Bovine mastitis caused by Streptococcus uberis is a common problem in pasture-based dairying systems. This study examines both the ecology of S. uberis and infection of the bovine mammary gland on a New Zealand dairy farm. Initially, the REP-PCR strain typing method was developed and the potential of MALDI-TOF mass spectrometry evaluated as a strain typing method. While strain-specific mass spectra were obtained with MALDI-TOF mass spectrometry, the irreproducibility of spectra was its major downfall. With further work, this rapid method could be very useful for strain typing S. uberis on a large scale. Using optimised REP-PCR and anchored typing methods, multiple S. uberis strains were isolated and strain typed from the dairy environment, including farm races and paddocks, faeces, teat skin, the cow body and from intramammary infections. High strain diversity was observed in all sampled locations; however, some strains were found at more than one site, suggesting transmission may occur between the environment and cows. The most likely means of S. uberis distribution throughout the dairy farm was via excretion with faeces and, although not all cow faeces contained this pathogen, the gastrointestinal tract of some cows appeared to be colonised by specific strains, resulting in persistent shedding of this bacteria in the faeces. Infection of the mammary gland is likely to occur through contamination of the teat skin with highly diverse environmental strains of S. uberis. However, only one or two strains are generally found in milk from mastitis cases, suggesting that, although infection may arise from a random or opportunistic event, a strain selection process may take place. Intramammary challenge with multiple strains of S. uberis revealed that selection of a single infective strain can occur within the mammary gland. The predominance of one strain over others may be related to production of virulence factors allowing enhanced ability to establish in the gland and evade the immune response, or due to direct competition between strains through the production of antimicrobial factors such as bacteriocins. In addition to strain-specific factors, the individual cow and quarter response may play an important role in the development of infection and selection of the infective strain. Using results from this study, a model of S. uberis strain transmission has been proposed, which includes potential mechanisms of infection and persistence of S. uberis within the mammary gland.

Streptococcus uberis, an environmental organism also associated with dairy animals, is a common and persistent cause of bovine mastitis. New approaches to control these infections need to be identified. One such strategy may be the application of bacteriocins; proteinaceous antimicrobials elaborated by bacteria that typically inhibit the growth of strains closely related to the producer organism. The well-characterized lactococcal bacteriocin nisin is the active ingredient in two commercial products currently in use for the prevention of mastitis. However, reports of resistance development have prompted the investigation of alternative bacteriocins to be used in conjunction with nisin in 'bacteriocin cocktails' designed to have more comprehensive inhibitory activity against mastitis pathogens. The bacteriocins of gram-positive bacteria have been divided into four distinct classes: (I) lantibiotics, (II) non-lantibiotic peptides, (III) large proteins, and (IV) circular peptides. Although it has been known for more than twenty years that S. uberis commonly produce bacteriocin-like inhibitory substances (BLIS), none had been characterised prior to the present study. The first step in the current investigation was a survey of the BLIS activities of a set of fifteen S. uberis and S. bovis strains against a set of standard indicators as well as common gram-positive mastitis pathogens. Additional tests using a deferred antagonism agar plate-based assay showed that some of the BLIS activities were heat-sensitive and their production was influenced by the presence of either blood or a fermentable carbohydrate source in the test medium. On the basis of the results obtained from these tests it became apparent that S. uberis and S. bovis may commonly produce more than a single inhibitory agent. S. uberis 42 became the focus of this study because (a) it had broad inhibitory activity against mastitis-associated bacteria, (b) it did not display cross-resistance to nisin, and (c) from the preliminary screening results it appeared to produce both heat-stable and heat-labile inhibitory agents. Acid extracts of S. uberis 42 cells yielded inhibitory activity that, when fractionated by reversed-phase HPLC, yielded a peptide of 3029 Da. Although this peptide was blocked to Edman degradation at position 2, following propanethiol-modification a 20-amino acid sequence was obtained. Degenerate primers to lantibiotic biosynthesis gene homologs were used to initiate inverse PCR and primer walking, ultimately yielding a 15-kb contiguous sequence encompassing 11 genes typical of those involved in lantibiotic synthesis, regulation and immunity. Due to the close similarities to nisin of the S. uberis 42 lantibiotic precursor (78%), and the organisation and composition of the locus, this inhibitor was named nisin U. Nucleotide sequences homologous to insertion sequences were detected in the vicinity of the nisin U locus, and indicate a possible mechanism of acquisition of this locus by S. uberis. The locus was detected in ten other S. uberis, and also in two S. agalactiae and two S. thoraltensis strains, and in one S. porcinus and one S. pluranimalium strain. The amino acid sequences of some of these differed in one or two amino acids, and these variants were named nisin U2 and nisin U3 accordingly. Nisin U, the two nisin U variants, and nisin A exhibited cross-immunity (i.e. all of the producer strains were insensitive to each form of nisin) and cross-inducibility (i.e. all of the producer strains displayed enhanced production when exposed to each form of nisin). Nisin U did not contribute to the entire spectrum of inhibitory activity of S. uberis 42. Freeze thaw extracts of S. uberis 42 agar cultures yielded heat-labile inhibitory activity that was inhibitory to L. lactis A5, a producer of nisin Z. Subsequent purification by cation-exchange chromatography, gel filtration, and reversed-phase HPLC yielded a peptide of mass 7048 Da, which was resistant to Edman degradation. Digestion with chymotrypsin released an 819 Da peptide fragment of sequence NH₂-KAQAVIW-COOH. Tn916 mutagenesis of S. uberis 42 enabled the identification of the genetic locus of the inhibitor, comprising six genes potentially involved in its biosynthesis and immunity. The detection of a pair of flanking 159-bp direct repeats indicates possible acquisition of the locus by 'long target duplication'. The inhibitor was inferred to be a circular peptide, on the basis of its behaviour to Edman degradation, and by comparison of its locus with that of other circular bacteriocins. On the basis that the purified peptide appears to induce lysis in sensitive bacteria, although by an as-yet unidentified mechanism, the inhibitor was named uberolysin. The uberolysin structural gene was detected in eight other strains of S. uberis, however not all of these appeared to be producing active inhibitor. No bacteriocins closely resembling the two reported in this thesis have been demonstrated previously to be produced by members of the genus Streptococcus. The remarkable diversity in the structures, activity spectra and basic modes of action of these two bacteriocins produced by a single strain of S. uberis, combined with the observation of apparent greater heterogeneity in properties of a preliminary sampling of BLIS-producing strains, indicates that these bacteria may be an important source of novel antimicrobials of potential value for the treatment of mixed bacterial infections and for minimising potential resistance development.

Bovine mastitis is one of the most costly dairy-based diseases worldwide. Streptococcus uberis is a prevalent causative organism of mastitis and resides naturally in the environment of the dairy cow making prevention of the disease difficult. New strategies need to be developed to control this pathogen. However, a fundamental understanding of the complex relationships that exist between the cow, the pathogen and the environment are required in order to advance the development of prevention strategies. Microarray technology was used to evaluate the complex transcriptional changes which occur in the bovine mammary gland following the onset of clinical S. uberis mastitis. A 22,000 bovine cDNA microarray indicated that S. uberis mastitis led to the up-regulation of 1,283 genes and the down-regulation of 1,237 genes by greater than 1.5 fold. Gene ontology analysis demonstrated that S. uberis mastitis was typically associated with the up-regulation of genes that are involved in the immune response and homeostasis and a down-regulation of genes involved in lipid metabolism. Quantitative real-time analyses for a selection of genes associated with the immune response validated the microarray data. Mammary epithelial cell cultures did not show an increase in the expression of any of these immune factors in response to the same S. uberis strain used to induce clinical mastitis. This indicates that the expression of immune-related genes by mammary epithelial cells may be initiated by host factors and not S. uberis. The application of bacteriocins, proteinaceous antimicrobials produced by bacteria which typically inhibit the same or closely-related species to that of the producer organism, has been suggested as one possible approach in the control of mastitis. S. uberis have been previously found to commonly produce bacteriocin-like inhibitory substances (BLIS). The BLIS activities of a set of fifteen S. uberis and S. bovis strains were assessed. The results confirmed the prolific and varied nature of BLIS production by S. uberis and S. bovis and also indicated that these strains may commonly produce more than one inhibitory agent. This survey of BLIS production led to the detection and characterisation of a novel circular bacteriocin, uberolysin, produced by S. uberis strains 233 and 42. The structural gene of uberolysin was subsequently identified in nine (64%) of the fifteen test strains. Multiplex PCR analysis showed that 93% of 158 New Zealand S. uberis isolates contained the structural genes of at least one of the four known S. uberis bacteriocins (uberolysin, nisin U, ubericin A and ubericin 63). However, no apparent direct association was identified between any one of these bacteriocin-related loci and apparent ability to cause mastitis on New Zealand dairy farms. The uberolysin structural gene was detected in 91% of the isolates and this widespread distribution prompted the advancement and evaluation of a potential role for uberolysin in immunomodulation within the bovine mammary gland. Two different preparations of uberolysin were found to have different stimulatory effects on monocytes, neutrophils and epithelial cells. The less highly purified preparation appeared to diminish the production of TNF-α by monocytes in the presence of a bacterial stimulus and to decrease neutrophil phagocytosis. By contrast, the relatively more highly purified preparation of uberolysin itself induced a significant immune response by monocytes. Consistent with this, the purer preparation of uberolysin induced an increase in C3, IL-1β, IL-6, IL-8, the β-defensin LAP, the acute-phase protein MSAA, the calcium-binding protein S100A12 and TLR2 by quantitative real-time analysis. Although currently only two S. uberis bacteriocins (uberolysin and nisin U) have been fully characterised, the present study has shown that this species may be an important source of novel antimicrobials. Furthermore, bacteriocin production by S. uberis may have an immunomodulation role within the mammary gland. A better understanding of the complex immune response initiated at the onset of clinical S. uberis mastitis and of the role that bacteriocins have in S. uberis pathogenesis may lead to development of improved strategies to combat this disease.

Streptococcus uberis is an important cause of intramammary infection in dairy cattle. Strains of S. uberis appear to differ in their ability to cause disease based on previous epidemiological studies. We explored the pathogenicity of 2 strains of S. uberis, where one strain represented a putatively host-adapted type based on its ability to cause persistent infection and to spread from cow to cow in a lactating herd. This type was part of a clonal complex that is commonly associated with bovine mastitis. The other strain, which was isolated from a transient infection in a single animal in the same herd and did not belong to any known clonal complex, was selected as putatively non-adapted type. Cows (6 per strain) were experimentally challenged in a single hind quarter and the adjacent hind quarter was used as mock challenged control quarter. Both strains showed an equal ability grow in milk of challenge animals in vitro. All cows that were challenged with the putatively host-adapted strain developed clinical signs of mastitis, including fever and milk yield depression as well as elevated somatic cell count due to influx of polymorphonuclear leucocytes and lymphocytes. The cytokine response followed a specific order, with an increase in IL-1β, IL-6 and IL-8 levels at the time of first SCC elevation, followed by an increase in IL-10, IL-12p40 and TNF-α levels approximately 6 h later. In 4 of 6 animals, IL-17A was detected in milk between 57 and 168 h post challenge. The increase in IL-17A levels coincided with inversion of the pre-challenge CD4+:CD8+ T lymphocyte ratio, and was observed from 96 h post challenge. This was followed by normalisation of the CD4+:CD8+ ratio due to continued increase of the CD8+ concentration up to 312 h post challenge. Spontaneous resolution of infection was observed in 5 animals and coincided with a measurable IL-17A response in 4 animals, suggesting that IL-17 may be involved in the resolution of intramammary infection. To explore the mechanism of action of IL-17A we stimulated bovine PMN and bovine blood derived macrophages with recombinant IL-17A in vitro. IL-17A enhanced the killing ability of phagocyte toward the challenge strain. With the exception of minor elevation of IL-8 levels, no clinical, cytological or immunological response was detected in quarters challenged with the non-adapted strain. The observed strain specific pathogenicity was consistent across animals, implying that it is determined by pathogen factors rather than host factors. We further studied in vitro possible mechanisms involved in the differences observed between the two strains such as ability to adhere to the mammary epithelial cells, ability to resist to killing by phagocytes and ability to form biofilm. The adapted strain FSL Z1-048 showed an increased ability to adhere to the epithelial cells and an increase ability to resist to killing of monocyte derived macrophages. These mechanisms thus could potentially explain the in vivo observations.

Mémoire numérisé par la Direction des bibliothèques de l'Université de Montréal.

The second messenger cyclic diguanylate monophosphate (c-di-GMP) has been proven to be a central regulator for physiological and metabolic processes including biofilm formation and sessile to motile transitioning (1,2). The synthesis and degradation of c-di-GMP are regulated by GGDEF- respectively EAL-domain proteins. Recently, c-di-GMP has been discovered in the Gram-positive Streptococcus genus including Streptococcus gallolyticus, which showed to have diguanylate cyclase activity (3). Characterisation of the c-di-GMP network in other Streptococcus is of relevance. Hence, the aim of this project was the assessment of the GGDEF- and EAL domains from the animal pathogenic Streptococcus uberis and Streptococcus henryi. In vivo recombination cloning was used for the analysis of the GGDEF, EAL and GGDEF-EAL domain proteins from S. uberis and S. henryi. The cloning was unsuccessful for most of the domain proteins, except, for EAL 2 of S. uberis. However, analysis of the sequencing results for the cloned EAL 2 presented mutations. Further studies testing alternative cloning methods should be applied. Research regarding probiotic streptococci is also of interest. Therefore, isolation of Streptococcus salivarius from human oral samples using Streptococcus Selection Agar was conducted. Isolation of S. salivarius from human saliva and tongue samples was successful using Streptococcus Selection Agar. Other Streptococcus spp., Lactobacillus, Staphylococcus, and additional bacterial species were also isolated.