Previous Article | Next Article ![]()
Journal of Virology, February 2007, p. 1586-1591, Vol. 81, No. 4
0022-538X/07/$08.00+0 doi:10.1128/JVI.01220-06
Medical Virology Section, Laboratory of Clinical Infectious Diseases, National Institutes of Health, Bethesda, Maryland
Received 11 June 2006/ Accepted 1 October 2006
|
|
|---|
|
|
|---|
ORF29 encodes a 130-kDa protein that binds to single-stranded DNA and localizes predominantly to the nucleus of virus-infected cells in vitro (24). ORF29 contains an atypical nuclear localization signal within amino acids 9 to 154, and transport to the nucleus requires Ran and karyopherins (37). While ORF29 protein is present in the nucleus of lytically infected cells, the protein is in the cytoplasm of neurons from human ganglia (17, 25). ORF29 protein localizes to the cytoplasm of guinea pig enteric ganglia neurons (5) and in an astrocyte-like cell line (38). Treatment with a proteosome inhibitor or expression of herpes simplex virus type 1 (HSV-1) ICP0 or VZV ORF61 results in translocation of ORF29 protein to the nucleus in both guinea pig enteric ganglia neurons and the astrocyte-like cell line.
ORF29 protein is secreted from VZV-infected fibroblasts and is endocytosed by human neurons in vitro (1). The protein is present in endothelial and epithelial cells in the skin of patients with varicella and zoster; the protein is also located in nerves in the dermis of patients with varicella. ORF29 protein is not present in virions (23).
ORF29 is expressed from a bidirectional promoter that it shares with ORF28 (27, 28, 40). Expression is activated during lytic infection by ORF62 protein and the USF transcription factor. During lytic infection, ORF29 transcripts of 4.1 and 4.2 kb are expressed, while during latency only the 4.2-kb transcript is present (26). Expression of ORF29 in T cells enhances the ability of ORF62 protein to transactivate the gI promoter; however, the effect of ORF29 is more modest in fibroblasts and melanoma cells, and expression of ORF29 in neuronal cells inhibits the ability of ORF62 protein to activate the gI promoter (2, 18). ORF29 protein also augments the ability of ORF62 to activate the ORF28 and ORF29 promoters; however, ORF29 protein does not activate transcription of these promoters directly (10).
Here we describe an ORF29 deletion mutant and repaired virus and characterize their growth in vitro as well as their ability to establish latency in rodents.
|
|
|---|
Plasmids and cosmids. Plasmid pCI-29 was constructed by performing PCR on VZV cosmid MstII B with primers GCCTAGCTAGCCAAAATGGAAAATACTCAGAAGACTGTG and GTCAGAATGCGGCCGCGGGAGGTTAAATCATTTCCATTG that amplify the ORF29 open reading frame, cutting the PCR product with NheI and NotI, and inserting the fragment into the corresponding sites of pCI (Promega, Madison, WI). Plasmid pAc-CMV (41) contains the human cytomegalovirus (CMV) immediate-early (IE) promoter inserted into the XhoI-EcoRI site of pAcSG2 (PharMingen). Plasmids pAc-CMV29StuI and pAc-CMV29EcoRV were constructed to produce baculoviruses expressing ORF29. Plasmid pCI-29 was cut with NheI, blunted with the Klenow fragment of Escherichia coli DNA polymerase, and cut with BamHI, and the fragment containing ORF29 and the simian virus 40 (SV40) polyadenylation sequence was inserted into the StuI-BglII site of pAc-CMV to create plasmid pAc-CMV29StuI. This plasmid is predicted to express ORF29 from both the baculovirus polyhedron promoter and the human immediate-early (IE) CMV promoter. Plasmid pCI-29 was cut with BglII, blunted with Klenow, and cut with BamHI, and the fragment containing ORF29 driven by the human CMV IE promoter and followed by the SV40 polyadenylation sequence was inserted into the EcoRV-BglII site of pAc-CMV to create plasmid pAc-CMV29EcoRV. This plasmid is predicted to express ORF29 from only the human IE CMV promoter.
VZV cosmids NotI A, NotI B, MstII A, and MstII B encompass the VZV genome (Fig. 1). VZV ORF29, encoded by nucleotides 50857 to 54468 of the VZV genome, is predicted to express a protein of 1,204 amino acids (14). To construct a virus deleted for ORF29, VZV cosmid MstII B was partially digested with HpaII using the RecA-assisted restriction endonuclease cleavage procedure (15). Two single-stranded oligonucleotides, CGGGGCCCCTGGGTTACGTTTATGCGTGCCGGGTTGAAGATTTGGATCTGGAGGAAATTT and GGCGCTTCTTGAAAAAACGGAAAACTTACCGGAATTATGGACTACGGCTTTTACTTCAAC, centered around HpaII sites at nucleotides 50919 and 53725 in the VZV genome were annealed to cosmid MstII B using the E. coli RecA protein. Additional HpaII sites in the cosmid were methylated using HpaII methylase and S-adenosylmethinone, and the reaction was heated to 65°C to remove the oligonucleotide-RecA complexes. The DNA was precipitated and cut with HpaII, and the large fragment, which lacks most of the ORF29 gene, was ligated to itself and was inserted into E. coli to produce cosmid VZV MstII B-29D (Fig. 1).
![]() View larger version (15K): [in a new window] |
FIG. 1. Construction of recombinant VZV deleted for ORF29 and an ORF29-repaired virus. The VZV genome (line 1) consists of unique long (UL), unique short (US), terminal repeat (TR), and internal repeat (IR) regions (line 2). Cosmids NotI A and NotI B (line 3), MstII A, and MstII B (line 4) encompass the VZV genome. Cosmid MstII B-29D is deleted for most of ORF29 (line 5). Cosmid MstII A-29 has a cassette with ORF29 driven by the human CMV promoter inserted into the AvrII site of the cosmid (line 6). Numbers indicate nucleotide positions based on the sequence of VZV strain Dumas (14).
|
Transfections, Southern blotting, immunoblotting, and virus growth studies. VZV cosmids were linearized with NotI or Bsu36I and transfected along with plasmid pCMV62 into human melanoma cells using the calcium phosphate procedure. Cells were passaged each week by treatment with trypsin, and cytopathic effects were noted.
Virion DNA was isolated from nucleocapsids, digested with restriction enzymes, fractionated on 1% agarose gels, transferred to nylon membranes, and probed with a radiolabeled fragment containing ORF29.
Lysates of baculovirus or VZV-infected cells were fractionated on sodium dodecyl sulfate-polyacrylamide gel electrophoresis gels, transferred to nylon membranes, and incubated with rabbit antibody to VZV ORF29 protein (24), thymidine kinase (a gift from Christine Talarico), IE4 (29), IE63 (31), IE62 (23), or mouse monoclonal antibody to glycoprotein E (gE) (Chemicon, Temecula, Calif.). The blots were incubated with horseradish peroxidase-conjugated anti-rabbit or anti-mouse antibodies and developed with enhanced chemiluminescence (Pierce Chemical Company, Rockford, Ill.). Intensity of bands was quantified by densitometry.
Duplicate flasks of melanoma cells were infected with 200 PFU of VZV recombinants, and on days 1 to 5 after infection the cells were treated with trypsin and serial dilutions were titrated on melanoma cells. VZV deleted for ORF29 was titrated on melanoma cells that had been infected with Baculo 29 at a multiplicity of infection of 25 the day before. One week after infection, the cells were fixed and stained with crystal violet and plaques were counted. The mean number of plaques obtained from duplicate wells was determined.
Animal experiments. Four- to 6-week-old female cotton rats were inoculated intramuscularly along the sides of the spine with virus-infected melanoma cells containing 1.75 x 105 PFU of recombinant VZV. For analysis of acute infection, animals were sacrificed 3 days after infection; for latent infection, animals were sacrificed 5 to 6 weeks after infection. Dorsal root ganglia from the left thoracic and lumbar spine were pooled, DNA was isolated, and PCR was performed using 500 ng of ganglia DNA from infected animals or serial dilutions of cosmid NotI A in 500 ng of ganglia DNA from uninfected animals (to generate a standard curve) and primers corresponding to ORF21 (3). The PCR products were fractionated by electrophoresis on agarose gels, transferred to nylon membranes, and probed with a radiolabeled ORF21 probe, and copy numbers were determined using a PhosphorImager. The lower limit of reliable detection was 10 copies per 500 ng of ganglia DNA. PCR was also performed using 500 ng of ganglia DNA and ORF29 primers CATTTGACCCTGCCAACAAC and TAGTGCGTGCTCCAGAAACC (the latter sequence is located within the region absent from the ORF29 deletion mutant). Southern blotting was performed, and the membrane was hybridized to a radiolabeled ORF29 probe.
RNA from dorsal root ganglia was isolated using Trizol (Invitrogen, Carlsbad, Calif.), treated with DNase I, and heated to inactivate DNase, and cDNA was prepared using oligo(dT)12-18 and reverse transcriptase. PCR was performed using ORF63 primers (35), and Southern blotting of the amplified DNA was performed using a radiolabeled ORF63 probe.
|
|
|---|
To complement a VZV ORF29 deletion mutant, we produced baculovirus expressing ORF29. Infection of Sf9 insect cells with Baculo 29 followed by immunoblotting with antibody to ORF29 protein yielded a 130-kDa band (Fig. 2, lane 3). A similar-sized band was not detected in cells infected with control baculovirus AcNPV. Infection of melanoma cells with Baculo 29 or control baculovirus failed to show a band corresponding to ORF29 protein; however, infection of the cells with VZV ROka showed a band of 130 kDa (Fig. 2, lane 1). Sodium butyrate is a histone deacetylase inhibitor that enhances expression of foreign genes in mammalian cells when expressed by baculovirus (11). Therefore, we treated baculovirus-infected melanoma cells with 5 mM sodium butyrate 1 day before preparing lysates of infected cells. Immunoblotting of Baculo 29-infected cells treated with sodium butyrate showed a band of 130 kDa (Fig. 2, lane 5); no band was detected in cells infected with control baculovirus that had been treated with the chemical.
![]() View larger version (59K): [in a new window] |
FIG. 2. Expression of ORF29 by recombinant baculovirus. Sf9 cells were infected with control baculovirus (Baculo; lane 2) or baculovirus expressing ORF29 (Baculo 29; lane 3). Melanoma cells were infected with VZV ROka (lane 1), Baculo 29 in the absence (lane 4) or presence (lane 5) of sodium butyrate (buty), or control baculovirus in the absence (lane 6) or presence (lane 7) of sodium butyrate. A prominent band of 110 kDa is present in Sf9 cells infected with Baculo 29. This band was not reduced in the presence of the proteosome inhibitor MG132 (J. I. Cohen and M. A. Ali, unpublished data) and could be due to protein degradation or initiation of translation at an internal site in insect cells. The numbers correspond to the sizes of the proteins in kilodaltons.
|
To verify that the deletion in ORF29 did not significantly affect expression of the genes adjacent to ORF29, we constructed cosmid MstII A-29, which contains the ORF29 gene driven by the human CMV promoter. Transfection of cells with cosmids NotI A, NotI B, MstII A-29, and MstII B-29D yielded infectious virus 7 days after transfection. This virus was termed ROka29DR.
To verify that VZV ROka29D and ROka29DR had the expected genomic structures, Southern blotting was performed. Virion DNA was digested with EcoRI and PacI and hybridized with a radiolabeled probe to ORF29. Virion DNA from cells infected with VZV ROka showed a band of 6.5 kbp, while cells infected with ROka29D had a band of 3.7 kbp due to the 2.8-kbp deletion in ORF29 (Fig. 3). Virion DNA from cells infected with VZV ROka29DR had the 2.8-kbp band due to the deletion in ORF29 and a new band of 22 kbp due to the insertion of ORF29 into the genome between ORFs 65 and 66.
![]() View larger version (31K): [in a new window] |
FIG. 3. Southern blot of virion DNA from cells infected with ORF29 mutants. Virion DNA from cells infected with VZV ROka, ROka29D, or ROka29DR was digested with EcoRI and PacI and hybridized with a radiolabeled ORF29 DNA probe. Numbers indicate the sizes of DNAs in kilobase pairs.
|
![]() View larger version (39K): [in a new window] |
FIG. 4. Immunoblot of lysates from cells infected with ORF29 mutants blotted with antibody to ORF29 protein, IE62, IE63, IE4, VZV thymidine kinase (TK), or gE. Lysates were obtained from cells infected with ROka29D in the presence of Baculo 29 (ROka29D), ROka, ROka29DR, and ROka29D after one passage in cells without Baculo 29 (ROka29D P1) or were not infected with any virus. The small amount of ORF29 protein in cells infected with ROka29D that has been passaged once in melanoma cells likely reflects some residual ORF29 protein in the inoculum. Equivalent amounts of lysates were loaded in each lane and in each panel.
|
Growth of VZV ORF29 deletion and repaired virus in cell culture. To study the growth of the ORF29 mutants in cell culture, melanoma cells were infected with the viruses and titers were measured for five consecutive days. VZV deleted for ORF29 was unable to grow in melanoma cells (Fig. 5). VZV ROka29DR, in which ORF29 was driven by the human CMV promoter at a non-native site in the virus genome, grew slower than ROka but eventually reached a peak titer that was nearly equivalent to that of ROka.
![]() View larger version (21K): [in a new window] |
FIG. 5. Growth of ORF29 mutants in melanoma cells. VZV ROka, ROka29D, and ROka29DR were grown in melanoma cells, and at various times the cells were treated with trypsin and virus titers were determined.
|
VZV ORF29 is critical for latent infection. To determine if VZV ORF29 is required for establishment of latent infection, cotton rats were inoculated with ROka29D, ROka29DR, or ROka, and 5 to 6 weeks later the animals were sacrificed, DNA was isolated from dorsal root ganglia, and PCR was performed with primers for ORF21 followed by Southern blotting. In the first experiment, 1 of 10 animals infected with VZV ROka29D, 3 of 11 animals infected with ROka 29DR, and 6 of 10 animals infected with ROka had viral DNA in ganglia (Fig. 6A). In the second experiment, none of 10 animals infected with ROka29D, 4 of 11 animals infected with ROka29DR, and 11 of 11 animals infected with ROka had VZV DNA in their ganglia (Fig. 6B). Taken together, 5% (1 of 20) of animals infected with ROka29D, 32% (7 of 22) of animals infected with ROka29DR, and 81% (17 of 21) of animals infected with ROka were latently infected. When the results of the two experiments were pooled, the differences between animals infected with ROka29D and ROka (P < 0.00001) and ROka29DR and ROka (P = 0.0044) were statistically significant, while the difference between animals infected with ROka29D and ROka29DR was barely significant (P = 0.045).
![]() View larger version (12K): [in a new window] |
FIG. 6. Copy number of VZV genomes in animals latently infected with VZV ROka, ROka29D, or ROka29DR in experiments 1 (A) and 2 (B). The lower limit of detection of viral DNA is 10 copies, and the geometric mean copy number per 500 ng of DNA for the PCR-positive ganglia is shown at the bottom.
|
![]() View larger version (18K): [in a new window] |
FIG. 7. Southern blot of cDNA corresponding to ORF63 RNA from animals latently infected with ROka29D, ROka29DR, or ROka. RNA was isolated from dorsal root ganglia of infected animals, cDNA was prepared, PCR was performed with primers to ORF63, and the blot was hybridized with a radiolabeled ORF63 probe. Cosmid MstII A, which encodes ORF63, is a positive control.
|
![]() View larger version (32K): [in a new window] |
FIG. 8. Southern blot for ORF29 in ganglia of animals latently infected with ROka29D or ROka. Numbers correspond to the size of DNA in kilobase pairs.
|
|
|
|---|
Cells infected with VZV deleted for ORF29 expressed similar levels of IE or putative early proteins as cells infected with parental virus, but the level of gE was reduced. Previous studies using transient transfections showed that ORF29 protein has no significant effect on the ability of IE62 to activate the ORF20 or ORF21 promoter and only a modest effect on the ORF28 promoter (10). While transient transfection using a plasmid expressing ORF29 protein alone did not upregulate expression from the gI promoter, ORF29 enhanced the ability of ORF62 protein to transactivate the gI promoter in transfections (18). In contrast, our studies show that in the context of the virus, ORF29 protein is important for expression of a late gene, gE.
HSV ICP8 is the homolog of VZV ORF29. Both proteins bind to single-stranded DNA (24, 33) and localize to punctate regions within the nucleus (4, 10), and ICP8, like ORF29, is essential for virus replication (16). Metabolic labeling studies with an HSV-2 ICP8 replication-defective mutant virus showed that proteins of all kinetic classes were expressed at levels similar to or slightly less than those of parental virus; gB and gD were expressed at lower levels than wild-type virus on immunoblot (12). Surprisingly, we found that overexpression of ORF29 protein during virus infection also resulted in reduced expression of gE. Thus, the level of ORF29 protein must be properly regulated for optimal late gene expression but not necessarily for immediate-early or early gene expression.
ORF29 is one of six proteins that are expressed during latency in human sensory or cranial nerve ganglia. Previously we showed that ORFs 21 and 66 are not required for establishment of latency (36, 39), while ORFs 4 and 63 (6, 7) have a critical role in latency in rodents. Here we show that while ORF29 is not required for VZV to enter ganglia, ORF29 is important for efficient establishment of latency in rodents. Similar studies with HSV-2 showed that an ICP8 null mutant was markedly impaired for latency in mice (19). Overexpression of ORF29 protein, as exemplified by the ROka29DR mutant, was also associated with a significant impairment of VZV latency in rodents. ORF29 protein is present in the nucleus of lytically infected cells but in the cytoplasm of human neurons during latency (17, 25). Interestingly, when an astrocytoma-derived cell line is infected with adenovirus which expresses ORF29, the protein is expressed in the cytoplasm; however, when these cells are treated with a proteosome inhibitor, the half-life of ORF29 protein is increased and the protein migrates to the nucleus (38). Thus, it is possible that overexpression of ORF29 protein in ROka29DR-infected neurons could result in both cytoplasmic and nuclear expression of the protein in the cells and thereby impair latency.
It is important to note that the cotton rat model does not reproduce all of the features of latency seen in humans. While the same VZV transcripts expressed during latency in humans are also expressed during latency in rodents experimentally infected with VZV (34), VZV reactivation from latency has not been described in rodents either in vivo or by cocultivation of ganglia on cell cultures in vitro. It is possible that reactivation occurs in rodents, but it has not been detected, since the animals are also asymptomatic during acute infection with VZV. It is important to note that VZV reactivation has also not been demonstrated by explant cocultivation of human ganglia. Thus, it will be important to study the effects of mutants in viruses similar to VZV, such as simian varicella virus.
VZV mutants of ORF29 might serve as useful vaccine candidates. Inoculation of mice with the HSV-1 d301 ICP8 deletion mutant virus induces HSV-specific T-cell proliferation and protects animals from lethal infection with wild-type virus (30, 32). Similarly, inoculation of animals with an HSV-2 IPC8 null mutant reduces acute and latent infection with a challenge virus and protects the animals from death by a challenge virus (13, 19). A VZV ORF29 deletion mutant might be useful as a replication-defective VZV vaccine, provided that a mutant is made that cannot recombine with the complementing cell line. Alternatively, ROka29DR, which overexpresses the ORF29 protein and is also impaired for latency, might serve as a vaccine candidate. This virus is impaired for latency but has the advantage that none of the viral proteins are deleted and all can be presented to the immune system, albeit at higher or lower levels than with wild-type virus.
We thank Paul Kinchington for antibodies to ORF29, IE62, and IE63 proteins, Christine Talarico for antibody to VZV thymidine kinase, Bernard Roizman for plasmid pAc-CMV, and Jing Qin for help with statistics.
Published ahead of print on 6 December 2006. ![]()
|
|
|---|
This article has been cited by other articles:
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Copyright © 2009 by the American Society for Microbiology. For an alternate route to Journals.ASM.org, visit: http://intl-journals.asm.org | More Info»