Search tips
Search criteria 


Logo of nihpaAbout Author manuscriptsSubmit a manuscriptHHS Public Access; Author Manuscript; Accepted for publication in peer reviewed journal;
Nat Struct Mol Biol. Author manuscript; available in PMC 2009 May 1.
Published in final edited form as:
PMCID: PMC2612995

The H3K36 demethylase Jhdm1b/Kdm2b regulates cell proliferation and senescence through p15Ink4b


The Ink4a/Arf/Ink4b locus plays a critical role in both cellular senescence and tumorigenesis. Jhdm1b/Kdm2b/Fbxl10, the mammalian paralogue of the histone demethylase Jhdm1a/Kdm2a/Fbxl11, has been implicated in cell cycle regulation and tumorigenesis. In this report, we demonstrate that Jhdm1b is an H3K36 demethylase. Knockdown of Jhdm1b in primary MEFs inhibits cell proliferation and induces cellular senescence in a pRb and p53 pathway-dependent manner. Importantly, the effect of Jhdm1b on cell proliferation and cellular senescence is mediated through de-repression of p15Ink4b as loss of p15Ink4b function rescues cell proliferation defects in Jhdm1b knockdown cells. Chromatin immunoprecipitation on ectopically expressed Jhdm1b demonstrates that Jhdm1b targets the p15Ink4b locus and regulates its expression in an enzymatic activity-dependent manner. Alteration of Jhdm1b level affects Ras-induced neoplastic transformation. Collectively, our results indicate that Jhdm1b is an H3K36 demethylase that regulates cell proliferation and senescence through p15Ink4b.


The Ink4a/Arf/Ink4b locus encodes three critical cell cycle inhibitors including p16Ink4a, p15Ink4b and Arf (p14 in human and p19 in mouse). The two members of the Ink4 protein family inhibit the binding of D-type cyclins to cyclin dependent kinase 4 and 6 (Cdk4/6), inhibiting the phosphorylation of retinoblastoma (Rb) family proteins and preventing G1/S phase transition in cells. The Arf protein, which shares two common p16Ink4a exons but contains a distinct open reading frame, is able to activate the p53 pathway through sequestration of the p53 negative regulator Mdm2 (reviewed in1,2).

The Ink4a/Arf/Ink4b locus plays a critical role in determining cellular response to oncogenic signals. In normal cells, inappropriate oncogenic stimulation activates this locus and leads to cellular senescence. However, dysregulation of this locus can facilitate tumorigenesis through multiple oncogenic signaling pathways. The importance of this locus in cellular defense against tumorigenesis is further supported by evidence that Ink4a/Arf/Ink4b is frequently deleted or mutated in a variety of human primary tumors36. In addition, combined deletion of Ink4a/Arf with Ink4b in mice results in a broader spectrum of tumors compared to mice with individual genetic deletions indicating that genes in this locus work synergistically to prevent tumor development and that p15Ink4b is a critical tumor suppressor in the absence of p16Ink4a 7. Since this locus controls both cellular senescence and tumorigenesis, tight regulation is crucial under physiological conditions. Although many oncoproteins and Polycomb group proteins have been shown to regulate Ink4a/Arf expression8,9, the mechanism that controls p15Ink4b expression remains unclear.

The JmjC-domain containing histone demethylase 1b (Jhdm1b) is a paralogue of the first identified histone lysine demethylase, Jhdm1a, which targets H3K36 for demethylation10. This fact, as well as their high homology within the catalytic JmjC domain (79%), led us to predict that the demethylase activity against the H3K36 methyl group is conserved between both paralogues (Supplementary Fig. 1). However, in a recent report Jhdm1b was implicated in the demethylation of H3K4me3 in vivo11. In addition, the reported biological functions of this protein are also controversial. Although two groups have identified Jhdm1b as a hotspot for proviral insertion in murine tumors generated by random MMLV mutagenesis, the locus has paradoxically been identified as an oncogene and a tumor suppressor depending on the screen and functional studies used12,13. In addition, subsequent studies have reported that Jhdm1b was a negative regulator of c-Jun14 or rRNA genes11, further implicating Jhdm1b in tumor suppression.

In an effort to resolve these apparent discrepancies, we set out to use methods well established in our previous studies to characterize the biochemical properties as well as biological function of Jhdm1b using primary MEF cells, which maintain normal cell cycle regulatory pathways, to study protein function. We report here that Jhdm1b is an H3K36-specific histone demethylase that functions to promote cellular proliferation and inhibit cellular senescence through the silencing of the p15Ink4b tumor suppressor gene.


Jhdm1b is an H3K36me2-specific demethylase

In an effort to explain the difference between activity prediction based on domain homology and reported in vivo substrate specificity11, we first investigated the in vitro and in vivo catalytic activity of Jhdm1b. To this end, recombinant protein was flag-affinity purified from baculovirus-infected Sf9 cells (Supplementary Fig. 2a), and subjected to an histone demethylase assay by measuring radioactive formaldehyde release10. Results shown in Figure 1a demonstrate that, of all the substrates tested, only the H3K36-specific SET2 labeled histone substrate can be demethylated by Jhdm1b, indicating that as predicted Jhdm1b has a similar substrate specificity to JHDM1A10.

Figure 1
Jhdm1b is an H3K36-specific demethylase in vitro and in vivo

Since the protein was implicated in the demethylation of H3K4me3 in vivo11, we tested the capacity of Jhdm1b to demethylate H3K4me3 in vitro. To this end, we purified a new batch of wild-type Jhdm1b along with a mutant Jhdm1b (H211A) predicted to kill the enzymatic activity (Supplementary Fig. 2b). As controls, we also purified the H3K36me2-specific demethylase, JHDM1A10, and the H3K4me3-specific demethylase, RBP215. Incubation of these proteins with SET2-labeled histone substrate confirmed efficient formaldehyde release for JHDM1A and Jhdm1b, but not Jhdm1b (H211A) (Fig. 1b). In addition, Jhdm1b exhibited about one third the activity of JHDM1A when equivalent amount of proteins were assayed. Furthermore, incubation of these proteins with substrate labeled by a mutated form of the SET7, SET7 (Y245A), which generates K4me2 and K4me316, did not show efficient release of formaldehyde above background. However, incubation of RBP2 with the same substrate resulted in formaldehyde release (Fig. 1b), demonstrating that Jhdm1b has no H3K4me3 demethylase activity under the assay conditions.

To further define the substrate specificity of Jhdm1b, we incubated the protein with core histones purified from HeLa cells and analyzed the product by Western blot using methylation state-specific antibodies. These experiments demonstrate that Jhdm1b, like its paralogue, JHDM1A, can specifically demethylate H3K36me2 and H3K36me1 histone substrates (Fig. 1c). However, it does not alter H3K4 methylation levels (Fig. 1c).

To characterize the substrate specificity of Jhdm1b in vivo, we utilized HEK293 cells selected for adherence (AD293) and overexpressed Jhdm1b and Jhdm1b (H211A) via retroviral infection. Quantitative reverse transcriptase PCR (qRT-PCR) verified similar levels of stable expression of wild-type and the mutant Jhdm1b (Fig. 1d). Western blot analysis demonstrated that overexpression of wild-type, but not the mutant, Jhdm1b resulted in marked decrease of H3K36me2 levels (Fig. 1e). Although a small decrease in H3K36me3, as well as H3K4me3 was observed, the minor decrease did not rely on a functional JmjC domain as it can also be seen in the cells over expressing the catalytic mutant. Given that overexpression of Jhdm1b in HeLa cells was previously reported to result in the decrease of H3K4me3 levels11, we overexpressed Jhdm1b in HeLa cells by lentiviral infection. Western blot analysis of the histones purified from the infected HeLa cells confirmed our observation in the HEK293 cells (Supplementary Fig. 3a). Taken together, our results indicate that Jhdm1b is an H3K36 rather than H3K4-specific demethylase.

Jhdm1b knockdown inhibits cell proliferation and induces senescence

We next explored the function of Jhdm1b in primary cells. Given that previous analyses of Jhdm1b tumor suppressor function had been carried out in transformed or immortalized cell lines, we postulated that the paradoxical conclusions regarding the role of Jhdm1b in cancer might be due to the many genetic alterations necessary for the establishment of these cell lines. We therefore designed an shRNA that can target all three murine isoforms of Jhdm1b and introduced it into primary mouse embryonic fibroblast (MEF) cells by lentiviral infection (Jhdm1b KD). After selection, the knockdown efficiency (90%) was verified by qRT-PCR (Fig. 2a). Interestingly, stable knockdown of Jhdm1b resulted in a substantial decrease in cell proliferation (Fig. 2b), a phenotype reminiscent of that displayed by MEF cells with reduced levels of the Polycomb group protein Ring1b20,21. Based on this observation, shRNA directed against Ring1b was introduced into MEF cells (Ring1b KD), and knockdown efficiency was determined to be 90% by qRT-PCR (Fig. 2a). Furthermore, knockdown of Jhdm1b had no substantial effect on Ring1b levels and vice versa (Fig. 2a). Comparison of the cell proliferation levels of Ring1b KD, Jhdm1b KD, and control cells confirmed Ring1b’s drastic effects on cell proliferation and indicated that Jhdm1b knockdown, while drastic, did not achieve the same level of cell growth inhibition as that of Ring1b knockdown (Fig. 2b). BrdU pulse labeling followed by flow cytometry analysis revealed that Jhdm1b KD resulted in a 2-fold reduction in the number of cells in S-phase as compared to control MEFs, while a 4-fold reduction is observed in the Ring1b KD cells (Fig. 2c). In addition, Ring1b has been demonstrated to play an important role in regulating cellular senescence as a component of the Polycomb repressive complex 1 (PRC1) which enacts negative regulation of the p16Ink4a locus20,22. To investigate whether Jhdm1b KD also contributes to pre-mature cellular senescence, Ring1b and Jhdm1b KD MEFs were analyzed for the presence of senescence-associated β-galactosidase (SA-β-galactosidase) at different passages. While only very few cells stained positive for SA-β-galactosidase in control MEFs at passage 5, Jhdm1b KD MEFs underwent cellular senescence at a markedly increased rate (Fig. 2d) while Ring1b knockdown had the most drastic ability to induce senescence. Taken together, these results suggest that Jhdm1b is involved in the positive regulation of cell cycle and negative regulation of passage-dependant cellular senescence.

Figure 2
Jhdm1b knockdown in primary MEFs inhibits cell proliferation and induces cellular senescence

Jhdm1b regulates cell proliferation through p15Ink4b

Cellular proliferation and senescence is tightly regulated through the p53 and pRb pathways, and both of these cellular pathways can be inhibited by the introduction of SV40 T antigen into primary cells23. To determine whether the function of Jhdm1b in cell proliferation and senescence is pRb and p53 dependent, we inactivated both pathways by retroviral expression of SV40 large T antigen in MEFs and then subjected the cells to Jhdm1b knockdown. Results shown in Supplementary Figure 4a demonstrate that knockdown of Jhdm1b does not alter the cellular proliferation of MEF cells when the p53 and pRb pathways are blocked by the SV40 large T antigen (Supplementary Fig. 4a). Consistent with this result, the percentage of S-phase cells, as indicated by BrdU incorporation, is not altered by Jhdm1b knockdown under these conditions (Supplementary Fig. 4b). In addition, Jhdm1b knockdown induced cellular senescence, as assessed by the appearance of SA-β-galactosidase staining, was also blocked in SV40 large T antigen transduced cells (Supplementary Fig. 4c). Collectively, these data suggest that Jhdm1b’s pro-growth and anti-senescence properties act upstream of these two pathways.

Previous studies have identified a number of genes whose expression is linked to Rb regulated cellular senescence and p53-dependant apoptosis24. These genes include members of the Ink4a/Arf/Ink4b tumor suppressor locus, p18 (also known as Ink4c), and the cell cycle regulators p27 (also known as Cdkn1b) and p21. To determine whether loss of Jhdm1b function affected expression of any of these genes, we performed qRT-PCR in control, Jhdm1b KD, and Ring1b KD cells. Results shown in Fig. 3a demonstrate that Jhdm1b KD resulted in a marked upregulation of p15Ink4b. In agreement with the crucial role of Ring1b in the PRC1 complex, Ring1b KD resulted in a substantial upregulation of p16Ink4a 20,22. In addition, we also observed a marked increase in p15Ink4b mRNA levels. The effect of Jhdm1b and Ring1b KD on p15Ink4b and p16Ink4a levels was also confirmed by Western blot (Fig. 3b). These data are consistent with the observation that the p16Ink4a protein has a more potent inhibitory effect on cell proliferation, as well as the notion that p15Ink4b functions to back-up the function of p16Ink4a 7. Importantly, similar to Ring1b KD, the expression of p21, a p53 pathway target gene, was not altered by Jhdm1b KD (Fig. 3a). To analyze whether p15Ink4b is a key mediator of Jhdm1b’s function in cellular proliferation, we derived primary MEFs from p15Ink4b null mice25 and performed knockdown of Jhdm1b or Ring1b. Results shown in Figure 3c and 3d demonstrate that loss of p15Ink4b function could largely rescue the slow proliferation and low BrdU incorporation caused by Jhdm1b knockdown. Thus, we conclude that Jhdm1b regulates cellular proliferation and senescence by negatively regulating the expression of the p15Ink4b tumor suppressor gene in primary cells.

Figure 3
Jhdm1b regulates cell proliferation through p15Ink4b

Jhdm1b regulates p15Ink4b expression through H3K36 demethylation

H3K36 methylation, which has been linked to active gene transcription in organisms from yeast to humans, is present within the coding regions of genes being actively transcribed, and tends to peak towards the 3’ end of transcribed regions2528. The mechanism by which high levels of H3K36me2 are excluded from the promoter of active genes, as well as the functional significance of this observation is still not clear in higher eukaryotes. However, it has been demonstrated that SET2 association with elongating RNAPII is at least partly responsible for this phenomenon in yeast27. To investigate whether the demethylase activity of Jhdm1b directly contributes to p15Ink4b regulation and cellular proliferation, we attempted to rescue the Jhdm1b KD MEFs with siRNA resistant wild-type F-Jhdm1b as well as catalytically deficient mutant, F-Jhdm1b (H211A). After confirming equal expression of the rescue constructs (Supplementary Fig. 5), we analyzed their effects on cellular proliferation. Results shown in Figure 4a demonstrate that re-introduction of wild-type Jhdm1b rescued the growth defects, while re-introduction of the catalytic-defective mutant did not. In addition, re-introduction of wild-type Jhdm1b, but not the catalytic mutant, restored p15Ink4b expression to control levels (Fig. 4b). Thus, both normal cell proliferation and p15Ink4b repression depends on the demethylase activity of Jhdm1b.

Figure 4
Jhdm1b regulates cell proliferation and p15Ink4b expression in a histone demethylase activity-dependent manner

Having established that transcriptional repression of p15Ink4b depends on the H3K36-demethylase activity of Jhdm1b, we used ChIP assays to decipher whether p15Ink4b is a direct target of Jhdm1b. Because none of the available Jhdm1b antibodies worked in immunoprecipitation in our hands, we resolved to retrovirally express F-Jhdm1b in primary MEF cells and performed ChIP assays using Flag antibodies. F-Jhdm1b was found to be specifically enriched 3-fold above background in regions just upstream (amplicon 1) and surrounding the transcription start site (amplicon 2) of p15Ink4b when compared with mock infected cells (Fig. 4c, d). However, the protein was not enriched within the p15Ink4b intron (amplicon 3) (Fig. 4c, d). ChIP analysis of H3K36me2 levels at these same regions in Jhdm1b KD and control cells revealed that KD of Jhdm1b resulted in an increase in the H3K36me2 levels across the locus when compared to control cells (Fig. 4e), suggesting that Jhdm1b might also play an active role in the demethylation of H3K36me2 downstream of the transcription start site. However, it has not escaped our attention that increased transcription of the gene may itself contribute to increases in H3K36me2 within the coding region. Furthermore, rescue of the KD by wild-type Jhdm1b resulted in H3K36me2 levels below those of control cells (Fig. 4e). In addition, H3K36me2 levels within the Gapdh locus were unchanged for all samples (data not shown). Taken together, these data suggest that p15Ink4b is a direct Jhdm1b target and that Jhdm1b regulates p15Ink4b expression through active demethylation of H3K36.

Jhdm1b cooperates with Ras to induce oncogenic transformation

All of the data presented above indicates that Jhdm1b functions as a proto-oncogene through its ability to repress transcription of the p15Ink4b locus. In an effort to further establish the oncogenic potential of Jhdm1b, we assessed the contribution of the protein to colony formation using knockdown or overexpression in p53 null MEFs followed by superinfection of retroviral H-Ras12V virus. Results of soft-agar colony formation analysis indicate that co-expression of RAS and Jhdm1b in p53 null MEF cells results in an increase in Ras induced oncogenic transformation (Fig. 5a). Conversely, knockdown of Jhdm1b inhibited Ras-induced colony formation (Fig. 5b). Together these data support the notion that Jhdm1b acts as a proto-oncogene in primary fibroblast cells.

Figure 5
Jhdm1b facilitates Ras induced neoplastic transformation


In this report, we demonstrate that Jhdm1b, like its paralogue Jhdm1a, functions as an H3K36 demethylase in vitro and in vivo. Recombinant Jhdm1b purified from baculovirus infected Sf9 cells demethylates H3K36me2 in vitro in a radioactive formaldehyde release assay and in histone Western blot analysis (Fig. 1) and Western blotting and ChIP analysis indicate that overexpression of Jhdm1b results in global H3K36me2 demethylation (Fig. 1e), as well as gene-specific H3K36me2 demethylation (Fig. 4) in different cell types including HeLa and HEK293 cells (Supplementary Fig. 3a). In addition, the previously suggested nucleolar staining pattern of Jhdm1b could not be observed in either HeLa cells11 (Supplementary Fig. 3b), or AD293 cells (Supplementary Fig. 5b).

In addition to the substrate specificity of Jhdm1b, its role in tumorigenesis has also been a point of contention. One of the earliest reports of Jhdm1b function came out of a genetic screen for tumor suppressor genes in mouse lymphomas13. The authors identified several bi-allelic retroviral insertion events at the Jhdm1b locus when several of the induced lymphomas were analyzed. However, further analysis of additional tumor samples revealed locus insertion that left the Jhdm1b coding region intact. This leaves open the possibility that the tumor manifesting retroviral insertions could either activate or suppresses Jhdm1b function. Therefore, the potential of Jhdm1b to serve as an oncogene or tumor suppressor was not resolved in this study. In addition some indirect evidence suggested Jhdm1b may act as a tumor suppressor, including a link to the negative regulation of c-Jun14, as well as a description of its role in the negative regulation of rRNA genes11. However, these experiments were carried out in various cell lines in which critical pathways important for cellular senescence and tumorigenesis (such as Ink4a/Arf/Ink4b) are also disrupted. Thus further evidence would be required to support the relevance of previously identified gene targets in a normal context. Arecent report that screened 44 random MMLV induced T cell lymphomas identified the Jhdm1b locus as an insertion hotspot12. This study revealed multiple, orientated provirus insertions upstream of the canonical Jhdm1b promoter and further implicated a role for Jhdm1b overexpression in the immortalization of primary MEF cells. In this study, we provide several lines of evidence that support Jhdm1b may indeed be a proto-oncogene. First, we demonstrate in primary MEFs that knockdown of Jhdm1b resulted in cell proliferation defects and increased senescence (Fig. 2). Furthermore, we demonstrate that Jhdm1b contributes to the regulation of cell proliferation and senescence by directly repressing the expression of the p15Ink4b tumor suppressor (Fig. 3 and Fig 4). Importantly, p15Ink4b appears to be a major target that mediates Jhdm1b function in cellular proliferation since loss of p15Ink4b function can largely rescue the Jhdm1b knockdown effects (Fig. 4). Finally, Jhdm1b can cooperate with Ras to transform primary MEF cells (Fig. 5). Therefore, we conclude that Jhdm1b is indeed a proto-oncogene which functions at least partially by controlling the expression of the p15Ink4b tumor suppressor through removal of H3K36me2.


Cloning, cell line generation

All of the cell lines used in this study were maintained in Dulbeccos’s modified Eagle’s medium supplemented with 10% (v/v) fetal bovine serum and 1% (v/v) penicillin/streptomycin. Primary MEFs harvested from day 13.5 C57BL/6 and p15Ink4b null mouse embryos were plated into a P75 flask. The confluent cells were frozen down and considered passage 1. The cells were split at 1:5 for each passage. Cell growth was measured by plating 5×105 MEFs on a 100mm plate in triplicate for each group. The cell number was counted using a hemocytometer at the time points indicated. Selection and stable maintenance of exogenous plasmids was accomplished in the presence of either 1µg ml−1 (AD293 cells) or 4µg ml−1 (MEF cells) puromycin. PCR amplification of an EST clone corresponding to full length Jhdm1B (NM_001003953) was cloned into a modified pFastBacHT B vector (Invitrogen) containing an N terminal flag epitope tag for baculovirus generation as per manufacturer’s protocol (Invitrogen). Full length Jhdm1B was subcloned into pMSCVpuro vector (Clontech) for stable cell line generation using the manufacturer’s protocol. Jhdm1b (H211A) mutant was generated by employing overlapping PCR/subcloning. Stable knockdown was achieved using a lentiviral system obtained from the NIH AIDS Research and Reference Reagent Program. The mouse U6 promoter was cloned from mouse genomic DNA and inserted into the NotI site of pTY-EF1a-nLacZ. For the LV-U6 shRNA-Pgk-Pac construct, the Pgk-Pac cassette at NotI/EcoRI sites replaced the EF1a-nLacZ cassette. The hairpin RNA targeting Jhdm1b (5’-GCTCCAACTCAGTTACTGT-3’), Ring1b (5’-GCAGTACACCATTTACATA-3’) and control (5’-GTTCAGATGTGCGGCGAGT-3’) were cloned into BBSI/HindIII sties under the U6 promoter 29. To generate wild type and mutant Jhdm1b (H211A) rescue constructs, the siRNA target site of Jhdm1b was mutated to ATTGCAGTTGAGTTACTGT by PCR mutagenesis. The siRNA resistant wild type and mutant Jhdm1b cDNAs were PCR amplified and cloned into either SpeI/EcoRI site of LV vector or NotI/XbaI of RV vector.

In vitro histone demethylase formaldehyde release assay

Histone substrates were radiolabeled and formaldehyde release assays were performed as previously described10,30. Unless otherwise stated, 5µg of Flag purified recombinant protein was incubated in the presence of labeled substrate corresponding to 60,000 input counts and demethylase buffer (50 mM HEPES-KOH (pH 8.0), 70 mM Fe(NH4)2(SO4)2, 1 mM α-ketoglutarate, 2 mM ascorbate) for one hour at 37°C. Released, labeled formaldehyde was extracted using a modified NASH technique and subjected to scintillation counting. Data is presented as counts per minute in the extracted sample.

Western blot analysis and immunostaining

Total protein was extracted by RIPA buffer. Purified native histones from HeLa cells or acid extracted histones from the indicated cell lines were prepared as previously described31. Antibodies against specific methylation states were used at dilutions ranging from 1:250 – 1:1000: H3K36me1(Abcam 9048), H3K36me210 H3K36me3(Abcam 9050), H3K4me1(Abcam 8895), H3K4me2 (Abcam 7766), H3K4me3(Abcam 8580). Blots were normalized using an antibody against pan H3 (Abcam 1791). Anti-p21(Santa Cruz SC-397), anti-p15Ink4b (Cell Signaling 4822), and anti-p16Ink4a (Santa Cruz SC-1207) were used at dilution of 1:1000 for Western Blot. Indirect immunostaining was carried out using primary MEFs and Hela cells. The cells were plated onto cover slips in 6 well plates after LV transduction, fixed 48 hours post transduction for 20 min in 4% (w/v) paraformaldehyde, washed with three times with PBS, and subsequently permeabilized for 20 min in 0.5% (v/v) Triton-X-100/PBS. Permeablized cells were blocked in 3% (w/v) BSA/PBS for 30min and incubated with Flag monoclonal M2 antibody (Sigma) at 1:1000 dilution in a humidified chamber for 3 hours. After incubation, cells were washed 3 times and incubated with FITC or Rhodamine conjugated secondary antibodies (Jackson ImmunoResearch Laboratories) at dilution of 1:500. Cells were washed twice with PBS, stained with 4,6-diamidino-2-phenylindole dihydrochloride (DAPI), and mounted on glass slides in fluorescence mounting medium (DAKO). Slides were analyzed on an Axiovert 200 fluorescent microscope (Zeiss).

BrdU incorporation and cycle cycle analysis

MEFs were grown in the presence of 10 µM BrdU for 60min. The cells were harvested and fixed overnight using 75% ethanol. Fixed cells were stained with FITC conjugated mouse anti-BrdU antibody (BD biosciences, 347583) and propidium iodide. The stained cells were analyzed by flow cytometry (BD FACSCalibur) and data was analyzed using WinMDI version 2.9 (TSRI flowcytometry Software).

Senescence associated-β-galactosidase assay

Cells were washed twice with PBS and immersed in fixation buffer (2% (w/v) formaldehyde, 0.2% (w/v) glutaraldehyde in PBS) for 10 min. After two additional PBS washes the cells were allowed to stain overnight in staining solution (40 mM citric acid/sodium phosphate pH 6.0, 150 mM NaCl, 2.0 mM MgCl2, 1 mg ml−1 x-gal).

Reverse transcription / qPCR

RNA was extracted and purified from cell lines using Qiashredder (Qiagen) and RNeasy (Qiagen) spin columns; DNase treated (Promega RQ1 Dnase) and cleaned up using Qiagen RNeasy column (Qiagen). 1µg of RNA was subjected to reverse transcription using random primers (Promega) and Improm-II reverse transcription kit (Promega). cDNA levels were assayed via Real time PCR using SYBR GreenER (Invitrogen) and analyzed on an ABI 7300 Real Time PCR System with SDS software version 1.3.1. qRT-PCR primer sequences are available in Supplementary Table 1.

ChIP assay

ChIP assays employing flag-antibody were carried out as previously reported32 with the following modifications: 20µl of M2 agarose (Sigma) was used in the immunoprecipitation and chromatin-bound beads were washed three times each in TSEI (0.1% [w/v] SDS, 1% [v/v] Triton-X-100, 2 mM EDTA, 150 mM NaCl, 20 mM Tris [pH 8.1]) TSEII (0.1% [w/v] SDS, 1% [v/v] Triton-X-100, 2mM EDTA, 500 mM NaCl, 20 mM Tris [pH 8.1]) and TSEIII (0.25M LiCl, 1% [v/v] NP-40, 1% [w/v] deoxycholate, 1mM EDTA, 10mM Tris [pH 8.1]) followed by two washes in TE. Histone modification ChIPs were carried out as previously reported33. ChIP DNA was analyzed via qPCR and data are presented as percentage of input as determined using Applied Biosystems’s SDS software Absolute Quantification protocol. Primer sequences are available in Supplementary Table 1.

Soft agar colony assay

p53 null MEFs were transduced with different LVs. 48 hours after transduction, the cells were superinfected with retroviral H-Ras12V virus. 5000 cells were mixed in the 0.35% (w/v) top agar and plated onto 0.5% (w/v) basal agar. 14 days after plating, the cells were stained with 0.005% (w/v) crystal violet and colony number was counted.

Standard error reporting

All in vitro formaldehyde release assays and cell based assays were performed in triplicate and error bars represent the standard deviation of three independent experiments. ChIP assays and qRT-PCR experiments were repeated at least twice and data is reported for one of the biological replicates. Error bars represent the standard deviation of three qPCR reactions as determined by SDS 1.3.1 software (Applied Biosystems). Soft agar colony assays were performed and error bars represent the standard deviation of large colony numbers between three separate plating replicates.

Supplementary Material


We would like to thank Dr. Linda Wolff (National Cancer Institute) for the p15Ink4b null mice. This work is partially supported by NIH (GM068804). J.H. is a fellow of the Leukemia and Lymphoma Society. Y.Z. is an investigator of the Howard Hughes Medical Institute.


1. Gil J, Peters G. Regulation of the INK4b-ARF-INK4a tumour suppressor locus: all for one or one for all. Nat Rev Mol Cell Biol. 2006;7:667–677. [PubMed]
2. Ortega S, Malumbres M, Barbacid M. Cyclin D-dependent kinases, INK4 inhibitors and cancer. Biochim Biophys Acta. 2002;1602:73–87. [PubMed]
3. Kamb A, et al. A cell cycle regulator potentially involved in genesis of many tumor types. Science. 1994;264:436–440. [PubMed]
4. Kannengiesser C, et al. New founder germline mutations of CDKN2A in melanoma-prone families and multiple primary melanoma development in a patient receiving levodopa treatment. Genes Chromosomes Cancer. 2007;46:751–760. [PubMed]
5. Ogawa S, et al. Homozygous loss of the cyclin-dependent kinase 4-inhibitor (p16) gene in human leukemias. Blood. 1994;84:2431–2435. [PubMed]
6. Okuda T, et al. Frequent deletion of p16INK4a/MTS1 and p15INK4b/MTS2 in pediatric acute lymphoblastic leukemia. Blood. 1995;85:2321–2330. [PubMed]
7. Krimpenfort P, et al. p15Ink4b is a critical tumour suppressor in the absence of p16Ink4a. Nature. 2007;448:943–946. [PubMed]
8. Bracken AP, et al. The Polycomb group proteins bind throughout the INK4A-ARF locus and are disassociated in senescent cells. Genes Dev. 2007;21:525–530. [PubMed]
9. Kotake Y, et al. pRB family proteins are required for H3K27 trimethylation and Polycomb repression complexes binding to and silencing p16INK4alpha tumor suppressor gene. Genes Dev. 2007;21:49–54. [PubMed]
10. Tsukada Y, et al. Histone demethylation by a family of JmjC domain-containing proteins. Nature. 2006;439:811–816. [PubMed]
11. Frescas D, Guardavaccaro D, Bassermann F, Koyama-Nasu R, Pagano M. JHDM1B/FBXL10 is a nucleolar protein that represses transcription of ribosomal RNA genes. Nature. 2007;450:309–313. [PubMed]
12. Pfau R, et al. Members of a family of JmjC domain-containing oncoproteins immortalize embryonic fibroblasts via a JmjC domain-dependent process. Proc Natl Acad Sci U S A. 2008;105:1907–1912. [PubMed]
13. Suzuki T, Minehata K, Akagi K, Jenkins NA, Copeland NG. Tumor suppressor gene identification using retroviral insertional mutagenesis in Blm-deficient mice. Embo J. 2006;25:3422–3431. [PubMed]
14. Koyama-Nasu R, David G, Tanese N. The F-box protein Fbl10 is a novel transcriptional repressor of c-Jun. Nat Cell Biol. 2007;9:1074–1080. [PubMed]
15. Klose RJ, et al. The retinoblastoma binding protein RBP2 is an H3K4 demethylase. Cell. 2007;128:889–900. [PubMed]
16. Xiao B, et al. Structure and catalytic mechanism of the human histone methyltransferase SET7/9. Nature. 2003;421:652–656. [PubMed]
17. Klose RJ, et al. Demethylation of histone H3K36 and H3K9 by Rph1: a vestige of an H3K9 methylation system in Saccharomyces cerevisiae? Mol Cell Biol. 2007;27:3951–3961. [PMC free article] [PubMed]
18. Yamane K, et al. PLU-1 is an H3K4 demethylase involved in transcriptional repression and breast cancer cell proliferation. Mol Cell. 2007;25:801–812. [PubMed]
19. Yamane K, et al. JHDM2A, a JmjC-containing H3K9 demethylase, facilitates transcription activation by androgen receptor. Cell. 2006;125:483–495. [PubMed]
20. Cales C, et al. Inactivation of the polycomb group protein Ring1B unveils an antiproliferative role in hematopoietic cell expansion and cooperation with tumorigenesis associated with Ink4a deletion. Mol Cell Biol. 2008;28:1018–1028. [PMC free article] [PubMed]
21. Voncken JW, et al. Rnf2 (Ring1b) deficiency causes gastrulation arrest and cell cycle inhibition. Proc Natl Acad Sci U S A. 2003;100:2468–2473. [PubMed]
22. Jacobs JJ, Kieboom K, Marino S, DePinho RA, van Lohuizen M. The oncogene and Polycomb-group gene bmi-1 regulates cell proliferation and senescence through the ink4a locus. Nature. 1999;397:164–168. [PubMed]
23. Goodrich DW, Wang NP, Qian YW, Lee EY, Lee WH. The retinoblastoma gene product regulates progression through the G1 phase of the cell cycle. Cell. 1991;67:293–302. [PubMed]
24. Sherr CJ, Roberts JM. CDK inhibitors: positive and negative regulators of G1-phase progression. Genes Dev. 1999;13:1501–1512. [PubMed]
25. Bannister AJ, et al. Spatial distribution of di- and tri-methyl lysine 36 of histone H3 at active genes. J Biol Chem. 2005;280:17732–17736. [PubMed]
26. Martin C, Zhang Y. The diverse functions of histone lysine methylation. Nat Rev Mol Cell Biol. 2005;6:838–849. [PubMed]
27. Morris SA, et al. Histone H3 K36 methylation is associated with transcription elongation in Schizosaccharomyces pombe. Eukaryot Cell. 2005;4:1446–1454. [PMC free article] [PubMed]
28. Rao B, Shibata Y, Strahl BD, Lieb JD. Dimethylation of histone H3 at lysine 36 demarcates regulatory and nonregulatory chromatin genome-wide. Mol Cell Biol. 2005;25:9447–9459. [PMC free article] [PubMed]
29. Cao R, et al. Role of hPHF1 in H3K27 methylation and Hox gene silencing. Mol Cell Biol. 2008;28:1862–1872. [PMC free article] [PubMed]
30. Zhang X, et al. Structure of the Neurospora SET domain protein DIM-5, a histone H3 lysine methyltransferase. Cell. 2002;111:117–127. [PMC free article] [PubMed]
31. Wang H, et al. Role of histone H2A ubiquitination in Polycomb silencing. Nature. 2004;431:873–878. [PubMed]
32. Zeng PY, Vakoc CR, Chen ZC, Blobel GA, Berger SL. In vivo dual cross-linking for identification of indirect DNA-associated proteins by chromatin immunoprecipitation. Biotechniques. 2006;41:694, 696, 698. [PubMed]
33. Cao R, Tsukada Y, Zhang Y. Role of Bmi-1 and Ring1A in H2A ubiquitylation and Hox gene silencing. Mol Cell. 2005;20:845–854. [PubMed]