Seventeen Minnesota and Wisconsin dogs with granulocytic ehrlichosis were studied. The diagnoses were made by finding ehrlichia morulae in peripheral blood neutrophils. Eight dogs were studied retrospectively, and nine dogs were studied prospectively. The medical records of all dogs were reviewed. Eighty-eight percent of the dogs were purebred and 76% were spayed females. The median age was 8 years. Sixty-five percent of the cases were diagnosed in October and November. Fever and lethargy were the most common clinical signs. The most frequent laboratory findings were lymphopenia, thrombocytopenia, elevated activities of serum alkaline phosphatase and amylase, and hypoalbuminemia. No dogs seroreacted to Ehrlichia canis or Ehrlichia chaffeensis antigens, which are cross-reactive. Seventy-five percent of the dogs tested during the acute phase of disease and 100% of the dogs tested during convalescence were seropositive for E. equi antigens. Granulocytic ehrlichial 16S rRNA gene DNAs from six dogs were amplified by PCR. Sequence analysis of a 919-bp sequence of the ehrlichial 16S rRNA gene amplified by PCR from the blood of two dogs revealed the agent to be identical to the agent of human granulocytic ehrlichiosis in Minnesota and Wisconsin and to be very similar to E. equi and Ehrlichia phagocytophila and less similar to E. canis, Ehrlichia ewingii, and E. chaffeensis. The geographic, clinical, serologic, and molecular evidence indicates that granulocytic ehrlichiosis in Minnesota and Wisconsin dogs is not caused by E. ewingii, but suggests that it is a zoonotic disease caused by an agent closely related to E. equi and that dogs likely contribute to the enzootic cycle and human infection.

Human granulocytotropic ehrlichias are tick-borne bacterial pathogens that cause an acute, life-threatening illness, human granulocytic ehrlichiosis (HGE). Ehrlichias within neutrophil granulocytes that invade tick bite sites are likely ingested by the vector, to be transmitted to another mammalian host during the tick’s next blood meal. Thus, the cycle of replication and development in the vector is prerequisite to mammalian infection, and yet these events have not been described. We report tick cell culture isolation of two strains of the HGE agent directly from an infected horse and a dog and have also established a human isolate from HL60 culture in tick cells, proving that the blood stages of the HGE agent are infectious for tick cells, as are those replicating in the human cell line HL60. This required changes to the culture system, including a new tick cell line. In tick cell layers, the HGE agent induced foci of infection that caused necrotic plaques and eventual destruction of the culture. Using the human isolate and electron microscopy, we monitored adhesion, internalization, and replication in vector tick cells. Both electron-lucent and -dense forms adhered to and entered cells by a mechanism reminiscent of phagocytosis. Ehrlichial cell division was initiated soon after, resulting in endosomes filled with numerous ehrlichias. During early development, pale ehrlichias with a tight cell wall dominated, but by day 2, individual bacteria condensed into dark forms with a rippled membrane. These may become compacted into clumps where individual organisms are barely discernible. Whether these are part of an ehrlichia life cycle or are degenerating is unknown.

The natural reservoirs for the agent of human granulocytic ehrlichiosis (HGE) are suspected to be the small mammals that host immature stages of Ixodes scapularis ticks. To determine if such small mammals are naturally infected, we collected blood and serum samples from small mammal species in rural and suburban areas of Minneapolis and St. Paul, Minn. Samples were collected from white-footed mice (Peromyscus leucopus), eastern chipmunks (Tamias striatus), southern red-backed voles (Clethrionomys gapperi), and insectivorous shrews (Blarina brevicauda and Sorex cinereus). Blood samples were tested by PCR for active infection with the HGE agent, and sera from P. leucopus mice were tested for serologic evidence of infection by indirect immunofluorescence. PCR analyses revealed the presence of HGE agent DNA in 20 of the 190 samples (10.5%) tested. Of the 119 P. leucopus mouse serum samples that were analyzed, 12 (10.1%) contained Ehrlichia equi antibodies. In 3 of 119 (2.5%) P. leucopus mice from which both blood and serum were collected, HGE agent DNA and antibodies against E. equi were present. Animals with evidence of infection with the HGE agent are widely distributed around the Minneapolis-St. Paul area in regions with known I. scapularis tick activity. Small mammals that are frequent hosts for larval I. scapularis ticks and that are found in areas where HGE occurs are likely to be a major reservoir from which infected ticks that bite humans are derived.

Equine granulocytic ehrlichiosis (EGE) is caused by infection with Ehrlichia equi. EGE has been reported primarily in northern California, where E equi is transmitted by the tick Ixodes pacificus. Reports of EGE and the emergence of human granulocytic ehrlichia in Minnesota prompted a seroprevalence study of E equi in horses of Minnesota and Wisconsin. Tick (Ixodes scapularis) endemic areas of Minnesota and Wisconsin were compared to nonendemic regions of Minnesota. Indirect fluorescent antibody was used to detect the presence of serum antibodies to E equi. Serum samples from healthy horses, 375 samples from I scapularis endemic counties, and 366 samples from nonendemic counties were screened at a 1:40 dilution. Results demonstrated a seroprevalence of 17.6% in endemic areas versus 3.8% in nonendemic areas. Ehrlichial DNA from 2 samples was successfully amplified by polymerase chain reaction and 919 base pairs were sequenced. The DNA sequence of 1 Minnesota/Wisconsin strain differed from the GenBank strain (M73223) of E equi at positions 84 and 886 and from the MRK strain of E equi at position 84, and was identical to the human granulocytic ehrlichiosis (HGE) agent. The 2nd Minnesota/Wisconsin strain was identical to the 1st with the exception of a substitution of "A" at position 453 that is not present in E phagocytophila, E equi, or HGE agent strain sequences. Based on the results of this study, we concluded that E equi is present and causes infection in horses in Minnesota and Wisconsin. The occurrence of infection is higher in tick endemic regions.

Inflammatory lesions in bone marrow, observed during a 2 year period at the College of Veterinary Medicine, University of Minnesota, were reviewed. Of 24 bone marrow specimens with evidence of inflammation, six were classified as acute inflammation, nine as fibrinous inflammation, five as chronic inflammation/hyperplasia, three as granulomatous inflammation and one as nodular lymphoid hyperplasia. Acute inflammation commonly accompanied bacterial sepsis. Two patterns of acute inflammation were identified. One pattern consisted of multifocal microabscesses. The other pattern of acute inflammation consisted of perivascular infiltrates of neutrophils, fibrin, edema, and hemorrhage. The most common disorder associated with fibrinous inflammation was disseminated intravascular coagulopathy. Chronic inflammation was difficult to differentiate from chronic immune stimulation. Discrete granulomas were identified in the marrow of animals with systemic mycotic disease, idiopathic systemic granulomatous disease, and serous atrophy of fat. This study indicates that a broad variety of inflammatory disorders occur in animal bone marrow and that these disorders can be classified based on general categories of inflammation described in other tissues.