(19) TZZ¥ ZZ_T (11) EP 3 260 540 A1 (12) EUROPEAN PATENT APPLICATION (43) Date of publication: (51) Int Cl.: 27.12.2017 Bulletin 2017/52 C12N 15/113 (2010.01) A61K 9/127 (2006.01) A61K 31/713 (2006.01) C12Q 1/68 (2006.01) (21) Application number: 17000579.7 (22) Date of filing: 12.11.2011 (84) Designated Contracting States: • Sarma, Kavitha AL AT BE BG CH CY CZ DE DK EE ES FI FR GB Philadelphia, PA 19146 (US) GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO • Borowsky, Mark PL PT RO RS SE SI SK SM TR Needham, MA 02494 (US) • Ohsumi, Toshiro Kendrick (30) Priority: 12.11.2010 US 412862 P Cambridge, MA 02141 (US) 20.12.2010 US 201061425174 P 28.07.2011 US 201161512754 P (74) Representative: Clegg, Richard Ian et al Mewburn Ellis LLP (62) Document number(s) of the earlier application(s) in City Tower accordance with Art. 76 EPC: 40 Basinghall Street 11840099.3 / 2 638 163 London EC2V 5DE (GB) (71) Applicant: The General Hospital Corporation Remarks: Boston, MA 02114 (US) •Thecomplete document including Reference Tables and the Sequence Listing can be downloaded from (72) Inventors: the EPO website • Lee, Jeannie T •This application was filed on 05-04-2017 as a Boston, MA 02114 (US) divisional application to the application mentioned • Zhao, Jing under INID code 62. San Diego, CA 92122 (US) •Claims filed after the date of receipt of the divisional application (Rule 68(4) EPC). (54) POLYCOMB-ASSOCIATED NON-CODING RNAS (57) This invention relates to long non-coding RNAs (IncRNAs), libraries of those ncRNAs that bind chromatin modifiers, such as Polycomb Repressive Complex 2, inhibitory nucleic acids and methods and compositions for targeting IncRNAs. EP 3 260 540 A1 Printed by Jouve, 75001 PARIS (FR) (Cont. next page) EP 3 260 540 A1 2 (Cont. next page) EP 3 260 540 A1 d 3 EP 3 260 540 A1 Description FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT 5 [0001] This invention was made with Government support under Grant No. RO1-GM-090278 awarded by the National Institutes of Health. The Government has certain rights in the invention. RELATED APPLICATIONS 10 [0002] This application claims priority of U.S. provisional nos. 61/412,862, filed on November 12, 2010, 61/425,174 filed on December 20, 2010, and 61/512,754 filed on July 28, 2011, the entirety of each of which is incorporated by reference herein. TECHNICAL FIELD 15 [0003] This invention relates to long non-coding RNAs (IncRNAs) that function to modulate gene expression, and methods of using them or inhibitory nucleic acids that bind them, to modulate gene expression. BACKGROUND 20 [0004] Transcriptome analyses have suggested that, although only 1-2% of the mammalian genome is protein-coding, 70-90% is transcriptionally active (Carninci et al., 2005; Kapranov et al., 2007; Mercer et al., 2009). Ranging from 100 nt to >100 kb, these transcripts are largely unknown in function, may originate within or between genes, and may be conserved and developmentally regulated (Kapranov et al., 2007; Guttman et al., 2009). Recent discoveries argue that 25 a subset of these transcripts play crucial roles in epigenetic regulation. For example, genes in the human HOX-D locus are regulated in trans by HOTAIR RNA, produced by the unlinked HOX-C locus (Rinn et al., 2007), and during X- chromosome inactivation, Tsix, RepA, and Xist RNAs target a chromatin modifier in cis to control chromosome-wide silencing (Zhao et al., 2008). Interestingly, all four RNAs bind and regulate Polycomb Repressive Complex 2 (PRC2), the complex that catalyzes trimethylation of histone H3-lysine27 (H3-K27me3)(Schwartz and Pirrotta, 2008). These 30 observations support the idea that long ncRNAs are ideal for targeting chromatin modifiers to specific alleles or unique locations in the genome (Lee, 2009) (Lee, 2010). [0005] RNA-mediated recruitment is especially attractive for Polycomb proteins. First identified in Drosophila as home- otic regulators, Polycomb proteins are conserved from flies to mammals and control many aspects of development (Ringrose and Paro, 2004; Boyer et al., 2006; Lee et al., 2006; Schuettengruber et al., 2007; Pietersen and van Lohuizen, 35 2008; Schwartz and Pirrotta, 2008). Mammalian PRC2 contains four core subunits, Eed, Suz 12, RbAp48, and the catalytic Ezh2. In humans, aberrant PRC2 expression is linked to cancer and disease (Sparmann and van Lohuizen, 2006; Bernardi and Pandolfi, 2007; Miremadi et al., 2007; Rajasekhar and Begemann, 2007; Simon and Lange, 2008). Despite growing recognition of Polycomb’s role in health, little is known about their regulation in vivo. In flies, Polycomb complexes may contain sequence-specific DNA-binding factors, such as Zeste, Pipsqueak (PSQ), or Pho, to help bind 40 Polycomb-response elements (PRE) (Ringrose and Paro, 2004; Schwartz and Pirrotta, 2008). By contrast, mammalian Polycomb complexes are not thought to contain such subunits. Therefore, their mechanism of recruitment to thousands of genomic locations remains poorly understood, though PRE-like elements (Sing et al., 2009; Woo et al., 2010) and Jarid2 may facilitate binding (Li et al.; Pasini et al.; Peng et al., 2009; Shen et al., 2009). Interestingly, several PRC2 subunits have potential RNA-binding motifs (Denisenko et al., 1998; Bernstein and Allis, 2005; Bernstein et al., 2006b) 45 - a possibility borne out by postulated functional interactions between Tsix/RepA/Xist RNA and PRC2 for X-inactivation (Zhao et al., 2008) and by HOTAIR and PRC2 for HOX regulation (Rinn et al., 2007). Recent work also identified several short RNAs of 50-200 nt as candidate PRC2 regulators (Kanhere et al., 2010). Control of Polycomb repressive comp lex 1 (PRC1) may also involve RNA (Yap et al., 2010). [0006] In spite of their ubiquity, the structure and function of many long ncRNAs remain largely uncharacterized. Recent 50 studies suggest that some long ncRNAs may have a function as an epigenetic regulator/RNA cofactor in chromatin remodeling and tumor suppression. Although knockdown technologies employing siRNAs and shRNAs have become staples in functional analysis of microRNAs (miRNAs) and cytoplasmically localized messenger RNAs (mRNAs) (4-6), these methods have been reported in some instances to be less consistently effective for long ncRNAs localized to the nucleus (Jepsen et al., Oligonucleotides, 14, 130-146 (2004)). 55 SUMMARY [0007] A method, referred to herein as "RNA immunoprecipitation (RIP)-seq," was used to identify a genome-wide 4 EP 3 260 540 A1 pool of>9,000 polycomb repressive complex 2 (PRC2)-interacting RNAs in embryonic stem cells (referred to herein as the "PRC2 transcriptome"). The transcriptome includes antisense, intergenic, and promoter-associated transcripts, as well as many unannotated RNAs. A large number of transcripts occur within imprinted regions, oncogene and tumor suppressor loci, and stem-cell-related bivalent domains. Evidence for direct RNA-protein interactions, some via the Ezh2 5 subunit, is provided. Further evidence is provided that inhibitory oligonucleotides that specifically bind to these PRC2- interacting RNAs can successfully up-regulate gene expression in a variety of separate and independent examples, presumably by inhibiting PRC2-associated repression. Gtl2 RNA was identifed as a PRC2 cofactor that directs PRC2 to the reciprocally imprinted Dlk1 coding gene. Thus, Polycomb proteins interact with a genome-wide family of RNAs, some of which may be used as biomarkers and therapeutic targets for human disease. 10 [0008] In one aspect, the invention provides methods for preparing a plurality of validated cDNAs complementary to a pool of nuclear ribonucleic acids (nRNAs). The methods include providing a sample comprising nuclear ribonucleic acids, e.g., a sample comprising nuclear lysate, e.g., comprising nRNAs bound to nuclear proteins; contacting the sample with an agent, e.g., an antibody, that binds specifically to a nuclear protein that is known or suspected to bind to nuclear ribonucleic acids, under conditions sufficient to form complexes between the agent and the protein; isolating the com- 15 plexes; synthesizing DNA complementary to the nRNAs to provide an initial population of cDNAs; PCR-amplifying, if necessary, using strand-specific primers; purifying the initial population of cDNAs to obtain a purified population of cDNAs that are at least about 20 nucleotides (nt) in length, e.g., at least 25, 50, 75, 100, 150, 200, or 250 nt in length; sequencin g at least part or substantially all of the purified population of cDNAs; aligning reads to a reference genome and retaining only those that are aligned; selecting high-confidence cDNA sequences, e.g., based on two criteria-- (1) that the candidate 20 transcript has a minimum read density in reads per kilobase per million reads (RPKM) terms (e.g., above a desired threshold); and (2) that the candidate transcript is enriched in the wildtype library versus a suitable control library (such as an IgG pulldown library or a protein-null pulldown library); thereby preparing the plurality of cDNAs. [0009] Some examples of nuclear proteins that are known or suspected to bind to nuclear ribonucleic acids include Ezh2 (Zhao et al., Science. 2008 Oct 31;322(5902):750-6; Khalil ct al., Proc Natl Acad Sci U S A. 2009 Jul 25 14;106(28):11667-72. Epub 2009 Jul 1); G9a (Nagano et al., Science. 2008 Dec 12;322(5908):1717-20. Epub 2008 Nov 6); and Cbx7 (Yap et al., Mol Cell. 2010 Jun 11;38(5):662-74.) [0010] In some embodiments, the invention includes methods for preparing a plurality of validated cDNAs comple- mentary to a pool of nuclear