The purpose of the study was to compare the sensitivity of PCR with those of cell culture reisolation of Ehrlichia canis, the indirect fluorescent antibody test (IFA), and Western immunoblotting (WI) in the early diagnosis of canine ehrlichiosis. Five German shepherd dogs were intravenously inoculated with 107 E. canis-infected DH82 cells. Blood was collected on alternate days during a 2-week postinoculation period. Mononuclear cell fractions were harvested and used for E. canis reisolation and DNA extraction for PCR. The plasma was used for assaying antibodies against E. canis. By PCR, the 16S rRNA gene of E. canis was detected in the mononuclear cell specimens collected as early as day 4 to 10 postexposure (PE). E. canis was reisolated from the blood starting on day 2 PE from all five dogs. The indirect fluorescent antibody test and Western immunoblotting could detect E. canis antibodies as early as 2 to 8 days PE. Cell culture reisolation proved to be the most sensitive and definitive for early diagnosis of ehrlichiosis, but it is not very convenient, since it takes a long time (14 to 34 days) to show up positive. The sensitivity of PCR is comparable to or slightly less than that of other established methods; however, the convenience, quickness, and direct nature of detecting E. canis DNA is expected to make PCR more useful for clinical diagnosis.

Ehrlichia risticii is the causative agent of Potomac horse fever. Variations among the major antigens of different local E. risticii strains have been detected previously. To further assess genetic variability in this species or species complex, the sequences of the 16s rRNA genes of several isolates obtained from sick horses diagnosed as having Potomac horse fever were determined. The sequences of six isolates obtained from Ohio and three isolates obtained from Kentucky were amplified by PCR. Three groups of sequences were identified. The sequences of five of the Ohio isolates were identical to the sequence of the type strain of E. risticii, the Illinois strain. The sequence of one Ohio isolate, isolate 081, was unique; this sequence differed in 10 nucleotides from the sequence of the type strain (level of similarity, 99.3%). The sequences of the three Kentucky isolates were identical to each other, but differed by five bases from the sequence of the type strain (level of similarity, 99.6%). The levels of sequence similarity of isolate 081, the Kentucky isolates, and the type strain to the next most closely related Ehrlichia sp., Ehrlichia sennetsu, were 99.3, 99.2, and 99.2%, respectively. On the basis of the distinct antigenic profiles and the levels of 16s rRNA sequence divergence, isolate 081 is as divergent from the type strain of E. risticii as E. sennetsu is. Therefore, we suggest that strain 081 and the Kentucky isolates may represent two new distinct Ehrlichia species.Ehrlichia risticii is a small, gram-negative, obligately intracellular bacterium that causes Potomac horse fever, an acute systemic infectious disease of the family Equidae (11, 15). Since it was first recognized in 1979 along the Potomac River in Maryland and Virginia (7), Potomac horse fever has been shown to occur throughout North America (10, l l ) , in France (19), and possibly in India (15). E. risticii is antigenically and genetically most closely related to Ehrlichia sennetsu, the etiologic agent of human Sennetsu fever in Japan and Malaysia (1,11,12,14,16). E. sennetsu can establish infections in horses but is not pathogenic, and preinoculation of E. sennetsu can protect horses from E. risticii challenge (14).During clinical diagnostic work in our laboratory, many strains of E. risticii have been isolated from horses which have had clinical signs compatible with Potomac horse fever, and these isolates have been identified on the basis of morphological and serological criteria. The antigenic and morphological characteristics of six isolates obtained from infected horses residing in Ohio and three isolates obtained from horses residing in Kentucky have been studied previously (3). Antigenic diversity among these strains was detected by Western blot (immunoblot) analysis and indirect fluorescent-antibody (IFA) tests in which we used monoclonal antibodies to E. risticii Illinois or Maryland strains which were isolated from two different infected horses residing in Maryland in 1984. In particular, isolate 081, which was obtained from ...

Ehrlichia risticii causes an acute infectious disease in horses called Potomac horse fever. To investigate the biological diversity ofE. risticii organisms, nine E. risticii isolates derived from the peripheral blood monocytes of clinically sick horses in Ohio and Kentucky during the summers of 1991 and 1993 were compared with Illinois and Virginia isolates originally obtained from horses in Maryland in 1984. Seven of the nine isolates (081, 606, 380, 679, As, Co, and Ov) formed large morulae (tightly packed inclusions of ehrlichial organisms). The remaining isolates, including 1984 isolates, were individually dispersed or formed small morulae in the cytoplasm of P388D1 cells. In Western blot (immunoblot) analysis with four equine and one rabbit polyclonal anti-E. risticii sera, these recent E. risticii isolates showed patterns of antigenic proteins distinct from those of the 1984 isolates and could be divided into three groups: (i) 081; (ii) 606, 022, 067, 380, and 679; and (iii) As, Co, and Ov. By indirect fluorescent antibody labeling with two panels of murine anti-E. risticii (Illinois and Maryland isolates) monoclonal antibodies, isolate 081 was not labeled with any of 20 monoclonal antibodies tested. The remaining isolates were not labeled with several monoclonal antibodies. The digestion pattern with one of the restriction enzymes, Avall, of the PCR-amplified partial 16S rRNA gene of E. risticii from all Kentucky isolates (As, Co, and Ov) was different from that of Illinois, Virginia, and six Ohio isolates. These results indicate the presence of distinct variants of E. risticii which vary significantly in morphology, antigenic composition, and the base sequence of the 16S rRNA gene.