QUINAZOLINE COMPOUNDS AS INHIBITORS OF PREMATURE TERMINATION CODONS

Abstract

The present invention relates to the use of at least one compound of formula (I), or one of its pharmaceutically acceptable salts, for preventing and/or treating a disease caused by a nonsense mutation. It also relates to compounds of formula (II) and their uses.

Claims

1. A method for preventing and/or treating a disease caused by a nonsense mutation, said disease being chosen from genetic diseases caused by a nonsense mutation and cancers caused by a nonsense mutation which is present in a tumor-suppressor gene, in a subject, which comprises administering to said subject at least one compound of formula (I), or one of its pharmaceutically acceptable salts: ##STR00043## wherein: R1 is a C1-C6 alkyl radical; R2 is H, a halogen atom or a C1-C6 alkoxy radical; and R3 is H, a C1-C6 alkyl radical or a C1-C6 alkoxy radical.

2. The method according to claim 1, wherein the genetic disease or the cancer is caused by the presence of a nonsense mutation in a coding sequence of interest.

3. The method according to claim 1, wherein R2 and R3 are each not simultaneously H.

4. The method according to claim 1, wherein: R1 is methyl or ethyl; R2 is H, a halogen atom or a methoxy radical; and R3 is H, a methyl radical or a methoxy radical.

5. The method according to claim 1, wherein: R1 is methyl or ethyl; R2 a halogen atom or a methoxy radical; and R3 is H or a methyl radical.

6. The method according to claim 1, wherein the compound is chosen from the following compounds and their pharmaceutically acceptable salts: TLN68, which is ##STR00044## EC-18, which is ##STR00045## EC-35 which is ##STR00046## EC-30 which is ##STR00047## EC-141 which is ##STR00048## ##STR00049## EC-85 which is ##STR00050## EC-11 which is ##STR00051## EC-288 which is ##STR00052## EC-130 which is ##STR00053## EC-335 which is ##STR00054## EC-265 which is ##STR00055##

7. The method according to claim 1, wherein the compound is chosen from the following compounds and their pharmaceutically acceptable salts: EC-18, which is ##STR00056## EC-141 which is ##STR00057## EC-85 which is ##STR00058## EC-11 which is ##STR00059## EC-288 which is ##STR00060## EC-130 which is ##STR00061## EC-335 which is ##STR00062## EC-265 which is ##STR00063##

8. The method according to claim 1, wherein the genetic disease caused by a nonsense mutation is chosen from cystic fibrosis due to a nonsense mutation G542X in the cystic fibrosis transmembrane conductance regulator gene, Duchenne muscular dystrophy due to a nonsense mutation in dystrophin, beta thalassemia due to nonsense mutation in β-globin, Niemann-Pick disease type A, B or C due to nonsense L261X mutation in acid sphingomyelinase, Hurler syndrome, Dravet syndrome, spinal muscular atrophy, myoadenylate deaminase deficiency, antithrombin III deficiency, alpha-1 antitrypsin deficiency, apolipoprotein deficiency of apolipoprotein Al, B, CII or E, adenine phosphoribosyltransferase (APRT) deficiency, haemophilia A due to nonsense mutation in Factor VIII), haemophilia B due to nonsense mutation in Factor IX), Von Willebrand disease due to a nonsense mutation in Von Willebrand factor, Fanconi anemia-group C, Marfan syndrome, Gaucher disease, Donohue syndrome, oculocerebrorenal syndrome of Lowe and Xeroderma pigmentosum.

9. Compound of formula (II) or one of its pharmaceutically acceptable salts: ##STR00064## wherein: R1 is a C1-C6 alkyl radical, preferably methyl or ethyl; R2 is a halogen atom or a C1-C6 alkoxy radical, preferably methoxy; and R3 is H or a C1-C6 alkyl radical, preferably methyl, with the proviso that when R2 is a C1-C6 alkoxy radical, then R1 is ethyl, and when R2 is a halogen and R3 is H then the compound is in the form of a pharmaceutically acceptable salt such as a salt with formic acid or with iodine.

10. Compound according to claim 9, wherein it is chosen from the following compounds and their pharmaceutically acceptable salts: EC-141 which is ##STR00065## EC-85 which is ##STR00066## EC-288 which is ##STR00067## EC-130 which is ##STR00068## EC-335 which is ##STR00069## EC-265 which is ##STR00070##

11. A method for treating a disease, which comprises administering at least one compound according to claim 9.

12. Composition comprising, in a pharmaceutically acceptable carrier, at least one compound according to claim 9 or one of its pharmaceutically acceptable salts.

13. A method for treating a disease caused by a nonsense mutation, said disease being chosen from genetic diseases caused by a nonsense mutation and cancers caused by a nonsense mutation which is present in a tumor-suppressor gene, in a subject, which comprises administering to said subject at least one product comprising: a) a compound according to claim 1, or one of its pharmaceutically acceptable salts, and b) at least one other drug, said compound and said at least one other drug being formulated in the product for a simultaneous, separate or sequential administration.

14. A method for treating cancer, and/or for preventing cancer metastasis, and/or for preventing cancer recurrence, and/or for decreasing resistance to a chemotherapeutic drug, in a subject, comprising administering to said subject at least one product comprising: a) a compound according to claim 9, or one of its pharmaceutically acceptable salts, and b) at least one chemotherapeutic drug, said compound and said at least one chemotherapeutic drug being formulated in the product for a simultaneous, separate or sequential administration.

15. The method according to claim 13, or wherein the drug is chosen from ataluren, gentamicin, negamycin, clitocine, escin and a nonsense-mediated mRNA decay (NMD) inhibitors.

16. The method according to claim 14, wherein the drug is chosen from ataluren, gentamicin, negamycin, clitocine, escin and a nonsense-mediated mRNA decay (NMD) inhibitor.

Description

LEGENDS TO THE FIGURES

[0171] Note that in the figures, TLN68 is the same molecule as “TLN468”.

[0172] FIG. 1: Primary selection and Secondary quantification of the readthrough effect of the positive hits

[0173] A) Positives hits were selected according to their strictly standardized mean difference (SSMD) score and according to the increase factor of the luciferase activity between the treated condition and the median of the negative controls (untreated cells). The molecules from the Chembridge chemotherapy library and those of Prestwick are represented in left and right panels respectively. The red crosses correspond to the molecules whose SSMD ≥2 or the factor of increase is ≥1.4 for a total of 465 molecules.

[0174] B) The dual reporter system is indicated in the top part. The stop indicates the R213X sequence inserted between lacZ and F-luc. The right part shows the ability of the molecules to stimulate readthrough. NIH3T3 cells were treated for 24 hours with the gentamicin (2.5 mM) as control or with the 43 molecules selected (50 μM). Arrows represent the 4 molecules (TLN1399, TLN236, TLN309 and TLN68) inducing readthrough by a factor greater than 4.

[0175] FIG. 2: Western blots from HDQ-P1 cells harboring the endogenous nonsense mutation P53-R213X

[0176] A) The p53 protein is detected by the p53-DO1 antibody recognizing the N-terminal region. Actin is used as control. Control was only obtained in one of two Western blots.

[0177] B) HDQ-P1 cells were treated or not with G418 (400 μM), TLN68 (80 μM) or TLN309 (40 μM) for 48 hours. Mutant p53 mRNA levels were determined by quantitative PCR (n=3).

[0178] The results of each condition are expressed in relation to the normalized relative amount of mRNA in the absence of a treatment.

[0179] C) Structures of the TLN68 (Translectine, or TLN468) and TLN309 (Amodiaquine).

[0180] FIG. 3: TLN68 restores p53-R231X activity

[0181] A) The human H1299 (P53.sup.−/−) cells were cotransfected by the vectors pCMV-R213X, pCMV-LacZ and p53BS-luc. The pCMV-R213X vector expresses the P53 gene with the nonsense R213X mutation. The p53BS-luc vector expresses the luciferase gene under the control of a p53-dependent promoter and allow detection of a transcriptionally active p53. The vector pCMVLacZ was used to normalize the results.

[0182] B) The HDQ-P1 cells were treated or not with G418 (400 μM) or TLN68 (80 μM) for 48 hours. Bax gene mRNA levels, one of the major targets of p53, were determined by quantitative PCR (n=3). The results of each condition are expressed relative to the normalized relative amount of mRNA in the absence of a treatment set at 1.

[0183] FIG. 4: Sequence specificity of TLN68

[0184] A) Schematic representation of the dystrophin gene and the position of the selected 40 mutations.

[0185] B) Quantification of stop codon readthrough efficiency for the 40 more frequent nonsense mutations found in DMD gene (n=6), in presence or absence of TLN68 at 60 mM.

[0186] Gentamicin (2.5 mM) measurement has been done in the same conditions to compare the TLN68 effect with gentamicin.

[0187] FIG. 5: Additive effect between TLN68 and gentamicin

[0188] Hela cells are treated with gentamicine, TLN68 or both drugs during 24 h before stop codon readthrough quantification. The values for the untreated cells are setup to 1 and fold changes are calculated from this value. At least six independent measurements are performed for each condition.

[0189] FIG. 6: Selection of the recombinant cell-line for HTS

[0190] Recombinant NIH3T3 cells are cultured from 96-wells plates starting from 15,000 cells/well. Luciferase activity is quantified by adding directly the coelenterazine in the well (n=4). Black and white bars represent basal metridia-luciferase activity and gentamicine (1.6 mM) induced activity, respectively. The inventors retain the clone number C14 for further development.

[0191] FIG. 7: Statistical validation of the reporter cell line SSMD values obtained for the screening of the two libraries Chembridge and Prestwick are in panels A and B respectively. X-crosses indicate compound with a SSMD>2, † crosses represent negative controls added to the plates during the screening.

[0192] FIG. 8: TLN68 has no impact on translational frameshifting

[0193] TLN468 has no impact on both −1 and +1 programmed ribosomal frameshifting (PRF). No significant difference is observed between treated and untreated cells (right and left bars respectively). TLN468 has been tested on −1 PRF from HIV-1 and +1 PRF from OAZ1 gene.

[0194] FIG. 9: WST1 assay to quantify toxicity of TLN468 on the human cell line HELA, and the mouse cell line NIH3T3.

EXAMPLE 1: PREPARATION OF COMPOUNDS OF FORMULA (I) OR (II) ACCORDING TO THE INVENTION

[0195] The following examples are prepared according to the protocols described above:

1-(7-methoxy-4-methylquinazolin-2-yl)guanidine EC-18 (C.SUB.11.H.SUB.13.N.SUB.5.O, 231.26 g/mol)

[0196] ##STR00036##

[0197] .sup.1H NMR (400 MHz, (CD.sub.3).sub.2SO): 8.45 (s, 1H), 8.08 (d, J=9.1 Hz, 1H), 7.35 (d, J=2.5 Hz, 1H), 7.15 (dd, J=9.1, 2.6 Hz, 1H), 3.93 (s, 3H), 2.79 (s, 3H).

[0198] .sup.13C NMR (400 MHz, (CD.sub.3).sub.2SO): 169.7, 167.4, 164.2, 156.4, 155.5, 151.4, 127.5, 117.7, 115.8, 105.5, 55.8, 21.3.

[0199] IR (neat): 3240 (broad), 2974, 1686, 1562, 1340, 1215, 1018, 699.

[0200] HRMS calcd for C.sub.11H.sub.14N.sub.5O (M+H.sup.+): 232.1193 Found: 232.1194.

1-(4,8-dimethylquinazolin-2-yl)guanidine EC-35 (C.SUB.11.H.SUB.13.N.SUB.5., 215.26 g/mol)

[0201] ##STR00037##

[0202] .sup.1H NMR (400 MHz, (CD.sub.3).sub.2SO): 9.23 (br s, 3H), 8.45 (s, 1H), 8.03 (d, J=8.2 Hz, 1H), 7.79 (d, J=7.0 Hz, 1H), 7.49-7.42 (m, 1H), 2.86 (s, 3H), 2.55 (s, 3H).

[0203] .sup.13C NMR (400 MHz, (CD.sub.3).sub.2SO): 171.6, 167.6, 156.6, 154.2, 147.5, 134.6, 133.7, 125.1, 123.7, 120.4, 21.7, 17.2.

[0204] IR (neat): 3210 (broad), 2981, 1697, 1586, 1353, 1153, 759.

[0205] HRMS calcd for C.sub.11H.sub.14N.sub.5 (M+H.sup.+): 216.1244 Found: 216.1246.

1-(7-bromo-4,8-dimethylquinazolin-2-yl)guanidine EC-130 (C.SUB.11.H.SUB.12.BrN.SUB.5., 294.16 g/mol)

[0206] ##STR00038##

[0207] .sup.1H NMR (400 MHz, (CD.sub.3).sub.2SO): 8.42 (s, 1H), 7.93 (d, J=8.8 Hz, 1H), 7.67 (d, J=8.9, 1H), 2.83 (s, 3H), 2.63 (s, 3H).

[0208] .sup.13C NMR (400 MHz, (CD.sub.3).sub.2SO): 171.4, 166.9, 157.2, 156.3, 148.7, 133.3, 130.0, 128.4, 124.9, 119.4, 21.7, 17.1.

[0209] IR (neat): 3305 (broad), 2972, 1694, 1560, 1323, 1032, 780.

[0210] HRMS calcd for C.sub.11H.sub.13BrN.sub.5 (M+H.sup.+): 294.0349 Found: 294.0350.

1-(7-fluoro-4,8-dimethylquinazolin-2-yl)guanidine EC-288 (C.SUB.11.H.SUB.12.FN.SUB.5., 233.25 g/mol)

[0211] ##STR00039##

[0212] .sup.1H NMR (400 MHz, (CD.sub.3).sub.2SO): 8.41 (s, 1H), 8.11 (d, J=9.1, 6.1 Hz, 1H), 7.41 (d, J=9.3, 1H), 2.83 (s, 3H), 2.41 (s, 3H).

[0213] .sup.13C NMR (400 MHz, (CD.sub.3).sub.2SO): 171.5, 167.2, 164.4, 161.9, 156.0, 155.1, 149.5, 126.0 (.sup.3J.sub.CF=10.8 Hz), 117.9 (.sup.2J.sub.CF=16.1 Hz), 115.0 (.sup.2J.sub.CF=26.5 Hz), 38.9 (overlapping with the solvent residual peak), 8.58 (.sup.3J.sub.CF=4.0 Hz).

[0214] .sup.19F (376 MHz, (CD.sub.3).sub.2SO): −106.5.

[0215] IR (neat): 3318 (broad), 2922, 1695, 1587, 1336, 1059, 788.

[0216] HRMS calcd for C.sub.11H.sub.13FN.sub.5 (M+H.sup.+): 234.1150 Found: 234.1153.

1-(7-chloro-4,8-dimethylquinazolin-2-yl)guanidine EC-335 (C.SUB.11.H.SUB.12.ClN.SUB.5., 249.70 g/mol)

[0217] ##STR00040##

[0218] .sup.1H NMR (400 MHz, (CD.sub.3).sub.2SO): 8.41 (s, 1H), 8.02 (d, J=8.9 Hz, 1H), 7.52 (d, J=8.9 Hz, 1H), 2.84 (s, 3H), 2.60 (s, 3H).

[0219] .sup.13C NMR (400 MHz, (CD.sub.3).sub.2SO): 170.9, 166.6, 157.4, 149.0, 138.3, 131.0, 125.4, 124.8, 124.7, 118.7, 21.7, 14.0.

[0220] IR (neat): 3301 (broad), 2970, 1694, 1571, 1325, 1022, 782.

[0221] HRMS calcd for C.sub.11H.sub.13ClN.sub.5 (M+H.sup.+): 250.0854 Found: 250.0858.

Example 2: Translational Suppression of PTC Mutations in Duchenne and Becker Muscular Dystrophy (DMD)

[0222] Materials and Methods

[0223] Cell Lines and Plasmids

[0224] All cells were cultured in DMEM plus GlutaMAX (Invitrogen), except for H1299 cells, which were cultured in RPMI plus GlutaMAX (Invitrogen). The medium was supplemented with 10% foetal calf serum (FCS, Invitrogen) and 100 U/ml penicillin/streptomycin. Cells were kept in a humidified atmosphere containing 5.5% CO.sub.2, at 37° C. NIH3T3 cells are embryonic mouse fibroblasts. H1299 is a p53-null cell line established from a human lung carcinoma (provided by the ATCC). HDQ-P1 is homozygous for a nonsense mutation at codon 213 (CGA to TGA) in the p53 gene. This cell line was established from a human primary breast carcinoma and was provided by DSMZ-German collection of microorganisms and cell cultures. In order to generate a stable mammalian cell line the inventors used a secreted Metridia luciferase reporter gene derived from pMetLuc2 (Clonetech). The coding sequence of this gene was interrupted by a TP53 nonsense mutation R213X with its own nucleotide context and cloned at an Eco53Kl site created by directed mutagenesis at nucleotide 57. The final construction was named pML213 and was used to stably transfect NIH3T3 cells with JetPei reagent (Invitrogen). Seven neomycin resistant clones were tested for their capacity to express active metridiae after readthrough induction in the presence of 1.6 mM gentamicin during 24 hours. Then 50 μL of culture medium was taken from each well and incubated in the presence of substrate: coelenterazine, according to the conditions recommended by the supplier (Ready-To-Glow Secreted Luciferase Reporter Assay (Clonetech)). The photon emission generated by the reaction was measured in a plate luminometer (Tecan).

[0225] The clone presenting the best increase factor between the treated condition and control was chosen to realize the HTS.

[0226] HTS Screening

[0227] Each molecule from the two libraries was tested at 50 μM. In each plate 8 wells over 80 were reserved for positive control (gentamicin) and 8 wells for negative control (DMSO). The Prestwick Chemical Library is a unique collection of 1,280 small molecules already approved by FDA, EMA and other agencies. The second library is a subgroup of 16,480 compounds, selected on the criteria of chemical and pharmacological diversity from the ChemBridge library that includes more than one million so far. The inventors also checked for false positive hits by testing all selected drugs on NIH3T3 parental cell line without reporter gene in order to eliminate drugs capable of artificially increasing the photon emission.

[0228] Statistical Validation of the New Reporter Cell Line

[0229] To validate the screening strategy the inventors first applied the screening protocol to five 96 wells plates from the library using gentamicin as positive control and DMSO as negative control. For statistical validation the inventors used the SSMD parameter presented by Zhang et al. in 1999. The strictly standardized mean difference (SSMD) is robust to both measurement unit and strength of positive control. It takes into account data variability in both compared groups and has a probability interpretation. The inventors obtain SSMD 2 validating the screening protocol (FIG. 7).

[0230] Readthrough Quantification

[0231] Complementary oligonucleotides corresponding to TP53 R213X nonsense mutations and 9 nucleotides on both sides are ligated into the pAC99 dual reporter plasmid, as previously described (Premature stop codons involved in muscular dystrophies show a broad spectrum of readthrough efficiencies in response to gentamicin treatment, Bidou et al, Gene Therapy 1 Apr. 2004, 11(7):619-627). This dual reporter is used to quantify stop-codon readthrough through the measurement of luciferase and beta-galactosidase (internal calibration) activities, as previously described (Stahl et al, Versatile vectors to study recoding: conservation of rules between yeast and mammalian cells. Nucleic Acids Res 1995; 23:1557-60). Readthrough levels for nonsense mutations were analysed in the presence or absence of tested molecules. Cells were seeded in a 6-well plate. The next day, cells are transfected with the reporter plasmid in the presence of JetPei reagent (Invitrogen). The following day, they are rinsed with fresh medium, with or without readthrough inducers. Cells were harvested 24 hours later, with trypsin-EDTA (Invitrogen), lysed with Glo lysis buffer (Promega) and beta-galactosidase and luciferase activities were assayed as previously described (Stahl et al, Versatile vectors to study recoding: conservation of rules between yeast and mammalian cells. Nucleic Acids Res 1995; 23:1557-60). Readthrough efficiency was estimated by calculating the ratio of luciferase activity to beta-galactosidase activity obtained with the test construct, with normalisation against an in-frame control construct. At least five independent transfection experiments were performed for each assay.

[0232] RNA Extraction and RT-qPCR

[0233] For the analysis of mRNA levels for p53 and its transcriptional target genes, Bax, the inventors extracted total RNA from HDQ-P1 cells that had or had not been treated with G418 (400 μM) or TLN68 (80 μM) for 48 hours (RNeasy Mini Kit, Qiagen). The RNA was treated with DNAse I (RNase-free DNase) and quantified with a Nanodrop spectrometer (ThermoScientific). The absence of RNA degradation was confirmed by agarose gel electrophoresis. The first-strand cDNA was synthesized from 2 μg of total RNA, with random primers and the SuperScript II Reverse Transcriptase (Invitrogen), as recommended by the manufacturer. Quantitative PCR was then carried out on equal amounts of the various cDNAs, with a CFX96 thermocycler (Biorad), and the accumulation of products was monitored with the intercalating dye FastStart Universal SYBRGreen Master (ROX) reagent (Roche). The inventors quantified mRNA levels relative to three reference mRNAs: RPL32, Hprt1 and HMBS. In each experiment, results are expressed relative to those for untreated cells, for which the value obtained was set to 1. Relative levels of gene expression were calculated at early stages of PCR, when the amplification was exponential and might, therefore, be correlated with the initial number of copies of the transcript. The specificity of quantitative PCR was checked by agarose gel electrophoresis, which showed that a single product of the desired length was produced for each gene. A melting curve analysis was also performed. Single product-specific melting temperatures were identified for each gene. For the quantification of each mRNA, three independent experiments (from biological replicates) were performed in triplicate. The inventors used the following oligonucleotides pairs for amplification:

p53 forward: 5′CCGCAGT CAGATCCTAGCG 3′ (SEQ ID NO:1) and reverse: 5′CCATTGCTTGGGACGGCAAGG 3′ (SEQ ID NO:2);
Bax forward: 5′GCTGTTG GGCTGGATCCAAG 3′ (SEQ ID NO:3) and reverse 5′ TCAGCCCATCTTCTTCCAGA (SEQ ID NO:4).

[0234] Western-Blot Analysis

[0235] HDQ-P1 cells (R213X) were treated with G418 (10 and 20 mM), or the selected molecules (TLN68 50 μM, TLN1399 50 μM, TLN236 30 μM and TLN309 40 μM), for 48 h. Cells were harvested by treatment with trypsin—EDTA (Invitrogen), lysed in 350 mM NaCl, 50 mM Tris—HCl pH 7.5, 1% NP-40, and protease inhibitor cocktail (Roche) and disrupted by passage through a syringe.

[0236] TLN1399 (out of the present invention) has the following structure:

##STR00041##

[0237] TLN236 (out of the present invention) has the following structure:

##STR00042##

[0238] Total proteins were quantified with Bradford reagent (Biorad) and extracts were denatured by incubation in Laemmli buffer for 5 minutes at 90° C. The inventors subjected 30 μg of total protein from HDQ-1 cells to SDS—PAGE in 4/10% Bis—Tris gels. Proteins were transferred onto nitrocellulose membranes, according to the manufacturer's instructions (Biorad). Membranes were saturated by overnight incubation in 5% skimmed milk powder in PBS, and incubated for 1 hour with the primary monoclonal antibody, DO-1 (N-terminal epitope mapping between amino acid residues 11 and 25 of p53; Santa Cruz Biotechnologies, 1/400) or a monoclonal antibody against mouse actin (Millipore, 1/2000). After three washes in PBS supplemented with 0.1% Tween, the membranes were incubated with the secondary antibody [horseradish peroxidase-conjugated anti-mouse IgG ( 1/2500)] for 45 minutes. The membranes were washed five times and chemiluminescence was detected with ECL Prime Western Blotting Detection Reagents (Amersham, GE Healthcare). The signal was quantified with ImageJ software.

[0239] TP53 Protein Activity Assays

[0240] The inventors investigated the transcriptional activity of the p53 protein in H1299, a p53-null cell line. Cells were cotransfected, by the JetPei method, with the p53BS-luc reporter plasmid containing the firefly luciferase gene downstream from seven p53 binding sites, the pCMVLacZ and the pCMVp53R213X containing the p53 cDNA interrupted by the stop mutation R213X. TLN68 (20, 50, 80 and 100 mM) was added to the medium just before transfection for a total of 20 hours of treatment. Protein extracts were then prepared and enzymatic activities were measured. Transfection with pCMVLacZ was used to normalise transfection efficiency, cell viability and protein extraction. At least six independent transfection experiments were performed for each set of conditions.

[0241] Results

[0242] Development and Validation of Stable Cell Line for Luciferase Assay

[0243] In order to identify new readthrough compounds that are not related to already known drugs the inventors choose an HTS approach. To select the drugs on their readthrough activity, the inventors first generated a stable mammalian cell line (from NIH3T3) by integrating a secreted Metridia luciferase reporter gene interrupted by a nonsense mutation (R213X). This system combines the advantages of a live-cell assay with the sensitivity of an enzyme-based system. The coding sequence of this gene is interrupted by a TP53 nonsense mutation R213X embedded in its own nucleotide context. This mutation has the advantage of presenting an easily measurable basal readthrough level.

[0244] The inventors reasoned that if the expression of luciferase is actually due to the readthrough of R213X stop codon, the addition of gentamicin should significantly increases the activity of luciferase. So they tested the 7 independent clones with a stable integration of the reporter gene with 1.6 mM of gentamicine. A significant induction was obtained for all of the clones, and they selected the one displaying the strongest induction (C14) for further development (FIG. 6).

[0245] Screening of Two Chemical Libraries

[0246] The inventors have screened 17,760 molecules (16,480 from ChemBridge library and 1,280 from Prestwick library) for their ability to efficiently stimulate stop codon readthrough. From this first screening they selected 465 molecules that display an increase factor greater or equal to 1.4 or a SSMD greater than or equal to 2 (FIG. 1.A). These first hits were then submitted to a second round of screening in the same conditions using duplicates for each molecule. The inventors retained 43 molecules for their ability to induce luciferase activity at least two fold. It is well-known that such screening can lead to the identification of a high number of false positive hits. To limit this and select only bona fide readthrough inducers, each of these 43 molecules were subjected to several independent assays to retain only the ones with a clear effect on stop codon readthrough.

[0247] Stop Codon Readthrough Quantification

[0248] Although highly sensitive and very convenient, the initial screening is subjected to several inherent biases. Indeed any molecule increasing the production of the Met-luciferase (mRNA transcription, stability, translation) or its secretion will lead to the identification of a false positive hit. To circumvent this the inventors checked the ability of each drug to stimulate readthrough in a dual reporter system to quantify stop codon readthrough efficiency (FIG. 1.B).

[0249] First this reporter system carries enzymatic activities (β-galactosidase and Firefly-luciferase) different from the one used in the initial screen. Second the β-galactosidase is used as an internal control to normalise expression level. So the use of this second reporter system eliminates many false positive hits potentially selected during the initial screen. For each tested molecule three independent measurements have been realised (FIG. 1.B). Only four molecules induce more than 2-fold change in PTC readthrough (data not shown). This second reporter system was very efficient to eliminate false positive hits, however it is still a reporter system using enzymatic activity. The inventors decided to use a more physiological system to test the last six potential hits.

[0250] Restoration of TP53 Protein Expression in HDQP-1 Cells

[0251] The inventors used the human cell line HDQ-P1 that carries the nonsense mutation R213X in its endogeneous TP53 gene. The expression of the full-length p53 is done by western-blot. The advantages of this third system are the use of an endogenous nonsense mutation and a direct visualisation of the final product induced by the potential hit (i.e. the full-length protein). Results are presented FIG. 2.A, and clearly reveal that two molecules stimulate the production of the full-length p53: TLN68 and TLN309. The level of induction is even higher to the one obtained in presence of G418 (one of the more efficient known readthrough inducers). Interestingly the inventors simultaneously observed an accumulation of the truncated form that could corresponds to the stabilisation of the TP53 mRNA through the inhibition of NMD. To confirm this they performed a RT-qPCR on the TP53-R213X mRNA and shown that TLN68 stabilises 6-fold this mRNA, whereas TLN309 shows a marginal stabilisation of the mRNA (FIG. 2.B).

[0252] A structural review of these two last candidates (FIG. 2.C) indicates that TLN309 is a 4-aminoquinoline similar to chloroquine that displays an autophagy-lysosomal inhibitory activity and promotes a ribosome biogenesis stress. The inventors decided to further pursue this analysis only with TLN68 that is a 2-guanidino quinazoline, named translectine (FIG. 2.C).

[0253] Functional TP53 Expression Restored by Stop Codon Readthrough

[0254] TLN68 is a very promising hit to restore the expression of the protein from a gene interrupted by a nonsense mutation. However restoring the expression of a full-length protein is a prerequisite to correct a defective gene, but this is not sufficient to warrant the functionality of the readthrough protein. Indeed, the inventors have shown that at least three tRNA (Tyr, Gln, Lys) can be used to readthrough UAG codon. Obviously the amino acid identity may strongly modify the activity of the restored full-length protein. So they tested with two different systems whether the full-length p53 expressed in presence of TLN68 is functional or not. First they used a plasmid carrying a luciferase gene under the control of a p53 promoter (FIG. 3.A), second they quantified by RT-qPCR the expression level of Bax mRNA that is one of the major targets of p53 (FIG. 3.B). The treatment by TLN68 induces a 3.5-fold increase of the p53-dependent luciferase, which is coherent with the 3-fold induction observed with bax mRNA level. This confirms that the readthrough p53 is at least partially active. Altogether these independent assays indicate that TLN68 is a promising readthrough inducer. During all these assays the inventors systematically used the same mutation (R213X). So, now that it is established that TLN68 induces stop codon readthrough on this premature stop codon, one may ask about its spectrum of action on various stop codons.

[0255] To the contrary, no protein is revealed in the presence of TLN236, suggesting that this drug does not act efficiently on stop codon readthrough.

[0256] Specificity of TLN68

[0257] To answer the question of TLN68 specificity the inventors selected the 40 more frequent premature nonsense mutations found in DMD gene responsible for Duchenne and Becker muscular dystrophy (Table 1; FIG. 4.A).

TABLE-US-00001 TABLE 1 Frequency of the 40 most frequent DMD   nonsense mutations Fre- SEQ   quency ID Name Nonsense mutation (%) NO: R145X AGC TGG GTC TGA CAA TCA ACT 0.75  5 S147X GTC CGA CAA TGA ACT CGT AAT 0.14  6 Q194X TCA GCC ACA TAA CGA CTG GAA 2.63  7 R195X GCC ACA CAA TGA CTG GAA CAT 0.61  8 Q267X GAA CAT TTT TAG TTA CAT CAT 2.63  9 R539X TTG GGA GAT TGA TGG GCA AAC 0.34 10 Q555X GTT CTT TTA TAA GAC ATC CTT 0.07 11 L654X TGG GAT ATT TAA CAT CAA AAA 0.07 12 E761X GAC TTA AAA TAA AAA GTC AAT 0.55 13 R768X GCC ATA GAG TGA GAA AAA GCT 0.82 14 R1051X AAT AAA CTC TGA AAA ATT CAG 0.55 15 W1075 AAG GAG GAA TAG CCT GCC CTT 2.63 16 El182X GAG TAT CTT TAG AGA GAT TTT 0.07 17 W1268X TGG GCA TGT TGA CAT GAG TTA 0.07 18 R1577X CGT AAG ATG TGA AAG GAA ATG 0.68 19 R1666X GTC ACC TCC TGA GCA GAA GAG 0.75 20 R1844X GAG AGA AAG TGA GAG GAA ATA 0.27 21 R1868X AGG TCT CAA TGA AGA AAA AAG 0.61 22 W1879X TCT CAT CAG TGA TAT CAG TAC 0.07 23 Y1882X TGG TAT CAG TAA AAG AGG CAG 0.14 24 W1956X AGC AAG CGC TAG CGG GAA ATT 0.07 25 R1967X GCT CAG TTT TGA AGA CTC AAC 0.61 26 E2035X TTT AAG CAA TAG GAG TCT CTG 5.26 27 R2095X TAC AAG GAC TGA CAA GGG CGA 0.55 28 E2286X ATA AGC CCA TAA GAG CAA GAT 2.63 29 Q2526X AGG CGT CCC TAG TTG GAA GAA 0.07 30 R2553X ATT ACG GAT TGA ATT GAA AGA 0.82 31 Q2574X CGG AGG CAA TAG TTG AAT GAA 2.63 32 K2791X GAA CTT CGG TAA AAG TCT CTC 2.63 33 R2870X GAG ACT GTA TGA ATA TTT CTG 0.82 34 R2905X CGG CTT CTA TGA AAG CAG GCT 0.34 35 W2925X TCC GCT GAC TGA CAG AGA AAA 0.14 36 R2982X AAG GCA CTT TGA GAA GAA AAT 0.55 37 R3034X GTC GAG GAC TGA GTC GTC CAG 0.55 38 S3127X TTG AGC CTG TGA GCT GCA TGT 2.63 39 R3190X GAT ACG GGA TGA ACA GGG AGG 0.89 40 R3345X TTT TCT GGT TGA GTT GCA AAA 0.48 41 R3370X GAA GAT GTT TGA GAC TTT GCC 1.02 42 R3381X AAC AAA TTT TGA ACC AAA AGG 1.57 43 R3391X AAG CAT CCC TGA ATG GGC TAC 1.5 44

[0258] Each nonsense mutation (+9 nucleotides in each side) has been cloned into the dual reporter previously used. Stop codon readthrough efficiency has been quantified either without drug or in presence of gentamicin or TLN68, in 6 independent experiments. The results shown FIG. 4.B indicate that TLN68 promotes stop codon readthrough not only onto a large variety of sequences (28 out of 40 tested) but also outperforms gentamicin on this set of sequences. The inventors conclude TLN68 is active onto a broad variety of sequences, although its action is sequence-dependent as most of the known readthrough inducers.

[0259] TLN68 has an Additive Effect with Gentamicin to Induce Readthrough

[0260] The inventors then test effect of a combined treatment of TLN68 with one of the most used readthrough molecule, gentamicin. NIH3T3 cells were transfected with dual reporter vector harboring R213X mutation and treated 24 h with each drug separately or together. Gentamicin and TLN68 induce a readthrough level of 10 and 11% respectively and the combination of the two drugs allow a readthrough level of 21% suggesting an additive effect between these molecules (FIG. 5).

[0261] TLN68 Effect on Translation is Specific of Readthrough Events

[0262] To determinate if TLN68 alter global translational fidelity or if it is specific of readthrough event the inventors tested capacity of this molecule to induce frameshifting. They used a frameshift target which have already been used to study the minus1 frameshift signal of the retrovirus HIV-1 (Stahl et al, 1995). This 54 nucleotide sequence was inserted in the dual reporter vector and used to transfect NIH3T3 cells. Transfected cells were treated or not with TLN68 at 80 μM. Results demonstrate that TLN68 has no impact on translational frameshifting suggesting that this molecule does not have an overall impact on translation fidelity but might rather act specifically on readthrough (FIG. 8).

[0263] TLN68 has Moderate Toxicity on Mammalian Cell Lines

[0264] The inventors tested the effect of TLN68 on viability of NIH3T3, HDQ-P1 and HeLa cell lines by using a tetrazolium salt which is cleaved only by metabolically active cells. Cells were treated or not with a range of TLN68 doses during 24 h and viability was measured. FIG. 9 shows that cell viability slowly decreases as the TLN68 concentration increases to reach 80% of viable cells for NIH3T3 which are the most resistant cell line cells and 70% for HDQ-P1 cells at the maximal dose used.

Example 3: Tests of the Ability of Different Compounds According to the Invention to Induce PTC Readthrough

[0265] Protocol:

[0266] Different compounds of the invention have been tested for their ability to induce readthrough on the nonsense mutation W1268X or R213X in HeLa cells.

[0267] To this aim the inventors used the pAC99 dual reporter plasmid. This dual reporter is used to quantify stop-codon readthrough through the measurement of luciferase and beta-galactosidase (internal normalisation) activities. Readthrough levels for nonsense mutations were analysed in the presence or absence of tested molecules. Cells are seeded in a 6-well plate. The next day, cells are transfected with the reporter plasmid using the JetPei reagent (Invitrogen). The following day, they are rinsed with fresh medium, with or without readthrough inducers. Cells were harvested 24 hours later, with trypsin—EDTA (Invitrogen), lysed with Glo lysis buffer (Promega) and beta-galactosidase and luciferase activities were assayed as previously described. Readthrough efficiency was estimated by calculating the ratio of luciferase activity to beta-galactosidase activity obtained with the test construct, with normalisation against an in-frame control construct. At least five independent transfection experiments were performed for each assay.

[0268] Results:

[0269] The results are as follows:

TABLE-US-00002 Inducting factor of Inducting factor of readthrough as readthrough as compared to compared to untreated cells untreated cells IC50 in (measured on (measured on HeLa cells nonsense mutation nonsense mutation Molecule (in μM) W1268X) R213X) TLN68 9 14 EC-18 60 69 50 EC-141 40 16 — EC-85 20 15 16 EC-11 40 11 — EC-35 20 11 14 EC-30 80 10 — EC-265 60 19 40 EC-288 20 68 55 EC-130 10 21 33 EC-335 10 33 38

[0270] As a conclusion, the compounds of the invention are PTC readthrough promoters which are very efficient.

Example 4: Ribosome Profiling Experiment Using TLN68

[0271] Objective

[0272] It is important to demonstrate that TLN68 acts specifically on premature termination codons (PTC) and has no effect on natural stop codons. To answer this question, the inventors performed a Ribosome profiling (RiboSeq) experiment that allows genome-wide mapping of all ribosome footprints. They tested three different conditions: untreated cells, 0.5 mg/ml G418, 80 mM TLN468 for 24 h. Each condition has been performed in triplicate for statistical reasons.

[0273] Ribosome Profiling Experiments

[0274] HeLa cells were plated at JO at 1 million cells per plate with 10 ml of MEM supplemented with 10% fœtal Calf Bovine Serum, 1% glutamine, non-essential amino acids and antibiotics-antimycotic (Gibco). At J+1, 80 mM TLN468 is added. At J+2, cells are collected. The medium was removed and then directly laid on liquid nitrogen bath, put to −70° C. before being scratched. The cells were collected with addition of Polysome Extraction Buffer (10mMTris CH.sub.3COONa pH7.6; 10 mM (CH.sub.3COO).sub.2Mg; 10 mM NH.sub.4Cl; 1% Triton; 2 mM DTT).

[0275] Polysomes were extracted with the addition of 2× complete EDTA-free protease inhibitor Roche and 1U Rnase Inhibitor Murine (Biolabs). The Ribosome Protected Fragments (RPF) were generated by 1 h of digestion with 15 U Rnase I (Ambion)/OD260.sub.nm at 25° C.). Monosomes are separated on 24% sucrose cushion at +4° C. then treated by DNasel. RNA is extracted by phenol at 65° C., CHCL.sub.3 and precipitated by ethanol in 0.3M CH.sub.3COONa pH5.2 then loaded on a 17% acrylamide-bisacrylamide (19:1) gel with 7 M urea and 1× Tris-acetate-EDTA (TAE). RPF at 28- to 34-nt are excised from gel and precipitated in ethanol in presence of glycogen. RPF are depleted from ribosomal RNA with the Ribo-zero Human kit (Illumina) following the manufacturer's recommendations. The RPF libraries are made using the Transcriptome TruSeq modified kit and sequenced on NextSeq 500 High Single Read 75 bases.

[0276] Results

[0277] Data are analyzed by homemade scripts. Cells are flash-frozen in liquid nitrogen and polysomes extracted. Once the data obtained, the inventors informatically removed all contaminating rRNA fragments and used a homemade docker package (RiboDoc) to map reads on the human transcriptome using the human genome (hg38) as reference. They obtained a total of 122 497 388, 160 481 925 and 51 860 606 uniquely mapped reads for untreated, G418 and TLN68 respectively. They first checked the ribosomal footprints length distribution. The results show that the majority of footprints are 30 nucleotides (30nts) long for the untreated cells (Hela) and cells treated with G418, whereas the footprints are 29 nucleotides (29nts) long for cells treated with TLN68. This length is perfectly what is expected for a ribosome footprint that must be comprised between 28-31 nucleotides. The observed difference between TLN68 and the other samples probably simply reflects variations during RPF preparation, without biological meaning.

[0278] The inventors selected only the 29nts and 30nts long ribosome footprints for TLN68 and HeLa/G418 samples respectively to perform a metagene analysis in which all annotated CDS are pooled and the number of ribosome footprint 5′ ends at each nucleotide position is determined. This step allows confirming that the analysis is performed on actively translating ribosomes. Indeed, the inventors can clearly observe a signal with a periodicity of 3 (data not shown). This periodicity represents ribosomes moving codon by codon along the mRNA. The inventors classically observe a peak at the start and stop codons because initiation and termination steps are kinetically slower than the elongation steps, promoting the accumulation of ribosomes at the start and stop codons. They can also observe weak signals upstream the start codon in the 5′UTR, which could correspond to uORFs. Immediately downstream of the stop codon the translation signal disappears, except in samples treated with G418. This signal represents genome-wide stop codon readthrough promoted by G418. The signal quickly disappears because readthrough efficiency is low and ribosomes quickly encounters a second stop (the median distance between the normal stop codon and the next in-frame codon is 18 codons for the human genome). Interestingly the inventors did not observe any genome-wide readthrough for the untreated cells and cells treated with TLN68.

[0279] This indicates that TLN68 specifically targets premature stop codons and has no effect on natural terminating codons.

[0280] The same experiment has been performed using other compounds of formula (I) or (II) or one of their salts of the invention, and preferably compound EC-18 or EC-288, instead of TLN68.