Transcriptional reactivation of murine cytomegalovirus ie gene ...

7 downloads 0 Views 313KB Size Report
histone deacetylase inhibitor trichostatin A results in more rapid induction of ie gene expression and reactivation of virus. Despite these results, which suggest a ...
Journal of General Virology (2007), 88, 1097–1102

Short Communication

DOI 10.1099/vir.0.82696-0

Transcriptional reactivation of murine cytomegalovirus ie gene expression by 5-aza-29-deoxycytidine and trichostatin A in latently infected cells despite lack of methylation of the major immediate-early promoter Mary Hummel,1,2 Shixian Yan,1 Zhigao Li,1 Thomas K. Varghese3 and Michael Abecassis1,2 1

Correspondence

Transplant Division, Department of Surgery, Northwestern University, Feinberg School of Medicine, Chicago, IL, USA

Michael Abecassis [email protected]

2

Department of Microbiology and Immunology, Northwestern University, Feinberg School of Medicine, Chicago, IL, USA

3

Department of Surgery, University of Michigan Medical School, Ann Arbor, MI, USA

Received 8 November 2006 Accepted 29 November 2006

We have used a spleen explant model to investigate mechanisms of murine cytomegalovirus latency and reactivation. Induction of immediate-early (ie) gene expression occurs in explants after approximately 9 days in culture and virus reactivation follows induction of ie gene expression with kinetics similar to that of productive infection in vitro. This occurs independently of TNF receptor signalling. Treatment with the DNA methylation inhibitor 5-aza-29-deoxycytidine and the histone deacetylase inhibitor trichostatin A results in more rapid induction of ie gene expression and reactivation of virus. Despite these results, which suggest a role for DNA methylation in maintenance of viral latency, we find that the major immediate-early promoter/enhancer is not methylated in latently infected mice. Our results support the hypothesis that latency is maintained by epigenetic control of ie gene expression, and that induction of ie gene expression leads to reactivation of virus, but suggest that these are not controlled by DNA methylation.

Cytomegalovirus (CMV) is a ubiquitous herpesvirus which latently infects the majority of adults. Although the molecular nature of latent infection and the mechanisms by which reactivation occurs have been controversial, recent studies suggest that there is a true molecular latency in which genes associated with productive infection are transcriptionally silent and that reactivation is initiated by transcriptional activation of immediate-early (ie) genes (Hummel et al., 2001; Kurz et al., 1999; Reeves et al., 2005). ie gene expression is controlled by the major immediateearly promoter/enhancer (MIEP/E) (Dorsch-Hasler et al., 1985). Activation of transcription factors which bind to the MIEP/E (Mocarski & Courcell, 2001) is likely to play an important role in reactivation of ie gene expression. In addition, recent studies indicate that epigenetic factors which control access of the chromatin to transcription factors are also important in controlling CMV latency and reactivation (Meier, 2001; Reeves et al., 2005). We have used a spleen explant model for reactivation of virus (Jordan & Mar, 1982) to further investigate the regulation of murine CMV (MCMV) latency and reactivation. The CMV ie genes are the first set of genes expressed during productive infection and encode transcriptional 0008-2696 G 2007 SGM

Printed in Great Britain

transactivator proteins whose expression is required for progression to later stages of virus replication (Mocarski & Courcell, 2001). In spleens, of which approximately 90 % were negative for ie gene expression prior to explant, induction of ie gene expression generally did not occur until 9–10 days in culture (Fig. 1). Release of infectious virus in the medium was rarely detectable until approximately 12 days after explant (Fig. 2). CMV is a slowly replicating virus and plaques are detectable 3–5 days after infection of permissive fibroblasts in vitro. These studies suggest that induction of ie gene expression is the ratedetermining step in reactivation and that, once induction of ie gene expression has occurred, reactivation recapitulates productive infection in vitro. Our previous studies suggested that TNF plays an important role in inducing reactivation in vivo (Hummel et al., 2001). To determine whether TNF is required for reactivation ex vivo, we infected mice deficient in both TNF receptors (TNFR) (B6;129S-Tnfrsf1atm1ImxTnfrsf1btm1Imx mice, The Jackson Laboratory) with MCMV and tested the ability of the virus to reactivate from latency in explants. Because the TNFR-deficient mice were generated on a B6;129S background and B6 mice are resistant to MCMV infection 1097

M. Hummel and others

Fig. 1. Kinetics of induction of ie gene expression in explants of spleens from latently infected mice. (a) Spleens from BALB/c mice latently infected with MCMV (Smith strain) were removed and divided in half. One half was used as a latent control (lanes C); the other was co-cultured with murine embryo fibroblasts and harvested at various times after explant (lanes E), as indicated. RNAs were analysed for expression of IE-1+3 with primers from exons 2 and 3, for IE-3 with primers from exons 3 and 5, and for b-actin RNA as previously described (Hummel et al., 2001). Data shown are representative of analyses of two to ten mice per time point. (b) Analysis of MCMV DNA in latent mice. DNA was isolated from mice used in (a) and amplified with nested ie primers as previously described (Koffron et al., 1998) to verify that all mice were latently infected.

due to NK cell recognition of the m157 protein, we infected TNFR-deficient mice with the Dm157 strain of MCMV (Bubic et al., 2004). No differences were observed in the ability of this virus to replicate in wild-type (WT) or TNFR-deficient mice during acute infection (data not shown). Spleens from three WT or TNFR-deficient mice latently infected with Dm157 were explanted and monitored for the production of virus. No difference in the ability of MCMV to reactivate from explants of WT or TNFR-deficient mice was observed (data not shown). Recent studies with human CMV-infected cells indicate that changes in histone modifications are associated with reactivation from latency (Meier, 2001; Reeves et al., 2005). Activation of genes silenced through epigenetic mechanisms has been achieved by incubation with 5-aza-29deoxycytidine (5-aza-dC), an inhibitor of de novo DNA methylation, and/or with the histone deacetylase inhibitor trichostatin A (TSA) (Chen et al., 2001; El Kharroubi et al., 2001; Grassi et al., 2003; Magdinier & Wolffe, 2001; Masucci et al., 1989). We therefore investigated the effect of adding TSA, 5-aza-dC or both to cultures of spleen 1098

Fig. 2. (a) Induction of ie gene expression in spleen explants with 5-aza-dC and TSA. One half spleen from each latent mouse was snap-frozen for use as a pre-explant control (lanes C); the other half was explanted (lanes E) and incubated for 3 days in medium (mice 1–3), for 3 days in the presence of 300 nM TSA (mice 4–6), for 24 h in the presence of 1 mM 5-aza-dC and then for 48 h in 0.3 mM 5-aza-dC (mice 7–9), or in the presence of both 5-aza-dC and TSA (mice 10–12) as previously described (El Kharroubi et al., 2001). Explants were cultured without fibroblasts, since this was found to be unnecessary for reactivation. Control and explanted spleens are shown paired for each mouse. IE-1+3 RNAs were incubated with (+) or without (”) RT and cDNAs amplified as above. Amplification of the DNA is detectable in some samples, but is distinguishable from the RNA by size. cDNA products are detectable only in reactions with RT. (b) DNA isolated from mice used in (a) was analysed as above to verify that all mice were latently infected. I, RNA from infected fibroblasts; N, no template. (c) Plaque assay of virus released into medium from nine explants treated with or without 5-aza-dC and TSA. *, P,0.05. Journal of General Virology 88

http://vir.sgmjournals.org MCMV reactivation induction with TSA and 5-aza-dC

1099

Fig. 3. (a) Frequency of CpG dinucleotides in the MCMV genome. Numbers represent nucleotide coordinates in the published sequence (Rawlinson et al., 1996). (b) Frequency of CpG dinucleotides in the ie-1 gene and ratios of observed to expected CpGs (obs./exp.). The overall ratio of observed to expected CpGs in the MCMV genome is 1.22. Below, schematic of the ie-1 gene showing the location of the primers used for analysis of methylation by methylationsensitive restriction enzymes. Primer sequences and conditions available upon request. (c–e) Analysis of CpG methylation in the MIEP/E (c), first intron (d) and exons 2 and 3 (e) by restriction enzyme analysis. Locations of CpG dinucleotides (grey circles), CpGs analysed for methylation (black circles) and primers used for amplification are indicated, along with potential transcription factor-binding sites in the MIEP/E. DNAs were digested and then divided into separate PCR and amplified with primers specific for the region being tested for the presence of methylation or, as a positive control, from regions which do not contain restriction sites. Regions amplified: E, exons 2 and 3; P, promoter region; I, intron 1; A, Ant4 promoter. P+, I+, E+: Uncut DNA amplified from the promoter, intron 1, and exon 2 and 3 regions, respectively. Arrows indicate size of products expected from methylated MCMV DNA. M, 100 bp Molecular mass marker.

M. Hummel and others

explants from latently infected mice (Fig. 2a, lanes E). In this and additional experiments, induction of ie gene expression was observed in 1 of 21 explants cultured for 3 days without addition of drug and 1 of 6 explants cultured with TSA alone (x2 analysis, not significant). Incubation for 3 days in the presence of 5-aza-dC alone or the combination of 5-aza-dC and TSA induced ie gene expression in 2 of 6 and 12 of 15 of the cultures, respectively. The effect of 5-aza-dC alone is statistically significant (x2 analysis, P,0.05), and the effect of TSA and 5-aza-dC together is greater than that of 5-aza-dC alone (x2 analysis, P,0.05). These studies demonstrate that 5-aza-dC can induce ie gene expression and that 5-aza-dC and TSA act synergistically to enhance transcriptional reactivation of ie gene expression. The kinetics of virus production in medium from explants of nine mice treated with or without 5-aza-dC and TSA was analysed at 7, 10 and 12 days post-explant (Fig. 2c). Consistent with the observation that ie gene expression is not detectable until 9–10 days post-explant (Fig. 1), virus was rarely detectable prior to 10 days in untreated cultures, but increased exponentially starting at 12 days postexplant. Treatment of explants with 5-aza-dC and TSA caused a statistically significant increase in virus production over untreated controls at 7 and 10 days post-explant (t-test, P,0.05). Thus, treatment of explants with 5-aza-dC and TSA results in more rapid reactivation of both ie gene expression and virus production. Analysis of the sequence of CMV DNAs shows a marked suppression in the frequency of CpG dinucleotides in the ie region, suggesting the presence of DNA methylation and its associated transcriptional silencing (Herman & Baylin, 2003; Honess et al., 1989). In addition, the observation that treatment of explants with 5-aza-dC was sufficient to induce transcriptional reactivation of ie gene expression suggested that latency might be controlled in part through DNA methylation-induced silencing of ie gene expression. We therefore looked directly for evidence of methylation of ie-1 DNA. The greatest deficit in the ratio of observed to expected CpGs in MCMV DNA occurs within exon 4 of the ie-1 gene and in the enhancer region (Fig. 3a, b). Methylation of coding regions does not interfere with transcription (Jones, 1999). We therefore used the bisulfite modification method (Clark et al., 1994) to analyse methylation of CpG dinucleotides in the promoter/ enhancer region spanning nt –251 to +42 relative to the ie-1 transcription start site in DNAs from organs of nine latently infected mice (Fig. 3b). Bisulfite treatment converts unmethylated C to U, which are converted to T after PCR amplification. Methylated Cs are protected from conversion. Bisulfite-modified DNAs were amplified by strand-specific nested PCR (Herman et al., 1996), cloned and sequenced. Methylated CpGs remain as CpG, while unmethylated CpGs are converted to TpG. Because incomplete conversion of Cs will result in a false positive, conversion of Cs in non-CpGs is used as an internal control for completeness of the reaction. Only clones in 1100

which .95 % of the Cs in non-CpGs are converted were analysed. RNA was isolated from the same organs and analysed for expression of ie transcripts to ensure that the mice were negative for ie gene expression. In many cases, MCMV DNA could not be amplified after bisulfite treatment because the DNA copy number is very low in MCMV-latent mice, and bisulfite treatment results in substantial degradation of the DNA (Raizis et al., 1995). However, we were able to analyse 3–12 clones of 10 independent PCR from DNA isolated from various organs of nine mice. One hundred per cent of the C residues in CpG dinucleotides were converted to T residues in clones in which more than 95 % of the Cs in non-CpGs were converted to T. As positive controls, we analysed four of these DNAs for methylation of the Ant4 promoter, which is highly methylated in adult mouse organs (Rodic et al., 2005). As expected, CpGs in the Ant4 promoter were protected from conversion in bisulfite-modified DNA. These results indicate that CpG dinucleotides in the ie promoter/enhancer region are not methylated in latently infected mice. Since we were able to analyse only a small number of independent PCR, due to the technical limitations of the method, we also analysed DNA from the kidneys of latently infected mice by digestion with methylation-sensitive restriction enzymes prior to amplification by nested PCR. In this method, methylated DNA is protected from digestion and therefore can be amplified, while unmethylated DNA is digested and cannot be amplified. We analysed methylation of CpG residues in the BstUI site at positions 217 and 219 and in the HpaII, XhoI and AatII sites in the promoter region at positions 292, 2147 and 2229 relative to the transcription start site (nt 182912, 182914, 182987, 183042 and 183124 in the MCMV DNA sequence; Rawlinson et al., 1996) (Fig. 3c). Since partial digestion of the DNA will result in a false positive, and since the DNA must be amplified by nested PCR in order to detect low-copy-number MCMV DNA in latently infected tissues, it is essential to achieve complete digestion of the DNA. DNAs were therefore digested with a large excess of enzyme, precipitated and redigested prior to amplification. As a positive control to ensure that MCMV DNA could still be amplified after these manipulations, an aliquot of the digested DNA was amplified with primers from regions of MCMV DNA which did not contain these restriction sites (lanes E and I). Undigested DNA was also analysed in parallel as a positive control (lanes P+). In cases where these sites occurred in the Ant4 promoter (HpaII and BstUI), we also amplified DNA from this region as a positive control to demonstrate detection of methylated DNA (lanes A). Although the DNA could be amplified in positive controls, none of the DNAs subjected to digestion with these enzymes could be amplified with promoter-specific primers (lanes P). These data confirm the results of our bisulfite analysis, which demonstrate that MIEP/E DNA is not methylated in latently infected mice. Journal of General Virology 88

MCMV reactivation induction with TSA and 5-aza-dC

In addition, we investigated methylation of the intron between exons 1 and 2 and in the region encoding exons 2 and 3 by methylation-sensitive restriction enzyme analysis. No methylation of CpGs in the MluI, SacII or HhaI sites in intron 1 or the BsaAI, HincII or XhoI sites in exons 2 and 3 was observed (Fig. 3d, e; BsaA1 data not shown). Although this method allows analysis of a very limited set of CpG dinucleotides within the gene, our results suggest that methylation of other regions of the ie-1 gene is unlikely.

El Kharroubi, A., Piras, G. & Stewart, C. L. (2001). DNA demethyl-

In this study, we find that transcriptional reactivation of ie gene expression in explants is followed by reactivation of infectious virus with kinetics similar to that observed with infection in vitro. These results suggest that induction of ie gene expression is the rate-determining step in reactivation of the virus. In addition, we find that induction of ie gene expression in explants occurs much more rapidly in the presence of 5-aza-dC and TSA. Although the CpG deficit in the enhancer region and the observation that 5-aza-dC alone can induce ie gene expression would seem to suggest a role for DNA methylation in transcriptional silencing, the promoter/enhancer appears to be unmethylated. The lack of CpGs in the enhancer may be due to a selective pressure in favour of multiple AP-1- and NF-kB-binding sites, which lack this sequence (Fig. 3c). The effect of 5-aza-dC may be due to DNA methylation-independent effects on histone acetylation (Nguyen et al., 2002; Takebayashi et al., 2001). While we cannot exclude the possibility that the effects of 5-aza-dC and TSA in this system are unrelated to epigenetic control of ie gene expression, our studies are consistent with previous studies suggesting that latency is achieved through an active mechanism of transcriptional silencing involving histone modifications (Reeves et al., 2005).

Herman, J. G., Graff, J. R., Myohanen, S., Nelkin, B. D. & Baylin, S. B. (1996). Methylation-specific PCR: a novel PCR assay for methylation

ation reactivates a subset of imprinted genes in uniparental mouse embryonic fibroblasts. J Biol Chem 276, 8674–8680. Grassi, G., Maccaroni, P., Meyer, R., Kaiser, H., D’Ambrosio, E., Pascale, E., Grassi, M., Kuhn, A., Di Nardo, P. & other authors (2003).

Inhibitors of DNA methylation and histone deacetylation activate cytomegalovirus promoter-controlled reporter gene expression in human glioblastoma cell line U87. Carcinogenesis 24, 1625–1635. Herman, J. G. & Baylin, S. B. (2003). Gene silencing in cancer in

association with promoter hypermethylation. N Engl J Med 349, 2042–2054.

status of CpG islands. Proc Natl Acad Sci U S A 93, 9821–9826. Honess, R. W., Gompels, U. A., Barrell, B. G., Craxton, M., Cameron, K. R., Staden, R., Chang, Y. N. & Hayward, G. S. (1989). Deviations

from expected frequencies of CpG dinucleotides in herpesvirus DNAs may be diagnostic of differences in the states of their latent genomes. J Gen Virol 70, 837–855. Hummel, M., Zhang, Z., Yan, S., DePlaen, I., Golia, P., Varghese, T., Thomas, G. & Abecassis, M. I. (2001). Allogeneic transplantation

induces expression of cytomegalovirus immediate-early genes in vivo: a model for reactivation from latency. J Virol 75, 4814–4822. Jones, P. A. (1999). The DNA methylation paradox. Trends Genet 15,

34–37. Jordan, M. C. & Mar, V. L. (1982). Spontaneous activation of latent

cytomegalovirus from murine spleen explants. Role of lymphocytes and macrophages in release and replication of virus. J Clin Invest 70, 762–768. Koffron, A. J., Hummel, M., Patterson, B. K., Yan, S., Kaufman, D. B., Fryer, J. P., Stuart, F. P. & Abecassis, M. I. (1998). Cellular

localization of latent murine cytomegalovirus. J Virol 72, 95–103. Kurz, S. K., Rapp, M., Steffens, H. P., Grzimek, N. K., Schmalz, S. & Reddehase, M. J. (1999). Focal transcriptional activity of murine

cytomegalovirus during latency in the lungs. J Virol 73, 482–494. Magdinier, F. & Wolffe, A. P. (2001). Selective association of the

methyl-CpG binding protein MBD2 with the silent p14/p16 locus in human neoplasia. Proc Natl Acad Sci U S A 98, 4990–4995.

Acknowledgements

Masucci, M. G., Contreras-Salazar, B., Ragnar, E., Falk, K., Minarovits, J., Ernberg, I. & Klein, G. (1989). 5-Azacytidine up

The authors thank Ullrich Koszinowski for the Dm157 virus. This work is supported by USPHS grant R01 AI42898 to M. A.

regulates the expression of Epstein-Barr virus nuclear antigen 2 (EBNA-2) through EBNA-6 and latent membrane protein in the Burkitt’s lymphoma line rael. J Virol 63, 3135–3141. Meier, J. L. (2001). Reactivation of the human cytomegalovirus major

References Bubic, I., Wagner, M., Krmpotic, A., Saulig, T., Kim, S., Yokoyama, W. M., Jonjic, S. & Koszinowski, U. H. (2004). Gain of virulence

caused by loss of a gene in murine cytomegalovirus. J Virol 78, 7536– 7544. Chen, J., Ueda, K., Sakakibara, S., Okuno, T., Parravicini, C., Corbellino, M. & Yamanishi, K. (2001). Activation of latent Kaposi’s

sarcoma-associated herpesvirus by demethylation of the promoter of the lytic transactivator. Proc Natl Acad Sci U S A 98, 4119–4124. Clark, S. J., Harrison, J., Paul, C. L. & Frommer, M. (1994). High

sensitivity mapping of methylated cytosines. Nucleic Acids Res 22, 2990–2997.

immediate-early regulatory region and viral replication in embryonal NTera2 cells: role of trichostatin A, retinoic acid, and deletion of the 21-base-pair repeats and modulator. J Virol 75, 1581–1593. Mocarski, E. S. & Courcell, C. T. (2001). Cytomegaloviruses and their

replication. In Fields Virology, 4th edn, pp. 2629–2673. Edited by D. M. Knipe & P. M. Howley. Philadelphia, PA: Lippincott Williams & Wilkins. Nguyen, C. T., Weisenberger, D. J., Velicescu, M., Gonzales, F. A., Lin, J. C., Liang, G. & Jones, P. A. (2002). Histone H3-lysine 9

methylation is associated with aberrant gene silencing in cancer cells and is rapidly reversed by 5-aza-29-deoxycytidine. Cancer Res 62, 6456–6461. Raizis, A. M., Schmitt, F. & Jost, J. P. (1995). A bisulfite method of

Dorsch-Hasler, K., Keil, G. M., Weber, F., Jasin, M., Schaffner, W. & Koszinowski, U. H. (1985). A long and complex enhancer activates

5-methylcytosine mapping that minimizes template degradation. Anal Biochem 226, 161–166.

transcription of the gene coding for the highly abundant immediateearly mRNA in murine cytomegalovirus. Proc Natl Acad Sci U S A 82, 8325–8329.

Rawlinson, W. D., Farrell, H. E. & Barrell, B. G. (1996). Analysis of

http://vir.sgmjournals.org

the complete DNA sequence of murine cytomegalovirus. J Virol 70, 8833–8849. 1101

M. Hummel and others

Reeves, M. B., MacAry, P. A., Lehner, P. J., Sissons, J. G. & Sinclair, J. H. (2005). Latency, chromatin remodeling, and reactivation of

translocase selectively expressed in mouse embryonic stem cells and germ cells. Stem Cells 23, 1314–1323.

human cytomegalovirus in the dendritic cells of healthy carriers. Proc Natl Acad Sci U S A 102, 4140–4145.

Takebayashi, S., Nakao, M., Fujita, N., Sado, T., Tanaka, M., Taguchi, H. & Okumura, K. (2001). 5-Aza-29-deoxycytidine induces histone

Rodic, N., Oka, M., Hamazaki, T., Murawski, M. R., Jorgensen, M., Maatouk, D. M., Resnick, J. L., Li, E. & Terada, N. (2005). DNA

hyperacetylation of mouse centromeric heterochromatin by a mechanism independent of DNA demethylation. Biochem Biophys Res Commun 288, 921–926.

methylation is required for silencing of Ant4, an adenine nucleotide

1102

Journal of General Virology 88