We screened 31,615 S. enterica serovar Enteritidis isolates that had been collected from clinical specimens during 2005–2010 and deposited in the culture collection of the Health Protection Agency Salmonella Reference Unit. We screened the isolates for R-type ACSSuTTm. A total of 14 serovar Enteritidis isolates showing this resistance phenotype were detected and subsequently examined for the presence of integron-located dfrA7. Of the 14 isolates, 11 were positive and their plasmid content was analyzed by S1 pulsed-field gel electrophoresis (PFGE) (2) and by the Kado and Liu methods (7); we used serovar Enteritidis strains NRL-Salm-PT4 and CNM4839/03 as controls for pSEV– and pUO-SeVR1–carrying isolates, respectively. The 11 isolates harbored 1 plasmid of variable size (60–95 kb); among these, 9 isolates hybridized with dfrA7-specific and spvC-specific probes (with plasmids of 85–95 kb) (Figure). These 9 isolates contained a VR-hybrid plasmid similar to pUO-SeVR1 and were selected for further analyses (Table 1, Table 2). The remaining 2 isolates carried the normal pSEV plasmid (60 kb), in which spvC hybridized; dfrA7 was chromosomally located.

In the 9 isolates carrying VR-hybrid plasmids, the R-type ACSSuTTm was encoded by the R-genes bla_TEM-1, catA2, strA-strB, sul1, sul2, tet(A), and dfrA7, which were located on the pUO-SeVR1–like plasmids as determined by Southern blot hybridization. By PCR amplification, using previously described primers and conditions (5,8) (Table 2), and by Southern blot hybridization (5), we tested for the presence of IncFII and IncFIB replicons, parAB (partition), spvRABCD, rck (resistance to complement killing), mig-5 (macrophage-induced gene), pefABCDI (Pef fimbriae operon), and srgA (SdiA-regulated gene; next to orf7), all carried by pSEV. The 9 plasmids were positive for the 2 replicons and for all genes screened except pefC and pefD (absent in H070360201 and H070420137), pefI-orf7 (absent in all), and srgA (either absent [H051860415, H073180204, and H091800482] or truncated [in the remaining isolates]) (Table 2).

Isolate subtyping was conducted by phage typing, multilocus variable number tandem repeat analysis (MLVA), multilocus sequence typing (MLST), and XbaI-PFGE (9,10) (http://mlst.ucc.ie/mlst/dbs/Senterica; www.pulsenetinternational.org). The 9 S. enterica serovar Enteritidis isolates belong to phage type (PT) 42, in contrast with CNM4839/03, which belongs to PT14b (Table 2). We identified 4 MLVA profiles, which were all single-locus variants of the highly variable locus SENTR5, indicating that the isolates are closely related (Table 2). The isolate from Spain and 3 of the 9 isolates from the United Kingdom, selected as representative of each MLVA profile, were also analyzed by MLST (Table 2). CNM4839/03 and H091340084 were assigned to sequence type (ST) 11, the most commonly found ST in serovar Enteritidis (http://mlst.ucc.ie/mlst/). The remaining 2 isolates could be ascribed to ST1479, the first examples of this single-locus variant of ST11 in the MLST database (Table 2). According to XbaI-PFGE, the control strain NRL-Salm-PT4 showed a clearly distinct profile in comparison with CNM4839/03 and the 9 isolates containing pUO-SeVR1–like plasmids, which generated 6 closely related patterns (Figure). Most isolates differed by 1 band of variable size (85–95 kb), which corresponded to the XbaI-linearized VR-hybrid plasmids. As an exception, isolate H070420137 showed additional differences in chromosomal bands of ≈150 and ≈300 kb. Because isolates H070420137 and H070360201 came from the same patient (Table 1) and shared the same V and R genotypes and other typing markers, 1 isolate could have evolved from the other; however, co-infection of the patient with 2 closely related strains cannot be ruled out. In addition, considering the typing results of H095100307, H100240198, and H101700366, the identical size of their plasmids, and the fact that they were isolated from the same patient (Table 1), these 3 isolates can be considered the same strain.

All except 2 of the UK isolates carrying pUO-SeVR1–like plasmids were recovered from the blood of patients who had recently returned from an African country or who had an African name (Table 1). Supporting the possible African origin, similar XbaI-PFGE profiles and MDR phenotypes (ampicillin, trimethoprim, sulfonamides, and tetracycline) have been identified in clinical S. enterica serovar Enteritidis isolates involved in outbreaks and community infections in different African countries. These isolates caused bacteremia, meningitis, diarrhea (11–13), and high case-fatality rates; they affected mainly children, whereas most clinical isolates analyzed in our study were obtained from adults with bacteremia (Table 1). The detection of the 700-bp/dfrA7 integron in S. enterica serovar Enteritidis isolates from Africa (14) also supports an African origin of the MDR serovar Enteritidis isolates harboring pUO-SeVR1–like plasmids. Of note, resistance derivatives of pSLT, the V-plasmid specific to S. Typhimurium, have been found in the epidemic ST313 clone of this serovar, which has been considered a major cause of invasive disease in sub-Saharan Africa (15).

Closely related MDR S. enterica serovar Enteritidis isolates carrying pUO-SeVR1–like plasmids were recovered in the United Kingdom. Most were isolated from the blood of patients linked to Africa, and they showed common features with serovar Enteritidis isolates involved in outbreaks on that continent. The possibility that this potentially invasive clone of S. enterica serovar Enteritidis can be spread through human travel, together with the detection of VR-plasmids in the serovar most frequently associated with human infections, is of public health concern and requires surveillance.