Herpes simplex virus type-1 (HSV-1) has been used for gene delivery in
the nervous system for the treatment of brain tumors and other neurol
ogical diseases. In most protocols, recombinant viruses containing the
gene of interest are directly injected into the brain. Since many peo
ple harbor latent wild-type HSV-1 virus in sensory ganglia and other r
egions of the nervous system, there is a potential risk that the injec
ted recombinant virus may reactivate the latent wild-type virus to cau
se severe encephalitis. The present study used two rat latent infectio
n models to evaluate this risk. Adult rats were infected with wild-typ
e kos by cornea scarification or by intracerebral injection, and after
the establishment of latency, the ICP6(-) strain hrR3 was injected in
tracerebrally. In the control group, the latent virus was reactivated
by treatment of cadmium sulfate Viral shedding from tears was detected
by incubation with Vero cells, and the trigeminal ganglia, cortical t
issue and the eyes were collected to detect reactivated wild-type viru
s by RT-PCR. Our results showed that while the reactivated wild-type v
irus was readily detectable in the cadmium-sulfate-treated animals, in
tracerebrally infected hrR3 did not reactivate the latent virus in eit
her the cornea model or the cerebral model. These results indicate tha
t intracranial infection of partially defective recombinant virus may
bear little risk of reactivation latent wild-type virus harbored in th
e sensory ganglia or the brain in our animal model.