Characterisation of Streptococcus uberis from bovine milk

Abstract: Salem Abureema

The goal of this thesis was to investigate three main aspects of mastitis due to Streptococcus uberis; (i) the epidemiology of clinical bovine mastitis, including the relative importance of cow-to-cow transmission and acquisition from an environmental source as well as the occurrence of persistent infection and re-infection, (ii) factors affecting biofilm formation of S. uberis, and (iii) the virulence factors of S. uberis.

Two hundred and twelve Streptococcus uberis isolates were obtained from clinical cases of mastitis in dairy cattle from different farms in Gippsland, Victoria, Australia. Approximately 28% (n = 60) of the isolates were from recurrent infections. Among these 60, S. uberis isolates from 27 cows, with 34 different DNA restriction patterns were observed, indicating a wide variety of S. uberis strains causing clinical mastitis. Most recurrences (n = 20) were due to a new strain, indicating that treatment of the initial infection was usually successful. There were five examples of the same strain being isolated from different cows on the same farm, suggesting either cow-to-cow transmission at milking or acquisition from a common environmental source.

Twenty-seven isolates of S. uberis were tested for their ability to form biofilms on a polystyrene surface in Todd Hewitt Broth, and to produce extracellular polysaccharide, assessed on Congo red agar. Biofilm production by S. uberis occurred in about half of the isolates from cows with or without mastitis, but most produced lower amounts of biofilm than the reference strain, Staphylococcus epidermidis RP62A. Most isolates produced polysaccharide according to their colonial morphology on Congo red agar. The addition of skim or whole pasteurised milk markedly increased the amount of biofilm formation of S. uberis isolates, suggesting that milk or its components contribute to biofilm production. Isolates from cows without evidence of mastitis produced more biofilm in the presence of milk than isolates from cows with clinical or subclinical mastitis. Casein and β-lactoglobulin over a range of concentrations found in vivo, did not enhance biofilm formation under the test conditions used. Lipids in milk are unlikely to contribute to biofilm formation as skim milk and whole milk behaved is a similar way. Lactoferrin did not appear to play a major role in biofilm formation by most isolates of S. uberis, but did enhance biofilms of three isolates from cows with low milk somatic cell counts. Based on our findings, S. uberis seems to be capable of biofilm formation, and the amount produced depends on culture conditions and differs between individual isolates.

A microarray assay was developed to analyse putative virulence factors among different S. uberis isolates. Suppressive subtractive hybridization, using a pool of isolates from cases of clinical mastitis (tester pool) and a pool from cows with low somatic cell counts in the milk (driver pool), was used to generate a library of tester-specific clones that were printed onto the array. Following hybridization with DNA from 29 S. uberis isolates, a dendogram showed five major clusters, which clearly differentiated 20 clinical isolates from nine low cell count isolates. The first cluster contained seven S. uberis isolates; among them just one isolate from a cow with clinical mastitis. This clearly indicated that these isolates lack a statistically significant linkage with disease status and makes them unlikely candidates for S. uberis virulence-associated factors. The second cluster was mixed, consisting of seven isolates; four from cows with clinical mastitis and three from cows with low cell counts in their milk. These S. uberis isolates may represent opportunistic pathogens, possibly able to cause infections in particular groups of cows with increased susceptibility to infection. The third, fourth or fifth clusters contained 15 S. uberis isolates from cows with clinical infection, but none from cows with low milk somatic cell counts. The strain-specific features together with the fourteen features chosen by principal component analysis were selected for sequencing. Importantly, the sequence of the probe W2-17, which is specific for S. uberis clinical isolates, had no significant match in the database and therefore, this probe is potentially a novel virulence marker gene.