Array Analysis of Simian Varicella Virus Gene Tran
http://www.100md.com
病菌学杂志 2005年第9期
Departments of Neurology
Microbiology, University of Colorado Health Sciences Center, Denver, Colorado 80262
ABSTRACT
Simian varicella virus (SVV) is a neurotropic alphaherpesvirus of monkeys that is a model for varicella pathogenesis and latency. Like human varicella-zoster virus (VZV), SVV causes chicken pox (varicella), becomes latent in ganglia along the entire neuraxis, and reactivates to produce shingles (zoster). We developed macroarrays to determine the extent of viral transcription from all 70 predicted SVV open reading frames (ORFs) in infected cells in tissue culture. Cloned fragments (200 to 400 bp) from the 5' and 3' ends of each ORF were PCR amplified, quantitated, spotted onto nylon membranes, and fixed by UV cross-linking. Using a cDNA probe prepared from poly(A)+ RNA extracted from SVV-infected Vero cells at the height of the cytopathic effect (3 days after infection) and chemiluminescence for detection, transcripts corresponding to all SVV ORFs were identified. The abundance of each SVV transcript was compared with that previously demonstrated for VZV in infected tissue culture cells.
INTRODUCTION
Natural infection of humans with varicella-zoster virus (VZV) or monkeys with simian varicella virus (SVV) causes chicken pox (varicella) in their natural hosts. Both viruses spontaneously reactivate years later to produce zoster (shingles). Like VZV, SVV becomes latent in cranial nerve and dorsal root ganglia along the entire neuraxis exclusively in ganglionic neurons (5). The mechanisms of varicella reactivation are not known, although in humans the incidence of zoster correlates with a decline in cell-mediated immunity to VZV during aging and immunosuppression. The cascade of events leading to varicella reactivation cannot be determined in living humans, but it is possible to study ganglia from latently infected monkeys.
Transcriptional analysis applied to ganglia will provide valuable information about SVV gene expression during latency but must first be standardized and quantified in productively infected cells. In tissue culture, SVV and VZV are highly cell associated and do not grow to high titers, and synchronous infection is not possible. Nevertheless, even with unsynchronized infection, a uniform cytopathic effect can readily be demonstrated 72 h after cocultivation of uninfected cells with VZV-infected cells in tissue culture. Our earlier studies which used macroarrays to study VZV gene expression in tissue culture (3) revealed that the optimal time for analysis was at the height of the cytopathic effect (3 days after infection). Thus, we focused our efforts on this single time point and conducted triplicate independent analyses with SVV. SVV macroarrays were constructed, and chemiluminescence was used to detect and quantitate viral transcription from every SVV open reading frame (ORF) in SVV-infected cells in tissue culture.
MATERIALS AND METHODS
Virus and cells. SVV was propagated by cocultivation of infected and uninfected Vero (African green monkey kidney) cells. SVV-infected cells were scraped, washed, and centrifuged at 1,000 x g for 5 min. Cell pellets were immediately frozen in liquid nitrogen and stored at –80°C.
DNA extraction and labeling. SVV nucleocapsids were prepared and DNA was extracted as described previously (1). Virus DNA was digested with restriction enzymes BamHI, BglII, EcoRI, and NcoI. The integrity of SVV DNA was determined by agarose gel electrophoresis. Restriction enzyme-digested SVV DNA (1 μg in 16 μl of double-distilled water) was labeled with digoxigenin using the DIG High Prime DNA labeling and detection starter kit II (Roche Applied Science, Mannheim, Germany).
RNA extraction and PCR. Total RNA was extracted from SVV-infected cells using the RNeasy Midi kit (QIAGEN, Valencia, Calif.). Poly(A)+ SVV mRNA was purified using a mRNA purification kit (Amersham Biosciences, Buckinghamshire, England), treated with 1 U/μg of RQ1 RNase-free DNase (Promega, Madison, Wis.) at 37°C for 30 min, and determined to be DNA free by PCR. All PCRs were performed as described previously (6).
Reverse transcription and cDNA labeling. Poly(A)+ SVV mRNA (2 μg) was mixed with 2 μg of oligo(dT) and 0.3 μg of random primers (Invitrogen, Carlsbad, Calif.), and the mixture (39.6 μl) was heated to 65°C for 5 min. The reaction temperature was decreased to 43°C over 10 min, after which 12 μl of 5x avian myeloblastosis virus buffer (Promega) and 1.4 μl of avian myeloblastosis virus reverse transcriptase (high concentration) (600 U) (Promega), 6 μl of PCR nucleotide mix (Roche Applied Science), and 1 μl of 10-mg/ml bovine serum albumin were added. After incubation at 43°C for 130 min, the mixture was heated to 95°C for 5 min. Four tubes containing 2 μg each of SVV mRNA in 60 μl were reverse transcribed to yield a total of 8 μg of SVV cDNA/RNA hybrid. The SVV cDNA/RNA hybrid was treated with 1 μl each of RNase H (1.5 U/μl) (Promega) and RNase-ONE RNase (10 U/μl) (Promega) at 65°C for 30 min to digest the RNA strand, extracted with phenol-chloroform, and alcohol precipitated. Single-stranded SVV cDNA was labeled with digoxigenin using the DIG High Prime DNA labeling and detection starter kit II (Roche). Unincorporated nucleotides were removed by phenol and chloroform extraction and alcohol precipitation.
Cloning of SVV DNA fragments. SVV DNA fragments (200 to 600 bp) from the 5' and 3' ends of each ORF were PCR amplified with forward primers (5'-TTTTCCTTTAGCGGCCGC-SVV DNA-3' [NotI]) and reverse primers (5'-AGGTTCAATTGGAGCTC-SVV DNA-3' [SstI]). A 284-bp DNA fragment was also amplified from pGEM3zf– using forward primers (5'-TTTTCCTTTAGCGGCCGCGGCGCTTTCTCATAGCTCAC-3' [NotI]) and reverse primers (5'-AGGTTCAATTGGAGCTCCGTCTCGCGTCTATGGTTT-3' [SstI]). Table 1 lists the primer sequences and their location on the SVV genome (4) of oligonucleotide primers for all SVV ORFs, as well as the G+C content of each amplified segment. Computer analysis of DNA sequences was performed using DNAMax (MiraiBio, Inc., Alameda, Calif.). PCR products were digested with NotI and SstI and inserted directionally in the multiple cloning sites of pGEM11zF (Promega). The concentrations of all recombinant plasmids were determined by absorbance at 260 nm and diluted to 30 ng/μl. The cloned SVV or pGEM3zf–-specific inserts were amplified using vector-specific primers (GEMF [5'-CCCAGTCACGACGTTGTAAA-SVV DNA-3'] and GEMR [5'-TCACACAGGAAACAGCTATG-SVV DNA-3']). The closed segment of pGEM3zf– and no DNA were used as negative controls. Actin, a positive control for cellular transcription, was amplified as described previously (2).
Array construction. SVV DNA fragments and positive and negative controls were amplified, quantitated, spotted (40 ng/4 μl) onto a 200-cm2 neutral BioBond nylon membrane (Sigma-Aldrich, St. Louis, Mo.) and fixed by UV cross-linking twice at 125 mJ using the GS gene linker (Bio-Rad, Hercules, Calif.).
Hybridization and detection. The UV-fixed target SVV DNA fragments were prehybridized in a hybridization oven (Boekel Scientific, Feasterville, Pa.) for 3 h at 42°C in 35 ml of Digoxigenin Easy hybridization solution (Roche Applied Science) in glass cylinders (35 by 300 mm) (VWR Scientific Products, Brisbane, Calif.). Digoxigenin-labeled DNA (2 μg) or cDNA (8 μg) was denatured at 95°C for 10 min and quenched on ice for 5 min. The prehybridization solution was replaced with 20 ml of fresh hybridization solution containing probe and hybridized for 48 to 72 h at 42°C. The nylon membrane was washed in 35 ml of 0.1x SSC (1x SSC is 0.15 M NaCl plus 0.015 M sodium citrate) and 0.1% sodium dodecyl sulfate at room temperature (three times for 5 min each and twice for 15 min each) and then in 90 ml of washing buffer (100 mM maleic acid, 150 mM NaCl) for 5 min at room temperature. The membrane was then placed in 30 ml of blocking solution (3 ml of 10x blocking solution [DIG High Prime DNA labeling and detection starter kit II; Roche Applied Science] with 27 ml of maleic acid buffer [0.1 M maleic acid, 0.15 M sodium chloride]) for 2 h at room temperature. The membranes were treated using one of the following two methods. (i) The membranes were incubated in 28 ml of alkaline phosphatase-conjugated antidigoxigenin Fab fragments (2.8 ml of blocking solution, 25.2 ml of maleic acid buffer, 1.4 μl of antidigoxigenin antibodies [1:20,000]; Roche Applied Science) for 1 h at room temperature, washed at room temperature with 140 ml of blocking solution (14 ml of 10x blocking solution with 126 ml of maleic acid buffer) three times for 8 min each time, and washed at room temperature with 140 ml of washing buffer (99.62 mM Tris-HCl, 99.25 mM NaCl, pH 9.5) twice for 8 min. (ii) The membranes were incubated with 40 ml of peroxidase-conjugated antidigoxigenin poly-Fab fragments (4 ml of blocking solution, 36 ml of 100 mM Tris-HCl, pH 7.5, 150 mM NaCl, 120 μl of antidigoxigenin antibodies [1:333]; Roche Applied Science) for 2 h at room temperature and washed by the washing protocol described above. Chemiluminescence detection was performed twice using two different preparations of mRNA with the CDP-Star detection reagent (New England BioLabs, Beverly, Mass.) and once using another independent preparation of mRNA with the ECL Western Blotting Detection Reagents and Analysis System (Amersham Bioscience, Piscataway, N.J.). Hybridization signals were detected using Kodak Biomax Light film. Uninfected Vero cells were treated by identical protocols.
Data analysis. Desktop optical scanning was used to digitize each radiogram. Individual ORF intensities were quantitated with Quantity One densitometry software (Bio-Rad, Hercules, Calif.). Optical density (OD) for each SVV target was used to calculate the relative expression (RE) of each SVV ORF according to the following formula:
where REi is the relative expression of the ith SVV ORF, ODi is the OD of the ith SVV ORF, no DNA is the average OD for the no-DNA targets, and act is the average OD for the actin targets. The denominator standardizes each radiogram for variations in background intensity or specific activity of the probe. To allow comparison of SVV transcription data with those published for VZV, the average relative SVV ORF expression (REi) obtained from all individual arrays was expressed as a percentage of the most abundant SVV ORF.
RESULTS
Macroarrays containing PCR fragments representing the entire SVV genome were constructed and probed with labeled SVV DNA or cDNA fragments. Figure 1 illustrates the specificity of the arrays examined with DNA probes as well as cDNA probes prepared from uninfected and SVV-infected cells. DNA probes from uninfected cells detected only actin (Fig. 1A), while DNA probes prepared from SVV-infected cells detected all 129 SVV DNA targets as well as actin (Fig. 1B). The signal intensities of actin after hybridization of DNA from both infected and uninfected cells were similar (Fig. 1A and B). cDNA probes from uninfected cells detected only actin (Fig. 1C), while cDNA probes from SVV-infected cells detected each of the 70 viral ORFs and actin (Fig. 1D). None of the probes hybridized to the negative controls (pGEM3zf– and no DNA). The signal intensities of actin after hybridization of cDNA from both infected and uninfected cells to actin were similar (Fig. 1C and D). Table 2 lists every ORF, region, location (row and column) on the array, relative expression of transcription, standard deviation for each SVV DNA target, and relative order of abundance 3 days after infection. For example, the 3' end of ORF 9 is located in row 2 and column 7 on the array and is the most abundant transcript.
Figure 2 graphically displays the average spot intensity and standard deviation of each ORF in SVV-infected cells compared to those of a cellular transcript (actin). The values are the averages of three independent experiments in which the spot intensity for each SVV ORF was first divided by the product of the no-DNA and actin spot intensities and then normalized to the spot intensities obtained for the no-DNA and actin targets on three individual control arrays. All signal intensities of SVV ORFs are >1 standard deviation above the negative controls. Not unexpectedly, some variation in virus transcription was seen in the three samples of independently obtained RNA from SVV-infected cells (error bars in Fig. 2), but the relative expression of individual SVV genes was not affected (Table 2). Table 3 lists the average signal intensity of each SVV ORF 3 days after infection and their predicted gene function (4). The most abundant SVV transcript detected during productive infection is ORF 9 (tegument protein). Figure 3 graphically shows a comparison for the transcriptional abundance of each ORF for both SVV and VZV in cells 3 days after infection.
DISCUSSION
This study is the first to use macroarrays with chemiluminescence detection to study varicella virus gene expression during lytic infection in culture. Chemiluminescent probes are safer than radiolabeled probes for the investigator and the environment and can be stored for prolonged periods. Moreover, hybridization signals require shorter exposure times, and as little as 0.1 pg (700 genome equivalents) of labeled SVV DNA can be detected (data not shown). However, chemiluminescence applied to the study of cDNA can be capricious, and it often took multiple experiments to yield quantifiable results.
Digoxigenin-labeled DNA from SVV-infected cells hybridized to all array targets. Similarly, digoxigenin-labeled cDNA from SVV-infected cells hybridized to all array targets but with various signal intensities. Several factors, including the abundance of mRNA, RNA stability, and the efficiency of the reverse transcription reaction, may have influenced our observations. The longer or shorter half-life of SVV transcripts is probably compensated for by the decrease or increase in their abundance. A similar pattern of global transcription for the two varicella viruses was seen (Fig. 3). The four most abundant transcripts in SVV (i.e., ORFs 9, 32, 63, and 23) were found to be greater than the relative expression of actin. The levels of transcription for two of these (e.g., ORF 9 and 63) correlated well with that previously reported by array analysis for VZV (3). As in VZV, the most abundant ORF found in SVV during lytic infection was ORF 9. ORF 9 is predicted to encode a tegument protein. The herpes simplex virus type 1 homolog of varicella virus ORF 9, VP22 protein (herpes simplex virus type 1 UL49) has been shown to be one of the four proteins responsible for mediating capsid binding to the nuclear pore complex (7). Therefore, the ORF 9 protein in the tegument of SVV may be necessary for cell-to-cell infection. VZV ORF 63 is an immediate-early gene. SVV ORFs 32 and 23 were found to have a greater transcriptional abundance during the height of the cytopathic effect, while these ORFs in VZV were not as abundantly transcribed. VZV ORF 32 is predicted to encode a phosphoprotein and ORF 23 a capsid protein. Both of these VZV ORFs have less than 50% homology to SVV, which may explain the variations in transcriptional abundance. It is also possible that the stability of the mRNA transcribed from these VZV ORFs may be more stable than their SVV homologs. The SVV ORFs 62 to 64 and 69 to 71 map within the inverted repeat segment of the virus genome. By design, the array targets cannot differentiate between transcripts originating from either of the diploid genes. Therefore, ORF 62 to 64 expression levels determined by array analysis may be overrepresented by twofold (the difference attributed to ORF 69 to 71 transcription, respectively). The implication is that the promoter activity for these three diploid SVV genes may be lower than shown on Table 2. However, this report describes the steady-state levels of all SVV genes transcribed and not the specific promoter activities.
Overall, transcription from every SVV ORF could be identified in lytically infected cells using array technology and chemiluminescence detection.
ACKNOWLEDGMENTS
This work was supported in part by Public Health Service grants AG 06127 and NS 32623 from the National Institutes of Health. Steven Deitch is supported by Public Health Service grant NS 07321 from the National Institutes of Health.
We thank Marina Hoffman for editorial assistance and Cathy Allen for manuscript preparation.
REFERENCES
Clarke, P., S. D. Rabkin, M. V. Inman, R. Mahalingam, R. Cohrs, M. Wellish, and D. H. Gilden. 1992. Molecular analysis of simian varicella virus DNA. Virology 190:597-605.
Cohrs, R. J., J. Randall, J. Smith, D. H. Gilden, C. Dabrowski, H. van der Keyl, and R. Tal-Singer. 2000. Analysis of individual human trigeminal ganglia for latent herpes simplex virus type 1 and varicella-zoster virus nucleic acids using real-time PCR. J. Virol. 74:11464-11471.
Cohrs, R. J., M. P. Hurley, and D. H. Gilden. 2003. Array analysis of viral gene transcription during lytic infection of cells in tissue culture with varicella-zoster virus. J. Virol. 77:11718-11732.
Gray, W. L., B. Starnes, M. W. White, and R. Mahalingam. 2001. The DNA sequence of the simian varicella virus genome. Virology 284:123-130.
Kennedy, P. G. E., E. Grinfeld, V. Traina-Dorge, D. H. Gilden, and R. Mahalingam. 2004. Neuronal localization of simian varicella virus DNA in ganglia of naturally infected African green monkeys. Virus Genes 28:273-276.
Mahalingam, R., M. Wellish, W. Wolf, A. N. Dueland, R. Cohrs, A. Vafai, and D. Gilden. 1990. Latent varicella zoster viral DNA in human trigeminal and thoracic ganglia. N. Engl. J. Med. 323:627-631.
Ojala, P. M., B. Sodeik, M. W. Ebersold, U. Kutay, and A. Helenius. 2000. Herpes simplex virus type 1 entry into host cells: reconstitution of capsid binding and uncoating at the nuclear pore complex in vitro. Mol. Cell. Biol. 20:4922-4931.(Steven B. Deitch, Donald )
Microbiology, University of Colorado Health Sciences Center, Denver, Colorado 80262
ABSTRACT
Simian varicella virus (SVV) is a neurotropic alphaherpesvirus of monkeys that is a model for varicella pathogenesis and latency. Like human varicella-zoster virus (VZV), SVV causes chicken pox (varicella), becomes latent in ganglia along the entire neuraxis, and reactivates to produce shingles (zoster). We developed macroarrays to determine the extent of viral transcription from all 70 predicted SVV open reading frames (ORFs) in infected cells in tissue culture. Cloned fragments (200 to 400 bp) from the 5' and 3' ends of each ORF were PCR amplified, quantitated, spotted onto nylon membranes, and fixed by UV cross-linking. Using a cDNA probe prepared from poly(A)+ RNA extracted from SVV-infected Vero cells at the height of the cytopathic effect (3 days after infection) and chemiluminescence for detection, transcripts corresponding to all SVV ORFs were identified. The abundance of each SVV transcript was compared with that previously demonstrated for VZV in infected tissue culture cells.
INTRODUCTION
Natural infection of humans with varicella-zoster virus (VZV) or monkeys with simian varicella virus (SVV) causes chicken pox (varicella) in their natural hosts. Both viruses spontaneously reactivate years later to produce zoster (shingles). Like VZV, SVV becomes latent in cranial nerve and dorsal root ganglia along the entire neuraxis exclusively in ganglionic neurons (5). The mechanisms of varicella reactivation are not known, although in humans the incidence of zoster correlates with a decline in cell-mediated immunity to VZV during aging and immunosuppression. The cascade of events leading to varicella reactivation cannot be determined in living humans, but it is possible to study ganglia from latently infected monkeys.
Transcriptional analysis applied to ganglia will provide valuable information about SVV gene expression during latency but must first be standardized and quantified in productively infected cells. In tissue culture, SVV and VZV are highly cell associated and do not grow to high titers, and synchronous infection is not possible. Nevertheless, even with unsynchronized infection, a uniform cytopathic effect can readily be demonstrated 72 h after cocultivation of uninfected cells with VZV-infected cells in tissue culture. Our earlier studies which used macroarrays to study VZV gene expression in tissue culture (3) revealed that the optimal time for analysis was at the height of the cytopathic effect (3 days after infection). Thus, we focused our efforts on this single time point and conducted triplicate independent analyses with SVV. SVV macroarrays were constructed, and chemiluminescence was used to detect and quantitate viral transcription from every SVV open reading frame (ORF) in SVV-infected cells in tissue culture.
MATERIALS AND METHODS
Virus and cells. SVV was propagated by cocultivation of infected and uninfected Vero (African green monkey kidney) cells. SVV-infected cells were scraped, washed, and centrifuged at 1,000 x g for 5 min. Cell pellets were immediately frozen in liquid nitrogen and stored at –80°C.
DNA extraction and labeling. SVV nucleocapsids were prepared and DNA was extracted as described previously (1). Virus DNA was digested with restriction enzymes BamHI, BglII, EcoRI, and NcoI. The integrity of SVV DNA was determined by agarose gel electrophoresis. Restriction enzyme-digested SVV DNA (1 μg in 16 μl of double-distilled water) was labeled with digoxigenin using the DIG High Prime DNA labeling and detection starter kit II (Roche Applied Science, Mannheim, Germany).
RNA extraction and PCR. Total RNA was extracted from SVV-infected cells using the RNeasy Midi kit (QIAGEN, Valencia, Calif.). Poly(A)+ SVV mRNA was purified using a mRNA purification kit (Amersham Biosciences, Buckinghamshire, England), treated with 1 U/μg of RQ1 RNase-free DNase (Promega, Madison, Wis.) at 37°C for 30 min, and determined to be DNA free by PCR. All PCRs were performed as described previously (6).
Reverse transcription and cDNA labeling. Poly(A)+ SVV mRNA (2 μg) was mixed with 2 μg of oligo(dT) and 0.3 μg of random primers (Invitrogen, Carlsbad, Calif.), and the mixture (39.6 μl) was heated to 65°C for 5 min. The reaction temperature was decreased to 43°C over 10 min, after which 12 μl of 5x avian myeloblastosis virus buffer (Promega) and 1.4 μl of avian myeloblastosis virus reverse transcriptase (high concentration) (600 U) (Promega), 6 μl of PCR nucleotide mix (Roche Applied Science), and 1 μl of 10-mg/ml bovine serum albumin were added. After incubation at 43°C for 130 min, the mixture was heated to 95°C for 5 min. Four tubes containing 2 μg each of SVV mRNA in 60 μl were reverse transcribed to yield a total of 8 μg of SVV cDNA/RNA hybrid. The SVV cDNA/RNA hybrid was treated with 1 μl each of RNase H (1.5 U/μl) (Promega) and RNase-ONE RNase (10 U/μl) (Promega) at 65°C for 30 min to digest the RNA strand, extracted with phenol-chloroform, and alcohol precipitated. Single-stranded SVV cDNA was labeled with digoxigenin using the DIG High Prime DNA labeling and detection starter kit II (Roche). Unincorporated nucleotides were removed by phenol and chloroform extraction and alcohol precipitation.
Cloning of SVV DNA fragments. SVV DNA fragments (200 to 600 bp) from the 5' and 3' ends of each ORF were PCR amplified with forward primers (5'-TTTTCCTTTAGCGGCCGC-SVV DNA-3' [NotI]) and reverse primers (5'-AGGTTCAATTGGAGCTC-SVV DNA-3' [SstI]). A 284-bp DNA fragment was also amplified from pGEM3zf– using forward primers (5'-TTTTCCTTTAGCGGCCGCGGCGCTTTCTCATAGCTCAC-3' [NotI]) and reverse primers (5'-AGGTTCAATTGGAGCTCCGTCTCGCGTCTATGGTTT-3' [SstI]). Table 1 lists the primer sequences and their location on the SVV genome (4) of oligonucleotide primers for all SVV ORFs, as well as the G+C content of each amplified segment. Computer analysis of DNA sequences was performed using DNAMax (MiraiBio, Inc., Alameda, Calif.). PCR products were digested with NotI and SstI and inserted directionally in the multiple cloning sites of pGEM11zF (Promega). The concentrations of all recombinant plasmids were determined by absorbance at 260 nm and diluted to 30 ng/μl. The cloned SVV or pGEM3zf–-specific inserts were amplified using vector-specific primers (GEMF [5'-CCCAGTCACGACGTTGTAAA-SVV DNA-3'] and GEMR [5'-TCACACAGGAAACAGCTATG-SVV DNA-3']). The closed segment of pGEM3zf– and no DNA were used as negative controls. Actin, a positive control for cellular transcription, was amplified as described previously (2).
Array construction. SVV DNA fragments and positive and negative controls were amplified, quantitated, spotted (40 ng/4 μl) onto a 200-cm2 neutral BioBond nylon membrane (Sigma-Aldrich, St. Louis, Mo.) and fixed by UV cross-linking twice at 125 mJ using the GS gene linker (Bio-Rad, Hercules, Calif.).
Hybridization and detection. The UV-fixed target SVV DNA fragments were prehybridized in a hybridization oven (Boekel Scientific, Feasterville, Pa.) for 3 h at 42°C in 35 ml of Digoxigenin Easy hybridization solution (Roche Applied Science) in glass cylinders (35 by 300 mm) (VWR Scientific Products, Brisbane, Calif.). Digoxigenin-labeled DNA (2 μg) or cDNA (8 μg) was denatured at 95°C for 10 min and quenched on ice for 5 min. The prehybridization solution was replaced with 20 ml of fresh hybridization solution containing probe and hybridized for 48 to 72 h at 42°C. The nylon membrane was washed in 35 ml of 0.1x SSC (1x SSC is 0.15 M NaCl plus 0.015 M sodium citrate) and 0.1% sodium dodecyl sulfate at room temperature (three times for 5 min each and twice for 15 min each) and then in 90 ml of washing buffer (100 mM maleic acid, 150 mM NaCl) for 5 min at room temperature. The membrane was then placed in 30 ml of blocking solution (3 ml of 10x blocking solution [DIG High Prime DNA labeling and detection starter kit II; Roche Applied Science] with 27 ml of maleic acid buffer [0.1 M maleic acid, 0.15 M sodium chloride]) for 2 h at room temperature. The membranes were treated using one of the following two methods. (i) The membranes were incubated in 28 ml of alkaline phosphatase-conjugated antidigoxigenin Fab fragments (2.8 ml of blocking solution, 25.2 ml of maleic acid buffer, 1.4 μl of antidigoxigenin antibodies [1:20,000]; Roche Applied Science) for 1 h at room temperature, washed at room temperature with 140 ml of blocking solution (14 ml of 10x blocking solution with 126 ml of maleic acid buffer) three times for 8 min each time, and washed at room temperature with 140 ml of washing buffer (99.62 mM Tris-HCl, 99.25 mM NaCl, pH 9.5) twice for 8 min. (ii) The membranes were incubated with 40 ml of peroxidase-conjugated antidigoxigenin poly-Fab fragments (4 ml of blocking solution, 36 ml of 100 mM Tris-HCl, pH 7.5, 150 mM NaCl, 120 μl of antidigoxigenin antibodies [1:333]; Roche Applied Science) for 2 h at room temperature and washed by the washing protocol described above. Chemiluminescence detection was performed twice using two different preparations of mRNA with the CDP-Star detection reagent (New England BioLabs, Beverly, Mass.) and once using another independent preparation of mRNA with the ECL Western Blotting Detection Reagents and Analysis System (Amersham Bioscience, Piscataway, N.J.). Hybridization signals were detected using Kodak Biomax Light film. Uninfected Vero cells were treated by identical protocols.
Data analysis. Desktop optical scanning was used to digitize each radiogram. Individual ORF intensities were quantitated with Quantity One densitometry software (Bio-Rad, Hercules, Calif.). Optical density (OD) for each SVV target was used to calculate the relative expression (RE) of each SVV ORF according to the following formula:
where REi is the relative expression of the ith SVV ORF, ODi is the OD of the ith SVV ORF, no DNA is the average OD for the no-DNA targets, and act is the average OD for the actin targets. The denominator standardizes each radiogram for variations in background intensity or specific activity of the probe. To allow comparison of SVV transcription data with those published for VZV, the average relative SVV ORF expression (REi) obtained from all individual arrays was expressed as a percentage of the most abundant SVV ORF.
RESULTS
Macroarrays containing PCR fragments representing the entire SVV genome were constructed and probed with labeled SVV DNA or cDNA fragments. Figure 1 illustrates the specificity of the arrays examined with DNA probes as well as cDNA probes prepared from uninfected and SVV-infected cells. DNA probes from uninfected cells detected only actin (Fig. 1A), while DNA probes prepared from SVV-infected cells detected all 129 SVV DNA targets as well as actin (Fig. 1B). The signal intensities of actin after hybridization of DNA from both infected and uninfected cells were similar (Fig. 1A and B). cDNA probes from uninfected cells detected only actin (Fig. 1C), while cDNA probes from SVV-infected cells detected each of the 70 viral ORFs and actin (Fig. 1D). None of the probes hybridized to the negative controls (pGEM3zf– and no DNA). The signal intensities of actin after hybridization of cDNA from both infected and uninfected cells to actin were similar (Fig. 1C and D). Table 2 lists every ORF, region, location (row and column) on the array, relative expression of transcription, standard deviation for each SVV DNA target, and relative order of abundance 3 days after infection. For example, the 3' end of ORF 9 is located in row 2 and column 7 on the array and is the most abundant transcript.
Figure 2 graphically displays the average spot intensity and standard deviation of each ORF in SVV-infected cells compared to those of a cellular transcript (actin). The values are the averages of three independent experiments in which the spot intensity for each SVV ORF was first divided by the product of the no-DNA and actin spot intensities and then normalized to the spot intensities obtained for the no-DNA and actin targets on three individual control arrays. All signal intensities of SVV ORFs are >1 standard deviation above the negative controls. Not unexpectedly, some variation in virus transcription was seen in the three samples of independently obtained RNA from SVV-infected cells (error bars in Fig. 2), but the relative expression of individual SVV genes was not affected (Table 2). Table 3 lists the average signal intensity of each SVV ORF 3 days after infection and their predicted gene function (4). The most abundant SVV transcript detected during productive infection is ORF 9 (tegument protein). Figure 3 graphically shows a comparison for the transcriptional abundance of each ORF for both SVV and VZV in cells 3 days after infection.
DISCUSSION
This study is the first to use macroarrays with chemiluminescence detection to study varicella virus gene expression during lytic infection in culture. Chemiluminescent probes are safer than radiolabeled probes for the investigator and the environment and can be stored for prolonged periods. Moreover, hybridization signals require shorter exposure times, and as little as 0.1 pg (700 genome equivalents) of labeled SVV DNA can be detected (data not shown). However, chemiluminescence applied to the study of cDNA can be capricious, and it often took multiple experiments to yield quantifiable results.
Digoxigenin-labeled DNA from SVV-infected cells hybridized to all array targets. Similarly, digoxigenin-labeled cDNA from SVV-infected cells hybridized to all array targets but with various signal intensities. Several factors, including the abundance of mRNA, RNA stability, and the efficiency of the reverse transcription reaction, may have influenced our observations. The longer or shorter half-life of SVV transcripts is probably compensated for by the decrease or increase in their abundance. A similar pattern of global transcription for the two varicella viruses was seen (Fig. 3). The four most abundant transcripts in SVV (i.e., ORFs 9, 32, 63, and 23) were found to be greater than the relative expression of actin. The levels of transcription for two of these (e.g., ORF 9 and 63) correlated well with that previously reported by array analysis for VZV (3). As in VZV, the most abundant ORF found in SVV during lytic infection was ORF 9. ORF 9 is predicted to encode a tegument protein. The herpes simplex virus type 1 homolog of varicella virus ORF 9, VP22 protein (herpes simplex virus type 1 UL49) has been shown to be one of the four proteins responsible for mediating capsid binding to the nuclear pore complex (7). Therefore, the ORF 9 protein in the tegument of SVV may be necessary for cell-to-cell infection. VZV ORF 63 is an immediate-early gene. SVV ORFs 32 and 23 were found to have a greater transcriptional abundance during the height of the cytopathic effect, while these ORFs in VZV were not as abundantly transcribed. VZV ORF 32 is predicted to encode a phosphoprotein and ORF 23 a capsid protein. Both of these VZV ORFs have less than 50% homology to SVV, which may explain the variations in transcriptional abundance. It is also possible that the stability of the mRNA transcribed from these VZV ORFs may be more stable than their SVV homologs. The SVV ORFs 62 to 64 and 69 to 71 map within the inverted repeat segment of the virus genome. By design, the array targets cannot differentiate between transcripts originating from either of the diploid genes. Therefore, ORF 62 to 64 expression levels determined by array analysis may be overrepresented by twofold (the difference attributed to ORF 69 to 71 transcription, respectively). The implication is that the promoter activity for these three diploid SVV genes may be lower than shown on Table 2. However, this report describes the steady-state levels of all SVV genes transcribed and not the specific promoter activities.
Overall, transcription from every SVV ORF could be identified in lytically infected cells using array technology and chemiluminescence detection.
ACKNOWLEDGMENTS
This work was supported in part by Public Health Service grants AG 06127 and NS 32623 from the National Institutes of Health. Steven Deitch is supported by Public Health Service grant NS 07321 from the National Institutes of Health.
We thank Marina Hoffman for editorial assistance and Cathy Allen for manuscript preparation.
REFERENCES
Clarke, P., S. D. Rabkin, M. V. Inman, R. Mahalingam, R. Cohrs, M. Wellish, and D. H. Gilden. 1992. Molecular analysis of simian varicella virus DNA. Virology 190:597-605.
Cohrs, R. J., J. Randall, J. Smith, D. H. Gilden, C. Dabrowski, H. van der Keyl, and R. Tal-Singer. 2000. Analysis of individual human trigeminal ganglia for latent herpes simplex virus type 1 and varicella-zoster virus nucleic acids using real-time PCR. J. Virol. 74:11464-11471.
Cohrs, R. J., M. P. Hurley, and D. H. Gilden. 2003. Array analysis of viral gene transcription during lytic infection of cells in tissue culture with varicella-zoster virus. J. Virol. 77:11718-11732.
Gray, W. L., B. Starnes, M. W. White, and R. Mahalingam. 2001. The DNA sequence of the simian varicella virus genome. Virology 284:123-130.
Kennedy, P. G. E., E. Grinfeld, V. Traina-Dorge, D. H. Gilden, and R. Mahalingam. 2004. Neuronal localization of simian varicella virus DNA in ganglia of naturally infected African green monkeys. Virus Genes 28:273-276.
Mahalingam, R., M. Wellish, W. Wolf, A. N. Dueland, R. Cohrs, A. Vafai, and D. Gilden. 1990. Latent varicella zoster viral DNA in human trigeminal and thoracic ganglia. N. Engl. J. Med. 323:627-631.
Ojala, P. M., B. Sodeik, M. W. Ebersold, U. Kutay, and A. Helenius. 2000. Herpes simplex virus type 1 entry into host cells: reconstitution of capsid binding and uncoating at the nuclear pore complex in vitro. Mol. Cell. Biol. 20:4922-4931.(Steven B. Deitch, Donald )