v w V p f V w A V a 1 c i p P The primordial varicella-zoster virus (VZV) is estimated to have emerged 70 million years ago (McGeoch and Davison, 1999). Even with this ancient history, VZV has long been considered to have a stable genome. The complete DNA sequence (124,884 bp) was published by Davison and Scott in 1986. The VZV genomic sequence with its 69 open reading frames (ORFs) is frequently cited, even for strains other than Dumas, with the presumption of identity. The recent discovery of a genetically variant virus has led us to reconsider the presumption of VZV immutability (Santos et al., 1998). In the case of VZV-MSP, we discovered a virus with a missense mutation in a major glycoprotein called gE. Because this point mutation eliminates a B cell epitope, we consider this polymorphism to have been the likely end result of an antibody selection in an infected human host (Grose, 1999). This mutant virus exhibited an accelerated cell-tocell spread phenotype in both infected cells and the SCID-hu animal model of VZV infection (Santos et al., 000). Prior to this observation, all VZV wild-type isolates nd strains were considered to have a phenotype so imilar as to be indistinguishable. Based upon its likely pathogenesis, VZV undergoes nly about 20 cycles of replication during primary infecion (Grose, 2000); even though this period provides imited opportunity for genetic diversification, some hanges obviously occurred to generate VZV-MSP. herefore, we postulated that a slowly coevolving VZV enome could be investigated by a similar genetic aproach as directed toward the human genome (Collins et l., 1998). The Human Genome Project has concentrated normous effort over the past 3 years on the identifica-