High-dose ocular infection with a herpes simplex virus type 1 ICP34.5 deletion mutant produces no corneal disease or neurovirulence yet results in wild-type levels of spontaneous reactivation

We report here that in the rabbit ocular model of herpes simplex virus type 1 (HSV-1) latency, spontaneous reactivation of the HSV-1 ICP34.5 deletion mutant d34.5 increased significantly in response to increasing infections doses. At the highest infectious dose of d34.5, the spontaneous reactivation rate was indistinguishable from that of wild-type virus (average spontaneous reactivation rates for d34.5, 0.3 to 1.4% at 2 x 10⁵ PFU per eye, 3.4% at 2 x 10⁶ PFU per eye, and 6.3 to 11.5% at 1 x 10⁸ PFU per eye; average spontaneous reactivation rates for marker-rescued virus, 7.7 to 19.6% at 2 x 10⁵ PFU per eye). The percentage of latency-associated transcript (LAT) RNA-positive neurons in sections from trigeminal ganglia (TG) of rabbits latently infected with d34.5 demonstrated a similar dose-response effect as estimated by in situ hybridization (0.05% LAT RNA-positive neurons at 2 x 10⁵ PFU per eye and 0.1% LAT RNA-positive neurons at 1 x 10⁸ PFU per eye; P = 0.002). In contrast, even at the highest infectious dose (1 x 10⁸ PFU per eye), d34.5 was less virulent (23 of 23 survivors) than the normal infectious dose (2 x 10⁵ PFU per eye) of marker-rescued virus (14 of 27 survivors; P < 0.0001). In addition, at 1 x 10⁸ PFU per eye, d34.5 produced virtually no corneal disease, compared with the production of severe corneal disease by 2 x 10⁵ PFU of marker-rescued virus per eye (P < 0.0001). Thus, at increasing infectious doses of d34.5, both spontaneous reactivation and the percentage of neurons expressing LAT appeared to increase, without a corresponding increase in virulence. These results strongly suggest that (i) the phenotypes of neurovirulence and spontaneous reactivation are separable, (ii) the phenotypes of corneal disease and spontaneous reactivation are separable, and (iii) the decreased rate of spontaneous reactivation previously reported for d34.5 (G. C. Perng, R. L. Thompson, N. M. Sawtell, W. E. Taylor, S. M. Slanina, H. Ghiasi, R. Kaiwar, A. B. Nesburn, and S. L. Wechsler, J. Virol. 69:3033-3041, 1995) is at least partially due to a reduced rate of establishing latency.