Clinical, pathological, immunological, and virological evidence suggests that simian varicella virus (SW) infection of primates is the counterpart ofvaricella-zoster virus infection of humans. To determine whether these two viruses share similrities in their properties during latency, we analyzed ganglia and brain of an African green monkey experimentally infected with SW for the presence of viral nucleic acid using the polymerase chain reaction technique. We detected SW DNA in dorsal root ganglia but not in brain of this monkey, which demonstrated no apparent clinical signs of SW infection. Ohr results suggest that SW becomes latent in monkey ganglia and that latency can develop in the absence of clinical varicella (chickenpox). These studies provide an animal model system to study varicella virus latency.Varicella-zoster virus (VZV) causes chickenpox (varicella) in childhood, becomes latent in dorsal root ganglia, and reactivates decades later to produce shingles (zoster) in adults. The host range of VZV is restricted exclusively to humans,' so that no animal model has been developed to study VZV pathogenesis and latency. However, there have been various reports of spontaneous epizootic outbreaks of varicella in housed monkeys. Like human varicella, simian varicella frequently occurs in epidemics, has an incubation period of 1 or more weeks, and histologic examination of skin lesions reveals necrosis and hemorrhagic foci of epidermal cells containing intranuclear inclusions (1). The causative agent of simian varicella is a member of the herpesvirus family, simian varicella virus (SVV), which is antigenically crossreactive with VZV (2, 3). SVV also reactivates in monkeys to produce disseminated varicella (3). If the'site of SVV latency is ganglionic as has been shown for human VZV (4, 5), then studies of SVV pathogenesis and latency in primates would be relevant to human VZV infection. In the present study, we demonstrate the presence of latent SVV DNA in ganglia and the establishment of latent infection in a monkey which seroconverted after experimental infection with SVV but failed to develop clinical varicella. METHODS SW. The deltaherpesvirus (DHV) strain of SVV was originally isolated from a naturally infected patas monkey (Erythocebus patas) as described (6) and propagated in Vero cells, and a virus stock was prepared as described (7,8).Virus Inoculation. A 6-year-old female African green monkey (Cercopithecus aethiops) was inoculated with SVV as described (8). Briefly, 105 plaque-forming units of SVV in 1.5 ml was administered by intratracheal catheter into the upper bronchi, and 1.5 ml was injected subcutaneously over the abdomen. Two days after inoculation, the monkey was treated intramuscularly with a purine nucleoside, B722U (Burroughs Wellcome), at 150 mg/kg of body weight per day for 10 days. The monkey was examined daily for a varicelliform eruption. Blood was collected every few days for detection of viremia and on days 14 and 21 after inoculation for determination of antibody titer to SV...