|
The second session with presentations of General Interest, on Saturday 6 June, started with a contribution from Sweden, presented by Eva Litrup (SSI, also on behalf of BfR contributors). A microarray hybridization analysis of various Salmonella serotypes had been carried out using a chip containing 282 probes, most of which were indicative of either pathogenicity or antimicrobial resistance. The same strains that had been hybridized were also genotyped by MLST, and from both sets of data a dendrogram was created to visualize the relatedness of the isolates. Most tested isolates clustered according to their serotype with both analyses, but Salmonella paratyphi B isolates split apart in two clusters. Since the two trees were not identical, one could ask ‘which tree is correct’ but of course both trees reflect the similarity of these isolates correctly, but the data on which similarity is calculated differs (7 MLST genes on the one hand, where every SNP counts, vs. >100 genes that have to be only generally conserved in order to hybridize). Either analysis did not separate blood isolates (from bacteraemic patients) from faecal samples, which was maybe what was hoped for when the study was set up. However, once again this demonstrates that the clinical outcome of infection is multifactorial, to which the gene content of the pathogen is only one contributor of many. The second presentation, by Jonas Larsson from SSI in collaboration with partners from DTU and the Norwegian Institute of Public Health in Oslo, dealt specifically with Salmonella Typhimurium. This group had thought about a rather Shakespearian theme: how to name the various genotypes that are recognizable, within this serotype, by the popular typing method multi-locus tandum repeat analysis (MLVA). “What’s in a name” is not at all a trivial question, and a nomenclature of genotypes requires careful and systematic consideration. The work presented here was a major step forward in this direction, though it remains to be seen whether their good and sound suggestions will be taken up by the scientific community. During the few minutes of discussion there was a lot of feedback from the audience, which can be taken as a good sign. Next, work from the UK conducted by partners from VLA in collaboration with the University of Bristol was presented by Anne Ridley. A longitudinal study had been carried out on a local conventional broiler farm, in order to follow the different types of Campylobacter that come and go in such a dynamic environment. It may come as a surprise that the prominent animal featuring on one of the slides was a cow: since the study included sampling the environment and any animals living at close proximity, the calves housed next-door were also tested, and were found to be persistently positive for Campylobacter. The strain found in these calves did not show up in the birds, so the researchers hypothesized that these strains were maybe adapted to the bovine host. However, when this hypothesis was put to the test, by experimentally infecting chickens with the strain, colonization was the result. Dismissing a hypothesis is also scientific progress, and the researchers were quite right to present these data. Finally, Montserrat Gutierrez presented, also on behalf of her collaborator from CVRL, Ireland, an extensive model for enhancing food safety surveillance. A very well presented story summarized the meticulous work that had been done to get the industry on board in a collaborative effort to improve surveillance. Their model showed that this can be done, and that industrial partners can overcome their constraints to share the vast amount of microbial testing data that are generated in‑house on a regular basis. Though some skepticism was expressed during the discussion, the Irish have shown how to tackle this sensitive problem, and by doing so they set an example for the rest of Europe. Chair: Trudy Wassenaar (MMGC)
|