Compounds for treatment of diseases related to DUX4 expression

Abstract

The present invention relates to compounds for the treatment of diseases related to DUX4 expression, such as muscular dystrophies, wherein the disease is facioscapulohumeral muscular dystrophy (FSHD). It also relates to use of such compounds, or to methods of use of such compounds.

Claims

1. A casein kinase 1 inhibitor for use in the treatment of a disease or condition associated with DUX4 expression, wherein the casein kinase 1 inhibitor reduces DUX4 expression.

2. A casein kinase 1 inhibitor for use according to claim 1, wherein said disease or condition associated with DUX4 expression is a muscular dystrophy or cancer, preferably wherein said disease or condition associated with DUX4 expression is a muscular dystrophy, most preferably facioscapulohumeral muscular dystrophy (FSHD).

3. A casein kinase 1 inhibitor for use according to claim 1 or 2, characterized in that it is administered to a subject 4, 3, 2, or 1 times per day or less, preferably 1 time per day.

4. A casein kinase 1 inhibitor for use according to any one of claims 1-3, wherein the casein kinase inhibitor inhibits at least casein kinase 1δ.

5. A casein kinase 1 inhibitor for use according to any one of claims 1-4, characterized in that it is administered to a subject in an amount ranging from 0.1 to 1500 mg/day, preferably from 0.1 to 400 mg/day, more preferably from 0.25 to 150 mg/day.

6. A casein kinase 1 inhibitor for use according to any one of claims 1-5, characterized in that it is administered orally, sublingually, intravascularly, intravenously, subcutaneously, or transdermally, preferably orally.

7. A casein kinase 1 inhibitor for use according to any one of claims 1-6, wherein DUX4 expression is reduced by at least 20%, 40%, 60%, 80%, or more.

8. A casein kinase 1 inhibitor for use according to any one of claims 1-7, wherein the casein kinase 1 inhibitor reduces DUX4 expression in muscle cells, immune cells, or cancer cells.

9. A casein kinase 1 inhibitor for use according to any one of claims 1-8, wherein the reduction of DUX4 expression is determined using PCR or immunostaining.

10. A casein kinase 1 inhibitor for use according to any one of claims 1-9, wherein the casein kinase 1 inhibitor is from the class comprising an azole core.

11. A casein kinase 1 inhibitor for use according to any one of claims 1-10, wherein the casein kinase 1 inhibitor is selected from the group consisting of compounds A, B, C, D, E, F, G, H, I, J, K, L, M, N, O, SR-3029, PF-670462, and PF-5006739.

12. A composition comprising at least one casein kinase 1 inhibitor as defined in any one of claims 1-11, and a pharmaceutically acceptable excipient, for use as defined in any one of claims 1-11.

13. A composition for use according to claim 12, wherein the composition is formulated for oral, sublingual, parenteral, intravascular, intravenous, subcutaneous, or transdermal administration, preferably for oral administration.

14. An in vivo, in vitro, or ex vivo method for reducing DUX4 expression, the method comprising the step of contacting a cell with a casein kinase 1 inhibitor as defined in any one of claims 1-11, or with a composition as defined in claim 12 or 13.

15. A method for reducing DUX4 expression in a subject in need thereof, the method comprising the step of administering an effective amount of a casein kinase 1 inhibitor as defined in any one of claims 1-11, or a composition as defined in claim 12 or 13.

Description

SHORT DESCRIPTION OF DRAWINGS

[0164] FIG. 1—(A): Illustration of a DUX4 immunocytochemistry staining in FSHD myotubes from 2 different donors after 3 days of differentiation. DUX4-positive nuclei clusters are clearly stained, while DUX4-negative nuclei are not stained. The histograms show the intensity of the immunofluorescent signals (increasing intensity on the X-axis) after staining with the DUX4 and secondary antibody (top) or the secondary antibody alone (bottom); the arrows on top show the background signal (leftward arrow) or specific DUX4 signal (rightward arrow); (B): Illustration of a DUX4-stained FSHD myotube after 3 days of differentiation. The dotted pattern results from the applied filter settings to deplete the background from the secondary antibody control. Note that the threshold settings prohibit detection of the weaker DUX4 signal in the nuclei more distant from the sentinel nucleus.

[0165] FIG. 2—Script-based image analysis includes nuclei identification, myotube identification, detection of nuclei inside or outside myotube borders (used to calculate fusion index), DUX4 positive nuclei and clusters, myotube area, myotube width, and myotube skeleton length.

[0166] FIG. 3—Validation of the primary screening assay format in 384-well format. Three independent experiments are shown, illustrating the assay window obtained using script-based quantification of the number of DUX4-expressing nuclei in differentiating primary myotubes after 3 days in differentiation medium. The assay window is defined by the DUX4 signal and the background signal of the secondary antibody (representing the signal in total absence of DUX4).

[0167] FIG. 4—(A): Schematic representation of the screening assay protocol. Myoblasts were seeded at day −1 and medium was changed to differentiation medium at day zero. Cells were allowed to differentiate for 3 days. Compounds were added 15 h prior to fixation. (B): Correlation of duplicated results from primary screening of an annotated compound library using 2 different readouts for DUX4 expression (Number of DUX4-positive nuclei and DUX4 intensity) and 2 different readouts to monitor potential toxicity (fusion index, nuclei count). Hit calling thresholds (high stringency) are indicated by a dashed line, and the upper right quadrants contain the hit compounds for the different readouts. Axes of the scatter plots are symmetrical.

[0168] FIG. 5—Concentration-response curves for various CK1 inhibitors for the different readouts. The DUX4 nuclei count, DUX4 intensity, fusion index, and total nucleus count were measured after 15 hour of compound exposure. (A): results for PF-670462; (B): results for PF-5006739; (C): results for compound C; (D): results for compound D; (E): results for compound E; (F): results for compound F; (G): results for compound G; Structural formulae are shown in example 5.

[0169] FIG. 6—(A): Schematic representation of the assay protocol. Myoblasts were seeded at day −1 and medium was changed to differentiation medium at day zero. Cells were allowed to differentiate for 3 days. Compounds were added for 15 h or 72 h prior to fixation. For the 15 h treatment, compounds are administered when differentiation already progressed significantly. In case of 72 h treatment, compounds were incubated during the full differentiation phase. The other panels show concentration-response curves for a BET inhibitors (B, C) or for beta2 adrenoreceptor agonists (D, E, F, G, H, I) for the different readouts. DUX4 nuclei count, DUX4 intensity, fusion index, and total nuclei count were assessed after 15 h or after 72 h of treatment. (B, C): (+)JQ1; (D, E): formoterol; (F, G): salbutamol; (H, I): salmeterol; (J): micrographs of myotubes after 72 hours in differentiation medium while exposed to the a beta2 adrenoreceptor agonist (formoterol); (K, L): results for both 15 hour and 72 hour exposure to a CK1 inhibitor (PF-670462).

EXAMPLES

Example 1—Primary FSHD Muscle Cells Express DUX4 in a Small Fraction of Myonuclei

[0170] The inventors succeeded in establishing a sensitive DUX4 detection method in primary myotubes and used this to build a high-content assay for quantitative assessment of endogenous DUX4 expression. The method was developed into a validated phenotypic screening platform for automated detection and quantification of endogenous DUX4 expression. Mechanisms underlying DUX4 repression may involve many interacting proteins, favouring such a phenotypic approach. Furthermore, it is pathway/target independent (and thus not hypothesis-driven) and provides additional information on cell toxicity or interference with muscle differentiation.

[0171] Significant differences in the levels of DUX4 expression between cells obtained from different donors have been reported. Therefore, muscle cell lines derived from different donors were thoroughly characterised and an optimal cell line was selected for primary screening. MyoD staining of myoblasts confirmed solid myogenicity of all cell lines (Rudnicki et al., 1993; cell 75(7):1351-9). After optimisation of parameters, a DUX4 detection procedure was established that could be applied in a screening assay which resulted in the expected DUX4 pattern in FSHD cells, but not in myotubes from healthy donors. As shown in FIG. 1, this included a nuclear DUX4 localization, with only few positive cells, and an intensity gradient through DUX4-positive nuclear clusters, as also described by Rickard et al., (2015, DOI: 10.1093/hmg/ddv315).

Example 2—Screening Assay to Identify DUX4 Repression

[0172] A quantitative assay readout was developed based on script-based image analysis. Cells were stained according to example 1, also using DAPI to detect myonuclei and an antibody against myosin heavy chain (MHC) to visualize the formation of myotubes. To analyse the images, an automated script was developed, enabling the detection of nuclei, myotube borders and DUX4 signals, with the script also detecting artefacts to reduce false positive signals. The script enabled multiple validated readouts including the number of DUX4 positive nuclei and nuclei clusters, the fusion index, myotube area, myotube width and myotube skeleton length (see FIG. 2). Additionally, the total nuclei count was included as a measure of cell loss or compound toxicity. The script was validated by evaluating endogenous DUX4 expression in the primary myotubes, and results were in line with literature values, with the number of DUX4 expressing nuclei being <0.5%.

[0173] The assay has been further matured to make it suitable for screening purposes. The assay quality was dependent on the donor cell line. The number of DUX4 positive nuclei was characteristic for each donor cell line, and was consistent between experiments. The best performing cell lines in terms of number of DUX4 expressing nuclei, reproducibility and Z-factor have been selected for miniaturization of the assay to a 384-well format, thus allowing for automated screening of large compound libraries. A cell line with 2 D4Z4 repeats was selected for the primary screening, while a cell line with 6 D4Z4 repeats was selected for later validation. The primary screening assay had a Z-factor of 0.6, which represents an excellent assay (Zhang et al., 1999, doi:10.1177/108705719900400206; see FIG. 3).

[0174] A compound library containing approximately 5000 annotated compounds was screened in the high-content assay. For this purpose, primary myoblasts were seeded in 384 well plates after which the growth medium was replaced with differentiation medium. After 3 days of differentiation, cells were treated with library compounds (in duplicate on different screening plates) for 15 h, after which they were fixed and stained with antibodies against DUX4, antibodies against myosin heavy chain (MHC), and with DAPI (4′,6-diamidino-2-phenylindole). Script-based analysis provided readouts for DUX4 expression (count of DUX4-positive nuclei or DUX4 intensity) and for potential toxicity (fusion index and nuclei count). Results are shown in FIG. 4. The majority of the approximately 200 hits was confirmed in an experiment using the same assay and 5 replicates. These compounds were selected for further concentration-response profiling.

[0175] Half of these hits were validated using RT-PCR. Based on mRNA expression of DUX4 and the downstream target genes Trim43 & ZScan4, using housekeeping genes hGUSB, GAPDH, hRPL27 as a reference, a very good correlation between DUX4 repression in the immunocytochemistry assay (protein level) and the RT-PCR assay (mRNA level) was observed. This suggests that the vast majority of the hits have an upstream mode of action, i.e. they act by inhibiting the expression of DUX4 (as opposed to accelerating degradation of DUX4).

[0176] RT-PCR was performed as described by Lemmers et al., (2010, DOI: 10.1126/science.1189044) using oligonucleotides ordered from Applied Biosystems (Foster City, USA), possibly as part of assay kits (for hGAPDH (app): AssayID Hs02758991_g1; for hTRIM43(app): Assay ID Hs00299174_m1; for hMYH2_tv1-2(app): AssayID Hs00430042_m1). Other oligonucleotides are shown in table 1.

TABLE-US-00001 TABLE 1 primers and probes for use in PCR Name Sequence SEQ ID NO: hDUX4 forward CCCGGCTGACGTGCAA 1 hDUX4 reverse AGCCAGAATTTCACGGAAGAAC 2 hDUX4 probe AGCTCGCTGGCCTCTCTGTGCC 3 hGUSB forward TTCCCTCCAGCTTCAATGACA 4 hGUSB reverse CCACACCCAGCCGACAA 5 hGUSB probe AGGACTGGCGTCTGCGGCA 6 hRPL27 forward TGTCCTGGCTGGACGCTACT 7 hRPL27 reverse GAGGTGCCATCATCAATGTTCTT 8 hRPL27 probe CGGACGCAAAGCTGTCATCGT 9 hZSCAN4 forward AGGCAGGAATTGCAAAGACTTT 10 hZSCAN4 reverse AATTTCATCCTTGCTGTGCTTTT 11 hZSCAN4 probe TAGGATCTTTCACTCATGGCTGC 12 AACCA hMYOG forward GCTCACGGCTGACCCTACA 13 hMYOG reverse CACTGTGATGCTGTCCACGAT 14 hMYOG probe CCCACAACCTGCACTCCCTCACCT 15

Example 3—CK1 Inhibitors Act as DUX4 Repressors

[0177] The validated assay was used for screening an annotated compound library containing approximately 5000 compounds, to identify novel mechanisms of action for DUX4 repression. This library contained compounds with annotated pharmacology, not only entailing the primary pharmacology of the compounds but also potential known polypharmacology. The primary screening achieved multiple hits, identifying compounds that reduced the number of DUX4 positive nuclei. Hits were further profiled by establishing concentration-response curves. By applying a bioinformatics approach on the screening and profiling dataset, the inventors surprisingly discovered that compounds with a CK1 annotation were significantly enriched in the phenotypically active compound population, i.e. in the group of compounds inducing a repression of DUX4. Interestingly, none of the original compounds with a CK1 annotation had CK1 as its primary pharmacological target, each having other high potency targets from other protein families. Thus the bioinformatics analysis was essential in identifying the association between CK1 and DUX4 repression.

[0178] Profiled compounds were annotated as being phenotypically active when they showed a concentration-dependent effect on DUX4 (inhibition or activation). Of these, compounds which showed inhibition of the fusion index or of the total number of nuclei by more than 10% were excluded unless the effect on these readouts was at least 5-fold less potent than the effect on DUX4. As such, from the 4790 unique compounds, 188 compounds were classified as being phenotypically active, 162 of which were DUX4 inhibitors.

[0179] For the phenotypically active compounds, the original target annotations were complemented with additional information that is publically available (literature, patent applications, supplier databases, etc.). All human proteins, and non-human orthologues where a mapping to the human proteome can be established, were considered. Each of the 4790 compounds was then evaluated against these target annotations, classifying the target as being active or inactive for a given compound. For the phenotypically active compounds, the annotated targets were classified as being active if the compound's potency on the target was <10 times the phenotypic potency, otherwise the target was classified as inactive. This analysis revealed that approximately 201 targets were associated with phenotypic activity at a False Discovery Rate of 0.05. An enrichment of compounds annotated as CK1 inhibitors was detected in the group of phenotypically active compounds.

Example 4—CK1 Isoforms are Expressed in FSHD Primary Muscle Cells

[0180] To confirm target expression in both healthy and FSHD muscle cells, an RNA sequencing approach was followed to determine the expression of the different CK1 isoforms in primary myotubes from 4 different FSHD donors and from 4 different healthy donors. The results show expression of all CK1 isoforms, both in FSHD and in healthy muscle cells. The highest expression is of CK1 α, CK1 δ and CK1 ε (see table 2).

TABLE-US-00002 TABLE 2 expression of casein kinase 1 isoforms in 4 healthy primary cell lines, and in 4 FSHD primary cell lines as determined by RNA sequencing of differentiated myotubes CSNK1A1 CSNK1D CSNK1E CSNK1G1 CSNK1G2 CSNK1G3 FSHD 134 159.1 160.1 49.9 81.8 37.9 FSHD 122.5 138.4 136.8 4.2 79.1 32.7 FSHD 176.7 170.6 120.5 69.8 65.8 41.3 FSHD 118.2 134 105.6 41.8 63.5 38.1 Healthy 138.9 168.5 188 45.8 75.9 35.8 Healthy 143.3 174.1 200.7 49.6 81.8 36.3 Healthy 139.2 192.8 176.1 51.9 71.4 33.2 Healthy 119.1 132.4 122.4 40.6 65.9 40.1

Example 5—Inhibition of CK1 Represses DUX4

[0181] The DUX4 repression of CK1 inhibitors was assayed following the protocol of Example 2, illustrated in FIG. 4A. Table 3 shows the structures of the CK1 inhibitors that are used in FIG. 5. Compounds were incubated with primary FSHD cells for 15 hours, as indicated by the arrow in FIG. 4A. Results are shown in FIG. 5, while table 3 shows half maximal effective concentrations (EC.sub.50) values. Table 3 also shows determined IC.sub.50 values in nM for CK1α, CK1δ, CK1ε, and p38α, denoted as CK1 a, d, e, and p38a, respectively.

TABLE-US-00003 TABLE 3 Examplary CK1 inhibitors for use according to the invention, along with half maximal effective concentrations (EC.sub.50) [00003] [00004] [00005] [00006] [00007] [00008] [00009] [00010] [00011] [00012] [00013] [00014] [00015] [00016] [00017] [00018] [00019] [00020]

[0182] Several of these compounds were also tested in vivo in a mouse model. The model was based on human FSHD-affected myoblasts engrafted onto a mouse thigh muscle. These human FSHD myoblasts then fused and developed into myotubes, which produce DUX4. This model approximates natural FSHD biology as much as possible by using primary FSHD-affected muscle cells. The diseases cells are engrafted in one thigh, and healthy human myoblasts in the other thigh, so that each mouse serves as its own control. The compounds also showed repression of DUX4 in these in vivo models, as established by RT-PCR and histological examination.

Example 6—CK1 Inhibitors do not Inhibit Myotube Fusion

[0183] Because DUX4 expression increases upon in vitro differentiation of proliferating FSHD myoblasts into multinucleated myotubes (Balog et al., 2015 Epigenetics. 2015; 10(12):1133-42), inhibition of differentiation might lead to a false positive effect on DUX4 repression.

[0184] Bromo- and Extra-Terminal domain (BET) inhibitors such as the non-selective inhibitor (+)JQ1 or the BRD4-selective inhibitor RVX-208 can inhibit the expression of DUX4 in immortalised differentiated myotube cultures (see US2015087636A1). It was shown there that when differentiating myotubes were exposed to (+)JQ1 at the start of the differentiation process, i.e. from the moment when the growth medium was changed to the differentiation medium, the expression of myosin heavy chain (MYH2, a differentiation marker) was decreased, suggesting that the inhibitor also impacted the differentiation process. Both (+)JQ1 and RVX-208 have been evaluated in the phenotypic assay described in this application. Agonists of the beta2 adrenoreceptor have also been reported to inhibit DUX4 expression in differentiating myotubes (Campbell et al., 2017). We evaluated the effect of both BET inhibitors and beta2 adrenoreceptor agonists on the fusion process and compared in to the effect of a CK1 inhibitor.

[0185] FIG. 6A shows the experimental setup of Example 2. Compounds are administered either 15 h before fixation, resembling the original screening protocol, or 72 h before fixation (grey arrow). In the latter case, compounds are present during the whole differentiation process. The inventors found that early administration of the BET inhibitor (+)JQ1 (FIG. 6B, C) and agonists of the beta2 adrenoreceptor (FIGS. 6D, E, F, G, H, I) inhibit the fusion process and the differentiation of myoblasts into myotubes. FIG. 6J shows that no myotube formation can be observed after treatment with a beta2 adrenoreceptor agonist (formoterol). This leads to a false positive readout when assessing the DUX4 signal. The BET inhibitor RVX-208 did not show any effect on DUX4 expression, irrespective of treatment time (not shown). While the fusion index did not appear to be affected at the 15 h timepoint, also with this treatment time the myotube fusion process was affected by these compounds as determined by RT-PCR showing inhibition of the expression of the late differentiation marker myosin heavy chain (Myh; not shown; primers were from hMYH2 kit described above).

[0186] As illustrated in example 5, inhibition of CK1 inhibits DUX4. This effect occurs without inhibiting myotube fusion, neither after 15 h nor after 72 h of compound treatment (FIG. 6K, L).

Example 7—Ck1 Inhibitors Inhibition Profile

[0187] Compounds PF-670462, PF-5006739, Compound E, Compound F, Compound D, Compound H, Compound A, and SR3029 were assayed for their inhibition of CK1 α, CK1 δ, CK1 ε, and of p38, and their concurrent repression of DUX4. Table 4 shows inhibitory results.

TABLE-US-00004 TABLE 4 inhibition of CK1 and p38 by CK1 inhibitors, in nM IC.sub.50 PF- PF- SR- EC.sub.50 670462 5006739 E F D H A 3029 CK1 α 320 123   592 561 644 33 30 >10k CK1 δ  29 20  31  18   33.1 22 19 346 CK1 ε 100 27  84  72   51.6 16 12 381 p38  32 74 1110 677 569 25 13 >10k DUX4 470 820  1890 2590  1410  10 50  50 (n = 4) (n = 12) (n = 4) (n = 2) (n = 2) (n = 2) (n = 2)

REFERENCES

[0188] Balog et al., 2015 Epigenetics. 2015; 10(12):1133-42; Bergerat et al., 2017, DOI: 10.1016/j.prp.2016.11.015; Van den Boogaard et al., 2016, DOI: 10.1016/j.ajhg.2016.03.013; Brockschmidt et al., 2008, DOI: 10.1136/gut.2007.123695; Campbell et al., 2017, DOI: 10.1186/s13395-017-0134-x; Chebib and Jo, 2016, DOI: 10.1002/cncy.21685; Eide E J, Virshup D M, 2001, DOI:10.1081/CBI-100103963; Etchegaray J P et al., 2009, DOI:10.1128/MCB.00338-09; Geng et al., 2012, DOI: 10.1016/j.devcel.2011.11.013; Kowaljow et al., 2007, DOI: 10.1016/j.nmd.2007.04.002; Lang et al., 2014, DOI: 10.14205/2310-8703.2014.02.01.1; Lemmers et al., 2010, DOI: 10.1126/science.1189044; Lilljebjörn & Fioretos, 2017, DOI: 10.1182/blood-2017-05-742643; Oyama et al., 2017 DOI: 10.1038/s41598-017-04967-0; Paz et al., 2003, DOI: 10.1093/hmg/ddg226; Rickard et al., 2015, DOI: 10.1093/hmg/ddv315; Rudnicki et al., 1993; cell 75(7):1351-9; Sharma et al., 2016, DOI:10.4172/2157-7412.1000303; Snider et al., 2010, DOI: 10.1371/journal.pgen.1001181; Stadler et al., 2013, DOI: 10.1038/nsmb.2571; Tawil et al., 2014, DOI: 10.1186/2044-5040-4-12; Vanderplanck et al., 2011, doi: 10.1371/journal.pone.0026820; Wallace et al., 2011, DOI: 10.1002/ana.22275; Yao et al., 2014, DOI: 10.1093/hmg/ddu251; Yasuda et al., 2016, doi: 10.1038/ng.3535; Young et al., 2013, doi:10.1371/journal.pgen.1003947; Zhang et al., 1999, doi:10.1177/108705719900400206; Zhang et al., 2017, DOI:10.1038/ng.3691 WO2011051858/WO2012085721/WO2015119579/EP2949651/WO2009016286/US2005/0131012/WO2015195880/WO2014081923/US20140221313/US2015087636A1