Human infections attributable to the D-tartrate-fermenting variant of Salmonella enterica serovar Paratyphi B in Germany originate in reptiles and, on rare occasions, poultry

Abstract

In this study, the population structure, incidence, and potential sources of human infection caused by the d-tartrate-fermenting variant of Salmonella enterica serovar Paratyphi B [S. Paratyphi B (dT+)] was investigated. In Germany, the serovar is frequently isolated from broilers. Therefore, a selection of 108 epidemiologically unrelated S. enterica serovar Paratyphi B (dT+) strains isolated in Germany between 2002 and 2010 especially from humans, poultry/poultry meat, and reptiles was investigated by phenotypic and genotypic methods. Strains isolated from poultry and products thereof were strongly associated with multilocus sequence type ST28 and showed antimicrobial multiresistance profiles. Pulsed-field gel electrophoresis XbaI profiles were highly homogeneous, with only a few minor XbaI profile variants. All strains isolated from reptiles, except one, were strongly associated with ST88, another distantly related type. Most of the strains were susceptible to antimicrobial agents, and XbaI profiles were heterogeneous. Strains isolated from humans yielded seven sequence types (STs) clustering in three distantly related lineages. The first lineage, comprising five STs, represented mainly strains belonging to ST43 and ST149. The other two lineages were represented only by one ST each, ST28 and ST88. The relatedness of strains based on the pathogenicity gene repertoire (102 markers tested) was mostly in agreement with the multilocus sequence type. Because ST28 was frequently isolated from poultry but rarely in humans over the 9-year period investigated, overall, this study indicates that in Germany S. enterica serovar Paratyphi B (dT+) poses a health risk preferentially by contact with reptiles and, to a less extent, by exposure to poultry or poultry meat.

Fig 1

UPGMA dendrogram of PFGE profiles identified in 108 S. enterica serovar Paratyphi B (dT+) strains after digestion with XbaI. Profiles were numbered serially from 1 to 73. The number of strains belonging to each source (total, reptile, poultry, human, and other), corresponding MLSTs, and present (1) or absent (0) O:5 antigen are shown on the right side. Assigned clusters A to C are indicated by curly brackets.

Fig 2

Minimal spanning tree of MLST data on 108 S. enterica serovar Paratyphi B (dT+) isolates. Each circle refers to one ST subdivided into one pie slice per strain. STs that share six identical gene alleles are linked by a black line, and STs sharing only one or no common gene allele are linked by a gray dotted line. Based on their similarity, STs were grouped in three complexes (C1, C2, and C3). According to the nomenclature of Achtman et al. (1), C1 is designated eBG5, C2 is eBG19, and C3 is eBG59. Pathogenicity array types (PATs) found in each ST are shown below ST designations.

Fig 3

UPGMA dendrograms of 35 S. enterica serovar Paratyphi B (dT+) strains based on the merged DNA sequences. On the left side, the MLST dendrogram with corresponding sequence types (STs) is shown, and on the right side the sop dendrogram with corresponding sop-ST pattern types (sopA-sopB-sopD) is shown. The scale indicates the similarity of merged DNA sequences.

Fig 4

Virulence determinants of 35 S. enterica serovar Paratyphi B (dT+) strains analyzed. On the left side the analyzed genes are indicated and grouped according to their particular genomic location (SPI-1 to SPI-7; prophages Gifsy-1, Gifsy-2, Gifsy-3 and Fels-1; plasmids; and islets) or function (fimbrial). At the top, assigned PATs and corresponding STs are indicated. Immediately below this, an UPGMA dendrogram shows the similarity of PATs in percentages. The hybridization result of each PAT is shown by column. A white box indicates the absence, and a gray box indicates the presence of the target sequence.