General and specific alterations in programming of global viral gene expression during infection by VP16 activation-deficient mutants of herpes simplex virus type 1

Abstract

During productive infection by herpes simplex virus 1 (HSV-1), viral gene expression occurs in a temporally regulated cascade in which transcription of the viral immediate-early (IE) genes is strongly stimulated by the virion protein VP16. We have employed an oligonucleotide microarray to examine the effect of VP16 mutations on the overall pattern of viral gene expression following infection of HeLa cells. This microarray detects essentially all HSV-1 transcripts with relative and absolute levels correlating well with known kinetics of expression. This analysis revealed that deletion of the VP16 activation domain sharply reduced overall viral gene expression; moreover, the pattern of this reduced expression varied greatly from the pattern of a wild-type (wt) infection. However, when this mutant virus was delivered at a high multiplicity of infection or in the presence of the cellular stress inducer hexamethylene bisacetamide, expression was largely restored to the wt levels and pattern. Infection with virions that deliver wt VP16 protein at the start of infection but synthesize only truncated VP16 resulted in a normal kinetic cascade. This suggests that newly synthesized VP16 does not play a significant role in the expression of later classes of transcripts. The VP16 activation domain comprises two subregions. Deletion of the C-terminal subregion resulted in minimal changes in the level and profile of gene expression compared to a normal (wt) cascade. In contrast, deletion of the N-terminal subregion reduced the overall expression levels and skewed the relative levels of IE transcripts but did not significantly alter the kinetic pattern of early and late transcript expression. We conclude that the general activation of IE gene transcription by VP16, but not the specific ratios of IE transcripts, is necessary for the subsequent ordered expression of viral genes. Moreover, this report establishes the feasibility of microarray analysis for globally assessing viral gene expression programs as a function of the conditions of infection.

FIG. 1.

Gene expression levels during infection with wt (RP5R) or VP16 activation domain-deficient (RP5) isolates of HSV-1. Microarray chips were hybridized with probes prepared from RNA harvested from HeLa cells infected with wt or RP5 virus at 2, 4, or 8 hpi. The levels of expression for IE (open), E (shaded), L (solid), and latency-associated (LAT) (hatched) transcripts are in arbitrary units. The numbers corresponding to individual transcripts are defined in Table 1.

FIG. 2.

Correlation of transcript abundances between expression levels of RP5R (wt) and RP5 at 2, 4, and 8 hpi. The relative abundances of transcripts for RP5R infection are displayed on the abscissa, and the relative abundances of transcripts from an RP5 infection are on the ordinate. Each point represents a signal for IE, E, or L transcripts at 2, 4, or 8 hpi. For each plot, the ordinate and abscissa have the same units, and the expected locations of equivalent signals under both conditions (i.e., a correlation of 1.0) are indicated by the 45° diagonals.

FIG. 3.

Gene expression levels during infection with RP3 or RP4. Microarray chips were hybridized with probes prepared from RNA harvested from HeLa cells infected with RP3 or RP4 at 2, 4, or 8 hpi. The expression levels of IE (open), E (shaded), L (solid), and latency-associated (LAT) (hatched) transcripts are in arbitrary units.