Herpes Simplex Virus Latency in Cultured Cells

The molecular basis of herpes simplex virus (HSV) latency has been investigated by analysis of a latent interaction between HSV and cells in culture. An in vitro latency system for HSV, based on previous work by E. Notarianni and C. M. Preston, has been characterised, in which incubation at a supraoptimal temperature converts HSV to a latent state within tissue culture cells. Human foetal lung (HFL) cells were infected at low multiplicity with HSV and, following adsorption, the cultures were shifted to 42C for 6 days, then downshifted to a temperature permissive for HSV replication for a further 4 to 6 days. During the latter incubation period no virus was usually detectable and the HSV was considered to be in a latent state. HSV could be reactivated from this latent state at high efficiency by intertypic superinfection of the cultures with HSV mutants or with human cytomegalovirus. To define the HSV gene products involved in latency, the behaviour of various temperature-sensitive (ts), insertion (in) and deletion (dl) mutants of HSV in the in vitro latency system was examined. The rationale behind this approach is that mutants which fail to become established in a latent state, or which fail to reactivate latent HSV, must lack functions involved in establishment or reactivation, respectively. Two mutants of HSV type 1 (HSV-1) used in these studies, tsK and inl 411, do not synthesise active immediate early (IE) polypeptide Vmw175 and are blocked at a very early stage of the virus replication cycle, and a third mutant of HSV-1, d11403, does not produce IE polypeptide Vmw110, but otherwise exhibits a pattern of protein synthesis indistinguishable from that of wt HSV-1. All mutants tested were able to establish latency in HFL fibroblasts and could be reactivated by intertypic superinfection with HSV or with human cytomegalovirus, showing that no viral DNA synthesis and little or no viral gene expression is necessary for the establishment of latency in vitro, and that at most the viral proteins involved are IE polypeptides Vmw12, Vmw63, Vmw68, Vmw175 or Vmw110, the early polypeptide Vmw136, and, possibly, components of the input virion. Reactivation of latent HSV-2 was achieved by superinfection with tsK or in1411. However, superinfection with d11403 failed to reactivate latent HSV-2 as a consequence of a deletion in the region of the genome encoding Vmw110, strongly suggesting that Vmw110, which is known to regulate gene expression by trans-activation, is required for reactivation in the in vitro latency system. The results presented do not indicate whether Vmw110 acts alone or in conjunction with one or more of the virion components and/or the other IE polypeptides, excluding Vmwl75. Harris et al. (1989) have recently shown that Vmw110 alone can reactivate latent HSV in vitro. One possibility is that a block in viral gene expression occurs at a very early stage in the viral cycle, either as the direct cause or as a consequence of establishment of latency, and that the block can be released by the Vmw110 gene product, thereby allowing HSV to continue into the lytic cycle. The latent state of HSV DNA in vivo appears to be Endless' and is therefore either circular, concatemeric or integrated via regions of the genome other than the termini. A recent report shows that the majority of latent HSV DNA in vivo is extrachromosomal, suggesting that latent HSV DNA in vivo is not likely to be integrated into cellular DNA. The physical nature of the HSV DNA in the in vitro latency system described has been determined. The relative proportion of latent HSV genomes, initially present in vitro at 0.03-0.1 copies per cell, was selectively increased and the presence of joint and terminal fragments of HSV in latent HSV DNA was investigated by the use of a modified Southern hybridisation procedure. During the 4