A varicella-zoster virus mutant impaired for latency in rodents, but not impaired for replication in cell culture

Abstract

While trying to generate a site-directed deletion in the ORF63 latency-associated gene of varicella-zoster virus (VZV) Oka, we constructed a virus with an unexpected rearrangement. The virus has a small deletion in both copies of ORF63 and two copies of a cassette inserted between ORFs 64/65 and 68/69 containing (a) truncated ORF62, (b) ORF63 with a small deletion, and (c) full-length ORF64. The virus was not impaired for growth in human cells, induced higher levels of neutralizing antibodies in guinea pigs, and was impaired for latency in cotton rats compared with parental virus (p=0.0022). Additional mutants containing the same truncation in ORF62, with or without the ORF63 deletion, were less impaired for latency. A VZV Oka mutant, replicating to similar titers and inducing a comparable immune response as parental virus, but impaired for latency, might serve as a safer vaccine and be less likely to reactivate to cause zoster.

Fig. 1

Structures of VZV ROka, ROka-NLS, ROka63-AatII-62t, and ROka-62t. Compared with VZV ROka, ROka-NLS has deletions in both copies of ORF63 (ORF63 and ORF70) and insertion of a cassette containing a truncated form of ORF62, a deleted version of ORF63, and full-length ORF64 between ORF 64 and ORF65. ROka-62t has the truncated form of ORF62 present in ROka-NLS inserted between ORF65 and ORF66. ROka63-AatII-62t has the truncated form of ORF62 present in ROka-NLS inserted between ORF65 and ORF66 and the deletion in both copies of ORF63 that are present in ROka-NLS.

Fig. 2

Southern blots of VZV ROka, ROka-NLS, ROka63-AatII-62t, and ROka-62t. Virion DNAs were cut with Not I (A) or BamHI (B), separated on agarose gels, transferred to nylon membranes, and hybridized with probes to ORF63 (A) or to ORF62 (near the 5′ end, corresponding to amino acids 140-360).

Fig. 3

Growth of VZV ROka, ROka-NLS, ROka63-AatII-62t, and ROka-62t in cell culture. Melanoma cells were infected with each virus and at days 1 to 5 after infection the cells were treated with trypsin and the titer of cell-associated virus was determined on melanoma cells. Each data point shown is the average of two separate values.

Fig. 4

VZV ROka-NLS, ROka63-AatII-62t, and ROka-62t each express a truncated ORF62 transcript at similar levels. (A) Total RNA was isolated from melanoma cells infected with each virus, separated on an agarose gel, transferred to a nylon membrane, and hybridized with a radiolabeled probe to ORF62. The experiment was performed four times and a representative result is shown. (B) The ratio of the intensity of the ORF62 full-length transcript (4.2 kb band) to the ORF62 truncated transcript (2.2 kb band) is plotted for each of viruses. The results shown were derived from four separate experiments and the vertical lines show standard deviations.

Fig. 5

Expression of IE62 is similar in cells infected with VZV ROka, ROka-NLS, ROka63-AatII-62t, and ROka-62t by immunoprecipitation or immunoblotting. Immunoprecipitation of cell lysates infected with the viruses probed with antibody to 782 amino acids of IE62 (A) or to full-length IE62 (B). Immunoblot of cells infected with the viruses probed with antibody to 782 amino acids of IE62 (C) or to full-length IE62 (D).

Fig. 6

Estimated copy number of VZV genomes in latently infected cotton rats from experiments 1 (A) and 2 (B) in Table 1. The geometric mean number of VZV genome copy numbers per 500 ng of ganglia DNA in PCR-positive ganglia is shown at the bottom of the figure. Open circles represent samples whose copy numbers were below the limit of detection (<10 copies per 500 ng of DNA), and filled circles show the viral copy number for samples above the limit of detection.