In a recent paper, Huehn et al. (6) examined a large number of poultry-associated Salmonella enterica subspecies enterica type I serovar 4,12:d:− strains which predominate in poultry flocks in Germany but are poorly represented in cases of human disease. The authors examined the clonality and pathogenic gene repertoire of these strains to shed light on this contradiction. They concluded that the epidemiology of this highly clonal strain may be due to a lack of several genes which contribute to human pathogenicity and the presence of “genetic factors” which facilitate broiler colonization. The data presented in this study, together with a number of other published reports from around the world which were not cited by these authors, may provide further insight into this important issue.

In Australia, subspecies type II Salmonella enterica subsp. salamae serovar Sofia is also highly clonal and has similar antigenic determinants (4,12:b−) to those described by Huehn et al. and has long been established as the predominant Salmonella serovar isolated from poultry (8). Furthermore, S. Sofia, like the German Salmonella 4,12:d− strain, is regarded as having relatively low virulence for humans as it is rarely associated with human disease in comparison with subspecies type I serovars such as Salmonella enterica serovar Typhimurium (5). The natural habitat of Salmonella subspecies type II is generally believed to be coldblooded animals and the environment (7), and it is therefore unusual to encounter the bacteria in poultry. Furthermore, in both Senegal and Demark, Salmonella enterica serovar Brancaster, which also has a similar antigenic structure (4,12:b−) to the strains found in Germany, has also been found to commonly occur in poultry. This serotype had an apparently low virulence for humans and displayed a relatively high degree of clonality (3, 4).

Salmonella strains possessing fimbrial markers, such as long polar fimbriae (Lpf) and plasmid-encoded fimbriae (Pef), are often associated with human diseases (1, 2). It has been established that plasmid-encoded phase 2 flagellar regions, which contain the genes lpf and pef, are absent in poultry-associated S. Sofia isolates (7). This correlates with the finding that these regions are absent in the German Salmonella 4,12:d− serovar strains. The complete loss of, or inability to harbor, plasmids in many of the Salmonella 4,12d:− strains (6) and in S. Brancaster (4) could be the main reason for the absence of lpf and pef in these strains. As pointed out by Huehn et al., the lack of fimbrial clusters such as Lpf might be an important reason for the successful spread of the serovar in poultry.

These data taken together add weight to the suggestion of Huehn et al. (6) that the inability to produce a phase 2 flagellar antigen has led to the successful colonization of these serovars in poultry. The presence of the 4,12 somatic antigen in serovars with similar traits from diverse continents, however, introduces a new dimension to understanding these issues. We hypothesize that the 4,12 antigen, in combination with the absence of phase 2 flagellar antigens, strongly favors the persistence of these serovars in poultry. Further work in this area, perhaps taking an international comparative approach, is required to fully test this hypothesis. The dominance of Salmonella serovars with a low human virulence potential and a high prevalence in poultry could result in significant public health benefits.

• Copyright © 2009 American Society for Microbiology