Canine granulocytic ehrlichiosis was diagnosed in 37 dogs by finding ehrlichial morulae in 0.1 to 26.2% of their blood neutrophils and eosinophils. All 37 dogs had clinical signs of arthritis or muscular stiffness. Titer to Ehrlichia canis was determined in sera from 31 of the 37 dogs; 25 dogs had titer ranging from 1:20 to 1:5,120. In the other 6 dogs, titer to E canis was < 1:10. The most common hematologic abnormality in these dogs, other than rickettsiemia, was thrombocytopenia. Granulocytes infected with ehrlichial organisms were not found in another 10 dogs that had clinical signs of arthritis or muscular stiffness. Of these 10 dogs, 3 had titer to E canis ranging from 1:40 to 1:320. Titer in the other 7 dogs was < 1:10. Ehrlichial morulae were not found in the granulocytes of 18 healthy dogs. Of these 18 dogs, 9 had titer to E canis ranging from 1:20 to 1:5,120. Titer in the other 9 dogs was < 1:10. Titer to Borrelia burgdorferi was determined in dogs with granulocytic ehrlichiosis, arthritic dogs without detected rickettsiemia, and in healthy dogs. Low titer determined by 2 laboratories was considered to be nonspecific reaction in all 3 groups of dogs and, thus, did not indicate that the arthritic disorders were attributable to canine borreliosis.

Sixteen healthy ponies were inoculated IV with Ehrlichia risticii-infected P388D1 mouse monocytes. Of the 16 ponies, 15 developed clinical signs of equine ehrlichial colitis. Twenty-four hours after onset of fever (rectal temperature > 38.8 degrees C), 7 ponies were treated with 25 mg of erythromycin stearate/kg of body weight and 10 mg of rifampin/kg, given orally every 12 hours for 5 days. The remaining 8 ill ponies served as nontreated controls. All ponies were observed for progression of clinical signs typical of equine ehrlichial colitis. Within 12 hours of initiation of treatment, 4 of the 7 treated ponies had rectal temperature < 38.4 C and, within 24 hours, 6 of the 7 ponies had rectal temperature < 38.3C. In contrast, all control ponies had rectal temperature > 39.2 C at 24 hours (P < 0.05). Of the 7 treated ponies, 4 no longer had signs of mental depression after the second day of treatment, and only 1 of the 7 ponies had mild signs of depression after the third day of treatment. In contrast, control ponies had high mental depression score during the observation period (P < 0.05). Feed intake improved in ponies of the treatment group, with feed intake of 4 of the 7 ponies returning to normal; the other 3 ponies were only mildly anorectic by the second day of treatment. Control ponies progressively decreased their feed intake during the observation period (P < 0.05). One control pony and 2 treated ponies developed diarrhea before the treatment/observation period began. Only 1 treated pony developed diarrhea after treatment began. Of the 8 control ponies, 7 developed diarrhea. Profound decrease in borborygmal sounds with silent periods lasting longer than 3 minutes was observed in 7 of the 8 control ponies. Only 1 of the 7 treated ponies had such profound decrease in borborygmi (P < 0.05). The decrease in borborygmal sounds progressed in the control ponies during the observation period. None of the treated ponies continued to have decreased borborygmi after treatment day 2 (P < 0.05). Of the 8 control ponies, 2 were euthanatized; all treated ponies survived. In survivors, signs lasted 8 to 17 (mean, 10) days in control ponies but only 1 to 5 (mean, 2.9) days in treated ponies.

Urinary protein-to-creatinine ratios and serum albumin concentrations were measured in 8 adult male dogs experimentally inoculated with Ehrlichia canis. Urinary protein concentration increased significantly, but transiently, during the acute phase of infection. Urinary protein-to-creatinine ratios were highest (mean, 8.6) during the third and fourth weeks after infection, and decreased to < 0.5 by 6 weeks after infection. Correspondingly, albumin concentration decreased significantly during the acute phase. Serum albumin concentrations were lowest (mean, 2.1 g/dl) the fourth week after infection and increased to > 3.0 g/dl by 11 weeks after infection. There was an inverse linear correlation between urinary protein-to-creatinine ratio and serum albumin concentration. The magnitude of proteinuria and its inverse relationship with serum albumin concentration suggested that hypoalbuminemia associated with acute E canis infection may be attributable primarily to increased renal loss of protein, rather than decreased hepatic synthesis as previously suggested. Another dog was subsequently inoculated with E canis from 1 of the experimentally infected dogs and a renal biopsy was performed during peak proteinuria (urinary protein-to-creatinine ratio = 22 and serum albumin = 1.1 g/dl). Immunofluorescent staining revealed mild to moderate deposits of anti-canine IgM, and to a lesser extent, anti-canine IgG and complement factor C3 in the glomerular tufts and mesangium. Ultrastructural evaluation revealed distortion and fusion of podocyte foot processes and increased microvilli on podocytes. These morphologic changes were consistent with transient glomerular leakage of protein of a magnitude that would significantly contribute to hypoalbuminemia during acute E canis infection. An underlying immunologic mechanism was suggested by positive glomerular immunofluorescence and previously described histologic findings.

Eight adult cats were inoculated IV (n = 6) or SC (n = 2) with Ehrlichia risticii-infected P388Dl (continuous murine macrophage) cells or with E risticii released from P388D1 cells. Three additional cats were inoculated with organism-free P388D1 cultured monocytes, and 1 cat, which served as a medium control, was inoculated with balanced salt solution. Clinical signs of illness were observed in the IV inoculated cats from which E risticii was isolated. One cat developed intermittent diarrhea between postinoculation days (PID) 8 and 18, and the other cat developed lymphadenopathy, acute depression, and anorexia between PID 20 and 24. Ehrlichia risticii was isolated in cultures from 2 of 6 IV inoculated cats on PID 6, 10, and 17. Both cats were inoculated with E risticii released from the P388D1 cells. Ehrlichia risticii was not isolated from SC inoculated cats or from control cats. All 8 cats inoculated with E risticii seroconverted between PID 10 and 23. A pony inoculated with E risticii isolated from 1 of the inoculated cats developed clinical signs of equine monocytic ehrlichiosis including fever, anorexia, depression, and mild colic. Ehrlichia risticii was isolated from the blood of this pony on PID 7, 9, 11, and 16.

Ehrlichia chaffeensis, the newly recognized agent of human ehrlichiosis, is closely related to E canis, the causative agent of canine ehrlichiosis. Eight pups were inoculated IV with E chaffeensis-, or with E canis-infected DH82 cells, or organisms released from these host cells. Two additional pups served as nonexposed controls. Marked thrombocytopenia was observed in the E canis-infected pups, but not in those infected with E chaffeensis. Homologous serologic response was observed in the E chaffeensis-exposed pups by postinoculation day (PID) 14 and in the E canis-exposed pups by PID 21. Ehrlichia chaffeensis and E canis were reisolated from the respective inoculated pups on each of 8 attempts from PID 7 to 26. One E chaffeensis-exposed pup that was challenge exposed with E canis via blood transfusion, developed fever, anorexia, and thrombocytopenia, suggesting lack of cross protection against E canis.

Canine granulocytic ehrlichiosis was diagnosed in 37 dogs by finding ehrlichial morulae in 0.1 to 26.2% of their blood neutrophils and eosinophils. All 37 dogs had clinical signs of arthritis or muscular stiffness. Titer to Ehrlichia canis was determined in sera from 31 of the 37 dogs; 25 dogs had titer ranging from 1:20 to 1:5,120. In the other 6 dogs, titer to E canis was < 1:10. The most common hematologic abnormality in these dogs, other than rickettsiemia, was thrombocytopenia. Granulocytes infected with ehrlichial organisms were not found in another 10 dogs that had clinical signs of arthritis or muscular stiffness. Of these 10 dogs, 3 had titer to E canis ranging from 1:40 to 1:320. Titer in the other 7 dogs was < 1:10. Ehrlichial morulae were not found in the granulocytes of 18 healthy dogs. Of these 18 dogs, 9 had titer to E canis ranging from 1:20 to 1:5,120. Titer in the other 9 dogs was < 1:10. Titer to Borrelia burgdorferi was determined in dogs with granulocytic ehrlichiosis, arthritic dogs without detected rickettsiemia, and in healthy dogs. Low titer determined by 2 laboratories was considered to be nonspecific reaction in all 3 groups of dogs and, thus, did not indicate that the arthritic disorders were attributable to canine borreliosis.

The antigenic profile of Ehrlichia canis, E risticii, E sennetsu, and E equi was investigated by the use of protein (western) immunoblot technique. Results of analysis of serum from acutely and chronically infected animals indicated that the 4 Ehrlichia species share a unique 25-kD polypeptide in addition to other peptides. Immune sera from dogs inoculated with E canis recognized a wide range of E canis polypeptide antigens, as determined by western blot analysis. A larger number of E sennetsu polypeptides were detected when homologous antiserum and antiserum to E equi were used. The latter antiserum did not recognize antigens of E canis or E risticii. Antisera to E canis, E risticii, and E sennetsu detected E equi antigens. Data indicate that a 25-kD protein is a common antigen among the species of the genus Ehrlichia and that the ascending order of abundance of immunodominant determinants in the 4 species of Ehrlichia studied would be: E risticii leads to E equi leads to E sennetsu leads to E canis. Implications of these findings for diagnosis of ehrlichial infections and prophylaxis are evident.

Fetal infectivity of Ehrlichia risticii was investigated in 19 ponies that were E risticii negative on the basis of results of an indirect fluorescent antibody (IFA) test. Thirteen pregnant ponies were infected by IV administration of E risticii between 90 and 180 days of gestation. Six pregnant ponies served as noninfected controls. Each infected pony had clinical signs of equine monocytic ehrlichiosis, was confirmed to be ehrlichemic, and developed an IFA titer to E risticii. Two infected ponies became recumbent, were unresponsive to supportive care, and were euthanatized. After recovery from clinical illness, the remaining ponies were observed throughout gestation for reproductive abnormalities. On abortion, each fetus was necropsied and tissue specimens from the liver, bone marrow, spleen, colon, and mesenteric lymph nodes were inoculated into canine monocyte cell cultures. Six infected ponies aborted at a mean 217 days of gestation, which was between postinoculation days 65 and 111. Five fetuses were recovered for evaluation, and E risticii was isolated from 4 of them. All 5 fetuses recovered had similar histologic findings, including enterocolitis, periportal hepatitis, and lymphoid hyperplasia with necrosis of the mesenteric lymph nodes and spleen. All fetuses tested negative for IgG to E risticii, although 3 had low IgM titer to E risticii. The remaining 5 infected ponies had normal parturition. Presuckle IFA titer to E risticii was measured in 4 of the term foals, and results for 3 were positive. Two foals from infected ponies were monitored for 6 months and daily gain in body weight was comparable to that of a control foal. None of the control ponies became ill or seroconverted during the clinical illness phase, and none aborted throughout gestation. Two control ponies seroconverted to E risticii 6 weeks before parturition. Results of this study indicate that E risticii is a primary abortifacient under experimental conditions.

Fourteen ponies and 3 horses were inoculated with Ehrlichia risticii 2 to 20 months after a similar initial inoculation. Although all 17 had clinical signs of equine ehrlichial colitis after the first inoculation, 16 of 17 remained clinically normal following the second inoculation. The remaining pony had a transient fever and developed signs of depression. Before the initial inoculation, none of the animals had a detectable antibody titer to E risticii. All animals developed titers after the initial infection; however, a significant change of titer did not develop after reinoculation in most animals.

Ten adult dogs (5 Beagles and 5 mixed-breed dogs) were inoculated IV with canine platelets containing Ehrlichia platys. Inclusions and morulae of E platys developed in platelets of infected dogs at 10 to 14 days after inoculation, followed by marked thrombocytopenia at 14 to 21 days. Parasitemia and marked thrombocytopenia recurred at 24 to 28 days after inoculation. Increased numbers of megakaryocytes were observed in marrow aspirate smears from infected dogs, indicative of regenerative thrombocytopenia. Prior to infection, platelet-rich plasma from these dogs was determined to have similar aggregatory response to arachidonate. After infection with E platys, the aggregatory response of platelet-rich plasma to collagen or 3 dilutions of adenosine diphosphate was evaluated. A statistically significant (P < 0.05) inhibition of platelet aggregatory response to the lowest dilution of adenosine diphosphate was detected for mixed-breed dogs, whereas aggregation responses were unchanged in Beagles. Results indicate that platelet activation may occur in dogs with acute ehrlichial infection.