SummaryAfter replication at sites of initial inoculation, herpes simplex virus type 1 and 2 (HSV-1 and HSV-2) establish lifelong latent infections of the sensory and autonomic neurons of the ganglia serving those sites. Periodically, the virus reactivates from these neurons, and travels centripetally along the neuronal axon to cause recurrent epithelial infection. The major clinically observed difference between infectious with herpes simplex virus type 1 and type 2 is the anatomic site specificity of recurrence. HSV-1 reactivates most efficiently and frequently from trigeminal ganglia, causing recurrent ocular and oral-facial lesions, while HSV-2 reactivates primarily from sacral ganglia causing recurrent genital lesions. An intertypic recombinant virus was constructed and evaluated in animal models of recurrent ocular and genital herpes. Substitution of a 2.8-kbp region from the HSV-1 latency-associated transcript (LAT) for native HSV-2 sequences caused HSV-2 to reactivate with an HSV-1 phenotype in both animal models. The HSV-2 phenotype was restored by replacing the mutated sequences with wild-type HSV-2 LAT-region sequences. These sequences or their products must act specifically in the cellular environments of trigeminal and sacral neurons to promote the reactivation patterns characteristic of each virus. p primary or initial infections with herpes simplex virus type 1 and 2 (HSV-1 and HSV-2) ~ are clinically indistinguishable. While social habits and other factors contribute to varying sites of initial infections, the predilection of virus to reactivate from specific ganglionic sites maintains the well-known associations of HSV-2 with genital herpes and of HSV-1 with ocular and oral-facial herpes (1, 2). The HSV latency-associated transcripts (LATs) are a family of transcripts specified by the genomic long repeat regions. The more abundant nuclear LAT introns (2.2 kbp in HSV-2 [3-5], 2.0 kbp and 1.5 kbp in HSV-1 [6-9]) are processed via a splicing mechanism from less stable ~8.5-kbp primary LAT transcripts (10). The LATs of HSV-1 and HSV-2 (11-13) are required for efficient reactivation from latency in vivo, but have not been shown to influence acute replication of virus or establishment of latency (14-19) in most 1Abbreviations used in this paper: HSV-1 and HSV-2, herpes simplex virus type 1 and 2; LAT, latency-associated transcript; PI postinoculation; PRK, primary rabbit kidney. animal models, although there is some evidence suggesting an effect on establishment of latency in a mouse model of infection with HSV-1 (20). The sequences of the LAT regions differ significantly between HSV-1 and HSV-2 (4, 21, 22), leading to the hypothesis that differences in the LATs could be responsible for site-specificity of virus recurrence. To test this hypothesis, we introduced an alteration into HSV-2 strain 333 that substituted the LAT region from HSV-1 strain 17syn+ for native HSV-2 LAT sequences, and constructed a rescuant of this mutant. We then tested the ability of parent, mutant, and rescuant viruses to replic...