WO2009114011A1 - Histone demethylation proteins and methods of use thereof - Google Patents

Abstract

UTX and JMJD3 act as histone H3 demethylases, acting specifically on H3K27. Histone H3K27 methylation is critical for repression of gene expression of HOX and other genes. Misregulation of Η3K27 methylation is associated with diseases including cancer.

Description

HISTONE DEMETHYLATION PROTEINS AND METHODS OF USE THEREOF

[0001] GOVERNMENT INTEREST

This invention was made using funds from grants from the U.S. National Institutes of Health. The U.S. government therefore retains certain rights in the invention.

[0002] TECHNICAL FIELD OF THE INVENTION

This invention is related to the areas of post-translational modifications and gene regulation. In particular, it relates to the area of modification of chromosome structure, specifically histone methylation, as a means of regulating gene transcription.

[0003] BACKGROUND OF THE INVENTION

Histone amino-terminal tails are subject to multiple post-translational modifications including methylation that determine chromatin structure and regulate gene transcription. Methylation of histone H3, particularly H3K27 methylation, has been recognized as important in epigenetic repression of transcription. H3K27 methylation is catalyzed by the histone methyltransferase EZH2, a mammalian homolog of Drosophila polycomb group protein Ez (enhancer of zeste). H3K27 methylation regulates transcription of tumor suppressor genes and has been implicated in tumorigenesis and cancer development. H3K27 methylation also impacts stem- cell pluripotency regulation. Thus, there is a continuing need in the art to regulate histone methylation status in order to control gene expression.

[0004] SUMMARY OF THE INVENTION

[0005] These and other embodiments which will be apparent to those of skill in the art upon reading the specification provide the art with reagents and methods for drug screening and therapy relating to histone methylation, neurological diseases and cancer.

[0006] BRIEF DESCRIPTION OF THE DRAWINGS

[0007] Figure 1 demonstrates histone demethylation mediated by the UTX. (A) Demethylation of

H3K27me3 by UTX. Tri-methylated synthetic histone peptides were incubated with UTX purified from Sf9 cells. Demethylated peptides were detected as mass peaks with the molecular weight that is 14 daltons (Da) smaller than the input. Demethylated peptides are marked with stars. Among the tri-methylated residues examined (H3K4, 9, 27, and 36, and H4K20), UTX only catalyzed demethylation of H3K27me3. (B) UTX demethylates H3K27me3/2 on bulk histones. Calf thymus histones were incubated with purified, full-length UTX, and subjected to Western blot analysis using antibodies that specifically recognize methylated histones. Mock-purified material was used as a control. Reduced signals were found only for H3K27me3 and H3K27me2. (C) JMJD3 catalytic domain demethylates H3K27me3/2 on nucleosomal substrates. Mono-nucleosomes purified from HeLa cells were incubated with purified catalytic domain of JMJD3 (aa 1164-1682). The demethylation of H3K27me3/2 was detected by Western blotting. (D) UTX and JMJD3 but not UTY catalytic domains mediate demethylation of H3K27me3/2/l histone peptides. Methylated H3K27 histone peptides were incubated with 2 Dg of the catalytic domains of UTX (aa 919-1401), JMJD3 (aa 1164-1682) and UTY (aa 866-1347) purified from bacteria. Demethylated peptides were detected as mass peaks with the molecular weights correspond to the loss of either one (14 Da), two (28 Da) or three (42 Da) methyl groups. Demethylated peptides are marked with stars. (E) Demethylation of H3K27me3 and K27me2 on bulk histones by the catalytic domains of UTX and JMJD3. Calf thymus histones were incubated with purified, catalytic domains of UTX, JMJD3 and UTY, respectively, and subjected to Western blot analysis using antibodies that specifically recognize methylated histone on H3K4, 9, 27 and H4K20. Decreased methylation was found only for H3K27me3 and H3K27me2 but not for trimethylated H3K4, H3K9, H3K36 or H4K20.

[0008] Figure 2 demonstrates that UTX regulates H3K27me3 at the Hox gene locus (A) Schematic diagram of the mammalian HoxD gene cluster. The primer sets used for ChIP are shown. The direction of transcription of the Hox genes shown in the diagram is from right to left. (B) Specific shRNAs effectively knocked down the transcripts of UTX and JMJDi in HeLa cells. (C) ChIP analysis of H3K27me3 at the HoxD gene cluster. ChIP results with standard deviation using the H3K27me3 antibody and the primer sets in (A) are shown. ChIP results are expressed as relative fold of enrichment over input, by comparing the Hox gene locus with a control ChIP of an unaffected region.

[0009] Figure 3 demonstrates that UTX binds near Hox gene start sites to facilitate H3K27 demethylation and gene activation. (A) UTX occupancy over -100 kilobases of the HoxA locus in lung fibroblasts. High confidence UTX occupancy sites (FDR <0.01, ChIP/IgG > 1.5) are shown as tick marks over the locus. Hox transcriptional start sites and locus-wide profiles of H3K27me3, PRC2 (i.e. Suzl2), H3K4me2, and RNA pol II occupancy (shown as Iog2 ratios of ChIP/input), and transcription are shown for comparison. (B) Hox genes represent -20% of probes on the Hox tiling array but account for 92% of UTX binding events (p<10-64). (C) Average UTX occupancy profiles for Hox genes bound by UTX aligned by Hox transcriptional start sites (left axis). Average of H3K27me3 occupancy profiles of Hox genes bound by both UTX and H3K27me3 (right axis). The genes and ChIP profiles used are detailed in Table 1. (D) UTX is excluded from Hox loci of ES cells. Occupancy of UTX, PRC2, and H3K27me3 in ES cells and differentiated fibroblasts is shown as a matrix; red indicates occupancy. (E) siRNA-mediated depletion of UTX decreases HoxA9 transcription. Relative transcript levels (mean + s.e.) by Taqman qRT-PCR are shown. Top: corresponding UTX and actin protein levels.

[0010] Figure 4 demonstrates that knockdown of zebrafish UTXl results in improper development of the posterior trunk. A). Prior to 24 hours post- fertilization (hpf) development of UTX morpholino (MO) injected embryos is slightly delayed. After 48hpf, lack of posterior extension is obvious and the embryos have slightly larger heads than controls. B). Evidence of notochord degeneration (between arrows) can be observed at 72hpf between approximately somites 13 to 23. However, the neural floor plate in this area (arrowheads) remains. The somites in the affected region are very short and compact. In most knockdown embryos, the most distal portion of the notochord (between somites 24-30) remains fairly normal. C) and D) Injection of human UTX mRNA together with zUTX MO results in a partial rescue of the truncation phenotype where the posterior notochord does not degenerate and the somites appear relatively normal (P=5.1E-14). However, injection of the catalytic inactive dead version of hUTX Hl 126A (hUTXc.d.) fails to significantly rescue the knockdown phenotype (P=O.26). Error bars represent s.e.m. calculated from 50 embryos of each group. E) Quantitative RT-PCR analysis shows clear, though modest, reduction of several posterior fish Hox genes. F) Whole mount zHoxC8a in situ experiment showing reduction of zHoxC8a transcript level in 6 out of 10 zUTXl morphants and posterior shift of zHoxC8a, which was observed in 4 of the 10 embryos. The control morphant has stronger staining across somite 1-7, but in the zUTXl morphants the staining shifted to somite 2-8. Arrows point to the boundary between 1st and 2nd somites. Purple color indicates zHoxC8a signal while orange color demarcates the somite boundaries (Dystrophin).

[0011] Figure 5 shows a phylogenic tree of the UTX and JMJD3 family proteins. All Amino acid sequences are retrieved from NCBI GenBank, except for ENSDART00000088452, which is from Ensembl database. The conserved domains are identified by the NCBI conserved domain search. The scale bar denotes 100 amino acids (aa).

[0012] Figure 6 demonstrates that over-expression of JMJD3 results in reduction of H3K27me3 and H3K27 me2 signal in 293T cells. Expression construct carrying HA-JMJD3 was transiently transfected into 293T cells, stained by anti-HA (green) and appropriate antibodies against methylated histone (H3K27me3, H3K27me2, H3K27mel, and H3K36me3; red). The nuclei were counter-stained by Hoechst (blue). The percentages of change in cell numbers with moderate to high HA-JMJD3 expression were listed in the Table. (A-H) Moderate to high level of over- expression of HA-tagged JMJD3 causes decreased levels of H3K27me3 and H3K27me2 (indicated with arrows). (I-L) In HA-JMJD3-transfected cells, the level of H3K27mel was either unchanged (indicated with arrows) or increased (indicated with arrowheads). (M-P) Over-expression of HA- JMJD3 had no effects on H3K36me3. (Table): JMJD3 over-expression decreased H3K27me3 and H3K27me2 levels in 293T cells. Cells that expressed moderate to high levels of JMJD3 were scored for the changes in signals of H3K27me3, H3K27me2, H3K27mel and H3K36me3.

[0013] Figure 7 demonstrates that UTX binding to the Hox loci and relationship with H3K4 methylation. (A) UTX occupancy over -100 kilobases of the HoxA locus in foot fibroblasts. High confidence UTX occupancy sites (FDR <0.01, ChIP/IgG > 1.5) are shown as tick marks over the locus. Hox transcriptional start sites and locus-wide profiles of raw UTX chlP-chip hybridization signal (shown as Iog2 ratios of UTX ChIP/IgG ChIP) are shown for comparison. Note that at the locus-wide scale it is difficult to resolve true UTX binding peaks (which are supported by strong hybridization signal over multiple contiguous probes, such as at HoxA9) versus spurious noise (which typically show high signal in a single probe but not by other neighboring probes, such as at HoxA3). The ability to examine ChIP signals from multiple contiguous probes is a significant technical advantage of ChlP-chip, and our analysis algorithm uses a permutation-based strategy to identify high confidence binding peaks versus spurious peaks. While both true and spurious peaks look rather similar at low resolution, they are resolved by peak calling and by detailed analysis at high resolution views. (B) Validation of UTX occupancy by qPCR. ChIP signals are presented as the percentage of input DNA (mean + standard deviation). Three predicted UTX occupancy events at the start of HoxA9 and HoxD13 in two different cell types show strong enrichment over IgG chIP while a predicted negative control in the intergenic region between HoxA9 and HoxAlO show no such enrichment. (C) Comparison of average UTX occupancy profiles versus H3K4me2 profiles in the Hox loci. Hox genes are aligned by their transcriptional start sites. ChlP-chip profiles used for this analysis are detailed in Table 1.

[0014] Figure 8 shows the identification of UTX occupancy sites in mES cells. (A) Summary of

UTX occupancy genome-wide. We focused our analysis on 21,600 promoters corresponding to well-annotated genes. The most significant class of UTX-occupied genes was the odorant receptor genes (OR). Of the 454 high confidence occupancy sites (> 1.5 fold enrichment over IgG ChIP and FDR<0.05), 81 UTX occupancy sites were in genes encoding OR. (B) Gene Ontology enrichment of UTX occupied genes in mES cells. Consistent with the over-representation of odorant receptor genes, the most significant p-values were for biological processes associated with the sensory perception of smell.

[0015] Figure 9 demonstrates that the UTX H1226A catalytic domain is inactive towards

H3K27me3/2/l histone peptides. The critical His 1226 residue involved in Fe2+ chelating was changed to Ala by site direct mutagenesis, and this mutant protein was purified and assayed under the same condition as the wt catalytic domain in Figure ID. No demethylated peptides were detected as mass peaks with the molecular weight that is 14 Da (removal of one methyl group) smaller than the input. Comparable amount of the wildtype UTX (analyzed alongside the mutant) showed clear demethylation activity (not shown). Note, the small peaks are background, and they are not 14Da less than the input.

[0016] Figure 10 provides a whole mount in situ experiment showing mis-regulation of zHox genes in UTXl morphants at 36 hpf. (a) Loss of expression zHoxA9b in the pectoral fin buds in the zUTXl morphants (5 out of 12 embryos). Arrows point to the zHoxA9b gene signal in the pectoral fin bud staining in the control morphants, which was lost in the zUTXl morphants (box arrowheads), (b) Mis-regulation of zHoxC 12b in the zUTXl morphants. Ectopic expression of zHoxC12b was observed in the zUTXl morphant embryos (5 out of 16, middle panel, marked by arrows), and reduced zHoxC12b expression in the tails (6 out of 16, right panel, marked by box arrowheads), (c) zHoxD9x expression is shifted posteriorly by 2 somites in the zUTXl morphants. Arrows demarcate the boundary between of somites 9 and 10, which was visualized by Dystrophin staining, (d) Reduced zHoxClla expression in the zUTXl morphants. zHoxClla staining is significantly stronger in the control morphants in the tail regions (arrows) than in the zUTXl morphants (box arrowheads), (e) Reduced expression of zHoxDl 2a in the pectoral fin buds in zUTXl morphants, which was observed in 5 out of 10 embryos. Arrows mark the pectoral fin bud staining in the control morphants, which was significantly reduced in the zUTXl morphants (box arrowheads), while the overall tail staining was comparable between the control and the zUTXl morphants.

[0017] Figure 11 shows the quantitative ChIP analyses of UTX, ALR, RBQ3 and H3K4me3 at selected Hox loci in lung fibroblasts. Co-localization of UTX, ALR and RBQ3, as well as H3K4me3, was identified at HoxA9, AlO and Dl 3 transcription start sites, but not at other neighboring regions. ChIP signals are presented as fold of enrichment (mean + standard deviation).

[0018] Figure 12 illustrates that no global change of H3K27me and H3K4me3 levels occurs in

UTX knocking down cells and no in vitro activity change of immunoprecipitated ALR/MLL2 from UTX knocking down cells. (A) UTX in mES cells and 293T cells were knocked down by siRNAs and shRNAs, respectively. Whole cell lysates were subjected to Western blotting analyses by indicated antibodies recognizing specific histone modifications. (B) ALR/MLL2 was immunoprecipitated from 293T cells treated with control and UTX shRNAs, and bulk histone and histone H3 peptides were used as substrates in the in vitro HMT assay.

[0019] DETAILED DESCRIPTION OF THE INVENTION

[0020] Studies have identified JmjC domain-containing proteins as histone demethylases that mediate the reversal of methylation at histone H3K4, H3K9 and H3K36. UTX, UTY and JMJD3 comprise a subfamily of JmjC domain-containing proteins, which are evolutionarily conserved from C. elegans to human. In addition to the JmjC domain, UTX and UTY, but not JMJD3, also contain tetratricopeptide repeats (TPR) at their N-terminal regions, which are predicted protein interaction motifs (Figure 5). All three proteins contain a Treble-clef zinc finger at their C-terminus. UTX resides on the X chromosome, escapes X-inactivation and is ubiquitously expressed. UTY is a male-specific protein and may contribute to sex-specific tissue transplantation rejection response. It is a discovery of the present inventors that UTX and JMJD3 function as transcriptional activators of gene expression. Using recombinant UTX and a collection of methylated histone peptides as substrates, it has been determined that UTX specifically mediates the methylation status of H3K27, particularly H3K27me3. See Lan et al. Nature (2007) 449: 689-94, the contents of which are hereby incorporated by reference in their entirety.

[0021] The terms "histone demethylase protein," and " JmjC domain-containing protein" (and similar terms) as used herein encompass any JmjC domain-containing demethylase (including histone demethylases), which includes without limitation proteins in the UTX, UTX and JMJD3 families, and further includes variants, functional fragments and "catalytic analogs" of any of the foregoing that retain substantial demethylase activity (e.g., at least about 5%, 10%, 25%, 50%, 60%, 75%, 80%, 85%, 90%, 95%, 100% or more demethylase activity as compared with the native protein). UTX and JMJD3 are preferred histone demethylase proteins. See, e.g., the Examples, Figures, and the protein and nucleic acid sequences of UTX and JMJD3 found at NCBI Accession Nos. NP_066963, NP_001073893, XP_043272, XM_941783, 015550, 070546, NP_872601, NP 872600, and NP 009056, as well as homologues and orthologs thereof, including but not limited to homologues from mammals (e.g., rat, mouse), Xenopus, D. rerio, C. elegans, S. pombe and S. cerevisiae. See, e.g., NCBI Accession Nos. NM_001080193, ENSDART00000088462, and XP_684619. UTY sequences include BC071744, NM_001093305, NM_001009002, EF491796, EF491815, NM_009484, NM_007125, NM_182659, and NM_182660. The amino acid sequence of human UTX is provided in Accession No. NP 066963 is shown below as SEQ ID NO: 1. mkscgvslat aaaaaaafgd eekkmaagka sgeseeasps ltaeerealg gldsrlfgfv rfhedgartk allgkavrcy eslilkaegk vesdffcqlg hfnllledyp kalsayqryy slqsdywkna aflyglglvy fhynafqwai kafqevlyvd psfcrakeih lrvglmfkvn tdyesslkhf qlalvdcnpc tlsnaeiqfh iahlyetqrk yhsakeayeq llqtenlsaq vkatvlqqlg wmhhtvdllg dkatkesyai qylqkslead pnsgqswyfl grcyssigkv qdafisyrqs idkseasadt wcsigvlyqq qnqpmdalqa yicavqldhg haaawmdlgt lyescnqpqd aikcylnatr skscsntsal aarikylqaq lcnlpqgslq nktkllpsie eawslpipae ltsrqgamnt aqqntsdnws gghavshppv qqqahswclt pqklqhleql ranrnnlnpa qklmleqles qfvlmqqhqm rptgvaqvrs tgipngptad sslptnsvsg qqpqlaltrv psvsqpgvrp acpgqplang pfsaghvpcs tsrtrgstdt ilignnhitg ngsngnvpyl qrnaltlphn rtnltssake pwknqlsnst qglhkgqssh sagpngerpl sstgpsqhlq aagsgiqnqn ghptlpsnsv tqgaalnhls shtatsggqq gitltkeskp sgniltvpet srhtgetpns tasveglpnh vhqmtadavc spshgdsksp gllssdnpql sallmgkann nvgtgtcdkv nnihpavhtk tdnsvassps saistatpsp ksteqtttns vtslnsphsg lhtingegme esqspmktdl llvnhkpspq iipsmsvsiy pssaevlkac rnlgknglsn ssilldkcpp prppsspypp lpkdklnppt psiylenkrd affpplhqfc tnpnnpvtvi rglagalkld lglfstktlv eannehmvev rtqllqpade nwdptgtkki whcesnrsht tiakyaqyqa ssfqeslree nekrshhkdh sdsestssdn sgrrrkgpfk tikfgtnidl sddkkwklql heltklpafv rvvsagnlls hvghtilgmn tvqlymkvpg srtpghqenn nfcsvninig pgdcewfvvp egywgvlndf ceknnlnflm gswwpnledl yeanvpvyrf iqrpgdlvwi nagtvhwvqa igwcnniawn vgpltacqyk laveryewnk lqsvksivpm vhlswnmarn ikvsdpklfe mikycllrtl kqcqtlreal iaagkeiiwh grtkeepahy csicevevfd llfvtnesns rktyivhcqd carktsgnle nfvvleqykm edlmqvydqf tlapplpsas s The nucleic acid sequence of human UTX is provided in Accession No. NM 021140 is shown below as SEQ ID NO: 2. aaagcaaaag aattcgctgc gtttccatga aatcctgcgg agtgtcgctc gctaccgccg ccgctgccgc cgccgctttc ggtgatgagg aaaagaaaat ggcggcggga aaagcgagcg gcgagagcga ggaggcgtcc cccagcctga cagccgagga gagggaggcg ctcggcggac tggacagccg cctctttggg ttcgtgagat ttcatgaaga tggcgccagg acgaaggccc tactgggcaa ggctgttcgc tgctatgaat ctctaatctt aaaagctgaa ggaaaagtgg agtctgattt cttttgtcaa ttaggtcact tcaacctctt attggaagat tatccaaaag cattatctgc ataccagagg tactacagtt tacagtctga ctactggaag aatgctgcct ttttatatgg tcttggtttg gtctacttcc attataatgc atttcagtgg gcaattaaag catttcagga ggtgctttat gttgatccca gcttttgtcg agccaaggaa attcatttac gagttgggct tatgttcaaa gtgaacacag actatgagtc tagtttaaag cattttcagt tagctttggt tgactgtaat ccctgcactt tgtccaatgc tgaaattcaa tttcacattg cccacttata tgaaacccag aggaaatatc attctgcaaa agaagcttat gaacaacttt tgcagacaga gaatctttct gcacaagtaa aagcaactgt cttacaacag ttaggttgga tgcatcacac tgtagatctc ctgggagata aagccaccaa ggaaagctat gctattcagt atctccaaaa gtccttggaa gcagatccta attctggcca gtcctggtat ttcctcggaa ggtgctattc aagtattggg aaagttcagg atgcctttat atcttacagg cagtctattg ataaatcaga agcaagtgca gatacatggt gttcaatagg tgtgctatat cagcagcaaa atcagcccat ggatgcttta caggcctata tttgtgctgt acaattggac catggccatg ctgcagcctg gatggaccta ggcactctct atgaatcctg caaccagcct caggatgcca ttaaatgcta cttaaatgca actagaagca aaagttgtag taatacctct gcacttgcag cacgaattaa gtatttacag gctcagttgt gtaaccttcc acaaggtagt ctacagaata aaactaaatt acttcctagt attgaggagg cgtggagcct accaattccc gcagagctta cctccaggca gggtgccatg aacacagcac agcagaatac ttctgacaat tggagtggtg gacatgctgt gtcacatcct ccagtacagc aacaagctca ttcatggtgt ttgacaccac agaaattaca gcatttggaa cagctccgcg caaatagaaa taatttaaat ccagcacaga aactgatgct ggaacagctg gaaagtcagt ttgtcttaat gcaacaacac caaatgagac caacaggagt tgcacaggta cgatctactg gaattcctaa tgggccaaca gctgactcat cactgcctac aaactcagtc tctggccagc agccacagct tgctctgacc agagtgccta gcgtctctca gcctggagtc cgtcctgcct gccctgggca gcctttggcc aatggaccct tttctgcagg ccatgttccc tgtagcacat caagaacgcg gggaagtaca gacactattt tgataggcaa taatcatata acaggaaatg gaagtaatgg aaacgtgcct tacctgcagc gaaacgcact cactctacct cataaccgca caaacctgac cagcagcgca aaggagccgt ggaaaaacca actatctaac tccactcagg ggcttcacaa aggtcagagt tcacattcgg caggtcctaa tggtgaacga cctctctctt ccactgggcc ttcccagcat ctccaggcag ctggctctgg tattcagaat cagaacggac atcccaccct gcctagcaat tcagtaacac agggggctgc tctcaatcac ctctcctctc acactgctac ctcaggtgga caacaaggca ttaccttaac caaagagagc aagccttcag gaaacatatt gacggtgcct gaaacaagca ggcacactgg agagacacct aacagcactg ccagtgtcga gggacttcct aatcatgtcc atcagatgac ggcagatgct gtttgcagtc ctagccatgg agattctaag tcaccaggtt tactaagttc agacaatcct cagctctctg ccttgttgat gggaaaagcc aataacaatg tgggtactgg aacctgtgac aaagtcaata acatccaccc agctgttcat acaaagactg ataactctgt tgcctcttca ccatcttcag ccatttcaac agcaacacct tctccaaaat ccactgagca gacaaccaca aacagtgtta ccagccttaa cagccctcac agtgggctac acacaattaa tggagaaggg atggaagaat ctcagagccc catgaaaaca gatctgcttc tggttaacca caaacctagt ccacagatca taccatcaat gtctgtgtcc atatacccca gctcagcaga agttctgaag gcatgcagga atctaggtaa aaatggctta tctaacagta gcattttgtt ggataaatgt ccacctccaa gaccaccatc ttcaccatac cctcccttgc caaaggacaa gttgaatcca cctacaccta gtatttactt ggaaaataaa cgtgatgctt tctttcctcc attacatcaa ttttgtacaa atccgaacaa ccctgttaca gtaatacgtg gccttgctgg agctcttaag ttagacctgg gacttttctc tactaaaact ttggtggaag ctaacaatga acatatggta gaagtgagga cacagttgtt gcagccagca gatgaaaact gggatcccac tggaacaaag aaaatctggc attgtgaaag taatagatct catactacaa ttgctaaata tgcacagtac caggcctcct cattccagga atcattgaga gaagaaaatg aaaaaagaag tcatcataaa gaccactcag atagtgaatc tacatcgtca gataattctg ggaggaggag gaaaggaccc tttaaaacca taaagtttgg gaccaatatt gacctatctg atgacaaaaa gtggaagttg cagctacatg agctgactaa acttcctgct tttgtgcgtg tcgtatcagc aggaaatctt ctaagccatg ttggtcatac catattgggc atgaacacag ttcaactata catgaaagtt ccagggagca gaacaccagg tcatcaggaa aataacaact tctgttcagt taacataaat attggcccag gtgactgtga atggtttgtt gttcctgaag gttactgggg tgttttgaat gacttctgtg aaaaaaataa tttgaatttc ctaatgggtt cttggtggcc caatcttgaa gatctttatg aagcaaatgt tccagtgtat aggtttattc agcgacctgg agatttggtc tggataaatg caggcactgt tcattgggtt caggctattg gctggtgcaa caacattgct tggaatgttg gtccacttac agcctgccag tataaattgg cagtggaacg gtacgaatgg aacaaattgc aaagtgtgaa gtcaatagta cccatggttc atctttcctg gaatatggca cgaaatatca aggtctcaga tccaaagctt tttgaaatga ttaagtattg tcttctaaga actctgaagc aatgtcagac attgagggaa gctctcattg ctgcaggaaa agagattata tggcatgggc ggacaaaaga agaaccagct cattactgta gcatttgtga agtggaggtt tttgatctgc tttttgtcac taatgagagt aattcacgaa agacctacat agtacattgc caagattgtg cacgaaaaac aagcggaaac ttggaaaact ttgtggtgct agaacagtac aaaatggagg acctgatgca agtctatgac caatttacat tagctcctcc attaccatcc gcctcatctt gatattgttc catggacatt aaatgagacc ttttctgcta ttcaggaaat aacccagttc tgcaccactg gtttttgtag ctatctcgta aggctgctgg ctgaaaactg tgtctatgca accttccaag tgcggagtgt caaccaactg gacgggagag agtactgctc ctactccagg actctcacaa agctgatgag ctgtacttca gaaaaaaata ataatttcca tgttttgtat atatctgaca aaactggcaa catcttacag actactgact tgaagacaac ctcttttata tttctctatt tctgggctga tgaatttgtt ttcatctgtc ttttccccct tcagaatttt ccttggaaaa aaaatactag cctagctggt catttctttg taaggtagtt agcaatttta agtctttctt tggtcaactt ttttttaatg tgaaaagtta ggtaagacac ttttttactg cttttatgtt tttctgtctt gttttgagac catgatggtt acacttttgg ttcctaaata aaatttaaaa aattaacagc caagtcacaa aggtaatgga ttgcacatag actaaggaat aaacttcaga tttgtgattt ttgtttctaa tcttgatgta aatttacact attataatac atatttattg cttgaaaata tttgtgaatg gaatgctgtt attttttcca gatttacctg ccattgaaat tttaaggagt tctgtaattt caaacactac tcctattaca ttttctatgt gtaaataaaa ctgcttagca ttgtacagaa acttttatta aaattgttta atgtttaaag agtttctatt gtttgagttt aaaaaagact ttatgtacag tgcccagttt tgttcatttt gaaatctgat aaatatattt atatatactt atgtatgtat atataatata tatagaaatc tggatatata tgtataaatc tttagaactt aaatttttct cgtttagttc acatctatgg tagatttttg aggtgtctac tgtaaagtat tgcttacaaa aagtatgatt atttttaaag aaatatatat ggtatgtatc ctcaagacct aaaatgtcag actggtttat tgttaagttg caattactgc aatgacagac caataaacaa ttgctgccaa aaaaaaaa

The amino acid sequence of human JMJD3 is provided in Accession No. NP OO 1073893 is shown below as SEQ ID NO: 3.

mhravdppga raareafalg glscagawss cpphppprsa wlpggrcsas igqpplpapl ppshgsssgh pskpyyapga ptprplhgkl eslhgcvqal lrepaqpglw eqlgqlyese hdseeatrcy hsalryggsf aelgprigrl qqaqlwnfht gscqhrakvl ppleqvwnll hlehkrnyga krggppvkra aeppvvqpvp paalsgpsge eglspggkrr rgcnseqtgl ppglplpppp lppppppppp pppplpglat sppfqltkpg lwstlhgdaw gperkgsapp erqeqrhslp hpypypapay tahppghrlv paappgpgpr ppgaeshgcl patrppgsdl resrvqrsrm dssvspaatt acvpyapsrp pglpgtttss ssssssntgl rgvepnpgip gadhyqtpal evshhgrlgp sahssrkpfl gapaatphls lppgpssppp ppcprllrpp pppawlkgpa craaredgei leelffgteg pprpappplp hregflgppa srfsvgtqds htpptpptpt tsssnsnsgs hssspagpvs fppppylars idplprppsp aqnpqdpplv pltlalppap psschqntsg sfrrpesprp rvsfpktpev gpgpppgpls kapqpvppgv gelpargprl fdfpptpled qfeepaefki lpdglanimk mldesirkee eqqqheagva pqpplkepfa slqspfptdt aptttapava vttttttttt ttatqeeekk pppalppppp lakfpppsqp qpppppppsp asllkslasv legqkycyrg tgaavstrpg plpttqyspg ppsgatalpp tsaapsaqgs pqpsassssq fstsggpwar errageepvp gpmtptqppp plslpparse sevleeisra cetlvervgr satdpadpvd taepadsgte rllppaqake eaggvaavsg sckrrqkehq kehrrhrrac kdsvgrrpre grakakakvp keksrrvlgn ldlqseeiqg reksrpdlgg askakpptap appsapapsa qptppsasvp gkkareeapg ppgvsradml klrslsegpp kelkirlikv esgdketfia seveerrlrm adltishcaa dvvrasrnak vkgkfresyl spaqsvkpki nteeklprek lnpptpsiyl eskrdafspv llqfctdprn pitvirglag slrlnlglfs tktlveasge htvevrtqvq qpsdenwdlt gtrqiwpces srshttiaky aqyqassfqe slqeekesed eeseepdstt gtppssapdp knhhiikfgt nidlsdakrw kpqlqellkl pafmrvtstg nmlshvghti lgmntvqlym kvpgsrtpgh qennnfcsvn inigpgdcew favhehywet isafcdrhgv dyltgswwpi lddlyasnip vyrfvqrpgd lvwinagtvh wvqatgwcnn iawnvgplta yqyqlalery ewnevknvks ivpmihvswn vartvkisdp dlfkmikfcl lqsmkhcqvq reslvragkk iayqgrvkde payycnecdv evfnilfvts engsrntylv hcegcarrrs aglqgvvvle qyrteelaqa ydaftlvrar rargqrrral gqaagtgfgs paapfpeppp afspqapast sr

The nucleic acid sequence of human JMJD3 is provided in Accession No. NM OO 1080424 is shown below as SEQ ID NO: 4. ggcaacatgc cagccccgta gcactgccca ccccacccac tgtggtctgt tgtaccccac tgctggggtg gtggttccaa tgagacaggg cacaccaaac tccatctggc tgttactgag gcggagacac gggtgatgat tggctttctg gggagagagg aagtcctgtg attggccaga tctctggagc ttgccgacgc ggtgtgagga cgctcccacg gaggccggaa ttggctgtga aaggactgag gcagccatct gggggtagcg ggcactctta tcagagcggc tggagccgga ccatcgtccc agagagctgg ggcagggggc cgtgcccaat ctccagggct cctggggcca ctgctgacct ggctggatgc atcgggcagt ggaccctcca ggggcccgcg ctgcacggga agcctttgcc cttgggggcc tgagctgtgc tggggcctgg agctcctgcc cgcctcatcc ccctcctcgt agcgcatggc tgcctggagg cagatgctca gccagcattg ggcagccccc gcttcctgct cccctacccc cttcacatgg cagtagttct gggcacccca gcaaaccata ttatgctcca ggggcgccca ctccaagacc cctccatggg aagctggaat ccctgcatgg ctgtgtgcag gcattgctcc gggagccagc ccagccaggg ctttgggaac agcttgggca actgtacgag tcagagcacg atagtgagga ggccacacgc tgctaccaca gcgcccttcg atacggagga agcttcgctg agctggggcc ccgcattggc cgactgcagc aggcccagct ctggaacttt catactggct cctgccagca ccgagccaag gtcctgcccc cactggagca agtgtggaac ttgctacacc ttgagcacaa acggaactat ggagccaagc ggggaggtcc cccggtgaag cgagctgctg aacccccagt ggtgcagcct gtgcctcctg cagcactctc aggcccctca ggggaggagg gcctcagccc tggaggcaag cgaaggagag gctgcaactc tgaacagact ggccttcccc cagggctgcc actgcctcca ccaccattac caccaccacc accaccacca ccaccaccac caccacccct gcctggcctg gctaccagcc ccccatttca gctaaccaag ccagggctgt ggagtaccct gcatggagat gcctggggcc cagagcgcaa gggttcagca cccccagagc gccaggagca gcggcactcg ctgcctcacc catatccata cccagctcca gcgtacaccg cgcacccccc tggccaccgg ctggtcccgg ctgctccccc aggcccaggc ccccgccccc caggagcaga gagccatggc tgcctgcctg ccacccgtcc ccccggaagt gaccttagag agagcagagt tcagaggtcg cggatggact ccagcgtttc accagcagca accaccgcct gcgtgcctta cgccccttcc cggccccctg gcctccccgg caccaccacc agcagcagca gtagcagcag cagcaacact ggtctccggg gcgtggagcc gaacccaggc attcccggcg ctgaccatta ccaaactccc gcgctggagg tctctcacca tggccgcctg gggccctcgg cacacagcag tcggaaaccg ttcttggggg ctcccgctgc cactccccac ctatccctgc cacctggacc ttcctcaccc cctccacccc cctgtccccg cctcttacgc cccccaccac cccctgcctg gttgaagggt ccggcctgcc gggcagcccg agaggatgga gagatcttag aagagctctt ctttgggact gagggacccc cccgccctgc cccaccaccc ctcccccatc gcgagggctt cttggggcct ccggcctccc gcttttctgt gggcactcag gattctcaca cccctcccac tcccccaacc ccaaccacca gcagtagcaa cagcaacagt ggcagccaca gcagcagccc tgctgggcct gtgtcctttc ccccaccacc ctatctggcc agaagtatag acccccttcc ccggcctccc agcccagcac agaaccccca ggacccacct cttgtacccc tgactcttgc cctgcctcca gcccctcctt cctcctgcca ccaaaatacc tcaggaagct tcaggcgccc ggagagcccc cggcccaggg tctccttccc aaagaccccc gaggtggggc cggggccacc cccaggcccc ctgagtaaag ccccccagcc tgtgccgccc ggggttgggg agctgcctgc ccgaggccct cgactctttg attttccccc cactccgctg gaggaccagt ttgaggagcc agccgaattc aagatcctac ctgatgggct ggccaacatc atgaagatgc tggacgaatc cattcgcaag gaagaggaac agcaacaaca cgaagcaggc gtggcccccc aacccccgct gaaggagccc tttgcatctc tgcagtctcc tttccccacc gacacagccc ccaccactac tgctcctgct gtcgccgtca ccaccaccac caccaccacc accaccacca cggccaccca ggaagaggag aagaagccac caccagccct accaccacca ccgcctctag ccaagttccc tccaccctct cagccacagc caccaccacc cccacccccc agcccggcca gcctgctcaa atccttggcc tccgtgctgg agggacaaaa gtactgttat cgggggactg gagcagctgt ttccacccgg cctgggccct tgcccaccac tcagtattcc cctggccccc catcaggtgc taccgccctg ccgcccacct cagcggcccc tagcgcccag ggctccccac agccctctgc ttcctcgtca tctcagttct ctacctcagg cgggccctgg gcccgggagc gcagggcggg cgaagagcca gtcccgggcc ccatgacccc cacccaaccg cccccacccc tatctctgcc ccctgctcgc tctgagtctg aggtgctaga agagatcagc cgggcttgcg agacccttgt ggagcgggtg ggccggagtg ccactgaccc agccgaccca gtggacacag cagagccagc ggacagtggg actgagcgac tgctgccccc cgcacaggcc aaggaggagg ctggcggggt ggcggcagtg tcaggcagct gtaagcggcg acagaaggag catcagaagg agcatcggcg gcacaggcgg gcctgtaagg acagtgtggg tcgtcggccc cgtgagggca gggcaaaggc caaggccaag gtccccaaag aaaagagccg ccgggtgctg gggaacctgg acctgcagag cgaggagatc cagggtcgtg agaagtcccg gcccgatctt ggcggggcct ccaaggccaa gccacccaca gctccagccc ctccatcagc tcctgcacct tctgcccagc ccacaccccc gtcagcctct gtccctggaa agaaggctcg ggaggaagcc ccagggccac cgggtgtcag ccgggccgac atgctgaagc tgcgctcact tagtgagggg ccccccaagg agctgaagat ccggctcatc aaggtagaga gtggtgacaa ggagaccttt atcgcctctg aggtggaaga gcggcggctg cgcatggcag acctcaccat cagccactgt gctgctgacg tcgtgcgcgc cagcaggaat gccaaggtga aagggaagtt tcgagagtcc tacctttccc ctgcccagtc tgtgaaaccg aagatcaaca ctgaggagaa gctgccccgg gaaaaactca acccccctac acccagcatc tatctggaga gcaaacggga tgccttctca cctgtcctgc tgcagttctg tacagaccct cgaaatccca tcacagtgat ccggggcctg gcgggctccc tgcggctcaa cttgggcctc ttctccacca agaccctggt ggaagcgagt ggcgaacaca ccgtggaagt tcgcacccag gtgcagcagc cctcagatga gaactgggat ctgacaggca ctcggcagat ctggccttgt gagagctccc gttcccacac caccattgcc aagtacgcac agtaccaggc ctcatccttc caggagtctc tgcaggagga gaaggagagt gaggatgagg agtcagagga gccagacagc accactggaa cccctcctag cagcgcacca gacccgaaga accatcacat catcaagttt ggcaccaaca tcgacttgtc tgatgctaag cggtggaagc cccagctgca ggagctgctg aagctgcccg ccttcatgcg ggtaacatcc acgggcaaca tgctgagcca cgtgggccac accatcctgg gcatgaacac ggtgcagctg tacatgaagg tgcccggcag ccgaacgcca ggccaccagg agaataacaa cttctgctcc gtcaacatca acattggccc aggcgactgc gagtggttcg cggtgcacga gcactactgg gagaccatca gcgctttctg tgatcggcac ggcgtggact acttgacggg ttcctggtgg ccaatcctgg atgatctcta tgcatccaat attcctgtgt accgcttcgt gcagcgaccc ggagacctcg tgtggattaa tgcggggact gtgcactggg tgcaggccac cggctggtgc aacaacattg cctggaacgt ggggcccctc accgcctatc agtaccagct ggccctggaa cgatacgagt ggaatgaggt gaagaacgtc aaatccatcg tgcccatgat tcacgtgtca tggaacgtgg ctcgcacggt caaaatcagc gaccccgact tgttcaagat gatcaagttc tgcctgctgc agtccatgaa gcactgccag gtgcaacgcg agagcctggt gcgggcaggg aagaaaatcg cttaccaggg ccgtgtcaag gacgagccag cctactactg caacgagtgc gatgtggagg tgtttaacat cctgttcgtg acaagtgaga atggcagccg caacacgtac ctggtacact gcgagggctg tgcccggcgc cgcagcgcag gcctgcaggg cgtggtggtg ctggagcagt accgcactga ggagctggct caggcctacg acgccttcac gctggtgagg gcccggcggg cgcgcgggca gcggaggagg gcactggggc aggctgcagg gacgggcttc gggagcccgg ccgcgccttt ccctgagccc ccgccggctt tctcccccca ggccccagcc agcacgtcgc gatgaggccg gacgccccgc ccgcctgcct gcccgcgcaa ggcgccgcgg ggccaccagc acatgcctgg gctggaccta ggtcccgcct gtggccgaga agggggtcgg gcccagccct tccaccccat tggcagctcc cctcacttaa tttattaaga aaaacttttt tttttttttt agcaaatatg aggaaaaaag gaaaaaaaat gggagacggg ggagggggct ggcagcccct cgcccaccag cgcctcccct caccgacttt ggccttttta gcaacagaca caaggaccag gctccggcgg cggcgggggt cacatacggg ttccctcacc ctgccagccg cccgcccgcc cggcgcagat gcacgcggct cgtgtatgta catagacgtt acggcagccg aggtttttaa tgagattctt tctatgggct ttacccctcc cccggaacct ccttttttac ttccaatgct agctgtgacc cctgtacatg tctctttatt cacttggtta tgatttgtat tttttgttct tttcttgttt ttttgttttt aatttataac agtcccactc acctctattt attcattttt gggaaaaccc gacctcccac acccccaagc catcctgccc gcccctccag ggaccgcccg tcgccgggct ctccccgcgc cccagtgtgt gtccgggccc ggcccgaccg tctccacccg tccgcccgcg gctccagccg ggttctcatg gtgctcaaac ccgctcccct cccctacgtc ctgcactttc tcggaccagt ccccccactc ccgacccgac cccagcccca cctgagggtg agcaactcct gtactgtagg ggaagaagtg ggaactgaaa tggtattttg taaaaaaaat aaataaaata aaaaaattaa aggttttaaa gaaagaacta tgaggaaaag gaaccccgtc cttcccagcc ccggccaact ttaaaaaaca cagaccttca cccccacccc cttttctttt taagtgtgaa acaacccagg gccagggcct cactggggca gggacacccc ggggtgagtt tctctggggc tttattttcg ttttgttggt tgttttttct ccacgctggg gctgcggagg ggtggggggt ttacagtccc gcaccctcgc actgcactgt ctctctgccc caggggcaga ggggtcttcc caaccctacc cctattttcg gtgatttttg tgtgagaata ttaatattaa aaataaacgg agaaaaaaaa tcct A homolog of a protein of interest, such as UTX and JMJD3, includes proteins comprising or consisting of an amino acid sequence that has at least about 70%, 80%, 90%, 95%, 98% or 99% identity with the amino acid sequence of the proteins described herein. A homolog may also be a protein that is encoded by a nucleic acid that has at least about 70%, 80%, 90%, 95%, 98% or 99% identity with a nucleotide sequence described herein. A homolog may also be a protein that is encoded by a nucleic acid that hybridizes, e.g., under stringent hybridization conditions, to a nucleic acid consisting of a nucleotide sequence described herein or the coding sequence thereof. For example, homologs may be encoded by nucleic acids that hybridize under high stringency conditions of 0.2 to 1 x SSC at 65oC followed by a wash at 0.2 x SSC at 65oC to a nucleic acid containing a sequence described herein. Nucleic acids that hybridize under low stringency conditions of 6 x SSC at room temperature followed by a wash at 2 x SSC at room temperature to nucleic acid consisting of a sequence described herein or a portion thereof can be used. Other hybridization conditions include 3 x SSC at 40 or 50oC, followed by a wash in 1 or 2 x SSC at 20, 30, 40, 50, 60, or 65oC. Hybridizations can be conducted in the presence of formaldehyde, e.g., 10%, 20%, 30% 40% or 50%, which further increases the stringency of hybridization. Theory and practice of nucleic acid hybridization is described, e.g., in S. Agrawal (ed.) Methods in Molecular Biology, volume 20; and Tijssen (1993) Laboratory Techniques in biochemistry and molecular biology- hybridization with nucleic acid probes, e.g., part I, chapter 2 "Overview of principles of hybridization and the strategy of nucleic acid probe assays," (Elsevier, New York) provide a basic guide to nucleic acid hybridization.

[0023] Homologs of a protein of interest also include portions thereof, such as portions comprising one or more conserved domains, such as those described herein.

[0024] The demethylase proteins of the invention can be derived from any species of interest, including without limitation, mammalian {e.g., human, non- human primate, mouse, rat, lagomorph, bovine, ovine, caprine, porcine, equine, feline, and canine), insect (e.g., Drosophila), avian, fungal, plant, yeast {e.g., S. pombe or S. cerevisiae), C. elegans, D. rerio (zebrafish), etc. as well as allelic variations, isoforms, splice variants and the like. The demethylase sequences can further be wholly or partially synthetic.

[0025] A catalytic analog of UTX or JMJD3 may be a portion of the wild type UTX or JMJD3 protein including one or more of the conserved domains. A catalytic analog of UTX or JMJD3 may comprise at least a portion of the JmjC domain, the Treble-clef zinc finger domain, and/or the TPR repeat. In particular embodiments, a catalytic analog of a histone demethylase protein includes a JmjC domain and, optionally, further includes a JmjN domain, a zinc finger domain (e.g., a Treble- clef zinc finger motif), zinc finger-like domain, a PHD domain, a tetratricopeptide repeat (TPR), an FBOX domain, a Tudor domain, an AT-Rich Interactive Domain (Arid/Bright), a coiled coil motif and/or a Leucine Rich Repeat (LRR) domain. For a further description of other JmjC proteins, see Klose et al. (2006) Nature Reviews/Genetics 7:715-727. [0026] Generally, analogs can differ from naturally occurring proteins by conservative amino acid sequence differences or by modifications which do not affect sequence, or by both. Any number of procedures may be used for the generation of mutant, derivative or variant forms of a protein of interest using recombinant DNA methodology well known in the art such as, for example, that described in Sambrook et al. (1989, Molecular Cloning: A Laboratory Manual, Cold Spring Harbor Laboratory Press, New York) and Ausubel et al. (1997, Current Protocols in Molecular Biology, Green & Wiley, New York).

[0027] For example, conservative amino acid changes may be made, which although they alter the primary sequence of the protein or peptide, do not normally alter its function, e.g., its demethylase activity. Conservative amino acid substitutions typically include substitutions within the following groups: glycine, alanine; valine, isoleucine, leucine; aspartic acid, glutamic acid; asparagine, glutamine; serine, threonine; lysine, arginine (in positions other than proteolytic enzyme recognition sites); phenylalanine, tyrosine.

[0028] Whether an analog is a catalytic analog can be determined according to methods known in the art. For example, a demethylase activity can be determined as described in the Examples. An illustrative example for determining whether a demethylase analog has demethylase activity includes contacting the demethylase analog with a target peptide that is methylated, and determining whether the demethylase analog is capable of demethylating the target peptide. The assay may further comprise one or more other components, such as other proteins. A target peptide may be a histone peptide. Any histone peptide can be used. Preferably it is used with a histone demethylase enzyme that recognizes the histone peptide as a substrate. The full histone protein can be used or a peptide comprising only a portion of the histone protein can be used, so long as that portion contains the methylated residue upon which the demethylase enzyme acts and the portion contains sufficient contextual residues to permit its recognition by the enzyme. Typically at least 3, at least 4, at least 5, at least 6, or at least 7 residues on either side of the methylated residue are believed to be sufficient for recognition. The methylated residue is preferably a lysine. Preferably the histone peptide and the histone demethylase are derived from the same species of organism.

[0029] Measurement of the reaction between a histone and a eukaryotic histone demethylase protein can be accomplished by any means known in the art.

[0030] The term protein or histone "substrate" as used herein refers to a starting reagent in an enzymatic reaction that is acted upon to produce the reaction product(s). The protein or histone substrate can be directly acted upon by the demethylase (typically by binding to the active site and undergoing a chemical reaction catalyzed by the enzyme) or can first be modified prior to being acted upon by the enzyme. A typical a histone substrate is a H3K27 histone peptide substrate (e.g., H3K27me3). "H3K27me" indicates methylation of a histone H3 peptide at lysine residue 27, when numbered in accordance with the amino acid sequence of the H3 protein. Methylated H3K27 includes mono-, di- and trimethylated H3K27 (i.e., H3K37mel, H3K37me2, and H3K37me3).

[0031] As used herein, the terms "modulate," "modulates" or "modulation" or grammatical variations thereof refer to enhancement (e.g., an increase) or inhibition (e.g., a reduction) in the specified activity. The terms "increases," "enhancement," "enhance," "enhances," or "enhancing" or grammatical variations thereof refers to an increase in the specified activity (e.g., at least about a 1.1 -fold, 1.25-fold, 1.5-fold, 2-fold, 3-fold, 4-fold, 5-fold, 6-fold, 8-fold, 10-fold, 12-fold, or even 15- fold or more increase). The terms "decreases," ""inhibition," "inhibit", "reduction," "reduce," "reduces" or grammatical variations thereof as used herein refer to a decrease or diminishment in the specified activity of at least about 5%, 10%, 25%, 35%, 40%, 50%, 60%, 75%, 80%, 90%, 95% or more. In particular embodiments, the decrease, inhibition or reduction results in no or essentially no (i.e., an insignificant amount, e.g., less than about 10% or even 5%) detectible activity. By the terms "treat," "treating" or "treatment of it is meant that the severity of the subject's condition is reduced or at least partially improved or ameliorated and/or that some alleviation, mitigation or decrease in at least one clinical symptom is achieved and/or there is a delay in the progression of the condition and/or prevention or delay of the onset of a disease or illness. Thus, the terms "treat," "treating" or "treatment of refer to both prophylactic and therapeutic regimens.

[0032] The phrase "reversing a tumorigenic state" of a cell or plurality of cells includes the act of a providing some modification to a cell such that it ceases to become tumorigenic or reduces the tumorigenic potential of the cell. Such a cell may become quiescent, senescent, or apoptotic.

[0033] A "tumor suppressor gene" includes a gene that protects a cell from one or more modifications that result in an increased tumorigenic potential of the cell. Tumor suppressor genes may promote apoptosis or repress cell cycle progression, or both.

[0034] An "embryonic stem cell" includes a pluripotent stem cell derived from the inner cell mass of an embryo, such as a blastocyst.

[0035] A "retinoblastoma susceptibility gene" includes the human retinoblastoma susceptibility gene (RB) and its gene product (pRB). See Lee et al. (1987), Science 235: 1394 - 1399.

[0036] Exemplary compositions and methods

[0037] Provided herein are compositions and complexes containing one or more proteins described herein. A composition may be a pharmaceutical composition. For example, the invention provides an isolated complex including a UTX protein or a catalytic analog thereof (such as a catalytic analog containing a JmjC domain and a treble-clef zinc finger domain), and a JMJD3 protein or a catalytic analog thereof (such as a catalytic analog containing a JmjC domain and a treble-clef zinc finger domain). The isolated complex may also include a histone H3 peptide. The invention also provides an isolated complex including a histone H3 peptide and either a UTX protein (or a catalytic analog thereof) or a JMJD3 protein (or a catalytic analog thereof).

[0038] Nucleic acids, e.g., those encoding a protein of interest or functional homolog thereof, or a nucleic acid intended to inhibit the production of a protein of interest (e.g., siRNA or antisense RNA) can be delivered to cells, e.g., eukaryotic cells, in culture, to cells ex vivo, and to cells in vivo. The cells can be of any type including without limitation cancer cells, stem cells, neuronal cells, and non-neuronal cells. The delivery of nucleic acids can be by any technique known in the art including viral mediated gene transfer, liposome mediated gene transfer, direct injection into a target tissue, organ, or tumor, injection into vasculature which supplies a target tissue or organ.

[0039] Polynucleotides can be administered in any suitable formulations known in the art. These can be as virus particles, as naked DNA, in liposomes, in complexes with polymeric carriers, etc. Polynucleotides can be administered to the arteries which feed a tissue or tumor. They can also be administered to adjacent tissue, whether tumor or normal, which could express the demethylase protein.

[0040] Nucleic acids can be delivered in any desired vector. These include viral or non- viral vectors, including adenovirus vectors, adeno-associated virus vectors, retrovirus vectors, lentivirus vectors, and plasmid vectors. Exemplary types of viruses include HSV (herpes simplex virus), AAV (adeno associated virus), HIV (human immunodeficiency virus), BIV (bovine immunodeficiency virus), and MLV (murine leukemia virus). Nucleic acids can be administered in any desired format that provides sufficiently efficient delivery levels, including in virus particles, in liposomes, in nanoparticles, and complexed to polymers.

[0041] The nucleic acids encoding a protein or nucleic acid of interest may be in a plasmid or viral vector, or other vector as is known in the art. Such vectors are well known and any can be selected for a particular application. In one embodiment of the invention, the gene delivery vehicle comprises a promoter and a demethylase coding sequence. Preferred promoters are tissue-specific promoters and promoters which are activated by cellular proliferation, such as the thymidine kinase and thymidylate synthase promoters. Other preferred promoters include promoters which are activatable by infection with a virus, such as the α- and β-interferon promoters, and promoters which are activatable by a hormone, such as estrogen. Other promoters which can be used include the Moloney virus LTR, the CMV promoter, and the mouse albumin promoter. A promoter may be constitutive or inducible.

[0042] In another embodiment, naked polynucleotide molecules are used as gene delivery vehicles, as described in WO 90/11092 and U.S. Patent 5,580,859. Such gene delivery vehicles can be either growth factor DNA or RNA and, in certain embodiments, are linked to killed adenovirus. Curiel et al., Hum. Gene. Ther. 3:147-154, 1992. Other vehicles which can optionally be used include DNA- ligand (Wu et al., J. Biol. Chem. 264:16985-16987, 1989), lipid-DNA combinations (Feigner et al., Proc. Natl. Acad. Sci. USA 84:7413 7417, 1989), liposomes (Wang et al., Proc. Natl. Acad. Sci. 84:7851-7855, 1987) and microprojectiles (Williams et al., Proc. Natl. Acad. Sci. 88:2726-2730, 1991).

[0043] A gene delivery vehicle can optionally comprise viral sequences such as a viral origin of replication or packaging signal. These viral sequences can be selected from viruses such as astrovirus, coronavirus, orthomyxovirus, papovavirus, paramyxovirus, parvovirus, picornavirus, poxvirus, retrovirus, togavirus or adenovirus. In a preferred embodiment, the growth factor gene delivery vehicle is a recombinant retroviral vector. Recombinant retroviruses and various uses thereof have been described in numerous references including, for example, Mann et al., Cell 33:153, 1983, Cane and Mulligan, Proc. Nat'l. Acad. Sci. USA 81 :6349, 1984, Miller et al., Human Gene Therapy 1 :5-14, 1990, U.S. Patent Nos. 4,405,712, 4,861,719, and 4,980,289, and PCT Application Nos. WO 89/02,468, WO 89/05,349, and WO 90/02,806. Numerous retroviral gene delivery vehicles can be utilized in the present invention, including for example those described in EP 0,415,731; WO 90/07936; WO 94/03622; WO 93/25698; WO 93/25234; U.S. Patent No. 5,219,740; WO 9311230; WO 9310218; Vile and Hart, Cancer Res. 53:3860-3864, 1993; Vile and Hart, Cancer Res. 53:962-967, 1993; Ram et al., Cancer Res. 53:83-88, 1993; Takamiya et al., J. Neurosci. Res. 33:493-503, 1992; Baba et al., J. Neurosurg. 79:729-735, 1993 (U.S. Patent No. 4,777,127, GB 2,200,651, EP 0,345,242 and WO91/02805).

[0044] A polynucleotide of interest can also be combined with a condensing agent to form a gene delivery vehicle. The condensing agent may be a polycation, such as polylysine, polyarginine, polyornithine, protamine, spermine, spermidine, and putrescine. Many suitable methods for making such linkages are known in the art.

[0045] In an alternative embodiment, a polynucleotide of interest is associated with a liposome to form a gene delivery vehicle. Liposomes are small, lipid vesicles comprised of an aqueous compartment enclosed by a lipid bilayer, typically spherical or slightly elongated structures several hundred Angstroms in diameter. Under appropriate conditions, a liposome can fuse with the plasma membrane of a cell or with the membrane of an endocytic vesicle within a cell which has internalized the liposome, thereby releasing its contents into the cytoplasm. Prior to interaction with the surface of a cell, however, the liposome membrane acts as a relatively impermeable barrier that sequesters and protects its contents, for example, from degradative enzymes. Additionally, because a liposome is a synthetic structure, specially designed liposomes can be produced which incorporate desirable features. See Stryer, Biochemistry, pp. 236-240, 1975 (W.H. Freeman, San Francisco, CA); Szoka et al., Biochim. Biophys. Acta 600:1, 1980; Bayer et al., Biochim. Biophys. Acta. 550:464, 1979; Rivnay et al., Meth. Enzymol. 149:119, 1987; Wang et al., PROC. NATL. ACAD. SCI. U.S.A. 84: 7851, 1987, Plant et al., Anal. Biochem. 176:420, 1989, and U.S. Patent 4,762,915. Liposomes can encapsulate a variety of nucleic acid molecules including DNA, RNA, plasmids, and expression constructs comprising growth factor polynucleotides such those disclosed in the present invention.

[0046] Liposomal preparations for use in the present invention include cationic (positively charged), anionic (negatively charged) and neutral preparations. Cationic liposomes have been shown to mediate intracellular delivery of plasmid DNA (Feigner et al., Proc. Natl. Acad. Sci. USA 84:7413-7416, 1987), mRNA (Malone et al., Proc. Natl. Acad. Sci. USA 86:6077-6081, 1989), and purified transcription factors (Debs et al., J. Biol. Chem. 265:10189-10192, 1990), in functional form. Cationic liposomes are readily available. For example, N[l-2,3-dioleyloxy)propyl]-N,N,N- triethylammonium (DOTMA) liposomes are available under the trademark LIPOFECTIN®, from GIBCO BRL, Grand Island, NY. See also Feigner et al, Proc. Natl. Acad. Sci. USA 91 : 5148- 5152.87, 1994. Other commercially available liposomes include TransfectACE (DDAB/DOPE) and DOTAP/DOPE (Boerhinger). Other cationic liposomes can be prepared from readily available materials using techniques well known in the art. See, e.g., Szoka et al., Proc. Natl. Acad. Sci. USA 75:4194-4198, 1978; and WO 90/11092 for descriptions of the synthesis of DOTAP (1,2- bis(oleoyloxy)-3-(trimethylammonio)propane) liposomes.

[0047] Similarly, anionic and neutral liposomes are readily available, such as from Avanti Polar

Lipids (Birmingham, AL), or can be easily prepared using readily available materials. Such materials include phosphatidyl choline, cholesterol, phosphatidyl ethanolamine, dioleoylphosphatidyl choline (DOPC), dioleoylphosphatidyl glycerol (DOPG), dioleoylphosphatidyl ethanolamine (DOPE), among others. These materials can also be mixed with the DOTMA and DOTAP starting materials in appropriate ratios. Methods for making liposomes using these materials are well known in the art.

[0048] One or more protein (e.g., a demethylase) or nucleic acid (e.g., siRNA) of interest may be encoded by a single nucleic acid delivered. Alternatively, separate nucleic acids may encode different protein or nucleic acids of interest. Different species of nucleic acids may be in different forms; they may use different promoters or different vectors or different delivery vehicles. Similarly, the same protein or nucleic acid of interest may be used in a combination of different forms.

Antisense molecules, siRNA or shRNA molecules, ribozymes or triplex molecules may be contacted with a cell or administered to an organism. Alternatively, constructs encoding these may be contacted with or introduced into a cell or organism. Antisense constructs, antisense oligonucleotides, RNA interference constructs or siRNA duplex RNA molecules can be used to interfere with expression of a protein of interest, e.g., a histone demethylase. Typically at least 15, 17, 19, or 21 nucleotides of the complement of the mRNA sequence are sufficient for an antisense molecule. Typically at least 19, 21, 22, or 23 nucleotides of a target sequence are sufficient for an RNA interference molecule. Preferably an RNA interference molecule will have a 2 nucleotide 3 ' overhang. If the RNA interference molecule is expressed in a cell from a construct, for example from a hairpin molecule or from an inverted repeat of the desired histone demethylase sequence, then the endogenous cellular machinery will create the overhangs. siRNA molecules can be prepared by chemical synthesis, in vitro transcription, or digestion of long dsRNA by Rnase III or Dicer. These can be introduced into cells by transfection, electroporation, or other methods known in the art. See Hannon, GJ, 2002, RNA Interference, Nature 418: 244-251; Bernstein E et al., 2002, The rest is silence. RNA 7: 1509-1521; Hutvagner G et al., RNAi: Nature abhors a double-strand. Curr. Opin. Genetics & Development 12: 225-232; Brummelkamp, 2002, A system for stable expression of short interfering RNAs in mammalian cells. Science 296: 550-553; Lee NS, Dohjima T, Bauer G, Li H, Li M-J, Ehsani A, Salvaterra P, and Rossi J. (2002). Expression of small interfering RNAs targeted against HIV-I rev transcripts in human cells. Nature Biotechnol. 20:500- 505; Miyagishi M, and Taira K. (2002). U6-promoter-driven siRNAs with four uridine 3' overhangs efficiently suppress targeted gene expression in mammalian cells. Nature Biotechnol. 20:497-500; Paddison PJ, Caudy AA, Bernstein E, Hannon GJ, and Conklin DS. (2002). Short hairpin RNAs (shRNAs) induce sequence-specific silencing in mammalian cells. Genes & Dev. 16:948-958; Paul CP, Good PD, Winer I, and Engelke DR. (2002). Effective expression of small interfering RNA in human cells. Nature Biotechnol. 20:505-508; Sui G, Soohoo C, Affar E-B, Gay F, Shi Y, Forrester WC, and Shi Y. (2002). A DNA vector-based RNAi technology to suppress gene expression in mammalian cells. Proc. Natl. Acad. Sci. USA 99(6):5515-5520; Yu J-Y, DeRuiter SL, and Turner DL. (2002). RNA interference by expression of short-interfering RNAs and hairpin RNAs in mammalian cells. Proc. Natl. Acad. Sci. USA 99(9):6047-6052. [0050] Antisense or RNA interference molecules can be delivered in vitro to cells or in vivo, e.g., to tumors of a mammal. Typical delivery means known in the art can be used. For example, delivery to a tumor can be accomplished by intratumoral injections. Other modes of delivery can be used without limitation, including: intravenous, intramuscular, intraperitoneal, intra-arterial, local delivery during surgery, endoscopic, subcutaneous, and per os. In a mouse model, the antisense or RNA interference can be administered to a tumor cell in vitro, and the tumor cell can be subsequently administered to a mouse. Vectors can be selected for desirable properties for any particular application. Vectors can be viral or plasmid. Adenoviral vectors are useful in this regard. Tissue-specific, cell-type specific, or otherwise regulatable promoters can be used to control the transcription of the inhibitory polynucleotide molecules. Non-viral carriers such as liposomes or nanospheres can also be used.

[0051] Exemplary methods of treatment and diseases

Provided herein are methods of treatment or prevention of conditions and diseases that can be improved by modulating the methylation status of histones, and thereby, e.g., modulate the level of expression of methylation activated and methylation repressed target genes. A method may comprise administering to a subject, e.g., a subject in need thereof, a therapeutically effective amount of an agent described herein.

[0052] Diseases such as cancers can be treated by administration of modulators of histone methylation, e.g., modulators of histone demethylase enzyme activity. H3K27 methylation has been reported to be involved in overexpression of certain genes in cancers. Modulators that are identified by the disclosed methods or modulators that are described herein can be used to treat these diseases, i.e., to restore normal methylation to affected cells.

[0053] Based at least on the fact that increased histone methylation has been found to be associated with certain cancers, a method for treating cancer in a subject may comprise administering to the subject a therapeutically effective amount of one or more agents that decrease methylation or restores methylation to its level in corresponding normal cells.

[0054] It is believed that modulators of methylation can be used for modulating cell proliferation generally. Excessive proliferation may be reduced with agents that decrease methylation, whereas insufficient proliferation may be stimulated with agents that increase methylation. Accordingly, diseases that may be treated include hyperproliferative diseases, such as benign cell growth and malignant cell growths. Exemplary cancers that may be treated include leukemias, e.g., acute lymphoid leukemia and myeloid leukemia, and carcinomas, such as colorectal carcinoma and hepatocarcinoma. Other cancers include Acute Lymphoblastic Leukemia; Acute Lymphoblastic Leukemia; Acute Myeloid Leukemia; Acute Myeloid Leukemia; Adrenocortical Carcinoma Adrenocortical Carcinoma; AIDS- Related Cancers; AIDS-Related Lymphoma; Anal Cancer; Astrocytoma, Childhood Cerebellar; Astrocytoma, Childhood Cerebral; Basal Cell Carcinoma, see Skin Cancer (non-Melanoma); Bile Duct Cancer, Extrahepatic; Bladder Cancer; Bladder Cancer; Bone Cancer, osteosarcoma/Malignant Fibrous Histiocytoma; Brain Stem Glioma; Brain Tumor; Brain Tumor, Brain Stem Glioma; Brain Tumor, Cerebellar Astrocytoma; Brain Tumor, Cerebral Astrocytoma/Malignant Glioma; Brain Tumor, Ependymoma; Brain Tumor, Medulloblastoma; Brain Tumor, Supratentorial Primitive Neuroectodermal Tumors; Brain Tumor, Visual Pathway and Hypothalamic Glioma; Brain Tumor; Breast Cancer; Breast Cancer and Pregnancy; Breast Cancer; Breast Cancer, Male; Bronchial Adenomas/Carcinoids; Burkitt's Lymphoma; Carcinoid Tumor; Carcinoid Tumor, Gastrointestinal; Carcinoma of Unknown Primary; Central Nervous System Lymphoma, Primary; Cerebellar Astrocytoma;Cerebral Astrocytoma/Malignant Glioma; Cervical Cancer; Childhood Cancers; Chronic Lymphocytic Leukemia; Chronic Myelogenous Leukemia; Chronic Myeloproliferative Disorders; Colon Cancer; Colorectal Cancer; Cutaneous T-CeIl Lymphoma, see Mycosis Fungoides and Sezary Syndrome; Endometrial Cancer; Ependymoma; Esophageal Cancer; Esophageal Cancer; Ewing's Family of Tumors; Extracranial Germ Cell Tumor; Extragonadal Germ Cell Tumor; Extrahepatic Bile Duct Cancer; Eye Cancer, Intraocular Melanoma; Eye Cancer, Retinoblastoma; Gallbladder Cancer; Gastric (Stomach) Cancer; Gastric (Stomach) Cancer; Gastrointestinal Carcinoid Tumor; Germ Cell Tumor, Extracranial; Germ Cell Tumor, Extragonadal; Germ Cell Tumor, Ovarian; Gestational Trophoblastic Tumor; Glioma; Glioma, Childhood Brain Stem; Glioma, Childhood Cerebral Astrocytoma; Glioma, Childhood Visual Pathway and Hypothalamic; Hairy Cell Leukemia; Head and Neck Cancer; Hepatocellular (Liver) Cancer, Adult (Primary); Hepatocellular (Liver) Cancer, Childhood (Primary); Hodgkin's Lymphoma; Hodgkin's Lymphoma; Hodgkin's Lymphoma During Pregnancy; Hypopharyngeal Cancer; Hypothalamic and Visual Pathway Glioma; Intraocular Melanoma; Islet Cell Carcinoma (Endocrine Pancreas); Kaposi's Sarcoma; Kidney (Renal Cell) Cancer; Kidney Cancer; Laryngeal Cancer; Laryngeal Cancer; Leukemia, Acute Lymphoblastic; Leukemia, Acute Lymphoblastic; Leukemia, Acute Myeloid; Leukemia, Acute Myeloid; Leukemia, Chronic Lymphocytic; Leukemia; Chronic Myelogenous; Leukemia, Hairy Cell; Lip and Oral Cavity Cancer; Liver Cancer, Adult (Primary); Liver Cancer, Childhood (Primary); Lung Cancer, Non-Small Cell; Lung Cancer, Small Cell; Lymphoma, AIDS- Related; Lymphoma, Burkitt's; Lymphoma, Cutaneous T-CeIl, see Mycosis Fungoides and Sezary Syndrome; Lymphoma, Hodgkin's; Lymphoma, Hodgkin's; Lymphoma, Hodgkin's During Pregnancy; Lymphoma, Non-Hodgkin's; Lymphoma, Non-Hodgkin's; Lymphoma, Non-Hodgkin's During Pregnancy; Lymphoma, Primary Central Nervous System; Macroglobulinemia, Waldenstrom's; Malignant Fibrous Histiocytoma of Bone/Osteosarcoma; Medulloblastoma; Melanoma; Melanoma, Intraocular (Eye); Merkel Cell Carcinoma; Mesothelioma, Adult Malignant; Mesothelioma; Metastatic Squamous Neck Cancer with Occult Primary; Multiple Endocrine Neoplasia Syndrome; Multiple Myeloma/Plasma Cell Neoplasm' Mycosis Fungoides; Myelodysplastic Syndromes; Myelodysplastic/Myeloproliferative Diseases; Myelogenous Leukemia, Chronic; Myeloid Leukemia, Adult Acute; Myeloid Leukemia, Childhood Acute; Myeloma, Multiple; Myeloproliferative Disorders, Chronic; Nasal Cavity and Paranasal Sinus Cancer; Nasopharyngeal Cancer; Nasopharyngeal Cancer; Neuroblastoma; Non-Hodgkin's Lymphoma; Non-Hodgkin's Lymphoma; Non-Hodgkin's Lymphoma During Pregnancy; Non-Small Cell Lung Cancer; Oral Cancer; Oral Cavity Cancer, Lip and; Oropharyngeal Cancer; Osteosarcoma/Malignant Fibrous Histiocytoma of Bone; Ovarian Cancer; Ovarian Epithelial Cancer; Ovarian Germ Cell Tumor; Ovarian Low Malignant Potential Tumor; Pancreatic Cancer; Pancreatic Cancer; Pancreatic Cancer, Islet Cell; Paranasal Sinus and Nasal Cavity Cancer; Parathyroid Cancer; Penile Cancer; Pheochromocytoma; Pineoblastoma and Supratentorial Primitive Neuroectodermal Tumors; Pituitary Tumor; Plasma Cell Neoplasm/Multiple Myeloma; Pleuropulmonary Blastoma; Pregnancy and Breast Cancer; Pregnancy and Hodgkin's Lymphoma; Pregnancy and Non-Hodgkin's Lymphoma; Primary Central Nervous System Lymphoma; Prostate Cancer; Rectal Cancer; Renal Cell (Kidney) Cancer; Renal Cell (Kidney) Cancer; Renal Pelvis and Ureter, Transitional Cell Cancer; Retinoblastoma; Rhabdomyosarcoma; Salivary Gland Cancer; Salivary Gland Cancer; Sarcoma, Ewing's Family of Tumors; Sarcoma, Kaposi's; Sarcoma, Soft Tissue; Sarcoma, Soft Tissue; Sarcoma, Uterine; Sezary Syndrome; Skin Cancer (non-Melanoma); Skin Cancer; Skin Cancer (Melanoma); Skin Carcinoma, Merkel Cell; Small Cell Lung Cancer; Small Intestine Cancer; Soft Tissue Sarcoma; Soft Tissue Sarcoma; Squamous Cell Carcinoma, see Skin Cancer (non-Melanoma); Squamous Neck Cancer with Occult Primary, Metastatic; Stomach (Gastric) Cancer; Stomach (Gastric) Cancer; Supratentorial Primitive Neuroectodermal Tumors; T- CeIl Lymphoma, Cutaneous, see Mycosis Fungoides and Sezary Syndrome; Testicular Cancer; Thymoma; Thymoma and Thymic Carcinoma; Thyroid Cancer; Thyroid Cancer; Transitional Cell Cancer of the Renal Pelvis and Ureter; Trophoblastic Tumor, Gestational; Unknown Primary Site, Carcinoma of; Unknown Primary Site, Cancer of; Unusual Cancers of Childhood; Ureter and Renal Pelvis, Transitional Cell Cancer; Urethral Cancer; Uterine Cancer, Endometrial; Uterine Sarcoma; Vaginal Cancer; Visual Pathway and Hypothalamic Glioma; Vulvar Cancer; Waldenstrom's Macroglobulinemia; Wilms' Tumor; and Women's Cancers.

Any other disease in which epigenetics, in particular methylation, plays a role is likely to be treatable or preventable by applying methods described herein.

[0057] Screening methods

[0058] Also provided herein are screening methods for identifying agents that modulate methylation of a target protein, such as a histone, e.g., lysine 27 (K27) of histone 3. One method involves screening for an enhancer or inhibitor of histone demethylase activity, including the steps of contacting a histone H3 peptide with a histone demethylase protein (such as UTX, JMJD3, or a catalytic analog thereof), in the presence and in the absence of a test substance; determining the methylation status of the histone H3 peptide at a lysine 27 position; and identifying a test substance as an enhancer of histone demethylase activity if less mono-, di- or trimethylated H3K27 is found in the presence than in the absence of the test substance, and identifying a test substance as an inhibitor of histone demethylase protein activity if more mono-, di- or trimethylated H3K27 is found in the presence than in the absence of the test substance. Test agents (or substances) for screening as inhibitors or enhancers of the demethylase enzymes can be from any source known in the art. They can be natural products, purified or mixtures, synthetic compounds, members of compound libraries, etc. The compounds to be tested may be chosen at random or may be chosen using a filter based on structure and/or mechanism of the enzymes. The test substances can be selected from those that have previously identified to have biological or drug activity or from those that have not. In some embodiments a natural substrate is the starting point for designing an inhibitor. Modifications to make the substrate non-modifiable by the enzyme can be used to make an inhibitor.

[0059] All publications, including patents, applications, and Accession numbers mentioned herein are hereby incorporated by reference in their entirety as if each individual publication or patent was specifically and individually indicated to be incorporated by reference. In case of conflict, the present application, including any definitions herein, will control.

[0060] The invention now being generally described, it will be more readily understood by reference to the following examples, which are included merely for purposes of illustration of certain aspects and embodiments of the present invention, and are not intended to limit the invention. [0061] EXAMPLE 1. A histone H3 27 demethylase regulates animal posterior development.

[0062] This example corresponds to the experiments described in Lan et al. Nature (2007) 449:

689-94, the contents of which are hereby incorporated by reference in their entirety.

[0063] Using recombinant UTX purified from Sf9 cells and a collection of methylated histone peptides as substrates, we found that UTX mediated demethylation of H3K27me3, but had no effect on the trimethylated H3K4, K9, K36, or H4K20 (Figure IA). UTX also mediated demethylation of H3K27me2 and H3K27mel peptides, indicating that UTX converts H3K27me3 to un-methylated products when assayed on peptide substrates in vitro. Consistently, Western blot analysis also showed UTX-mediated demethylation of H3K27me3/2 on bulk histones (Figure IB). However, unlike the peptide substrates, a reduction of H3K27mel levels was not detectable on the bulk histones (Figure IB, top panel).

[0064] Similar to the full-length UTX, the UTX JmjC catalytic domain alone also mediated demethylation of H3K27me3, me2 and mel, when the methylated histone peptides were used as substrates (Figure ID). When bulk histones were analyzed, only H3K27me3 and me2, but not H3K27mel, levels were reduced (Figure IE). JMJD3 catalytic domain displayed similar specificity, demethylating H3K27me3, me2 and mel of the histone peptides (Figure ID) but only H3K27me3 and me2 on native histones and nucleosomal substrates (Figure ID and 1C). In contrast, the UTY JmjC domain or the full-length UTY purified from Sf9 cells was inactive under the same assay conditions (Figure ID and E, and data not shown). Importantly, these enzymes showed no activity towards trimethylated H3K4, K9, K36 or H4K20 (Figure IE). The JMJD3- mediated demethylation reaction of the nucleosomal substrates is significantly less efficient than on bulk histones. Under the same assay condition, UTX had barely detectable activity. This suggests that additional cofactors and/or posttranslational modifications may be required for efficient demethylation of nucleosomal substrates. This is not unprecedented as the H3K4 demethylase LSDl requires CoREST for demethylation of nucleosomes 25,26. Importantly, we show that genomic sites bound by UTX in primary cells are correspondingly depleted of H3K27me3 (Figure 3C), suggesting that UTX can demethylate H3K27me3 on nucleosomes in vivo.

[0065] When over-expressed in 293T cells, JMJD3 significantly reduced the levels of H3K27me3 and me2 in approximately 78% and 56% of the transfected cells, respectively (Figure 6, marked by arrowheads), but not H3K36me3, H3K9me3 or H3K4me3 (Figure 6 and data not shown). Over- expression of JMJD3 did not reduce the H3K27mel level, and in fact, an increase in H3K27mel was observed in 24% of the transfected cells (Figure 6). This accumulation of H3K27mel is probably a result of the conversion of H3K27me3 and me2 to H3K27mel due to over-expression of JMJD3. Over-expression of JMJD3 in HeLa cells, however, appeared to have less uniform effects on the global H3K27me3 level, i.e., some transfected cells showed clear reduction of H3K27me3 but others showed less robust or marginal effects. In contrast, over-expression of UTX in 293T cells had little effect, if any at all, on the global level of H3K27me3. These findings suggest that JMJD3 and UTX may have differential ability to impact global H3K27 trimethylation in a cell type- specific manner and UTX action may be targeted to selective genomic loci in cells. Taken together, these findings suggest that UTX and JMJD3 are histone demethylases with specificities largely directed towards H3K27me3 and H3K27me2. Whether UTY is an active H3K27 demethylase remains to be determined by further experimentation.

[0066] We next asked whether UTX regulates H3K27 methylation at endogenous genes (Figure 2).

Because H3K27 trimethylation has been shown to be critical for the regulation of the Hox gene cluster, we investigated H3K27 methylation at the Hox gene locus and compared their levels in HeLa cell in the presence and absence of shRNA plasmids that inhibited the expression of UTX or JMJD3 (Figure 2B). We found that RNAi inhibition of the endogenous UTX resulted in H3K27me3 increases in some but not all Hox D genes. Specifically, in the promoters of HoxDIO (primer #3), HoxDl 1 (primer #7) and D12 (primer #10 and 11), H3K27me3 level was clearly elevated in the UTX knockdown cells (Figure 2C). We also observed one instance in the HoxD12 gene where H3K27me3 was elevated at the 3' end of the gene in UTX knockdown cells (primer #9). In contrast, knockdown of JMJD3 had very little effects on H3K27me3 levels within the HoxD cluster (Figure 2C). These findings suggest that UTX is involved in regulating the H3K27me3 level at some but not all posterior Hox genes in HeLa cells. However, the increase in H3K27me3 levels was not accompanied by further repression of the corresponding Hox genes. This may be due to the fact that these Hox genes are already highly silenced in HeLa cells. In support of a role for UTX in transcriptional activation, we observed further decrease of the transcript levels of a number of Hox genes in human primary fibroblast cells upon RNAi and in Zebrafish upon morpholino inhibition of UTX (see below).

[0067] To determine how general a role UTX plays in the regulation of the Hox gene locus, we carried out loci-specific and genome-wide analysis of UTX localization. We studied two primary human fibroblast cell types (foot and lung) that preserve distinct patterns of embryonic Hox expression 8, as well as in mouse ES cells where Hox genes are silent 11. The endogenous UTX in fibroblasts was isolated by chromatin immunoprecipitation (ChIP) followed by hybridization to ultra-dense tiling microarrays (ChIP-chip) that interrogated all four human Hox loci at 5 base pair resolution and 2 megabases of control regions including portions of X chromosome, chromosome 22, the beta-globin locus, and many transcribed genes in fibroblasts 27. Strikingly, UTX was selectively localized to the Hox loci of fibroblasts; over 90% of all UTX binding events were in the Hox loci (Figure 3B, p<10-64, hypergeometric distribution). In contrast to the broad domains of H3K27me3, PRC2, H3K4me2, and RNA polymerase II occupancy over Hox genes in these cells, UTX was selectively targeted to narrow windows within 500 base pairs downstream of the transcriptional start site of HOX genes (Figure 3 A and C). The raw UTX ChIP-chip profiles at both locus-wide and gene-specific resolutions are shown in Figure 7. Although focal, these UTX binding events are supported by hybridization of multiple contiguous probes and are thus of high statistical confidence and validated by additional quantitative PCR experiments (Figure 7). UTX binds the start of both transcriptionally active and silent Hox genes in a manner largely independent of anatomic origins of cells (Figure 3 A and D). Importantly, when UTX binding occurred within domains of H3K27me3, H3K27me3 was concomitantly diminished (Figure 3C). This result supports the idea that UTX acts as an H3K27me-specific demethylase in cells to mark transcriptional start sites, and is consistent with the preferential effect of UTX depletion on H3K27me3 in gene-proximal promoters (Figure 2).

[0068] Unlike fibroblasts with distinct positional identities, HOX genes in embryonic stem (ES) cells are largely occupied by H3K27me3 and transcriptionally silent 9. We performed ChIP-chip analysis of UTX in mouse ES cells on tiling arrays which interrogated 4 kilobases around -48000 annotated transcriptional start sites in the mouse genome. UTX was entirely excluded from the Hox loci (Figure 3D), although UTX is expressed and appeared to bind selected sets of genes at other genomic locations (Figure 8). The lack of UTX occupancy suggests a potential deficiency in the mechanisms important for targeting of UTX to the Hox gene locus in mES cells. To address the role of UTX binding to HOX genes, we depleted UTX in fibroblasts and examined the effects on ΗoxA9, a posterior Hox gene, which bound UTX in a position- invariant manner. Depletion of UTX led to decreased expression of H0XA9 in both lung and foot fibroblasts (Figure 3E). Taken together, these results suggest that endogenous UTX is actively and dynamically localized to the start of HOX genes in differentiation, potentially to reverse Η3K27 methylation and to enable transcriptional activation.

[0069] Given that UTX is located predominately at the HOX gene locus to regulate Η3K27 methylation, we next asked whether UTX plays a role in animal body patterning, which is controlled by proper temporal and spatial expression of the developmentally regulated Hox genes. We turned to zebrafish, which possesses two UTX homologs (Figure 5). Knockdown of zebrafish UTXl (NM OO 1080193) by two non- overlapping antisense morpholino oligonucleotides resulted in minor developmental delay by 24 hours post-fertilization (hpf). By 48hpf, posterior somites failed to retain the characteristic chevron shape (Figure 4A). Inter-segmental distances are dramatically decreased in the caudal trunk resulting in a much-shortened embryo, suggesting defects in posterior somitogenesis. Degeneration of the posterior notochord is also observed, approximately from somite 13 to 23, which results in a highly compact and shortened body at this affected region and a slight dorsal curvature of the tail. (Figure 4B). These phenotypes persist until 8dpf. Anterior portions of the embryo appear mostly unaffected, and the head appears to develop normally, although it seems to be slightly larger than the control. Morpholino inhibition of fish UTX2 expression also affected posterior development, albeit to a much smaller extent. Importantly, co- injection of 200pg of mRNA encoding fish zUTXl or human wildtype, but not a catalytically inactive UTX (Figure 9), resulted in a significant rescue of the development defects (P=5.1E-14, Figure 4C and D and data not shown). Consistently, there was no significant difference between the catalytically inactive mutant and the control in their ability to rescue the fish phenotype (P=O.26). These findings further indicate that the phenotype was specifically due to UTX knockdown and its demethylase activity is required for proper posterior patterning.

Given the finding of UTX occupancy at the HOX gene promoters and its role in Η3K27me3 demethylation (Figures 2 and 3), and the fact that HOX genes are known to be important for body patterning during vertebrate development 1, we investigated whether zΗox genes are mis-regulated in zUTXl morphants. To avoid potential complications that may be caused by the size difference between the wild and zUTX morpholino-treated embryos in later time points, we carried out Hox gene expression analysis at time points (12, 24 and 36 hrs post-injection of the morpholinos) prior to the appearance of the posterior truncation phenotype. Importantly, we found that transcript levels of several 3' posterior Ηox genes, such as zHoxA9b, zHoxC8a, zHoxClla, zHoxC12b, and zHoxDl 2a are modestly but consistently reduced at 36 hours dpf (Figure 4E). However, the expression of more anterior zHox genes (such as zHoxCόa, zHoxCόb and zHoxA3) and the most posterior Hox genes (zHoxA13, zHoxC13a and zHoxClib) appears unaffected (Figure 4E and data not shown), which correspond well to the normal appearance of the anterior and most posterior tail regions of the mutant embryos (Figure 4E and data not shown). Consistent with the RT-PCR results, RNA in situ analysis showed modestly reduced transcript level of zHoxC8 and a posterior shift of its expression domain in most of the UTX morpholino-treated embryos (from somite 1-7 to somite 2-8) (Figure 4F). Similarly, we also observed reduced expression of zHoxCl Ia and zHoxC12b, loss of expression of zHoxDl 2a and zHoxA9b at the pectoral fin bud, and a shift of the zHoxD9 expression domain (Figure 10). Taken together, these findings indicate that UTX plays an important role in regulating Hox gene expression and body patterning during vertebrate development through its demethylase activity, consistent with our ChIP-chip analyses, which identified UTX occupancy at HOX loci in differentiated cells (Figure 3). [0071] The ALR/MLL2 protein complex that mediates H3K4 trimethylation has been identified 28.

This complex contains both UTX and WDR5, which is important for MLL complex regulating H3K4 methylation 29. Interestingly, a posterior developmental defect and an improper expression pattern of xHoxC8, similar to what we have observed for the UTX morpholino-treated embryos, have also been noted in the WDR5 MO treated Xenopus embryos 29. These findings suggest a potential collaboration of H3K4 methylation and demethylation of H3K27me in regulating Hox gene expression, and possibly evolutionary conservation of the role of UTX in body patterning from zebrafish to human. Consistently, we found co-localization of UTX with some of the other components of the ALR/MLL2 complex, including ALR/MLL2 and RbQ3, as well as H3K4me3 at the H0XA9, HOXAlO and HOXDl 3 loci in lung fibroblasts, but not other neighboring regions (Figure 11), suggesting a potential cooperation between MLL2 and UTX in regulating HoxA9 gene in lung fibroblasts. However, in HeLa cells, UTX and ALR/MLL2 complex do not occupy H0XA9 promoter, and consequently depletion of ALR/MLL2 does not affect H3K4 methylation and transcription of H0XA9 28. This discrepancy may be due to cell type specific regulation of HOX genes.

[0072] In addition, it is likely that H3K4 and H3K27 methylation can be toggled independently for certain genes. RNAi inhibition of UTX in 293T or mouse ES cells had no effect on global H3K4 methylation or the in vitro H3K4 methylation activity of ALR/MLL2 (Figure 12), and UTX occupancy did not correspond to bulk H3K4 methylation level in the Hox loci (Figure 7C). This suggests that, in some cases, UTX occupancy of the HOX locus may occur independently of ALR function. Even when ALR and UTX are co-localized to the same genomic loci, UTX may act independently of ALR function in the HOX locus. Indeed, in 293 cells, H0XA9 is also regulated by the MLLl complex, which is related to ALR/MLL2 but lacks UTX 30,31. Taken together, these findings set the stage for future studies to delineate the relationship between MLL proteins and UTX, and the mechanism integrating the functions of the ALR/MLL2 complex in chromatin regulation.

[0073] It has been proposed that Η3K27me3 demethylases were either absent or were tightly regulated in order to protect H3K27me3, which is an important epigenetic modification 32. Recent studies showed that a large set of developmentally important genes are enriched for both H3K27me3 and H3K4me2 in ES cells (bivalent domains), which represent the repressive and activating histone modifications, respectively 4. The bivalent domains appear to segregate upon differentiation, and the corresponding genes are either repressed or activated, suggesting that the bivalent modifications may be important for poising these genes for activation and repression in a cell type- and fate- dependent manner. Importantly, activation of some of these genes is correlated with a loss of H3K27me3 9, suggesting possible dynamic regulation through demethylation during differentiation. Our finding of the H3K27me demethylases, combined with the observation that the JMJD3 expression is up-regulated during ES cell differentiation 9, supports the hypothesis that H3K27me demethylases play a role in the resolution of the "bivalent domain" and in the regulating transcription of these genes during ES differentiation. Furthermore, our observation that UTX is excluded from the Hox gene locus identifies a possible mechanism that may help protect the bivalent domains at the Hox gene locus in ES cells. At the same time, UTX is involved in HOX gene locus regulation during development and differentiation. Given the fact that UTX is in the same protein complex with enzymes that mediate Η3K4 trimethylation, we speculate that important mechanisms must be in place to facilitate differential regulation of the K4 and K27 methylation states in ES versus differentiated cells. These mechanisms may involve, but not limited to, possible differential MLL/UTX complex composition and/or differential regulation of MLL and/or UTX enzymatic activities at the target loci at different stages of cell differentiation. Finally, a recent study has identified a link between Rb, PcG proteins and H3K27 methylation in the transcriptional silencing of the pl6INK4Dtumor suppressor 33. HOX genes are also candidate oncogenes and tumor suppressor genes in several types of human cancer 34,35. Thus, UTX and or JMJD3 may play an antagonistic role to that of the PcG proteins in pi 6 and/or HOX regulation and therefore may function as putative tumor suppressors.

[0074] Methods

[0075] Peptides and antibodies. Biotinylated histone peptides were purchased from AnaSpec.

Antibodies (Ab) that recognize different histone modifications were purchased from Upstate Group INC.

[0076] Plamid construction. Full-length UTX, UTY and JmjD3 as well as the JmjC domains were

PCR amplified from HeLa cell and a male cDNA libraries using Phusion polymerase (Finnenzymes) and cloned into the Gateway Entry system (Invitrogen).

[0077] Purification of recombinant proteins and demethylation reactions. The FLAG-tagged UTX and UTY were expressed and purified from Sf9 cells using BAC-N-BLUE baculoviral expression system (Invitrogen) as previously described 36. GST fused JmjC domains were purified from bacteria. For demethylation reactions, 2 μg full-length proteins and 4 μg JmjC domains were incubated with 1 μg peptides or 4 μg bulk histone in a 100 μl reaction for 4 hours at 37°C.

[0078] MALDI-TOF mass spectrometry. One microliter of the demethylation reaction mixture was desalted through a Cl 8 ZipTip (Millipore). The ZipTip was activated, equilibrated, and loaded as previously described by Shi et al. (2004). The bound material was then eluted with 10 mg/ml D- cyano-4-hydroxycinnamic acid MALDI matrix in 70% acetonitrile/0.1% TFA before being spotted and co-crystallized. The samples were analyzed by a MALDI-TOF/TOF mass spectrometer.

[0079] Immunofluorescence microscopy. 293T and HeLa cells were transfected, fixed with 3% paraformaldehyde and immunostained as previously described 36. The samples were mounted with Vectashield (Vector Laboratories) and analyzed by fluorescence microscopy (Nikon E600) using a 6Ox objective. Images were acquired and processed with Openlab 3.1.5 software.

[0080] RT-PCR. Total RNA was isolated with Trizol (Invitrogen) and reverse transcribed with

Superscript III (Invitrogen). PCR reactions were conducted with equal amounts cDNA and kept in the linear range. A complete list of primers and conditions will be provided upon request.

[0081] Chromatin immunoprecipitation. HeLa cells were transfected with shRNA vectors targeting

UTX and JmjD3 and selected by puromycin for 48 hours. Chromatin prepared from 107 cells was used in one ChIP experiment as previously described 4. Eluted DNA samples were used in semiquantitative radioactive PCR reactions and the results were obtained and analyzed by phosphor image system. An unrelated region with hypo-H3K27me3 was used to normalize the folds of enrichment against input DNA.

[0082] ChIP-chip. Primary human lung, foot, and foreskin fibroblasts, custom human Hox tiling microarray, ChIP-chip analysis, and Hox loci-wide occupancy of Suzl2, H3K27me3, H3K4me2, and RNA pol II were as described (PMID: 17604720). ChIP of mouse ES cells (gift of A. Wright and M. Scott) were hybridized to mouse promoter tiling array set (Nimblegen Systems, WI) which tiles 3.25 kilobases upstream and 0.75 kilobases downstream of promoters genome-wide. Affinity purified UTX antiserum was used for ChIP as described above and compared with isotype-matched IgG in parallel ChIP-chip experiments. UTX occupancy was determined by SignalMap peak calling algorithm comparing binding event to simulated data on shuffled probe sets; we chose peaks that had estimated false discovery rate of less than 0.01 and had signal intensity at least 1.5-fold over control chIP-chip experiment with IgG.

[0083] Knockdown of zebrafish UTX homolog and phenotypic analysis. Knockdown of zebrafish

UTX was performed by injection of 2nL of a stock concentration of 250 μM antisense morpholinos (Gene-Tools, LLC) and 75ng/μl of mRNA from human UTX constructs or a control EGFP into one- cell stage zebrafish embryos using a gas driven microinjector (Medical Systems Corp.). The zUTXl morpholino sequence used in Figure 4 was 5'- AGCTCCGAGCGTCCAAAAGCCACAA -3' covering bases -55 to -31 in the 5' UTR, and a similar phenotype was observed at a lower frequency with a second morpholino (5'- CCACCGAC ACTCGGCACGGCTTCAT -3') covering ORF bases +1 to +25. The sequence of the control morpholino was 5'-

CCTCTTACCTCAGTTACAATTTATA -3'. Whole-mount in situ hybridization was done using digoxigenin and/or fluorescein-labeled antisense RNA probes (Roche).

] Table 1. Summary of UTX occupancy in the Hox loci. Columns F-I give the genes that were used to create the average occupancy profiles shown in Figure 3C and Figure 7C. For example, 23 UTX occupancy profiles of 16 unique Hox genes from two sources of fibroblasts were averaged for the left axis of Fig. 3C. 12 H3K27me3 profiles of 10 unique Hox genes from the same two types of fibroblasts were averaged for the right axis of Fig. 3C.

TABLE 1. Fig7-

Figure 3C- Figure 3C- left

Occupancy left axis right axis axis Fig7- right axis

U

T Pol K27me

Sample Gene X II 3

Foot Fb HOXAl y

Foot Fb H0XA2 y

Foot Fb H0XA3 y y y y

Foot Fb H0XA4 y

Foot Fb H0XA5 y

Foot Fb H0XA6 y

Foot Fb H0XA7 y

Foot Fb H0XA9 y y y

Foot Fb HOXAlO y y y

Foot Fb HOXAI l y y y

Foot Fb HOXAl 3 y

Foot Fb HOXBl y

Foot Fb H0XB2 y

Foot Fb H0XB3 y

Foot Fb H0XB4 y y y y

Foot Fb H0XB5 y

Foot Fb H0XB6 y

Foot Fb H0XB7 y

Foot Fb H0XB8 y y y y

Foot Fb H0XB9 y

Foot Fb HOXB 13 y

Foot Fb H0XC4 y

Foot Fb H0XC5 y

Foot Fb H0XC6 y

Foot Fb H0XC8 y

Foot Fb H0XC9 y

Foot Fb HOXClO y

Foot Fb HOXCI l y y y

Foot Fb HOXC 12 y y y y Foot Fb HOXC 13 y

Foot Fb HOXDl y

Foot Fb H0XD3 y

Foot Fb H0XD4 y

Foot Fb H0XD8 y y y y

Foot Fb H0XD9 y

Foot Fb HOXDlO y

Foot Fb HOXDI l y

Foot Fb HOXD 12 y

Foot Fb H0XD13 y y y y

Lung Fb HOXAl y

Lung Fb H0XA2 y

Lung Fb H0XA3 y y y

Lung Fb H0XA4 y

Lung Fb H0XA5 y

Lung Fb H0XA6 y y y

Lung Fb H0XA7 y

Lung Fb H0XA9 y y y y y y

Lung Fb HOXAlO y y y y y y

Lung Fb HOXAl 1 y

Lung Fb HOXAl 3 y

Lung Fb HOXBl y

Lung Fb H0XB2 y

Lung Fb H0XB3 y y y

Lung Fb H0XB4 y y y

Lung Fb H0XB5 y

Lung Fb H0XB6 y y y

Lung Fb H0XB7 y y y

Lung Fb H0XB8 y y y

Lung Fb H0XB9 y

Lung Fb HOXB 13 y

Lung Fb H0XC4 y

Lung Fb H0XC5 y

Lung Fb H0XC6 y

Lung Fb H0XC8 y

Lung Fb H0XC9 y y y y y y

Lung Fb HOXClO y

Lung Fb HOXCI l y

Lung Fb HOXC 12 y

Lung Fb HOXC 13 y

Lung Fb HOXDl y

Lung Fb H0XD3 y

Lung Fb H0XD4 y y y y y y

Lung Fb H0XD8 y y y y y y

Lung Fb H0XD9 y

Lung Fb HOXDlO y

Lung Fb HOXDI l y

Lung Fb HOXD 12 y

Lung Fb H0XD13 y y y y y y [0085] Table 2. Sequences of synthetic C-terminal biotinylated histone peptides used in this study.

1. H3-K9 triMe ARTKQTARK(3Me)STGGKAPRKQLA

2. H3-K9 diMe ARTKQTARK(2Me)STGGKAPRKQLA

3. H3-K9 monoMe ARTKQTARK(IMe)STGGKAPRKQLA

4. H3-K4 triMe ARTK(3Me)QTARKSTGGKAPRKQLA

5. H3-K4 diMe ARTK(2Me)QTARKSTGGKAPRKQLA

6. H3-K4 monoMe ARTK(IMe)QTARKSTGGKAPRKQLA

7. H3-K27 triMe RKQLATKAARK(3Me)SAP ATGGVKKP

8. H3-K27 diMe RKQLATKAARK(2Me)SAP ATGGVKKP

9. H3-K27 monoMe RKQLATKAARK(IMe)SAPATGGVKKP

10. H3-K36 triMe RKSAPATGGVK(3Me)KPHRYRPGTV

11. H3-K36 diMe RKSAPATGGVK(2Me)KPHRYRPGTV

12. H3-K36 monoMe

RKS APATGGVK( 1 Me)KPHRYRPGTV

13. H4-K20 triMe LGKGGAKRHRK(3Me) VLRDNIQGIT

14. H4-K20 diMe LGKGGAKRHRK(2Me) VLRDNIQGIT

15. H4-K20 monoMe LGKGGAKRHRK( 1 Me)VLRDNIQGIT Table 3. Sequence and location information of the ChIP oligonucleotides used in Figure 2C.

Amplicon

Primer # Sequence (5' - 3') sizes Relative locations to the closest genes

IF C ICCIiA TC 1 IiC I CCA 1 CCIi 197bp 0.2kb upstream of HoxD9 transcription start

IR CTCGCCTCTCTCCCTACTCA

0.4kb downstream to the end of HoxDIO

2F CAATGGGTGTGAGCTTTCCT 207bp 3'UTR

2R CTTCCAGGTGGGCAAGACTA

3F CGCTGTTGTCCGTGCTTACC 227bp across HoxDIO transcription start

3R AGAGCTGTTGGGAAAGGACA

CCTTAATTTCCCTGCAAACG

4F 213bp 1.5kb downstream to the end of HoxD 11 3'UTR

4R AGGCTGCACAGAACTTCCTC

0.6kb downstream to the end of HoxD 11

GACGATGTGACCAAGCAATG

5F 281bp 3'UTR

5R CAAGCCCTCTCCACCTACTG

6F TTGGCGAGCGTTGATATAGA 307bp across HoxD 11 transcription start

6R GGCTTGCTAGCGAAGTCAGA

7F CGGAAAGAGCCAAGTCACTC 215bp 0.6kb upstream of HoxDll transcription start

7R GGTGTGAGGGTGTGAGGTTC

2.3kb downstream to the end of HoxDl 2

8F 265bp

CAAGCTCCTAGCTGGTGGAG 3'UTR

8R CCCAGCCAGGTGGAGAATCC

9F ATGGCTTTCCAGTTGAGGTG 223bp lkb downstream to the end of HoxDl 2 3'UTR

9R CTGGGCCTCTGGTAGAGTTG

1OF GGAAACCCGCAGACAGTTAG 214bp across HoxD12 transcription start

1OR CAACCGGGAATAAGAGGACA

HF CTCGGCTTGGAAATAACGAG 250bp 1.5kb upstream of HoxDl 2 transcription start

HR GCCCAGGCATAGAGACTCAC

0.5kb downstream of HoxD 13 transcription

12F 194bp

TCCTCTTCTGCCGTTGTAGC start

12R GATGACTTGAGCGCATTCTG

Ctrl IF 446bp a geneless region -10 kb upstream of ActG

CAATTGACAGGCAATGATGG gene, very low H3K4me3 and K27me3

Ctrl IR ACAAGAGAGGCCTTGGGAAT intronic region of RNA pol2 gene, very low

Ctrl 2F gcaccacgtccaatgacat 632bp H3K4me3 and K27me3

Ctrl 2R gtgcggctgcttccataa [0086] References

1. Pearson, J. C, Lemons, D. & McGinnis, W. Modulating Hox gene functions during animal body patterning. Nat Rev Genet 6, 893-904 (2005).

2. Plath, K. et al. Role of histone H3 lysine 27 methylation in X inactivation. Science 300, 131-5 (2003).

3. Heard, E. Delving into the diversity of facultative heterochromatin: the epigenetics of the inactive X chromosome. Curr Opin Genet Dev 15, 482-9 (2005).

4. Bernstein, B. E. et al. A bivalent chromatin structure marks key developmental genes in embryonic stem cells. Cell 125, 315-26 (2006).

5. Lee, T. I. et al. Control of developmental regulators by Polycomb in human embryonic stem cells. Cell 125, 301-13 (2006).

6. Varambally, S. et al. The polycomb group protein EZH2 is involved in progression of prostate cancer. Nature 419, 624-9 (2002).

7. Kleer, C. G. et al. EZH2 is a marker of aggressive breast cancer and promotes neoplastic transformation of breast epithelial cells. Proc Natl Acad Sci USA lQQ, 11606-11 (2003).

8. Rinn, J. L., Bondre, C, Gladstone, H. B., Brown, P. O. & Chang, H. Y. Anatomic demarcation by positional variation in fibroblast gene expression programs. PLoS Genet 2, el 19 (2006).

9. Boyer, L. A. et al. Polycomb complexes repress developmental regulators in murine embryonic stem cells. Nature 441, 349-53 (2006).

10. Klose, R. J., Kallin, E. M. & Zhang, Y. JmjC-domain-containing proteins and histone demethylation. Nat Rev Genet 7, 715-27 (2006).

11. Shi, Y. & Whetstine, J. R. Dynamic regulation of histone lysine methylation by demethylases. MoI Cell 25, 1-14 (2007).

12. Yamane, K. et al. PLU-I is an H3K4 demethylase involved in transcriptional repression and breast cancer cell proliferation. MoI Cell 25, 801-12 (2007).

13. Iwase, S. et al. The X-linked mental retardation gene SMCX/JARIDIC defines a family of histone H3 lysine 4 demethylases. Cell 128, 1077-88 (2007).

14. Tahiliani, M. et al. The histone H3K4 demethylase SMCX links REST target genes to X- linked mental retardation. Nature (2007).

15. Secombe, J., Li, L., Carlos, L. & Eisenman, R. N. The Trithorax group protein Lid is a trimethyl histone H3K4 demethylase required for dMyc-induced cell growth. Genes Dev 21, 537-51 (2007).

16. Klose, R. J. et al. The retinoblastoma binding protein RBP2 is an H3K4 demethylase. Cell 128, 889-900 (2007). 17. Eissenberg, J. C. et al. The trithorax-group gene in Drosophila little imaginal discs encodes a trimethylated histone H3 Lys4 demethylase. Nat Struct MoI Biol 14, 344-346 (2007).

18. Christensen, J. et al. RBP2 belongs to a family of demethylases, specific for tri-and dimethylated lysine 4 on histone 3. Cell 128, 1063-76 (2007).

19. Lee, M. G., Norman, J., Shilatifard, A. & Shiekhattar, R. Physical and functional association of a trimethyl H3K4 demethylase and Ring6a/MBLR, a polycomb-like protein. Cell 128, 877-87 (2007).

20. Liang, G., Klose, R. J., Gardner, K. E. & Zhang, Y. Yeast Jhd2p is a histone H3 Lys4 trimethyl demethylase. Nat Struct MoI Biol 14, 243-5 (2007).

21. Seward, D. J. et al. Demethylation of trimethylated histone H3 Lys4 in vivo by JARIDl JmjC proteins. Nat Struct MoI Biol 14, 240-2 (2007).

22. Ginalski, K., Rychlewski, L., Baker, D. & Grishin, N. V. Protein structure prediction for the male-specific region of the human Y chromosome. Proc Natl Acad Sci USA 101, 2305-10 (2004).

23. Greenfield, A. et al. The UTX gene escapes X inactivation in mice and humans. Hum MoI Genet 7, 737-42 (1998).

24. Greenfield, A. et al. An H-YDb epitope is encoded by a novel mouse Y chromosome gene. Nat Genet U, 474-8 (1996).

25. Shi, Y. J. et al. Regulation of LSDl Histone Demethylase Activity by Its Associated Factors. MoI Cell 19, 857-64 (2005).

26. Lee, M. G., Wynder, C, Cooch, N. & Shiekhattar, R. An essential role for CoREST in nucleosomal histone 3 lysine 4 demethylation. Nature 437, 432-5 (2005).

27. Rinn, J. L. et al. Functional demarcation of active and silent chromatin domains in human HOX loci by noncoding RNAs. Cell 129, 1311-23 (2007).

28. Issaeva, I. et al. Knockdown of ALR (MLL2) reveals ALR target genes and leads to alterations in cell adhesion and growth. MoI Cell Biol 27, 1889-903 (2007).

29. Wysocka, J. et al. WDR5 associates with histone H3 methylated at K4 and is essential for H3 K4 methylation and vertebrate development. Cell 121, 859-72 (2005).

30. Dou, Y. et al. Regulation of MLLl H3K4 methyltransferase activity by its core components. Nat Struct MoI Biol 13, 713-9 (2006).

31. Cho, Y. W. et al. PTIP Associates with MLL3- and MLL4-containing Histone H3 Lysine 4 Methyltransferase Complex. J Biol Chem 282, 20395-406 (2007).

32. Trojer, P. & Reinberg, D. Histone lysine demethylases and their impact on epigenetics. Cell 125, 213-7 (2006). 33. Kotake, Y. et al. pRB family proteins are required for H3K27 trimethylation and Polycomb repression complexes binding to and silencing pl6INK4alpha tumor suppressor gene. Genes Dev 2\, 49-54 (2007).

34. Ayton, P. M. & Cleary, M. L. Transformation of myeloid progenitors by MLL oncoproteins is dependent on Hoxa7 and Hoxa9. Genes Dev 17, 2298-307 (2003).

35. Raman, V. et al. Compromised HOXA5 function can limit p53 expression in human breast tumours. Nature 405, 974-8 (2000).

36. Whetstine, J. R. et al. Reversal of histone lysine trimethylation by the JMJD2 family of histone demethylases. Cell 125, 467-81 (2006).

37. S. Lomvardas, G. Barnea, D. J. Pisapia et al., Cell 126 (2), 403 (2006).

38. B. Schuettengruber, D. Chourrout, M. Vervoort et al., Cell 128 (4), 735 (2007).

[0087] Equivalents

Those skilled in the art will recognize, or be able to ascertain using no more than routine experimentation, many equivalents to the specific embodiments of the invention described herein. Such equivalents are intended to be encompassed by the following claims.

Claims

We claim:

1. A method of screening for an enhancer or inhibitor of histone demethylase activity, comprising: contacting a histone H3 peptide with a histone demethylase protein selected from the group consisting of UTX, JMJD3, and a catalytic analog thereof, in the presence and in the absence of a test substance; determining the methylation status of the histone H3 peptide at a lysine 27 position; and identifying a test substance as an enhancer of histone demethylase activity if less mono-, di- or trimethylated H3K27 is found in the presence than in the absence of the test substance, and identifying a test substance as an inhibitor of histone demethylase protein activity if more mono-, di- or trimethylated H3K27 is found in the presence than in the absence of the test substance.

2. The method of claim 1, wherein the histone H3 peptide comprises H3K27me3, H3K27me2, H3K27mel, or unmethylated H3 peptide.

3. The method of claim 1, wherein the UTX catalytic analog comprises a JmjC domain.

4. The method of claim 3, wherein the UTX catalytic analog further comprises a treble-clef zinc finger domain.

5. The method of claim 1, wherein the JMJD3 catalytic analog comprises a JmjC domain.

6. The method of claim 5, wherein the JMJD3 catalytic analog further comprises a treble-clef zinc finger domain.

7. An isolated complex comprising a UTX protein or a catalytic analog thereof, and a JMJD3 protein or a catalytic analog thereof.

8. The isolated complex of claim 7, wherein the UTX catalytic analog comprises a JmjC domain and optionally further comprises a treble-clef zinc finger domain, and wherein the JMJD3 catalytic analog comprises a JmjC domain and optionally further comprises a treble- clef zinc finger domain.

9. The isolated complex of claim 7, further comprising a histone H3 peptide.

10. A method for reversing a tumorigenic state of a cell comprising administering a UTX protein, a JMJD3 protein or a catalytic analog thereof, thereby decreasing methylation of a histone H3 peptide at a lysine 27 position in the cell.

11. The method of claim 10, wherein the cell has a mutated retinoblastoma susceptibility gene.

12. A method for reversing a tumorigenic state of a cell comprising administering an agent that increases the level or activity of a UTX protein or a JMJD3 protein, thereby decreasing methylation of a histone H3 peptide in the cell.

13. A method for differentiating an embryonic stem cell, comprising administering UTX protein, a JMJD3 protein, a catalytic analog of a UTX or JMJD3 protein, or an agent that increases the level or activity of a UTX protein or a JMJD3 protein, thereby decreasing methylation of a histone H3 peptide in the cell.

14. A method for increasing the transcription of a gene in a cell, comprising contacting the cell with, or administering into the cell, a UTX protein, a JMJD3 protein, a catalytic analog of a UTX or JMJD3 protein, or an agent that increases the level or activity of a UTX protein or a JMJD3 protein.

15. The method of claim 14, wherein the gene is a tumor suppressor gene or a gene involved in stem cell pluripotency.

16. The method of claim 14, wherein the gene is associated with a histone that is methylated by the histone methyltransferase EZH2.

17. A method for decreasing the transcription of a gene in a cell, comprising contacting the cell with an agent that decreases the protein or activity level of a UTX protein or a JMJD3 protein in the cell.

18. The method of claim 17, wherein the agent is a siRNA.

19. A method of identifying a compound that modulates an activity of UTX or JMJD3 demethylase, comprising:

(a) contacting the demethylase with a H3K27 histone peptide substrate in the presence of a test compound; and

(b) detecting the level of demethylation of the H3K27 histone protein substrate under conditions sufficient for demethylation, wherein a change in demethylation of the protein substrate as compared with the level of demethylation in the absence of the test compound indicates that the test compound is a modulator of the demethylase activity of the demethylase.

20. A method of altering the level of a UTX or JMJD3 demethylase in a eukaryotic cell, comprising:

(a) transforming the cell with a recombinant DNA construct encoding the UTX or JMJD3 demethylase; and

(b) growing the transformed cell under conditions that are suitable for expression of the recombinant DNA construct, wherein expression of the recombinant DNA construct results in production of an altered level of the UTX or JMJD3 demethylase in the transformed cell.

21. A method of screening for an enhancer or inhibitor of histone demethylase activity, comprising: contacting a histone H3 peptide with a histone demethylase protein in the presence and in the absence of a test substance, wherein the histone demethylase protein comprises

(a) an amino acid sequence having at least 95% identity to the amino acid sequence set forth in SEQ ID NO: 1 or 3;

(b) an amino acid sequence set forth in SEQ ID NO: 1 or 3 with 1 to 10 amino acid substitutions;

(c) an amino acid sequence comprising at least 10 consecutive residues of residues 1133-1241 and optionally at least 10 consecutive residues of residues 168-282 set forth in SEQ ID NO: 1 and which is less than 1401 amino acids in length;

(d) an amino acid sequence comprising at least 10 consecutive residues of residues 1339-1410 or least 10 consecutive residues of residues 1377-1485, or the combination thereof, as set forth in SEQ ID NO: 3 and which is less than 1682 amino acids in length; wherein the polypeptides of (a), (b), (c) and (d) are capable of histone demethylation, and wherein the histone H3 peptide comprises H3K27me3, H3K27me2, H3K27mel, or unmethylated H3 peptide; determining the methylation status of the histone H3 peptide at a lysine 27 position; and identifying a test substance as an enhancer of histone demethylase activity if less mono-, di- or trimethylated H3K27 is found in the presence than in the absence of the test substance, and identifying a test substance as an inhibitor of histone demethylase protein activity if more mono-, di- or trimethylated H3K27 is found in the presence than in the absence of the test substance.