Search

The varicella-zoster virus (VZV) ORF61 protein is necessary for normal replication in vitro and virulence in human skin xenografts in the severe combined immunodeficiency mouse model in vivo . These experiments identify a hydrophobic domain that mediates ORF61 self-interaction. While not needed to inhibit host cell defenses, disruption of this domain (residues 250 to 320) severely impairs VZV growth, transactivation of the immediate early 63 and glycoprotein E genes, and the pathogenesis of VZV skin infection in vivo .

The tropism of herpes simplex virus (HSV-1) for human sensory neurons infected in vivo was examined using dorsal root ganglion (DRG) xenografts maintained in mice with severe combined immunodeficiency (SCID). In contrast to the HSV-1 lytic infectious cycle in vitro , replication of the HSV-1 F strain was restricted in human DRG neurons despite the absence of adaptive immune responses in SCID mice, allowing the establishment of neuronal latency. At 12 days after DRG inoculation, 26.2% of human neurons expressed HSV-1 protein and 13.1% expressed latency-associated transcripts (LAT). Some infected neurons showed cytopathic changes, but HSV-1, unlike varicella-zoster virus (VZV), only rarely infected satellite cells and did not induce fusion of neuronal and satellite cell plasma membranes. Cell-free enveloped HSV-1 virions were observed, indicating productive infection. A recombinant HSV-1-expressing luciferase exhibited less virulence than HSV-1 F in the SCID mouse host, enabling analysis of infection in human DRG xenografts for a 61-day interval. At 12 days after inoculation, 4.2% of neurons expressed HSV-1 proteins; frequencies increased to 32.1% at 33 days but declined to 20.8% by 61 days. Frequencies of LAT-positive neurons were 1.2% at 12 days and increased to 40.2% at 33 days. LAT expression remained at 37% at 61 days, in contrast to the decline in neurons expressing viral proteins. These observations show that the progression of HSV-1 infection is highly restricted in human DRG, and HSV-1 genome silencing occurs in human neurons infected in vivo as a consequence of virus-host cell interactions and does not require adaptive immune control.

Analyses of varicella-zoster virus (VZV) protein expression during latency have been discordant, with rare to many positive neurons detected. We show that ascites-derived murine and rabbit antibodies specific for VZV proteins in vitro contain endogenous antibodies that react with human blood type A antigens in neurons. Apparent VZV neuronal staining and blood type A were strongly associated (by a χ 2 test, α = 0.0003). Adsorption of ascites-derived monoclonal antibodies or antiserum with type A erythrocytes or the use of in vitro -derived VZV monoclonal antibodies eliminated apparent VZV staining. Animal-derived antibodies must be screened for anti-blood type A reactivity to avoid misidentification of viral proteins in the neurons of the 30 to 40% of individuals who are blood type A.

Varicella-zoster virus (VZV) causes varicella and establishes latency in sensory nerve ganglia, but the characteristics of VZV latency are not well defined. Immunohistochemical detection of the VZV immediate-early 63 (IE63) protein in ganglion neurons has been described, but there are significant discrepancies in estimates of the frequency of IE63-positive neurons, varying from a rare event to abundant expression. We examined IE63 expression in cadaver ganglia using a high-potency rabbit anti-IE63 antibody and corresponding preimmune serum. Using standard immunohistochemical techniques, we evaluated 10 ganglia that contained VZV DNA from seven individuals. These experiments showed that neuronal pigments were a confounding variable; however, by examining sections coded to prevent investigator bias and applying statistical analysis, we determined that IE63 protein, if present, is in a very small proportion of neurons (

Varicella-zoster virus (VZV) glycoprotein E (gE) is the most abundant glycoprotein in infected cells and, in contrast to those of other alphaherpesviruses, is essential for viral replication. The gE ectodomain contains a unique N-terminal region required for viral replication, cell-cell spread, and secondary envelopment; this region also binds to the insulin-degrading enzyme (IDE), a proposed VZV receptor. To identify new functional domains of the gE ectodomain, the effect of mutagenesis of the first cysteine-rich region of the gE ectodomain (amino acids 208 to 236) was assessed using VZV cosmids. Deletion of this region was compatible with VZV replication in vitro, but cell-cell spread of the rOka-ΔCys mutant was reduced significantly. Deletion of the cysteine-rich region abolished the binding of the mutant gE to gI but not to IDE. Preventing gE binding to gI altered the pattern of gE expression at the plasma membrane of infected cells and the posttranslational maturation of gI and its incorporation into viral particles. In contrast, deletion of the first cysteine-rich region did not affect viral entry into human tonsil T cells in vitro or into melanoma cells infected with cell-free VZV. These experiments demonstrate that gE/gI heterodimer formation is essential for efficient cell-cell spread and incorporation of gI into viral particles but that it is dispensable for infectious varicella-zoster virion formation and entry into target cells. Blocking gE binding to gI resulted in severe impairment of VZV infection of human skin xenografts in SCIDhu mice in vivo, documenting the importance of cell fusion mediated by this complex for VZV virulence in skin.

The assembly of herpesvirus capsids is a complex process involving interactions of multiple proteins in the cytoplasm and in the nucleus. Based on comparative genome analyses, varicella-zoster virus (VZV) open reading frame 23 (ORF23) encodes a conserved capsid protein, referred to as VP26 (UL35) in other alphaherpesviruses. Mutagenesis using a VZV bacterial artificial chromosome system showed that ORF23 was dispensable for replication in vitro. However, the absence of ORF23 disrupted capsid assembly in a melanoma cell line. Expression of ORF23 as a red fluorescent protein (RFP) fusion protein appeared to have a dominant negative effect on replication that was rescued by ORF23 expression from a nonnative site in the VZV genome. In contrast to its VP26 homolog, ORF23 has an intrinsic nuclear localization capacity that was mapped to an SRSRVV motif at residues 229 to 234 in the extreme C terminus of ORF23. In addition, coexpression with ORF23 resulted in nuclear import of the major capsid protein, ORF40. VZV ORF33.5 also translocated ORF40, which may provide a redundant mechanism in vitro but appears insufficient to overcome the dominant negative effect of the monomeric RFP-ORF23 (mRFP23) fusion protein. ORF23 was required for VZV infection of human skin xenografts, indicating that ORF33.5 does not compensate for lack of ORF23 in vivo. These observations suggest a model of VZV capsid assembly in which nuclear transport of the major capsid protein and associated proteins requires ORF23 during VZV replication in the human host. If so, ORF23 expression could be a target for a novel antiviral drug against VZV.

Varicella-zoster virus (VZV) glycoprotein E (gE) is essential for viral replication and is involved in cell-to-cell spread, secondary envelopment, and entry. We created a set of mutations in the gE promoter to investigate the role of viral and cellular transcriptional factors in regulation of the gE promoter. Deletion or point mutation of the two Sp1 sites in the gE promoter abolished Sp1 binding and IE62-mediated transactivation of the gE promoter in vitro. Incorporation of the deletion or the point mutations disrupting both of the Sp1 binding sites into the VZV genome was not compatible with viral replication. A point mutation altering the atypical Sp1 binding site was lethal, while altering the second site impaired VZV replication significantly, indicating functional differences between the two Sp1 binding sites. Deletions in the gE promoter that abolished putative binding sites for cellular transcriptional factors other than Sp1, identified by bioinformatics analysis, were inserted in the VZV genome. Replication of the viruses with mutations of the gE promoter did not differ from control recombinants in melanoma cells or primary human tonsil T cells in vitro. These deletions did not affect infection of human skin xenografts in SCIDhu mice. These results indicate that Sp1 is required for IE62-mediated transactivation of the gE promoter and that this transcriptional factor appears to be the only cellular factor essential for regulation of the gE promoter.

The immediate-early 63-kDa (IE63) protein of varicella-zoster virus (VZV) is a phosphoprotein encoded by open reading frame (ORF) ORF63/ORF70. To identify functional domains, 22 ORF63 mutations were evaluated for effects on IE63 binding to the major VZV transactivator, IE62, and on IE63 phosphorylation and nuclear localization in transient transfections, and after insertion into the viral genome with VZV cosmids. The IE62 binding site was mapped to IE63 amino acids 55 to 67, with R59/L60 being critical residues. Alanine substitutions within the IE63 center region showed that S165, S173, and S185 were phosphorylated by cellular kinases. Four mutations that changed two putative nuclear localization signal (NLS) sequences altered IE63 distribution to a cytoplasmic/nuclear pattern. Only three of 22 mutations in ORF63 were compatible with recovery of infectious VZV from our cosmids, but infectivity was restored by inserting intact ORF63 into each mutated cosmid. The viable IE63 mutants had a single alanine substitution, altering T171, S181, or S185. These mutants, rOKA/ORF63rev[T171], rOKA/ORF63rev[S181], and rOKA/ORF63rev[S185], produced less infectious virus and had a decreased plaque phenotype in vitro. ORF47 kinase protein and glycoprotein E (gE) synthesis was reduced, indicating that IE63 contributed to optimal expression of early and late gene products. The three IE63 mutants replicated in skin xenografts in the SCIDhu mouse model, but virulence was markedly attenuated. In contrast, infectivity in T-cell xenografts was not altered. Comparative analysis suggested that IE63 resembled the herpes simplex virus type 1 U S 1.5 protein, which is expressed colinearly with ICP22 (U S 1). In summary, most mutations of ORF63 made with our VZV cosmid system were lethal for infectivity. The few IE63 changes that were tolerated resulted in VZV mutants with an impaired capacity to replicate in vitro. However, the IE63 mutants were attenuated in skin but not T cells in vivo, indicating that the contribution of the IE63 tegument/regulatory protein to VZV pathogenesis depends upon the differentiated human cell type which is targeted for infection within the intact tissue microenvironment.

To investigate the role of the ORF47 protein kinase of varicella-zoster virus (VZV), we constructed VZV recombinants with targeted mutations in conserved motifs of ORF47 and a truncated ORF47 and characterized these mutants for replication, phosphorylation, and protein-protein interactions in vitro and for infectivity in human skin xenografts in the SCID-hu mouse model in vivo. Previous experiments showed that ROka47S, a null mutant that makes no ORF47 protein, did not replicate in skin in vivo (J. F. Moffat, L. Zerboni, M. H. Sommer, T. C. Heineman, J. I. Cohen, H. Kaneshima, and A. M. Arvin, Proc. Natl. Acad. Sci. USA 95:11969-11974, 1998). The construction of VZV recombinants with targeted ORF47 mutations made it possible to assess the effects on VZV infection of human skin xenografts of selectively abolishing ORF47 protein kinase activity. ORF47 mutations that resulted in a C-terminal truncation or disrupted the DYS kinase motif eliminated ORF47 kinase activity and were associated with extensive nuclear retention of ORF47 and IE62 proteins in vitro. Disrupting ORF47 kinase function also resulted in a marked decrease in VZV replication and cutaneous lesion formation in skin xenografts in vivo. However, infectivity in vivo was not blocked completely as long as the capacity of ORF47 protein to bind IE62 protein was preserved, a function that we identified and mapped to the N-terminal domain of ORF47 protein. These experiments indicate that ORF47 kinase activity is of critical importance for VZV infection and cell-cell spread in human skin in vivo but suggest that it is the formation of complexes between ORF47 and IE62 proteins, both VZV tegument components, that constitutes the essential contribution of ORF47 protein to VZV replication in vivo.

Varicella-zoster virus (VZV) glycoprotein I is dispensable in cell culture but necessary for infection of human skin and T cells in SCIDhu mice in vivo. The gI promoter contains an activating upstream sequence that binds the cellular transactivators specificity factor 1 (Sp1) and upstream stimulatory factor (USF) and an open reading frame 29 (ORF29)-responsive element (29RE), which mediates enhancement by ORF29 DNA binding protein of immediate-early 62 (IE62)-induced transcription. Recombinants, rOKAgI-Sp1 and rOKAgI-USF, with two base pair substitutions in Sp1 or USF sites, replicated like rOKA in vitro, but infectivity of rOKAgI-Sp1 was significantly impaired in skin and T cells in vivo. A double mutant, rOKAgI-Sp1/USF, did not replicate in skin but yielded low titers of infectious virus in T cells. The repaired protein, rOKAgI:rep-Sp1/USF, was as infectious as rOKA. Thus, disrupting gI promoter sites for cellular transactivators altered VZV virulence in vivo, with variable consequences related to the cellular factor and the host cell type. Mutations in the 29RE of the gI promoter were made by substituting each of four 10-bp blocks in this region with a 10-bp sequence, GATAACTACA, that was predicted to interfere with enhancer effects of the ORF29 protein. One of these mutants, which was designated rOKAgI-29RE-3, had diminished replication in skin and T cells, indicating that ORF29 protein-mediated enhancement of gI expression contributes to VZV virulence. Mutations within promoters of viral genes that are nonessential in vitro should allow construction of recombinant herpesviruses that have altered virulence in specific host cells in vivo and may be useful for designing herpesviral gene therapy vectors and attenuated viral vaccines.

Varicella-zoster virus (VZV) open reading frame 63 (ORF63), located between nucleotides 110581 and 111417 in the internal repeat region, encodes a nuclear phosphoprotein which is homologous to herpes simplex virus type 1 (HSV-1) ICP22 and is duplicated in the terminal repeat region as ORF70 (nucleotides 118480 to 119316). We evaluated the role of ORFs 63 and 70 in VZV replication, using recombinant VZV cosmids and PCR-based mutagenesis to make single and dual deletions of these ORFs. VZV was recovered within 8 to 10 days when cosmids with single deletions were transfected into melanoma cells along with the three intact VZV cosmids. In contrast, VZV was not detected in transfections carried out with a dual deletion cosmid. Infectious virus was recovered when ORF63 was cloned into a nonnative Avr II site in this cosmid, confirming that failure to generate virus was due to the dual ORF63/70 deletion and that replication required at least one gene copy. This requirement may be related to our observation that ORF63 interacts directly with ORF62, the major immediate-early transactivating protein of VZV. ORF64 is located within the inverted repeat region between nucleotides 111565 and 112107; it has some homology to the HSV-1 Us10 gene and is duplicated as ORF69 (nucleotides 117790 to 118332). ORF64 and ORF69 were deleted individually or simultaneously using the VZV cosmid system. Single deletions of ORF64 or ORF69 yielded viral plaques with the same kinetics and morphology as viruses generated with the parental cosmids. The dual deletion of ORF64 and ORF69 was associated with an abnormal plaque phenotype characterized by very large, multinucleated syncytia. Finally, all of the deletion mutants that yielded recombinants retained infectivity for human T cells in vitro and replicated efficiently in human skin in the SCIDhu mouse model of VZV pathogenesis.

The SCID-hu mouse implanted with human fetal tissue is a novel model for investigating human viral pathogenesis. Infection of human skin implants was used to investigate the basis for the clinical attenuation of the varicella-zoster virus (VZV) strain, V-Oka, from which the newly licensed vaccine is made. The pathogenicity of V-Oka was compared with that of its parent, P-Oka, another low-passage clinical isolate, strain Schenke (VZV-S), and VZV-Ellen, a standard laboratory strain. The role of glycoprotein C (gC) in infectivity for human skin was assessed by using gC-negative mutants of V-Oka and VZV-Ellen. Whereas all of these VZV strains replicated well in tissue culture, only low-passage clinical isolates were fully virulent in skin, as shown by infectious virus yields and analysis of implant tissues for VZV DNA and viral protein synthesis. The infectivity of V-Oka in skin was impaired compared to that of P-Oka, providing the first evidence of a virologic basis for the clinical attenuation of V-Oka. The infectivity of V-Oka was further diminished in the absence of gC expression. All strains except gC-Ellen retained some capacity to replicate in human skin, but cell-free virus was recovered only from implants infected with P-Oka or VZV-S. Although VZV is closely related to herpes simplex virus type 1 (HSV-1) genetically, experiments in the SCID-hu model revealed differences in tropism for human cells that correlated with differences in VZV and HSV-1 disease. VZV caused extensive infection of epidermal and dermal skin cells, while HSV-1 produced small, superficial lesions restricted to the epidermis. As in VZV, gC expression was a determinant for viral replication in skin. VZV infects human CD4 + and CD8 + T cells in thymus/liver implants, but HSV-1 was detected only in epithelial cells, with no evidence of lymphotropism. These SCID-hu mouse experiments show that the clinical attenuation of the varicella vaccine can be attributed to decreased replication of V-Oka in skin and that tissue culture passage alone reduces the ability of VZV to infect human skin in vivo. Furthermore, gC, which is dispensable for replication in tissue culture, plays a critical role in the virulence of the human alphaherpesviruses VZV and HSV-1 for human skin.

Glycoprotein B (gB), the most conserved protein in the family Herpesviridae , is essential for the fusion of viral and cellular membranes. Information about varicella-zoster virus (VZV) gB is limited, but homology modeling showed that the structure of VZV gB was similar to that of herpes simplex virus (HSV) gB, including the putative fusion loops. In contrast to HSV gB, VZV gB had a furin recognition motif ([R]-X-[KR]-R-|-X, where | indicates the position at which the polypeptide is cleaved) at residues 491 to 494, thought to be required for gB cleavage into two polypeptides. To investigate their contribution, the putative primary fusion loop or the furin recognition motif was mutated in expression constructs and in the context of the VZV genome. Substitutions in the primary loop, W180G and Y185G, plus the deletion mutation Δ 491 RSRR 494 and point mutation 491 GSGG 494 in the furin recognition motif did not affect gB expression or cellular localization in transfected cells. Infectious VZV was recovered from parental Oka (pOka)-bacterial artificial chromosomes that had either the Δ 491 RSRR 494 or 491 GSGG 494 mutation but not the point mutations W180G and Y185G, demonstrating that residues in the primary loop of gB were essential but gB cleavage was not required for VZV replication in vitro. Virion morphology, protein localization, plaque size, and replication were unaffected for the pOka-gBΔ 491 RSRR 494 or pOka-gB 491 GSGG 494 virus compared to pOka in vitro. However, deletion of the furin recognition motif caused attenuation of VZV replication in human skin xenografts in vivo. This is the first evidence that cleavage of a herpesvirus fusion protein contributes to viral pathogenesis in vivo, as seen for fusion proteins in other virus families.

Several functions have been attributed to the serine/threonine protein kinase encoded by open reading frame 66 (ORF66) of varicella-zoster virus (VZV), including modulation of the apoptosis and interferon pathways, down-regulation of major histocompatibility complex class I cell surface expression, and regulation of IE62 localization. The amino acid sequence of the ORF66 protein contains a recognizable conserved kinase domain. Point mutations were introduced into conserved protein kinase motifs to evaluate their importance to ORF66 protein functions. Two substitution mutants were generated, including a G102A substitution, which blocked autophosphorylation and altered IE62 localization, and an S250P substitution, which had no effect on either autophosphorylation or IE62 localization. Both kinase domain mutants grew to titers equivalent to recombinant parent Oka (pOka) in vitro. pOka66G102A had slightly reduced growth in skin, which was comparable to the reduction observed when ORF66 translation was prevented by stop codon insertions in pOka66S. In contrast, infection of T-cell xenografts with pOka66G102A was associated with a significant decrease in infectious virus production equivalent to the impaired T-cell tropism found with pOka66S infection of T-cell xenografts in vivo. Disrupting kinase activity with the G102A mutation did not alter IE62 cytoplasmic localization in VZV-infected T cells, suggesting that decreased T-cell tropism is due to other ORF66 protein functions. The G102A mutation reduced the antiapoptotic effects of VZV infection of T cells. These experiments indicate that the T-cell tropism of VZV depends upon intact ORF66 protein kinase function.

The open reading frame 10 (ORF10) of varicella-zoster virus (VZV) encodes a tegument protein that enhances transactivation of VZV genes and has homology to herpes simplex virus type 1 (HSV-1) VP16. While VP16 is essential for HSV replication, ORF10 is dispensable for vaccine OKA (VOKA) growth in vitro. We used parent OKA (POKA) cosmids to delete ORF10, producing POKAΔ10; point mutations that disrupted the acidic activation domain and the putative motif for binding human cellular factor 1 (HCF-1) in ORF10 protein yielded POKA10-Phe28Ala, POKA10-Phe28Ser, and POKA10-mHCF viruses. Deleting ORF10 or mutating these two functional domains had no effect on VZV replication, immediate-early gene transcription, or virion assembly in vitro. However, deleting ORF10 reduced viral titers and the extent of cutaneous lesions significantly in SCIDhu skin xenografts in vivo compared to POKA. Epidermal cells infected with POKAΔ10 had significantly fewer DNA-containing nucleocapsids and complete virions compared to POKA; extensive aggregates of intracytoplasmic viral particles were also observed. Altering the activation or the putative HCF-1 domains of ORF10 protein had no consequences for VZV replication in vivo. Thus, the decreased pathogenic potential of POKAΔ10 in skin could not be attributed to absence of these ORF10 protein functions. In contrast to skin cells, deleting ORF10 did not impair VZV T-cell tropism in vivo, as assessed by infectious virus yields. We conclude that ORF10 protein is required for efficient VZV virion assembly and is a specific determinant of VZV virulence in epidermal and dermal cells in vivo.

Although genes related to varicella-zoster virus (VZV) open reading frame 35 (ORF35) are conserved in the herpesviruses, information about their contributions to viral replication and pathogenesis is limited. Using a VZV cosmid system, we deleted ORF35 to produce two null mutants, designated rOkaΔ35(#1) and rOkaΔ35(#2), and replaced ORF35 at a nonnative site, generating two rOkaΔ35/35@Avr mutants. ORF35 Flag-tagged recombinants were made by inserting ORF35-Flag at the nonnative Avr site as the only copy of ORF35, yielding rOkaΔ35/35Flag@Avr, or as a second copy, yielding rOka35Flag@Avr. Replication of rOkaΔ35 viruses was diminished in melanoma and Vero cells in a 6-day analysis of growth kinetics. Plaque sizes of rOkaΔ35 mutants were significantly smaller than those of rOka in melanoma cells. Infection of melanoma cells with rOkaΔ35 mutants was associated with disrupted cell fusion and polykaryocyte formation. The small plaque phenotype was not corrected by growth of rOkaΔ35 mutants in melanoma cells expressing the major VZV glycoprotein E, gE. The rOkaΔ35/35@Avr viruses displayed growth kinetics and plaque morphologies that were indistinguishable from those of rOka. Analysis with ORF35-Flag recombinants showed that the ORF35 gene product localized predominantly to the nuclei of infected cells. Evaluations in the SCIDhu mouse model demonstrated that ORF35 was required for efficient VZV infection of skin and T-cell xenografts, although the decrease in infectivity was most significant in skin. These mutagenesis experiments indicated that ORF35 was dispensable for VZV replication, but deleting ORF35 diminished growth in cultured cells and was associated with attenuated VZV infection of differentiated human skin and T cells in vivo.

The protein product of varicella-zoster virus (VZV) ORF47 is a serine/threonine protein kinase and tegument component. Evaluation of two recombinants of the Oka strain, rOka47ΔC, with a C-terminal truncation of ORF47, and rOka47D-N, with a point mutation in the conserved kinase motif, showed that ORF47 kinase function was necessary for optimal VZV replication in human skin xenografts in SCID mice but not in cultured cells. We now demonstrate that rOka47ΔC and rOka47D-N mutants do not infect human T-cell xenografts. Differences in the growth of kinase-defective ORF47 mutants allowed an examination of requirements for VZV pathogenesis in skin and T cells in vivo. Although virion assembly was reduced and no virion transport to cell surfaces was observed, epidermal cell fusion persisted, and VZV polykaryocytes were generated by rOka47ΔC and rOka47D-N in skin. Virion assembly was also impaired in vitro, but VZV-induced cell fusion continued to cause syncytia in cultured cells infected with rOka47ΔC or rOka47D-N. Intracellular trafficking of envelope glycoprotein E and the ORF47 and IE62 proteins, components of the tegument, was aberrant without ORF47 kinase activity. In summary, normal VZV virion assembly appears to require ORF47 kinase function. Cell fusion was induced by ORF47 mutants in skin, and cell-cell spread occurred even though virion formation was deficient. VZV-infected T cells do not undergo cell fusion, and impaired virion assembly by ORF47 mutants was associated with a complete elimination of T-cell infectivity. These observations suggest a differential requirement for cell fusion and virion formation in the pathogenesis of VZV infection in skin and T cells.

Varicella-zoster virus (VZV) glycoprotein E (gE) is essential for VZV replication. To further analyze the functions of gE in VZV replication, a full deletion and point mutations were made in the 62-amino-acid (aa) C-terminal domain. Targeted mutations were introduced in YAGL (aa 582 to 585), which mediates gE endocytosis, AYRV (aa 568 to 571), which targets gE to the trans -Golgi network (TGN), and SSTT, an “acid cluster” comprising a phosphorylation motif (aa 588 to 601). Substitutions Y582G in YAGL, Y569A in AYRV, and S593A, S595A, T596A, and T598A in SSTT were introduced into the viral genome by using VZV cosmids. These experiments demonstrated a hierarchy in the contributions of these C-terminal motifs to VZV replication and virulence. Deletion of the gE C terminus and mutation of YAGL were lethal for VZV replication in vitro. Mutations of AYRV and SSTT were compatible with recovery of VZV, but the AYRV mutation resulted in rapid virus spread in vitro and the SSTT mutation resulted in higher virus titers than were observed for the parental rOka strain. When the rOka-gE-AYRV and rOka-gE-SSTT mutants were evaluated in skin and T-cell xenografts in SCIDhu mice, interference with TGN targeting was associated with substantial attenuation, especially in skin, whereas the SSTT mutation did not alter VZV infectivity in vivo. These results provide the first information about how targeted mutations of this essential VZV glycoprotein affect viral replication in vitro and VZV virulence in dermal and epidermal cells and T cells within intact tissue microenvironments in vivo.