RE-IIBP Methylates H3K79 and Induces MEIS1-mediated Apoptosis via H2BK120 Ubiquitination by RNF20

Histone lysine methylation contributes to transcriptional regulation by serving as a platform for the recruitment of various cofactors. Intense studies have been conducted for elucidating the functional meaning of H3K79 methylation, and to date, the only known HMTase responsible for the modification was DOT1L. In this study, we report that the MMSET isoform RE-IIBP has HMTase activity for H3K79. It was uncovered that RE-IIBP up-regulates MEIS1 transcription through H3K79 methylation via recruitment to the MEIS1 promoter. By means of proteomic and biochemical analysis, association of RE-IIBP with the E3 ubiquitin ligase RNF20 was demonstrated for synergistic activation of MEIS1 transcription via H3K79 HMTase activity. Furthermore, It was observed that RE-IIBP induces MEIS1-mediated apoptosis, which was dependent on H2BK120 ubiquitination by RNF20. These findings suggest RE-IIBP as another candidate for further studies to elucidate the mechanism of H3K79 methylation and its biological functions.


RE-IIBP Has HMTase Activity for Histone H3 Lysine 79. Previous studies indicated that SET
domain containing MMSET and isoform RE-IIBP may possess a number of different histone lysine specificities for methylation, including H3K4, H3K27, H3K36 and H4K20 [10][11][12][13][14] . Given that dysregulation of MMSET proteins and subsequent alteration of the epigenetic modifications thereafter have been linked to several human diseases, we decided to further characterize the unidentified lysine specificity of RE-IIBP HMTase activity.
In order to investigate the additional lysine specificity of RE-IIBP, the methylation levels of purified histones were screened after RE-IIBP transfection. H3K79 di-methylation (H3K79me2) was elevated, besides the previously identified H3K27me2 (Supplementary Figure S1A and Fig. 1A). When transfection with the RE-IIBP R477A, HMTase activity defective point mutant, was carried out, increase of the H3K79me2 levels did not occur (Fig. 1A). In contrast to H3K79 methylation, RE-IIBP did not change H3K9 and H3K4 methylation levels in our assay condition (Supplementary Figure S1A). RE-IIBP stably knocked-down 293T cell line was created by lenti-viral infection using two independent shRNAs, and the H3K79me2 levels were found to be significantly down-regulated (Fig. 1B). To further confirm the H3K79 HMTase activity of RE-IIBP, in vitro HMTase assays were performed using GST-RE-IIBP fusion protein and core histones or nucleosome, for which consistent results were obtained (Fig. 1C). RE-IIBP had similar HMTase activity toward core histones and necleosomes. In addition, confocal analysis was conducted using transiently RE-IIBP overexpressed 293T cells. This analysis also showed that RE-IIBP expression increased the H3K79me2 levels, confirming the H3K79 HMTase activity of RE-IIBP (Fig. 1D). However, in RE-IIBP R477A overexpressed cells, the change in H3K79me2 level was not detected (Fig. 1D). Both H3K79me2 images contain speckles showing similar patterns with γ -tubulin and they are speculated as non-histone proteins in centrosome detected non-specifically (data not shown). Upon examination of whether the RE-IIBP might mono-, di-, or tri-methylate H3K79, it was found that RE-IIBP can induce H3K79me1, H3K79me2, and H3K79me3 (Fig. 1E). Next, a more specific HMTase assay was conducted with histone H3 peptides, in which each six-amino-acid peptide contained one lysine residue. As a negative control, peptide H3N5 incubated with only GST protein was analyzed. Incubation with peptide H3N5 enhanced the strong methylation activity along with other peptides tested, which indicates that histone H3K79 is also a methylation target residue for RE-IIBP (Fig. 1F). To further confirm the lysine specificity of RE-IIBP, histone peptides were analyzed by LC-MS after the HMTase assay. The calculated molecular mass of the H3N5 peptide is 866 Da, for which each addition of a methyl group causes an increase of 14 to 15 Da. The nonmethylated H3N5 peptide had a main peak at 866.4 Da, while the di-and trimethylated peptides appeared at 919.4 and 911.4 Da, respectively, with the 23 Da mass of Na + during the peptide sample preparation after the HMTase assay (Fig. 1G). The spectroscopy results clearly indicated that histone peptides H3N5 were di-and tri-methylated at H3K79. No peaks corresponding to the methylated forms of the H3N5 peptides were observed without the addition of RE-IIBP.
To prove that H3K79 methylation activity of RE-IIBP in the absence of DOT1L, 293T cells were treated with the SGC0946, DOT1L specific inhibitor, after which a significant decrease of H3K79me2 was observed (Fig. 1H). Consequently, when RE-IIBP was overexpressed in the SGC0946-treated cells, H3K79 methylation was successfully restored, proving the H3K79 HMTase activity of RE-IIBP (Fig. 1H). A DOT1L knocked-down stable cell line was created using lenti-viral infection, and overexpression of RE-IIBP successfully up-regulated the levels of H3K79me2 (Supplementary Figure S1B). Taken together, these results strongly indicate that RE-IIBP has H3K79 HMTase activity.   H3K79me2 and H3K27me2 levels were analyzed in RE-IIBP or RE-IIBP R477A transfected 293T cells and stably knocked-down 293T cells using two independent shRNAs. Purified histones were immunoblotted with anti-H3K79me2, anti-H3K27me2, anti-RE-IIBP and anti-GAL4 antibodies. β -actin and H3 were used as loading controls. (c), Core histones and nucleosomes were used as substrates in the in virto HMTase assay with GST and GST-RE-IIBP. Incubated core histones and nucleosomes were subjected to Western blot. The amount of GST-RE-IIBP was determined by Coomassie staining. (d), 293T cells transiently transfected with His-RE-IIBP or His-RE-IIBP R477A were fixed, permeabilized, and immunostained with anti-His and anti-H3K79me2 antibodies. Nuclei were counterstained with DAPI. (e), Immunoblot analyses show the relative methylation levels of H3K79 in RE-IIBP transfected 293T cells. Purified histones using histone purification assay were immunoblotted with anti-H3K79me1, anti-H3K79me2, and anti-H3K79me3 antibodies. The methylation levels are normalized to H3. (f), HMTase assays were performed with four synthesized peptides, shown above, as substrates. Peptide methylations were measured using a scintillation counter. (g), Synthesized peptides (H3N5) were used as substrates in the HMTase assay with purified GST or GST-RE-IIBP. After 3 h, the peptides were analyzed by LC-MS. (h), SGC0946, DOT1L inhibitor, mediated H3K79me2 levels were restored by RE-IIBP. 293T cells treated with 1 nM SGC0946 were transiently transfected with RE-IIBP. Cells were lysed or used for histone purification assay. Each extracts were immunoblotted with anti-H3K79me2, anti-H3, anti-HA and anti-β -actin antibodies (left). The relative expression levels of H3K79me2 were quantified (right). Results are shown as means ± SDs n = 3; **p < 0.01, ***p < 0.001 been reported 24 . Therefore, we sought to screen the potential target genes of RE-IIBP. Selected genes were identified using real-time PCR. Transcription levels of JMJD6, HSPA1A, WDR6, and MEIS1 were up-regulated by RE-IIBP overexpression and USP36 level was not changed (Supplementary Figure S2). A member of the TALE family MEIS1 is a cofactor of HOX and increases HOX-DNA binding affinity. Increasing of HOX transcriptional activity by MEIS1 is known to contribute to accelerated development of leukemia. Besides, conjugation of MEIS1 with HOXA7 or HOXA9 has been found in various types of human cancer 25,26 . Similar to proto-oncogene c-Myc-mediated proliferation and apoptosis, overexpression of MEIS1 was found to induce apoptosis in various types of cells via interaction with cofactor PBX1 23,27,28 .
In addition, we previously reported that the RE-IIBP is up-regulated in the blood cells of various leukemic patients 10 . We supposed that RE-IIBP may regulate leukemia related gene expression. We previously screened RE-IIBP target genes in leukemia cell line and reported that the transcription of MEIS1 is regulated by RE-IIBP 17 . Extensive studies have been conducted for the role of MEIS1 in normal hematopoiesis and leukemia 22 . To further elucidate the detailed molecular mechanism of MEIS1 regulation by RE-IIBP, the effect of RE-IIBP overexpression on the levels of MEIS1 was first tested using real-time PCR and western blot in 293T cell lines. As expected, overexpression of RE-IIBP caused up-regulation of MEIS1 expression ( Fig. 2A). In RE-IIBP stably knocked-down 293T cell line, MEIS1 expression was significantly down-regulated (Fig. 2B). In order to determine more precisely whether the transactivation of MEIS1 by RE-IIBP is attributed to the H3K79 HMTase activity of RE-IIBP, a reporter assay was performed in 293T cells using a MEIS1-driven luciferase (luc) reporter system. Consistent with our other studies, MEIS1 promoter activity was increased when RE-IIBP was transfected in a dose-dependent manner (Fig. 2C). Using the RE-IIBP R477A, the increase of MEIS1 promoter activity by RE-IIBP was shown to be dependent on the HMTase activity, possibly through H3K79 methylation (Fig. 2C). As expected, decrease of MEIS1 promoter activity was identified when RE-IIBP was knocked down using two different shRE-IIBP RNAs (Fig. 2C). Since RE-IIBP antibodies recognize both RE-IIBP and MMSET, a K562 stable cell line which only targets MMSET, but not RE-IIBP, was generated using shRNA against the MMSET HMG domain through lenti-viral infection (Supplementary Figure S3). Interestingly, knocked-down of MMSET did not affect MEIS1 expression (Fig. 2D).
To further elucidate the mechanism underlying the transcriptional regulation of MEIS1 by H3K79 HMTase activity of RE-IIBP, a ChIP analysis with real-time PCR was performed. As expected, RE-IIBP recruitment and H3K79me2 levels at the MEIS1 promoter were increased when RE-IIBP was overexpressed ( Fig. 2E). Together, these findings suggest that RE-IIBP up-regulates MEIS1 transcription through H3K79 methylation via recruitment to the MEIS1 promoter.

RE-IIBP
Associates with RNF20 to Synergistically Activate MEIS1 Transcription. To gain a better understanding of the proteins that interact with RE-IIBP, GFP-RE-IIBP was overexpressed in 293T cells and GFP affinity purification was performed. The immunoprecipitated proteins were analyzed by liquid chromatography-mass spectrometry (LC-MS). Among the proteins found to interact with RE-IIBP, RNF20, a known E3 ubiquitin ligase, was observed (Fig. 3A).
To confirm the interaction between RE-IIBP and RNF20, the GST pull-down assay was first performed with purified GST-RE-IIBP and RNF20 overexpressed cell extracts (Fig. 3B). In addition, we confirmed this interaction by performing the experiment in reverse; that is, lysates from ectopically expressing RE-IIBP were incubated with GST-RNF20 (Fig. 3B). These experiments revealed noticeable interaction between RE-IIBP and RNF20. The interaction between the two proteins was confirmed in vivo via reciprocally conducted co-immunoprecipitation (IP) assays (Fig. 3C). It was next examined whether RNF20 affects the transcriptional regulation of MEIS1 through RE-IIBP. It is interesting that the RE-IIBP-mediated up-regulation of MEIS1 expression was further activated, as shown by increasing expressions of MEIS1 in real-time PCR and western blot analysis (Fig. 3D). To further delineate the effects of RNF20 on RE-IIBP activity, a reporter assay was performed using a MEIS1-luc promoter. Consistent results were obtained in that RNF20 further increased the RE-IIBP-mediated transactivation of MEIS1 (Fig. 3E). To investigate the role of the RE-IIBP/RNF20 complex at the MEIS1 promoter, a ChIP analysis for the MEIS1 promoter was performed in the DOT1L knocked-down stable cell line. Again, the recruitment of RE-IIBP to MEIS1 promoter was higher when RNF20 was co-transfected ( Fig. 3F). It was further identified that the H3K79 HMTase activity by RE-IIBP was also up-regulated in the presence of RNF20 (Fig. 3F). Together, these results suggest that RNF20 interacts with RE-IIBP and potentiates the RE-IIBP-mediated MEIS1 transcription through increasing the H3K79 HMTase activity of RE-IIBP.

Induction of MEIS1-Mediated Apoptosis by RE-IIBP is Dependent on H2BK120
Ubiquitination. In a previous study, RNF20-mediated H2BK120ub was found to be prerequisite for H3K79 methylation 3 . Since RE-IIBP interacts with RNF20, we hypothesized that RNF20 might recruit RE-IIBP to target loci and increase the H2BK120ub-mediated H3K79 methylation. To investigate this possibility, confocal analysis was conducted using RE-IIBP and RNF20 co-transfected H1299 cells, we showed that overexpression of RE-IIBP up-regulated H3K79 methylation in Fig. 1D. In addition, RNF20 was co-localized with RE-IIBP in nucleus and further increased RE-IIBP-mediated H3K79 methylation levels ( Fig. 4A and Supplementary Figure S4A). Consistent results were obtained that RNF20 further Scientific RepoRts | 5:12485 | DOi: 10.1038/srep12485 increased the RE-IIBP-mediated H3K79 methylation (Supplementary Figure S4B). To exclude the DOT1L-mediated H3K79 methylation, 293T cell line was treated with SGC0946, after which the H3K79 methylation levels were decreased (Fig. 4B). When RE-IIBP was overexpressed in the SGC0946-treated cells, the H3K79 methylation levels were restored (Fig. 4B). Intriguingly, overexpression of RNF20 increased not only the levels of H2bK120ub, but also that of H3K79 methylation (Fig. 4B). In addition, consistent results were confirmed using DOT1L knocked-down stable cell line (Supplementary Figure  S4C). This strongly indicates that RNF20 increased RE-IIBP-mediated H3K79 methylation independent of H3K79 methylation by DOT1L. To determine whether RNF20 can regulate recruitment of RE-IIBP to the target gene, MEIS1 promoter, a ChIP analysis was conducted with real-time PCR. The ChIP analysis demonstrated that the recruitment of RE-IIBP and the levels of H3K79me2 at the MEIS1 promoter were increased when RNF20 was overexpressed (Fig. 4C). When RNF20Δ RING, RING domain deleted mutant, was overexpressed, levels of H3K79me2 were not changed. However, the recruitment of RE-IIBP  remained increased in both RNF20 and RNF20Δ RING overexpression (Fig. 4C). In our preliminary data, RE-IIBP interacts with intermediate domain of RNF20 (Supplementary Figure S4D). This indicates that RNF20 interaction with RE-IIBP is RING domain independent. Still, RE-IIBP-mediated H3K79 methylation is dependent on H2BK120ub by RNF20. Taken together, these results indicate that RNF20 interacts with RE-IIBP and guides it to the target gene promoter for the enhancement of H2BK120ub-mediated H3K79 methylation by RE-IIBP.
Given that MEIS1 overexpression is known to induce apoptosis in a variety of cell types through a caspase-dependent process 23 , we next hypothesized that RE-IIBP might induce apoptosis through the activation of MEIS1 expression. To verify RE-IIBP induced apoptosis, RE-IIBP dependent cell survival rate was measured at indicated time points. The cell survival rate was decreased by RE-IIBP overexpression but increased by RE-IIBP knocked-down (Supplementary Figure S4E). To further confirm the effects of RE-IIBP on viability, the MTT assay was performed to measure cell survival at different time points, and consistent results were obtained as for the cell survival rate (Fig. 4D). Co-transfection of RNF20 and RE-IIBP further decreased for the cell survival rate (Fig. 4D).
Since RNF20 further enhanced the RE-IIBP-mediated induction of MEIS1, we next hypothesized that RNF20 might have an effect on RE-IIBP-mediated apoptosis. When RE-IIBP and RNF20 was overexpressed individually, increase of apoptosis marker caspase3 was detected. In addition, further up-regulation of the caspases3 was verified after co-transfection of RE-IIBP and RNF20, confirming that MEIS1-mediated apoptosis was induced by RE-IIBP and further enhanced by RNF20 (Fig. 4E). To further elucidate the biological role of RE-IIBP in cellular function, FACS analysis was performed to measure apoptosis. FACS analysis revealed that apoptosis was induced by the expression of RE-IIBP (sub-G1 6.27% increased) and RNF20 (sub-G1 7.78% increased), respectively. As a positive control, cells treated with etoposide were analyzed. Also, co-transfection of RE-IIBP and RNF20 resulted in a further increase in the number of cells entering apoptosis (sub-G1 11.64% increased) (Fig. 4F). Consistent results were obtained in FACS analysis using shRE-IIBP K562 stable cell, in which knocked-down of RE-IIBP inhibited etoposide-mediated apoptosis (24.64% or 16.89% decreased) (Fig. 4G). These results indicated that RE-IIBP induced MEIS1 expression via methylation of H3K79, and that interaction with RNF20 further enhanced RE-IIBP recruitment to the MEIS1 promoter and activating MEIS1-mediated apoptosis.

Discussion
In this study, we report that RE-IIBP has HMTase activity against H3K79. It was discovered that RE-IIBP induces apoptosis through transcriptional activation of MEIS1 via its H3K79 HMTase activity. RE-IIBP interacts with RNF20, and H2BK120 ubiquitination by RNF20 further increases the RE-IIBP-mediated H3K79 methylation. Interaction with RNF20 increases the recruitment of RE-IIBP to the MEIS1 promoter, suggesting that RNF20 has a synergistic effect on H3K79 methylation and the activation of target genes (Supplementary Figure S4F).
MMSET and its isoform RE-IIBP are known for their diverse lysine specificities towards H3K27, H3K36 and H4K20 methylation 10,11,14,15 . In addition, MMSET increases H3K4me3 in vitro, although this has not significantly been confirmed in vivo, indicating their wide repertoire of lysine specificities 29 . Recent study indicates that carboxyl terminal domain (CTD) of NSD family (NSD1, NSD2/MMSET/ RE-IIBP, and NSD3/WHSC1L) can recognize and methylate H3K4, H3K9, H3K27, H3K36, H3K79, and H4K20 in vitro 30 . RNF20 targets H2BK120 monoubiquitination and stimulates the H3K79 methylation activity of DOT1L 31,32 . Similar histone crosstalk also applies to RE-IIBP-mediated H3K79 methylation and H2BK120 ubiquitination by RNF20. Using confocal and ChIP analysis, it was demonstrated herein that H2BK120 ubiquitination is prerequisite for H3K79 methylation by RE-IIBP, suggesting that the H3K79 methylation and H2BK120ub crosstalk might be universal.
We previously reported that RE-IIBP was overexpressed in leukemia cell lines and leukemia patient samples 10 . It is intriguing that screening of RE-IIBP target genes identified both up-and down-regulated genes. GATA1 and HOXA9 genes were repressed and MEIS1 gene was activated by RE-IIBP in leukemia cell line 17 . It is possible that RE-IIBP may play dual role in regulating transcription of target genes using H3K27 methylation and H3K79 methylation depends on specific condition. On the other hand, MLL fusion proteins recruit DOT1L to MLL target genes loci resulting in H3K79 methylation increase and leads to aberrant expression of a characteristic set of genes including HOXA9 and MEIS1 that drive leukemogenesis 33,34 In this study, we report that RE-IIBP induces MEIS1-mediated apoptosis via H3K79 methylation and H2BK120 ubiquitination by RNF20 give synergistic effect to this process. It is possible that RE-IIBP maintains cells from undergoing transformation through the MEIS1-mediated apoptosis induction mechanism in normal stages but has other functions during abnormal stages, such as transformation. Dot1-mediated H3K79 methylation plays an important role in cell cycle regulation, DNA damage response, and transcription. Furthermore, mammalian DOT1L is an essential gene required for embryogenesis, hematopoiesis, and cardiac function 24 . In HeLa cells, there are high levels of H3K79me2 in G1 phase, which decrease in S phase, reach the lowest levels in G2 phase, and increase again in the M phase 35 . Our observations suggest that RE-IIBP, as another H3K79 HMTase, has a possible role in these biological pathways and in different cell cycles. The relationship between the only known H3K79 HMTase DOT1L and RE-IIBP in these processes will require more detailed investigation. Along the way, Scientific RepoRts | 5:12485 | DOi: 10.1038/srep12485 the identification of an H3K79 demethylase will further extend the understanding of the physiological role of H3K79 methylation. The plasmids pCDNA3.1-RE-IIBP-His, pCDNA6-RE-IIBP-HA, GAL4-RE-IIBP,   GAL4-RE-IIBP R477A, GFP-RE-IIBP and GST-RE-IIBP were previously described 10 . The plasmid pcD-NA3.1-RNF20-myc-His was previously described 36 . RNF20, RNF20Δ RING (1-919 a.a), RNF20Δ 2 (1-750 a.a), and RNF20Δ 1 (1-300 a.a) coding sequences were subcloned into the bacterial expression vector pGEX-4T1 (Amersham Biosciences), thereby generating constructs for the production of glutathione S-transferase (GST)-tagged fusion proteins. RNF20Δ RING (1-919 a.a) coding sequences were subcloned into the mammalian expression vector pCDNA6 (Invitrogen). The MEIS1 promoter region (− 1345/+ 238) was amplified from human genomic DNA using primer pairs (Supplementary Table S1) and inserted into the KpnI/HindIII sites of the pGL3-basic vector (Promega). The short hairpin RNAs (shRNA) against human MMSET, RE-IIBP, DOT1L were designed using the siRNA sequence designer software (Clontech). A double-stranded oligonucleotide for shRNA plasmid construction was produced using primers from the 5' to the 3' end (Supplementary Table S1). These oligonucleotides were inserted into the AgeI/EcoRI site of the pLKO.1 TRC vector. Histone Purification. Acid extraction of histones was performed as described in the literature 38 .

Antibodies. Antibodies against
Briefly, approximately 5 × 10 6 cells were collected and washed once with PBS, then the cell pellet was resuspended in 1 ml of TEB lysis buffer (0.5% Triton X-100, 2 mM PMSF, and 1X protease inhibitor cocktail) and incubated for 30 min at 4 °C to promote hypotonic swelling. The cells were then recovered by centrifugation, resuspended in 400 μ l of 0.2 M H 2 SO 4 , and incubated overnight on a rotator at 4 °C. After centrifugation, the supernatant was transferred and was treated with 132 ul 100% TCA for 30 min at 4 °C. The pellet was recovered by centrifugation, and washed with acetone, after which it was resuspended in deionized water.
HMTase Assay. HMTase assays were carried out at 30 °C for 2 h in 30 ul volumes containing 50 mM Tris-HCl [pH 8.5], 20 mM KCl, 10 mM MgCl 2 , 10 mM beta-mercaptoethanol, 1.25 M sucrose, 100 μ M cold-S-adenosyl-methionine (SAM) (Sigma) or 100 nCi of 14 C-SAM (Perkin Elmer) and 1 μ g core histones form calf thymus (Sigma) or nucleosome or histone peptides, and 2 μ g of GST-RE-IIBP and GST. Peptides (H3N1, H3N2, H3N3, and H3N5) were synthesized based on the N-terminal amino acid sequences of H3 histone (Peptron). Proteins were separated via 15% sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) and analyzed by immunobloting with the indicated antibodies. The peptides were filtered using p81 filter paper (Upstate) and washed three times with cold 10% trichloroacetic acid (TCA) and 95% ethanol for 5 min at room temperature. The filters were allowed to air dry, after which 2 ml of Ultima Gold (Perkin Elmer) was added, and the 14 C-SAM was quantified using a scintillation counter. The HMTase assay, using cell lysates from RE-IIBP transiently transfected cells, was analyzed by immunobloting with the indicated antibodies.
Immunocytochemistry. 293T and H1299 cells were seeded in slides which were covered with collagen.
Next day, cells were transfected with indicated plasmids. Cells were then washed with phosphate-buffered saline (PBS) and fixed with 4% paraformaldehyde and permeabilized with 0.2% triton X-100. After blocking with 1% BSA, cells were incubated with indicated antibodies, followed by incubation with FITC-conjugated anti-mouse, and Cy3-conjugated anti-rabbit antibodies (Jackson ImmunoResearch Laboratories). Finally, slides were mounted with GEL/MOUNT (Biomeda), and images were obtained by confocal laser scanning microscopy (Zeiss LSM 700).
Immunoprecipitation and Mass Spectrometry. The 293T cells were transfected with GFP-RE-IIBP.
After 48 h, cell lysates were immunopreicipitated with anti-GFP antibody. Protein A/G agarose beads were then added and rotated for 2 h at 4 °C. Bound proteins were resolved on SDS-PAGE, and visualized with silver staining. The eluted gel was subjected to LC-MS/MS sequencing and data analysis at the Korea Basic Science Institute. For the RNF20 interaction assay, the ectopically expressed GFP-RE-IIBP and Scientific RepoRts | 5:12485 | DOi: 10.1038/srep12485 myc-RNF20 cell lysates were immunoprecipitated with anti-GFP and anti-myc antibodies. Protein A/G agarose beads (GenDEPOT) were then added and rotated for 2 h at 4 °C. Bound proteins were analyzed by immunoblotting with anti-GFP and anti-myc antibodies.
Luciferase Assay. Luciferase assays were conducted using a MEIS1 promoter reporter system. 293T cells were co-transfected with the MEIS1 promoter reporter construct and the indicated DNA constructs using PEI. Cells were harvested after 48 h and assayed for luciferase activity using a luciferase assay system (Promega). Each value was expressed as the mean of five replicates from a single assay. All experiments were performed at least three times.
Reverse Transcription and Real-time PCR. Total RNA was isolated from transfected 293T cells using RNAiso Plus (TaKaRa). After cDNA synthesis, the cDNA was quantified and then subjected to analysis of mRNA expression. The PCR primers used are presented in Supplementary Table S1. Dissociation curves were examined after each PCR run to ensure amplification of a single product of the appropriate length. The mean threshold cycle (C T ) and standard error values were calculated from individual C T values obtained from triplicate reactions per stage. The normalized mean C T value was estimated as Δ C T by subtracting the mean C T of β -actin. The value Δ Δ C T was calculated as the difference between the control Δ C T and the values obtained for each sample. The n-fold change in gene expression, relative to an untreated control, was calculated as 2 -ΔΔCT .
Liquid Chromatography-Mass Spectrometry. Synthetic peptides (H3N5) (100 mM) were used as substrates in the HMTase assay with GST or GST-RE-IIBP. The reaction was stopped with 10% TCA precipitation for 10 min at 4 °C. After removing the precipitates by centrifugation for 10 min at 4000 rpm, the supernatants were retrieved and the methylated peptides were analyzed by liquid chromatographymass spectrometry (LC-MS), carried out at the Korea Basic Science Institute. The eluted peptides were separated on a Luna column (C18 PepMap 100, 150 × 1 mm 5 micron) with a linear gradient (A: 5% ACN, 0.1% formic acid; B: 95% ACN, 0.1% formic acid) at a flow rate of 50 ml/min. Typically, 5 ml of sample was injected. Mass spectrometry was performed on a linear ion trap mass spectrometer (LCQ DECA XP, Thermo Finnigan) coupled to a nano-LC system (NANOSPACE SI-2, Shiseido). The MS scan range was 160-2000 m/z. Chromatin Immunoprecipitation Analysis. ChIP analysis was performed as described in the literature 10 . Briefly, 293T and the DOT1L knocked-down stable 293T cells were transfected with indicated plasmids. All cells were harvested after 48 h. The cells were cross-linked with the addition of 1% formaldehyde to the medium for 10 min at 37 °C, followed by the addition of 125 mM glycine for 5 min at room temperature. The cells were then lysed in SDS lysis buffer, and the samples were sonicated and immunoprecipitated using the indicated antibodies. The immunoprecipitates were eluted and reverse cross-linked. The DNA fragments were then purified and PCR amplified for quantification using each PCR primer pair (Supplementary Table S1). Disassociation curves were generated after each PCR run to ensure amplification of a single product of the appropriate length. The mean threshold cycle (C T ) and standard error values were calculated from individual C T values, obtained from duplicate reactions per stage. The normalized mean C T value was estimated as Δ C T by subtracting the mean C T of the input from that of MEIS1.

FACS Analysis.
To measure the effect of RE-IIBP on apoptosis, K562 shRE-IIBP stable cells treated 5 μ M etoposide (SIGMA). Also, K562 cells were transfected with indicated plasmids using Lipofectamine