Characterisation of bacteriophage-host interactions in Lactococcus lactis

Lactococcus lactis is used extensively world-wide for the production of fermented dairy products. Bacteriophages (phages) infecting L. lactis can result in slow or incomplete fermentations, or may even cause total fermentation failure. Therefore, bacteriophages disrupting L. lactis fermentation are of economic concern. This thesis employed a multifaceted approach to investigate various molecular aspects of phage-host interaction in L. lactis. The genome sequence of an Irish dairy starter strain, the prophage-cured L. lactis subsp. cremoris UC509.9, was studied. The 2,250,427 bp circular chromosome represents the smallest among its sequenced lactococcal equivalents. The genome displays clear genetic adaptation to the dairy niche in the form of extensive reductive evolution. Gene prediction identified 2066 protein-encoding genes, including 104 which showed significant homology to transposase-specifying genes. Over 9 % of the identified genes appear to be inactivated through stop codons or frame shift mutations. Many pseudogenes were found in genes that are assigned to carbohydrate and amino acid transport and metabolism orthologous groups, reflecting L. lactis UC509.9’s adaptation to the lactose and casein-rich dairy environment. Sequence analysis of the eight plasmids of L. lactis revealed extensive adaptation to the dairy environment. Key industrial phenotypes were mapped and novel lactococcal plasmid-associated genes highlighted. In addition to chromosomally-encoded bacteriophage resistance systems, six functional such systems were identified, including two abortive infection systems, AbiB and AbiD1, explaining the observed phage resistance of L. lactis UC509.9 Molecular analysis suggests that the constitutive expression of AbiB is not lethal to cells, suggesting the protein is expressed in an un/inactivated form. Analysis of 936 species phage sk1-escape mutants of AbiB revealed that all such mutants harbour mutations in orf6, which encodes the major capsid protein. Results suggest that the major capsid protein is required for activation of the AbiB system, although this requires furrther investigations. Temporal transcriptomes of L. lactis UC509.9 undergoing lytic infection with either one of two distinct bacteriophages, Tuc2009 and c2, was determined and compared to the transcriptome of uninfected UC509.9 cells. Whole genome microarrays performed at various time-points post-infection demonstrated a rather modest impact on host transcription. Alterations in the UC509.9 transcriptome during lytic infection appear phage-specific, with a relatively small number of differentially transcribed genes shared between infection with either Tuc2009 or c2. Transcriptional profiles of both bacteriophages during lytic infection was shown to generally correlate with previous studies and allowed the confirmation of previously predicted promoter sequences. Bioinformatic analysis of genomic regions encoding the presumed cell wall polysaccharide (CW PS) biosynthesis gene cluster of several strains of L. lactis was performed. Results demonstrate the presence of three dominant genetic types of this gene cluster, termed type A, B and C. These regions were used for the development of a multiplex PCR to identify CW PS genotype of various lactococcal strains. Analysis of 936 species phage receptor binding protein phylogeny (RBP) and CW PS genotype revealed an apparent correlation between RBP phylogeny and CW PS type, thereby providing a partial explanation for the observed narrow host range of 936 phages. Further analysis of the genetic locus encompassing the presumed CW PS biosynthesis operon of eight strains identified as belonging to the CW PS C (geno)type, revealed the presence of a variable region among the examined strains. The obtained comparative analysis allowed for the identification of five subgroups of the C type, named C1 to C5. We purified an acidic polysaccharide from the cell wall of L. lactis 3107 (C2 subtype) and confirmed that it is structurally different from the CW PS of the C1 subtype L. lactis MG1363. Combinations of genes from the variable region of C2 subtype were amplified from L. lactis 3107 and introduced into a mutant of the C1 subtype L. lactis NZ9000 (a direct derivative of MG1363) deficient in CW PS biosynthesis. The resulting recombinant mutant synthesized a CW PS with a composition characteristic for that of the C2 subtype L. lactis 3107 and not the wildtype C1 L. lactis NZ9000. The recombinant mutant exhibited a changed phage resistance/sensitivity profile consistent with that of L. lactis 3107, which unambiguously demonstrated that L. lactis 3107 CW PS is the host cell surface receptor of two bacteriophages belonging to the P335 species as well as phages that are member of the 936 species. The research presented in this thesis has significantly advanced our understanding of L. lactis bacteriophage-host interactions in several ways. Firstly, the examination of plasmidencoded bacteriophage resistance systems has allowed inferences to be made regarding the mode of action of AbiB, thereby providing a platform for further elucidation of the molecular trigger of this system. Secondly, the phage infection transcriptome data presented, in addition to previous work, has made L. lactis a model organism in terms of transcriptomic studies of bacteriophage-host interactions. And finally, the research described in this thesis has for the first time explicitly revealed the nature of a carbohydrate bacteriophage receptor in L. lactis, while also providing a logical explanation for the observed narrow host ranges exhibited by 936 and P335 phages. Future research in discerning the structures of other L. lactis CW PS, combined with the determination of the molecular interplay between receptor binding proteins of these phages and CW PS will allow an in depth understanding of the mechanism by which the most prevalent lactococcal phages identify and adsorb to their specific host.

Preclinical atherosclerosis in rheumatoid arthritis

Introduction: There is accumulating evidence of an increased risk of cardiovascular morbidity and mortality in rheumatoid arthritis patients. A combination of both traditional cardiovascular risks and rheumatoid specific factors appear to be responsible for driving this phenomenon. Rheumatoid arthritis has been an orphan of cardiologists in the past and rheumatologists themselves are not good at CVD screening. Identifying the extent of preclinical atherosclerosis in RA patients will help us to appreciate the magnitude of this serious problem in an Irish population. Methods: We undertook a cross-sectional study of 63 RA patients and 48 OA controls and compared the 2 groups with respect to 1) traditional CV risks factors, 2) serum biomarkers of inflammation, including CRP, TNFα, IL6 and PAI-1, 3) carotid intima-media thickness (cIMT), carotid plaque and ankle-brachial index (ABI) as markers of pre-clinical atherosclerosis, 4) biochemical and ultrasonic measures of endothelial dysfunction and 5) serum and echocardiographic measures of diastolic dysfunction. Within the RA group, we also investigated for associations between markers of inflammation, subclinical atherosclerosis and diastolic dysfunction. Results: Prevalence of traditional CV risks was similar in the RA and OA groups. A number of biomarkers of inflammation were significantly higher in the RA group: CRP, fibrinogen, IL- 2, -4, -6, TNFα. PAI-1, a marker of thrombosis, correlated with disease activity and subclinical atherosclerosis in RA patients. With regard to subclinical atherosclerosis measures, RA patients had a significantly lower ABI than OA patients. Carotid plaque and cIMT readings were similar in RA and OA patients. Assessment of endothelial function revealed that RA patients had significantly higher concentrations of adhesion molecules, in particular sero-positive RA patients and RA smokers. Adhesion molecule concentrations were associated with markers of diastolic dysfunction in RA. Urine PCR, another marker of endothelial dysfunction also correlated with diastolic dysfunction in RA. Assessment of endothelial function with flow mediated dilatation (FMD) found no difference between the RA and OA groups. Disease activity scores in RA patients were associated with endothelial dysfunction, as assessed by FMD. Conclusions: We did not find significant differences in measures of subclinical atherosclerosis, flow mediated dilatation or diastolic function between RA and OA patients. This is most likely in part due to the fact that there is increasing evidence that OA has an inflammatory component to its pathogenesis and is associated with metabolic syndrome and increased CV risk. We reported a significant association between urinary PCR and measures of diastolic dysfunction. Urinary PCR may be a useful screening tool for diastolic dysfunction in RA. The association between RA disease activity and measures of vascular function supports the theory that the excess cardiovascular burden in RA is linked to uncontrolled inflammation.