PATHOGENIC ESCHERICHIA-COLI STRAINS FROM DOGS AND CATS - I) DETECTIONOF ENTEROTOXIGENIC (ETEC), ENTEROPATHOGENIC (EPEC), VEROTOXIGENIC (VTEC), ENTEROHEMORRHAGIC (EHEC) AND NECROTOXIGENIC (NTEC) STRAINS

Seven hundred and ninety-eight Escherichia coli isolates from dogs and 113 from cats of fecal, intestinal, urinary or blood origin were stud ied by DNA colony hybridization with 23 gene probes for specific virul ence factors and/or properties of enterotoxigenic (ETEC), enteropathog enic (EPEC), verotoxigenic (VTEC), enterohaemorrhagic (EHEC), necrotox igenic (NTEC), diffusely adherent (DAEC), and enteroaggregative (EAggE C) strains. Three hundred and ten canine (39%) and 69 feline (61%) iso lates tested positive with at least one gene probe. Genes homologous t o those coding for the LT2a enterotoxin, for the SLT1 verocytotoxin an d for the F6, AF/R1 and Agg adhesins were not detected. Isolates posit ive with the probes were grouped in ETEC, EPEC, VTEC, EHEC, NTEC, and other strains, which did not fit in any of the first five groups. The most represented groups were the NTEC and the ((other strains,, groups , ''other strains'' which were positive with probes specific for adhes ins only: 102 isolates, i.e. 13%, and 134 isolates, i.e. 17%, respecti vely for canine isolates; 33 isolates, i.e. 29%, and 27 isolates, i.e. 24%, respectively for feline isolates. All NTEC isolates, but three c anine (98%) had the NTEC1 hybridization profile, i.e. they are potenti al producers of the Cytotoxic Necrotizing Factor type 1 (CNF1). Most N TEC1 isolates were also positive with gene probes for fimbrial adhesin s of P family (26 isolates, i.e. 20% of NTEC1 isolates), of S family ( 11 isolates, i.e. 8% of NTEC1 isolates), and for both families (86 iso lates, i.e. 65% of NTEC1 isolates). The group of isolates positive wit h probes for adhesins could be further subdivided: ETEC adhesins (one canine isolate, i.e. <1%), EPEC adhesins (17 canine isolates, i.e. 2%, and four feline isolates, i.e. 3%), VTEC adhesins (two canine isolate s, i.e. <1%), and NTEC adhesins (112 canine isolates, i.e. 14%, and 23 feline isolates, i.e. 21%). EPEC isolates were present in the same pr oportions among canine and feline isolates: 44 canine isolates, i.e. 5 .5%, and six feline isolates, i.e. 5%. Two pathotypes were predominant : EaeA+ (34 canine and two feline isolates; i.e. 72% of EPEC isolates) and EaeA+BFP+ (10 canine and one feline isolate; i.e. 22% of EPEC iso lates). On the opposite VTEC isolates were very two canine EHEC (EaeASLT2+) and two feline VTEC (SLT2+) isolates. Canine ETEC isolates were present (28 isolates, i.e. 3.5%) but feline ETEC isolates were almost absent: one isolate (<1%). The canine ETEC isolates belonged to three main pathotypes: STa+ (four isolates, i.e. 14% of ETEC isolates), STa +STb+ (17 isolates, i.e. 61% of ETEC isolates), and STb+ (five isolate s, i.e. 18 % of ETEC isolates). Sixty percent of canine and 40% of fel ine isolates were thus negative to the 23 probes used. They may be iso lates belonging to other uninvestigated or still unknown pathotypes or non pathogenic E. coli. The problem of the animal species specificity must also be questionned, as several isolates belonged to pathotypes regularly detected in human beings or in other animal species.