|
|
Proc Natl Acad Sci U S A. 2008 August 26; 105(34): 12521–12526. Published online 2008 August 19. doi: 10.1073/pnas.0805540105. | PMCID: PMC2518094 |
Copyright © 2008 by The National Academy of Sciences of the USA Microbiology TRIM28 mediates primer binding site-targeted silencing of Lys1,2 tRNA-utilizing retroviruses in embryonic cells Daniel Wolf, Kevin Hug, and Stephen P. Goff* Howard Hughes Medical Institute, Department of Biochemistry and Molecular Biophysics, College of Physicians and Surgeons, Columbia University, New York, NY 10032 Received May 14, 2008. Freely available online through the PNAS open access option. Murine leukemia viruses (MLVs) and related retroelements are potently restricted in embryonic cells by postintegration transcriptional silencing, likely to protect the germ line from insertional mutagenesis. This silencing is in large part attributable to the presence of a nuclear repression complex, which targets a sequence element of the proviral DNA, the repressor-binding site. The repressor-binding site closely overlaps the tRNA primer binding site, a highly conserved sequence essential for virus replication and defining the site of initiation of DNA synthesis during reverse transcription. We have recently demonstrated that the cellular corepressor TRIM28 is recruited to the proline tRNA primer-binding site used by many MLVs and is required to mediate this silencing. Here, we show that TRIM28 is also required for the restriction of retroviruses using a completely distinct tRNA for the priming of their DNA synthesis, namely Lys-1,2 tRNA. These results generalize the role of TRIM28 in retroviral restriction and suggest that this system has evolved to restrict multiple retroviruses. Keywords: HP1, KAP-1, MLV, PBS Murine leukemia viruses (MLVs) are unable to replicate in embryonic cells ( 1– 3). The block to replication occurs after the viral DNA has integrated into the host genome and is caused by transcriptional silencing of the proviruses ( 2, 4). This silencing is the result of several cumulative effects that include poor enhancer function of the 5′ LTR ( 5– 7), the binding of several cell-type specific transacting transcriptional repressors ( 6, 8), and de novo DNA methylation ( 4). A critical target of this repression is the DNA element known as the repressor binding site (RBS) ( 2, 5, 9, 10). The RBS element comprises 17 bp that overlap closely with the 18-bp primer binding site (PBS) of MLV ( 2), a sequence complementary to the cellular proline tRNA (tRNA Pro) used to prime minus-strand DNA synthesis during reverse transcription ( 11). This silencing can be abrogated by a single base-pair substitution in the PBS known as the B2 mutation ( 2). Characterization of RBS-mediated restriction demonstrated that the 17-bp sequence was able to induce transcriptional silencing of reporter constructs in EC cells independently of position and orientation of the RBS sequence ( 10, 12). Furthermore, exonuclease III protection assay and EMSA experiments revealed the presence of a DNA-binding activity that is specifically enriched in stem cell nuclear extracts ( 10, 12). Based on these experiments, a trans-acting DNA-binding factor was postulated to exist; we have recently demonstrated that a high molecular weight complex binds the RBS sequence, and that a key component of this complex is the TRIM28 protein ( 13). TRIM28 (also known as Kap-1, or Tif1-beta) is a well characterized transcriptional corepressor that is recruited to its target genes by interactions with the Krüppel associated box (KRAB) zinc-finger DNA-binding proteins ( 14). This interaction is mediated via the KRAB box domain and leads to these DNA-binding proteins, serving as sequence-specific transcriptional repressors ( 15). TRIM28 functions as a transcriptional corepressor by, in turn, recruiting several other factors involved in transcriptional silencing and heterochromatin formation, including the NuRD histone deacetylase complex, the histone H3 K9 methyltransferase ESET, and the heterochromatin-associated protein HP1 ( 16– 18). As a component of the RBS-binding complex, TRIM28 is recruited to the PBS of MLV in EC cells, and functional TRIM28 is absolutely required for RBS-mediated restriction of MLV ( 13, 19). TRIM28 recruitment to integrated silenced proviruses in EC cells leads to the recruitment of HP1γ ( 13), and disruption of the HP1-binding domain in TRIM28 leads to inactivation of the RBS-mediated silencing machinery ( 19). This suggests that recruitment of HP1 to proviruses is critical for their subsequent silencing. Retroviruses, such as visna, spuma, and Mason-Pfizer monkey viruses, contain a distinct PBS sequence corresponding to tRNA Lys-1,2 (PBS Lys-1,2) ( 20– 22). MLV-based vectors using this sequence have been shown to be specifically restricted in embryonic cells ( 9, 23). Furthermore, it has been demonstrated that a single point mutation in the PBS Lys-1,2 sequence is able to relieve this restriction in a fashion similar to the B2 mutation in the PBS Pro sequence ( 2, 23). These observations suggest that the PBS Lys-1,2 may recruit silencing machinery similar to the PBS Pro, and thus might similarly involve TRIM28. We sought to investigate directly whether TRIM28 is also required for restriction of PBS Lys-1,2 sequences in embryonic cells and thus determine whether TRIM28-mediated silencing of PBS sequences is a more generalized mechanism of retroviral restriction. Previous work from many laboratories has documented the presence of strong sequence-specific DNA-binding activity for the PBS Pro of MLV in lysates of embryonic cells ( 9, 12, 13, 24). Similar tests for DNA-binding activity in F9 EC cell extracts, when performed using the DNA sequence of PBS Lys-1,2, the PBS used by retroviruses such as visna, spuma, and Mason-Pfizer monkey virus ( 20– 22), yielded a complex detected by EMSA with a shift indicative of similar size as the complex that binds to the PBS Pro sequence ( 9) ( A). The PBS Lys-1,2 has also been shown to specifically induce the functional restriction of viral gene expression in EC cells ( 9, 23). We sought to determine if the complex binding to this PBS sequence also contained TRIM28. Nuclear extracts were prepared from the PCC4 EC cell line, as well as from a pool of cells that had been stably transduced with a retroviral vector expressing an anti-TRIM28 siRNA hairpin (PCC4 111), and an equivalent pool of PCC4 cells expressing a scrambled siRNA control (PCC4 SCRAM) ( 13). These nuclear extracts were then used in EMSA reactions with a 28-bp probe containing either 17 bp of the PBS Pro or of the PBS Lys-1,2, as described ( 9). Nuclear extracts from PCC4 and PCC4 SCRAM cells both produced an EMSA shift of the PBS Lys-1,2 probe of approximately the same size as for the PBS Pro probe (see A). Unlike the PBS Pro probe, several additional smaller shift complexes were also detected with the PBS Lys-1,2 sequence. The identity of these bands is unknown. In the absence of TRIM28 (PCC4 111), the major shifted species was no longer detected. To confirm that TRIM28 is a component of this complex, an anti-TRIM28 antibody was added to each EMSA reaction and was able to specifically supershift the complex formed on the PBSPro sequence and the complex binding to the PBSLys-1,2 sequence (see A). Upon knockdown of TRIM28 in PCC4 111 cells, the supershift band was not observed. The effectiveness of the knockdown was confirmed by assaying total cellular levels of TRIM28 by Western blot (see A). To demonstrate that specific depletion of TRIM28 from PCC4 cells by RNAi was responsible for the disappearance of the PBSLys-1,2 RBS shift, an adenovirus was used to transiently re-express an RNAi-resistant TRIM28-myc/HIS construct in PCC4 111 cells. The transient expression of TRIM28-myc/HIS restored the RBS shift with the 28-bp PBSLys-1,2 probe, and infection with an “empty” adenovirus vector not expressing TRIM28-Myc/HIS did not (B). To demonstrate that the RBS band shift seen was derived from the re-expressed tagged TRIM28-Myc/HIS construct, the same EMSA was performed in the presence of an anti-myc antibody; this antibody was able to supershift this recapitulated myc tagged complex but did not shift the endogenous complex (see B). Western blots were performed to confirm the expression of TRIM28-myc/HIS in the adenovirus-infected lines (see B). Because TRIM28 is not a DNA-binding protein, it likely depends on other sequence-specific DNA-binding proteins for its function. If it is recruited to different viral PBS sequences by two different DNA-binding proteins, then two corresponding different DNA probes should not cross-compete with each other. To test this notion, we performed EMSA reactions with either the PBSLys-1,2 or PBSPro 28-bp probes in the presence of increasing quantities of either cold probe acting as a competitor (). As would be expected, PBSPro and PBSLys-1,2 were both effectively competed by addition of unlabeled DNAs of their own sequence, even at low DNA concentrations (5 μg/ml) (see ). There was, however, no cross-competition between probes even at concentrations as high as 125 μg/ml, an ≈500-fold excess of cold probe (see ). This suggests strongly that these DNA sequences are indeed recognized by two separate DNA-binding activities. A single point mutation in the PBS Lys-1,2 sequence is able to relieve the repression it induces in EC cells ( 23). This mutated PBS, named PBS Lys-1,2(m), would be predicted to be unable to bind TRIM28 if TRIM28 were also responsible for the restriction of the PBS Lys-1,2 sequence in EC cells. To test this hypothesis, EMSA reactions were performed with 28-bp PBS Lys-1,2 and PBS Lys-1,2(m) probes (). As predicted, the PBS Lys-1,2(m) probe did not show the TRIM28 complex RBS shift. To further investigate the relative affinities of the TRIM28 PBS Lys-1,2 binding complex, DNA competitions were performed with either WT PBS Lys-1,2 or PBS Lys-1,2(m) cold probe (see ). It was observed that the PBS Lys-1,2(m) cold probe was not able to significantly compete for binding of the TRIM28 RBS complex, whereas the PBS Lys-1,2 probe did (see ), demonstrating that the Lys-1,2(M) mutation has a considerably lower affinity for the TRIM28 RBS binding complex. These results are consistent with the fact that the mutation relieves restriction in EC cells. Having determined that the TRIM28 binding to the PBS Lys-1,2 sequence correlated with its ability to cause repression in EC cells, we sought to determine whether knock-down of TRIM28 relieved this repression. An M-MLV based construct expressing the neomycin resistance gene from the 5′LTR was modified by replacing the PBS Pro with the 17-bp corresponding to PBS Lys-1,2 (named MLV-K2-NEO), which has previously been shown to be sufficient to induce repression in an Akv-MLV based vector ( 23). An identical vector with the 17-bp corresponding to PBS Lys-1,2(m), which should be resistant to this restriction, was also constructed (MLV-K2M-NEO]. It has been demonstrated that in the case of HIV-1, the tRNA used to prime reverse transcription can be influenced by sequences outside of the PBS ( 25), and that artificially introduced PBS sequences quickly revert back to the WT tRNA Lys-3 sequence ( 26). To ensure that these newly constructed M-MLV vectors were both using tRNA Lys-1,2 to prime reverse transcription, stocks of both viruses pseudotyped with the vesticular stomatitis virus (VSV) G envelope protein were used to infect Rat2 cells. The PBS sequences from both pools of infected cells were amplified by PCR and the products were sequenced. ( A). The DNA sequences confirmed that, as expected, the MLV-K2-NEO construct produced proviruses that solely contained the tRNA Lys-1,2 sequence. The MLV-K2M-NEO construct produced a mixture of integrants containing both the Lys-1,2(M) and Lys-1,2 sequence. This partial reversion, also as expected, takes place during reverse transcription, as one strand of the double-stranded viral DNA derives its sequence from the host tRNA and not the viral genome sequence ( 27). After integration of the provirus and cell division, without mismatch repair, half the cells will contain the WT PBS and half the mutant PBS. These results validate the expected structures of the reporter constructs. We next asked whether the RBS-mediated repression manifest in PCC4 cells was attenuated in TRIM28-depleted cells. M-MLV particles pseudotyped by the VSV G protein and containing either the MLV-K2-NEO or the MLV-K2M-NEO vectors were produced, and the infectivity of each of these virus preparations was determined by colony formation assays in medium containing G418. The Lys-1,2 vector (K2) is repressed by RBS binding activity and the Lys-1,2(M) vector (K2M) is not ( 23), and therefore the ratio of the titers of these viruses in a particular cell line is a measure of the RBS activity in that cell line. Cells infected with K2M contain a mixture of unrepressible (M) and repressible (WT) proviruses; the unrepressed mutant proviruses predominate in terms of their contribution to the expression in repressive cells. Because of the partial reversion, the true ratio of K2M/K2 repression should be approximately twice the observed ratio of titers of the virus preparations. However, here we simply report the observed ratio of titers as the simplest and adequate measure of repression of the fully WT PBS-containing over that of the (partially) mutant PBS-containing viruses. Rat2 cells lack TRIM28-mediated repression activity, and therefore the MLV-K2-NEO and MLV-K2M-NEO viruses infect with these cells with very similar efficiencies. To correct for variations in titers between virus preparations, we chose to use this cell line as the standard, and the ratios of the titers of the K2M/K2 viruses were all normalized to the ratio of titers in this cell line. In PCC4 and PCC4 SCRAM cells, the titer of the K2 virus was strongly repressed (9- and 7.8-fold, respectively) over that of the K2M virus ( B), demonstrating that these cells possess RBS silencing activity. As has been reported, the level of repression mediated by the PBS Lys-1,2 was considerably lower than that of repression mediated by the PBS Pro, which has a repression in these cells of ≈70-fold ( 13, 23). In the TRIM28 knockdown cells (PCC4 111), however, there was a near complete relief of repression, with a repression of only 1.7-fold (see B). The datasets for each of several individual experiments are presented in supporting information (SI) Table S1. These results confirm that knockdown of TRIM28 causes a significant decrease in RBS-dependent silencing of PBS Lys-1,2, and thus strongly suggest that TRIM28 is required for this process. We have demonstrated that MLV PBS Pro-mediated restriction in embryonic cells requires TRIM28 ( 13, 19). In this article we have extended these observations to show that TRIM28 is also required for the specific restriction of PBS Lys-1,2-containing viruses in embryonic cells (see B). We observe that the TRIM28-containing complex that recognizes the PBS Lys-1,2 sequence is of a similar size (as judged by EMSA migration) as the PBS Pro binding complex (see A), and can be similarly supershifted with an anti-TRIM28 antibody (see A). The DNA recognition components of the complex, however, appear to be different (see ), suggesting that TRIM28 is interacting with at least two separate DNA-binding factors, each of which recognizes a different PBS sequence. Because both of these DNA-binding components presumably interact with TRIM28 and subsequently lead to transcriptional repression, KRAB-box zinc-finger proteins make attractive candidates for this activity. KRAB-box zinc-finger proteins are a large class of zinc-finger proteins (290 in the human genome) that remain largely uncharacterized. The KRAB box has previously been shown to specifically interact with TRIM28, and this interaction has been shown to induce transcriptional repression of relevant target genes ( 14, 15, 28). In the context of retrovirus and retroelement silencing, we expect that further biochemical analysis of the TRIM28 complex should allow identification of these different DNA-binding factors. These data as a whole lead us to suggest a generalized model for PBS-mediated restriction of retroviruses, in which viruses with either PBSPro and PBSLys-1,2, or indeed other as yet uncharacterized PBS sequences, are silenced in embryonic cells by the specific recruitment of TRIM28 (). We speculate that each PBS sequence will be recognized by a different DNA-binding protein, and that these proteins will all have the ability to recruit TRIM28. TRIM28 recruitment to the newly integrated provirus will lead to silencing of LTR-driven transcription through the subsequent recruitment of HP1 and possibly also silencing machinery, such as the NuRD histone deacetylase complex and the histone H3 K9 methyltransferase ESET (see ), although the involvement of these last two components remains to be formally established. As TRIM28 is ubiquitously expressed, it also seems plausible that other components of the system, likely the DNA-binding factors themselves, will be under tight developmental control to maintain this restriction only in embryonic and stem cells. The biological purpose of PBS-mediated silencing of retroviruses is still as yet not fully elucidated. The stem cell specificity of this process suggests that it may have evolved to prevent proviral insertions into the germ line of organisms, or possibly prevent the activation of retrotransposons during germ cell development. As TRIM28 is essential for early mouse development ( 29), genetic analysis of this question has so far proved difficult. Identification of the PBS-DNA-binding factors, however, which presumably will be required for this process, should lead to a better understanding of this process. The discovery that multiple PBS sequences are specifically silenced in embryonic cells through a common mechanism suggests that this mode of restriction has evolved to specifically inhibit multiple retrovirus types. To date, most studies of PBS-mediated restriction have been performed in mouse cells. However, it has been reported that PBS Pro-mediated restriction also occurs in human cells ( 24). Thus it is likely that other mammals will also induce embryonic silencing of other PBS-containing retroviruses. Identification of the PBS-DNA binding factors from multiple species should give insight into whether different mammalian species have evolved the ability to restrict sequences other than PBS Pro and PBS Lys-1,2, based presumably as a result of selective pressure to exposure to such viruses at some time during their evolution. The inherited legacy of such repressive systems provide a historical record of the viral epidemics occurring during the course of mammalian evolution, and an indication of the importance of the restriction of virus replication as a selective force. Cell Culture and Stable RNAi-Expressing Cell Line Production and Transduction. PCC4, Rat2, and 293T cells were cultured in DMEM with 10% FBS at 37°C in 5% CO 2. RNAi knockdown lines were as previously described ( 13). Viral transduction assays and virus production were also performed as described ( 13). Briefly, retroviral preparations were serially diluted (for titer determination), and added to PCC4 and Rat2 cells (seeded at [PCC4] 3.5 × 10 3 and [Rat2] 2 × 10 3 cells per cm 2 the day before transduction) in the presence of 8 μg of Polybrene/ml. G418-containing selective media was added 48 h after transduction at 1 mg/ml for Rat2 and at 0.5 mg/ml PCC4 cells (for PCC4 cells, media were changed to 1 mg/ml after 10 days), and colonies were counted after 14–18 days of selection. Each experiment was performed in triplicate. Preparation of Nuclear Extracts. Nuclear extracts were prepared from PCC4 cells, as described ( 13). EMSAs. EMSAs were performed as before ( 13) with the following probes:
- PBSPro: GGGGGCTCGT CCGGGATCGGGAGACCCC
- PBSLys-1,2: GGCGCCCAACGTGGGGCCGGGAGAC CCC
- PBSLys-1,2(m): GGCGCCtAACGTGGGGCCGGGAGACCCC. Sense strands only are shown.
Adenoviral Production/Infections. Adenoviral production and infections were performed as described ( 30) with modifications ( 31). We thank Eric Barklis (Oregon Health & Science University, Portland, OR) for his generosity with reagents. We are grateful to Martha de los Santos and Mathew Walters for technical assistance. This work was supported by Public Health Service Grant R37 CA 30488 from the National Cancer Institute. D.W. is an Associate and S.P.G. is an Investigator of the Howard Hughes Medical Institute. 1. Linney E, Davis B, Overhauser J, Chao E, Fan H. Non-function of a Moloney murine leukaemia virus regulatory sequence in F9 embryonal carcinoma cells. Nature. 1984;308:470–472. [PubMed] 2. Barklis E, Mulligan RC, Jaenisch R. Chromosomal position or virus mutation permits retrovirus expression in embryonal carcinoma cells. Cell. 1986;47:391–399. [PubMed] 3. Teich NM, Weiss RA, Martin GR, Lowy DR. Virus infection of murine teratocarcinoma stem cell lines. Cell. 1977;12:973–982. [PubMed] 4. Stewart CL, Stuhlmann H, Jahner D, Jaenisch R. De novo methylation, expression, and infectivity of retroviral genomes introduced into embryonal carcinoma cells. Proc Natl Acad Sci USA. 1982;79:4098–4102. [PubMed] 5. Feuer G, Taketo M, Hanecak RC, Fan H. Two blocks in Moloney murine leukemia virus expression in undifferentiated F9 embryonal carcinoma cells as determined by transient expression assays. J Virol. 1989;63:2317–2324. [PubMed] 6. Tsukiyama T, Niwa O, Yokoro K. Mechanism of suppression of the long terminal repeat of Moloney leukemia virus in mouse embryonal carcinoma cells. Mol Cell Biol. 1989;9:4670–4676. [PubMed] 7. Loh TP, Sievert LL, Scott RW. Proviral sequences that restrict retroviral expression in mouse embryonal carcinoma cells. Mol Cell Biol. 1987;7:3775–3784. [PubMed] 8. Gorman CM, Rigby PW, Lane DP. Negative regulation of viral enhancers in undifferentiated embryonic stem cells. Cell. 1985;42:519–526. [PubMed] 9. Yamauchi M, Freitag B, Khan C, Berwin B, Barklis E. Stem cell factor binding to retrovirus primer binding site silencers. J Virol. 1995;69:1142–1149. [PubMed] 10. Petersen R, Kempler G, Barklis E. A stem cell-specific silencer in the primer-binding site of a retrovirus. Mol Cell Biol. 1991;11:1214–1221. [PubMed] 11. Harada F, Peters GG, Dahlberg JE. The primer tRNA for Moloney murine leukemiavirus DNA synthesis. Nucleotide sequence and aminoacylation of tRNAPro. J Biol Chem. 1979;254:10979–10985. [PubMed] 12. Loh TP, Sievert LL, Scott RW. Evidence for a stem cell-specific repressor of Moloney murine leukemia virus expression in embryonal carcinoma cells. Mol Cell Biol. 1990;10:4045–4057. [PubMed] 13. Wolf D, Goff SP. TRIM28 mediates primer binding site-targeted silencing of murine leukemia virus in embryonic cells. Cell. 2007;131:46–57. [PubMed] 14. Friedman JR, et al. KAP-1, a novel corepressor for the highly conserved KRAB repression domain. Genes Dev. 1996;10:2067–2078. [PubMed] 15. Urrutia R. KRAB-containing zinc-finger repressor proteins. Genome Biol. 2003;4:231. [PubMed] 16. Schultz DC, Ayyanathan K, Negorev D, Maul GG, Rauscher FJ., 3rd SETDB1: a novel KAP-1-associated histone H3, lysine 9-specific methyltransferase that contributes to HP1-mediated silencing of euchromatic genes by KRAB zinc-finger proteins. Genes Dev. 2002;16:919–932. [PubMed] 17. Schultz DC, Friedman JR, Rauscher FJ., 3rd Targeting histone deacetylase complexes via KRAB-zinc finger proteins: the PHD and bromodomains of KAP-1 form a cooperative unit that recruits a novel isoform of the Mi-2alpha subunit of NuRD. Genes Dev. 2001;15:428–443. [PubMed] 18. Le Douarin B, et al. A possible involvement of TIF1 alpha and TIF1 beta in the epigenetic control of transcription by nuclear receptors. EMBO J. 1996;15:6701–6715. [PubMed] 19. Wolf D, Cammas F, Losson R, Goff SP. Primer binding site-dependent restriction of murine leukemia virus requires HP1 binding by TRIM28. J Virol. 2008;82:4675–4679. [PubMed] 20. Maurer B, Bannert H, Darai G, Flugel RM. Analysis of the primary structure of the long terminal repeat and the gag and pol genes of the human spumaretrovirus. J Virol. 1988;62:1590–1597. [PubMed] 21. Sonigo P, et al. Nucleotide sequence of the visna lentivirus: relationship to the AIDS virus. Cell. 1985;42:369–382. [PubMed] 22. Sonigo P, Barker C, Hunter E, Wain-Hobson S. Nucleotide sequence of Mason-Pfizer monkey virus: an immunosuppressive D-type retrovirus. Cell. 1986;45:375–385. [PubMed] 23. Modin C, Lund AH, Schmitz A, Duch M, Pedersen FS. Alleviation of murine leukemia virus repression in embryonic carcinoma cells by genetically engineered primer binding sites and artificial tRNA primers. Virology. 2000;278:368–379. [PubMed] 24. Haas DL, et al. The Moloney murine leukemia virus repressor binding site represses expression in murine and human hematopoietic stem cells. J Virol. 2003;77:9439–9450. [PubMed] 25. Kleiman L, Cen S. The tRNALys packaging complex in HIV-1. Int J Biochem Cell Biol. 2004;36:1776–1786. [PubMed] 26. Wakefield JK, Wolf AG, Morrow CD. Human immunodeficiency virus type 1 can use different tRNAs as primers for reverse transcription but selectively maintains a primer binding site complementary to tRNA(3Lys). J Virol. 1995;69:6021–6029. [PubMed] 27. Gilboa E, Mitra SW, Goff S, Baltimore D. A detailed model of reverse transcription and tests of crucial aspects. Cell. 1979;18:93–100. [PubMed] 28. Margolin JF, et al. Kruppel-associated boxes are potent transcriptional repression domains. Proc Natl Acad Sci USA. 1994;91:4509–4513. [PubMed] 29. Cammas F, et al. Mice lacking the transcriptional corepressor TIF1beta are defective in early postimplantation development. Development. 2000;127:2955–2963. [PubMed] 30. Ng P, Cummings DT, Evelegh CM, Graham FL. Yeast recombinase FLP functions effectively in human cells for construction of adenovirus vectors. BioTechniques. 2000;29:524–526. 528. [PubMed] 31. Kyratsous CA, Silverstein SJ. BAG3, a host cochaperone, facilitates varicella-zoster virus replication. J Virol. 2007;81:7491–7503. [PubMed]
|
This article has been 
