Research Article

Epstein-Barr virus nuclear antigens 3C and 3A maintain lymphoblastoid cell growth by repressing p16INK4A and p14ARF expression

Seiji Maruo, Bo Zhao, Eric Johannsen, Elliott Kieff, James Zou, and Kenzo Takada
PNAS February 1, 2011 108 (5) 1919-1924; https://doi.org/10.1073/pnas.1019599108

  1. Contributed by Elliott Kieff, December 29, 2010 (sent for review December 21, 2010)

Abstract

Epstein-Barr virus (EBV) nuclear antigen 3C (EBNA3C) and EBNA3A are each essential for EBV conversion of primary human B lymphocytes into continuously proliferating lymphoblast cell lines (LCLs) and for maintaining LCL growth. We now find that EBNA3C and EBNA3A's essential roles are to repress p16INK4A and p14ARF. In the absence of EBNA3C or EBNA3A, p16INK4A and p14ARF expression increased and cell growth ceased. EBNA3C inactivation did not alter p16INK4A promoter CpG methylation, but reduced already low H3K27me3, relative to resting B cells, and increased H3K4me3 and H3-acetylation, linking EBNA3C inactivation to histone modifications associated with increased transcription. Importantly, knockdown of p16INK4A or p14ARF partially rescued LCLs from EBNA3C or EBNA3A inactivation-induced growth arrest and knockdown of both rescued LCL growth, confirming central roles for p16INK4A and p14ARF in LCL growth arrest following EBNA3C or EBNA3A inactivation. Moreover, blockade of p16INK4A and p14ARF effects on pRb and p53 by human papilloma virus type 16 E7 and E6 expression, sustained LCL growth after EBNA3C or EBNA3A inactivation. These data indicate that EBNA3C and EBNA3A joint repression of CDKN2A p16INK4A and p14ARF is essential for LCL growth.

Epstein-Barr virus (EBV) is causally associated with lymphoproliferative diseases, endemic Burkitt lymphomas, Hodgkin lymphomas, other B- and T-cell lymphomas, anaplastic nasopharyngeal carcinomas, and gastric carcinomas (reviewed in ref. 1). In primary human infection, and in vitro, EBV latency III infection converts B cells into proliferating lymphoblast cell lines (LCLs) (2, 3). In LCLs, EBV expresses six nuclear antigen proteins (EBNA2, EBNALP, EBNA3A, EBNA3B, EBNA3C, and EBNA1), two latent infection membrane proteins (LMP1 and LMP2), two small RNAs (EBER1 and EBER2), and BamHI A rightward transcripts (BARTs). EBNA and BART transcripts also encode multiple micro-RNAs. Reverse genetic and biochemical analyses indicate that EBNA2, EBNALP, EBNA3A, EBNA3C, and LMP1 mediate LCL growth, whereas EBNA1 maintains and enhances transcription from EBV episome, and LMP2 is important for immortalization of less mature B cells. EBNA2 activates and EBNALP coactivates MYC, EBNA3A, EBNA3B, EBNA3C, EBNA1, LMP1, and LMP2 transcription (reviewed in ref. 4).

The experiments described here investigate the underlying biochemical mechanisms and significance of initial findings that conditional EBNA3A or EBNA3C inactivation in LCLs induces G1 growth arrest, which for EBNA3C is associated with induction of CDKN2A p16INK4A expression, consistent with the possibility that EBNA3A and EBNA3C's essential role is to repress CDKN2A p16INK4A (5–8). EBNA3A, EBNA3B, and EBNA3C likely arose as a consequence of gene triplication, after the evolution of EBNA2 and EBNALP (9). EBNA3A, EBNA3B, and EBNA3C have the same promoter, similar exons, codons, introns, and a homologous domain, which strongly interacts with RBPJ, the sequence-specific DNA binding transcription factor that mediates EBNA2 up-regulation of transcription (10–14). Despite these similarities, EBNA3A, EBNA3B, and EBNA3C have divergent roles in maintaining LCL growth, because EBNA3B is dispensable, and only EBNA3A can replace a conditionally inactivated EBNA3A, and only EBNA3C can replace a conditionally inactivated EBNA3C (5, 6, 15–17). Because EBNA2 activates MYC expression through RBPJ, and associates less stably with RBPJ than EBNA3A, EBNA3B, or EBNA3C, some EBNA3 effects on transcription and LCL growth may be in limitation of EBNA2 access to RBPJ (10–14, 18–21). Indeed, all mutations in EBNA3A or EBNA3C that inhibit association with RBPJ are null for LCL growth (22–24), whereas EBNA3A or EBNA3C mutations that affect binding to the adenovirus E1a C-terminal binding protein (CtBP) corepressor result in continuous, albeit slower, growth (22–26). EBNA3C or EBNA3A have many specific and potentially significant interactions with other transcription factors or modifiers, including PU.1, Spi-B, HDAC1, DP103, prothymosin-α, p300, Nm23-H1, SUMO1, and SUMO3, cyclin A, SCFSKP2 ubiquitin ligase, pRb, Chk2, Mdm2, and MRS18-2, and these interactions could be relevant to CDKN2A p16INK4A or p14ARF regulation and LCL growth (27–38). Furthermore, EBNA3C up-regulates TCL1A and ITGA4, down-regulates JAG1 and NCALD RNAs, and cooperates with EBNA3A in repressing Bim, a proapoptotic Bcl-2 family protein (24, 39–41). Indeed, EBNA3C and EBNA3A may cooperatively repress the p16INK4A and p14ARF tumor suppressors to allow cell cycle progression, as is required for MYC conversion of tissue cells to stem cells, for HPV16- and HPV18-induced cervical cancer, and for other enforced cell proliferations (42–44).

Results

EBNA3C Inactivation Induces p16INK4A and p14ARF in LCLs.

LCLs infected with recombinant EBV genomes that express a conditionally active EBNA3C fused to a 4-hydroxytamoxifen (4HT)-dependent mutant estrogen-receptor hormone-binding domain stably express E3CHT in media with 4HT and grow at rates similar to wild-type LCLs (6, 45). In media without 4HT, E3CHT levels substantially decrease by 5–10 d and the LCLs stop growing by day 10, or soon thereafter (Fig. 1A, Fig. S1A, and ref. 6), unless wild-type EBNA3C is expressed, in trans, to maintain LCL growth (Fig. S1B and ref. 6). In media without 4HT, at day 6 and 11, CDKN2A p16INK4A and p14ARF RNAs significantly, substantially, and progressively increased and at day 5, 10, and 15 CDKN2A p16INK4A and p14ARF protein expression also substantially and progressively increased (Fig. 1 A–C and refs. 6 and 8). In contrast, p16INK4A and p14ARF RNA remained low in the E3CHT LCLs cultured with 4HT at day 6 and 11, as did protein levels at day 5, 10, and 15 (Fig. 1 A–C and refs. 6 and 8). Further, wild-type EBNA3C expression from an oriP plasmid transfected into E3CHT LCLs, before transfer to media without 4HT, prevented p16INK4A and p14ARF RNA induction after E3CHT inactivation and sustained LCL growth in the absence of 4HT (Fig. 1D and Fig. S1B). These results indicate that E3CHT inactivation induces p16INK4A and p14ARF RNA and protein expression and LCL growth arrest, whereas wild-type EBNA3C expressed in the same cells prevents p16INK4A and p14ARF RNA and protein expression and maintains LCL growth.

Fig. 1.
Fig. 1.

EBNA3C inactivation results in the induction of p16INK4A and p14ARF in LCLs. (A) Lysates of E3CHT LCLs cultured in the absence (−) or presence (+) of 4HT for the indicated days were Western blotted with EBV-immune human or p16INK4A, p14ARF, or actin antibodies. n.s., nonspecific. (B) Drawing of the CDKN2A locus p16INK4A and p14ARF mRNAs and p16INK4A (dark gray) and p14ARF (light gray) ORFs. Dashes below p16INK4A and p14ARF unique exons indicate shRNA targets and arrows, primers for real-time qRT-PCR. p14ARF exon 1β is upstream of p16INK4A exon 1α. (C) RNA was extracted from E3CHT LCLs in medium with (+) or without (−) 4HT for 6 or 11 d. Real-time p16INK4A and p14ARF qRT-PCR was normalized to GAPDH. Each bar is a three determination-based SEM. (D) E3CHT LCLs transfected with an oriP plasmid expressing wild-type EBNA3C (E3C) or a control oriP plasmid (Cont) and 5 d later (day 0), were washed and transferred to medium with (+) or without (−) 4HT. RNA was extracted at day 29 and real-time qRT-PCR analyzed for p16INK4A and p14ARF levels.

EBNA3C Inactivation Does Not Change p16INK4A and p14ARF Promoter CpG Methylation.

Because the p16INK4A locus is frequently silenced by CpG promoter methylation in human tumors (46), we examined whether E3CHT inactivation decreases CDKN2A locus CpG methylation in LCLs, using methylation-specific PCR (MSP). Methylated CpG-specific amplifications were readily detected at the p16INK4A and p14ARF promoters in Burkitt lymphoma (BL) cell lines (Fig. 2A). In contrast, unmethylated CpG-specific amplifications, but not methylated CpG-specific amplifications, were detected at the p16INK4A and p14ARF promoters in E3CHT LCLs, in the absence or presence of 4HT (Fig. 2A). Using bisulfite sequencing, CpGs in the p16INK4A promoter were almost fully methylated in BL cells, whereas most CpGs were unmethylated in E3CHT LCLs in the absence or presence of E3CHT (Fig. 2B), consistent with the methylation-specific PCR results (Fig. 2A). Thus, EBNA3C represses CDKN2A p16INK4A and p14ARF expression without significantly altering the mostly nonmethylated CpG state in LCLs.

Fig. 2.
Fig. 2.

EBNA3C inactivation does not alter p14ARF or p16INK4A promoter CpG methylation in LCLs. (A) Akata Burkitt lymphoma (BL), P3HR-1 BL, and E3CHT LCL DNAs cultured in the absence (−) or presence (+) of 4HT for 41 d were subjected to methylation-specific PCR using p16INK4A- or p14ARF-specific primers for the methylated (M) or unmethylated (U) promoter sequences. (B) The methylation state of CpGs in the p16INK4A promoter in P3HR-1 BL and E3CHT LCLs cultured in the absence or presence of 4HT for 41 d was determined by bisulfite sequencing.

EBNA3C Inactivation Changes Chromatin at the p16INK4A Locus from Repressive to Active.

High-level histone H3K27 trimethylation (H3K27me3) and low level H3K4me3 are associated with p16INK4A silencing (47). ChIP analyses indicated that E3CHT inactivation significantly decreased H3K27me3 and increased H3K4me3 within 0.6 kb upstream and 0.5 kb downstream of the p16INK4A promoter in E3CHT LCLs (Fig. 3), indicative of a shift from repressive to active chromatin. In further support of a more active chromatin state, E3CHT inactivation increased H3 acetylation at the same sites (Fig. 3, AcH3).

Fig. 3.
Fig. 3.

EBNA3C inactivation increases p16INK4A histone H3 acetylation and H3K4me3 and reduces of H3K27 trimethylation in LCLs. E3CHT LCLs cultured in the absence (−) or presence (+) of 4HT for 20 d were ChIP and qRT-PCR analyzed with the indicated antibodies and primer sets are shown and listed in Table S1. Bars indicate the mean and SE of three reactions.

Comparison of active transcription-associated H3K4me3 and H3K36me3 histone modifications with Polycomb repression-associated H3K27me3 marks at the CDKN2A p16INK4A and p14ARF in LCLs relative to primary CD19 positive B cells using normalized Encyclopedia of DNA Elements (ENCODE) data (Fig. S2) revealed extensive H3K27me3 through the CDKN2A p16INK4A and p14ARF locus in resting CD19+ primary B cells, except at the p16INK4A and p14ARF promoters versus very little H3K27me3 throughout the locus in LCLs. Whereas H3K36me3 was low through the locus in both resting B cells and LCLs, H3K4me3 levels were high at the p16INK4A and p14ARF promoters. The coupling of low overall H3K27me3 and high H3K4me3 at the p16INK4A and p14ARF promoters, with above background PolII through these loci in LCLs (Fig. S2) is consistent with p16INK4A and p14ARF being poised for activation in LCLs.

p16INK4A and p14ARF Knockdown Rescues LCLs from Growth Arrest Following EBNA3C Inactivation.

To evaluate whether p16INK4A and p14ARF are responsible for the LCL growth arrest induced by EBNA3C inactivation, oriP plasmids that express shRNAs targeting p16INK4A or p14ARF were verified to knock down their target RNAs in 293T cells by Western blot for p16INK4A and p14ARF (Fig. S3). When E3CHT LCLs were transfected with oriP plasmids that express shRNA targeting p16INK4A or p14ARF or p16INK4A and p14ARF, and transfected LCLs were cultured in medium with or without 4HT, knockdown of p16INK4A or p14ARF partially sustained E3CHT LCL growth in medium without 4HT and knockdown of p16INK4A and p14ARF fully rescued LCL growth (Fig. 4A). Western blot confirmed the knockdown of p16INK4A and p14ARF proteins in E3CHT LCLs following transfection with both p16INK4A and p14ARF shRNAs and culture in medium without 4HT (Fig. 4B). These data indicate that EBNA3C suppression of p16INK4A and p14ARF is the principal basis for EBNA3C's essential role in LCL growth.

Fig. 4.
Fig. 4.

Knockdown of p16INK4A and p14ARF rescues LCLs from EBNA3C inactivation-induced growth arrest. (A) E3CHT LCLs were transfected with an oriP plasmid expressing EBNA3C (E3C) or an oriP plasmid expressing control shRNA (sh-cont), shRNA targeting p16INK4A (sh-p16), shRNA targeting p14ARF (sh-p14), or both sh-p16– and sh-p14–targeting plasmids. Five days after transfection (day 0), cells were washed and transferred to medium with (+) or without (−) 4HT, and cell growth was determined. (B) Total cell lysates were prepared in A at day 41 and Western blot analyzed with EBV-immune human serum, p16INK4A-, p14ARF-, or actin-specific antibodies. Arrowhead indicates transfected E3C and n.s. indicates nonspecific. (C) E3CHT LCLs were transfected with an oriP plasmid expressing EBNA3C (E3C), HPV16 E7, HPV16 E6, both oriP plasmids expressing E7 and E6 (E7 + E6), or a control oriP plasmid (Cont). Cell growth was determined in medium with (+) or without (−) 4HT.

Because human papilloma virus type 16 (HPV16) E7 and E6 inactivate pRb and p53, downstream targets of p16INK4A and p14ARF, respectively, in cell growth inhibition (43, 44), E6 and E7 were used to confirm the key role for p16INK4A and p14ARF and to investigate the importance of pRB and p53 inactivation in EBNA3C effects. Expression of either E7 or E6 in E3CHT LCLs partially rescued LCLs from growth arrest after E3CHT inactivation, whereas E7 and E6 expression fully rescued LCL growth (Fig. 4C). E7 and E6 rescue of LCL growth was similar to that of shRNAs targeting p16INK4A and p14ARF (Fig. 4 A and C), except that pRb was further decreased and p16INK4A and p14ARF were further increased (Fig. S4). Thus, these data indicate that EBNA3C repression of p16INK4A and p14ARF or their effects on pRb and p53 is essential for LCL growth and implicate pRb and p53 in p16INK4A and p14ARF effects.

EBNA3A Is also Required for p16INK4A and p14ARF Repression.

Because conditional E3AHT inactivation in LCLs has similar effects on LCL growth and p16INK4A is also elevated in EBNA3A knockout LCLs relative to wild-type LCLs (5, 7, 8), EBNA3A may also have a role in p16INK4A and p14ARF repression. Indeed, incubation of E3AHT LCLs in media without 4HT, resulted at day 6 in substantially reduced E3AHT, slowed cell growth, increased p16INK4A and p14ARF protein expression, and ∼1.4-fold significantly increased p16INK4A and p14ARF RNA levels (Fig. 5 A and B). E3AHT LCL growth stopped by day 10 and at day 12, p16INK4A and p14ARF RNA were 2- and 1.7-fold increased and p16INK4A and p14ARF proteins were at very high levels (Fig. 5). Importantly, shRNA knockdown of p16INK4A and p14ARF restored EBNA3A-dependent LCL growth (Fig. 5C). Further, HPV16 E7 and E6 expression similarly sustained LCLs following E3AHT inactivation (Fig. 5D), confirming the importance of EBNA3A repression of the p16INK4A and p14ARF pathways for continued LCL growth.

Fig. 5.
Fig. 5.

EBNA3A inactivation induces p16INK4A and p14ARF expression and LCL growth arrest. (A) Lysates of E3AHT LCLs, cultured in the absence (−) or presence (+) of 4HT, for the indicated days, were Western blotted with EBV-immune human serum or p16INK4A-, p14ARF-, or actin-specific antibodies. n.s., nonspecific. (B) RNA was extracted from E3AHT LCLs in medium with (+) or without (−) 4HT for the indicated days. Real-time qRT-PCRs for p16INK4A and p14ARF RNA levels were normalized against GAPDH. Each bar is the mean of three reactions and the SE. (C) E3AHT LCLs were transfected with an oriP plasmid expressing EBNA3A (E3A), control shRNA (sh-cont), shRNA targeting p16INK4A (sh-p16), shRNA targeting p14ARF (sh-p14), both oriP plasmids expressing sh-p16 and sh-p14 (sh-p16 + sh-p14), or a negative control oriP plasmid (Cont). Seven days after transfection (day 0), cells were washed and transferred to medium with (+) or without (−) 4HT, and subsequent cell growth was determined. (D) E3AHT LCLs were transfected with an oriP plasmid expressing EBNA3A (E3A), HPV16 E7, HPV16 E6, both oriP plasmids expressing E7 and E6 (E7 + E6), or a control oriP plasmid (Cont). Cell growth was determined in medium with (+) or without (−) 4HT.

Discussion

The data presented here demonstrate that the E3CHT or E3AHT inactivation in LCLs induces CDKN2A p16INK4A and p14ARF expression and that knockdown of p16INK4A or p14ARF partially sustains LCL growth, whereas knockdown of both p16INK4A and p14ARF maintains LCL growth. These data, together with previous data that EBNA3A and EBNA3C are each essential for LCL growth (5, 6) and cannot be replaced by increased expression of the other, indicate that EBNA3A and EBNA3C have cooperative and independent essential roles in maintaining LCL growth by repressing CDKN2A p16INK4A and p14ARFexpression.

These effects on LCL growth are likely at least partially mediated by p16INK4A effects on pRb stability because pRb levels fall as EBNA3C inactivation induces p16INK4A and p14ARF expression (Fig. S4). Consistent with an important role for p16INK4A effects on pRB and of p14ARF on p53, HPV E7 or E6 partially maintained LCL growth following EBNA3C or EBNA3A inactivation and HPV E7 and E6 restored LCL growth.

The mechanisms through which EBNA3A and EBNA3C suppress CDKN2A p16INK4A and p14ARF are incompletely understood, but may be determined by the fundamental mechanisms through which CDKN2A p16INK4A and p14ARF are induced and regulated in latency III EBV infection. EBNA2 is expressed first and strongly up-regulates MYC (48–50). MYC induces cyclin D2, and cyclin D2-enforced cell cycle entry likely induces CDKN2A p16 and p14 expression (48–52). CDKN2A p16INK4A and p14ARF are partially regulated by different promoters and have different first exons (exon 1α for p16INK4A and 1β for p14ARF), but share exons 2 and 3 (53, 54). The CDKN2A p16INK4A-pRb and the p14ARF-p53 pathways are inactivated in many cancers and by DNA viruses, which force infected cells into S phase to enable virus DNA replication. High-risk HPV E7 and E6 oncoproteins disrupt the p16INK4A-pRb and p14ARF-p53 pathways by targeting downstream components, pRB and p53 (55). EBNA proteins are also expressed in virus replication and may induce MYC to support virus replication (56). Nevertheless, EBNA2 induction of MYC and EBNA3A and EBNA3C suppression of CDKN2A p16INK4A and p14ARF expression are essential for latency III EBV infected cell proliferation, which enables long-term latency in human lymphoid organs (for review, see ref. 1).

This study indicates that EBNA3C and EBNA3A's principal essential role in LCL growth is to suppress cell proliferation–induced activation of the CDKN2A p16INK4A and p14ARF pathways to prevent p16INK4A and p14ARF inhibition of cell growth. Although E7 and E6 expression rescued LCL cell growth, E6 and E7 further increased p16INK4A and p14ARF expression (Fig. S4), as described with HPV infection (57), probably as a consequence of E7-induced Rb degradation (Fig. S4 and ref. 58).

DNA methylation was not evident at the CDKN2A p16INK4A and p14ARF promoters in E3CHT on or off LCLs, indicating that DNA methylation does not have a major role in EBNA3C repression of this locus. Although Polycomb group (PcG) protein-associated H3K27me3 modifications and JMJD3 demethylase are implicated in CDKN2A locus regulation (54, 59–61), the ENCODE LCL data shown in Fig. S2 indicate that the CDKN2A p16INK4A and p14ARF locus in LCLs is already deficient in H3K27me3 and has over-background Pol II, as well as promoter-associated H3K4me3, consistent with significant potential for further derepression and activation. Indeed, EBNA3C inactivation further reduced H3K27me3 at the p16INK4A promoter and increased histone H3 acetylation and H3K4me3 consistent with recent findings of EBNA3C- and EBNA3A-related histone modification at the p16INK4A locus (8).

Knockdown of p14ARF was also required for restoration of LCL growth in EBNA3C- or EBNA3A-inactivated LCLs, indicating the significance of EBNA3C and EBNA3A p14ARF suppression. Interestingly, p14ARF protein was near or below detection levels in EBNA3A− LCLs (7), which required months on feeder layers for long-term outgrowth. Perhaps selection for p14ARF suppression occurred during the initial growth on fibroblast feeder cells. The EBNA3A− LCLs continuously grew with reduced proliferation rates, perhaps due to p16INK4A induction.

The mechanisms by which EBNA3C and EBNA3A repress CDKN2A p16INK4A and p14ARF expression remain to be elucidated. Certainly, constitutive EBNA2-mediated c-myc up-regulation has a central role in EBV-mediated LCL growth (49, 62) and almost certainly underlies p16INK4A and p14ARF induction in EBV-infected B cells in the absence of EBNA3C or EBNA3A. Forced MYC expression induces p16INK4A and p14ARF in normal human diploid fibroblasts and megakaryocytes (42, 63). Alternatively, aberrant cell cycle progression induced by EBNA2 may induce p16INK4A and p14ARF through E2F or CDC6 (50, 53, 64, 65). EBNA3C and EBNA3A association with RBPJ may competitively modulate EBNA2 effects through RBPJ on MYC up-regulation so that EBV-infected B cells proliferate at a lower rate or with a compensatory EBNA3C or EBNA3A RBPJ-mediated effect on CDKN2A p16INK4A and p14ARF suppression.

Because only EBNA3C can transcomplement a conditionally inactivated EBNA3C and only EBNA3A can transcomplement a conditionally inactivated EBNA3A, both EBNA3C and EBNA3A are necessary for effective p16INK4A and p14ARF repression. Also, EBNA3B, which interacts with RBPJ is not required for LCL proliferation (17). EBNA3C and EBNA3A association with CtBP has been implicated in epigenetic repression of p16INK4A (8), although deletion of the EBNA3C or EBNA3A binding sites for CtBP is only associated with slower LCL growth, whereas even point mutations in the EBNA3C or EBNA3A binding sites for RBPJ are null for LCL growth (22–24). EBNA3C and/or EBNA3A may cooperatively modulate transcription of regulators of the CDKN2A locus or more directly regulate suppression of p16INK4A and p14ARF through another interaction with RBPJ.

Materials and Methods

Cell Lines and Plasmids.

E3CHT and E3AHT LCLs are infected with a recombinant EBV, which express 4-hydroxy-tamoxifen (4HT)-dependent EBNA3C (E3CHT) or EBNA3A (E3AHT) (5, 6). In some experiments, E3CHT or E3AHT LCLs were transcomplemented with EBNA3A or EBNA3C expressed from OriP plasmids (5, 6). HPV16 E6 or E7 (65) were kindly provided by T. Kiyono (National Cancer Center Research Institute, Tokyo, Japan) and cloned into an oriP plasmid. The oriP plasmids expressing shRNA have a hygromycin resistance cassette derived from pSilencer 2.1-U6 hygro (Ambion) and a human U6 RNA pol III promoter-driving shRNA targeting p16INK4A, p14ARF, or control shRNA (Ambion), subcloned into an XbaI–NruI-digested pCEP4 vector (Invitrogen). The shRNA-targeted sequences were as follows: p16INK4A, 5′-TGCCCAACGCACCGAATAGTTACGGTC-3′ and p14ARF, 5′-GAACATGGTGCGCAGGTTC-3′.

Western Blot Analysis.

Western blot analyses (6) used anti-p16 (Santa Cruz), anti-p15 (Santa Cruz), anti-p14 (Lab Vision), anti-Rb (PharMingen), and anti–β-actin (Sigma) antibodies.

Quantitative RT-PCR.

Total RNA extraction, cDNA synthesis, and real-time qRT-PCR (6) used primers for amplification of specific targets including p16 forward (5′-AGCATGGAGCCTTCGGCTGA-3′, located on exon 1α) and reverse (5′-CCATCATCATGACCTGGATCG-3′, located on the junction of exon 1α and exon 2); p14 forward (5′-TACTGAGGAGCCAGCGTCTA-3′, located on exon 1β) and reverse (5′-TGCACGGGTCGGGTGAGAGT-3′, located on exon 2); and GAPDH forward (5′-GCCTCCTGCACCACCAACTG-3′) and reverse (5′-CGACGCCTGCTTCACCACCTTCT-3′).

MSP Analysis.

Genome DNA was subjected to bisulfite modification using an EpiTect Bisulfite kit (Qiagen). Modified DNA template was used to amplify the p16INK4A and p14ARF promoters with methylated- or unmethylated-specific primer pairs (66, 67).

Bisulfite Sequencing of p16INK4A Promoter.

Bisulfite-converted DNA was used to amplify the p16INK4A promoter (68). PCR products were cloned using pGEM-T Easy vector system (Promega). Colonies were picked and sequenced.

ChIP Assays.

ChIPs used EZ-ChIP kit (Millipore) and ChIP Ab+ acetyl-histone H3 (17–615; Millipore), ChIP Ab+ trimethyl-histone H3 (Lys27) (17–622; Millipore), and ChIP Ab+ trimethyl-histone H3 (Lys4) (17–614; Millipore). Immunoprecipitated DNAs were quantified by real-time qPCR using a SYBR Premix Ex Taq II kit (Takara) and a LightCycler (Roche) and were normalized to input DNA. Primers are listed in Table S1 (69).

Transcomplementation Assay.

Five to 10 million E3CHT LCLs or E3AHT LCLs were electroporated with 20–30 μg of oriP plasmid DNA. Transfected LCLs were cultured in medium with 4HT for 5–7 d, washed, and resuspended at 5–10 × 105 cells/5 mL of complete medium with or without 4HT in a 25-cm2 culture flask. Every 5–8 d, viable cells were counted using a hemocytometer and trypan blue exclusion and cultures were split. Total viable cells were calculated relative to initial cultures.

Acknowledgments

We thank M. Noguchi for suggestions for ChIP analyses. We also acknowledge Brad Bernstein and the National Human Genome Research Institute ENCODE Chromatin Group at Massachusetts General Hospital/Broad Institute, and the National Institutes of Health Roadmap for Epigenomics Mapping Center at the Broad Institute for the ENCODE data. This work was supported by a Grant-in-Aid for Scientific Research from the Japanese Ministry of Education, Culture, Sports, Science, and Technology (to S.M. and K.T.) and by National Cancer Institute Grant CA47006 (to E.K.).

Footnotes

  • 1To whom correspondence may be addressed. E-mail: ekieff{at}rics.bwh.harvard.edu or smaruo{at}igm.hokudai.ac.jp.

  • Author contributions: S.M., E.K., and K.T. designed research; S.M. performed research; B.Z., E.J., J.Z., and K.T. contributed new reagents/analytic tools; S.M. and E.K. analyzed data; and S.M. and E.K. wrote the paper.

  • The authors declare no conflict of interest.

  • This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1019599108/-/DCSupplemental.

View Abstract
Back to top
Article Alerts
Email Article
Citation Tools
Request Permissions
Epstein-Barr virus nuclear antigens 3C and 3A maintain lymphoblastoid cell growth by repressing p16INK4A and p14ARF expression
Seiji Maruo, Bo Zhao, Eric Johannsen, Elliott Kieff, James Zou, Kenzo Takada
Proceedings of the National Academy of Sciences Feb 2011, 108 (5) 1919-1924; DOI: 10.1073/pnas.1019599108
Digg logo Reddit logo Twitter logo Facebook logo Google logo Mendeley logo
Proceedings of the National Academy of Sciences: 108 (5)
Table of Contents

Submit

Article Classifications

Jump to section

You May Also be Interested in

Similar Articles