Herpesvirus quiescence (QIF) in neuronal cells VI: Correlative analysis demonstrates usefulness of QIF-PC12 cells to examine HSV-1 latency, reactivation and genes implicated in its regulation

Abstract

Purpose. To compare the usefulness of the in vitro quiescently infected (QIF)-PC12 cell model with the in vivo rabbit eye model of latency for the study of herpes simplex virus (HSV) genes implicated in reactivation from latency. Methods. HSV-1 strains 17+/pR20.5/5 and 17+/pR20.5/ 5/LAT, that were previously constructed by insertion of genes encoding beta-galactosidase, green fluorescent protein (GFP) or the latency associated transcript (LAT) open reading frame in the U S 5 region, were used to examine viral growth and inducible reactivation in the two models. Results. 17+/pR20.5/5 exhibited diminished reactivation phenotype when compared with wild type 17+ in neuronal cells (i.e., QIF-PC12 cell model) and the rabbit eye model of latency. 17+/pR20.5/5/LAT, which contains the deregulated LAT gene, reactivated at wild type levels. Analysis of growth in neurally differentiated (ND)-PC12 cells demonstrated a low proportion of QIF cells expressed virus-encoded signals during the quiescent infection and a direct relationship between lytic viral growth in neuronal cells and reactivation phenotype. Even though 17+/pR20.5/5/LAT produced a more severe acute infection in the rabbit cornea, the different reactivation efficiency of 17+/pR20.5/5 and 17+/pR20.5/5/LAT in vivo and in vitro was not attributed to different viral genome copy number in the cells harboring cryptic genomes. Conclusions. We conclude that 1) viral growth in neuronal cells correlates with reactivation phenotype in vivo and in vitro, 2) 17+/pR20.5/5 is attenuated in viral growth and reactivation in both models, and 3) 17+/pR20.5/5/LAT demonstrates wild-type phenotype for reactivation in both models. Attenuation of 17+/pR20.5/5 could be the result of the disruption of U S 5 or a second site mutation. If the attenuation is the result of U S 5 disruption, a gene that provides anti-apoptotic functions, this attenuation is more than compensated for by the expression of the LAT ORF. Overall, the findings indicate that the QIF-PC12 cell model is useful for segregating phases of reactivation, and particularly studying the inductive events involved in reactivation of a cryptic viral genome in neurally differentiated cells.