. epidermidis were the species of coagulase-negative staphylococci (CNS) most frequently involved in subclinical mastitits of dairy cattle. The same biochemical features characterized the successive isolates of a particular chronically infected udder quarter. Occurrence rate of the various species differed between the herds examined. In cases of clinical mastitis of cows S. xylosus, S. simulans, S. haemolyticus, and S. chromogenes predominated. A comparison between strains of the particular CNS species from infections in man and cattle did not show host attributable biochemical features or profiles obtained by polymerase chain reaction using arbitrary primers (AP-PCR). Bovine mastitis, Coagulase-negative staphylococci, biochemical features, polymerase chain reaction Coagulase-negative staphylococci (CNS) can by frequently isolated from milk samples of dairy cows, often in association with subclinical mastitis (T i m m s et al. 1987; H arm 0 n and Lan g I 0 i s 1989; Watt s et al. 1989; Davidson et al. 1992; Matthews et al. 1992; Todhunter et al. 1993). For differentiation between species of CNS several approaches have been supposed. Typing schemes primarily based on panels of biochemical tests have often been applied in tandem with other methods including composition of cell wall, isoenzyme pattern, polymorphism of whole cell protein and penicillin binding protein, pattern of rRNA gene and of endonuclease cleaved DNA, polymerase chain reaction using arbitrary primers (AP-PCR). By AP-PCR Welsh and McClelland (1990) obtained species specific patterns of AP-PCR products. Moreover, Staphylococcus (S.) haemolyticus of human and non-human primate origin produced different AP-PCR patterns related to their ecological origin and supposedly to host specifity of strains. The purpose of this work is to study (1) involvement of the various CNS species in udder infections of cows and (2) comparison between strains within the particular CNS species originating from man and cattle by means of AP-PCR. Materials and Methods Bacterial strains The CNS examined in this study originated from cases of clinical mastitis (54 strains) and of subclinical mastitits (57 strains) in cattle. For comparison, 8 strains from infections in man were included in these studies. Of the strains from bovine clinical mastitis 41 were isolated from udder secretions in 39 dairy herds submitted for bacteriological examination to the regional veterinary laboratory. An additional 13 strains originated from Belgium. The 57 strains from subclinical mastitis were selected in five large herds of cattle (Table 1) with 300 to 500 cows each. An infection of the mammary gland means isolation of CNS in pure culture together with an elevated total cell count (>300 000 cells/ml milk). Sampling of udder secretions was carried out before and after milking. Repeated isolates from a particular udder quarter possessed of the same biochemical features were considered a single strain. The 57 strains represented a total of 580 isolates. All of the herds practised post milking teat dipping but only two of them (herds Np and Lu) applied a proper hygiene at milking. There was an occasional history of Streptococcus agalacriae in these herds and the infected cows were removed for slaughter. In herd Np quarter milk samples from 463 cows at freshening were investigated for the presence of CNS infections. 21 cows with 29 chronically CNS infected quarters in herd Np were resampled monthly or bimonthly to the next freshening. The 29 strains represented a total of 502 isolates. In herds Mo and Zi, respectively, 7 out of 197 and 3 out of 293 cows could be found chronically infected. Three consecutive samplings were performed and a total of 42 isolates gathered. Herds Ka and Lu have been sampled only once.


Bacterial strains
The CNS examined in this study originated from cases of clinical mastitis (54 strains) and of subclinical mastitits (57 strains) in cattle.For comparison, 8 strains from infections in man were included in these studies.
Of the strains from bovine clinical mastitis 41 were isolated from udder secretions in 39 dairy herds submitted for bacteriological examination to the regional veterinary laboratory.An additional 13 strains originated from Belgium.
The 57 strains from subclinical mastitis were selected in five large herds of cattle (Table 1) with 300 to 500 cows each.An infection of the mammary gland means isolation of CNS in pure culture together with an elevated total cell count (>300 000 cells/ml milk).Sampling of udder secretions was carried out before and after milking.Repeated isolates from a particular udder quarter possessed of the same biochemical features were considered a single strain.The 57 strains represented a total of 580 isolates.All of the herds practised post milking teat dipping but only two of them (herds Np and Lu) applied a proper hygiene at milking.There was an occasional history of Streptococcus agalacriae in these herds and the infected cows were removed for slaughter.
In herd Np quarter milk samples from 463 cows at freshening were investigated for the presence of CNS infections.21 cows with 29 chronically CNS infected quarters in herd Np were resampled monthly or bimonthly to the next freshening.The 29 strains represented a total of 502 isolates.In herds Mo and Zi, respectively, 7 out of 197 and 3 out of 293 cows could be found chronically infected.Three consecutive samplings were performed and a total of 42 isolates gathered.Herds Ka and Lu have been sampled only once.

41
MST, strains from cases of clinical mastitis from 39 herds of cattle; x, somatic cell count from quarter ntilk samples of CNS infected cows only; y, chronically infected quarters; z. quarters sampled but once; ne.not examined.

Sampling procedures
Prior to sampling, the udder of each cow was washed, dried, and desinfected thoroughly, and about IS ml offoremilk from each quarter were discarded.Than the two individual quarter samples, one before, one after automatic milking, were collected in separate test tubes.Samples were kept on ice until starting the laboratory examination within the next 6 to 15 hours.

Microbiology and somatic cell count
Each milk sample (0,1 ml) was streaked onto the quadrant of a 3% blood agar plate.Plates were incubated for 48 hours at 37°C.A single colony of each primary culture of abundant growth (at least 10 colonies) yielding pure cultures with only one morphologically homogenous type of CNS was selected for further studies.Somatic cell counts (SCC) were done using a Fossomatic cell counter (NSN, Hillerod, Denmark).

Biochemical characterization of CNS
Methods described by K I 0 0 s et al. (1974) were applied for biochemical characterization of strains.In addition, sensitivity to furazolidone was tested according to Rheinha ben and Hadlock (1981); sensitivity to novobiocin according to De v ri e s e (1979).Production of phosphatase, urease, hyaluronidase, heat stable DNase and splitting of Tween 80 was studied as described by Devriese et al. (1883).Production of glucosidase and glucoronidase was determined according to W hit e et al. (1990).Inhibition of delta-hemolysin was detected as described by S kalka (1991).Strains were identified to genus level on the basis of culture and cell morphology, pigmentation, Gram stain, catalase activity, hemolysis, clumping factor, coagulase of rabbit and human plasmas and sensitivity to furazolidone.The identification to species level was based on features contained in table 2. Production of heat stable DNase, hyaluronidase, splitting of Tween 80 and inhibition of delta-hemolysin were only determined for differentiation between S. hyicus and S. chromogenes.From the reaction patterns thus obtained CNS species were determined by using the data base of Geary et al. (1989) modified at the Institute for Experimental Epidemiology, Wernigerode.

Micrococcus luteus (CCM 169).
AP-PCR A total of 50 strains of bovine and 8 of human origin were examined.Table 3 shows the origin of the strains.Genomic DNA of the strains was isolated by the methodofB ialkow ska-Hobrzan ska etal.(1990).AP-PCR procedure was carried out according to Williams et al. (1990) using the following three IOmerprimers already applied by Williams et al. (1990): namely 5'TCACGATGCA(designatedprimerl),5'ACGGTACACT(primer2),5'CACATGCTTC(primer3).Theprimers have been synthesized by Dr. Birch-Hirschfeld, Institute for Molecular Biotechnology, Jena.Concentration of the primers used was 4 JAM and 0,2 JAM for both primers I and 2, while 1 f.l.M was applied for primer 3. Two runs of PCR experiments in duplicate tubes were carried out with each of the template DNA.The reagent control that should not produce any amplification product contained all components used for AP-PCR but without addition of template DNA.Amplification products were analyzed by electrophoresis in 1,4% agarose gel and detected by staining with ethidium bromid.Size markers were the 1 kb DNA ladder from GmCO BRL.

Mastitis and CNS species
The occurrence of subclinical mastitis associated with CNS in 1257 cows in the five herds of cattle, mean values of somatic cell count (SSC), and CNS species involved are presented in Table 1.On the basis of biochemical features six CNS species were identified among the 57 strains recovered from the five dairy herds.The same biochemical features characterized the successive isolates of a particular quarter, and the successive isolates were considered a single strain.Occurence and prevalence of the various species differed between the five herds of cattle, but, S. simulans, S. chromogenes, and S. epidennidis predominated in chronical infections.Among the 41 CNS strains from 39 farms isolated from quarters with clinical mastitis, the four species S. xylosus, S. simulans, S. haemolyticus, and S. chromogenes were found in relatively high frequency (Table 1), while another four species (S. warneri, S. epidennidis, S. saprophyticus, and S. cohniO were infrequent.

Characteristics of the CNS strains
The features of the CNS strains investigated by biochemical methods are shown in Table 2.All of the strains could be identified to species level with high certainty except for some of the S. chromogenes strains.9 out of the 26 S. chromogenes strains were classified being S. chromogenes despite their ability to split Tween 80 and their lack in pigment production.
The various patterns of biochemical reactions within the particular CNS species were detected among the strains originating from the same herd of cattle, from different dairy herds as well as among the strains from diseases in man.

AP-PCR
The AP-PCR experiments performed with 58 strains representing four different CNS species resulted in reproducible patterns of amplified DNA.
Origin and species of the strains tested, the patterns of amplified DNA, and the number of different AP-PCR profiles observed within each of the species are listed in Table 3.Here, an AP-PCR profile includes all patterns of amplified DNA generated with a given strain by each of the three primers at the various concentrations applied.
For the species S. chromogenes, S. epidennidis, S. simulans and S. xylosus two, three, five, and six AP-PCR profiles, respectively, were observed.There was no correlation between the groups of AP-PCR profiles within a CNS species and the origin of strains neither from cattle of different dairy herds nor from man or cattle.
Between one and five amplificates resulted from each fingerprinting.The size of the amplificates increased in most of the cases as the concentration of a primer in the AP-PCR reaction decreased.Each primer at each of the concentrations used gave different patterns of DNA amplificates with a potential of detecting polymorphisms between species and between groups of strains.Some of these patterns observed within a particular species, however, were rather similar or even the same.
In identification the eNS strains studied by biochemical methods we met no difficulties except for 9 strains assigned to S. chromogenes.They take a somewhat intermediate position between S. chromogenes and S. hyicus, a finding already ruled out by Skalka (1991).Two of the crucial features discriminating both species were atypical for S. chromogenes: the strains produced pigment weakly or not at all and splitted Tween 80.The AP-PCR studies with our strains revealed only weakly pronounced species specific profiles, while WeI shand M c C 1 e 11 and (1990) clearly distinguished CNS species by AP-PCR although a few of their AP-PCR products were shared between species.Host attributable amplificate profiles within the CNS species studied could not be detected with our strains indicating a close relationship between the strains from man and cattle.Especially, the amplificate patterns and profiles were quite similar within the species S. epidermidis.Findings yielded by ribotyping seem to provide a comparable view.DeB u y s e r et al. (1992) obtainedanumberof equalhybridization patterns with s1Iainsof dift'erenthostorigin belonging to a same CNS species by ribotyping using ribosomal RNA gene probes and restriction endonuclease cleaved DNA.For example, equal hybridization patterns were observed for S. xylosus strains isolated from man, goat, and horse, or for S. epidermidis, S. warneri, and S. lugdunensis strains from man and goat.On the contrary, in strains from human and .nonhuman primates, WeI s chand M c C 1 ella n d (1990) observed different AP-PCR pat-terns indicating the existence of host adapted subspecies.For eNS of human and non other primate origin Kloos and W olfshohl (1979;1983) confirmed the hypothesis of a conjugate evolution of microorganisms with their host by DNA hybridization.The question rises whether the conditions in animal husbandry favour a mutual exchange of eNS strains between man and animals, thus being the reason for a close relationship between them.To answer this question, DNA homology studies with human and bovine strains of the particular eNS species are needed that can provide the most exacting data to determine the taxonomic status and relatedness of the species and the existence of subspecies.

Table I
CNS iDtectioos In herds of cattle I I