We have analyzed the prevalence of Shiga toxin-producing Escherichia coli (STEC) in stool specimens of patients with diarrhea or other gastrointestinal alterations from the Xeral-Calde Hospital of Lugo City (Spain). STEC strains were detected in 126 (2.5%) of 5,054 cases investigated, with a progressive increase in the incidence from 0% in 1992 to 4.4% in 1999. STEC O157:H7 was isolated in 24 cases (0.5%), whereas non-O157 STEC strains were isolated from 87 patients (1.7%). STEC strains were (after Salmonella and Campylobacter strains) the third most frequently recovered enteropathogenic bacteria. A total of 126 human STEC isolates were characterized in this study. PCR showed that 43 (34%) isolates carried stx 1 genes, 45 (36%) possessed stx 2 genes and 38 (30%) carried both stx 1 and stx 2 . A total of 88 (70%) isolates carried an ehxA enterohemolysin gene, and 70 (56%) isolates possessed an eae intimin gene (27 isolates with type ␥1, 20 with type ␤1, 8 with type , 5 with type ␥2, and 3 with type ). STEC isolates belonged to 41 O serogroups and 66 O:H serotypes, including 21 serotypes associated with hemolytic uremic syndrome and 30 new serotypes not previously reported among human STEC strains in other studies. Although the 126 STEC isolates belonged to 81 different seropathotypes (associations between serotypes and virulence genes), only four accounted for 31% of isolates. Seropathotype O157:H7 stx 1 stx 2 eae-␥1 ehxA was the most common (13 isolates) followed by O157:H7 stx 2 eae-␥1 ehxA (11 isolates), O26:H11 stx 1 eae-␤1 ehxA (11 isolates), and O111:H-stx 1 stx 2 eae-␥2 ehxA (4 isolates). Our results suggest that STEC strains are a significant cause of human infections in Spain and confirm that in continental Europe, infections caused by STEC non-O157 strains are more common than those caused by O157:H7 isolates. The high prevalence of STEC strains (both O157:H7 and non-O157 strains) in human patients, and their association with serious complications, strongly supports the utilization of protocols for detection of all serotypes of STEC in Spanish clinical microbiology laboratories.

In Spain, as in many other countries, verotoxin-producing Escherichia coli (VTEC) strains have been frequently isolated from cattle, sheep, and foods. VTEC strains have caused seven outbreaks in Spain (six caused by E. coli O157:H7 and one by E. coli O111:H- [nonmotile]) in recent years. An analysis of the serotypes indicated serological diversity. Among the strains isolated from humans, serotypes O26:H11, O111:H-, and O157:H7 were found to be more prevalent. The most frequently detected serotypes in cattle were O20:H19, O22:H8, O26:H11, O77:H41, O105:H18, O113:H21, O157:H7, O171:H2, and OUT (O untypeable):H19. Different VTEC serotypes (e.g., O5:H-, O6:H10, O91:H-, O117:H-, O128:H-, O128:H2, O146:H8, O146:H21, O156:H-, and OUT:H21) were found more frequently in sheep. These observations suggest a host serotype specificity for some VTEC. Numerous bovine and ovine VTEC serotypes detected in Spain were associated with human illnesses, confirming that ruminants are important reservoirs of pathogenic VTEC. VTEC can produce one or two toxins (VT1 and VT2) that cause human illnesses. These toxins are different proteins encoded by different genes. Another virulence factor expressed by VTEC is the protein intimin that is responsible for intimate attachment of VTEC and effacing lesions in the intestinal mucosa. This virulence factor is encoded by the chromosomal gene eae. The eae gene was found at a much less frequency in bovine (17%) and ovine (5%) than in human (45%) non-O157 VTEC strains. This may support the evidence that the eae gene contributes significantly to the virulence of human VTEC strains and that many animal non-O157 VTEC strains are less pathogenic to humans.

CTX-M-type ESBLs, primarily CTX-M-14 and CTX-M-15, have emerged as the predominant types of ESBL produced by E. coli isolates in Lugo. In view of the reported findings, long-term care facilities for elderly people may represent a significant reservoir for E. coli clone O25b:H4-ST131 producing CTX-M-15. The triplex PCR developed in this work will be useful for rapid and simple detection of this clone.

To discern the possible spread of the Escherichia coli O25b:H4-ST131 clonal group in poultry and the zoonotic potential of avian strains, we made a retrospective search of our strain collection and compared the findings for those strains with the findings for current strains. Thus, we have characterized a collection of 19 avian O25b:H4-ST131 E. coli strains isolated from 1995 to 2010 which, interestingly, harbored the ibeA gene. Using this virulence gene as a criterion for selection, we compared those 19 avian strains with 33 human O25b:H4-ST131 ibeA-positive E. coli strains obtained from patients with extraintestinal infections (1993 to 2009). All 52 O25b:H4-ST131 ibeA-positive E. coli strains shared the fimH, kpsMII, malX, and usp genes but showed statistically significant differences in nine virulence factors, namely, papGIII, cdtB, sat, and kpsMII K5, which were associated with human strains, and iroN, kpsMII K1, cvaC, iss, and tsh, which were associated with strains of avian origin. The XbaI macrorestriction profiles of the 52 E. coli O25b:H4-ST131 ibeA-positive strains revealed 11 clusters (clusters I to XI) of >85% similarity, with four clusters including strains of human and avian origin. Cluster VII (90.9% similarity) grouped 10 strains (7 avian and 3 human strains) that mostly produced CTX-M-9 and that also shared the same virulence profile. Finally, we compared the macrorestriction profiles of the 12 CTX-M-9-producing O25b:H4-ST131 ibeA strains (7 avian and 5 human strains) identified among the 52 strains with those of 15 human O25b:H4-ST131 CTX-M-14-, CTX-M-15-, and CTX-M-32-producing strains that proved to be negative for ibeA and showed that they clearly differed in the level of similarity from the CTX-M-9-producing strains. In conclusion, E. coli clonal group O25b:H4-ST131 ibeA has recently emerged among avian isolates with the new acquisition of the K1 capsule antigen and includes CTX-M-9-producing strains. This clonal group represents a real zoonotic risk that has crossed the barrier between human and avian hosts.

The aim of this study was to determine the clinical significance of Streptococcus salivarius isolates recovered from blood cultures and compare them with isolates of Streptococcus bovis biotypes I and II. Seventeen of the 52 (32%) S. salivarius isolates recovered were considered clinically significant, compared with 62 of the 64 (97%) S. bovis isolates (p<0.0001). Bacteremia caused by S. salivarius occurred mostly in patients who showed relevant disruption of the mucous membranes and/or serious underlying diseases. Patients with S. salivarius bacteremia were younger than those with S. bovis bacteremia (57 vs. 67 years; p<0.01). Patients with S. salivarius bacteremia and patients with S. bovis II bacteremia had similar rates of endocarditis, colon tumors, and non-colon cancer. On the other hand, when compared with S. bovis I bacteremia, S. salivarius bacteremia was associated with lower rates of endocarditis (18% vs. 74%, respectively) (p<0.01) and colon tumors (0% vs. 57%, respectively) (p<0.005) and higher rates of non-colon cancer (53% vs. 9.5%, respectively) (p<0.01). Bacteremia caused by S. bovis II had a hepatobiliary origin in 50% of the patients, while, in contrast, that due to S. salivarius or S. bovis I was less frequently associated with a hepatobiliary origin (12% and 5%, respectively) (p<0.00001). The rate of penicillin resistance was 31% among S. salivarius isolates and 0% among S. bovis isolates (p<0.0001). In conclusion, the clinical characteristics of S. salivarius bacteremia and S. bovis II bacteremia are similar, and the isolation of S. salivarius in blood should not be systematically regarded as contamination.