GTF2i inhibitors and uses thereof

Abstract

The present invention refers to a molecule for use in the treatment and / or prevention of DUX4- r positive leukemia wherein said molecule is selected from the group consisting of: - an inhibitor of GTF2I expression and / or function; - an inhibitor of HDAC expression and / or function; and - an inhibitor of c-Myc expression and / or function.

Description

[0001] GTF2I INHIBITORS AND USES THEREOF

[0002] Field of the Invention

[0003] The present invention, by identifying the pathway regulating GTF2I expression, provides selective drugs reducing GTF2I expression and blocking DUX4-r leukemogenic activity. The drugs reducing GTF2I expression provide a targeted therapy for the treatment of B-ALL, in particular for DUX4-r+ B- ALL.

[0004] Background

[0005] B-cell acute lymphoblastic leukemia (B-ALL) is the most common pediatric malignancy and the most frequent cause of death from cancer at young age (I-3). B-ALL is a highly heterogeneous disease driven by recurrent chromosomal alterations giving rise to multiple disease subtypes. Therefore, the characterization of the leukemic pathways associated to specific B-ALL drivers could allow the discovery of new therapeutic vulnerabilities in a personalized medicine perspective (4).

[0006] Recurrent translocations of the double homeobox 4 (DUX4) gene to the immunoglobulin heavy chain (IGH) locus have been reported in up to 10% of B-ALL patients and define a new B-ALL subtype, characterized by ETS-Related Gene (ERG) deregulation, positivity for CD371, a distinctive gene expression profile and widespread hypomethylation compared to healthy B cells or other ALL subtypes (5-9).

[0007] DUX4 is a sequence-specific transcription factor whose expression is normally confined to early embryonic development, where it activates the expression of cleavage-stage genes (10-12). DUX4 possesses an N-terminal DNA binding domain and activates gene expression by recruiting CREB- binding protein / ElA binding protein p300 (CBP / EP300) coactivators through its C-terminal transcription activation domain (CTD) (13).

[0008] All DUX4 translocations in B-ALL maintain the region encoding for the DNA-binding domain but result in random truncations of the CTD, which, in most cases, is substituted by random ammo acids encoded by the IGH locus giving rise to DUX4-IGH chimeras (5-7). The resulting rearranged DUX4 (DUX4-r) is expressed specifically in B-cell precursors (5-7), where wild-type (wt) DUX4 is never expressed.

[0009] The DUX4 rearrangement is a clonal event occurring early in leukemogenesis and maintained in relapsed leukemia. Importantly, DUX4-r silencing in patient-derived xenograft (PDX) or the NALM6 cell line derived from DUX4-r positive (DUX4-r+) B-ALL patients supports an essential role of DUX4- r in the development and maintenance of this leukemia subtype (5, 14). Nevertheless, it was recently proposed that wt DUX4 and DUX4-r function interchangeably and that leukemia development requires “just right” levels of either one of them. Too much wt DUX4 or DUX4-r would result in apoptosis and too little would just block cell proliferation (75).

[0010] In addition to DUX4 rearrangements, ERG alterations are observed almost exclusively in the DUX4-r B-ALL subtype (16-18). However, these alterations are absent in up to 40% of DUX4-r patients, are often subcl onal and inconsistent between diagnosis and relapse, thus indicating that ERG alterations do not drive leukemogenesis but represent a secondary event in disease progression (16, 19). Hence, additional factors in addition to ERG alterations may be required for DUX4-r induced leukemogenesis. The proposed models for DUX4-r driven B-ALL, including cooperation with ERG alterations or just right DUX4 levels, are not fully supported by the literature (5, 6, 10, 15). Additional studies suggest a novel activity of the DUX4-r oncogene (7, 8, 20), but are based on very different genetic background or cellular systems. Functional enrichment analyses on DUX4-r transcriptomic data so far failed to identify specifically dysregulated pathways (7-9). RNA sequencing (RNA-seq) has been performed upon DUX4-r overexpression in NALM6 cells, which nevertheless already express it endogenously, or 72h after DUX4-r knockdown (9). Moreover, chromatin immunoprecipitation sequencing (ChlP-seq) data for endogenous DUX4-r in NALM6 cells have been compared to those of ectopically overexpressed wt DUX4 in pluripotent stem cells or human myoblasts (6, 9, 10, 21, 22). Thus, it is very difficult to discern primary DUX4-r transcriptional targets from secondary events and there remains a need to compare the activity of wt DUX4 and DUX4-r in a relevant system in order to define the molecular mechanism through which DUX4-r promotes B-ALL and identify vulnerabilities that could be exploited for therapeutic purposes.

[0011] Summary of the invention

[0012] Within the present invention, it was compared side-by-side the molecular and functional activity of wt DUX4 and representative patient-derived DUX4-r variants in a B-ALL context. Despite sharing the same DNA-binding domain, the inventors found that wt DUX4 and DUX4-r regulate different gene sets and control very differently cell behavior. Indeed, only DUX4-r activates genes belonging to the DUX4- r+ B-ALL patient gene signature and stimulates B-ALL cell proliferation. The inventors found that DUX4-r acquires these abilities through a gain of interaction with the transcription factor GTF2I (General Transcription Factor 2 1, also defined as GTF2-I / TFII-I / BAP135 / WBSCR6).

[0013] The present inventors further demonstrated that the activity of DUX4-r strictly depends on the cellular availability of GTF2I. GTF2I was identified as a key co-factor of rearranged double homeobox 4 (DUX4-r) in B-cell acute lymphoblastic leukemia (B-ALL). The inventors found that GTF2I is required for DUX4-r leukemogenic activity. Indeed, genetic or pharmacological GTF2I silencing block DUX4- r activity inducing cell death selectively in DUX4-r positive cells. Accordingly, genetic or pharmacological GTF2I targeting inhibits DUX4-r activity in cellular and animal leukemia models. Within the present invention it was also found that HDAC6, known to modulate the expression of the oncogene c-Myc, is required for the expression of GTF2I and for the activity of DUX4-r.

[0014] Based on these results, the HDAC6 -> c-Myc -> GTF2I pathway can be targeted with drugs inhibiting HDAC6 or c-MYC for the treatment of DUX4-r positive leukemia.

[0015] Collectively, these data clarify the molecular mechanism responsible for DUX4-r+ B-ALL and identify a therapeutic option for this leukemia subtype.

[0016] There is currently no targeted treatment for DUX4-r positive leukemia. For B-ALL, there have been little improvements in cure rate over the last 20 years due to a lack of new treatment options. Current therapies for B-ALL are associated with a number of side-effects including anemia, infections, mucositis, hair loss, infertility, cognitive decline, heart problems and secondary cancer. Moreover, B- ALL in older patients tend to be more resistant to treatment. Patients older than 35 years have decreased remission duration and a harder time tolerating chemotherapy. Although the 5-year survival rate for people under age 20 is 89%, for people aged 20 and older is only 38%. Chemotherapy is associated with life-long health sequelae and relapse occurs in 50% of adult and 20% of pediatric cases. A targeted therapy for B-ALL caused by DUX4-r can increase therapeutic safety and efficacy. For the first time the present invention, by identifying the pathway regulating GTF2I expression, provides selective drugs reducing GTF2I expression and blocking DUX4-r leukemogenic activity.

[0017] It is therefore an object of the invention a molecule for use in the treatment and / or prevention of DUX4- r positive leukemia, wherein said molecule is selected from the group consisting of: an inhibitor of GTF2I expression and / or function; an inhibitor of HD AC expression and / or function; and an inhibitor of c-Myc expression and / or function.

[0018] Preferably, the DUX4-r positive leukemia is acute lymphoblastic leukemia, preferably B-cell acute lymphoblastic leukemia (B-ALL), preferably said acute lymphoblastic leukemia or B-cell acute lymphoblastic leukemia (B-ALL) is caused by DUX4-r.

[0019] Preferably said inhibitor of GTF2I expression and / or function is an inhibitor of HD AC expression and / or function, preferably of HDAC6.

[0020] Preferably, said inhibitor of GTF2I expression and / or function is an inhibitor of c-Myc.

[0021] Preferably said inhibitor of GTF2I expression and / or function is an inhibitor of the DUX4-r / GTF2I interaction. Preferably said inhibitor of the DUX4-r / GTF2I interaction is a GTF2I-denved peptide, preferably a peptide comprising or consisting of GTF2I Domain B of SEQ ID No: 66 or a derivative thereof, wherein said derivative has at least 80% or at least 90% sequence identity with respect to SEQ ID No: 66 and / or is functionalized at the C- or N-terminus.

[0022] Preferably the inhibitor of the invention is selected from the group consisting of: a) polynucleotide capable of inhibiting the expression of the GTF2I gene and / or of HD AC gene and / or of c-Myc gene or a polynucleotide coding for said polynucleotide; b) a polypeptide, preferably an antibody or a synthetic or recombinant derivative thereof; c) a polynucleotide coding for said polypeptide or a functional derivative thereof; d) a small molecule; e) a component of the CRISPR / Cas9 system, for example a sgRNA (single strand guide RNA, ("single guide RNA")), a gRNA (guide RNA) or a guide sequence and f) a component of the protein degradation system, for example by proteolysis, of the cell, for example of proteolysis targeting chimeras (for example, proteolysis targeting chimera “PROTAC” or molecular glue).

[0023] Preferably, said inhibitor of HDAC is selected from Vorinostat (SAHA), belinostat (PXD-101), panobinostat (LBH-589), Dacinostat (LAQ824), CUDC-101, Abexinostat (PCI-24781), Ricolinostat (ACY-1215), Fimepinostat (CUDC-907), Resminostat and Trichostatin A (TSA).

[0024] Preferably, said inhibitor of HDAC is a selective inhibitor of HDAC6, preferably said selective inhibitor of HDAC6 is MPT0G211, Tubastatin-A or HDAC6 degraders.

[0025] Preferably said inhibitor of c-Myc is selected from MYCi975 and WBC100.

[0026] Preferably, said inhibitor is a polynucleotide capable of inhibiting the expression of the GTF2I gene and / or of HDAC gene and / or of c-Myc gene, preferably an antisense construct, an RNA inhibitor, more preferably a siRNA, a shRNA, a microRNA, a sgRNA, a gRNA, a guide sequence or an antisense oligonucleotide, or a polynucleotide coding for it.

[0027] The molecule of the invention is preferably used in combination with a further therapeutic agent and / or a therapeutic intervention.

[0028] The further therapeutic agent is preferably a chemotherapy drug.

[0029] The further therapeutic agent is also preferably a drug antagonizing the activity of inhibitors of apoptosis, preferably an inhibitor of cIAP and XIAP, more preferably the further therapeutic agent is Birinapant.

[0030] Another object of the invention is a pharmaceutical composition for use as defined above, said pharmaceutical composition comprising a molecule as defined above and at least one pharmaceutical acceptable carrier and / or vehicle, preferably said pharmaceutical composition is for use in the treatment and / or prevention of B-cell acute lymphoblastic leukemia (B-ALL), preferably said B-cell acute lymphoblastic leukemia (B-ALL) is caused by DUX4-r.

[0031] A further object of the invention is a vector comprising or expressing the polynucleotide as defined above or herein for use as disclosed above or herein.

[0032] Another object of the invention is a genetically modified host cell capable of expressing in suitable conditions the polypeptide or the polynucleotide as defined above or herein or comprising the polynucleotide as defined above or herein or at least a component of e) or f) as defined above or comprising the vector as defined above for use as defined above.

[0033] Another object of the invention is an in vitro method for evaluating the risk and / or for diagnosis and / or for prognosis and / or for monitoring progression and / or for monitoring the efficacy of a therapeutic treatment and / or for screening of DUX4-r positive leukemia in a subject or in an ex vivo or in vitro system, preferably an organoid as a bone marrow organoid, comprising the steps of: a) detect or measure the amount or activity of the GTF2I protein and / or of HD AC protein and / or of c- Myc protein or of fragments thereof or of the polynucleotide encoding said protein or its fragments in an isolated biological sample obtained from the subject, which may be optionally comprised in an ex vivo or in vitro system, and b) comparison with a proper control.

[0034] DETAILED DESCRIPTION OF THE INVENTION

[0035] Translocations producing rearranged versions of the transcription factor DUX4 (DUX4-r) are one of the most frequent causes of B-cell acute lymphoblastic leukemia (B-ALL). DUX4-r retains the DNA binding domain of wt DUX4, but is truncated on the C-terminal transcription activation domain. The mechanism through which DUX4-r causes leukemia is unknown, and no targeted therapy is currently available. In the present invention it was found that the rearrangement leads to both a loss and a gain of function in DUX4-r. Loss of CBP / EP300 transcriptional co-activators interaction leads to inability to bind and activate repressed chromatin. Concurrently, gain of interaction with the transcription factor GTF2I redirects DUX4-r toward leukemogenic targets. Importantly, this neomorphic activity exposes an Achilles' heel whereby DUX4-r positive leukemia cells are exquisitely sensitive to GTF2I targeting, which inhibits DUX4-r leukemogenic activity. Elucidating the molecular mechanism through which DUX4-r causes leukemia, the invention provides a novel therapeutic avenue tailored to this B-ALL subtype. Importantly, despite having the same DNA-binding domain, wt DUX4 and DUX4-r associate to and activate highly different gene sets. Only DUX4-r is capable of activating genes belonging to the DUX4-r signature in B-ALL patients, which encode for proteins involved in cellular adhesion, migration and cancer pathways. Such DUX4-r activity requires GTF2I, whose targeting abrogates DUX4-r chromatin binding, target gene activation and leukemogenic activity.

[0036] Therefore the present invention provides an inhibitor of GTF2I for use in the treatment of B-cell acute lymphoblastic leukemia (B-ALL), preferably said B-cell acute lymphoblastic leukemia (B-ALL) is caused by DUX4-r.

[0037] The present invention also refers to an inhibitor of GTF2I expression and / or function; an inhibitor of HD AC expression and / or function; or an inhibitor of c-Myc expression and / or function for use in the treatment and / or prevention of diseases or pathologies caused by or resulting from the presence and / or expression of DUX4-r.

[0038] The objects of the present invention may therefore be used to treat and / or prevent any disease or pathology caused by or resulting from the presence and / or expression of DUX4-r, such as acute lymphoblastic leukemia caused by DUX4-r or B-cell acute lymphoblastic leukemia (B-ALL) caused by DUX4-r.

[0039] In the context of the present invention a disease or pathology “caused by DUX4-r” means a disease caused by rearranged versions of DUX4 affecting the C-terminal portion of the protein.

[0040] In the context of the present invention a “DUX4-r positive leukemia” means a leukemia wherein at least one cell is positive for DUX4-r or a leukemia caused by rearranged versions of DUX4 affecting the C- terminal portion of the protein, e g. it comprises acute lymphoblastic leukemia or B-cell acute lymphoblastic leukemia preferably caused by DUX4-r.

[0041] As used herein an inhibitor of GTF2I or HDAC or c-Myc expression and / or activity refers to any molecule or agent which reduces the expression of GTF2I or HDAC or c-Myc gene respectively or reduces the activity of GTF2I or HDAC or c-Myc respectively, i.e. the downstream signaling of GTF2I or HDAC or c-Myc respectively. Said inhibitor can be e.g. a nucleic acid, a small molecule, an antibody or any agent which inhibits the expression of GTF2I or of HDAC or of c-myc.

[0042] The term “molecule” in the present invention includes the term “inhibitor of ... ” as defined herein. In the present invention, the herein defined molecules or inhibitors may be used in combinations. In a preferred embodiment of the invention, GTF2I is a protein and it preferably has essentially the amino acid sequences disclosed as NCBI Accession no. NP_127492.1, NP_127493.1, NP_127494.1, NP_001509.3, NP_001157108.1 or NP_001267729.1.

[0043] In another preferred embodiment, GTF2I is a gene and it preferably has essentially the nucleotide sequence disclosed as NCBI Accession no. NG_008179.2.

[0044] In another preferred embodiment, GTF2I is a mRNA and it preferably has essentially the nucleotide sequences disclosed as NCBI Accession no. NM_033001.4, NM_032999.4, NM_033000.4, NM_001518.5, NM_001163636.3or NM 001280800.2.

[0045] In preferred embodiment, c-Myc is a protein and it preferably has essentially the amino acid sequences disclosed as NCBI Accession no. NP_002458.2 or NP_001341799.1.

[0046] In another preferred embodiment, c-Myc is a gene and it preferably has essentially the nucleotide sequence disclosed as NCBI Accession no. NG_007161.2.

[0047] In another preferred embodiment, c-Myc is a mRNA and it preferably has essentially the nucleotide sequences disclosed as NCBI Accession no. NM_002467.6 or NM_001354870.1.

[0048] In preferred embodiment, HDAC6 is a protein and it preferably has essentially the amino acids sequences disclosed as NCBI Accession no. NP_001308154. 1, NP_001308155.1, NP_001308156.1, NP 001308157.1, NP 001308158.1, NP 001308159.1, NP 001308160.1 or NP_006035.2.

[0049] In another preferred embodiment, HDAC6 is a gene and it preferably has essentially the nucleotide sequence disclosed as NCBI Accession no. NG_012560.2.

[0050] In another preferred embodiment, HDAC6 is a mRNA and it preferably has essentially the nucleotide and amino acids sequences disclosed as NCBI Accession no. NM_001321225.2, NM_001321227.2, NM 001321226.2, NM 001321228.2, NM 001321229.1, NM 001321230.2, NM 006044.4 or NM_001321231.2.

[0051] The term "histone deacetylase" or "HD AC", as used herein, refers to an enzyme that removes acetyl groups from histones. There are currently 18 known HDACs, which are classified into four groups. Class I HDACs, includes HDAC 1-3 and HDAC8. Class II HDACs include HDAC4-7 and HD AC9- 10. Class III HDACs (also known as the sirtuins) include SIRT1-7. Class IV HDACs, which contains only HDAC11, has features of both Class I and II HDACs.

[0052] The term "HDAC inhibitor" or "HDACi" as used herein, refers to a compound natural or synthetic that inhibits histone deacetylase activity. There exist different classes of HDACi in function of their selectivity for their substrates divided in classical HDACi, selective class I HDACi and selective class II HDACi. A "classical HDACi" (also known as pan-HDACi) refers thus to a compound natural or not which has the capability to inhibit the histone deacetylase activity independently of the class of HDACs. Therefore a classical HDACi is a non-selective HDACi. By "non-selective" it is meant that said compound inhibits the activity of classical HDACs (i.e. class I, II and IV) with a similar efficiency independently of the class of HD AC. Examples of classical HDACi include, but are not limited to, Belmostat (PDX-101), Vormostat (SAHA; Cat. No.: HY- 10221. CAS No. : 149647-78-9) and Panobinostat (LBH-589).

[0053] A "selective class I HDACi" is selective for class HDACs (i.e. HD AC 1-3 and 8) as compared with class II HDACs (i.e. HDAC4-7, 9 and 10). By "selective" it is meant that selective class I HDACi inhibits class I HDACs at least 5-fold, preferably 10-fold, more preferably 25-fold, still preferably 100- fold higher than class II HDACs. Selectivity of HDACi for class I or class II HDACs may be determined according to previously described method (Kahn et al. 2008). Examples of selective class I HDACi include, but are not limited to, valproic acid (VP A), Romidepsin (FK-228) and Entinostat (MS-275).

[0054] A "selective class II HDACi" is selective for class II HDACs (i.e. HDAC4-7, 9 and 10) as compared with class I HDACs (i.e. HD AC 1-3 and 8). By "selective" it is meant that selective class II HDACi inhibits class II HDACs at least 5 -fold, preferably 10-fold, more preferably 25- fold, still preferably 100-fold higher than class I HDACs. Examples of selective class II HDACi include, but are not limited to, include MC- 1568 (aryloxopropenyl)pyrrolyl hydroxamate).

[0055] Selective inhibitors for an HD AC family or for a specific isoform, especially HDAC6, may be particularly useful for treating pathologies related to proliferative disorders and protein accumulation, immune system disorders and neurological and neurodegenerative disease, such as stroke, Huntington's disease, AES and Alzheimer's disease. Particularly for the HDAC6 isoform, different substrates have been identified, such as a-tubulin, Hsp90 (Heat Shock Protein 90), cortactin, b-catenin. Modulation of these proteins’ acetylation by HDAC6 has been correlated with several important processes, such as immune response (J. Med. Chem. (2012), 55, 639-651 ; Mol. Cell. Biol. (2011 ), 31 (10), 2066-2078), regulation of microtubule dynamics, including cell migration and cell-cell interaction (Aldana- Masangkay et al., J. Biomed. Biotechnol. (2011 ), 2011 , 875824), and degradation of degenerated proteins. In addition, HDAC6 is involved in the process of catabolism of degraded proteins through the complex known as aggresome: HDAC6 is able to bind polyubiquitinated proteins and dynein, thus activating a kind of delivery of denatured proteins along the microtubules to the aggresome (Kawaguchi et al., Cell (2003) 115 (6), 727-738). Alteration of this HDAC6 cytoprotective activity has been correlated with various neurodegenerative pathologies such as Parkinson's disease (Outerio et al., Science (2007), 317 (5837), 516-519) and Huntington's disease (Dompierre et al., J. Neurosci. (2007), 27(13), 3571-3583), wherein the accumulation of degraded proteins is a common pathological feature. Furthermore, HDAC6 is involved in regulating many oncological proteins, especially in hematologic tumors, such as various types of leukemia (Fiskus et al., Blood (2008), 112(7), 2896-2905) and multiple myeloma (Hideshima et al., Proc. Natl. Acad. Sci. USA (2005), 102(24), 8567-8572). Regulation of α- tubulin acetylation by HDAC6 may be implicated in metastasis onset, wherein cellular motility plays an important role (Sakamoto et al., J. Biomed. Biotechnol. (2011 ), 2011 , 875824). Recently, HDAC6 has attracted interest as a novel immune-oncology target, due to the fact that this enzyme was shown to be an obligate regulator of the expression of the immune-checkpoint protein PD-L1 (Lienlaf et al. Mol Oncol 2016 May; 10(5): 735- 750). HDAC6 inhibitors were shown to be effective in preclinical immuno- oncology models and to enhance the activity of anti PD-1 antibodies (Ray et al., Leukemia 2018 Mar;32(3):843-846. Keremu et al Cancer Chemother Pharmacol. 2019 Feb;83(2):255-264. Knox et al. Sci Rep. 2019 Apr 16;9(1 ) :6136).

[0056] HD AC inhibition relies mainly on a mechanism based on the inhibition of the HD AC enzymatic activity which can be determined by a variety of methods well known by the skilled person. Usually, these methods comprise assessing the lysine deacetylase activity of HD AC enzymes using colorimetric HD AC assays. Commercial kits for such techniques are available (see for example, Histone Deacetylase (HD AC) Activity Assay Kit (Fluorometric) purchased from Abeam or Sigma- Aldrich). These methods are ideal for the determination of IC50 values of known or suspected HD AC inhibitors.

[0057] Many HDAC inhibitors are known and, thus, can be synthesized by known methods from starting materials that are known, may be available commercially, or may be prepared by methods used to prepare corresponding compounds in the literature.

[0058] A preferred class of HDAC inhibitors are hydroxamic acid inhibitors which are disclosed e. g. in WO 97 / 35990, US-A 5, 369, 108, US-A 5, 608, 108, US-A 5, 700, 811, WO 01 / 18171, WO 98 / 55449, WO 93 / 12075, WO 01 / 49290, WO 02 / 26696, WO 02 / 26703, JP 10182583, WO 99 / 12884, WO 01 / 38322, WO 01 / 70675, WO 02 / 46144, WO 02 / 22577 and WO 02 / 30879. All HDAC inhibitors disclosed in these publications are included herein by reference.

[0059] Other HDAC inhibitors which can be included within the compositions of the present invention are cyclic peptide inhibitors, and here it can be referred e. g. to US-A 5,620, 953, US- A 5, 922, 837, WO 01 / 07042, WO 00 / 08048, WO 00 / 21979, WO 99 / 11659, WO 00 / 52033 and WO 02 / 0603. All HDAC inhibitors disclosed in these publications are included herein by reference. Suitable HD AC inhibitors are also those which are based on a benzamide structure which are disclosed e. g. in Proc. Natl. Acad. Sci. USA (1999), 96: 4592-4597, but also in EP- A 847 992, US 6, 174, 905, JP 11269140, JP 11335375, JP 11269146, EP 974 576, WO 01 / 38322, WO 01 / 70675 and WO 01 / 34131. All HD AC inhibitors, which are disclosed in these documents, are included herein by reference.

[0060] The HDAC inhibitors may be used under any pharmaceutically acceptable form, including without limitation, their free form and their pharmaceutically acceptable salts or solvates.

[0061] More particularly, examples of suitable HDAC inhibitors according to the invention include, but are not limited to, the compounds listed in Table 1 below.

[0062] Table 1: Examples of HDAC inhibitors useful within the context of the invention

[0063] In a preferred embodiment, the HD AC inhibitor is selected from the group consisting of Vorinostat (SAHA), belinostat (PXD-101), panobinostat (LBH-589), Dacinostat (LAQ824), CUDC-101, Abexinostat (PCI-24781), Ricolinostat (ACY-1215), Fimepinostat (CUDC-907), Resminostat, BRD73954 and Trichostatin A (TSA).

[0064] Examples of HDAC6 inhibitors include, but are not limited to, Tubastatin A HC1, tubacin, NN-390, ACY-241 (citarinostat) and A452.

[0065] Particularly preferred HD AC inhibitors of the invention are the HDAC6 inhibitors including, but not limited to, KT-531, KA2507, ACY-775, Tubastatin A TFA, ACY-738, CAY10603, Nexturastat A, HDAC6-IN-10, tubacin, NN-390, ACY-241 (citarinostat), A452 and preferably including MPT0G211 (Cat. No.: HY-123976. CAS No. : 2151853-97-1) and Tubastatin A (CAS No 1252003-15-8). Chemical structures of the two compounds are indicated below:

[0066] Further preferred HDAC6 inhibitor are the HDAC6 degraders, a class of PROTACs that comprises a selective HDAC6 inhibitor as the HDAC6 binder, a linker and a ligand for recruiting E3 ligase and induce significant degradation of HDAC6. Non limitative examples of HDAC6 degraders within the scope of the invention are HDAC6 degrader-1 (https: / / www.medchemexpress.com / hdac6-degrader- l.html), J22352 (https: / / www.medchemexpress.com / j22352.html) and compounds described in Yang K, Wu H, Zhang Z, Leisten ED, Nie X, Liu B, Wen Z, Zhang J, Cunningham MD, Tang W. Development of Selective Histone Deacetylase 6 (HDAC6) Degraders Recruiting Von Hippel-Lindau (VHL) E3 Ubiquitin Ligase. ACS Med Chem Lett. 2020 Mar 18;11(4): 575-581. doi: 10.1021 / acsmedchemlett.0c00046. PMID: 32292566; PMCID: PMC7153272.

[0067] The term "c-Myc" as used herein means the transcription factor encoded by the proto- oncogene c-Myc that controls cell proliferation. c-Myc also plays a role in regulating cell cycle, cell growth, angiogenesis, apoptosis, and oncogenesis. The c-Myc transcription factor is of the helix-loop-helix leucine zipper class and plays a role in the modulation and initiation of transcription. c-Myc binds to E- boxes (CACGTG) in the vicinity of target genes, which are then activated. The DNA binding activity requires dimerization with another hehx-loop-helix leucine zipper protein called Max. Max can also interact with transcriptional repressors such as Mad and Mxil, which presumably down-regulate expression of c-Myc target genes. c-Myc, when activated, can induce malignancy in a variety of tissues, most notably hematopoietic tissues (Leder et al., 222 Science 765, 1983). Myc's activity can increase in tumors as a consequence of mutations, chromosomal rearrangements, increased expression, or gene amplification. Elevated or deregulated expression of c-Myc has been detected in a wide range of human cancers and is often associated with aggressive, poorly differentiated tumors. Such cancers include colon, breast, cervical, small-cell lung carcinomas, osteosarcomas, glioblastomas, melanoma and myeloid leukemias.

[0068] Several different pharmaceutical-based strategies have recently been described for targeting MYC with the aim of blocking tumor growth (see Michael J. Duffy, Shane O’Grady, Minhong Tang, John Crown, Cancer Treatment Reviews https: / / doi.Org / 10.1016 / j.ctrv.2021.102154). Of these, the most investigated are inhibition of MYC-MAX interaction, prevention of MYC expression and targeting of genes exhibiting synthetic lethality with overexpression of MYC, c-Myc degraders / PROTACs.

[0069] Table 2 below summarizes currently investigated inhibitors of Myc:

[0070] As used herein, a c-Myc inhibitor can be a small molecule, a peptide, a nucleic acid, or a PROTAC (proteolysis targeting chimera), a degrader, a molecular glue for targeted protein degradation.

[0071] Preferably the c-Myc inhibitor is MYCi975 and / or WBC100.

[0072] In a preferred aspect of the invention, the inhibitor is an antibody or synthetic or recombinant derivative thereof.

[0073] Said antibody is preferably a monoclonal or polyclonal antibody, or synthetic or recombinant derivatives thereof, more preferably said antibody being a humanized monoclonal antibody.

[0074] Preferably, said polynucleotide is an RNA or DNA.

[0075] Preferably said polynucleotide able to inhibit the expression of the GTF2I or c-Myc or HD AC gene is an RNA inhibitor, more preferably it is a siRNA, a shRNA, a microRNA or an antisense oligonucleotide.

[0076] In another preferred embodiment, the siRNA comprises or targets or consists of at least one of the sequences selected from the group consisting of the SEQ ID NO: 15-18, or a functional derivative thereof.

[0077] In another preferred embodiment, the shRNA comprises or targets or consists of the SEQ ID NO: 21, or a functional derivative thereof.

[0078] In a preferred embodiment, the above vector is an expression vector selected from the group consisting of: plasmids, viral particles and phages.

[0079] Preferably, said host cell is selected from the group consisting of: bacterial cells, fungal cells, insect cells, animal cells, plant cells, preferably being an animal cell, more preferably a human cell.

[0080] In a preferred embodiment, the molecule for use in the treatment and / or prevention of DUX4-r positive leukemia according to the invention is used in combination with a further therapeutic agent. Said further therapeutic agent is a chemotherapy drug. Preferably said further therapeutic agent is a drug antagonizing the activity of inhibitors of apoptosis, preferably an inhibitor of cIAP and XIAP, more preferably Birinapant. Birinapant (TL32711), a bivalent Smac mimetic, is a potent antagonist for XIAP and cIAPl with Kds of 45 nM and less than 1 nM, respectively. Birinapant (TL32711) induces the autoubiquity lation and proteasomal degradation of cIAPl and cIAP2 in intact cells, which results in formation of a RIPK1 : caspase-8 complex, caspase-8 activation, and induction of tumor cell death. Birinapant (TL32711) targets TRAF2-associated cIAPs and abrogates TNF-mduced NF-ĸB activation. Further agents known for use in the treatment of include Venetoclax, a BCL2 antagonist approved for use in the treatment of CLL and AML, Obataclax, a pan-BCL2 inhibitor, Navitoclax,a BCL family inhibitor, and AMG176 MCL-1 inhibitor.

[0081] By the term "inhibitor" it is meant a molecule that effects a change in the expression and / or function of the target. The change is relative to the normal or baseline level of expression and / or function in the absence of the molecule, but otherwise under similar conditions, and it represent a decrease in the normal / baseline expression and / or function.

[0082] The terms "function" and "activity" are herein interchangeable.

[0083] The inhibition (which includes blocking) of the expression and / or function of the target may be assessed by any means known to the skilled in the art. The assessment of the expression level or of the presence of the target is preferably performed using classical molecular biology techniques such as (real time Polymerase Chain Reaction) qPCR, microarrays, bead arrays, RNAse protection analysis or Northern blot analysis or cloning and sequencing.

[0084] The assessment of target function (or activity) is e.g. performed by analysis of targets genes, chromatin association, chromatin modifications, chromatin accessibility, cell adhesion, cell migration, cell proliferation, cell death, leukemia engraftment in animal models, leukemia burden in animal models.

[0085] In the context of the present invention, the target may be the gene, the mRNA, the cDNA, or the encoded protein thereof.

[0086] The polynucleotides as above described, as e.g. the siRNAs, may further comprise dTdT or UU 3'- overhangs, and / or nucleotide and / or polynucleotide backbone modifications as described elsewhere herein.

[0087] In the method of the invention, preferably step a) comprises measuring the amount of the protein GTF2I, HD AC and / or c-Myc or of fragments thereof or of the polynucleotide coding for said proteins or of fragments thereof in said isolated biological sample obtained from the subject and step b) comprises comparing the measured amount of step a) with a proper control amount.

[0088] Preferably, the in vitro method as above defined preferably comprises the steps of: a) measuring the amount or the activity of the protein GTF2I, HD AC and / or c-Myc or of fragments thereof or of the polynucleotide coding for said protein or fragments thereof in said isolated biological sample obtained from the subject and b) comparing the measured amount or activity of step a) with a proper control amount or activity, wherein an amount or activity of said protein GTF2I, HD AC and / or c-Myc or of fragments thereof or of said polynucleotide or of fragments thereof in the isolated biological sample obtained from the subj ect higher than the control amount or activity indicates that the subject is either at increased risk for developing or is affected by the pathologies herein mentioned.

[0089] Preferably, the in vitro method for monitoring the progression and / or for monitoring the efficacy of a therapeutic treatment of one of the pathologies herein mentioned comprises the steps of: a) measuring the alteration of the amount or the alteration of the activity of the GTF2I, HD AC and / or c-Myc proteins or of fragments thereof or of the polynucleotide coding for said protein or fragments thereof in said isolated biological sample obtained from the subject and b) comparing the measured alteration of step a) with a proper control alteration.

[0090] Preferably, the biological sample is a fluid, a cell or a tissue sample, more preferably said sample is plasma or serum or blood or bone marrow.

[0091] Another object of the invention is a kit for carrying out the above methods, comprising

[0092] - means to measure the amount or the activity of the GTF2I, HDAC and / or c-Myc proteins or of fragments thereof and / or means to measure the amount of the polynucleotide coding for said protein or of fragments thereof and optionally,

[0093] - control means.

[0094] Control means can be used to compare the amount or the increase of amount of the compound as above defined to a proper control. The proper control may be obtained for example, with reference to known standard, either from a normal subject or from normal population. The means to measure the amount of at least one compound as above defined are preferably at least one antibody, functional analogous or derivatives thereof. Said antibody, functional analogous or derivatives thereof are specific for said compound.

[0095] In a preferred embodiment, the kit of the invention comprises:

[0096] - a solid phase adhered antibody specific for said compound;

[0097] - detection means of the ligand specific-biomarker complex. Other means may be e.g. specific primers and probes for RT PCR. The kits according to the invention can further comprise customary auxiliaries, such as buffers, carriers, markers, etc. and / or instructions for use.

[0098] In the case of a method or a kit for assessing the risk and / or diagnosing and / or prognosing of a disease herein mentioned, the proper control may be a sample taken from a healthy patient or from a patient affected by another disorder or pathology not characterized by DUX-4 rearrangement, and the proper control amount or activity may be the amount or activity of the same protein or polynucleotide measured in a sample taken from a healthy patient or from a patient affected by another disorder or pathology not characterized by DUX-4 rearrangement.

[0099] In the case of a method or a kit for monitoring the progression of the disease, the progress of the disease is monitored and the proper control may be a sample taken from the same subject at various times or from another patient, and the proper control amount or activity may by the amount or activity of the same protein or polynucleotide measured in a sample taken from the same subject at various times or from another patient.

[0100] In the case of a method or a kit for monitoring the efficacy of a therapeutic treatment, the proper control may by a sample taken from the same subject before initiation of the therapy or taken at various times during the course of the therapy and the proper control amount or activity may be the amount or activity of the same protein or polynucleotide measured in a sample taken from the same subject before initiation of the therapy or taken at various times during the course of the therapy. In the case of a method or a kit for the screening of a therapeutic treatment, the proper control may be a sample taken from subjects without treatment and from subjects treated with a substance that is to be assayed or from subjects treated with a reference treatment and the proper control amount or activity may be the average of the amounts or activity of the same protein or polynucleotide measured in samples taken from subjects without treatment and from subjects treated with a substance that is to be assayed or from subjects treated with a reference treatment. In this case, if the amount or activity of said protein or of said polynucleotide in the isolated biological sample obtained from the subject is lower or equal than the control amount or activity, it may indicate that the tested substance is effective for the treatment of the herein mentioned disease.

[0101] In the present invention, the expression "measuring the amount" can be intended as measuring the amount (or the activity) or concentration or level of the respective protein and / or mRNA thereof and / or DNA thereof, preferably semi-quantitative or quantitative. Measurement of a protein can be performed directly or indirectly. Direct measurement refers to the amount or concentration measure of the biomarker, based on a signal obtained directly from the protein, and which is directly correlated with the number of protein molecules present in the sample. This signal - which can also be referred to as intensity signal - can be obtained, for example, by measuring an intensity value of a chemical or physical property of the biomarker. Indirect measurements include the measurement obtained from a secondary component (e.g., a different component from the gene expression product) and a biological measurement system (e.g. the measurement of cellular responses, ligands, "tags" or enzymatic reaction products).

[0102] The term "amount", as used in the description refers but is not limited to the absolute or relative amount of proteins and / or mRNA thereof and / or DNA thereof, and any other value or parameter associated with the same or which may result from these. Such values or parameters comprise intensity values of the signal obtained from either physical or chemical properties of the protein, obtained by direct measurement, for example, intensity values in an immunoassay, mass spectroscopy or a nuclear magnetic resonance. Additionally, these values or parameters include those obtained by indirect measurement, for example, any of the measurement systems described herein. Methods of measuring mRNA and DNA in samples are known in the art. To measure nucleic acid levels, the cells in a test sample can be lysed, and the levels of mRNA in the lysates or in RNA purified or semi-purified from lysates can be measured by any variety of methods familiar to those in the art. Such methods include hybridization assays using detectably labeled DNA or RNA probes ( i.e., Northern blotting) or quantitative or semi-quantitative RT-PCR methodologies using appropriate oligonucleotide primers. Alternatively, quantitative or semi-quantitative in situ hybridization assays can be carried out using, for example, tissue sections, or unlysed cell suspensions, and detectably labeled (e.g., fluorescent, or enzyme-labeled) DNA or RNA probes. Additional methods for quantifying mRNA include RNA protection assay (RPA), cDNA and oligonucleotide microarrays, representation difference analysis (RD A), differential display, EST sequence analysis, and serial analysis of gene expression (SAGE).

[0103] If by comparing the measured amount or activity of the protein or of the polynucleotide coding for said protein with the amount or activity obtained from a control sample, the amount or the activity of said compound in the sample isolated from the subject corresponds to a higher value, the subject may present the disease or go towards an aggravation of said disease.

[0104] If by comparing the measured amount or activity of the protein or of the polynucleotide coding for said protein with the amount or the activity obtained from a control sample, the amount or the activity of said compound in the sample isolated from the subject corresponds to a similar or lower value, the subject may be not affected by the disease or go toward an amelioration of the disease, respectively. Alternatively, the expression "detection" or "measuring the amount" is intended as measuring the alteration of the molecule. Said alteration can reflect an increase or a decrease in the amount or activity of the molecules as above defined. An increase of the protein or of the activity of the protein or of the polynucleotide coding for said protein can be correlated to an aggravation of the disease.

[0105] A decrease of the protein or of the activity of the protein or of the polynucleotide coding for said protein can be correlated to an amelioration of the disease or to recovery of the subject.

[0106] The expression "protein GTF2I " or " GTF2I " is intended to include also the corresponding protein encoded from a GTF2I orthologous or homologous genes, functional mutants, functional derivatives, functional fragments or analogues, isoforms thereof.

[0107] The expression "gene GTF2I " or " GTF2I " is intended to include also the corresponding orthologous or homologous genes, functional mutants, functional derivatives, functional fragments or analogues, isoforms thereof.

[0108] The same applies also for c-Myc and HD AC.

[0109] The term "biological sample" encompasses a clinical sample, and also includes tissue obtained by surgical resection, tissue obtained by biopsy, cells in culture, cell supernatants, cell lysates, tissue samples, organs, bone marrow, blood, plasma, serum, and the like.

[0110] A "sample" in the context of the present teachings refers to any biological sample that is isolated from a subject. A sample can include, without limitation an aliquot of body fluid, whole blood, serum, plasma, solid tissue samples such as tissue biopsies, or tissue cultures or cells derived therefrom and the progeny thereof, synovial fluid, lymphatic fluid, ascites fluid, and interstitial or extracellular fluid. The term "sample" also encompasses the fluid in spaces between cells, including gingival crevicular fluid, bone marrow, cerebrospinal fluid (CSF), saliva, mucous, sputum, semen, sweat, urine, or any other bodily fluids. "Blood sample" can refer to whole blood or any fraction thereof, including serum and plasma. Samples can be obtained from a subject by means including but not limited to venipuncture, excretion, ejaculation, massage, biopsy, needle aspirate, lavage, scraping, surgical incision, or intervention or other means known in the art.

[0111] In the methods of the invention, the subject (or patient) may be an animal, preferably a mammalian, or a human. As used herein, the term " genetically engineered host cell " relates to host cells which have been transduced, transformed or transfected with the polynucleotide or with the vector described previously. As representative examples of appropriate host cells, one can cite bacterial cells, such as E. coli, Streptomyces, Salmonella typhimurium, fungal cells such as yeast, insect cells such as Sf9, animal cells such as CHO or COS, plant cells, etc. The selection of an appropriate host is deemed to be within the scope of those skilled in the art from the teachings herein. Preferably, said host cell is an animal cell, and most preferably a human cell.

[0112] Further assessment methods include in vitro suppression assay, whole transcriptome analysis, mass spectrometry analysis to identify proteins interacting with the target.

[0113] The polynucleotide according to the invention includes also chemically and structurally modified polynucleotide, preferably siRNAs, to solve their intrinsic problems related to in vivo instability, off- target effects, and immunogenicity.

[0114] The vector according to the invention also includes novel vector-driven bifunctional short hairpm RNA (bi-shRNA) (Ku SH„ et al. 2015; Mohr SE,et al. 2014).

[0115] The vector of the invention also includes an RNAi-inducing vector whose presence within a cell results in production of an siRNA or shRNA or rniRNA, preferably siRNA.

[0116] One or more siRNAs or shRNAs (in any combination) may be used.

[0117] Such polynucleotides can be single or double stranded. Preferably, one strand of a double-stranded polynucleotide comprises at least a partial sequence complementary to a target mRNA. The nucleotides of the inhibitory nucleic acid can be chemically modified, natural or artificial. The sequence homology between the polynucleotide (e.g. a RNAi inducing agent) and the target mRNA may be 100% or less, but is ideally greater than about 50% and typically 90% or greater and even more preferably at least 98% and 99%. It will be understood that the percentage of sequence homology between RNAi inducing agent and the target mRNA should be sufficient to result in sequence specific association between the RNAi inducing agent, e g. siRNA, and the target mRNA, preferably under cytoplasmic conditions.

[0118] Such siRNAs comprise two RNA strands having a region of complementarity of approximately 20 or so nucleotides in length and optionally further comprises one or two single-stranded overhangs or loops. In mammalian cells, dsRNA longer than 30 base pairs can cause non-specific gene suppression by an interferon a response. However, cells transfected with 21 nucleotide synthetic double-stranded siRNA bearing two nucleotides protruding at both 3 '-ends have been found to escape an interferon response and effectively exert sequence-specific gene silencing function. The silencing effect of the synthetic siRNA, however, is transient. The double stranded siRNA molecule down regulates expression of the GTF2I or HDAC or c-Myc protein via RNAi, wherein each strand of said siRNA molecule is independently about 18 to about 28 nucleotides in length and one strand of the siRNA molecule comprises a nucleotide sequence having sufficient complementarity to the RNA of the target protein for the siRNA molecule to direct cleavage of the target RNA via RNA interference. In shRNA, the single RNA strand may form a hairpin structure with a stem and loop and, optionally, one or more unpaired portions at the 5' and / or 3' portion of the RNA.

[0119] The delivery of an effective amount of the polynucleotide according to the invention, may be carried out with any know method which is suitable for the delivery. For example, some delivery agents for the siRNA, microRNA, etc. are selected from the following non- limiting group of cationic polymers, modified cationic polymers, peptide molecular transporters, lipids, liposomes and / or non-cationic polymers. Viral vector delivery systems may also be used. For example, an alternative delivery route includes the direct delivery of the polynucleotide (including siRNA, shRNA and miRNA) and even anti-sense RNA (asRNA) in gene constructs followed by the transformation of cells with the resulting recombinant DNA molecules. This results in the transcription of the gene constructs encoding the polynucleotide, such as siRNA, shRNA and miRNA, or even asRNA and provides for the transient and stable expression of the polynucleotide in cells and organisms. For example, such an alternative delivery route may involve the use of a lentiviral vector comprising a nucleotide sequence encoding a shRNA which targets the GTF2I or HD AC or c-myc. Such a lentiviral vector may be comprised within a viral particle. Adeno-associated viruses (AAV) may also be used and the use of these as delivery vehicles is expanded on later. siRNA delivery formulations include e.g. siRNA conjugates, polymerized siRNA, and nucleic acid-based nanoparticles to improve in vivo delivery.

[0120] RNA interference (RNAi) is the process by which expression of a target gene is effectively silenced or knocked down by the selective inactivation of its corresponding mRNA by double-stranded RNA (dsRNA). RNAi is activated by dsRNA species delivered to the cytoplasm of cells. The silencing mechanisms can either lead to the degradation of a target mRNA, as induced by small interfering RNAs (siRNAs) or short hairpin RNAs (shRNAs), or the suppression of translation of specific mRNAs, as induced by microRNA (miRNA). Two key approaches to RNAi that have gained substantial interest for use in gene silencing are the double-stranded small interfering RNAs (siRNAs) and the vector-based short hairpin RNAs (shRNAs).

[0121] Both these methods can achieve robust and specific knockdown, but they have striking mechanistic differences with broad practical implications. shRNA can effectively target knockdown with low copy numbers and longer lasting effects than siRNA. The recently introduced bifunctional design has similar benefits to standard shRNA but with greatly enhanced potency.

[0122] The use of shRNA is suggested by the advantages of shRNA over siRNA including the ability to use viral vectors for delivery to overcome the difficulty of transfecting certain cell types, the option to control shRNA expression using inducible promoters, and the ability to co-express them with a reporter gene. Additionally, they may cause fewer off-target effects.

[0123] Furthermore, novel vector-driven bifunctional short hairpin RNA (bi-shRNA) technology that harnesses both cleavage-dependent and cleavage-independent RISC loading pathways to enhance knockdown potency may be used. Indeed, consequent advantages provided by the bi-shRNA include a lower effective systemic dose than comparator siRNA / shRNA to minimize the potential for off-target side effects, due to its ability to induce both a rapid (inhibition of protein translation) and delayed (mRNA cleavage and degradation) targeting effect depending on protein and mRNA kinetics, and a longer duration of effectiveness for clinical applications.

[0124] The RNAi inhibitors as above defined are preferably capable of hybridizing to all or part of specific target sequence of GTF2I or of HD AC or of c-myc. Therefore, RNAi inhibitors may be fully or partly complementary to all of or part of the target sequence

[0125] The RNAi inhibitors may hybridize to the specified target sequence under conditions of medium to high stringency.

[0126] RNAi inhibitors may be defined with reference to a specific sequence identity to the reverse complement of the sequence to which it is intended to target. The antisense sequences will typically have at least about 75%, preferably at least about 80%, at least about 85%, at least about 90%, at least about 95% or at least about 99% sequence identity with the reverse complements of then-target sequences. The antisense nucleic acid may be Gapmers.

[0127] A "derivative" may be a nucleic acid molecule, as a DNA molecule, coding the polynucleotide as above defined, or a nucleic acid molecule comprising the polynucleotide as above defined, or a polynucleotide of complementary sequence. In the context of the present invention the term "derivatives" also refers to longer or shorter polynucleotides and / or polynucleotides having e.g. a percentage of identity of at least 41 % , 50 %, 60 %, 65 %, 70 % or 75%, more preferably of at least 85%, as an example of at least 90%, and even more preferably of at least 95% with the sequences herein disclosed or with their complementary sequence or with their DNA or RNA corresponding sequence. The term "derivatives" and the term "polynucleotide" also include modified synthetic oligonucleotides. The modified synthetic oligonucleotides are preferably LNA (Locked Nucleic Acid), phosphoro-thiolated oligos or methylated oligos, morpholinos, 2'-0-methyl, 2 -0-methoxyethyl oligonucleotides and cholesterol-conjugated 2'-0- methyl modified oligonucleotides (antagomirs).

[0128] The term "derivative" may also include nucleotide analogues, i.e. a naturally occurring ribonucleotide or deoxyribonucleotide substituted by a non-naturally occurring nucleotide. The modified nucleotide analog may be located for example at the 5'-end and / or the 3 '-end of the nucleic acid molecule. Nucleotide analogs may be selected from sugar- or backbone-modified ribonucleotides or nucleobase- modified ribonucleotides, i.e. ribonucleotides, containing a non-naturally occurring nucleobase instead of a naturally occurring nucleobase.

[0129] The term "derivatives" also includes nucleic acids that may be generated by mutating one or nucleotide in their sequences, equivalents or precursor sequences. The term "derivatives" also includes at least one functional fragment of the polynucleotide.

[0130] In the context of the present invention "functional" is intended for example as "maintaining their activity".

[0131] As used herein "fragments" refers to polynucleotides having preferably a length of at least 1000 nucleotides, 1100 nucleotide, 1200 nucleotides, 1300 nucleotides, 1400 nucleotides, 1500 nucleotides. Where the fragments are referred to the siRNA as above defined, the length is preferably of 10, 11, 12, 13, 14, 15, 16, 17, 18,19, 20, 21 nucleotides.

[0132] The term "polynucleotide" also refers to modified polynucleotides.

[0133] According to the present invention, an "effective amount" of a composition is one that is sufficient to achieve a desired biological effect, in this case an amelioration or the treatment of disease herein mentioned.

[0134] It is understood that the effective dosage will be dependent upon the age, sex, health, and weight of the recipient, kind of concurrent treatment, if any, frequency of treatment, and the nature of the effect desired. The provided ranges of effective doses of the inhibitor or molecule of the invention (e g. from 1 mg / kg to 100 mg / kg, in particular systemically administered) are not intended to limit the invention and represent preferred dose ranges. However, the preferred dosage can be tailored to the individual subject, as is understood and determinable by one of skill in the art, without undue experimentation. The administration of oligonucleotides of the present invention may be carried out by known methods, wherein a nucleic acid is introduced into a desired target cell in vitro or in vivo.

[0135] An aspect of the present invention comprises a nucleic acid construct comprised within a delivery vehicle. A delivery vehicle is an entity whereby a nucleotide sequence can be transported from at least one media to another. Delivery vehicles may be generally used for expression of the sequences encoded within the nucleic acid construct and / or for the intracellular delivery of the construct. It is within the scope of the present invention that the delivery vehicle may be a vehicle selected from the group of RNA based vehicles, DNA based vehicles / vectors, lipid-based vehicles, virally based vehicles and cell- based vehicles. Examples of such delivery vehicles include: biodegradable polymer microspheres, lipid- based formulations such as liposome carriers, coating the construct onto colloidal gold particles, lipopolysaccharides, polypeptides, polysaccharides, pegylation of viral vehicles.

[0136] In one embodiment of the present invention may comprise a virus as a delivery vehicle, where the virus may be selected from: adenoviruses, retroviruses, lentiviruses, adeno-associated viruses, herpesviruses, vaccinia viruses, foamy viruses, cytomegaloviruses, Semliki forest virus, poxviruses, RNA virus vector and DNA virus vector. Such viral vectors are well known in the art.

[0137] Commonly used gene transfer techniques include calcium phosphate, DEAE-dextran, transfection, electroporation and micromjection and viral methods. Another technique for the introduction of DNA into cells is the use of cationic liposomes. Commercially available cationic lipid formulations are e.g. Tfx 50 (Promega) or Lipofectamin 2000 (Life Technologies).

[0138] The above pharmaceutical compositions are preferably for systemic, oral, locally, preferably rectally, or topical administration.

[0139] The compositions of the present invention may be in form of a solution, e.g. an injectable solution, a cream, ointment, tablet, suspension or the like. The composition may be administered in any suitable way, e.g. by injection, particularly by intraocular injection, by oral, topical, nasal, rectal application etc. The carrier may be any suitable pharmaceutical carrier. Preferably, a carrier is used, which is capable of increasing the efficacy of the RNA molecules to enter the target-cells. Suitable examples of such carriers are liposomes, particularly cationic liposomes.

[0140] The recombinant expression vector of the invention can be any suitable recombinant expression vector, and can be used to transform or transfect any suitable host. Suitable vectors include those designed for propagation and expansion or for expression or both, such as plasmids and viruses. The recombinant expression vectors of the invention can be prepared using standard recombinant DNA techniques. Constructs of expression vectors, which are circular or linear, can be prepared to contain a replication system functional in a prokaryotic or eukaryotic host cell. Replication systems can be derived, e.g., from CoIEl, 2 μ plasmid, λ, SV40, bovine papilloma virus, and the like.

[0141] Desirably, the recombinant expression vector comprises regulatory sequences, such as transcription and translation initiation and termination codons, which are specific to the type of host (e.g., bacterium, fungus, plant, or animal) into which the vector is to be introduced, as appropriate and taking into consideration whether the vector is DNA- or RNA- based. The recombinant expression vector can include one or more marker genes, which allow for selection of transformed or transfected hosts. Marker genes include biocide resistance, e.g., resistance to antibiotics, heavy metals, etc., complementation in an auxotrophic host to provide prototrophy, and the like. Suitable marker genes for the inventive expression vectors include, for instance, neomycin / G418 resistance genes, hygromycin resistance genes, histidinol resistance genes, tetracycline resistance genes, and ampicillin resistance genes. The recombinant expression vector can comprise a native or normative promoter operably linked to the nucleotide sequence encoding the GTF2I or HD AC or c-Myc inhibitor (including functional portions and functional variants thereof), or to the nucleotide sequence which is complementary to or which hybridizes to the nucleotide sequence encoding the RNA. The selection of promoters, e.g., strong, weak, inducible, tissue- specific and developmental-specific, is within the ordinary skill of the artisan. Similarly, the combining of a nucleotide sequence with a promoter is also within the skill of the artisan. The promoter can be a non- viral promoter or a viral promoter, e.g., a cytomegalovirus (CMV) promoter, an SV40 promoter, an RSV promoter and a promoter found in the long-terminal repeat of the murine stem cell virus.

[0142] The B-ALL cells positive for DUX4-r express CD371 (also known as CLL1 or CLEC12A) as a surface protein (https: / / www.ncbi.nlm.nih.gov / pmc / articles / PMC6669149 / ). Therefore, the molecule of the invention may be advantageously fused or conjugated to ligand specific for CD371, e.g. anti-CD371 antibodies, to deliver the molecules selectively to the cells which express DUX4-r, so avoiding or reducing adverse effects.

[0143] Advantageously, the molecule of the invention may also be included or fused or conjugated in nanoparticles, e.g. lipid nanoparticles, presenting CD371 ligands, e.g. antibodies, on their surface.

[0144] Preferably, the virus expressing the molecule of the invention, e.g. the shRNA herein disclosed, may expose CD371 ligands, e.g. antibodies, on its surface. The virus may preferably be a lentivirus.

[0145] The inventive recombinant expression vectors can be designed for either transient expression, for stable expression, or for both. Also, the recombinant expression vectors can be made for constitutive expression or for inducible expression.

[0146] In the above compositions further materials as well as processing techniques and the like may be set out in Part 5 of Remington's Pharmaceutical Sciences, 20th Edition, 2000, Marek Publishing Company, Easton, Pennsylvania, which is incorporated herein by reference.

[0147] The compounds of this invention can also be administered in sustained release forms or from sustained release drug delivery systems. A description of representative sustained release materials can also be found in the incorporated materials in Remington's Pharmaceutical Sciences. Furthermore, pharmaceutical formulations can be prepared using a process, which is generally known in the pharmaceutical art. In the present invention, when the molecule of the invention is administered with another therapeutic agent, it may be administered simultaneously or sequentially.

[0148] In the methods of the invention, the subject (or patient) may be an animal, preferably a mammalian, or a human.

[0149] The present invention will be described by means of non-limiting examples, referring to the following figures:

[0150] Figure 1. DUX4 and DUX4-r associate to and activate different gene sets.

[0151] A. Heatmap showing the 500 most variable genes identified in REH cells upon EV, wt DUX4, DUX4-IGH or DUX4-del50 expression. Expression levels are in RPKM. The heatmap reports Z scores scaled by row.

[0152] B. Venn diagram displaying: Left: The overlap between genes upregulated in the present study (RNA-seq data) and in DUX4-r B-ALL patients (from EGAS00001000654 dataset). Right: The overlap between genes downregulated in the present study (RNA-seq data) and in DUX4-r B-ALL patients (from EGAS00001000654 dataset). Two-tailed Fisher’s exact test was performed to test the significance of the gene intersection.

[0153] C. Dot plot displaying the functional families enriched in the overlap between: Left: genes positively regulated in DUX4-r B-ALL patients and by DUX4-r in REH cells. Right: Genes negatively regulated in DUX4-r B-ALL patients and by DUX4-r in REH cells. The dot size is proportional to the percentage of modulated genes enriched in each pathway.

[0154] D. Heatmap showing the 500 hundred most variable repetitive elements (REs) identified in REH cells upon EV, wt DUX4, DUX4-IGH or DUX4-del50 expression. Expression levels are in RPKM. The heatmap reports Z scores scaled by row.

[0155] E. Top: Venn diagram displaying the overlap between CUT&Tag peaks identified in wt DUX4, DUX4-IGH or DUX4-del50 expressing REH cells. Bottom: CUT&Tag profile plots of wt DUX4 (red) and DUX4-IGH (cyan) peaks on wt DUX4 (left) or DUX4-r (right) direct transcriptional targets.

[0156] F. De novo motif calling analyses of peaks selectively associated to wt DUX4 (DUX4 only) or DUX4-IGH and DUX4-del50 (DUX4-IGH / DUX4-del50 overlap).

[0157] G. CUT&Tag profile plot of H3K27Ac enrichment at wt DUX4 and DUX4-IGH peaks on respective direct transcriptional targets in the absence (left) or presence (right) of Dox. Wilcoxon Ranks- sum test was performed to compare the signal of H3K27Ac signal between -Dox / +Dox. H. CUTAC profile plot at wt DUX4 and DUX4-IGH peaks on respective direct transcriptional targets in the absence (left) or presence (right) of Dox. Wilcoxon Ranks-sum test was performed to compare the CUTAC signal between -Dox / +Dox conditions.

[0158] Figure 2. DUX4-r activates cell adhesion and migration, promoting cell proliferation.

[0159] A. Schematics of the HS5 bone marrow stroma adhesion (top) and transwell migration assays (bottom).

[0160] B. Quantification by FACS analysis of adhesion assay performed in GFP+ REH cells expressing EV, wt DUX4 or DUX4-r variants in -Dox / +Dox conditions. Bar plots represent the average of three biological replicates (n=3). Error bars represent the mean + / -SD. Student’s T-test ***p≤0,001; ****p≤0,0001.

[0161] C. Quantification by FACS analysis of migration assay performed in GFP+ REH cells expressing EV, wt DUX4 or DUX4-r variants in -Dox / +Dox conditions. Bar plots represent the average of three independent biological replicates (n=3). Error bars represent the mean + / -SD. Student’s T-test **p≤0,01; ****p≤0,0001.

[0162] D. Western blot validation of DUX4-r knockdown in NALM6 cells using Vinculin as loading control.

[0163] E. Quantification by FACS analysis of adhesion assay (top) and migration assay (bottom) performed in GFP+ NALM6 cells in control (shCTRL) or DUX4-r (shDUX4-r) knockdown conditions. Bar plots represent the average of three independent biological replicates (n=3). Error bars represent the mean + / -SD. Student’s T-test **p≤0,01; ****p≤0,0001.

[0164] F. Representative phase microscopy images of tumor spheroids generated in liquid cell culture specifically by DUX4-r expressing REH cells. Images were acquired at the 45h time point. Scalebar: 300 μm.

[0165] G. Real-time quantification of tumor spheroids, formed by NALM6 cells in control (shCTRL) or DUX4-r (shDUX4-r) knockdown conditions, by automated Incucyte Live-Cell analyses, performed by computing the phase object average as normalized to TO. Curves represent the average of six biological replicates (n=6). Error bars represent mean + / - SD. Two-Way analysis of variance (ANOVA) with Bonferrom’s correction to compare curves at the 45h time point. **** p≤0,0001.

[0166] H. Proliferation assays in low serum conditions of REH cells expressing EV, wt DUX4, DUX4- IGH or DUX4-del50. Curves represent the average of three independent biological replicates (n=3). Error bars represent mean + / - SD. One-Way ANOVA. **p≤0,01; ***p≤0,001. I. Proliferation assay of NALM6 cells in control (shCTRL) or DUX4-r (shDUX4-r) knockdown conditions. Curves represent the average of three independent biological replicates (n=3). Error bars represent mean + / - SD. One-Way ANOVA. *p≤0,05.

[0167] Figure 3. DUX4-r activity is prominent in B-ALL cells.

[0168] A. RT-qPCR expression analysis of the indicated DUX4-r (top) or wt DUX4 (bottom, right) direct target genes in REH cells expressing EV, wt DUX4, DUX4-IGH or DUX4-del50. Relative levels of wt DUX4 / DUX4-r are also shown (bottom, left). Relative mRNA levels were normalized to GAPDH. Bar plots represent the average of three independent biological replicates (n=3). Error bars represent mean + / -SD. One-way ANOVA with multiple comparisons. *p≤0,05; **p≤0,01; ***p≤0,001; ****p≤0,0001.

[0169] B. RT-qPCR analysis of the same genes as in Fig. 3A in HEK cells expressing EV, wt DUX4, DUX4-IGH or DUX4-del50. Relative mRNA levels were normalized to GAPDH. Bar plots represent the average of three independent biological replicates (n=3). Error bars represent mean + / -SD. One-way ANOVA with multiple comparisons. ***p≤0,001; ****p≤0,0001.

[0170] C. RT-qPCR analysis of the same genes as in Fig. 3A-B in Jurkat cells expressing EV, wt DUX4, DUX4-IGH or DUX4-del50. Relative mRNA levels were normalized to GAPDH. Bar plots represent the average of three independent biological replicates (n=3). Error bars represent mean + / -SD. One-way ANOVA with multiple comparisons. ***p≤0,001; ****p≤0,0001.

[0171] D. CUT&Tag profile plot of genomic enrichment at: Left: wt DUX4 on its direct transcriptional targets in HEK (dark red) and REH (red) cells; Right: DUX4-IGH on its direct transcriptional targets in HEK (blue) and REH (teal) cells. Wilcoxon Ranks-sum test was performed to compare the signal of wt DUX4 or DUX4-IGH between HEK and REH cells.

[0172] Figure 4. DUX4-r selectively interacts with GTF2I.

[0173] A. Heatmap showing DUX4-r selective interactors which are more expressed in REH cells. Protein levels were expressed as MS spectral counts. The heatmap reports unsealed MS / MS counts.

[0174] B. Tandem affinity purification followed by immunoblotting validation of selective GTF2I interaction with DUX4-IGH and DUX4-del50. 5% of input and 1 st elution and 10% of 2nd elution were loaded and incubated with antibodies specific for GTF2I or wt DUX4 and DUX4-r.

[0175] C. Western blot validation of nuclear protein levels of GTF2I in REH, Jurkat or REH cells using GAPDH as loading control.

[0176] D. Densitometric analysis of signals in Fig. 3D using the ImageLab software (Bio-Rad, ver. 6.1). Bar plots represent the mean of three independent biological replicates (n=3). Error bars represent + / - SD. One-way ANOVA with multiple comparisons against REH. *p≤0,05; ***p≤0,001 E. CUT&Tag profile plot of GTF2I enrichment on DUX4-r direct targets in HEK EV (red), REH EV (blue) or REH DUX4-IGH (teal) cells. Wilcoxon Ranks-sum test was performed to compare the signal of GTF2I between HEK EV, REH EV and REH DUX4-IGH cells.

[0177] F. Heatmap showing the expression levels (FPKM) of DUX4-r core genes in different DUX4-r B- ALL patients ordered based on decreasing GTF2I expression levels. The heatmap reports Z scores scaled by row.

[0178] G. Pearson’s correlation plot showing the positive correlation between GTF2I expression levels and the mean activation of DUX4-r core genes in different DUX4-r B-ALL patients. The correlation coefficient (R) and P-value are indicated.

[0179] Figure 5. GTF2I is required for DUX4-r biological activity in vitro and in vivo.

[0180] A. RT-qPCR (left) and western blot (right) validation of GTF2I knockdown in REH cells expressing ectopic DUX4-IGH (top) or NALM6 cells expressing endogenous DUX4-r (bottom). Relative mRNA levels were normalized to GAPDH. Bar plots represent the mean of three independent biological replicates (n=3). Error bars represent + / - SD. Student’s T-test *p≤0,05; ****p≤0,0001.

[0181] B. RT-qPCR expression analysis in REH DUX4-IGH (top) and NALM6 (bottom) cells of the indicated direct DUX4-r targets in control (shCTRL) or GTF2I (shGTF2I) knockdown conditions. Relative mRNA levels were normalized to GAPDH. Bar plots represent the average of three independent biological replicates (n=3). Error bars represent mean + / -SD. One-way ANOVA with multiple comparisons. **p≤0,01; ***p≤0,001; ****p≤0,0001.

[0182] C. Quantification by FACS analysis of adhesion (top) and migration (bottom) assays performed in GFP+ REH DUX4-IGH (left) and NALM6 (right) cells in shCTRL or shGTF2I conditions. Bar plots represent the average of three independent biological replicates (n=3). Error bars represent the mean + / - SD. Student’s T-test. **p≤0,01; ***p≤0,001; ****p≤0,0001.

[0183] D. Real-time quantification of tumor spheroid formation by REH DUX4-IGH plus control (shCTRL) or GTF2I (shGTF2I) knockdown performed by computing the phase object average as normalized to TO. Curves represent the average of six biological replicates (n=6). Error bars represent mean + / - SD. Two-Way ANOVA with Bonferrom’s correction to compare curves at the 45h time point. **** p≤0,0001.

[0184] E. Proliferation assays in low serum conditions of REH cells expressing DUX4-IGH in shCTRL or shGTF2I knockdown conditions. Curves represent the average of three independent biological replicates (n=3). Error bars represent the mean + / -SD. One-Way ANOVA. *p≤0,05. F. Apoptosis assays in DUX4-IGH expressing REH cells (left) and NALM6 cells (right) in shCTRL or shGTF2I knockdown conditions. Bar plots represent the average of four independent biological replicates (n=4). Error bars represent the mean + / -SD. Student’s T-test ***p≤0,001.

[0185] G. RT-qPCR expression analysis for GTF2I as well as the indicated DUX4-r direct targets in HEK cells expressing EV, wt DUX4 or DUX4-IGH plus EV or GTF2I. Relative mRNA levels were normalized to GAPDH. Bar plots represent the average of three independent biological replicates (n=3). Error bars represent mean + / -SD. One-way ANOVA with multiple comparisons was performed. **p≤0,01; ***p≤0,001.

[0186] H. CUT&Tag profile plot of DUX4-IGH enrichment on its direct targets in HEK DUX4-IGH cells transfected with EV (blue) or GTF2I (teal). Wilcoxon Ranks-sum test was performed to compare the signal of DUX4-IGH between EV and GTF2I conditions.

[0187] I. Bioluminescence detection of leukemia expansion in peripheral blood of NSG mice 7- and 14- days post NALM6-Lucia cell transplantation. Bar plots show the mean Luciferase signal + / - SD of shCTRL (n=5) or shGTF2I (n=5) NALM6-Lucia mice. Two-way ANOVA. **p≤0,01.

[0188] J. Kaplan-Meyer survival analysis of the above mice. *p≤0,05.

[0189] Figure 6. GTF2I pharmacological targeting inhibits DUX4-r leukemogenic activity.

[0190] A. RT-qPCR expression analysis of wt DUX4, DUX4-IGH, GTF2I and the indicated DUX4-IGH direct targets upon SAHA treatment in EV, wt DUX4 or DUX4-IGH REH cells. Relative mRNA levels were normalized to GAPDH. Bar plots represent the average of three independent biological replicates (n=3). Error bars represent mean + / -SD. One-way ANOVA with multiple comparisons. **p≤0,01; ***p≤0,001; ****p≤0,0001.

[0191] B. CUT&Tag profile plot of DUX4-IGH enrichment on its direct targets in REH DUX4-IGH cells treated with DMSO (teal) or SAHA (blue) for 24h. Wilcoxon Ranks-sum test was performed to compare the signal of DUX4-IGH between DMSO and SAHA conditions.

[0192] C. Quantification by FACS analysis of adhesion (left) and migration (right) assays performed in GFP+ REH DUX4-IGH (teal) or NALM6 (blue) cells treated with DMSO or SAHA. Bar plots represent the average of three independent biological replicates (n=3). Error bars represent the mean + / -SD. Student’s T-test. *p≤0,05; **p≤0,01; ****p≤0,0001.

[0193] D. Real-time quantification of tumor spheroid formation in REH DUX4-IGH cells after treatment with DMSO or SAHA. Quantification was performed by computing the phase object average as normalized to TO. Curves represent the average of six biological replicates (n=6). Error bars represent mean + / -SD. Two-Way ANOVA with Bonferroni’ s correction to compare curves at the 45h time point. **** p≤0,0001.

[0194] E. Proliferation assays in low serum conditions of control REH cells expressing EV (left) or DUX4- IGH (right) treated with DMSO or SAHA. Curves represent the average of three biological replicates (n=3). Error bars represent mean + / -SD. One-Way ANOVA. *p≤0,05.

[0195] F. Apoptosis assays in REH cells expressing EV or DUX4-IGH, and NALM6 cells after treatment with DMSO or SAHA. Bar plots represent the average of three independent biological replicates (n=3). Error bars represent the mean + / -SD. Student’s T-test. ****p≤0,0001.

[0196] G. Bioluminescence detection of leukemia expansion in peripheral blood of NSG mice at 7 days and 14 days post NALM6-Lucia transplantation. Bar plots show the average signal of DMSO- or SAHA-treated mice. Error bars represent the mean + / - SD Luciferase signal (n=8). Two-Way ANOVA. **p≤0,01.

[0197] H. Kaplan-Meyer survival analysis of the above mice. *p≤0,05.

[0198] Figure 7. DUX4 and DUX4-r associate to and activate different gene sets.

[0199] A. RT-qPCR expression analysis of wt DUX4, DUX4-IGH and DUX4-del50 upon 12h doxycycline induction in REH cells. Relative mRNA levels were normalized to GAPDH. Bar plots represent the average of three independent biological replicates (n=3). Error bars represent mean + / -SD. One-way ANOVA with multiple comparisons against EV. *p≤0.05; ****p≤0.0001 .

[0200] B. Western blot analysis of wt DUX4, DUX4-IGH and DUX4-del50 protein levels upon 12h of doxycycline induction in REH cells, using Vinculin as loading control.

[0201] C. Apoptosis assays in wt DUX4, DUX4-IGH or DUX4-del50 expressing REH cells upon 18h of doxycycline induction. Bar plots represent the average of four independent biological replicates (n=4). Error bars represent mean + / -SD. Student’s T-test. ****p≤0.0001. (RLU: relative luminescence unit)

[0202] D. Principal component analysis of the transcriptome of REH cells expressing EV, wt DUX4, DUX4-IGH or DUX4-del50.

[0203] E. Venn diagram displaying the overlap between genes upregulated by wt DUX4, DUX4-IGH or DUX4-del50 expressing REH cells as compared to control REH EV cells.

[0204] F. Venn diagram displaying: Left: The overlap between genes upregulated in the present study (RNA-seq data) and bona-fide wt DUX4 upregulated genes (from GSE85461 dataset); Right: the overlap between genes downregulated in the present study (RNA-seq data) and bona-fide wt DUX4 downregulated genes (from GSE85461 dataset). Two-tailed Fisher’s exact test was performed to test gene intersection. G. Dot plots displaying the functional families enriched in the overlap between: Left: genes positively regulated by wt DUX4 in REH cells and bona-fide wt DUX4 upregulated target genes (from GSE85461 dataset); Right: genes negatively regulated by wt DUX4 in REH cells and bona-fide wt DUX4 downregulated genes (from GSE85461 dataset).

[0205] H. Top: Venn diagram displaying the overlap between expressed genes showing >1 RPKM in REH EV control cells and DUX4-IGH direct target genes. Bottom: Venn diagram displaying the overlap between expressed genes showing >1 RPKM in REH EV control cells and wt DUX4 direct target genes. Fig. 8. DUX4-r promotes tumor spheroid formation.

[0206] A. Representative phase microscopy images of tumor spheroids generated in liquid cell culture specifically by DUX4-r expressing REH cells. Images were acquired at the 45h time point. Scalebar: 300 pm.

[0207] B. Representative phase microscopy images of reduced formation of tumor spheroids by NALM6 cells following knockdown of DUX4-r. Images were acquired at the 45h time point. Scalebar: 300 pm. Figure 9. DUX4-r selectively interacts with GTF2I.

[0208] A. Heatmap showing the proteins specifically interacting with wt DUX4 and or DUX4-r identified in our TAP-MS approach. The heatmap reports Z scores scaled by row.

[0209] B. Heatmap showing DUX4-r selective interactors which are more expressed in REH cells. Protein levels were expressed as MS spectral counts. The heatmap reports Z scores scaled by row.

[0210] C. Protein-protein ranked interaction networks generated using the SAINT package, by performing the empirical fold change score and saint (SP) score as compared to the EV control to compute the enrichment of wt DUX4, DUX4-IGH or DUX4-DEL50 with respect to the background. Selective DUX4, DUX4-IGH or DUX4-del50 interactors having a Fold Change (FC) score >1,5 over EV and probability score (SP) >0.9 are shown.

[0211] D. Endogenous interaction between DUX4-r and of GTF2I in NALM6 cells. Immunoblottmg of Input or either DUX4-r (IP) or control IgG immunoprecipitates analyzed with antibodies specific for GTF2I or DUX4-r

[0212] E. Genome-wide CUT&Tag profile plot and heatmap of GTF2I enrichment at TSS of genes in REH EV (left) or REH DUX4-IGH (right) cells.

[0213] Figure 10. GTF2I is required for DUX4-r biological activity in vitro and in vivo.

[0214] A. RT-qPCR expression analysis of the indicated wt DUX4 direct transcriptional targets in REH wt DUX4 cells in shCTRL or shGTF2I conditions. Relative mRNA levels were normalized to GAPDH. Bar plots represent the average of three independent biological replicates (n=3). Error bars represent mean + / -SD. One-way ANOVA with multiple comparisons.

[0215] B. Proliferation assays in low serum conditions in NALM6 cells in shCTRL or shGTF2I conditions. Curves represent the average of three independent biological replicates (n=3). Error bars represent mean + / -SD. One-Way ANOVA. *p≤0.05.

[0216] C. Representative phase microscopy images of the decreased tumor spheroids formation of DUX4- IGH expressing REH cells in shGTF2I as compared to shCTRL condition. Images were acquired at the 45h time point. Scale bar: 300pm.

[0217] D. Real-time quantification of tumor spheroid formation by NALM6 plus control (shCTRL) or GTF2I (shGTF2I) knockdown performed by computing the phase object average as normalized to TO. Curves represent the average of six biological replicates (n=6). Error bars represent mean + / - SD. Two- Way ANOVA with Bonferroni’s correction to compare curves at the 45h time point. **** p≤0,0001.

[0218] E. Real-time quantification of tumor spheroid formation by wt DUX4 REH plus control (shCTRL) or GTF2I (shGTF2I) knockdown performed by computing the phase object average as normalized to TO. Curves represent the average of six biological replicates (n=6). Error bars represent mean + / - SD. Two-Way ANOVA with Bonferroni’s correction to compare curves at the 45h time point.

[0219] F. Proliferation assays in low serum conditions in NALM6 cells in shCTRL or shGTF2I conditions. Curves represent the average of three independent biological replicates (n=3). Error bars represent mean + / -SD. One-Way ANOVA. *p≤0.05.

[0220] G. GTF2I protein levels analyzed by immunoblotting in HEK cells transfected with EV or GTF2I as compared to endogenous GTF2I in REH cells using GAPDH as loading control.

[0221] H. RT-qPCR expression analysis for GTF2I as well as the indicated wt DUX4 direct targets in HEK cells expressing EV, wt DUX4 or DUX4-IGH plus EV or GTF2I. Relative mRNA levels were normalized to GAPDH levels. Bar plots represent the average of three independent biological replicates (n=3). Error bars represent mean + / -SD. One-way ANOVA with multiple comparisons.

[0222] I. Left: Genome-wide CUT&Tag profile plot and heatmap of enrichment at TSS of genes of IgG or DUX4-IGH in HEK + EV, and DUX4-IGH in HEK + GTF2I conditions. Right: Boxplot representing the differential enrichment of normalized counts in IgG or DUX4-IGH in HEK + EV, and DUX4-IGH in HEK + GTF2I samples.

[0223] J. CUT&Tag profile plot of wt DUX4 enrichment on its direct targets in HEK wt DUX4 cells transfected with EV (blue) or GTF2I (teal). Wilcoxon Ranks-sum test was performed to compare the signal of wt DUX4 between EV and GTF2I conditions. K. Bioluminescence detection of leukemia expansion in the spleen (left) and bone marrow (right) at sacrifice in NSG animals transplanted with NALM6-Lucia cells expressing shCTRL (n=5) or shGTF 2I (n=5). Bar plots show mean + / -SD. Unpaired t-test. *p≤0.05.

[0224] Figure 11. GTF2I pharmacological targeting inhibits DUX4-r leukemogenic activity.

[0225] A. Left: RT-qPCR expression analysis of GTF2I upon REH cells treatment for 24h with the indicated SAHA concentrations. Relative mRNA levels were normalized to GAPDH. Bar plots represent the average of three independent biological replicates (n=3). Error bars represent mean + / -SD. One-way ANOVA with multiple comparisons against DMSO. *** p<0.001; **** p<0.0001. Right: immunoblotting validation of GTF2I downregulation upon REH cells treatment for 24h with the indicated SAHA concentrations.

[0226] B. RT-qPCR expression analysis of the indicated wt DUX4 direct targets in REH wt DUX4 cells upon 24h treatment with DMSO or SAHA. Relative mRNA levels were normalized to GAPDH. Bar plots represent the average of three independent biological replicates (n=3). Error bars represent mean + / -SD. One-way ANOVA with multiple comparisons.

[0227] C. Representative phase microscopy images of the decreased tumor spheroids formation of DUX4- IGH expressing REH cells in DMSO and SAHA conditions. Images were acquired at the 45h time point. Scalebar: 300 pm.

[0228] D. Real-time quantification of tumor spheroid formation in NALM6 cells after treatment with DMSO or SAHA. Quantification was performed by computing the phase object average as normalized to TO. Curves represent the average of six biological replicates (n=6). Error bars represent mean + / -SD. Two-Way ANOVA with Bonferroni’s correction to compare curves at the 45h time point. **** p≤0,0001.

[0229] E. Proliferation of NALM6 cells treated with DMSO or SAHA. Curves represent the average of three independent biological replicates (n=3). Error bars represent mean + / - SD. One-Way ANOVA. **** p≤0,0001.

[0230] F. Proliferation assays in low serum condition of REH cells expressing wt DUX4 cells treated with DMSO or SAHA. Curves represent the average of three independent biological replicates (n=3). Error bars represent mean + / - SD. One-Way ANOVA.

[0231] Figure 12. GTF2I is regulated by HDAC6

[0232] A. Left: RT-qPCR expression analysis of GTF2I upon REH cells treatment for 24h with the indicated SAHA or MPT0G211 concentrations. Relative mRNA levels were normalized to DUX4-r. Bar plots represent the average of three independent biological replicates (n=3). Error bars represent mean + / -SD. One-way ANOVA with multiple comparisons against DMSO. ** p<0.01; *** p<0.001; **** p<0.0001. Right: immunoblotting validation of GTF2I downregulation upon REH cells treatment for 24h with luM SAHA or MPT0G211 concentration.

[0233] B. Real-time quantification of tumor spheroid formation in REH DUX4-IGH cells after treatment with DMSO, SAHA or MPT0G211. Quantification was performed by computing the phase object average as normalized to TO. Curves represent the average of six biological replicates (n=6). Error bars represent mean + / -SD. Two-Way ANOVA with Bonferroni’ s correction to compare curves at the 45h time point. **** p≤0,0001.

[0234] Figure 13. The HDAC6-specific inhibitor Tubastatin A downregulates GTF2I and inhibits DUX4- r activity.

[0235] Left: RT-qPCR expression analysis of GTF2I mRNA levels upon treatment with the indicated concentrations of Tubastatin A. Relative mRNA levels were normalized to DUX4-r. Bar plots represent the average of three independent biological replicates (n=3). Error bars represent mean + / -SD. One-way ANOVA with multiple comparisons. *p≤0,05; **p≤0,01; ****p≤0,0001. Right: RT-qPCR expression analysis of CD34-S mRNA levels (a direct DUX4-r transcriptional target), upon treatment with the indicated concentrations of Tubastatin A. Relative mRNA levels were normalized to DUX4-r. Bar plots represent the average of three independent biological replicates (n=3). Error bars represent mean + / -SD. One-way ANOVA with multiple comparisons. **p≤0,01; ***p≤0,001.

[0236] Figure 14. SAHA might regulate GTF2I through c-Myc

[0237] A. CUTAC results on the GTF2I locus in REH cells treated with DMSO or SAHA.

[0238] B. c-Myc enrichment (as determined by chromatin immunoprecipitation followed by sequencing, ChlP-seq) showing that MYC / MAX is the most enriched transcription factor in the GTF2I locus.

[0239] C. Left. Real-time quantification of GTF2I expression in cells knockdown for c-Myc, STAT3 or SMAD3. ** p≤0,01. Right. Pearson correlation an analysis showing that GTF2I and c-Myc expression and significantly and positively correlated in DUX4-r positive B-ALL patients.

[0240] Figure 15. GTF2I and c-Myc regulate each other expression

[0241] A. RT-qPCR analysis of c-Myc (left) or GTF2I (right) mRNA levels upon control (siNT) or c-Myc (siMYC) siRNA-mediated knockdown. Relative mRNA levels were normalized to GAPDH. Bar plots represent the average of three independent biological replicates (n=3). Error bars represent mean + / -SD. One-way ANOVA with multiple comparisons against siNT. * p<0.05

[0242] B. RT-qPCR analysis of GTF2I (left) or c-Myc (right) mRNA levels upon control (siNT) or GTF2I (siGTF2I) siRNA-mediated knockdown. Relative mRNA levels were normalized to GAPDH. Bar plots represent the average of three independent biological replicates (n=3). Error bars represent mean + / -SD.

[0243] One-way ANOVA with multiple comparisons against siNT. * p<0.05

[0244] Figure 16. Kaplan- Meier curves of animals transplanted with PDXs from different DUX4-r+ B- ALL patients. One million live PDX cells were infused into 8 weeks NSG mice. Mice were sacrificed when showing signs of sickness, reached >50% PB engraftment or at the end of the experiment. The curves of animals transplanted with GTF2I high PDXs (SJERG016, ALL-811 and SJBALL230) are significantly different compared to those of animals transplanted with GTF2I low PDXs (SJERG020, SJERG014, SJERG011). Log-rank (Mantel-Cox) test.

[0245] Figure 17. HDAC6 is required for GTF2I expression and for DUX4-r transcriptional activity. A. RT-qPCR for HDAC6, GTF2I and the indicated DUX4-r direct targets in control and HDAC6 knockdown NALM6 cells. B. RT-qPCR for the indicated direct DUX4-r targets in NALM6 treated with DMSO or HDAC6i. Relative mRNA levels normalized to GAPDH. Bar plots represent the average of three independent biological replicates (n=3). Error bars represent mean + / -SD. Unpaired t-test. * p<0.05; ** pcO.Ol.

[0246] Figure 18. HDAC6 is required for DUX4-r biological activity. A.B. Quantification of cell adhesion (A) or migration (B) in NALM6 treated with DMSO or HDAC6i. Bar plots represent the average of three independent biological replicates (n=3). Error bars represent the mean + / -SD. Student’s T-test. **p≤0,01; ***p≤0,001. C. Tumor spheroid formation in NALM6 treated with DMSO or HDAC6i performed by computing the phase object average as normalized to TO. Curves represent the average of six biological replicates (n=6). Error bars represent mean + / - SD. Two-Way ANOVA with Bonferroni’s correction to compare curves at the 45h time point. ** p≤0,01; **** p≤0,0001. D-E. Proliferation in NALM6 (D) or apoptosis in REH empty vector, REH DUX4-r or NALM6 (E) treated with DMSO or HDAC6i. Graphs represent the average of three independent biological replicates (n=3). Error bars represent the mean + / -SD. One-Way ANOVA. . ** p<0.01; *** p≤0,001 ; **** p≤0,0001.

[0247] Figure 19. HDAC6 activates GTF2I expression by stabilizing the c-Myc protein. A. CUT&Tag in NALM6 cells performed with control (IgG) and c-Myc-specific antibodies and analyzed by qPCR with primers located in a control region (region 1) or on regions of the GTF2I locus containing c-Myc binding sites (region 2 and 3). IgG and MYC enrichment levels were normalized relative to Input. Barplots represent the average of three independent biological replicates (n=3). Error bars represent mean + / -SD. Unpaired t-test was performed for each region to compare IgG and MYC enrichment levels. *** p<0.001; **** p<0.0001. B. Immunoblotting with antibodies specific for MycK148Ac, total c-Myc, GTF2I or vinculin (as loading control) in NALM6 treated with DMSO or HDAC6i. Figure 20. HDAC6, c-Myc and GTF2I are overexpressed in DUX4-r leukemia cells. A. HDAC6, c-Myc and GTF2I gene expression levels in B cells from healthy donors (n=21) and DUX4-r+ B-ALL patients (n=107). Wilcoxon rank-sum test. P-values: HDAC6 = 5e-13, MYC = 1.2e-12, GTF2I = 5e- 13. B. The Cancer Dependency Map reveals a significantly higher GTF2I survival dependency in human lymphoid malignancies compared to other cancer cell types. Wilcoxon rank-sum test . **** p≤0,0001. Cancer Dependency Map database (https : / / depmap , or g / portal / 1.

[0248] Figure 21. c-Myc is required for DUX4-r activity. A. RT-qPCR for GTF2I in NALM6 cells treated with DMSO or MYCi-975. Relative mRNA levels normalized to DUX4-r. Bar plots represent the average of three independent biological replicates (n=3). Error bars represent mean + / -SD. Unpaired t- test. ** p<0.01; ***p≤0.001. B. Immunoblotting for c-Myc, GTF2I and VCL (as control) in NALM6 cells treated with DMSO or MYCi-975. C. RT-qPCRfor the indicated direct DUX4-r targets inNALM6 cells treated with DMSO or MYCi-975. Relative mRNA levels normalized to DUX4-r. Bar plots represent the average of three independent biological replicates (n=3). Error bars represent mean + / -SD. Unpaired t-test. ** p<0.01. D. Apoptosis assay in NALM6 cells treated with DMSO or MYCi-975. Bar plots represent the average of three independent biological replicates (n=3). Error bars represent mean + / -SD. Student’s T-test. ****p≤0.0001. (RLU: relative luminescence unit).

[0249] Figure 22. DUX4-r leukemia cells are specifically sensitive to Birinapant. A. Apoptosis assay in NALM6 cells treated with the indicated inhibitor of apoptosis antagonists. Error bars represent the mean + / - SEM. B. cIAPl and XIAP gene expression levels in B cells from healthy donors (n=21) and DUX4- r+ B-ALL patients (n=107). Wilcoxon rank-sum test. P-values: cIAPl = 5e-13; XIAP = 3.8e-04.

[0250] Figure 23. Identification of drug combinations killing preferentially leukemia cells expressing DUX4-r. A. DUX4-r negative (REH EV) or DUX4-r positive (REH DUX4-r and NALM6) cells treated with HDAC6i alone, Birinapant alone or HDAC6i plus Birinapant. 50nM and 25nM concentrations of MPT0G211 and Birinapant were used, respectively. Bar plots represent the average of 3 independent biological replicates (n=3). Error bars represent the mean + / -SD. One-way ANOVA with multiple comparisons against DMSO. ** p<0.01; *** p<0.001. B. DUX4-r negative (REH EV) or DUX4-r positive (REH DUX4-r and NALM6) cells treated with MYCi alone, Birinapant alone or MYCi plus Birinapant. 5uM and 5nM concentrations of MYCi-975 and Birinapant were used respectively. Bar plots represent the average of 3 independent biological replicates (n=3). Error bars represent the mean + / -SD. One-way ANOVA with multiple comparisons against DMSO. **** p<0.0001

[0251] Figure 24. DUX4-r region 160-230 is required for its transcriptional activity. A. RT-qPCR analysis of DUX4 variants (left) and the indicated DUX4-r direct transcriptional targets (right) in inducible REH EV, REH DUX4, REH DUX4-IGH, REH DUX4-del50 and REH DUX4-r 2-160 upon induction with lug / mL of Doxycycline for 12h. Relative mRNA levels normalized to GAPDH. Bar plots represent the average of three independent biological replicates (n=3). Error bars represent mean + / -SD. Unpaired t- test. * p<0.05; ** p<0.01; *** p<0.001; **** p<0.0001. B. RT-qPCR analysis of GTF2I (left) and the indicated DUX4-r direct transcriptional targets (right) in inducible HEK cells, co-transfected with GTF2I together with EV control, DUX4, DUX4-r, DUX4 2-160, DUX4 2-230. Relative mRNA levels normalized to GAPDH. Bar plots represent the average of three independent biological replicates (n=3). Error bars represent mean + / -SD. Unpaired t-test. * p<0.05; ** p<0.01; *** p<0.001; **** p<0.0001.

[0252] Figure 25. Only GTF2I isoforms containing domain B are able to work as DUX4-r co-activators. A. Schematic representation of the domains present in the GTF2I isoforms. B. Evaluation of the DUX4- r co-activator activity of the GTF2I isoforms. HEK cells were co-transfected with DUX4-r in combination with an empty vector control (EV) or the indicated GTF2I isoforms, and expression of GTF2I (left) and the indicated direct DUX4-r targets (right) was analyzed by RT-qPCR. Relative mRNA levels normalized to GAPDH. Bar plots represent the average of three independent biological replicates (n=3). Error bars represent mean + / -SD. Unpaired t-test. * p<0.05; ** p<0.01; *** p<0.001; **** p<0.0001.

[0253] Figure 26. GTF2I exon 12 is required for DUX4-r transcriptional activity. A. Schematic representation of the splice-switch antisense oligonucleotide. B. RT-PCR analysis of GTF2I exon 12 splicing and total GTF2I mRNA upon treatment with the exon 12 skipping ASO. C. RT-qPCR analysis the indicated DUX4-r direct transcriptional targets upon mock or exons-skipping ASO treatment in NALM6 cells. Relative mRNA levels normalized to GAPDH. Bar plots represent the average of three independent biological replicates (n=3). Error bars represent mean + / - SD. Unpaired t-test. * p<0.05; ** p<0.01.

[0254] Figure 27. Inhibition of DUX4-r activity with a GTF2I-derived dominant negative peptide. Expression of the indicated direct DUX4-r target genes analyzed by RT-qPCR in NALM6 cells with GTF2I domain A or domain B peptides in combination with TAT-HA2, an assist endosomal escape peptide consisting of a cell penetrating peptide derived from the HIV trans-activator of transcription (TAT) protein and an endosomal escape peptide from influenza A virus haemagglutinin (HA2) protein. Relative mRNA levels normalized to GAPDH. Bar plots represent the average of three independent biological replicates (n=3). Error bars represent mean + / -SD. Unpaired t-test. * p<0.05; ** p<0.01.

[0255] Figure 28. Maps of the plasmids

[0256]

[0257] EXAMPLES

[0258] MATERIALS AND METHODS

[0259] Plasmids and Cloning

[0260] Lentiviral vectors encoding inducible, N-terminally tagged DUX4, DUX4-IGH and DUX4-del50 were generated using Gateway Technology (Invitrogen) starting from the destination vector pCW57.1, a gift from David Root (Addgene plasmid # 41393; http: / / n2t.net / addgene:41393; RRID:Addgene_41393). The N-terminal Twin-Strep-2xHA (Strep-HA) double affinity tag was recovered by PCR (Table 3) from pCDNA5 / FRT / TO STREP-HA vector, kindly provided by Dr. Giulio Superti-Furga (CeMM Research Center for Molecular Medicine of the Austrian Academy of Sciences, 1090 Vienna, Austria), and inserted in the EcoRI site of pCW57. 1 generating the pCW57.1 Strep-HA vector, which was subsequently used as EV control. Codon altered DUX4 open reading frames (ORF) was recovered by PCR (Table 3) from pCW57.1-DUX4-CA, a gift from Stephen Tapscott (Addgene plasmid # 99281; http: / / n2t.net / addgene: 99281; RRID:Addgene_99281). DUX4-IGH ORF was recovered by PCR (Table 3) pMXs-DUX4-IGH JALSG-003, a gift of Hiroyuki Mano(5), respectively, using the same 5’-primer but different 3’ ones (Table 3) WT DUX4 ORF was recovered by PCR (Table 3) from pCS2- mkgDUX4, a gift from Stephen Tapscott (Addgene plasmid # 21156; http: / / n2t.nct / addgene:21156; RRID:Addgene_21156). Primers were designed to carry attBl / 2 recombination sites. The DUX4-del50 construct was generated through site-directed mutagenesis starting from the wt DUX4 construct, by introducing a premature stop codon 150bp upstream of DUX4 wild type stop codon.

[0261] Table 3. Primer List

[0262] List of oligonucleotides used for cloning and RT-qPCR experiments.

[0263] The constitutive shRNA vector targeting DUX4-r was generated by the GeneScript gene-synthesis service by introducing into pLKO.l-hPGK-Puro lentiviral vector, a gift from Bob Weinberg (Addgene plasmid # 8453; http: / / n2t.net / addgene: 8453; RRID:Addgene_8453), the shDUX4-r sequence described in (14). The non-target shRNA control plasmid pLKO. 1-hPGK-Puro shCTRL vector was obtained from Sigma Aldrich (Cat # SHC016). The GTF2I mammalian expression vector was a gift from Ananda Roy (Addgene plasmid # 22148; http: / / n2t.net / addgene:22148; RRID:Addgene_22148). The lentiviral vector encoding for the constitutively expressed GFP reporter was a gift from Eric Campeau & Paul Kaufman (Addgene plasmid # 17447; http: / / n2t.net / addgene: 17447; RRID:Addgene_17447). The constitutive shRNA vectors targeting GTF2I (pLV[shRNA]-EGFP:T2A:Neo-U6>hGTF2I[shRNA#l]) and control (pLV[shRNA]-EGFP:T2A:Neo-U6>Scramble_shRNA#l) were sourced from VectorBuilder.

[0264] Lentivirus production

[0265] Preparation of lentiviral particles was performed by calcium phosphate co-transfection of HEK293T cells with the lentiviral vector of interest in combination with a pCMV-VSV-G lentiviral envelop, a gift from Bob Weinberg (Addgene plasmid # 8454; http: / / n2t.net / addgene:8454; RRID:Addgene_8454), and the pCMV-dR8.2-dvpr lentiviral packaging, a gift from Bob Weinberg (Addgene plasmid # 8455; http: / / n2t.net / addgene:8455; RRID:Addgene_8455), plasmids. For each transfection, 9x10^6 HEK293T cells were plated 18h prior to transfection. Culture medium was replaced with fresh IMDM medium (Lonza-LOBE12726F), supplemented with 10% fetal bovine serum (FBS) (Gibco- 10270106), or Tetracycline Negative FBS when transfecting inducible constructs, 5mM L-Glutamine and 1% Penicillin / Streptomycin antibiotic cocktail (P / S), 2h before transfection. Respectively, 32ug, 9ug and 12,5ug of lentiviral, VSV-G and packaging vectors were diluted in 0.06X TE and brought to a final volume of l,125mL. 125uL of ice-cold CaC12 was added and the mix was allowed to incubate at RT for 10’, following drop-wise addition of l,250mL of in house-prepared 2X Hanks' Balanced Salt Solution (HBSS) while vortexing the mix at full speed. 2,5mL of the transfection mix was gently added dropwise on HEK293T cells. After 16h, the medium was replaced with fresh IMDM medium, and cells were incubated for 3 Oh before collection of the viral supernatant. Medium was then replaced and, after 24h of further incubation, a second round of virus collection was performed. Viral collections were then pooled and ultracentrifuged at 20000xg for 2h at 4°C. Viral particles were then resuspended in 75uL of Optimem (ThermoFisher-31985-047) for each tube and stored at -80°C. Cell culture, transfection and transduction

[0266] REH (DSMZ-ACC22), NALM6 (DSMZ-ACC128) and JURKAT cells were maintained in RPMI-1640 (Euroclone-ECM2001L) medium supplemented with 20%, 10% and 10% Tetracycline Negative FBS (Euroclone, ECS0182L), respectively, and 1% P / S. HEK293 (ATCC-crll573), HEK293T and HS-5 (ATCC-crll 1882) cells were maintained in DMEM medium supplemented with 10% FBS and 1% P / S. Transient transfection experiments were performed in 6-well plates using Lipofectamine LTX&PLUS reagent (Invitrogen 15338-100) according to manufacturer’s instructions. WT DUX4, DUX4-IGH, DUX4-r 2-160, DUX4-r 2-230, GTF2I -encoding and empty vector (EV) control plasmids were transfected into HEK cells using 2ug of inducible EV, wt DUX4 or DUX4-IGH, DUX4-r 2-160, DUX4- r 2-230 vectors and 250ng of EV or GTF2I vectors plus 750ng of control vector to balance the transfection reaction. After approximately 18h from transfection, lug / mL Doxycycline (dox) (SigmaAldrich-D9891-10G) was added to a final concentration of lug / mL. RNA was collected 8h post dox induction. For transient transfection experiments of GTF2I isoforms, lug of GTF2I-α, GTF2I-β, GTF2i-γ or GTF2I-Δ encoding or empty vectors (EV) were co-transfected together with lug of DUX4- IGH encoding vector. After approximately 18h from transfection, lug / mL Doxycycline (dox) was added to a final concentration of lug / mL.

[0267] Transfection of GTF2I exon 12-skipping antisense oligonucleotides (ASOs) or Mock control was performed in NALM6 cells in 6-well plates using Lipofectamine 3000 reagent (Invitrogen-L3000001) by incubation of 200nM 1 : 1 of selected ASOs for 72h, performing a second transfection after 24h from the first one.

[0268] Transduction of suspension cells was performed by using lx10^6 cells in ImL of RPMI-1640 medium containing 8ug / mL Polybrene (SigmaAldrich-TR-1003-G) and 100uL of viral particles by spinoculation, to increase transduction efficiency. Spinoculation was performed at 1290xg, for 1.30h at 34°C and cells were returned in the cell culture incubator for 48h before changing the medium.

[0269] Inducible REH cells were generated by transduction with pCW57.1 Strep-HA (EV), pCW57.1 Strep- HA codon optimized DUX4 (wt DUX4), pCW57.1 Strep-HA DUX4-IGH (DUX4-IGH) or pCW57.1 Strep-HA DUX4-del50 (DUX4-del50), pCW57.1 Strep-HA DUX4-r 2-160 (DUX4-r 2-160) lentiviruses as described above, followed by selection with 0.25ug / uL puromycin (ThermoFisher- A1113803) for 2 weeks. Then, cells were maintained at O.lug / uL of puromycm. Versions of these cells constitutively expressing GFP were generated by transduction with the pCMV-GFP-Neo lentivirus and selected with 500ug / mL of G418-Sodium Salt (SigmaAldrich-A1720-5G) dissolved in water for 1 month. Then, cells were maintained at 100ug / mL G418 and 0, lug / mL puromycin. NALM6 cells expressing shCTRL or shDUX4-r were generated by transduction with pLKO. 1-hPGK- Puro lentiviruses described above. Versions of these cells constitutively expressing GFP were generated by transduction with the pCMV-GFP-Neo lentivirus as above.

[0270] Inducible REH, NALM6 or NALM6 LUCIA cells expressing shCTRL or shGTF2I were generated by transduction with pLV[shRNA]-EGFP:T2A:Neo-U6 lentiviruses described above. Cells were selected with 500ug / mL of G418-Sodium Salt for 1 month.

[0271] Peptides delivery in NALM6 cells

[0272] For each biological replicate, 600,000 NALM6 cells were resuspended in 200uL of Opti-MEM® Reduced Serum Medium (LifeTechnologies- 31985-047) and incubated with 5uM of either GTF2I-A or GTF2I-B peptides (CASLO) and TAT-HA2 fusion peptide (DBA- HY-P4108) for 30’ at 37°C. Then, cells were spun for 5’ at 300xg and cells were resuspended in 500uL of fully supplemented RPMI medium and incubated in a 48-well plate (Corning- 3548) for 24h before collection of cells.

[0273] RNA extraction and RT-qPCR, RT-PCR and qPCR

[0274] For >2x10^6 cells, total RNA was extracted using the PureLink RNA Mini Kit (LifeTechnologies- 12183025), following the manufacturer’s instructions. For processing fewer cells, total RNA was extracted using the NucleoSpin RNA XS kit (Macherey-Nagel-FC140955N), following the manufacturer’s instructions. Reverse transcription was performed with SuperScript III First-Strand Synthesis SuperMix (LifeTechnologies- 11752-250), following the manufacturer’s instructions. As input, lug of DNA-free RNA was used, or the same amount of RNA for each sample in case of less abundant samples. Quantitative real-time PCR was performed with the iTaq Universal SYBR Green Supermix (Bio-Rad-1725122), using 5uL or 3uL of a 1 : 10 dilution of cDNA and 0,4uM or 0,2uM concentration of primers in the 96 or 384 well PCR plates, respectively. The real-time PCR program used was: 10’ @ 95°C, 30” @ 95°C, 30” @ 58°C, 30” @ 72°C, repeated 39 times, 5” @ 65°C and 30” @ 65°C to terminate the reaction. To evaluate effect on gene expression, inducible REH cells or NALM6 cells were treated with DMSO or 50nM SAHA, 50nM MPT0G211 or 20pM MYCi-975 dissolved in 100% DMSO for 12h before induction of transgenes (EV, wt DUX4 and DUX4-IGH) in REH cells, and additional 12h during transgene induction. RNA was collected after a total of 24h SAHA / MPT0G211 / MYCi-975 treatment and, in case of induction of REH cells, 12h transgene induction. To perform qPCR on MYC CUT&Tag samples, the produced IgG / MYC libraries were diluted 1 :5, and 4uL of DNA and 0,2uM concentration of primers specific for 3 regions within the GTF2I locus in the 384 well PCR plate. 4uL of each respective input sample was loaded and used to normalize IgG / MYC enrichment at the three representative regions of the GTF2I locus. To evaluate the GTF2I exon 12 skipping after ASO treatments, transfected NALM6 cells were collected 72h after the first transfection, lug of RNA was reverse-transcribed and all of the obtained cDNA was used as input for PCR, using 0,2uM isoform specific or total GTF2I primer pairs to evaluate exon 12 skipping as well as total GTF2I RNA levels. The PCR program used was: 2’ @ 95°C, 30” @ 95°C, 30” @ 58°C, 15” @ 72°C, repeated 30 times, 5’ @ 72°C to terminate the reaction. To evaluate the effect on gene expression upon ASOs treatment, three independent parallel replicates were processed in the same manner, and 5uL of a 1 : 10 dilution of cDNA and 0,4uM concentration of primers in the 96 well PCR plates, using the same protocol indicated before. To evaluate the effect of gene expression upon GTF2I exon- A (exon 10) or exon-B (exon 12) peptides delivery, total RNA was purified 24h post-delivery, 500ng of total RNA was used as input for reverse-transcription, and 4uL of a 1 :20 dilution of cDNA and 0,2uM concentration of primers in the 384 well PCR plates, using the same protocol indicated before.

[0275] Protein extraction and immunoblotting

[0276] Total proteins were extracted either via direct lysis by addition of 50uL of 2X Laemmli buffer (Bio- Rad-#1610747) to ~10^6 cells followed by boiling 10’ at 95°C. Alternatively protein extracts were performed in RIPA buffer (25mM Tri-HCl pH8, 150mM NaCl, 10mM NaF, 20mM Na3VO4, 0,1% SDS, 1% NP-40) by incubation for 10’ in ice, followed by centrifugation of the extract at 16000xg, for 10’ at 4°C, and the supernatant containing proteins was collected. RIPA-extracted proteins were quantified with the Protein Assay Dye Reagent Concentrate (Bio-Rad-#5000006) using lug / uL gamma- globin as standard. Absorbance was read at 595nm using a spectrophotometer. For immunoblotting analyses of REH, HEK and Jurkat cells, lul of whole cell extract was diluted to 10uL with Laemmli buffer IX. For the others immunoblotting analyses, 15ug of proteins were diluted in 4X Laemmli buffer to a concentration of IX and boiled for 5’ at 95°C. For immunoblotting analysis of NALM6 upon MPT0G21 l / MYCi-975 treatment, cells were collected after 24 of 50nM MPT0G211 or 20μM MYCi- 975, washed once in PBS and then proteins were extracted via direct lysis by the addition of 30 μl of 2x Laemmli buffer to ~ 0,5*10^6 cells followed by boiling for 10 mm at 95°C.

[0277] For immunoblotting analysis of tandem affinity purification material, 5%, 5% and 10% fractions of input, 1stelution and 2ndelution were loaded, respectively. Standard SDS-PAGE was performed at 10% polyacrylamide concentration, followed by wet-transferred onto a 0.45um nitrocellulose membrane using an 80% Tris-HCl, 10% Methanol solution for 1.30h. Membranes were stained with Ponceau solution and then incubated with a 5% Milk blocking solution in tris-buffered saline supplemented with 0,1% Tween20 (TBS-T) for >lh. Primary antibodies (Table 4) were incubated overnight, oscillating at 4°C and secondary, HRP-conjugated antibodies were incubated for Ihr at room temperature. The SuperSignal West Pico PLUS Chemiluminescent Substrate (LifeTechnologies-34580) was used for the visualization of proteins.

[0278] Table 4. Antibodies List

[0279] List of antibodies used for western blot, CUT&Tag and CUTAC experiments.

[0280] Apoptosis assay

[0281] Apoptosis was measured through Caspase-Gio 3 / 7 luminescent assay (Promega-#G8090), following the manufacturer’s instructions. Luminescence was quantified by Wallac 1420 multilabel Victor3 microplate reader (Perkin Elmer). In inducible REH cells, assays were performed 24h after doxycycline administration. In shCTRL / shGTF2INALM6 cells, assays were performed 72h after lentiviral infection. In shCTRL / shGTF2I REH DUX4-IGH cells, doxycycline was added 48h post lentiviral transduction and apoptosis was measured 24h after doxy administration. For SAHA treatment, NALM6 cells were treated with luM of SAHA (dissolved in 100% DMSO) or DMSO, as control, for 72h before measurement of apoptosis. For inducible REH cells, 1 uM of SAHA or DMSO were administered at the start of the experiment, while doxycycline was supplemented 24h after the start of the experiment. Apoptosis was measured after 72h of SAHA treatment. For MPT0G2211 treatment, REH EV, REH DUX4-IGH and NALM6 cells were treated with DMSO and either 50nM or 100nM concentration of the inhibitor (dissolved in 100% DMSO) for 48h before measurement of apoptosis. For the treatment of NALM6 with MYCi-975, NALM6 cells were treated with DMSO or 5uM of compound (dissolved in 100%DMSO) for 24h before measurement of apoptosis. F or the treatment with inhibitors of inhibitors of apoptosis, DMSO or 25nM concentration of Sepantronium Bromide, S63845, Birinapant, Navitoclax or Venetoclax (dissolved in 100% DMSO) were used to treat NALM6 cells for 24h before measurement of apoptosis. For the combinatorial MPT0G211 / Birinapant treatment, REH EV, REH DUX4-IGH and NALM6 cells were treated with 50nM and 25nM of MPT0G211 and Birinapant, respectively, for 24h before measurement of apoptosis. For the combinatorial MYCi-975 / Birinapant treatment, REH EV, REH DUX4-IGH and NALM6 cells were treated with 5uM and 5nM of MYCi-975 and Birinapant, respectively, for 24h before measurement of apoptosis.

[0282] Cell-stroma adhesion assay

[0283] HS-5 cells were maintained in fully supplemented RPMI medium for at least 72h prior to starting the experiment. Next, 250.000 HS-5 cells were plated into a 12 well plate 8-12h before co-culturing with 100.000 REH-GFP or NALM6-GFP cells for 18h. Cells were then washed twice with PBS. Subsequently, adhering cells were recovered with trypsin, washed, resuspended in PBS and analyzed by FACS for the percentage GFP-expressing cells.

[0284] To evaluate adhesion in knockdown cells, NALM6-GFP cells or GFP+ inducible REH were transduced with shCTRL, shDUX4-r, or shGTF2I lentiviral particles 48h prior the start of the experiment.

[0285] To evaluate adhesion in SAHA treated cells, NALM6-GFP cells or GFP+ inducible REH were treated with DMSO or luM SAHA dissolved in 100% DMSO before co-culture with HS-5 cells. At the same time, GFP+ REH cells were supplemented with 1 ug of Doxycycline to induce DUX4-IGH expression. To evaluate adhesion in MPT0G211 treated cells, NALM6-GFP cells were treated with DMSO or luM SAHA dissolved in 100% DMSO before co-culture with HS-5 cells.

[0286] Transwell migration assay

[0287] HS-5 cells were maintained in fully supplemented RPMI medium for at least 72h prior to starting the experiment. 100.000 HS-5 cells were plated into the bottom compartment of a 24 Transwell plate (Corning- CLS3399) 18-24h before the start of the co-culture. Once HS-5 were fully attached and well- elongated, 50.000 inducible REH-GFP cells, shCTRL or shDUX4-r NALM6-GFP, shCTRL or shGTF2I inducible REH-GFP were seeded in the upper compartment and doxycycline or vehicle (H2O) was added. After 16-18h incubation, cells in the bottom compartment were trypsinized and the amount of GFP+ (GFP-positive) cells that had migrated from the upper compartment was determined by flow cytometry.

[0288] To evaluate migration in SAHA treated cells, NALM6-GFP cells or GFP+ inducible REH were treated with DMSO or luM SAHA dissolved in 100% DMSO before incubation in the transwell upper compartment. At the same time, GFP+ REH cells were supplemented with 1 ug of Doxycycline to induce DUX4-IGH expression. To evaluate migration in MPT0G211 treated cells, NALM6-GFP cells were treated with DMSO or luM MPT0G211 dissolved in 100% DMSO before incubation in the transwell upper compartment.

[0289] Cell proliferation in low serum

[0290] 200.000 inducible REH or NALM6 cells were seeded in 2mL of RPMI medium supplemented with 5% of FBS in a 6 well plate to get a final concentration of cells of 100.000 / mL. Transgenes were induced with lug / mL of Doxycycline (for inducible REH cells) and cell proliferation was measured by automatic cell counting (Bio-Rad, TC20-1450102) every 48h, accompanied by replacement of the medium with fresh Doxycycline for up to 7 days. For NALM6 cell proliferation analysis upon HDAC6 inhibition, luM concentration of MPT0G211 was added at the start of the experiment and replaced every 48h, together with automatic cell counting. siRNA transfection siRNA transfection experiments were performed in 12-well plates using Lipof ectamine 3000 (Thermo Fisher Scientific, #L3000001) as follows: 30nM of siRNA and 3uL of Lipofectamine 3000 reagent were added in Optimem medium and incubated for 15’ before addition to the well. After 24h, the medium was replaced with fresh regular medium and incubated again for 24h before cell collection. For siRNA mediated HDAC6 KD in HEK cells, cells were collected 24h after transfection.

[0291] Endogenous GTF2I-DUX4-r Co-IP

[0292] 5x10^7 NALM6 cells were resuspended in 4mL of Buffer N (10mM HEPES-KOH pH7.9, 300mM sucrose, 10mM KC1, 0,lmM EDTA, 0,lmM EGTA, 0,lmM DTT, 0,75mM spermidine, 0,15mM spermine, 0,1% NP-40, 50mM NaF) and incubated in ice for 10’ to isolate intact nuclei. At this step, a 30uL aliquot was collected and lysed in Laemmli buffer to be used as input. After one wash in the same volume of Buffer N, nuclei were resuspended in the same volume of high-salt nuclear extraction buffer C420 (20mM HEPES-KOH pH7.9, 420mM NaCl, 25% Glycerol, ImM EGTA, 0,1 mM DTT 50mM NaF), and vigorously shaken in a thermomixer at 1400rpm for 30’ at 4°C in order to extract nuclear proteins. Samples were clarified by centrifugation at 16000xg for 10’ at 4°C, and the supernatant, which contains nuclear proteins, was collected. Nuclear extracts were then adjusted to 150mM salt by addition of HEPES buffer (20mM HEPES-KOH pH7.9, 50mM NaF), and incubated at 4°C while rotating with and 20U / mL of Benzonase (Sigma Aldrich- E8263-5KU) and 50ng / mL of RNase A (ThermoFisher- EN0531 ) to remove any remaining nucleic acids. After clearing at 16000xg for 10’ at 4°C, supernatants were pre-cleared with 200uL (bead volume) of cross-linked Sepharose (Sigma Aldrich- DCL6B100) for Ih while rotating at 4°C. Extracts where then quantified with the Protein Assay Dye Reagent Concentrate (Bio-Rad #5000006) using lug / uL gamma-globin as standard. Absorbance was read at 595nm using a spectrophotometer. 2mg of nuclear proteins were then incubated with 4ug of anti- DUX4c antibody or ChromPure Rabbit IgG (Jackson Immunoresearch - #011-000-003), as negative control, rotating O.N. at 4°C. The following day, Protein G Dynabeads (Invitrogen - #10004D) were blocked with 3x washes with 500uL of 0,5% BSA / PBS solution, and then 30uL were added to Ab- protem complexes and incubated for 3h, rotating at 4°C. Samples were then washed 3x times by incubation with TNN-HS (50 mM HEPES pH 8.0, 150 mM NaCl, 5 mM EDTA, 0.5% NEMO, 50 mM NaF, protease inhibitors) followed by 3’ of rotation at 4°C. Finally, beads were boiled with 30uL of 2X Laemmli buffer at 95°C for 10’, and centrifuged for 5’ at 1200xrpm to remove beads. 5uL of Input sample and 7uL of either DUX4-r IP or IgG were loaded in a 10% acrylamide gel for western blot analysis.

[0293] Live-cell spheroid assay

[0294] 40.000 inducible REH cells were seeded in 800uL (50.000 / mL) of RPMI medium containing 5% of FBS and 1 ug / mL of Doxycyhne into a 48well plate (Costar #3548). Cells were then incubated for 72h into the Incucyte Live-Cell Analysis Systems (Sartorius) for real-time analysis of cellular aggregation into spheroids. Clustering analysis was performed according to the Incucyte Handbook, Chapter 5a, “Kinetic Assays for Immune Cell Activation and Proliferation”. Label- free, phase-contrast imaging was used to detect morphological differences between cells expressing different DUX4 variants. Quantification of the average spheroid area was calculated using the following parameters: Analysis of Phase Area Object Average normalized to TO, cell / background ratio: -1, fill holes: -5.

[0295] RNA-sequencing

[0296] Total RNA was extracted as described above from ~3x10^6 inducible REH cells per sample, after 12h of doxycycline induction in 2 mL (1,5x10^6 / mL) of fully supplemented RPMI. RNA concentration, purity and integrity were determined through the BioAnalyzer 2100 system. All samples used for RNA- seq library preparation showed an RNA integrity number (RIN) greater than 9. RNA-seq libraries were generated from biological quadruplicates using the TruSeq Stranded mRNA library prep kit, using 100ng of total RNA as input. Libraries were run with a NovaSeq 6000 instrument and sequenced in single-end (SE) generating 75bp reads, for a total number of 50M (million) clusters per sample.

[0297] Transcriptomic analyses

[0298] Sequencing adapters were removed using trimmomatic (v0.39) (- / 7) and fastq files were then aligned to the human genome assembly GRCh38 (hg38) using the STAR aligner (v2.5.3a) (48). Annotation of genomic features was performed using the featureCounts tool (vl .6.4) (49), using the GENCODE v31 Gene transfer format (GTF). Differential gene expression analysis was carried out with DeSeq2 (vl.3) package (Bioconductor) (50).

[0299] To generate lists of DUX4, DUX4-IGH and DUX4-del50 up- and downregulated genes, DGEs showing Log2FC≥l or Log2FC≤-l for each condition have been compared with their relative expression levels (RPKM) in the given condition. Only genes with RPKM≥0,5 were used for further analyses.

[0300] Gene-Set enrichment analyses was performed using the GSeABase software (1.57.0) package (Bioconductor) (51), using the Canonical Pathways (CP), Gene Ontology (GO), from the Molecular Signatures Database MSigDB (v7.5.1) (52) and setting a False Discovery Rate value to 0,05.

[0301] To check the expression of DUX4 / DUX4-IGH direct transcriptional targets in EV control cells, direct target lists have been compared with the list of genes expressed in EV REH cells with RPKM>0,5.

[0302] To test the percentage of surface proteins activated by DUX4-IGH, the list of direct DUX4-IGH transcriptional targets has been compared with the list of surface proteins identified by Bausch-Fluck et al. (60).

[0303] Immunoblotting

[0304] 2x10^6 of REH cells in 2mL of RPMI were treated with luM of SAHA or MPT0G211, or the same volume of DMSO as negative control, for 24h prior to cells harvesting. Proteins were extracted by boiling cell pellets in 50uL of 2X Laemmli buffer for 10’ at 95°C. 2uL of each sample were loaded for western blot analysis.

[0305] RT-qPCR

[0306] 2x10^6 of inducible REH cells in 2mL of RPMI were treated with different concentrations of either SAHA or MPT0G211, or the same volume of DMSO as negative control, 12h before induction of DUX4-r with lug / mL of Doxycycline for 12h. After a total of 24h after treatments with various concentrations of SAHA / MPT0G211, cells were harvested and RNA was extracted using the PureLink RNA Mini Kit (LifeTechnologies-12183025), following the manufacturer’s instructions. Reverse transcription was performed with SuperScript III First-Strand Synthesis SuperMix (LifeTechnologies- 11752-250), following the manufacturer’s instructions. Quantitative real-time PCR was performed with the iTaq Universal SYBR Green Supermix (Bio-Rad-1725122), using 3uL of a 1 : 10 dilution of cDNA and 0,2uM concentration of primers in the 384 well PCR plate. DUX4-r mRNA levels was used to normalize GTF2I expression levels upon HDACi treatment.

[0307] CUT&Tag

[0308] Starting from >5x10^6 NALM6 cells or REH / HEK cells induced with doxycycline for 12h, nuclei from three or two independent replicates were prepared by resuspending cells in NE1 buffer (20mMHEPES- KOH pH7.9, 10mM KC1, 0,5mM Spermidine, 1% Triton X-100) at a cellular density of 10^6 cells / mL, followed by 10’ incubation in ice. Samples were spun for 4’ at 600xg at 4°C and the nuclear pellet was resuspended in 1 volume (V) of PBS. A light cross-linking was performed by adding formaldehyde to 0,1% final concentration for 2’, followed by quenching with a double molar concentration of glycine. Nuclei were spun and resuspended in ½ V of Wash Buffer 150 (20mM HEPES-KOH pH7.9, 150mM NaCl, 0,5mM Spermidine). Nuclei were then manually counted using the trypan blue exclusion assay, as described(27). For each CUT&Tag assay against different DUX4 versions, GTF2I and MY C, 100000 nuclei were immobilized to 4uL of Concanavahn A-conjugated beads (ConA-Epycypher-21-1401) for 10’ at RT.

[0309] For each CUT&Tag against H3K27 Ac, 50.000 nuclei were immobilized to 4uL of Concanavalin A for 10’ at RT.

[0310] The primary (1st) antibody (Ab) was diluted 1 :100 in antibody buffer (20mM HEPES-KOH pH7.9, 150mM NaCl, 0,5mM Spermidine, 0,5mg / mL BSA, 0,4mM EDTA pH7.5) and incubated with the ConA-bound nuclei ON at4°C while rotating. The following day, ConA-nuclei- 1 stAb were spun 100xg, 1 ” , the supernatant was removed, nuclei were resuspended in 100ul of Wash Buffer 150 and incubated with a secondary antibody (2nd) diluted 1 : 100 in Wash Buffer 150 for 1 h at RT. After removal of the 2nd Ab, nuclei were washed once with 200ul of Wash Buffer 150 and incubated with the protein AG- conjugated-Transposase (pAG-Tn5, Epycypher #23615-1117) preloaded with sequencing adapters, diluted 1:20 in Wash Buffer 300 (20mM HEPES-KOH pH7.9, 300mM NaCl, 0,5mM Spermidine) for Ih at RT. Nuclei were then washed twice in Wash Buffer 300 and tagmentation reaction was performed by addition of 10mM MgC12 following incubation at 37°C for Ih. After a wash with TAPS buffer (10mM TAPS in H2O), tagmented DNA was released by addition of 5uL 0.1% SDS release buffer (0,1% SDS, 10mM TAPS in H2O) following incubation at 58°C for Ih. Then, 15uL of 0.67% Triton- X-100 was added to allow PCR for library preparation. For the library prep, 2uL of 10uM i5 universal adapter primer (Nextera) and 2uL of 10uM i7 uniquely barcoded adapter primers were added to released, tagmented DNA, followed by addition of 25uL of the Q5 High-Fidelity 2X Master Mix (New England Biolabs, M0492L).

[0311] PCR was performed as follows: Cycle 1 : 72 °C for 5’ (gap filling), Cycle 2: 98 °C for 30”, Cycle 3: 98 °C for 10” Cycle 4: 63 °C for 10” . 72°C for 1 ’ and hold at 8°C. For H3K27Ac CUT&Tag and CUTAC, repeat Cycles 3-4 13 times; for CUT&Tag against different DUX4 versions, GTF2I and MYC, repeat cycles 3-4 19 times. Amplified DNA was then recovered through size exclusion purification by addition of 1.3V of AMPureXP magnetic beads (Beckman- A63881), followed by two washes in 80% ethanol and elution in 22uL of 10mM Tris-HCl (pH8). Quantification and average fragment size of libraries was performed via the 4200 TapeStation system (Agilent), using the D5000 high sensitivity kit (Agilent 5067-5592 / 5067-5593). 450bp of average library fragment size was selected for library quantification. A 5nM pool of the libraries was then run with the NovaSeq instrument and sequenced in paired end (PE), generating 2xl50bp reads. 4M-10M reads per sample for H3K27Ac CUT&Tag and CUT AC, and at least 10M reads per sample for TF (DUX4 / DUX4-r / GTF2I) were obtained. To generate the input samples for MYC CUT&Tag, 100,000 NALM6 nuclei were harvested at the start of the experiment, diluted 1:3 in wash buffer 150, and incubated Ih at 55°C after addition of 40ug and 5ug of Benzonase nuclease and RNase A respectively, followed by sonication at high intensity for 7x cycles, 15” ON / OFF before clearing at 16000xg for 10’ at 4°C.

[0312] Bioinformatic analyses of DUX4-r patient datasets

[0313] To evaluate expression levels selected genes (HDAC6, MYC, GTF2I, cIAPl and XIAP), raw HTSeq counts from N=106 DUX4-r patients and N=21 CD 19+ healthy donors were retrieved from St. Jude Cloud (PMID 30643249) and processed to replace NA values with 0. Then, gene lengths were extracted from Homo_sapiens.GRCh37.87.gtf annotation file and transcripts per million (TPM) values were calculated for both DUX4-r and patients and CD19+ healthy donors. Boxplots were generated using the TPM values of the indicated genes using the “geom_boxplot” ggplot2 function, while statistical significance was calculated using the “geom_signif” function of ggplot2 of RStudio, using the Wilcoxon rank-sum test.

[0314] CUTAC

[0315] Starting from >5x10^6 REH cells induced with doxycycline for 12h, nuclei from two independent replicates were prepared by resuspending cells in NE1 buffer (20mM HEPES-KOH pH7.9, 10mM KC1, 0,5mM Spermidine, 1% TritonX-100) ata cellular density of 10^6 cells / mL, followed by 10’ incubation in ice. Samples were spun for 4’ at 600xg at 4°C and the nuclear pellet was resuspended in 1 volume (V) of PBS. Nuclei were then manually counted using the trypan blue exclusion assay, as previously described(27). For each CUTAC assay against 50000 nuclei were immobilized to 4uL of Concanavalin A-conjugated for 10’ at RT.

[0316] The primary (1st) antibody (Ab) targeting Phospho_Rbpl CTD (Ser5) was diluted 1: 100 in antibody buffer (20mM HEPES-KOH pH7.9, 150mMNaCl, 0,5mM Spermidine, 0,5mg / mLBSA, 0,4mMEDTA pH7.5) and incubated with the ConA-bound nuclei ON at 4°C while rotating. The following day, ConA- nuclei-lstAb were spun 100xg, 1”, the supernatant was removed, nuclei were resuspended in 100ul of Wash Buffer 150 and incubated with a secondary antibody (2nd) diluted 1: 100 in Wash Buffer 150 for Ih at RT. After removal of the 2nd Ab, nuclei were washed once with 200ul of Wash Buffer 150 and incubated with the protein AG-conjugated-Transposase (pAG-Tn5, Epy cypher #23615-1117) preloaded with sequencing adapters, diluted 1 :40 in Wash Buffer 300 (20mMHEPES-KOH pH7.9, 300mMNaCl, 0,5mM Spermidine) for Ih at RT. 50uL of pre- warmed CUTAC buffer (5 mM MgC12, 10 mM TAPS) was added to each tube followed by incubation at 37°C for 20’. After a wash with TAPS buffer (10mM TAPS in H2O), tagmented DNA was released by addition of 5uL 0.1% SDS release buffer (0,1% SDS, 10mM TAPS in H2O) following incubation at 58°C for Ih. Then, 15uL of 0.67% Triton-X-100 was added to allow PCR for library preparation. For the library prep, 2uL of 10uM 15 universal adapter primer (Nextera) and 2uL of 10uM 17 uniquely barcoded adapter primers were added to released, tagmented DNA, followed by addition of 25uL of the Q5 High-Fidelity 2X Master Mix (New England Biolabs, M0492L).

[0317] PCR was performed as follows: Cycle 1 : 72 °C for 5’ (gap filling), Cycle 2: 98 °C for 30”, Cycle 3: 98 °C for 10” Cycle 4: 63 °C for 10”. 72°C for 1 ’ and hold at 8°C. Repeat Cycles 3-4 13 times. Amplified DNA was then recovered through size exclusion purification by the addition of 1.3V of AMPureXP magnetic beads (Beckman-A63881), followed by two washes in 80% ethanol and elution in 22uL of 1 OmM Tris-HCl (pH8). Quantification of the average fragment size of libraries was performed via the 4200 TapeStation system (Agilent), using the D5000 high sensitivity kit (Agilent 5067-5592 / 5067- 5593). 175bp-650bp of average library fragment size was selected for library quantification. A 5nM pool of the libraries was then run with the NovaSeq instrument and sequenced in paired end (PE), generating 2xl50bp reads.

[0318] CUT&Tag and CUTAC analyses

[0319] Sequencing adapters were removed via Cutadapt (v4. 1 ) and then trimmed fastq files were aligned with Bowtie2 (v2.2. 1) (53) to the human genome assembly GRCh38 (hg38) from Genome Reference Consortium, using the following parameters: —local -very-sensitive -no-mixed -no-discordant — phred33 -1 10 -X 700. Bam files containing uniquely mapped reads were created using samtools vl . 15. 1 (54).

[0320] CUT&Tag signals were called on using MACS (v2.2.7.1) (55) as normalized to IgG control signal, using the following settings: FDR cutoff of le-3, BAMPE format, keep all duplicates.

[0321] The MACS — bdg setting was used to generate IgG control scaled, normalized bigwig files, which were used to visualize normalized binding signals. The same bigwig files were also used to generate profile plots using deepTools plotprofile (v3.5.0) (56). Genes proximal to peaks were annotated against the hg38 genome using annotatePeaks.pl from HOMER v4(57) DUX4, DUX4-IGH and DUX4-del50 binding motifs were identified using findMotifsGenome.pl from HOMER v4, using standard parameters.

[0322] To identify regions associated by GTF2I both in REH and HEK cells, intersectBed from the BEDtools suite was used with default parameters.

[0323] DUX4 and DUX4-IGH direct targets lists were generated as follows: non-redundant, annotated genes associated to DUX4 or DUX4-IGH peaks were crossed with Log2FC≥l and RPKM≥0,5 DGEs in the respective lists. Only genes present in both peaks and DGEs lists were considered as DUX4 or DUX4- IGH direct target genes.

[0324] To assess difference between genomics signals (coverage) for selected regions sets (wt DUX4 or DUX4-IGH direct targets), per-bin corrected Wilcoxon Rank-sum tests was performed using the dsCompareCurves tool from deepStats (v0.4) (61).

[0325] To analyze genome-wide GTF2I DNA association in REH cells expressing EV or DUX4-IGH, reads per genome coverage (RPGC) normalized counts were generated using the deepTools bamCoverage function with standard parameters. Coverage files were then used as input for the deepTools plotHeatmap function, and were plotted against the human reference genome GENCODE v43. To analyze DUX4-IGH association to chromatin in HEK cells in presence / absence of GTF2I, RPGC- normalized counts of DUX4-IGH CUT&Tag in HEK +EV and HEK +GTF2I, as well as RPGC- normalized counts of IgG control CUT&Tag were generated using the deepTools bamCoverage function with standard parameters. The generated coverage files were then used as input for the deepTools plotHeatmap function, and were plotted against the human reference genome GENCODE v43. Furthermore, to generate boxplots of normalized counts signals, the generated coverage files were used as input for the multiBigWigSummary function of deepTools.

[0326] Identification of candidate transcriptional regulators of GTF2I

[0327] To identify differential CUT AC signal upon SAHA treatment, the getDifferentialPeaksReplicates.pl tool from HOMER v4 was used. Genomic regions of differential CUTAC signal neighboring the GTF2I locus were extracted and used as input for transcription factor binding analysis via the LOLA web tool (Nucleic Acids Research, vol. 46, no. wl, pp. wl94-wl99, Jul. 2018.), which identified significant enrichment of c-Myc binding sites in the input regions. Finally, database analysis of GTF2I transcriptional regulation identified c-Myc within the top 20 genes showing significant RNA-seq gene- gene co-expression with GTF2I as well as c-Myc enrichment (as determined by chromatin immunoprecipitation followed by sequencing, ChlP-seq) in the GTF2I genomic locus in both the ENCODE and ChEA databases showing that MYC / MAX is the most enriched transcription factor in the GTF2I locus.

[0328] Tandem affinity purification

[0329] Tandem affinity purification from REH cells was performed 12h after doxycycline induction. 2.5x10^8 cells were resuspended in 20mL of Buffer N (10mMHEPES-KOHpH7.9, 300mM sucrose, 10mMKCl, 0,lmM EDTA, 0,lmM EGTA, 0,lmM DTT, 0,75mM spermidine, 0,15mM spermine, 0,1% NP-40, 50mM NaF) and incubated in ice for 10’ to isolate intact nuclei. After one wash in the same volume of Buffer N, nuclei were resuspended in the same volume of high-salt nuclear extraction buffer C420 (20mM HEPES-KOH pH7.9, 420mM NaCl, 25% Glycerol, ImM EGTA, 0,1 mM DTT 50mM NaF), and vigorously shaken in a thermomixer at 1400rpm for 30’ at 4°C in order to extract nuclear proteins. Samples where then centrifuged at 16000xg for 10’ at 4°C, and the supernatant, which contains nuclear proteins, was collected. Nuclear extracts were then adjusted to 150mM salt by addition of HEPES buffer (20mM HEPES-KOH pH7.9, 50mM NaF), and incubated at 4°C while rotating with lug / mL of avidm (IBA #2-0204-015) to chelate biotinylated proteins, and 20U / mL of Benzonase nuclease (Sigma Aldrich- E8263-5KU) and 50ng / mL of RNase A (ThermoFisher- EN0531) to remove any remaining nucleic acids. After clearing at 16000xg for 10’ at 4°C, supernatants were pre-cleared with 2,5mL (bead volume) of cross-linked Sepharose (Sigma Aldrich- DCL6B100) for Ih while rotating at 4°C. Extracts where then quantified with the Protein Assay Dye Reagent Concentrate (Bio-Rad #5000006) using lug / uL gamma-globin as standard. Absorbance was read at 595nm using a spectrophotometer. For each purification, 80mg of pre-cleared nuclear extracts were incubated with 2mL (bead volume) of StrepTactin sepharose beads (IBA #2-1201-010), for 4h at 4°C while rotating. Then, three washes in TNN-HS (10mM HEPES-KOH pH7.9, 150mM NaCl, 0,1% NP-40, 50mM NaF) buffer were performed, and beads-bound proteins were specifically eluted with three subsequent incubations with 2V of 2.5mM biotin-containing TNN-HS buffer. Next, eluted proteins were incubated with 2mL (bead volume) of anti-HA agarose beads (SigmaAldrich-A2095) while rotating at 4°C overnight. The following day, two washes in TNN-HS were performed, followed by two additional washes in TNN buffer (10mM HEPES-KOH pH7.9, 150mM NaCl, 50mM NaF) in order to get the rid of any leftover NP-40 detergent in the buffer. Finally, HA beads- bound proteins were eluted with elution buffer (50mM HEPES-KOH pH7.9, 150mM NaCl, 1% SDS) and precipitated with Acetone. For this, 4V of ice-cold 100% acetone were added to IV of eluted proteins, followed by vortexing and incubation at - 20°C for Ih. Samples were then centrifuged at 15000xg for 10’. The whole procedure was repeated twice to get the rid of all the remaining SDS. One aliquot of each protein sample was kept for quantification via silver staining.

[0330] Mass spectrometry sample preparation

[0331] Samples were resolubilized in 50 uL of detergent-based buffer [1% sodium deoxycholate, 10 mM tris(2- carboxyethyl)phosphine, 10 mM Tris, and 40 mM chloroacetamide] with Complete mini EDTA-free protease inhibitor cocktail (Roche- 04693159001), sonicated for 15 cycles (30 sec on: 30 sec off) and heated at 95C for 10min. Trypsin digestion (1.5 h) and peptide cleanup were performed using the S- trap system (Protifi) following manufacturer’s specifications. Peptides were eluted and vacuum centrifuged to complete dryness. Purified peptides were resolubilized in 1% FA to a concentration of 1 ug / uL and 1 uL was injected to achieve a peptide load of 1 pg on column.

[0332] Mass spectrometry data acquisition

[0333] Spectra were acquired on an Orbitrap Exploris 480 mass spectrometer (Thermo Fisher Scientific) in data-dependent fashion (DDA) coupled to an Ultimate3000 liquid chromatography system (Thermo Fisher Scientific) and separated on a 50-cm reversed phase column packed in-house (Poroshell EC-C18, 2.7 pm, 50 cm x 75 μm; Agilent Technologies). Samples were eluted over a linear gradient ranging from 9-13% acetonitrile over 2 min and from 13-44% over 95min, 44-99% acetonitrile in 3 minutes, followed by 99% acetonitrile for 5 minutes with a flow rate of 300 nl / min. MSI spectra had a resolving power of 60.000 at 200 m / z with the AGC target set to “standard” and the maximum injection time set to “auto”. MS / MS spectra were acquired with HCD fragmentation, an HCD collision energy of 28%, a 1.4-m / z-wide isolation window, a resolving power of 30.000 at 200 m / z, the AGC target set to “standard” and the maximum injection time set to “auto”.

[0334] Mass spectrometry data processing

[0335] To quantify proteins, the raw data were loaded into the MaxQuant (58) software version 2.2.0 to search the human_proteome 20180425 (93,606 sequences; 37,037,628 residues). Searches were performed using default settings, namely: trypsin as proteolytic enzyme; 1 missed cleavage allowed; protein N- terminus-acetylation. Peptides and proteins were accepted with an FDR of less than 1 %. Label-free protein quantification was based on the spectral counts considering only proteins identified with minimum two peptides in any tandem affinity purification. Specific interactors of wt DUX4, DUX4- IGH or DUX4-del50 were selected as those being detected in all the three biological replicates and showing spectral counts (MS / MS) fold enrichment ≥ 2 with respect to the control tandem affinity purifications. To generate protein-protein interaction networks, Significance Analysis of INTeractome (SAINT v2.0) (59) was performed using as input the list of identified wt DUX4, DUX4-IGH or DUX4- del50 specific interactors. Only proteins showing a Fold Change score (FC) >1.5 and probability score (SAINT SP) >0.9 were included in the protein- protein interaction network.

[0336] Endogenous GTF2I-DUX4-r Co-IP

[0337] 5x10^7 NALM6 cells were resuspended in 4mL of Buffer N (10mM HEPES-KOH pH7.9, 300mM sucrose, 10mM KC1, 0,lmM EDTA, 0,lmM EGTA, 0,lmM DTT, 0,75mM spermidine, 0,15mM spermine, 0,1% NP-40, 50mM NaF) and incubated on ice for 10’ to isolate intact nuclei. At this step, a 30uL aliquot was collected and lysed in Laemmli buffer to be used as input. The remaining nuclei, after one wash in the same volume of Buffer N, were resuspended in the same volume of high-salt nuclear extraction buffer C420 (20mM HEPES-KOH pH7.9, 420mM NaCl, 25% Glycerol, ImM EGTA, 0,1 mM DTT 50mM NaF), and vigorously shaken in a thermomixer at 1400rpm for 30’ at 4°C in order to extract nuclear proteins. Samples were clarified by centrifugation at 16000xg for 10’ at 4°C, and the supernatant, which contains nuclear proteins, was collected. Nuclear extracts were then adjusted to 150mM salt by addition of HEPES buffer (20mM HEPES-KOH pH7.9, 50mM NaF), and incubated at 4°C while rotating with 20U / mL of Benzonase (Sigma Aldrich- E8263-5KU) and 50ng / mL of RNase A (ThermoFisher- EN0531) to remove any remaining nucleic acids. After clearing at 16000xg for 10’ at 4°C, supernatants were pre-cleared with 200uL (bead volume) of cross-linked Sepharose (Sigma Aldrich- DCL6B100) for Ih while rotating at 4°C. Extracts where then quantified with the Protein Assay Dye Reagent Concentrate (Bio-Rad #5000006) using lug / uL gamma-globin as standard. Absorbance was read at 595nm using a spectrophotometer. 2mg of nuclear proteins were then incubated with 4ug of anti-DUX4 antibody (R&D Systems - MAB95351) or ChromPure Rabbit IgG (Jackson Immuno - #011-000-003), as negative control, and rotated O.N. at 4°C. The following day, Protein G Dynabeads (Invitrogen - #10004D) were blocked with 3x washes with 500uL of 0,5% BSA / PBS solution, and then 30uL were added to each Ab-protein complex and incubated for 3h, rotating at 4°C. Samples were then washed 3x times by incubation with TNN-HS (50 mM HEPES pH 8.0, 150 mM NaCl, 5 mM EDTA, 0.5% NP-40, 50 mM NaF, protease inhibitors), followed by 3’ of rotation at 4°C. Finally, beads were boiled with 30uL of 2X Laemmli buffer at 95°C for 10’, and centrifuged for 5’ at 1200xrpm to remove beads. 5uL of Input sample and 7uL of either DUX4-r IP or IgG were loaded in a 10% acrylamide gel for western blot analysis.

[0338] In vivo studies

[0339] All mouse experiments were approved by the Institutional Animal Care and Use Committee (IACUC) of San Raffaele University Hospital and Scientific Institute and by the Italian Ministry of Health (Rome, Italy). Mice were kept in a specific-pathogen-free (SPF) facility within individually ventilated cages and were given irradiated food and water ad libitum. Eight to 10 weeks old NSG (NOD.Cg-Prkdcscid I12rgtmlWjI / SzJ) mice (Charles River Laboratories) were infused i.v. with 250.000 shCTRL or shGT F 2I NALM6-Lucia cells. After 1 week, blood was collected, and cell engraftment was evaluated by reading the secreted luciferase using the Berthold Mithras (model) Luminometer with a read time of 10s. Subsequently, blood was collected every 7 days and mice were sacrificed when relative light units (RLU) exceeded the threshold of 1.5 x10^6 or when manifesting clinical signs of suffering.

[0340] For the pharmacological treatment, NSG mice were transplanted with 250.000 NALM6-Lucia cells. One week after transplantation, once an efficient engraftment was reached, mice were randomly split in groups and treated intraperitoneally (i.p.) with either 50mg / Kg SAHA or DMSO, as vehicle control dissolved in 2% DMSO + 30% PEG300 + 5% Tween80 + ddH2O. SAHA or DMSO IP injections were repeated every day, changing side on every administration. Animals were sacrificed when relative light units (RLU) exceeded the threshold of 1.5 * 10^6 or when manifesting clinical signs of suffering.

[0341] Quantification and statistical analyses

[0342] Statistical analysis was performed with either Graphpad Prism 9 or R. For RNA-seq, CUT&Tag, CUTAC, RT-PCR, WB, FACS, spheroid formation assay, proliferation assay as well as in vivo experiments, statistical parameters including p value, replicates and S.D. are detailed on the figures or annotated in the respective figure legends.

[0343] RESULTS

[0344] DUX4 and DUX4-r associate to and activate different gene sets.

[0345] In vitro culture of primary B-ALL cells is very challenging. Hence, to compare the activity of wt DUX4 and DUX4-r variants, within the present invention REH cells were used, a near-diploid B-ALL cell line lacking endogenous expression of either wt DUX4 or its rearrangements (5, 6). The inventors focused on patient-derived DUX4-r variants representing the distinct types of mutations documented in B-ALL patients: the DUX4-IGH chimera mostly used in the literature (5, 6, 22) and a DUX4-r lacking the whole CTD without any extra amino acid (DUX4-del50) (6). Doxycycline (Dox)-inducible REH cells were generated with the empty vector (EV), wt DUX4, DUX4-IGH or DUX4-del50.

[0346] Upon Dox induction, wt DUX4 was expressed at consistently lower levels compared to DUX4-r variants (Fig. 7A-B). Surprisingly, only wt DUX4 was able to induce apoptosis (Fig. 7C), in contrast with the “just right” model, which posits that wt DUX4 or its variants induce apoptosis when expressed at high levels (75). It was decided to compare the transcriptional ability of wt DUX4 and DUX4-r variants. To this aim, RNA-seq upon 12h of Dox induction was performed, to avoid the toxic effect of prolonged wt DUX4 expression and enrich for primary targets of the transcription factors. Principal component analysis (PCA) showed very tight clustering of the biological replicates and that the gene expression profiles of cells expressing DUX4-IGH and DUX4-del50 cluster nearby and distantly from those of EV or DUX4 samples (Fig. 7D). Despite differential gene expression (DGE) analysis indicated that wt DUX4 or DUX4-r variants are mainly associated to activation of gene expression, DUX4-r variants regulate a very different set of genes compared to wt DUX4 (Fig. 1A). Notably, DUX4-IGH and DUX4-del50 regulated genes are qualitatively similar, even if DUX4-del50 appears to have a milder transactivation ability with respect to DUX4-IGH (Fig. 1 A). Accordingly, we found a nearly complete overlap between DUX4-IGH and DUX4-del50 regulated genes, which was instead minimal with wt DUX4 (Fig. 7E). Previous works failed to identify significantly deregulated pathways on the gene signature specific of DUX4-r B-ALL patients (S, 20). Using the publicly available datasets (6, 8), inventors found a highly significant overlap of the DUX4-r B-ALL patient gene signature with the DUX4-IGH or DUX4-del50 datasets, which was instead negligent with the wt DUX4 dataset (Fig. IB). Next gene set enrichment analysis (GSEA) was performed on the list of genes commonly deregulated in DUX4-r+ B-ALL patients and the DUX4-IGH and DUX4-del50 datasets. It is referred to these commonly deregulated genes as the DUX4-r core gene set from here onward. GSEA revealed that the upregulated DUX4-r core gene set is mostly enriched in molecules involved in cell-cell / cell-matrix interaction, cell migration, and cancer- related pathways (Fig. 1C). The downregulated DUX4-r core gene set is instead mostly involved in pre-B cell receptor activity, B-cell activation, as well as immune-related processes (Fig. 1C).

[0347] Using the previously reported wt DUX4-associated gene expression signature (23), opposite results were obtained: a significant overlap for only our wt DUX4 dataset (Fig. 7F), matching previously reported (23) GSEA pathways including regulation of transcription, RNA processing, and apoptosis (Fig. 7G).

[0348] Previous studies demonstrated that wt DUX4 activates the expression of several classes of repetitive elements (REs) including endogenous retroviruses (ERVs) and pericentromeric human satellite II (HSATII) repeats (10, 12, 21, 24, 25). To investigate the activity of DUX4-r variants in the context of the REs genome, inventors re-analyzed their RNA-seq results and performed DGE analysis focused on the REs transcriptome. In line with previous reports (10, 12, 21, 24, 26), they found that wt DUX4 activates the expression of hundreds of REs (Fig. ID). Instead, the REs transcriptome of both DUX4-r variants mostly resembles that of control EV cells (Fig. ID). Hence, contrary to wt DUX4, DUX4-r variants minimally affect the REs transcriptome.

[0349] To identify direct transcriptional targets of DUX4-r variants, the chromatin occupancy of wt DUX4 and DUX4-r variants was mapped by performing Cleavage Under Targets and Tagmentation (CUT&Tag) (27) using an antibody recognizing a DUX4 epitope retained in DUX4-r. It was found that wt DUX4 and its rearranged versions associate to largely non-overlapping genomic regions, while DUX4-IGH and DUX4-del50 mostly bind to the same genomic loci (Fig. IE). This result is surprising considering that all DUX4-r variants described to date maintain the DNA-bmding domain of wt DUX4 (5, 6, 9). Wt DUX4 recognizes the consensus motif 5'-TAAYBBAATCA-3' (SEQ ID NO: 19) (24, 28). Strikingly, de novo motif identification analyses indicated that for both wt DUX4 and DUX4-r variants the most enriched sequence motif under the CUT&Tag peaks was the DUX4 consensus (Fig. IF).

[0350] Wt DUX4 functions as a pioneer transcription factor, able to bind condensed chromatin and increase its accessibility, by recruiting the histone acetyltransferases CBP and EP300 via its CTD to promote histone H3 lysine 27 acetylation (H3K27Ac) (13). It was evaluated H3K27Ac by CUT&Tag and chromatin accessibility by Cleavage Under Targeted Accessible Chromatin (CUT AC) (29). In line with previously reported results (10, 11, 13), wt DUX4 induction was associated to significantly increased H3K27Ac and chromatin accessibility of its direct targets. In contrast, DUX4-IGH induction did not cause significant changes in H3K27Ac or chromatin accessibility on DUX4-r direct targets (Fig. 1G- H). Intriguingly, most wt DUX4 direct targets were not expressed in its absence, while nearly half of DUX4-r direct targets were already expressed, albeit at lower levels, also in its absence (Fig. 7H). Hence, DUX4-r fails to bind and activate repressed chromatin.

[0351] Altogether, these results indicate that wt DUX4 and DUX4-r variants regulate mutually exclusive gene sets despite displaying the same DNA-binding domain and sequence specificity, with only wt DUX4 behaving as a pioneer factor. Notably, only DUX4-r is capable of activating the expression of genes involved in leukemia, with an extensive overlap with the DUX4-r B-ALL patient gene signature.

[0352] DUX4-r activates cell adhesion and migration, promoting cell proliferation.

[0353] Several direct DUX4-r target genes (27%) encode for surface proteins involved in cell adhesion or migration. Adhesion molecules are critical regulators of leukemia development and chemoresistance, being involved in cell survival, differentiation, proliferation, and migration (30, 31). Bone marrow (BM) is considered a protective niche to ALL cells (32), which contributes to chemotherapy resistance (33). To directly address the functional relevance of our findings, the inventors tested the adhesive and migratory properties of DUX4-r inducible REH cells using human bone marrow stroma cells (HS-5). To facilitate cell tracking, REH cell were modified by introducing a constitutively expressed eGFP reporter (Fig. 2A). As shown in Fig. 2B-C, induction of both DUX4-r variants significantly increased REH cell adhesion and migration to HS-5 stromal cells as compared to the uninduced control. Adhesion and migration were unaffected by inducing EV or wt DUX4 (Fig. 2B-C). Accordingly, knockdown of endogenous DUX4-r in NALM6 cells, a B-ALL cell line that expresses DUX4-r and depends on it for its proliferation (5), decreased cell adhesion and migration as compared to control knockdown cells (Fig. 2D-E), strengthening the conclusion that DUX4-r promotes cell adhesion and migration.

[0354] Tumor spheroids are a useful tool for preclimcal cancer research (34). Usually, spheroid formation is induced by the use of extracellular matrix / scaffold components and / or treatments forcing the cells to aggregate. Surprisingly, within the present invention it was found that DUX4-r expressing REH cells migrate toward each other and perform homotypic adhesion, generating areas of high cellular density which result in the spontaneous formation of spheroids in liquid cell culture (Fig. 8A). Real time quantification of live cells demonstrated significantly enhanced tumor spheroid formation by DUX4- IGH and DUX4-del50 expressing cells compared to EV or wt DUX4 (Fig. 2F). Importantly, tumor spheroid formation by NALM6 cells was significantly reduced upon knockdown of endogenous DUX4- r (Fig. 2G and Fig. 8B).

[0355] Importantly, when grown in low serum (5% FBS), DUX4-r expressing cells proliferated significantly more than EV or wt DUX4 cells (Fig. 2H). Accordingly, proliferation of NALM6 cells was significantly reduced by knockdown of endogenous DUX4-r (Fig. 21) in agreement with (5).

[0356] Collectively, these results indicate that DUX4-r directly activates the expression of cell-adhesion molecules, mediating the ability of DUX4-r expressing cells to migrate, perform homotypic adhesion, and proliferate.

[0357] DUX4-r transcriptional activity is prominent in a B-cell setting.

[0358] To validate RNA-seq results, representative DUX4-r direct targets consistently upregulated in DUX4- r+ B-ALL patients were selected. RT-qPCR analysis using RNA extracted from induced REH cells confirmed the ability of DUX4-r and wt DUX4 to activate mutually exclusive target genes (Fig. 3A). For several genes, the milder transactivation ability of DUX4-del50 compared to DUX4-IGH was also confirmed (Fig. 3A). To further characterize the transcriptional activity of DUX4-r variants, we generated Dox-inducible human embryonic kidney 293 (HEK) cells with EV, wt DUX4, DUX4-IGH or DUX4-del50. RT-qPCR analysis revealed that, despite similar transgene induction levels, DUX4-r variants were unable to activate the expression of their direct targets in HEK cells, while wt DUX4 efficiently drove expression of its direct target genes (Fig. 3B). Similar outcomes were also obtained using human T-ALL Jurkat cells (Fig. 3C).

[0359] To begin exploring the molecular determinants of DUX4-r transcriptional inefficiency in non-B cells, we performed CUT&Tag for wt DUX4 and DUX4-IGH in HEK cells. In line with its comparable transcriptional proficiency in different cellular settings, wt DUX4 was associated to its direct targets similarly in HEK and REH cells (Fig. 3D). Instead, DUX4-IGH association to its target genes was significantly higher in REH compared to HEK cells (Fig. 3D).

[0360] Overall, these results indicate that DUX4-r genomic association and transactivation ability are prominent in B cells.

[0361] GTF2I is a DUX4-r selective interactor, which associates to DUX4-r direct targets.

[0362] Given the above results, inventors hypothesized that the activity of DUX4-r requires a cofactor preferentially expressed in REH cells. To test this, they took advantage of the fact that the inducible DUX4 versions are fused to a streptavidin-binding peptide and a hemagglutinin (Strep-HA) double affinity tag. To identify factors selectively associated with DUX4-r variants, tandem affinity purification was performed using nuclease-treated and pre-cleared nuclear extracts under high stringency followed by quantitative mass spectrometry (TAP-MS), to reduce nonspecific background binding and identify tight interactors. This resulted in the identification of 70 proteins specifically interacting with wt DUX4 and / or DUX4-r. Of importance, wt DUX4 and DUX4-r signals were comparable across the different TAP-MS, indicating that the differential protein association with wt DUX4 or DUX4-r variants was not due to different efficiency of wt DUX4 or DUX4-r affinity purification (Fig. 9A). In order to determine selective DUX4-r interactors preferentially expressed in REH cells, the nuclear proteomes of REH, HEK and Jurkat cells were analyzed by quantitative mass spectrometry. Subsequently results were filtered based on: I. enrichment of DUX4-r associated proteins over the EV control; II. enrichment of proteins in DUX4-r over wt DUX4; III. DUX4-r interactors enrichment in REH over HEK and Jurkat cells. This approach identified 12 proteins selectively interacting with DUX4-IGH and DUX4-del50 and more abundant in REH cells (Fig. 9B). In line with the fact that the DUX4 CTD is deleted in DUX4- r variants, the histone acetyltransferase EP300 was identified as wt DUX4 selective interactor (Fig. 9C), thus providing a possible molecular explanation for DUX4-r lack of pioneer activity (Fig. 1G-H). Among the selective DUX4-r mteractors preferentially expressed in REH cells, the most enriched in both DUX4-IGH and DUX4-del50 TAP-MS was General Transcription Factor 21 (GTF2I / TFII-I) (Fig. 4A). GTF2I encodes for a transcription factor regulated by several signaling pathways, including the pre-B cell receptor (35, 36). Intriguingly, GTF2I translocations are associated to ALL and acute promyelocytic leukemia (37, 38). Moreover, a recurrent GTF2I mutation is the most frequent oncogenic driver in thymic epithelial tumors (39).

[0363] New affinity purifications followed by immunoblotting confirmed GTF2I selective interaction with both DUX4-r variants, but not wt DUX4 (Fig. 4B). Moreover, immunoblotting of total protein extracts confirmed significantly higher levels of GTF21 in REH compared to HEK and Jurkat cells (Fig. 4C-D). Importantly, co-immunoprecipitation (Co-IP) confirmed the interaction between the endogenous DUX4-r and GTF2I in NALM6 cells (Fig. 9D).

[0364] CUT&Tag was performed for GTF2I in HEK and REH cells lacking expression of DUX4-r, and in REH cells expressing DUX4-IGH. Intriguingly, for direct DUX4-r targets a significantly higher GTF2I signal was observed in REH compared to HEK cells, which was further increased upon DUX4-IGH expression (Fig. 4E). Importantly, no significant difference in GTF2I chromatin association was observed in REH cells expressing DUX4-IGH as compared to EV expressing REH cells, indicating that DUX4-IGH expression enhances GTF2I binding selectively to DUX4-r target genes (Fig. (3E).

[0365] Intrigued by the correlation between GTF2I expression levels and DUX4-r transcription activation ability in cell lines, DUX4-r B-ALL patient datasets (6) were re-analyzed to determine the expression levels of GTF2I and the level of activation of the DUX4-r core gene set in each patient. Despite all patients analyzed overexpress the DUX4-r core gene set compared to healthy B cells or other ALL subtypes (6), they do so at variable levels (Fig. 4F). Strikingly, a significant and positive correlation between the expression levels of GTF2I and the relative activation of DUX4-r core genes was found (Fig. 4G) strongly supporting the relevance of above findings.

[0366] Taken together, the above results identify GTF2I as the strongest, selective DUX4-r interactor, which discriminates DUX4-r targets from non-target genomic regions and whose expression levels significantly correlate with the level of activation of DUX4-r targets in B-ALL patients.

[0367] DUX4-r biological activity depends on the cellular availability of GTF2I.

[0368] Given the above results, it was investigated whether genetic GTF2I targeting could impact the ability of DUX4-r to activate its direct target genes. To test this, GTF2I knockdown was performed in REH cells expressing ectopic DUX4-IGH or in NALM6 expressing endogenous DUX4-r (Fig. 5A). In line with a GTF2I requirement for DUX4-r transcriptional activity, significant downregulation of DUX4-r target genes was observed (Fig. 5B). Importantly, GTF2I silencing had no significant effect on the activation of wt DUX4 targets (Fig. 10A).

[0369] Next, it was investigated whether GTF2I is required for the cellular phenotypes induced by DUX4-r. GTF2I knockdown significantly reduced both the adhesion and migration of DUX4-r expressing B- ALL cells (Fig. 5C and Fig. 10B), leading to a significantly decreased tumor spheroid formation selectively in DUX4-r expressing cells (Fig. 5D and Fig. 10C-E). Importantly, GTF2I genetic downregulation significantly reduced the proliferation of REH expressing ectopic DUX4-IGH (Fig. 5E) or NALM6 expressing endogenous DUX4-r (Fig. 10F), significantly increasing their cell death (Fig. 5F).

[0370] The above results indicate that GTF2I is required for DUX4-r activity. To evaluate sufficiency, the effects of supplementing HEK cells with GTF2I were then investigated. Vectors encoding for GTF2I or a control empty vector (EV) were introduced in HEK cells together with EV, wt DUX4 or DUX4- IGH. GTF2I was expressed at comparable levels in all cell lines and did not affect wt DUX4 or DUX4- IGH expression levels (Fig. 5G). Notably, in HEK cells, ectopically expressed GTF2I reached levels comparable to endogenous GTF2I in REH cells (Fig. 10G). GTF2I expression in HEK cells significantly increased target gene activation by DUX4-IGH (Fig. 5G), while wt DUX4 activity was unaffected by GTF2I supplementation (Fig. 10H).

[0371] To investigate the molecular determinants of the observed DUX4-r transcriptional “rescue” by GTF2I, CUT&Tag was performed in HEK cells expressing DUX4-IGH in presence / absence of transfected GTF2I. In the absence of GTF2I, DUX4-IGH displayed negligent chromatin association genome-wide (Fig. 101). Instead, upon GTF2I supplementation DUX4-IGH displayed a significantly enhanced ability to associate to its target gene loci in HEK cells (Fig. 5H). In line with the identification of GTF2I as a selective DUX4-r interactor, wt DUX4 association to its target loci was unaffected by GTF2I overexpression (Fig. 10J).

[0372] Given the above results, the implications of the DUX4-r / GTF2I transcriptional axis in leukemia were explored. To this aim, it was used a modified version of NALM6 cells expressing Gaussia luciferase Lucia, which is actively secreted by cells and detectable in the blood facilitating tumor progression monitoring in mice (40). NALM6-Lucia where transduced with lentiviral vectors encoding for a shRNA against GTF2I or a control shRNA (shCTRL). NSG mice were then transplanted with knockdown cells and disease progression was followed over time. One week after transplantation, we observed a similar luciferase signal between groups, indicating a comparable leukemia cell engraftment (Fig. 51). Nonetheless, after 2 weeks, we detected significantly lower leukemia cell expansion in peripheral blood of GTF2I knockdown animals compared to controls (Fig. 51). This translated into significantly higher survival rates in mice transplanted with GTF21 shRNA compared to control shRNA leukemia cells (Fig. 5J). At sacrifice, GTF2I knockdown transplants showed also a significantly lower leukemia burden in both BM and spleen compartments (Fig. 10K). All in all, the above results indicate that GTF2I is required for DUX4-r transcriptional and leukemogenic activities.

[0373] GTF2I pharmacological targeting inhibits DUX4-r leukemogenic activity.

[0374] Recently, a high-throughput screening identified the histone deacetylase inhibitor vorinostat / suberylanilide hydroxamic acid (SAHA) as a selective repressor of GTF2I expression (41). By performing a SAHA dose-response analysis in REH cells, it was found that treatment with as low as 50nM SAHA was sufficient to downregulate GTF2I consistently and significantly at both mRNA and protein levels (Fig. 11 A). Similarly to GTF2I knockdown, SAHA treatment significantly reduced the ability of DUX4-IGH to activate its target genes (Fig. 6A). Importantly, SAHA caused no significant effect on the ability of wt DUX4 to activate its targets (Fig. 1 IB). In line with a GTF2I requirement for DUX4-r genomic association, by CUT&Tag a significantly lower DUX4-IGH enrichment at its direct targets in SAHA-treated was found compared to control-treated cells (Fig. 6B).

[0375] We next assessed the impact of SAHA treatment on the cellular phenotypes caused by DUX4-r. Similar to what was observed with DUX4-r or GTF2I genetic silencing, SAHA significantly reduced cell adhesion, transwell migration, tumor spheroid formation and cell proliferation of both REH cells expressing ectopic DUX4-IGH and NALM6 cells expressing endogenous DUX4-r (Fig. 6C and Fig. 11C-E). SAHA treatment also significantly reduced tumor spheroid formation and proliferation of DUX4-IGH REH cells (Fig. 6D-E, Fig. 10). Consistently, SAHA significantly increased cell death both of REH expressing ectopic DUX4-IGH and NALM6 expressing endogenous DUX4-r (Fig. 6F). Importantly, proliferation and survival of control EV REH cells and DUX4-expressing REH cells were unaffected by SAHA treatment (Fig. 6E-F, Fig. 1 IF), in line with the possibility that DUX4-r expression sensitizes leukemia cells to SAHA treatment.

[0376] To test the therapeutic potential of SAHA treatment for DUX4-r+ B-ALL, NALM6-Lucia cells were transplanted in NSG mice. One week after transplantation, when comparable NALM6 engraftment was observed in all mice (Fig. 6G), animals were randomly divided in two cohorts: one was treated with SAHA and the other with DMSO, as control. After 1 week (two weeks from cell transplantation), a significantly lower leukemia cell expansion was detected in peripheral blood of SAHA-treated compared to control mice (Fig. 6G). This translated in significantly higher survival rates in mice treated with SAHA compared to controls (Fig. 6H).

[0377] Collectively, the above results indicate that the DUX4-r / GTF2I circuit can be targeted pharmacologically as a potential therapeutic option tailored to the DUX4-r+ B-ALL subtype. SAHA is inhibiting the activity of several HDACs, in particular HDAC1, HDAC2, HDAC3, HDAC6 and HDAC8. To determine if a specific HDAC is required for the expression of GTF2I, cells were treated with compounds selectively inhibiting HDAC1 / 2, HDAC3, HDAC6 or HDAC8. Only HDAC6- specific inhibitors recapitulated the ability to downregulate GTF2I expression at both RNA and protein levels. In particular, as shown in Fig. 12A-B, the HDAC6-selective inhibitor MPT0G211 was found more potent than SAHA in decreasing GTF2I expression and in blocking DUX4-r activity (analyzed at the level of tumor spheroid formation). Similarly, the HDAC6-selective inhibitor Tubastatin-A decreased the expression of GTF2I and inhibited the activity of DUX4-r (analyzed at the level of the expression of DUX4-r targets) (Fig. 13).

[0378] Since HDACs reduce the level of GTF2I RNA, a possible hypothesis is that they act at transcriptional level. To test this, CUTAC that analyzes the accessibility of chromatin was performed. As shown in Fig. 14, panel A, a reduced accessibility of the GTF2I locus was found in REH cells treated with SAHA as compared to the DMSO control indicating that SAHA regulates GTF2I at the transcriptional level. This suggested that SAHA and thus HDAC6 could regulate the expression or activity of a transcription factor required for GTF2I expression. To test this hypothesis, bioinformatic analyses were performed by crossing: i. publicly available ChlP-seq datasets showing transcription factor (TF) association to the GTF2I locus; ii. RNA-seq datasets showing altered GTF2I expression following TF manipulations; iii. putative TF binding sites located in CUTAC peaks that was identified in the GTF2I locus; iv. published reports of TF whose expression or activity is inhibited by HDAC6i. These analyses identified three candidates: c-Myc, STAT3 and SMAD3. To determine which of these were required for GTF2I expression in the experimental systems the three candidates were silenced. Intriguingly, only c-Myc silencing caused a significant decrease in GTF2I expression. Notably, chromatin immunoprecipitation followed by sequencing (ChlP-seq) databases show that the MY C / MAX is the most enriched sequence- specific transcription factor in the GTF2I locus. In accordance with these results, published reports show that HDAC6 is required for c-Myc stability and HDAC6 inhibitors cause c-Myc protein degradation (PMID: 36068335). Finally, the expression of c-Myc and GTF2I is significantly and positively correlated in DUX4-r positive B-ALL patients. Altogether, these results indicate thatHDAC6 activates GTF2I expression by stabilizing c-Myc protein levels (Fig. 14 B-C).

[0379] Curiously, it was not only found that c-Myc knockdown decreases GTF2I levels, but also that GTF2I knockdown reduces c-Myc. Hence, these data indicate that GTF2I and c-MYC regulate each other expression (Fig. 15).

[0380] GTF2I levels are positively correlated with DUX4-r leukemia aggressiveness In DUX4-r+ B-ALL patients, the inventors recently found that the expression level of GTF2I is positively correlated to the level of activation of the DUX4-r oncogenic signature. Moreover, it was found that genetic or pharmacologic reduction of GTF2I levels reduce the aggressiveness of the DUX4- r+ B-ALL cell line NALM6 (Campolungo,D., Salome, M., Biferali,B., Tascini,A.S. and Gabellini,D. (2023) DUX4-r exerts a neomorphic activity that depends on GTF2I in acute lymphoblastic leukemia. Sci Adv, 9).

[0381] Patient-derived xenografts models of B-ALL faithfully reflect somatic mutations, copy-number alterations, RNA expression, gene fusions, whole-genome methylation patterns, and immunophenotypes found in primary tumor samples (Rogojina,A., KI esse, L. J., Butler, E., Kim, J., Zhang, H., Xiao,X., Guo,L., Zhou,Q., Hartshorne, T., Garcia, D., et al. (2023) Comprehensive characterization of patient-derived xenograft models of pediatric leukemia. iScience, 26). To evaluate the biological relevance of these findings, DUX4-r+ B-ALL PDXs expressing comparable DUX4-r levels but relatively high or low GTF2I levels were obtained (Campolungo,D., Salome, M., Biferali,B., Tascini,A.S. and Gabellini,D. (2023) DUX4-r exerts a neomorphic activity that depends on GTF2I in acute lymphoblastic leukemia. Sci Adv, 9). Equal cell number of each PDX was transplanted in immunocompromised NSG mice and leukemia development was followed over time. Intriguingly, the inventors found a significantly faster accumulation of leukemia cells in the peripheral blood (PB) of mice transplanted with GTF2I high compared to GTF2I low DUX4-r+ B-ALL PDXs. Accordingly, a significantly faster deterioration of animal health status requiring sacrifice was found in GTF2I high compared to GTF2I low PDX transplanted animals. In particular, all animals transplanted with DUX4- r+ PDXs expressing relatively high GTF2I levels had to be sacrificed by 98 days after transplantation, while the majority of DUX4-r+ PDXs expressing relatively low GTF2I levels where still alive and healthy at the end of the experiment, 8 months after the transplant (Figure 16).

[0382] Hence, these results demonstrate that the expression level of GTF2I is an important determinant of DUX4-r+ PDX aggressiveness.

[0383] Pharmacological targeting of HDAC6 blocks DUX4-r activity

[0384] Given that the expression level of GTF2I is a critical determinant of DUX4-r oncogenic activity, characterizing the molecular mechanism driving GTF2I expression in leukemia might provide therapeutic opportunities. Given that the histone deacetylase HDAC6 is required for GTF2I expression, the inventors found that silencing of HDAC6 in a cell line derived from a DUX4-r patient significantly decreases expression of GTF2I and consequently blocks the expression of direct DUX4-r target genes (Figure 17A). Similarly, treatment with the HDAC6-selective inhibitor MPT0G211 (which will be called HDAC6i for simplicity from here onward) also caused a significant decrease in the expression of DUX4-r targets (Figure 17B).

[0385] Importantly, treatment with HDAC6i significantly inhibited the biological activity of DUX4-r (Figure 18). Indeed, HDAC6i inhibited adhesion, migration, tumor spheroid formation and proliferation leading to apoptosis of a cell line derived from a DUX4-r patient (Figure 18A-D). Importantly, a significantly higher cell death following HDAC6i treatment was found in DUX4-r positive compared to DUX4-r negative cells (Figure 18E). Hence, DUX4-r expression sensitizes cells to HDAC6 targeting.

[0386] Characterization of the molecular mechanism through which HDAC6 activates GTF2I expression Given that the transcription factor c-Myc is required for the transcriptional activation of GTF2I expression downstream of HDAC6, to determine if c-Myc regulates GTF2I directly, CUT&Tag was performed using control antibodies or antibodies specific for c-Myc. As shown in Figure 19A, the inventors found that c-Myc is significantly enriched on the GTF2I genomic locus in line with the possibility that c-Myc directly activates GTF2I transcription downstream of HDAC6.

[0387] It was previously shown that c-Myc is a direct HDAC6 substrate and that HDAC6 inhibition or its downregulation causes hyperacetylation of c-Myc lysine 148 (MycK148Ac) leading to c-Myc degradation mediated by the proteasome (Winkler,R, Magdefrau,A.S., Piskor,E.M., Kleemann,M., Beyer,M., Linke, K., Hansen, L., Schaffer, A.M., Hoffmann, M.E., Poepsel,S., et al. (2022) Targeting the MYC interaction network in B-cell lymphoma via histone deacetylase 6 inhibition. Oncogene, 41, 4560-4572; Zeleke,T.Z., Pan,Q., Chiuzan,C., Onishi,M., Li,Y., Tan,H., Alvarez, M.J., Honan, E., Yang,M., Chia,P.L., et al. (2023) Network-based assessment of HDAC6 activity predicts preclinical and clinical responses to the HDAC6 inhibitor ricolinostat in breast cancer. Nat Cancer, 4, 257-275). Notably, the inventors found that treatment of a cell line derived from a DUX4-r patient with HDAC6i induces hyper-acetylation of c-Myc lysine 148, decreasing total c-Myc protein levels, which is associated to decreased GTF2I expression (Figure 19B).

[0388] Collectively, these results indicate that HDAC6 activates the transcription of the GTF2I gene by de- acetylating and stabilizing the c-Myc protein. c-Myc is required for DUX4-r activity

[0389] Using publicly available datasets (Gu,Z., Churchman, M L., Roberts,K.G, Moore,I., Zhou,X., Nakitandwe,J., Hagiwara,K., Pelletier, S., Gingras, S., Berns, H., et al. (2019) PAX5-driven subtypes of B-progenitor acute lymphoblastic leukemia. Nat Genet, 51, 296-307), the inventors found that HDAC6, c-Myc and GTF2I are all highly significantly overexpressed in DUX4-r+ B-ALL patients compared to healthy donors (Figure 20A), suggesting that targeting this pathway could have therapeutic relevance for this leukemia subtype. Notably, analysis of the Cancer Dependency Map database (https: / / depmap.org / portal / ) revealed a significantly higher GTF2I survival dependency in human lymphoid malignancies compared to other cancer cell types (Figure 20B), highlighting a potentially broader involvement of GTF2I in lymphoid cancers.

[0390] The small molecule inhibitor MYCi975 (MYCi from now onward) is a potent and selective inhibitor of c-Myc that disrupts MYC / MAX interaction, promotes c-Myc degradation, impairs c-Myc driven gene expression, is well tolerated and exhibits potent anti-tumor activities in vivo (Holmes, A. G, Parker, J.B., Sagar, V., Truica,M.I., Som,P.N., Han,H., Schiltz,G.E., Abdulkadir,S.A. and Chakravarti,D. (2022) A MYC inhibitor selectively alters the MYC and MAX cistromes and modulates the epigenomic landscape to regulate target gene expression. Sci Adv, 8; Han,H., Jain,A.D., Truica,M.I., Izquierdo- Ferrer,J., Anker, J.F., Lysy,B., Sagar, V., Luan,Y., Chalmers, Z.R., Unno,K., et al. (2019) Small- Molecule MYC Inhibitors Suppress Tumor Growth and Enhance Immunotherapy. Cancer Cell, 36, 483- 497. el 5). Intriguingly, the inventors found that MYCi significantly reduced GTF2I levels and significantly decreased the expression of direct DUX4-r target genes, leading to apoptosis of a cell line derived from a DUX4-r patient (Figure 21 A-D).

[0391] Hence, HDAC6 stabilizes c-Myc protein, which activates GTF2I expression that is required for the oncogenic activity of DUX4-r.

[0392] DUX4-r leukemia cells are specifically sensitive to Birinapant

[0393] Protein inhibitors of apoptosis are frequently involved in anti-cancer drug resistance. Hence, drugs antagonizing the activity of inhibitors of apoptosis could make cancer cells more sensitive to anti-cancer drugs.

[0394] To determine if DUX4-r positive leukemia cells display sensitivity to inhibitors of apoptosis antagonists, a focused small molecule screening was performed by treating a cell line derived from a DUX4-r patient with antagonists of Survivin (Sepantronium bromide), MCL1 (S63845), cIAPl and XIAP (Birinapant), Bcl-xL / Bcl-2 / Bcl-w (Navitoclax) orBcl-2 (Venetoclax). Intriguingly, the only drug able to increase apoptosis in DUX4-r expressing cells was Birinapant (Figure 22A). Accordingly, it was found that the Birinapant targets cIAPl and XIAP are significantly overexpressed in leukemia cells derived from DUX4-r patients compared to control cells (Figure 22B).

[0395] Identification of drug combinations killing preferentially leukemia cells expressing DUX4-r

[0396] Having identified drugs preferentially killing DUX4-r expressing cells, it was investigated if their combination could lead to synergistic effects. Notably, the inventors found that the combination of Birinapant with HDAC6i or MYCi leads to a significant increase in DUX4-r positive cell death (Figure 23A-B). Importantly, this synergy was present selectively in DUX4-r positive cells.

[0397] Thus, leukemia cells expressing DUX4-r are exquisitely sensitive to inhibitors of HD AC6, c-Myc and cIAPl / XIAP, and their combination produces synergistic effects allowing to use the drugs at very low concentrations.

[0398] Identification of the DUX4-r / GTF2I interaction domains

[0399] Given that GTF2I is an essential DUX4-r leukemogenic partner (Campolungo,D., Salome, M., Biferah,B., Tascim,A.S. and Gabellini,D. (2023) DUX4-r exerts a neomorphic activity that depends on GTF2I in acute lymphoblastic leukemia. Sci Adv, 9), characterizing the molecular determinants involved in their interaction could help designing targeted treatments.

[0400] The minimal DUX4 portion reported in B-ALL patients contains the first 228 amino acids (Zhang, J., McCastlain,K., Yoshihara,H., Xu,B., Chang, Y., Churchman, M.L., Wu,G, Li,Y., Wei,L., lacobucci, l., et al. (2016) Deregulation of DUX4 and ERG in acute lymphoblastic leukemia. Nat Genet, 48, 1481— 1489). Accordingly, the inventors found that a DUX4-r variant containing ammo acids 2-230 (DUX4- r2-230) activates direct DUX4-r targets genes while it cannot activate direct wt DUX4 targets. Intriguingly, a fragment containing just amino acids 2-160 (DUX4-r2-i6o) is unable to activate DUX4-r targets (Figure 24 and data not shown). Given that the interaction with GTF2I is required for DUX4-r transcriptional activity and that DUX4- r2-230is active while DUX4-r2-160is inactive, it is tempting to speculate that the domain required for GTF2I interaction is located between DUX4-r amino acids 161 and 230.

[0401] There are four GTF2I isoforms generated by alternative splicing resulting from skipping cassette exons 10 and / or 12. Isoform α lacks exon 12, whereas exon 10 is missing in isoform β. Isoform A lacks both exons, while isoform γ maintains both exons (Figure 25A; Cheriyath,V. and Roy,A.L. (2000) Alternatively spliced isoforms of TFII-I. Complex formation, nuclear translocation, and differential gene regulation. J Biol Chem, 275, 26300-26308) . Intriguingly, only GTF2I β and γ was found to support DUX4-r transcriptional activity when supplemented to non-B cells (Figure 25B).

[0402] Next, to test the relevance of the endogenous GTF2I splicing isoforms, a leukemia cell line derived from a DUX4-r patient was transfected with a splice-switch antisense oligonucleotide (ssASO) to induce GTF2I exon 12 skipping (Figure 26A). The treatment reduced the abundance of GTF2I mRNA isoforms containing exon 12 while maintaining the total GTF2I mRNA output unchanged (Figure 26B). Intriguingly, the ssASO significantly reduced the expression of direct DUX4-r targets compared to the control (Figure 26C). Hence, these results strongly indicate that the GTF2I region encoded by exon 12, which is called domain B and is just 21 amino acids long (Cheriyath,V. and Roy,A.L. (2000) Alternatively spliced isoforms of TFII-I. Complex formation, nuclear translocation, and differential gene regulation. J Biol Chem, 275, 26300-26308), is required for GTF2I interaction with DUX4-r.

[0403] Collectively, these data demonstrate that a DUX4-r region of 70 ammo acids (DUX4-r161-230; GGLCSAAPGGGHPAPSWVAFAHTGAWGTGLPAPHVPCAPGALPQGAFVSQAARAAPALQPS QAAP AEGIS; SEQ ID No. 67) and a GTF2I region of 21 amino acids (GTF2I domain B; SEQ ID No. 66) are responsible for the DUX4-r / GTF2I interaction.

[0404] Delivery of GTF2I domain B inhibits DUX4-r activity

[0405] Compared to small-molecule drugs, peptides are highly selective and efficacious and, at the same time, relatively safe and well-tolerated.

[0406] Based on the above results, that the inventors investigated if GTF2I domain B could function as a dominant negative inhibitor of the DUX4-r / GTF2I interaction. To test this, GTF2I domain A (as negative control) or GTF2I domain B were delivered to a leukemia cell line derived from a DUX4-r patient and evaluated the expression of DUX4-r target genes by RT-qPCR. Strikingly, a decreased expression of DUX4-r targets was selectively found in cells treated with GTF2I domain B (Figure 27). Hence, the inventors identified a dominant negative GTF2I-derived peptide that can be delivered to B- ALL cells to block DUX4-r transcriptional activity.

[0407] In summary, the present invention provides drug combinations synergistically and selectively killing leukemia cells expressing DUX4-r, and a GTF2I-derived peptide inhibiting DUX4-r activity.

[0408] DISCUSSION

[0409] By combining transcriptomics, genomics, and proteomics with functional assays in cellular and animal leukemia models, the present invention provides a novel molecular mechanism of B-cell acute lymphoblastic leukemia caused by DUX4-r. The chromosomal rearrangement producing DUX4-r leads to both a loss and a gain of function in the leukemogenic transcription factor. Deletion of the C-terminal transcription activation domain causes a loss of interaction with CBP / EP300 transcriptional co- activators leading to inability to bind and activate repressed chromatin. On the other hand, DUX4-r gains the ability to interact with the transcription factor GTF2I, which redirects DUX4-r toward binding and activation of targets involved in leukemogenesis. wt DUX4 and DUX4-r bind to and activate highly different genes sets, even though they share the same DNA-binding domain and recognize the same DNA consensus. Present results differ from those of a previous study (9), which reported a large overlap between wt DUX4 and DUX4-r genomic targets. Those authors performed ChlP-seq only for DUX4-IGH in NALM6 cells and compared their result to a previously published ChlP-seq dataset of wt DUX4 performed by others in induced pluripotent stem cells (9). To perform a side-by-side comparison and enrich for primary targets, present inventors instead used REH cells lacking endogenous wt DUX4 / DUX4-r expression, induced wt DUX4 / DUX4-r at levels comparable to those of DUX4-r+ B-ALL patients and performed CUT&Tag early upon DUX4 / DUX4- r induction. The technical differences between the previously published and the herein study are probably responsible for the different results.

[0410] While wt DUX4 shows essentially the same transcriptional activity in every tested cell type (10-12, 21, 23, 24, 42, 43), the present invention found that the activity of DUX4-r is intimately linked to GTF2I cellular availability. In DUX4-r+ B-ALL patients, the magnitude of activation of the DUX4-r signature positively correlates with GTF2I expression levels. At genomic level, DUX4-r enrichment at its targets is significantly higher in REH compared to HEK cells, in line with GTF2I expression levels. Also, GTF2I downregulation leads to significantly lower DUX4-r chromatin association and target activation. Similarly, in HEK cells, which express relatively low GTF2I endogenous levels, DUX4-r displays negligent chromatin association genome-wide. Notably, ectopic GTF2I expression in HEK cells is sufficient to drive DUX4-r recruitment selectively to its target genes. It thus tempting to speculate that the DNA-binding activity of DUX4-r is not solely determined by its DNA binding domain, but that it rely on GTF2I for target gene definition.

[0411] In REH cells, it was found that GTF2I is already associated to DUX4-r targets in its absence. Hence, GTF2I pre-marking in the vicinity of DUX4 binding sites could contribute to DUX4-r target gene definition. Interestingly, these results suggest that also GTF2I is affected by DUX4-r. Indeed, GTF2I enrichment at DUX4-r targets is significantly enhanced upon DUX4-r expression. While the precise mechanism at the basis of the apparent synergy between DUX4-r and GTF2I remains to be determined, it is tempting to speculate that DUX4-r binding tethers GTF2I to target loci and / or that the two factors stabilize each other chromatin association.

[0412] Previous work reported that treatment with the HDAC inhibitor SAHA downregulates GTF2I expression (41). This result was confirmed and extended by showing that SAHA treatment in the nM range is sufficient to downregulate GTF2I at both mRNA and protein levels. Notably, SAHA blocks activation of DUX4-IGH targets and the downstream phenotypic effects as effectively as direct GTF2I or DUX4-r knockdown. Importantly, it was found that DUX4-r confers enhanced SAHA to expressing leukemia cells in line with the possibility that DUX4-r+ B-ALL cells are addicted to the DUX4-r / GTF2I circuit. Hence, these results provide a rationale for the development of a pharmacological treatment for DUX4-r+ B-ALL. While SAHA is already approved for the treatment of cutaneous T-cell lymphoma (44) and is in clinical trial for several other types of cancers (45, 46), further studies concerning the molecular mechanisms of GTF2I downregulation by SAHA could identify additional targets to develop therapeutic strategies personalized to DUX4-r+ B-ALL.

[0413] Overall, the results obtained in the present invention indicate that DUX4-r acquires a neomorphic oncogenic activity depending on GTF2I that can be targeted for therapeutic purposes. The invention demonstrates an increased vulnerability of DUX4-r positive compared to DUX4-r negative cells to the treatments provided, thus providing the possibility of selectively targeting DUX4-r+ cells. More specifically the invention provides drug combinations synergistically and selectively killing leukemia cells expressing DUX4-r and a GTF2I-derived peptide specifically inhibiting DUX4-r activity.

[0414] References

[0415] 1. S. P. Hunger, C. G. Mullighan, Acute lymphoblastic leukemia in children. New England Journal of Medicine. 373, 1541-1552 (2015).

[0416] 2. G. R. Bunin, E. J. Feuer, P. A. Witman, A. T. Meadows, Increasing incidence of childhood cancer: report of 20 years experience from the greater Delaware Valley Pediatric Tumor Registry. Paediatr Perinat Epidemiol. 10, 319-338 (1996).

[0417] 3. M. S. Linet, L. A. Ries, M. A. Smith, R. E. Tarone, S. S. Devesa, Cancer surveillance series: recent trends in childhood cancer incidence and mortality in the United States. J Natl Cancer Inst. 91, 1051-1058 (1999).

[0418] 4. I. lacobucci, S. Kimura, C. G. Mullighan, Biologic and therapeutic implications of genomic alterations in acute lymphoblastic leukemia. J Clin Med. 10 (2021), doi:10.3390 / jcml0173792.

[0419] 5. T. Yasuda, S. Tsuzuki, M. Kawazu, F. Hayakawa, S. Kojima, T. Ueno, N. Imoto, S. Kohsaka, A. Kunita, K. Doi, T. Sakura, T. Yujiri, E. Kondo, K. Fujimaki, Y. Ueda, Y. Aoyama,

[0420] 5. Ohtake, J. Takita, E. Sai, M. Taniwaki, M. Kurokawa, S. Morishita, M. Fukayama, H. Kiyoi, Y. Miyazaki, T. Naoe, H. Mano, Recurrent DUX4 fusions in B cell acute lymphoblastic leukemia of adolescents and young adults. Nat Genet. 48, 569-574 (2016).

[0421] 6. J. Zhang, K. McCastlain, H. Yoshihara, B. Xu, Y. Chang, M. L. Churchman, G. Wu, Y. Li, L. Wei, I. lacobucci, Y. Y. Y. Y. Liu, C. Qu, J. Wen, M. Edmonson, D. Payne-Turner, K. B. Kaufmann, S. I. Takayanagi, E. Wienholds, E. Waanders, P. Ntziachristos, S. Bakogianni, J. Wang, I. Aifantis, K. G. Roberts, J. Ma, G. Song, J. Easton, H. L. Mulder, X. Chen, S. Newman, X. Ma, M. Rusch, P. Gupta, K. Boggs, B. Vadodaria, J. Dalton, Y. Y. Y. Y. Liu, M. L. Valentine, L. Ding, C. Lu, R. S. Fulton, L. Fulton, Y. Tabib, K. Ochoa, M. Devidas, D. Pei, C. Cheng, J. Yang, W. E. Evans, M. V. Relhng, C.-H. H. Pui, S. Jeha, R. C. Harvey, I.-M. M. L. Chen, C. L. Willman, G. Marcucci, C. D. Bloomfield, J. Kohlschmidt, K. Mrozek, E. Paietta, M. S. Tailman, W. Stock, M. C. Foster, J. Racevskis, J. M. Rowe, S. Luger, S. M. Kornblau, S. A. Shurtleff, S. C. Raimondi, E. R. Mardis, R. K. Wilson, J. E. Dick, S. P. Hunger, M. L. Loh, J. R. Downing, C. G. Mullighan, Deregulation of DUX4 and ERG in acute lymphoblastic leukemia. Nat Genet. 48, 1481-1489 (2016).

[0422] 7. H. Lilljebjorn, R. Hennmgsson, A. Hyrenius-Wittsten, L Olsson, C. Orsmark-Pietras, S. Von Palffy, M. Askmyr, M. Rissler, M. Schrappe, G. Cario, A. Castor, C. J. H. H. Pronk, M. Behrendtz, F. Mitelman, B. Johansson, K. Paulsson, A. K. Andersson, M. Fontes, T. Fioretos, Identification of ETV6-RUNXl-like and DUX4-rearranged subtypes in paediatric B-cell precursor acute lymphoblastic leukaemia. Nat Commun. 7, 11790 (2016).

[0423] 8. Y. Marincevic-Zuniga, J. Dahlberg, S. Nilsson, A. Raine, S. Nystedt, C. M. Lindqvist, E. C. Berglund, J. Abrahamsson, L. Cavelier, E. Forestier, M. Heyman, G. Lonnerholm, J. Nordlund, A.-C. C. Syvanen, Transcriptome sequencing in pediatric acute lymphoblastic leukemia identifies fusion genes associated with distinct DNA methylation profiles. J Hematol Oncol. 10, 148 (2017).

[0424] 9. Yosuke Tanaka, Masahito Kawazu, Takahiko Yasuda, Miki Tamura, Fumihiko Hayakawa, Shinya Kojima, Toshihide Ueno, Hitoshi Kiyoi, Tomoki Naoe, Hiroyuki Mano, Y. Tanaka, M. Kawazu, T. Yasuda, M. Tamura, F. Hayakawa, S. Kojima, T. Ueno, H. Kiyoi, T. Naoe, H. Mano, Transcriptional activities of DUX4 fusions in B-cell acute lymphoblastic leukemia. Haematologica. 103, e522-e526 (2018).

[0425] 10. P. G. Hendrickson, J. A. Dorais, E. J. Grow, J. L. Whiddon, J. W. Lim, C. L. Wike, B. D. Weaver, C. Pflueger, B. R. Emery, A. L. Wilcox, D. A. Nix, C. M. Peterson, S. J. Tapscott, D. T. Carrell, B. R. Cairns, Conserved roles of mouse DUX and human DUX4 in activating cleavage- stage genes and MERVL / HERVL retrotransposons. Nat Genet. 49, 925-934 (2017).

[0426] 11. A. De laco, E. Planet, A. Coluccio, S. Verp, J. Due, D. Trono, DUX- family transcription factors regulate zygotic genome activation in placental mammals. Nat Genet. 49, 941-945 (2017).

[0427] 12. J. L. Whiddon, A. T. Langford, C. J. Wong, J. W. Zhong, S. J. Tapscott, Conservation and innovation in the DUX4-family gene network. Nat Genet. 49, 935-940 (2017).

[0428] 13. S. H. Choi, M. D. Gearhart, Z. Cui, D. Bosnakovski, M. Kim, N. Schennum, M. Kyba, DUX4 recruits p300 / CBP through its C-terminus and induces global H3K27 acetylation changes. Nucleic Acids Res. 44, 5161-5173 (2016).

[0429] 14. M. Carlet, K. Voise, J. Vergalli, M. Becker, T. Herold, A. Arner, D. Senft, V. Jurinovic,

[0430] W.-H. Liu, Y. Gao, V. Dill, B. Fehse, C. D. Baldus, L. Bastian, L. Lenk, D. M. Schewe, J. W. Bagnoli, B. Vick, J. P. Schmid, A. Wilhelm, R. Marschalek, P. J. Jost, C. Miething, K. Riecken, M. Schmidt- Supprian, V. Binder, I. Jeremias, In vivo inducible reverse genetics in patients’ tumors to identify individual therapeutic targets. Nat Commun. 12, 5655 (2021).

[0431] 15. L. Tian, Y. Shao, S. Nance, J. Dang, B. Xu, X. Ma, Y. Li, B. Ju, L. Dong, S. Newman,

[0432] X. Zhou, P. Schreiner, E. Tseng, T. Hon, M. Ashby, C. Li, J. Easton, T. A. Gruber, J. Zhang, Long-read sequencing unveils IGH-DUX4 translocation into the silenced IGH allele in B-cell acute lymphoblastic leukemia. Nat Commun. 10, 2789 (2019). 16. E. Clappier, M. F. Auclerc, J. Rapion, M. Bakkus, A. Caye, A. Khemiri, C. Giroux, L. Hernandez, E. Kabongo, S. Savoia, T. Leblanc, K. Yakouben, G. Plat, V. Costa, A. Ferster, S. Girard, O. Fenneteau, J. M. Cayuela, F. Sigaux, N. Dastugue, S. Suciu, Y. Benoit, Y. Bertrand, J. Soulier, H. Cave, An intragenic ERG deletion is a marker of an oncogenic subtype of B-cell precursor acute lymphoblastic leukemia with a favorable outcome despite frequent IKZF1 deletions. Leukemia. 28, 70-77 (2014).

[0433] 17. M. Zaliova, M. Kotrova, S. Bresolin, J. Stuchly, J. Stary, O. Hrusak, G. Te Kronnie, J. Trka, J. Zuna, M. Vaskova, ETV6 / RUNXl-like acute lymphoblastic leukemia: A novel B-cell precursor leukemia subtype associated with the CD27 / CD44 immunophenotype. Genes Chromosomes Cancer. 56, 608-616 (2017).

[0434] 18. C. G. Mullighan, C. B. Miller, X. Su, I. Radtke, J. Dalton, G. Song, X. Zhou, C.-H. Pui, S. A. Shurtleff, J. R. Downing, ERG Deletions Define a Novel Subtype of B-Progemtor Acute Lymphoblastic Leukemia. Blood. 110, 691 (2007).

[0435] 19. M. Zaliova, E. Potuckova, L. Hovorkova, A. Musilova, L. Winkowska, K. Fiser, J. Stuchly, E. Mejstrikova, J. Starkova, J. Zuna, J. Stary, J. Trka, Marketa Zaliova, Eliska Potuckova, Lenka Hovorkova, Alena Musilova, Lucie Winkowska, Karel Fiser, Jan Stuchly, Ester Mejstrikova, Julia Starkova, Jan Zuna, Jan Stary, Jan Trka, ERG deletions in childhood acute lymphoblastic leukemia with DUX4 rearrangements are mostly polyclonal, prognostically relevant and their detection rate strongly depends on screening method sensitivity. Haematologica. 104, 1407-1416 (2019).

[0436] 20. Y.-F. F. Liu, B.-Y. Y. Wang, W. G. W. N. W.-G. W.-N. W. G. Zhang, J.-Y. Y. Huang, B.-S. S. Li, M. Zhang, L. Jiang, J. F. J. M. J. F. J.-M. J.-F. Li, M.-J. J. Wang, Y.-J. J. Dai, Z.-G. G. Zhang, Q. Wang, J. Kong, B. Chen, Y.-M. M. Zhu, X.-Q. Q. Weng, Z.-X. X. Shen, J. F. J. M. J. F. J.-M. J.-F. Li, J. Wang, X.-J. J. Yan, Y. Y. Y. Li, Y.-M. M. Liang, L. Liu, X.-Q. Q. Chen, W. G. W. N. W.-G. W.-N. W. G. Zhang, J.-S. S. Yan, J.-D. Da Hu, S.-H. H. Shen, J. Chen, L.-J. J. Gu, D. Pei, Y. Y. Y. Li, G Wu, X. Zhou, R.-B. B. Ren, C. Cheng, J. J. Yang, K -K. K. Wang, S.-Y. Y. Wang, J. Zhang, J.-Q. Q. Mi, C.-H. H. Pui, J.-Y. Y. Tang, Z. Chen, S.-J. J. Chen, Genomic Profiling of Adult and Pediatric B-cell Acute Lymphoblastic Leukemia. EBioMedicine. 8, 173-183 (2016).

[0437] 21. J. M. Young, J. L. Whiddon, Z. Yao, B. Kasinathan, L. Snider, L. N. Geng, J. Balog, R. Tawil, S. M. van der Maarel, S. J. Tapscott, DUX4 Binding to Retroelements Creates Promoters That Are Active in FSHD Muscle and Testis. PLoS Genet. 9 (2013), doi : 10.1371 / j ournal. pgen. 1003947.

[0438] 22. X. Dong, W. Zhang, H. Wu, J. Huang, M. Zhang, P. Wang, H. Zhang, Z. Chen, S. J. Chen, G. Meng, Structural basis of DUX4 / IGH-driven transactivation. Leukemia. 32, 1466-1476 (2018).

[0439] 23. S. Jagannathan, S. C. Shadle, R. Resnick, L. Snider, R. N. Tawil, S. M. van der Maarel, R K. Bradley, S. J. Tapscott, Model systems of DUX4 expression recapitulate the transcriptional profile of FSHD cells. Hum Mol Genet. 25, 4419-4431 (2016).

[0440] 24. L N. Geng, Z. Yao, L. Snider, A P. Fong, J. N. Cech, J. M Young, S. M. vanderMaarel, W. L. Ruzzo, R. C. Gentleman, R. Tawil, S. J. Tapscott, S. M. van der Maarel, W. L. Ruzzo, R. C. Gentleman, R. Tawil, S. M. vanderMaarel, W. L. Ruzzo, R. C. Gentleman, R. Tawil, S. J. Tapscott, DUX4 Activates Germline Genes, Retroelements, and Immune Mediators: Implications for Facioscapulohumeral Dystrophy. Dev Cell. 22, 38-51 (2012).

[0441] 25. P. Dmitriev, Y. Bou Saada, C. Dib, E. Ansseau, A. Barat, A. Hamade, P. Dessen, T. Robert, V. Lazar, R. A. N. Louzada, C. Dupuy, V. Zakharova, G. Carnac, M. Lipinski, Y. S. Vassetzky, DUX4-induced constitutive DNA damage and oxidative stress contribute to aberrant differentiation of myoblasts from FSHD patients. Free Radic Biol Med. 99, 244-258 (2016).

[0442] 26. P. Dmitriev, E. Kiseleva, O. Kharchenko, E. Ivashkin, A. Pichugin, P. Dessen, T. Robert, F. Coppee, A. Belayew, G. Carnac, A. Vasiliev, Y. S. Y. S. Vassetzky, D. Laoudj-Chenivesse, M. Lipinski, A. Vasiliev, Y. S. Y. S. Vassetzky, Dux4 controls migration of mesenchymal stem cells through the Cxcr4-Sdfl axis. Oncotarget. 7, 65090-65108 (2016).

[0443] 27. H. S. Kaya-Okur, S. J. Wu, C. A. Codomo, E. S. Pledger, T. D. Bryson, J. G. Henikoff, K. Ahmad, S. Henikoff, CUT&Tag for efficient epigenomic profiling of small samples and single cells. Nat Commun. 10, 1930 (2019).

[0444] 28. Y. Zhang, J. K. Lee, E. A. Toso, J. S. Lee, S. H. Choi, M. Slattery, H. Aihara, M. Kyba, DNA-binding sequence specificity ofDUX4. Skelet Muscle. 6, 8 (2016).

[0445] 29. S. Henikoff, J. G. Henikoff, K. Ahmad, Simplified epigenome profiling using antibody- tethered tagmentation. Bio Protoc. 11 (2021), doi: 10.21769 / BioProtoc.4043

[0446] 30. I. C. C. F. E. Fonseca, F. A. C. da Luz, I. A. Uehara, M. J. B. Silva, Cell-adhesion molecules and their soluble forms: Promising predictors of “tumor progression” and relapse in leukemia. Tumour Biol. 40, 1010428318811525 (2018).

[0447] 31. B. F. S. S. Scharff, S. Modvig, H. V. Marquart, C. Christensen, Integrin-Mediated Adhesion and Chemoresistance of Acute Lymphoblastic Leukemia Cells Residing in the Bone Marrow or the Central Nervous System . Frontiers in Oncology . 10 (2020), p. 775.

[0448] 32. R. H. Scheuermann, E. Racila, CD19 Antigen in Leukemia and Lymphoma Diagnosis and Immunotherapy. Leuk Lymphoma. 18, 385-397 (1995).

[0449] 33. R. Jacamo, Y. Chen, Z. Wang, M. Wencai, M Zhang, E. L. Spaeth, Y. Wang, V. L. Battula, P. Y. Mak, K. Schallmoser, P. Ruvolo, W. D. Schober, E. J. Shpall, M. H. Nguyen, D. Strunk, C. E. Bueso-Ramos, S. Konoplev, R. E. Davis, M. Konopleva, M. Andreeff, Reciprocal leukemia-stroma VCAM-l / VLA-4-dependent activation of NF-ĸB mediates chemoresistance. Blood. 123, 2691-2702 (2014).

[0450] 34. T. Ishiguro, H. Ohata, A. Sato, K. Yamawaki, T. Enomoto, K. Okamoto, Tumor-derived spheroids: Relevance to cancer stem cells and clinical applications. Cancer Sci. 108, 283-289 (2017).

[0451] 35. T. Chailangkarn, C. Noree, A. R. Muotri, The contribution of GTF2I haploinsufficiency to Williams syndrome. Mol Cell Probes. 40, 45-51 (2018). 36. A. L. Roy, Pathophysiology of TFII-I: Old Guard Wearing New Hats. Trends Mol Med. 23, 501-511 (2017).

[0452] 37. J. Li, H.-Y. Zhong, Y. Zhang, L. Xiao, L.-H. Bai, S.-F. Liu, G.-B. Zhou, G.-S. Zhang, GTF2I-RARA is a novel fusion transcript in a t(7; 17) variant of acute promyelocytic leukaemia with clinical resistance to retinoic acid. Br J Haematol. 168 (2015), pp. 904-908.

[0453] 38. I. Panagopoulos, M. Brunetti, M. Stoltenberg, R. A. U. Strandabø, J Staurseth, K. Andersen, I. Kostolomov, T. S. Hveem, S. Lorenz, T. A. Nystad, T. Flæsgstad, F. Micci, S. Heim, Novel GTF2I-PDGFRB and IKZF1-TYW1 fusions in pediatric leukemia with normal karyotype. Exp Hematol Oncol. 8 (2019), p. 12.

[0454] 39. I. Petrini, P. S. Meltzer, I.-K. Kim, M. Lucchi, K.-S. Park, G. Fontanini, J. Gao, P. A. Zucali, F. Calabrese, A. Favaretto, F. Rea, J. Rodriguez-Canales, R. L. Walker, M. Pineda, Y. J. Zhu, C. Lau, K. J. Killian, S. Bilke, D. Voeller, S. Dakshanamurthy, Y. Wang, G. Giaccone, A specific missense mutation in GTF2I occurs at high frequency in thymic epithelial tumors. Nat Genet. 46, 844-849 (2014).

[0455] 40. S. Arcangeli, L. Falcone, B. Camisa, F. De Girardi, M. Biondi, F. Giglio, F. Ciceri, C. Bonini, A. Bondanza, M. Casucci, Next-Generation Manufacturing Protocols Enriching TSCM CAR T Cells Can Overcome Disease-Specific T Cell Defects in Cancer Patients. Front Immunol. 11 (2020), doi: 10.3389 / fimmu.2020.01217.

[0456] 41. F. Cavallo, F. Troglio, G. Faga, D. Fancelli, R. Shyti, S. Trattaro, M. Zanella, G. D’Agostino, J. M. Hughes, M. R. Cera, M. Pasi, M. Gabriele, M. Lazzarin, M. Mihailovich, F. Kooy, A. Rosa, C. Mercurio, M. Varasi, G. Testa, High-throughput screening identifies histone deacetylase inhibitors that modulate GTF2I expression in 7ql l.23 microduplication autism spectrum disorder patient-derived cortical neurons. Mol Autism. 11, 88 (2020).

[0457] 42. S. Mitsuhashi, S. Nakagawa, M. Sasaki-Honda, H. Sakurai, M. C. Frith, H. Mitsuhashi, Nanopore direct RNA sequencing detects DUX4-Activated repeats and isoforms in human muscle cells. Hum Mol Genet. 30, 552-563 (2021).

[0458] 43. Z. Yao, L. Snider, J. Balog, R. J. L. F. Lemmers, S. M. Van Der Maarel, R. Tawil, S. J. Tapscott, DUX4-induced gene expression is the major molecular signature in FSHD skeletal muscle. Hum Mol Genet. 23, 5342-5352 (2014).

[0459] 44. O. Khan, N. B. La Thangue, Drug Insight: Histone deacetylase inhibitor-based therapies for cutaneous T-cell lymphomas. Nat Clin Pract Oncol. 5 (2008), pp. 714-726.

[0460] 45. A. Wawruszak, L. Borkiewicz, E. Okon, W. Kukula-Koch, S. Afshan, M. Halasa, Vorinostat (SAHA) and breast cancer: An overview. Cancers (Basel). 13 (2021), doi: 10.3390 / cancersl 3184700.

[0461] 46. L. Sanchez, D. H. Vesole, J. R. Richter, N. Biran, E. Bilotti, L. McBride, P. Anand, K. Ivanovski, D. S. Siegel, A phase lib trial of vorinostat in combination with lenalidomide and dexamethasone in patients with multiple myeloma refractory to previous lenalidomide- containing regimens. Br J Haematol. 176, 440-447 (2017). 47. A. M. Bolger, M. Lohse, B. Usadel, Trimmomatic: a flexible trimmer for Illumina sequence data. Bioinformatics. 30, 2114-2120 (2014).

[0462] 48. A. Dobin, C. A. Davis, F. Schlesinger, J. Drenkow, C. Zaleski, S. Jha, P. Batut, M. Chaisson, T. R. Gingeras, STAR: ultrafast universal RNA-seq aligner. Biomformatics 29, 15- 21 (2013).

[0463] 49. Y. Liao, G. K. Smyth, W. Shi, featureCounts: an efficient general purpose program for assigning sequence reads to genomic features. Biomformatics. 30, 923-930 (2014).

[0464] 50. J. Costa-Silva, D. Domingues, F. M. Lopes, RNA-Seq differential expression analysis: An extended review and a software tool. PLoS One. 12, 1-18 (2017).

[0465] 51. G. R. Morgan M, Falcon S, GSEABase: Gene set enrichment data structures and methods. R package version 1.56.0 (2021).

[0466] 52. A. Subramanian, P. Tamayo, V. K. Mootha, S. Mukherjee, B. L. Ebert, M. A. Gillette, A. Paulovich, S. L. Pomeroy, T. R. Golub, E. S. Lander, J. P. Mesirov, Proceedings of the National Academy of Sciences, in press, doi: 10.1073 / pnas.0506580102.

[0467] 53. B. Langmead, S. L. Salzberg, Fast gapped-read alignment with Bowtie 2. Nat Methods. 9, 357-359 (2012).

[0468] 54. P. Danecek, J. K. Bonfield, J. Liddle, J. Marshall, V. Ohan, M. O. Pollard, A. Whitwham, T. Keane,

[0469] S. A. McCarthy, & R. M. Davies Twelve years of SAMtools and BCFtools. GigaScience, 10(2) (2021). https: / / doi.org / 10.1093 / gigascience / giab008

[0470] 55. Y. Zhang, T. Liu, C. A. Meyer, J. Eeckhoute, D. S. Johnson, B. E. Bernstein, C. Nussbaum, R. M Myers, M. Brown, W. Li, & X. S. Shirley Model-based analysis of ChlP-Seq (MACS). Genome Biology, 9(9). (2008). https: / / doi.org / 10.1186 / gb-2008-9-9-rl37

[0471] 56. F. Ramirez, D. P. Ryan, B. Griming, V. Bhardwaj, F. Kilpert, A. S. Richter, S. Heyne,

[0472] F. Dündar, & T. Manke deepTools2: a next generation web server for deep-sequencing data analysis. Nucleic Acids Research, 44(W1 ), W160-W165. (2016). https: / / doi. org / 10, 1093 / NAR / GKW257

[0473] 57. S. Heinz, C. Benner, N. Spann, E. Bertolino, Y. C. Lin, P. Laslo, J. X. Cheng, C. Murre, H. Smgh, & C. K. Glass. Simple Combinations of Lineage-Determining Transcription Factors Prime cis-Regulatory Elements Required for Macrophage and B Cell Identities. Molecular Cell, 38(4), 576-589. (2010). https: / / doi.Org / 10.1016 / j.molcel.2010.05.004

[0474] 58. J. Cox, N. Neuhauser, A. Michalski, R. A. Scheltema, J. V. Olsen, & M. Mann. (2011). Andromeda: A peptide search engine integrated into the MaxQuant environment. Journal of Proteome Research, 10(4), 1794-1805. https: / / doi.org / 10.1021 / prl01065j

[0475] 59. D. Mellacheruvu, Z. Wright, A. L. Couzens, J. P. Lambert, N. A. St-Denis, T. Li, Y. V Miteva, S. Hauri, M. E. Sardiu, T. Y. Low, V. A. Halim, R. D. Bagshaw, N. C. Hubner, A. Al- Hakim, A. Bouchard, D. Faubert, D. Fermin, W. H. Dunham, M. Goudreault, ... A. I. Nesvizhskii. The CRAPome: A contaminant repository for affinity purification-mass spectrometry data. Nature Methods, 10(8), 730-736. (2013). https: / / doi.org / 10.1038 / nmeth.2557

[0476] 60. D. Bausch-Fluck, U. Goldmann, S. Muller, M. van Oostrum, M Muller, O. T. Schubert, & B. Wollscheid. The in silico human surfaceome. Proceedings of the National Academy of

[0477] Sciences of the United States of America, 115(46), E10988-E10997. (2018). https: / / d0i.0rg / l 0, 1073 / pnas.1808790115

[0478] 61. RICHARD, G. (2019). gtrichard / deepStats: deepStats 0.3.1. https: / / doi.org / l 0.5281 / ZENODO.3361799

Claims

CLAIMS1. A molecule for use in the treatment and / or prevention of DUX4-r positive leukemia wherein said molecule is selected from the group consisting of: an inhibitor of GTF2I expression and / or function; an inhibitor of HD AC expression and / or function; and an inhibitor of c-Myc expression and / or function.

2. The molecule for use according to claim 1, wherein the DUX4-r positive leukemia is acute lymphoblastic leukemia, preferably B-cell acute lymphoblastic leukemia (B-ALL), preferably said acute lymphoblastic leukemia or B-cell acute lymphoblastic leukemia (B-ALL) is caused by DUX4-r.

3. The molecule for use according to claim 1 or 2 wherein said inhibitor of GTF2I expression and / or function is an inhibitor of HD AC expression and / or function, preferably of HD AC 6.

4. The molecule for use according to any one of previous claims wherein said inhibitor of GTF2I expression and / or function is an inhibitor of c-Myc.

5. The molecule for use according to claims 1 or 2 wherein said inhibitor of GTF2I expression and / or function is an inhibitor of the DUX4-r / GTF2I interaction.

6. The molecule for use according to any one of previous claims, wherein said inhibitor is selected from the group consisting of: a) polynucleotide capable of inhibiting the expression of the GTF2I gene and / or of HD AC gene and / or of c-Myc gene or a polynucleotide coding for said polynucleotide; b) a polypeptide, preferably an antibody or a synthetic or recombinant derivative thereof; c) a polynucleotide coding for said polypeptide or a functional derivative thereof; d) a small molecule; e) a component of the CRISPR / Cas9 system, for example a sgRNA (single strand guide RNA, ("single guide RNA")), a gRNA (guide RNA) or a guide sequence and f) a component of the protein degradation system, for example by proteolysis, of the cell, for example of proteolysis targeting chimeras (for example, proteolysis targeting chimera “PROTAC” or molecular glue).

7. The molecule for use according to any one of claims 1-3 or 6 wherein said inhibitor of HD AC is selected from Vorinostat (SAHA), belinostat (PXD-101), panobinostat (LBH-589), Dacinostat (LAQ824), CUDC-101, Abexinostat (PCI-24781), Rrcolinostat (ACY-1215), Fimepinostat (CUDC-907), Resminostat and Trichostatin A (TSA).

8. The molecule for use according to any one of claims 1-3 or 6 wherein said inhibitor of HD AC is a selective inhibitor of HDAC6, preferably said selective inhibitor of HDAC6 is MPT0G211, Tubastatin-A or HDAC6 degraders.

9. The molecule for use according to any one of claims 1-2 or 4, 6 wherein said inhibitor of c-Myc is selected from MYCi975 and WBC100.

10. The molecule for use according to claim 5, wherein said inhibitor of the DUX4-r / GTF2I interaction is a GTF2I-derived peptide, preferably a peptide comprising or consisting of GTF2I Domain B of SEQ ID No: 66.

11. The molecule for use according to any one of previous claims wherein said inhibitor is a polynucleotide capable of inhibiting the expression of the GTF2I gene and / or of HDAC gene and / or of c-Myc gene, preferably an antisense construct, an RNA inhibitor, more preferably a siRNA, a shRNA, a microRNA, a sgRNA, a gRNA, a guide sequence or an antisense oligonucleotide, or a polynucleotide coding for it.

12. The molecule for use according to any one of previous claims wherein the molecule is used in combination with a further therapeutic agent and / or a therapeutic intervention.

13. The molecule for use according to claim 12 wherein the further therapeutic agent is a chemotherapy drug or a drug antagonizing the activity of inhibitors of apoptosis, preferably an inhibitor of cIAP and XIAP, more preferably Birinapant.

14. A pharmaceutical composition for use as defined in any one of previous claims said pharmaceutical composition comprising a molecule as defined in any one of previous claims and at least one pharmaceutical acceptable carrier and / or vehicle, preferably said pharmaceutical composition is for use in the treatment and / or prevention of B-cell acute lymphoblastic leukemia (B-ALL), preferably said B-cell acute lymphoblastic leukemia (B-ALL) is caused by DUX4-r.

15. Vector comprising or expressing the polynucleotide as defined in claims 6 or 11 for use according to any one of previous claims.

16. A genetically modified host cell capable of expressing in suitable conditions the polypeptide or the polynucleotide as defined in any one of claim 6 or 11 or comprising the polynucleotide as defined in any one of claims 6 or 11 or at least a component of e) or f) as defined in claim 6 or comprising the vector as defined in claim 15 for use according to any one of previous claims.

17. An in vitro method for evaluating the risk and / or for diagnosis and / or for prognosis and / or for monitoring progression and / or for monitoring the efficacy of a therapeutic treatment and / or for screening of DUX4-r positive leukemia in a subject or in an ex vivo or in vitro system, preferably an organoid as a bone marrow organoid, comprising the steps of: a) detect or measure the amount or activity of the GTF2I protein and / or of HD AC protein and / or of c-Myc protein or of fragments thereof or of the polynucleotide encoding said protein or its fragments in an isolated biological sample obtained from the subject, which may be optionally comprised in an ex vivo or in vitro system, and b) comparison with a proper control.

Images