Antisense oligonucleotides against nucleotide deamination in the treatment of Stargardt disease

Abstract

The present invention relates to RNA editing oligonucleotides (EONs) capable of effecting specific editing of a target nucleotide (adenosine) in a target RNA molecule in a eukaryotic cell, wherein said oligonucleotides are for use in the treatment of Stargardt's disease, more preferably for deamination of a target adenosine present in ABCA4 pre-mRNA or ABCA4 mRNA.

Description

[Technical Field]

[0001] The present invention relates to the fields of medicine and biotechnology. The present invention relates to an antisense oligonucleotide applicable to the deamination of nucleotides through the RNA editing process using (endogenous) ADAR enzymes by site-specifically targeting alterations (such as G>A mutations) in ABCA4 (pre-) mRNA in patients suffering from Stargardt disease. [Background technology]

[0002] Stargardt disease (STGD or STGD1) is the most common hereditary macular dystrophy, causing progressive impairment of central vision. It typically develops in childhood or young adulthood, with the lowest incidence in later adulthood. Later onset is generally associated with a better prognosis. The disease prevalence is 1 in 8,000–10,000 individuals and is inherited in an autosomal recessive manner associated with disease-causing mutations in the gene encoding the photoreceptor-specific ATP-binding cassette, subfamily A, member 4 protein (ABCA4, sometimes called ABCR). This protein contains 2,273 amino acids and is primarily expressed in the retina (photoreceptor cells and retinal pigment epithelium (RPE)), where it is localized in the periphery and cone outer segment discs. It is thought to flip N-retinylidene-phosphatidylethanolamine from the luminal side of the photoreceptor disc to the cytosolic side. Stargardt disease is closely associated with massive deposition of lipofuscin inclusions in the RPE, failure to remove toxic substances, and significant loss of photoreceptor cells. The di-retinoid pyridinium-ethanolamine, the main component of lipofuscin, is formed when ABCA4 is absent or dysfunctional. Indeed, multiple reports have identified ABCA4 as the gene underlying Stargardt disease, demonstrating a large number (approximately 1,000) of disease-causing variants, more than half of which have only been described once. Biallelic variants in ABCA4 have been identified in approximately 75% of Stargardt disease cases and approximately 30% of patients with autosomal recessive Camrod dystrophy (CRD). Most mutations are missense, followed by nonsense mutations, small insertions / deletions, and mutations affecting RNA splicing. An unusually high proportion (approximately 30%) of Stargardt disease cases from Northern Europe and the United States are the result of a single ABCA4 variant. The third most common ABCA4 variant, c.5461-10T>C, located in intron 38, is known to cause a severe form of Stargardt disease due to exon 39 skipping or exon 39 + exon 40 skipping in the ABCA4 mRNA.Skipping exon 39 results in a frameshift deletion of 124 nucleotides, whereas double skipping of exons 39 and 40 results in a frameshift deletion of 254 nucleotides. It is estimated that approximately 7000 individuals with Stargardt disease in Western countries are affected by this mutation.

[0003] Currently, the three main interventional pathways for treating Stargardt disease are stem cell therapy, gene replacement therapy, and various drug approaches. A relatively new therapeutic development for treating inherited eye diseases is the use of antisense oligonucleotides (AONs) that target precursor mRNA (pre-mRNA) transcribed from mutant genes. AONs are generally small polynucleotide molecules (16–25 mers) whose sequences are complementary to those of the target pre-mRNA molecule, thereby disrupting splicing. The postulated mechanism is that when the AON binds to its complementary target sequence, the targeted region within the pre-mRNA interferes with splicing factors, which in turn leads to altered splicing. Therapeutically, such methodologies can be used in two ways: a) redirecting normal splicing of genes in which the mutation activates cryptic splice sites, and b) skipping exons harboring (proteolytic) mutations in such a way that the mRNA reading frame remains intact and a functional (partially or fully) protein is produced. Both methods have already been successfully applied to patients. Regarding eye diseases, AONs have shown promise for the treatment of Leber congenital amaurosis (LCA) (WO 2012 / 168435; WO 2013 / 036105; WO 2016 / 034680; WO 2016 / 135334). Furthermore, WO 2016 / 005514 discloses exon-skipping AONs for targeting USH2A pre-mRNA to skip exons 13, 50, and PE40 and / or retain exon 12 for the treatment, prevention, or delay of Usher syndrome type II. WO 2015 / 004133 discloses the use of AONs to stimulate exon 10 skipping from ABCA4 pre-mRNA for the treatment of Stargardt disease.WO 2018 / 189376 discloses antisense oligonucleotides that inhibit the skipping of exons 39 and 40 caused by the c.5461-10T>C mutation. WO 2018 / 109011 discloses AONs to prevent the inclusion of pseudoexons erroneously introduced into mRNA due to some intronic mutations.

[0004] As shown, when considering treatments involving splice modulation, prevention of exon skipping, or prevention of pseudoexon inclusion, the resulting mRNA must be in-frame so that the translated protein is functional, or at least partially functional, and not prematurely terminated. However, there are many mutations in the human ABCA4 gene that cannot be resolved by splice modulation. For example, the resulting transcript may be out of frame or the resulting protein may lack essential parts required to perform its function. Therefore, although significant efforts have been made to treat Stargardt disease by introducing antisense oligonucleotides that modulate splicing, many Stargardt disease patients will not benefit from these efforts because they carry different types of mutations in the ABCA4 gene. The c.5882G>A mutation in exon 42 is one of the more common ABCA4 mutations in Western countries, with an estimated prevalence of 10,000-15,000 individuals (Lewis et al. Am J Hum Genet 64:422-434, 1999). This mutation leads to a substitution of the amino acid glycine (G; codon: GGA) at position 1961 of the ABCA4 protein with glutamic acid (E; codon: GAA). This amino acid is located in the nucleotide-binding domain 2 (NBD2), which is essential for providing energy for binding substrates for transport. Exon 42 encodes a portion of NBD2 that, although in-frame if skipped, cannot be disrupted for proper ABCA4 protein function. Biochemical evaluation showed that ABCA4 proteins carrying this substitution had reduced ATPase activity (Sun et al. Nature Genetics 26:242-246, 2000), also indicating that even small changes in NBD2 can cause protein dysfunction. Other mutations that can be targeted by RNA editing, as will be understood by those skilled in the art, are mutations that appear in exons that cannot be skipped because the exon is out of frame with the surrounding exons. Summary of the Invention [Means for solving the problem]

[0005] The present invention relates to an RNA editing oligonucleotide (EON) that can form a double-stranded complex with a target RNA molecule, and when the EON forms a complex with the target RNA molecule, it can recruit an "RNA-acting adenosine deaminase" (ADAR) enzyme and form a complex with the ADAR, thereby allowing the ADAR enzyme to deaminate the target adenosine in the target RNA molecule, wherein the EON does not form an intramolecular loop structure, and the target RNA molecule is human ABCA4 pre-mRNA or mRNA, or a part thereof. In one embodiment, the EON comprises a central triplet of three consecutive nucleotides, and the nucleotide directly opposite the target adenosine is the central nucleotide of the central triplet, which is cytidine. Preferably, the ADAR enzyme that is recruited to form a complex with the dsRNA complex is ADAR2. In particularly preferred embodiments, one, two, or three nucleotides in the central triplet contain modifications, provided that the central nucleotide does not have a 2'-O-methyl (2'-OMe) or 2'-methoxyethoxy (2'-MOE) modification in the sugar moiety. Furthermore, the EON preferably contains at least one non-naturally occurring internucleoside linkage modification, as outlined herein. In preferred embodiments, the target adenosine is part of a premature stop codon in the human ABCA4 pre-mRNA or mRNA; any one of the G>A mutations provided in Table 1; or, more preferably, the c.5882 G>A mutation in exon 42 of the human ABCA4 gene. In a preferred embodiment, the present invention relates to an EON according to the present invention, comprising or consisting of a sequence selected from the group consisting of SEQ ID NOs: 1, 2, 3, 4, 5, 6, 7, 8, 9, and 11. In a preferred embodiment, the EON comprises or consists of the sequence of SEQ ID NO: 8 or 9. In an even more preferred embodiment, the EON comprises or consists of the sequence of SEQ ID NO:9.

[0006] The present invention further relates to a pharmaceutical composition comprising an EON according to the present invention and a pharmaceutically acceptable carrier. The present invention also relates to a composition comprising a set of two single-stranded antisense oligonucleotides (AONs) for use in deaminating a target adenosine in human ABCA4 pre-mRNA or mRNA, or a portion thereof, wherein one AON is an EON according to the present invention and the other AON is a "helper AON," which is complementary to a stretch of nucleotides in human ABCA4 pre-mRNA or mRNA that is different from the stretch of nucleotides complementary to the EON, and the helper AON has a length of 16 to 22 nucleotides. In a preferred embodiment, the composition according to the present invention comprises, as the helper AON, an oligonucleotide comprising or consisting of the sequence of SEQ ID NO: 10, wherein the EON is an oligonucleotide comprising or consisting of the sequence of SEQ ID NO: 9.

[0007] The present invention also relates to an EON or composition according to the present invention for use in treating Stargardt's disease.The present invention also relates to a method for deaminating at least one specific target adenosine present in a target RNA molecule in a cell, wherein the target RNA molecule is human ABCA4 pre-mRNA or mRNA, or a part thereof, and the target adenosine that needs to be deaminated is preferably one of the target adenosines outlined herein, and the deamination method comprises the following steps: administering the EON or composition according to the present invention to a cell; allowing the cell to take up an oligonucleotide; annealing the oligonucleotide to the target RNA molecule; allowing a mammalian ADAR enzyme that contains a natural dsRNA-binding domain such as that found in wild-type enzymes to deaminate the target adenosine in the target RNA molecule to inosine; and optionally confirming the presence of inosine in the target RNA molecule. [Brief explanation of the drawings]

[0008] [Figure 1-1] Figure 1(A) shows the complementarity of the EONs designated "ABCA4-1 through ABCA4-20" used in the examples of this disclosure (ABCA4-1, 2, 3, 4, and 5 are SEQ ID NOs: 1, 2, 3, 4, and 5, respectively; ABCA4-6 through 12 are SEQ ID NO: 6; ABCA4-13 through 16 are SEQ ID NO: 7; and ABCA4-17 through 20 are SEQ ID NO: 8; all 3' to 5'). The human ABCA4 target (pre-)mRNA is shown above (5' to 3'; SEQ ID NO: 12). The target adenosine A in the target sequence (in exon 42) is shown in bold, and the 5' end of downstream intron 42 is underlined, preceded by the 3' end of exon 42. ABCA4-13 through 16 are complementary to the exon 42 / intron 42 boundary. [Figure 1-2]Figure 1(B) shows the complementarity set of EONs designated ABCA4-21-29 (here, 3' to 5'). ABCA4-21-27 (SEQ ID NO: 9), together with ABCA4-28 (SEQ ID NO: 10), each form a split EON set, whereas ABCA4-29 (SEQ ID NO: 11) is an EON with multiple mismatches. The target adenosine A in the target sequence is shown in bold, again underlined at the 5' end of downstream intron 42 and preceded by the 3' end of exon 42. 2'-O-Me RNA is shown in lowercase. 2'-O-Me nucleotides connected to each other by phosphorothioate linkages are italicized and in lowercase. DNA nucleotides connected to each other by phosphorothioate linkages are underlined and in uppercase italics. 2'-MOE modified nucleotides are italicized and in uppercase. 2'-MOE modified nucleotides connected to each other by phosphorothioate linkages are bold, italicized, and in uppercase. The methylphosphonate (MeP) modified adenosine is boxed in gray and is at linkage position -1 from the middle nucleotide of the central triplet (see Figure 5 in WO 2020 / 201406 for linkage numbering). The middle nucleotide of the central triplet is opposite the target adenosine. The 2'-F modified adenosine is bolded and boxed in gray. Mismatches are boxed in black. ABCA4-5 contains a hairpin sequence (lowercase). [Figure 2-1] Figure 2 shows the results of a biochemical assay using ABCA4-1 to ABCA4-29 (excluding ABCA4-5 and ABCA4-28) from Figure 1 in an RNA editing reaction with purified hADAR2, as outlined in the accompanying Examples. (A) to (G) show the results for four different EONs, respectively, for clarity. [Figure 2-2]Figure 2 shows the results of a biochemical assay using ABCA4-1 to ABCA4-29 (excluding ABCA4-5 and ABCA4-28) from Figure 1 in an RNA editing reaction with purified hADAR2, as outlined in the accompanying Examples. (A) to (G) show the results for four different EONs, respectively, for clarity. [Figure 3] Figure 3 shows the Pyromark sequencing results of RNA editing against the ABCA4 target sequence using EON ABCA4-5, 6, 8, 20, 21+28, 22+28, 23+28, 24+28, and ABCA4-21, 22, 23, 24, and 26 separately in a cell-based assay using midigenes and a plasmid overexpressing ADAR2. WT refers to transfection with midigenes carrying the wild-type target sequence, while "mutant" refers to transfection with the mutant midigene alone without any EON. RT is a negative control for reverse transcriptase. All transfections of EONs were accompanied by transfection with the mutant midigene. [Figure 4] Figure 4 shows the results of ddPCR of the cell-based assay using midigenes and ADAR2 overexpression. Using ddPCR, we observed very low RNA editing efficiency in EONs close to the boundary between exon 42 and intron 42, whereas short EONs complementary to sequences far from the exon / intron boundary showed significant levels of RNA editing. [Figure 5] Figure 5 shows the results of exon skipping confirmation experiments, revealing that transfection of EONs ABCA4-5~16 and ABCA-29 resulted in a certain level of exon 42 skipping, which explains the lower RNA editing efficiency observed in these EONs (closer to the exon 42 / intron 42 boundary) in pyromark and ddPCR sequencing results. DETAILED DESCRIPTION OF THE INVENTION

[0009] The present inventors sought a means other than splice regulation to correct defects in (pre-)mRNA within ABCA4. They contemplated using another type of RNA repair, often referred to as "RNA editing." RNA editing is a natural method by which eukaryotic cells alter the sequence of their RNA molecules, often in a site-specific and precise manner, thereby expanding the repertoire of RNA encoded by their genomes by several orders of magnitude. RNA editing enzymes have been described in eukaryotic species throughout the animal and plant kingdoms, and these methods play a critical role in managing cellular homeostasis in metazoans, from the simplest organisms such as Caenorhabditis elegans to humans. Examples of RNA editing are the conversion of adenosine (A) to inosine (I) and the conversion of cytidine (C) to uridine (U) by enzymes called adenosine deaminase and cytidine deaminase, respectively. The most widely studied RNA editing system is the adenosine deaminase enzyme, a multidomain protein containing a recognition domain and a catalytic domain. The recognition domain recognizes specific double-stranded RNA (dsRNA) sequences and / or conformations, while the catalytic domain converts adenosines to inosines at nearby, predefined positions in the target RNA by deamination of the nucleobase. Inosines are read as guanosines by the cellular translation machinery, meaning that edited adenosines can recode protein sequences when present in the coding region of mRNA or pre-mRNA. Therefore, RNA editing by adenosine deamination is a perfect method for repairing not only G>A mutations but also other mutations in which A-to-G conversion allows for the production of functional proteins. Adenosine deaminases are part of a family of enzymes called adenosine deaminases acting on RNA (ADARs), including the human deaminases hADAR1, hADAR2, and hADAR3.

[0010] The use of oligonucleotides to apply adenosine deaminase to edit target RNA is known in the art.Montiel-Gonzalez et al. (Proc Natl Acad Sci USA 2013, 110(45):18285-18290) described the editing of target RNA using a genetically engineered fusion protein comprising the adenosine deaminase domain of hADAR2 protein fused to the bacteriophage lambda N protein that recognizes boxB RNA hairpin sequence.The drawback of this method in therapeutic settings is the need for fusion protein.This is the main hurdle, but it requires cells to be transduced with fusion protein or to transfect target cells with a nucleic acid construct that encodes the adenosine deaminase fusion protein that is engineered for expression. Vogel et al. (2014. Angewandte Chemie Int Ed 53:267-271) disclosed the editing of RNA encoding eCFP and Factor V Leiden using a genetically engineered fusion protein (engineered O6-alkylguanosine-DNA-alkyltransferase) containing a benzylguanosine-substituted guide RNA and the adenosine deaminase domain of ADAR1 or 2 (lacking the dsRNA binding domain) genetically fused to a SNAP-tag domain. This system has similar drawbacks to the engineered ADARs described by Montiel-Gonzalez et al. (2013).Woolf et al. (1995. Proc Natl Acad Sci USA 92:8298-8302) described a simpler approach using relatively long, single-stranded antisense RNA oligonucleotides (25-52 nucleotides in length). In this case, due to the double-stranded nature of the oligonucleotide hybridizing with the target RNA, longer oligonucleotides (34-mer and 52-mer) could promote target RNA editing by endogenous ADARs. However, these oligonucleotides appeared to function only in cell extracts or in amphibian (Xenopus) oocytes via microinjection, and they suffered from a severe lack of specificity: nearly every adenosine in the target RNA strand that was complementary to the antisense oligonucleotide was edited. Woolf et al. (1995) did not achieve deamination of a specific target adenosine in the target RNA sequence. Nearly every adenosine opposite the unmodified nucleotide in the antisense oligonucleotide was edited through a process sometimes referred to as "random editing." WO 2016 / 097212 discloses RNA-editing oligonucleotides characterized by a sequence complementary to a target RNA sequence (the "targeting portion") and by the presence of a stem-loop structure (the "recruitment portion"). WO 2017 / 220751 discloses RNA-editing oligonucleotides lacking a recruitment portion but having a chain of nucleotides complementary to a target region for specific editing of a single adenosine, where the oligonucleotide contains one or more mismatches, wobbles, and / or bulges in combination with specific chemical modifications. Very specific locations for specific chemical modifications in such RNA-editing oligonucleotides were further disclosed in WO 2018 / 041973, WO 2019 / 158475, and WO 2019 / 219581.WO 2019 / 005884 discloses a system for targeting adenosine deamination, in which the targeting system comprises a targeting domain linked to adenosine deaminase or its catalytic domain (a CRISPR system comprising a CRISPR effector protein (e.g., Cas13) and a guide molecule, which is generally an oligonucleotide having a sequence complementary to the target sequence). The present invention relates to "naked" oligonucleotides or oligonucleotides expressed from (viral) vectors, but in either manner, do not comprise a targeting domain that generates a loop structure for complexing with an effector protein, such as those in CRISPR systems. While the RNA-editing oligonucleotides disclosed in WO 2016 / 097212 and WO 2019 / 005884, which have an intramolecular loop structure that recruits or connects an effector protein, can be considered first-generation RNA-editing oligonucleotides, the present invention relates to RNA-editing oligonucleotides that do not comprise such an intramolecular loop structure and therefore can be considered second-generation RNA-editing oligonucleotides. WO 2019 / 005884 discloses RNA-editing oligonucleotides for use in targeting the G>A mutation in the USH2A gene, which causes Usher syndrome type II, a degenerative disease of the retina and inner ear. WO 2019 / 005884 also discloses RNA-editing oligonucleotides that form loop structures and also complex with (mutated) ADAR enzymes. The present invention relates to RNA-editing oligonucleotides that do not contain sequences that form loop structures and do not form complexes (covalently or non-covalently) with ADAR enzymes or mutant ADAR enzymes, but rather utilize endogenous deaminase enzymes, such as ADAR, already present in target cells. The RNA-editing oligonucleotides of the present invention recruit such enzymes intracellularly after administration to cells or tissues, and can also recruit ADAR enzymes in biochemical assays, as shown in the accompanying examples.In a preferred embodiment, the deaminase is an ADAR enzyme, more preferably ADAR2, which is already present at endogenous levels in the target cell and does not need to be co-administered, such as via an expression vector or otherwise.

[0011] Most of the prior art on RNA editing concerns the general applicability of such phenomena to any type of disease or genetic disorder, where editing specific targeted adenosines to inosines is necessary to restore translation (if the adenosine was part of a stop codon) and / or to repair RNA if the adenosine was part of a codon that alters the protein and causes the genetic disease. Prior art documents did not specifically clarify the application of RNA editing oligonucleotides in ocular diseases such as Stargardt's disease, where genetic mutations are the cause of the disorder, or how this should be done specifically. In contrast, a large amount of prior art has accumulated demonstrating the utility of antisense oligonucleotides in downregulating protein expression or affecting splicing (e.g., WO 2012 / 168435, WO 2013 / 036105, WO 2016 / 005514, WO 2016 / 034680, WO 2016 / 138353). (See, for example, International Publication Nos. WO 2016 / 135334, WO 2017 / 060317, WO 2017 / 186739, WO 2018 / 055134, WO 2015 / 004133, WO 2018 / 189376, WO 2018 / 109011, and U.S. Patent No. 9,353,371.) To the best of the inventors' knowledge, the use of RNA-editing oligonucleotides to deaminate specific adenosines in genes that cause eye defects, and the application of such RNA-editing oligonucleotides to the treatment of eye disorders, particularly Stargardt's disease, has not been published.

[0012] The present invention relates to RNA-editing oligonucleotides (collectively abbreviated herein as "EONs") and their use in the treatment of ocular diseases, particularly Stargardt disease. The EONs of the present invention target specific adenosines in ABCA4 pre-mRNA or mRNA, deaminating them to inosines, which are then read as guanosines in translation. It should be noted that adenosine (A) itself is not necessarily the mutation that causes Stargardt disease, but may be part of a premature stop codon that is responsible for the disease (e.g., because a shorter ABCA4 protein product is produced), and that guanosine (G) or thymidine (T) are true mutations, for example. Deamination of adenosine to inosine may result in a wild-type protein or a protein with an altered amino acid in place of the stop codon (and altered relative to the original codon in the wild-type mRNA), but should allow continued translation.

[0013] The present invention relates to an EON capable of forming a double-stranded complex with a target RNA molecule, wherein the EON, upon complexing with the target RNA molecule, can recruit an "RNA-acting adenosine deaminase" (ADAR) enzyme and form a complex with the ADAR, thereby enabling the ADAR enzyme to deaminate a target adenosine in the target RNA molecule. The EON does not form an intramolecular loop structure, and the target RNA molecule is human ABCA4 pre-mRNA or mRNA, or a part thereof. It should be noted that when the EON is administered to a cell, tissue, or (human) subject, the ADAR, preferably ADAR2, is not complexed with the EON. It should be noted that the EON can be delivered to cells directly (naked) or via expression from a viral vector or other expression vector. Once inside the cell, the EON targets the ABCA4 target pre-mRNA or mRNA and hybridizes with its target. Due to its mismatch and the selective modification and content of EON, this double-stranded complex can attract (or recruit) endogenous ADAR enzymes, which then deaminate the target adenosine opposite the central nucleotide of the central triplet, preferably cytidine. This allows specific RNA editing of adenosine in human ABCA4 pre-mRNA or mRNA. When administered in a "naked" form, the EON of the present invention is not bound or (non-)covalently bound to any protein before entering cells. Nor does it form an intramolecular loop structure for binding to other effector proteins, as seen in the art for ADAR or CRISPR / Cas systems. The EON of the present invention comprises a central triplet of three consecutive nucleotides, and the nucleotide directly opposite the target adenosine is the central nucleotide of the central triplet, preferably cytidine. Preferably, the ADAR enzyme attracted by the dsRNA complex is ADAR2.As shown in the accompanying examples, those skilled in the art can design biochemical assays that are representative of the in vivo situation in that they involve contacting an EON with its target sequence, recruiting an ADAR enzyme, and causing RNA editing of a specific target adenosine in an ABCA4 target molecule. In preferred embodiments, one, two, or three nucleotides in the central triplet contain modifications, provided that the central nucleotide does not have a 2'-O-methyl (2'-OMe) or 2'-methoxyethoxy (2'-MOE) modification in the sugar moiety. In another preferred embodiment, the modifications of the nucleotides in the central triplet are selected from the group consisting of deoxyribose (DNA), unlocked nucleic acid (UNA), and 2'-fluororibose. As further outlined herein, DNA is therefore considered a chemical derivative of RNA. Most (but certainly not all, in some embodiments) nucleotides in the EONs of the present invention are RNA, which may be modified with non-naturally occurring substituents, as further detailed herein.The EONs of the present invention are preferably phosphorothioates, chirally pure phosphorothioates, Rp phosphorothioates, Sp phosphorothioates, phosphorodithioates, phosphonoacetates, tphosphonoacetates, phosphonacetamides, thiophosphonacetamides, phosphorothioate prodrugs, S-alkylated phosphorothioates, H-phosphonates, methyl phosphonates, methyl phosphonothioates, methyl phosphates, methyl phosphorothioates, ethyl phosphates, ethyl phosphorothioates, boranophosphates, boranophosphorothioates, methyl boranophosphates, methyl The EON comprises at least one non-naturally occurring internucleoside linkage modification selected from the group consisting of boranophosphorothioate, methylboranophosphonate, methylboranophosphonothioate, phosphorylguanidine, methylsulfonylphosphoramidate, phosphoramidite, phosphonamidite, N3'→P5' phosphoramidate, N3'→P5' thiophosphoramidate, phosphorodiamidate, phosphorothiodiamidate, sulfamate, dimethylenesulfoxide, sulfonate, triazole, oxalyl, carbamate, methyleneimino, thioacetamide, and derivatives thereof. Most preferred is the use of phosphorothioate linkages, although the EON need not be fully phosphorothioated. For example, in certain embodiments, at least 2, 3, 4, 5, or 6 terminal nucleotides at the 5' and 3' ends of the EON are linked by phosphorothioate linkages, and preferably the 5 terminal nucleotides at the 5' and 3' ends are linked by phosphorothioate linkages. Additional nucleotides closer to the central triplet may also be connected by non-naturally occurring linkages such as phosphorothioates. In a preferred embodiment, the EON is chemically modified to stabilize the EON against degradation by RNases in cells and other environmental conditions in vivo.To this end, one or more nucleotides in the EON outside the central triplet preferably contain mono- or di-substitutions at the 2', 3', and / or 5' positions of the sugar, where the substitutions are selected from the group consisting of: -OH; -F; substituted or unsubstituted, straight-chain or branched lower (C1-C10) alkyl, alkenyl, alkynyl, alkaryl, aryl, or aralkyl, optionally interrupted by one or more heteroatoms; O-, S-, or N-alkyl; O-, S-, or N-alkenyl; O-, S-, or N-alkynyl; O-, S-, or N-aryl; O-alkyl-O-alkyl; -methoxy; -aminopropoxy; -methoxyethoxy; -dimethylaminooxyethoxy; and -dimethylaminoethoxyethoxy. Other possible chemical modifications that can be introduced into the EONs of the invention, particularly those of the sugars, bases, and linkers, are discussed further herein.

[0014] RNA editing by the system outlined by the present invention preferably targets a single adenosine in the target molecule (the transcribed pre-mRNA or mRNA of the human ABCA4 gene). Preferably, the target adenosine is part of a premature stop codon in the pre-mRNA or mRNA of human ABCA4. In another preferred embodiment, the target adenosine is any one of the G>A mutations provided in Table 1. More preferably, the target adenosine is the c.5882 G>A mutation in exon 42 of the human ABCA4 gene. In the human ABCA4 gene, the nucleotide 3' to this G>A mutation is also an adenosine. When GGA (wild-type; glycine) is repaired to GGA or GGG, both of which result in glycine at that codon, the second adenosine may be edited simultaneously with the target adenosine.

[0015] The present invention further relates to an EON that targets the (pre-)mRNA sequence of SEQ ID NO: 12, or a portion thereof containing a G>A mutation. Preferably, the present invention relates to an EON comprising or consisting of a sequence selected from the group consisting of SEQ ID NOs: 1, 2, 3, 4, 5, 6, 7, 8, 9, and 11. In a preferred embodiment, the EON comprises or consists of the sequence of SEQ ID NO: 8 or 9. In an even more preferred embodiment, the EON consists of the sequence of SEQ ID NO: 9. In another preferred embodiment, the EON according to the present invention comprises at least one methylphosphonate (MeP or MP) linkage, preferably at linkage position -1, according to the nucleotide and linkage numbering as disclosed in Figure 5 of WO 2020 / 201406.

[0016] The present invention further relates to a pharmaceutical composition comprising an EON according to the present invention and a pharmaceutically acceptable carrier. The present invention also relates to a composition comprising a set of two single-stranded antisense oligonucleotides (AONs) for use in deaminating a target adenosine in human ABCA4 pre-mRNA or mRNA, or a portion thereof, wherein one AON is an EON according to the present invention and the other AON is a "helper AON," which is complementary to a stretch of nucleotides in human ABCA4 pre-mRNA or mRNA that is different from the stretch of nucleotides complementary to the EON, and the helper AON has a length of 16 to 22 nucleotides. Preferably, the EON comprises or consists of the sequence set forth in SEQ ID NO:9, and the helper AON comprises or consists of the sequence set forth in SEQ ID NO:10.

[0017] The present invention further relates to an EON according to the invention for use in the treatment of Stargardt's disease.

[0018] The present invention further relates to the use of an EON according to the present invention for the manufacture of a medicament for use in treating, ameliorating, preventing or slowing the progression of Stargardt's disease.

[0019] The present invention further relates to a method for deaminating at least one specific target adenosine present in a target RNA molecule in a cell, wherein the target RNA molecule is a pre-mRNA or mRNA of human ABCA4, or a portion thereof, and the deamination method comprises the following steps: providing an EON according to the present invention to a cell; allowing the EON to be taken up by the cell; annealing the EON to the target RNA molecule; allowing a mammalian ADAR enzyme comprising a natural dsRNA-binding domain as found in a wild-type enzyme to deamidate the target adenosine in the target RNA molecule to inosine; and, optionally, confirming the presence of inosine in the target RNA molecule. The optional steps preferably include (i) determining the sequence of the target RNA molecule; (ii) assessing the presence of functional, extended, full-length and / or wild-type ABCA4 protein; (iii) assessing whether splicing of the pre-mRNA is regulated by deamination; or (iv) using a functional readout, wherein the target RNA after deamination encodes functional, full-length, extended and / or wild-type ABCA4 protein.

[0020] The present invention further relates to a method for treating, ameliorating, preventing, and / or slowing the progression of Stargardt disease in a human subject in need thereof suffering from or at risk of suffering from Stargardt disease, comprising administering to the human subject an EON according to the present invention, or a pharmaceutical composition according to the present invention. In a preferred embodiment, the EON is administered to a human subject in need thereof by intravitreal administration, preferably by direct injection of naked (and chemically modified) EONs, and by enabling targeting of mutant ABCA4 pre-mRNA or mutant ABCA4 mRNA in retinal cells, preferably photoreceptor cells or RPE cells, to allow translation of the edited (i.e., preferably mostly wild-type) ABCA4 mRNA, thereby mitigating the disease-causing effects of the mutant ABCA4 protein.

[0021] The EON of the present invention does not contain the recruitment moiety described in WO2016 / 097212. The EON of the present invention does not contain a moiety that can form an intramolecular stem-loop structure. The EON of the present invention is shorter than those that form a loop structure (that recruits ADAR), which makes the EON of the present invention cheaper to produce, easier to use, and easier to manufacture. Furthermore, they are more likely to enter cells efficiently and are less likely to be degraded than longer oligonucleotides. WO2017 / 220751 and WO2018 / 041973 disclose EONs that are complementary to target RNAs to deaminate target adenosines present in the target RNA sequence to which the EON is complementary, but lack a recruitment moiety while still being able to use ADAR enzymes present in cells to edit target adenosines. The present invention aims to utilize such knowledge to solve the problem of targeting Stargardt disease mutations where exon skipping may not be a preferred treatment option or where alternative approaches are sought.

[0022] The EON of the present invention comprises one or more nucleotides having one or more non-naturally occurring sugar modifications. Thus, a single nucleotide of the EON can have one or more such sugar modifications. Within the EON, one or more nucleotides can have such sugar modifications. It is also an aspect of the present invention that the nucleotide in the EON of the present invention opposite the nucleotide that needs to be edited does not contain a 2'-O-methyl (2'-OMe) or 2'-methoxyethoxy (2'-MOE) modification. Often, the nucleotides immediately 3' and 5' of this nucleotide in the EON (the "flanking nucleotides") also lack such chemical modifications, although it is not essential that both flanking nucleotides not contain a 2'-O-alkyl group (e.g., 2'-OMe). One flanking nucleotide, both flanking nucleotides, or all three nucleotides of the "central triplet" can possess a 2'-OH group.

[0023] Those skilled in the art know that oligonucleotides, such as RNA oligonucleotides, are generally composed of repetitive monomers. These monomers are most often nucleotides or nucleotide analogs. The most common naturally occurring nucleotides in RNA are adenosine monophosphate (A), cytidine monophosphate (C), guanosine monophosphate (G), and uridine monophosphate (U). These consist of a pentose sugar, a ribose, a 5'-linked phosphate group linked via a phosphate ester, and a 1'-linked base. The sugar connects the base and the phosphate and is therefore often referred to as the "scaffold" of the nucleotide. Therefore, modifications of the pentose sugar are often referred to as "scaffold modifications." In severe modifications, the original pentose sugar may be replaced in its entirety with another moiety that also connects the base and the phosphate. Therefore, it is understood that while the pentose sugar is often the scaffold, the scaffold is not necessarily the pentose sugar.

[0024] The bases, sometimes called nucleobases, are generally adenine, cytosine, guanine, thymine, or uracil, or their derivatives. Cytosine, thymine, and uracil are pyrimidine bases, generally linked to the scaffold through their 1-nitrogen. Adenine and guanine are purine bases, generally linked to the scaffold through their 9-nitrogen.

[0025] Nucleotides are generally connected to adjacent nucleotides through condensation of their 5'-phosphate moiety with the 3'-hydroxyl moiety of an adjacent nucleotide monomer. Similarly, their 3'-hydroxyl moiety is generally connected to the 5'-phosphate of an adjacent nucleotide monomer, forming a phosphodiester linkage. The phosphodiester and scaffold form an alternating copolymer. Bases are grafted onto this copolymer, i.e., onto the scaffold moiety. Because of this property, the alternating copolymer formed by the linked monomers of an oligonucleotide is often referred to as the "backbone" of the oligonucleotide. Because phosphodiester linkages connect adjacent monomers together, they are often referred to as "backbone linkages." It is understood that if the phosphate group is modified to a similar moiety, such as a phosphorothioate, instead, such moiety is still referred to as the backbone linkage of the monomer. This is referred to as "modification of the backbone linkage." In general terms, the backbone of an oligonucleotide comprises alternating scaffolds and backbone linkages.

[0026] In one embodiment, the nucleobase of an EON of the invention is adenine, cytosine, guanine, thymine, or uracil. In another embodiment, the nucleobase is a modified form of adenine, cytosine, guanine, or uracil. In another embodiment, the modified nucleobase is hypoxanthine (nucleobase in inosine), pseudouracil, pseudocytosine, 1-methylpseudouracil, orotic acid, agmatidine, lycidine, 2-thiouracil, 2-thiothymine, 5-halouracil, 5-halomethyluracil, 5-trifluoromethyluracil, 5-propynyluracil, 5-propynylcytosine, 5-aminomethyluracil, 5-hydroxymethyluracil, 5-formyluracil, 5-aminomethylcytosine, 5-formylcytosine, 5-hydroxymethylcytosine, 7-deazaguanine, 7-deazaadenine, 7-deaza-2,6-diaminopurine, 8-aza- ... Azaguanine, 8-aza-7-deazaadenine, 8-aza-7-deaza-2,6-diaminopurine, pseudoisocytosine, N4-ethylcytosine, N2-cyclopentylguanine, N2-cyclopentyl-2-aminopurine, N2-propyl-2-aminopurine, 2,6-diaminopurine, 2-aminopurine, G-clamp, super A, super T, super G, amino-modified nucleobase or their derivatives; and degenerate or universal base such as 2,6-difluorotoluene or non-existent abasic site (for example, 1-deoxyribose, 1,2-dideoxyribose, 1-deoxy-2-O-methylribose, azaribose).The terms "adenine", "guanine", "cytosine", "thymine", "uracil" and "hypoxanthine" used herein refer to nucleobases in this way. The terms "adenosine," "guanosine," "cytidine," "thymidine," "uridine," and "inosine" refer to a nucleobase linked to a (deoxy)ribosyl sugar. The term "nucleoside" refers to a nucleobase linked to a (deoxy)ribosyl sugar. The term "nucleotide" refers to the respective nucleobase-(deoxy)ribosyl-phospholinker, as well as any chemical modification of the ribose moiety or its phospho group.Thus, the term includes nucleotides containing locked ribosyl moieties (including 2'-4' bridges containing methylene groups or any other group well known in the art), and nucleotides containing linkers including phosphodiester, phosphotriester, phosphoro(di)thioate, methylphosphonate, phosphoramidate, etc. The terms adenosine and adenine, guanosine and guanine, cytosine and cytidine, uracil and uridine, thymine and thymidine, and inosine and hypoxanthine may be used interchangeably to refer to the corresponding nucleobase, nucleoside, or nucleotide. Unless the context clearly dictates otherwise, the terms nucleobase, nucleoside, and nucleotide may be used interchangeably.

[0027] In one embodiment, the EON of the present invention comprises a 2'-substituted phosphorothioate monomer, preferably a 2'-substituted phosphorothioate RNA monomer, a 2'-substituted phosphate RNA monomer, or a 2'-substituted mixed phosphate / phosphorothioate monomer. It should be noted that DNA is considered an RNA derivative in terms of the 2' substitution. The EON of the present invention comprises at least one 2'-substituted RNA monomer connected through or by a phosphorothioate or phosphate backbone linkage, or a mixture thereof. The 2'-substituted RNA is preferably 2'-F, 2'-H (DNA), 2'-O-methyl, or 2'-O-(2-methoxyethyl). 2'-O-methyl is often abbreviated as "2'-OMe," and the 2'-O-(2-methoxyethyl) moiety is often abbreviated as "2'-MOE." More preferably, the 2'-substituted RNA monomer in the EON of the present invention is a 2'-OMe monomer, excluding the monomer opposite the target adenosine, which should not have a 2'-OMe substitution, as further outlined herein. In a preferred embodiment of this aspect, an EON according to the present invention is provided, wherein the 2'-substituted monomer can be a 2'-substituted RNA monomer, such as a 2'-F monomer, a 2'-NH2 monomer, a 2'-H monomer (DNA), a 2'-O-substituted monomer, a 2'-OMe monomer, or a 2'-MOE monomer, or a mixture thereof. Preferably, the monomer opposite the target adenosine is a 2'-H monomer (DNA), but can also be a monomer other than a 2'-OMe monomer that allows deamination of the target adenosine. Preferably, any other 2'-substituted monomer in the EON is a 2'-substituted RNA monomer, such as a 2'-OMe RNA monomer or a 2'-MOE RNA monomer, and these may also appear in combination within the EON.

[0028] Throughout this application, 2'-OMe monomers within the EONs of the invention can be replaced with 2'-OMe phosphorothioate RNA, 2'-OMe phosphate RNA, or 2'-OMe phosphate / phosphorothioate RNA. Throughout this application, 2'-MOE monomers can be replaced with 2'-MOE phosphorothioate RNA, 2'-MOE phosphate RNA, or 2'-MOE phosphate / phosphorothioate RNA. Throughout this application, oligonucleotides consisting of 2'-OMe RNA monomers linked by or through phosphorothioate, phosphate, or mixed phosphate / phosphorothioate backbone linkages can be replaced with oligonucleotides consisting of 2'-OMe phosphorothioate RNA, 2'-OMe phosphate RNA, or 2'-OMe phosphate / phosphorothioate RNA. Throughout this application, oligonucleotides consisting of 2'-MOE RNA monomers linked by or connected through phosphorothioate, phosphate or mixed phosphate / phosphorothioate backbone linkages can be replaced with oligonucleotides consisting of 2'-MOE phosphorothioate RNA, 2'-MOE phosphate RNA or 2'-MOE phosphate / phosphorothioate RNA.

[0029] In addition to certain preferred chemical modifications at certain positions of the compounds of the invention, the compounds of the invention may also comprise or consist of one or more (further) modifications to the nucleobase, scaffold and / or backbone linkages, which may or may not be present in the same monomer, for example, at the 3' and / or 5' positions. Scaffold modifications refer to the presence of modified versions of the ribosyl moiety (i.e., pentose moiety) naturally occurring in RNA, such as bicyclic sugars, tetrahydropyrans, hexoses, morpholinos, 2'-modified sugars, 4'-modified sugars, 5'-modified sugars and 4'-substituted sugars.Examples of suitable modifications include, but are not limited to, 2'-O-alkyl or 2'-O-(substituted) alkyl, such as 2'-O-methyl, 2'-O-(2-cyanoethyl), 2'-MOE, 2'-O-(2-thiomethyl)ethyl, 2'-O-butyryl, 2'-O-propargyl, 2'-O-allyl, 2'-O-(2-aminopropyl), 2'-O-(2-(dimethylamino)propyl), 2'-O-(2-amino) 2'-O-modified RNA monomers such as ethyl, 2'-O-(2-(dimethylamino)ethyl); 2'-deoxy (DNA); 2'-O-(haloalkyl)methyl such as 2'-O-(2-chloroethoxy)methyl (MCEM), 2'-O-(2,2-dichloroethoxy)methyl (DCEM); 2'-O-[2-(methoxycarbonyl)ethyl] (MOCE), 2'-O-[2-N-methylcarbamoyl)ethyl] (MCE), 2'- 2'-O-alkoxycarbonyl, such as O-[2-(N,N-dimethylcarbamoyl)ethyl] (DCME); 2'-halo, e.g., 2'-F, FANA; 2'-O-[2-(methylamino)-2-oxoethyl] (NMA); bicyclic or bridged nucleic acid (BNA) scaffold modifications, e.g., conformationally restricted nucleotide (CRN) monomers, locked nucleic acid (LNA) monomers, xylo-LNA monomers, α-LNA monomers, α- L-LNA monomer, β-D-LNA monomer, 2'-amino-LNA monomer, 2'-(alkylamino)-LNA monomer, 2'-(acylamino)-LNA monomer, 2'-N-substituted-2'-amino-LNA monomer, 2'-thio-LNA monomer, (2'-O,4'-C) constrained ethyl (cEt) BNA monomer, (2'-O,4'-C) constrained methoxyethyl (cMOE) BNA monomer, 2',4'-BNA. NC (NH) monomer, 2',4'-BNA NC (NMe) monomer, 2',4'-BNA NC(NBn) monomers, ethylene-bridged nucleic acid (ENA) monomers, carba-LNA (cLNA) monomers, 3,4-dihydro-2H-pyran nucleic acid (DpNA) monomers, 2'-C-bridged bicyclic nucleotide (CBBN) monomers, oxo-CBBN monomers, heterocyclic bridged BNA monomers (e.g., triazolyl- or tetrazolyl-linked), amide-bridged BNA monomers (e.g., AmNA), urea-bridged BNA monomers, sulfonamide-bridged BNA monomers, bicyclic carbocyclic nucleotide monomers, TriNA monomers, α-L-TriNA monomers, bicycloDNA (bcDNA) monomers, F-bcDNA monomers, tricycloDNA (tcDNA) monomers , F-tcDNA monomers, alpha-anomeric bicycloDNA (abcDNA) monomers, oxetane nucleotide monomers, 2'-amino-LNA-derived locked PMO monomers, guanidine-bridged nucleic acid (GuNA) monomers, spirocyclopropylene-bridged nucleic acid (scpBNA) monomers and their derivatives; cyclohexenyl nucleic acid (CeNA) monomers, altritol nucleic acid (ANA) monomers, hexitol nucleic acid (HNA) monomers, fluorinated HNA (F-HNA) monomers, pyranosyl-RNA (p-RNA) monomers, 3'-deoxypyranosyl DNA (p-DNA), unlocked nucleic acid (UNA); inverted versions of any of the above monomers. All of these modifications are known to those skilled in the art.

[0030] "Backbone modification" refers to the presence of modified versions of the ribosyl moiety ("scaffold modification"), as described above, and / or the presence of modified versions of the phosphodiesters naturally occurring in RNA ("backbone linkage modifications"). Examples of internucleoside linkage modifications are phosphorothioate (PS), chirally pure phosphorothioate, Rp phosphorothioate, Sp phosphorothioate, phosphorodithioate (PS2), phosphonoacetate (PACE), phosphonoacetamide (PACA), thiophosphonacetamide, phosphorothioate prodrug, S-alkylated phosphorothioate, H-phosphonate, methylphosphonate, methylphosphonothioate, methylphosphate, methylphosphorothioate, ethylphosphate, ethylphosphorothioate, boranophosphate, boranophosphorothioate, and the like. oate, methylboranophosphate, methylboranophosphorothioate, methylboranophosphonate, methylboranophosphonothioate, phosphorylguanidine (PGO), methylsulfonyl phosphoramidate, phosphoramidite, phosphonamidite, N3'→P5' phosphoramidate, N3'→P5' thiophosphoramidate, phosphorodiamidate, phosphorothiodiamidate, sulfamate, dimethylenesulfoxide, sulfonate, triazole, oxalyl, carbamate, methyleneimino (MMI), and thioacetamide (TANA); and derivatives thereof.

[0031] A preferred EON of the present invention does not contain a 5'-end O6-benzylguanosine or 5'-end amino modification and is not covalently linked to a SNAP tag domain (engineered O6-alkylguanosine-DNA-alkyltransferase). In one embodiment, the EON of the present invention contains 0, 1, 2, or 3 wobble base pairs with the target sequence and / or 0, 1, 2, or 3 mismatches with the target RNA sequence, where a single mismatch can include multiple consecutive nucleotides. Similarly, a preferred EON of the present invention does not contain a boxB RNA hairpin sequence. The EON of the present invention can specifically edit a target adenosine in a target RNA sequence by utilizing endogenous cellular pathways and naturally available ADAR enzymes. The EON of the present invention can recruit and complex with ADAR, and then deaminate a (single) specific target adenosine nucleotide in the target RNA sequence. Ideally, only one adenosine is deaminated. Alternatively, for example, when target adenosines are close to each other, one, two or three adenosine nucleotides are deaminated.For example, when mutation is from wild-type GGA (glycine) codon to mutant GAA (glutamic acid) codon, deamination of both adenosines will produce GGG, which also encodes glycine.When EON of the present invention is combined with ADAR, it preferably deaminates single target adenosine.

[0032] Analysis of natural targets of ADAR enzymes has shown that they generally contain mismatches between the two strands that form the RNA helices edited by ADAR1 or ADAR2. These mismatches have been suggested to enhance the specificity of the editing reaction (Stefl et al. 2006. Structure 14(2):345-355; Tian et al. 2011. Nucleic Acids Res 39(13):5669-5681). Characterization of the optimal pattern of paired / mismatched nucleotides between EONs and target RNAs also appears crucial for the development of efficient ADAR-based EON therapies.

[0033] The EON of the present invention uses specific nucleotide modifications at predefined locations to ensure stability and proper ADAR binding and activity.These modifications are various, and as detailed above, can include modifications of the backbone of EON, in the sugar moiety of nucleotides, and in the nucleic acid base or phosphodiester linkage.Such modifications can also be distributed in various ways throughout the sequence of EON.Specific modifications may be required to support the interaction of various amino acid residues in the RNA binding domain of ADAR enzymes and the interaction of various amino acid residues in the deaminase domain.For example, phosphorothioate linkages between nucleotides, or 2'-OMe or 2'-MOE modifications can be tolerated in some parts of EON, while in other parts they must be avoided to avoid disrupting the crucial interaction between phosphate and 2'-OH groups and the enzyme.If the target sequence is not optimal for ADAR editing, specific nucleotide modifications may be required to enhance the editing activity of substrate RNA.Previous studies have confirmed that certain sequence configurations are more suitable for editing. For example, the target sequence 5'-UAG-3' (with a central target A) contains the most favorable nearest-neighbor nucleotide for ADAR2, whereas the 5'-CAA-3' target sequence is not preferred (Schneider et al. 2014. Nucleic Acids Res 42(10):e87). Structural analysis of the ADAR2 deaminase domain suggested that careful selection of the nucleotide opposite the target trinucleotide may enhance editing. For example, the 5'-CAA-3' target sequence paired with the 3'-GCU-5' sequence on the opposite strand (with a central AC mismatch) is not preferred because the guanosine base sterically clashes with the amino acid side chain of ADAR2.

[0034] As used herein, a "central triplet" refers to the three nucleotides opposite a target adenosine in a target RNA, where the central nucleotide of the central triplet is directly opposite the target adenosine. The central triplet need not be located at the center of the EON, as it can be located not only at the more 3'-end of the EON, but also at the more 5'-end, whichever is preferred for a particular target. Thus, the term "central" in this embodiment refers more to the triplet at the center of catalytic activity when it comes to chemical modification and targeting adenosines. It should also be noted that EONs may be displayed 3' to 5', particularly when the target sequence is displayed 5' to 3' (see, for example, Figure 1). However, whenever the order of nucleotides within an EON is discussed herein, the order is from 5' to 3' of the EON. The position can also be expressed in terms of a particular nucleotide within the EON, still following a 5' to 3' directionality, where other nucleotides 5' of that nucleotide are marked as negative positions and other nucleotides 3' of that nucleotide are marked as positive positions.

[0035] As outlined herein, the nucleotides outside the central triplet are often 2'-OMe or 2'-MOE modified.However, this is not a strict requirement for the EON of the present invention.The use of such 2' substitutions ensures the appropriate stability of these parts of the EON, but other modifications may also be applied.

[0036] The EON of the present invention can be administered indirectly using suitable means known in the art. For example, the EON can be provided to an individual or a cell, tissue, or organ of the individual in the form of an expression vector, where the expression vector encodes a transcript comprising the oligonucleotide. The expression vector is preferably introduced into a cell, tissue, organ, or individual via a gene delivery vehicle. In a preferred embodiment, a viral-based expression vector is provided, comprising an expression cassette or transcription cassette that drives the expression or transcription of the EON identified herein. Thus, the present invention provides viral vectors capable of expressing the EON of the present invention when placed under conditions conducive to EON expression (and thus free of non-natural chemical modifications). Cells can be provided with the EON via plasmid-driven EON expression or viral expression provided by adenovirus or adeno-associated virus-based vectors. Expression can be driven by a polymerase II promoter (Pol II), such as the U7 promoter, or a polymerase III (Pol III) promoter, such as the U6 RNA promoter. Preferred delivery vehicles are retroviral vectors, such as AAV or lentiviral vectors. In addition, plasmids, artificial chromosomes, and plasmids that can be used for targeted homologous recombination and integration into the human genome of cells can be suitably applied to the delivery of EONs as defined herein. Preferred for the present invention are those vectors in which transcription is driven by a Pol III promoter and / or the transcript is in the form of a fusion with a U1 or U7 transcript, which provides good results for the delivery of small transcripts. Designing an appropriate transcript is within the skill of those skilled in the art. Preferred are Pol III-driven transcripts, preferably in the form of a fusion transcript with a U1 or U7 transcript, which are known to those skilled in the art.

[0037] Typically, when delivered by a viral vector, the EON is in the form of an RNA transcript, containing the sequence of the oligonucleotide of the present invention within a portion of the transcript. The resulting EON, active in cells, is therefore naturally expressed and therefore unmodified, whereas EONs produced in a "naked" form (i.e., without the use of a plasmid or viral vector expressing the EON) may contain single or multiple non-naturally occurring modifications. The AAV vector of the present invention refers to a recombinant AAV vector comprising a portion of an AAV genome containing the encoded EON of the present invention enclosed in a protein shell of capsid protein derived from an AAV serotype. The portion of the AAV genome can contain inverted terminal repeats (ITRs) derived from an adeno-associated virus serotype, such as AAV1, AAV2, AAV3, AAV4, AAV5, AAV6, AAV7, AAV8, AAV9, etc. The protein shell composed of capsid protein can be derived from an AAV serotype, such as AAV1, 2, 3, 4, 5, 6, 7, 8, 9, etc. The protein shell may also be referred to as a capsid protein shell. An AAV vector may lack one or preferably all wild-type AAV genes but still contain functional ITR nucleic acid sequences. Functional ITR sequences are necessary for AAV virion replication, rescue, and packaging. The ITR sequences may be wild-type sequences or may have at least 80%, 85%, 90%, 95, or 100% sequence identity with the wild-type sequence, and may be altered, for example, by nucleotide insertion, mutation, deletion, or substitution, as long as they remain functional. In this context, functionality refers to the ability to induce genome packaging into a capsid shell and subsequently enable expression in infected host or target cells. In the context of the present invention, the capsid protein shell may be of a different serotype from the AAV vector genome ITRs.Thus, an AAV vector according to the invention may be comprised of a capsid protein shell, i.e., an icosahedral capsid, that includes the capsid proteins (VP1, VP2, and / or VP3) of one AAV serotype, e.g., AAV serotype 2, while the ITR sequences contained in an AAV2 vector can be any of the AAV serotypes listed above, including AAV2 vectors. Thus, an "AAV2 vector" includes the capsid protein shell of AAV serotype 2, while, for example, an "AAV5 vector" includes the capsid protein shell of AAV serotype 5, both of which are capable of encapsidating any AAV vector genome ITR according to the invention. Preferably, a recombinant AAV vector according to the present invention comprises an AAV serotype 2, 5, 8 or AAV serotype 9 capsid protein shell, wherein the AAV genome or ITRs present in said AAV vector are derived from AAV serotype 2, 5, 8 or AAV serotype 9; such AAV vectors are referred to as AAV2 / 2, AAV2 / 5, AAV2 / 8, AAV2 / 9, AAV5 / 2, AAV5 / 5, AAV5 / 8, AAV5 / 9, AAV8 / 2, AAV8 / 5, AAV8 / 8, AAV8 / 9, AAV9 / 2, AAV9 / 5, AAV9 / 8 or AAV9 / 9 vectors.

[0038] More preferably, a recombinant AAV vector according to the invention comprises a capsid protein shell of AAV serotype 2, and the AAV genome or ITRs present in the vector are derived from AAV serotype 5; such a vector is referred to as an AAV2 / 5 vector. More preferably, a recombinant AAV vector according to the invention comprises a capsid protein shell of AAV serotype 2, and the AAV genome or ITRs present in the vector are derived from AAV serotype 8; such a vector is referred to as an AAV2 / 8 vector. More preferably, a recombinant AAV vector according to the invention comprises a capsid protein shell of AAV serotype 2, and the AAV genome or ITRs present in the vector are derived from AAV serotype 9; such a vector is referred to as an AAV2 / 9 vector. More preferably, a recombinant AAV vector according to the invention comprises a capsid protein shell of AAV serotype 2, and the AAV genome or ITRs present in the vector are derived from AAV serotype 2; such a vector is referred to as an AAV2 / 2 vector. The nucleic acid molecule encoding the EON of the present invention represented by the selected nucleic acid sequence is preferably inserted between the AAV genome or ITR sequences identified above; for example, the expression construct comprises an expression regulatory element and a 3' termination sequence operably linked to the coding sequence. "AAV helper functions" generally refer to the corresponding AAV functions required for AAV replication and packaging provided to an AAV vector in trans. AAV helper functions complement AAV functions missing from the AAV vector but lacking the AAV ITRs (provided by the AAV vector genome). AAV helper functions include the two major AAV ORFs, i.e., the rep and cap coding regions, or sequences substantially identical in function to them. The Rep and Cap regions are well known in the art. AAV helper functions can be provided on an AAV helper construct, which can be a plasmid.

[0039] The introduction of the helper construct into the host cell can be carried out, for example, by transformation, transfection, or transduction before or simultaneously with the introduction of the AAV genome present in the AAV vector specified herein.Therefore, the AAV helper construct of the present invention can be selected to provide a desired combination of the serotype of the capsid protein shell of the AAV vector on the one hand, and the serotype of the AAV genome present in the AAV vector replication and packaging on the other hand.An "AAV helper virus" provides additional functions required for AAV replication and packaging.

[0040] Suitable AAV helper viruses include adenovirus, herpes simplex virus (e.g., HSV types 1 and 2), and vaccinia virus. As described in U.S. Pat. No. 6,531,456, additional functions provided by helper viruses can also be introduced into host cells via vectors. Preferably, the AAV genome present in a recombinant AAV vector according to the present invention does not contain any nucleotide sequences encoding viral proteins, such as the AAV rep (replication) or cap (capsid) genes. The AAV genome may further contain a marker or reporter gene, such as an antibiotic resistance gene, a gene encoding a fluorescent protein (e.g., gfp), or a gene encoding a chemically, enzymatically, or otherwise detectable and / or selectable product known in the art (e.g., lacZ, aph, etc.). A preferred AAV vector according to the present invention is an AAV vector, preferably an AAV2 / 5, AAV2 / 8, AAV2 / 9, or AAV2 / 2 vector, that expresses an EON according to the present invention.

[0041] Whenever "antisense oligonucleotides" ("AONs"), "RNA editing oligonucleotides" ("EONs"), "oligonucleotides," or "oligos" are used, it means both oligoribonucleotides and deoxyoligoribonucleotides unless the context clearly dictates otherwise. Whenever "oligoribonucleotides" are used, it means oligonucleotides containing the ribonucleosides adenosine (A), guanosine (G), cytidine (C), 5-methylcytidine (m ... 5 C), uridine (U), 5-methyluridine (m 5Whenever "deoxyoligoribonucleotide" is mentioned, this may include the deoxyribonucleosides deoxyadenosine (A), deoxyguanosine (G), deoxycytidine (C), thymine (T), or deoxyinosine (I). In preferred embodiments, most of the EONs of the present invention are oligoribonucleotides that may contain chemical modifications and, at some specific positions, may contain deoxyribonucleosides (DNA). When referring to nucleotides in oligonucleotide constructs such as cytosine, 5-methylcytosine, 5-hydroxymethylcytosine, pyrrolocytidine, and β-D-glucosyl-5-hydroxymethylcytosine are included. References to adenine include 2-aminopurine, 2,6-diaminopurine, 3-deazaadenosine, 7-deazaadenosine, 8-azidoadenosine, 8-methyladenosine, 7-aminomethyl-7-deazaguanosine, 7-deazaguanosine, N6-methyladenine, and 7-methyladenine. References to uracil include 5-methoxyuracil, 5-methyluracil, dihydrouracil, pseudouracil, and thienouracil, dihydrouracil, 4-thiouracil, and 5-hydroxymethyluracil. References to guanosine include 7-methylguanosine, 8-aza-7-deazaguanosine, thienouracil, and 1-methylguanosine. Reference to nucleosides or nucleotides includes ribofuranose derivatives such as 2'-deoxy, 2'-hydroxy, 2-fluororibose, and 2'-O-substituted variants such as 2'-O-methyl, as well as other modifications, including 2'-4' bridged variants.

[0042] The term "comprising" encompasses "including" as well as "consisting of," e.g., a composition "comprising X" may consist exclusively of X or may include something additional, e.g., X+Y. The term "about," with respect to a numerical value x, is optional, and may mean, for example, x±10%. The term "substantially" does not exclude "completely," e.g., a composition "substantially free of Y" may be completely free of Y. Where appropriate, the term "substantially" may be omitted from the definition of the invention. The term "downstream," with respect to a nucleic acid sequence, means further along the sequence in the 3' direction; the term "upstream" means vice versa. Thus, in any sequence encoding a polypeptide, the start codon is upstream of the stop codon on the sense strand, but downstream of the stop codon on the antisense strand. Reference to "hybridization" typically refers to specific hybridization and excludes non-specific hybridization. Specific hybridization can be achieved under selected experimental conditions using techniques well known in the art to ensure that the most stable interaction between the probe and target occurs when the probe and target share at least 70%, preferably at least 80%, and more preferably at least 90% sequence identity. The term "mismatch" is used herein to refer to opposing nucleotides in a double-stranded RNA complex that do not form a perfect base pair according to the Watson-Crick base pairing rules. Mismatched base pairs include GA, CA, UC, AA, GG, CC, and UU base pairs. In some embodiments, the EONs of the present invention contain zero, one, two, or three mismatches, where a single mismatch can include several consecutive nucleotides. In some embodiments, the EONs of the present invention contain zero, one, two, or three wobble base pairs. Wobble base pairs include GU, IU, IA, and IC base pairs.

[0043] The usual internucleoside linkage between nucleotides can be modified by mono- or di-thiolation of the phosphodiester bond to obtain phosphorothioate or phosphorodithioate esters, respectively. Other modifications of the internucleoside linkage are possible, including amidation and peptide linkers. In a preferred embodiment, the EON of the present invention has one, two, three, four, or more phosphorothioate linkages between nucleotides at the terminal ends of the EON (thus, preferably at both the 5' and 3' ends), meaning that the final five nucleotides are linked accordingly, in the case of four phosphorothioate linkages, which is a particularly preferred embodiment. Those skilled in the art will understand that the number of such linkages can vary at each end depending on the target sequence or other aspects, such as stability, toxicity, and / or efficiency. In one embodiment of the present invention, the EON of the present invention contains a substitution of one of the non-bridging oxygens in the phosphodiester linkage. This modification slightly destabilizes base pairing but confers significant resistance to nuclease degradation. The exact chemical nature and format may vary depending on each oligonucleotide construct and each application, and can be determined according to the desires and preferences of those skilled in the art. Four or more consecutive DNA nucleotides in an oligonucleotide create a so-called "gapmer," which is believed in the art to induce RNase H cleavage of target RNA when annealed to its RNA cognate sequence. According to the present invention, RNase H cleavage of target RNA should generally be avoided as much as possible.

[0044] EONs according to the present invention are typically longer than 10 nucleotides, preferably longer than 11, 12, 13, 14, 15, or 16 nucleotides, and even more preferably longer than 17 nucleotides. In one embodiment, EONs according to the present invention are longer than 20 nucleotides. Oligonucleotides according to the present invention are preferably shorter than 100 nucleotides, even more preferably shorter than 60 nucleotides, and even more preferably shorter than 50 nucleotides. In preferred embodiments, oligonucleotides according to the present invention comprise 18 to 70 nucleotides, more preferably 18 to 60 nucleotides, and even more preferably 18 to 50 nucleotides. Thus, in particularly preferred embodiments, oligonucleotides according to the present invention comprise 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, or 50 nucleotides. In another preferred embodiment, an EON according to the invention can incorporate an inverted deoxy-T or dideoxy-T nucleotide at either or both termini.

[0045] It is known in the art that RNA editing entities (e.g., human ADAR enzymes) edit dsRNA structures with various specificities depending on several factors. One important factor is the degree of complementarity of the two strands that make up the dsRNA sequence. Perfect complementarity of the two strands usually causes the catalytic domain of hADAR to non-discriminately deaminate adenosine, and hADAR reacts with any adenosine it encounters. The specificity of hADAR1 and 2 can be improved by ensuring several mismatches in dsRNA, which are presumed to help position the dsRNA binding domain in a manner that has not yet been clarified. In addition, the deamination reaction itself can be enhanced by providing an EON containing a mismatch on the opposite side of the adenosine to be edited. Mismatches are preferably created by providing a target moiety with cytidine on the opposite side of the adenosine to be edited. Alternatively, uridine can also be used on the opposite side of adenosine, but this will not result in a "mismatch", since U and A are paired. When adenosine in the target strand is deaminated, the target strand acquires inosine, which in most biochemical processes is "read" as G by the cellular biochemical machinery. Therefore, after the conversion of A to I, the mismatch is resolved, because I is fully capable of base pairing with the opposite C in the target portion of the oligonucleotide construct of the present invention. After the mismatch is resolved by editing, the substrate is released, and the oligonucleotide construct-editing entity complex is released from the target RNA sequence, which then becomes available for downstream biochemical processes such as splicing and translation. The desired level of specificity for editing the target RNA sequence can depend on the application. Following the instructions of this disclosure, those skilled in the art can design the complementary portion of the oligonucleotide to suit their needs and, with a little trial and error, achieve the desired results.

[0046] The teachings of the present invention can also be used to edit target RNA sequences in cells in so-called organoids, such as in vitro-generated optic cups.Cells treated in organoids, or in vivo, or in vitro or ex vivo situations, generally have adenosine-targeted gene mutations.Mutations can be heterozygous or homozygous.The present invention can typically be used to modify point mutations, such as N to A mutations, where N can be G, C, U (T at the DNA level), preferably G to A mutations, or N to C mutations, where N can be A, G, U (T at the DNA level), preferably U to C mutations.Table 1 lists the most relevant pathogenic G>A mutations found in ABCA4 gene (reported in the art), all of which can potentially be targeted by EON according to the present invention. It should be noted that not all (potentially possible) mutations are given here because the EONs of the present invention can also be applied to target adenosines that are not the result of a G>A mutation, but are actually part of a premature stop codon resulting from a mutation that results in the appearance of a T or G. As an example, if a wild-type TCA codon (serine) mutates to TGA (stop) due to a C>G mutation, targeting the A using an EON of the present invention will change the stop codon to TGG (tryptophan), which can be tolerated in the ABCA4 protein. As another example, a wild-type AAA codon (lysine) can mutate to TAA (stop) due to an A>T mutation, and then targeting the 5' A using an EON of the present invention will change this stop codon to GAA (glutamic acid), which can be tolerated in the ABCA4 protein.

[0047] [Table 1-1]

[0048] [Table 1-2]

[0049] In one aspect of the present invention, a composition is provided comprising at least one EON according to the present invention, wherein preferably the composition comprises at least one excipient, and / or the EON comprises at least one conjugated ligand that can further help enhance targeting and / or delivery of the composition and / or the EON to and / or into tissues and / or cells. The compositions described herein are referred to herein as compositions according to the present invention. The compositions according to the present invention can comprise one or more EONs according to the present invention. In the context of the present invention, the excipient can be a separate molecule, but it can also be a conjugated moiety. In the first case, the excipient can be a filler such as starch. In the latter case, the excipient can be, for example, a targeting ligand linked to an EON according to the present invention.

[0050] In a preferred embodiment of this aspect, such a composition can further comprise a cationic amphiphilic compound (CAC) or a cationic amphiphilic drug (CAD).CAC is generally a lysosomotropic agent, and is a weak base that can buffer endosomes and lysosomes (Mae et al. J Contr Rel 2009, 134: 221).Compared with a similar composition that does not contain the CAC, the composition preferably has improved parameters for RNA editing.Examples of CAC can be found, for example, in WO2018 / 007475 or WO2018 / 134310.

[0051] In a preferred embodiment, the composition according to the present invention is for use as a medicament. In that case, the composition according to the present invention is a pharmaceutical composition. Pharmaceutical compositions typically comprise pharmaceutically acceptable carriers, diluents and / or excipients. In a preferred embodiment, the composition according to the present invention comprises an EON as defined herein, and optionally further comprises pharmaceutically acceptable formulations, fillers, preservatives, solubilizers, carriers, diluents, excipients, salts, adjuvants and / or solvents. Such pharmaceutically acceptable carriers, fillers, preservatives, solubilizers, diluents, salts, adjuvants, solvents and / or excipients can be found, for example, in Remington: The Science and Practice of Pharmacy (20 th edition, Baltimore, MD; Lippincott, Williams & Wilkins, 2000). EONs according to the present invention may have at least one ionizable group. The ionizable group may be basic or acidic and may be charged or neutral. The ionizable group may exist as an ion pair with a suitable counterion carrying the opposite charge. Examples of cationic counterions are sodium, potassium, cesium, tris, lithium, calcium, magnesium, trialkylammonium, triethylammonium, and tetraalkylammonium. Examples of anionic counterions are chloride, bromide, iodide, lactate, mesylate, besylate, triflate, acetate, trifluoroacetate, dichloroacetate, tartrate, phosphate, and citrate.

[0052] The pharmaceutical composition according to the present invention may contain an excipient capable of forming a complex, nanoparticle, microparticle, nanotube, nanoparticle, nanogel, virosome, exosome, hydrogel, poloxamer or pluronic, polymersome, colloid, microbubble, vesicle, micelle, lipoplex, and / or liposome to enhance the stability, solubility, absorption, bioavailability, activity, pharmacokinetics, pharmacodynamics, cellular uptake, and / or intracellular transport of EON. Examples of nanoparticles include polymeric nanoparticles, (mixed) metal nanoparticles, carbon nanoparticles, gold nanoparticles, lipid nanoparticles, magnetic nanoparticles, and peptide nanoparticles, as well as combinations thereof. An example of a combination of nanoparticles and oligonucleotides is spherical nucleic acid (SNAs; Barnaby et al. Cancer Treat. Res. 2015, 166: 23).

[0053] A preferred composition according to the present invention comprises at least one excipient that can further enhance the targeting and / or delivery of EON to tissues and / or cells. Preferred tissues or cells are muscle cells, retinal cells such as photoreceptor cells or cells of the RPE layer, corneal cells, retinal tissue, or corneal tissue. Preferably, the EON according to the present invention is present in a pharmaceutical composition according to the present invention, which is administered intravitreally (e.g., by direct injection of naked EON using a syringe) to target photoreceptor cells or RPE in the retina to edit the target adenosine in human ABCA4 mRNA and / or ABCA4 pre-mRNA carrying a mutation that causes Stargardt disease.

[0054] Many potential excipients are known in the art and can be classified as type 1 excipients. Examples of type 1 excipients include polymers (e.g., polyethyleneimine (PEI), polypropyleneimine (PPI), dextran derivatives, butyl cyanoacrylate (PBCA), hexyl cyanoacrylate (PHCA), poly(lactic-co-glycolic acid) (PLGA), polyamines (e.g., spermine, spermidine, putrescine, cadaverine), chitosan, poly(amidoamine) (PAMAM), poly(esteramine), polyvinyl ether, polyvinylpyrrolidone (PVP), polyethylene glycol diol (PEG), polyethylene glycol terephthalate (PEG ... glycol (PEG), cyclodextrin, hyaluronic acid, colominic acid, and their derivatives), dendrimers (e.g., poly(amidoamine)), lipids (e.g., 1,2-dioleoyl-3-dimethylammonium propane (DODAP), dioleoyldimethylammonium chloride (DODAC), phosphatidylcholine derivatives (e.g., 1,2-distearoyl-sn-glycero-3-phosphocholine (DSPC)), lyso-phosphatidylcholine derivatives (e.g., 1-stearoyl-2-lyso-sn- glycero-3-phosphocholine (S-LysoPC)), sphingomyelin, 2-(3-bis-(3-aminopropyl)amino)propylamino)-N-ditetradecylcarbamoylmethylacetamide (RPR209120), phosphoglycerol derivatives (e.g., 1,2-dipalmitoyl-sn-glycero-3-phosphoglycerol, sodium salt (DPPG-Na), phosphatidic acid derivatives (e.g., 1,2-distearoyl-sn-glycero-3-phosphatidic acid, sodium salt (DSPA)), phospho Ethanolamine derivatives (e.g., dioleoyl-LR-phosphatidylethanolamine (DOPE), 1,2-distearoyl-sn-glycero-3-phosphoethanolamine (DSPE), 2-diphytanoyl-sn-glycero-3-phosphoethanolamine (DPhyPE)), N-(1-(2,3-dioleoyloxy)propyl)-N,N,N-trimethylammonium (DOTAP), N-(1-(2,3-dioleyloxy)propyl)-N,N,N-trimethylammonium (DOTMA), 1,3-Di-oleoyloxy-2-(6-carboxy-spermyl)propylamide (DOSPER), (1,2-dimyristoyloxypropyl-3-dimethylhydroxyethylammonium (DMRIE), (N1-cholesteryloxycarbonyl-3,7-diazanonane-1,9-diamine) (CDAN), dimethyldioctadecylammonium bromide (DDAB), 1-palmitoyl-2-oleoyl-sn-glycerol-3-phosphatidylcholine Phocoline (POPC), (bL-arginyl-2,3-L-diaminopropionic acid N-palmityl-N-oleylamide trihydrochloride (AtuFECT01), N,N-dimethyl-3-aminopropane derivatives (e.g., 1,2-distearoyl-N,N-dimethyl-3-aminopropane (DSDMA), 1,2-dioleyloxy-N,N-dimethyl-3-aminopropane (DoDMA), 1,2-dilinoleyloxy-N,N-3-dimethylaminopropane (DiDMA)), Examples of excipients include 2,2-dilinoleyl-4-dimethylaminomethyl(1,3)-dioxolane (DLin-K-DMA), DLinKC2DMA, DLinMC3DMA (MC3), phosphatidylserine derivatives (1,2-dioleyl-sn-glycero-3-phospho-L-serine, sodium salt (DOPS)), transfection reagents, proteins (e.g., albumin, gelatin, atelocollagen), and linear or cyclic peptides (e.g., protamine, PepFects, NickFects, polyarginine, hexa-arginine, polylysine, polyornithine, CADY, MPG, cell-penetrating peptides (CPPs), cell transport peptides (CTPs), targeting peptides, and endosomal escape peptides). Carbohydrates and carbohydrate clusters, when used as separate compounds, are also suitable for use as the first type of excipient.

[0055] Another preferred composition according to the present invention may contain at least one excipient classified as a second type of excipient. The second type of excipient may comprise or contain a conjugate group as described herein to enhance targeting and / or delivery to tissues and / or cells, such as retinal (photoreceptor or RPE) or corneal tissues or cells. The conjugate group may present one or more different or identical ligands. Examples of conjugate group ligands include peptides, vitamins, aptamers, carbohydrates or carbohydrate mixtures, proteins, small molecules, antibodies, polymers, and drugs. Examples of carbohydrate conjugate group ligands include glucose, mannose, fructose, maltose, galactose, N-galactosamine (GalNAc), glucosamine, N-acetylglucosamine, glucose-6-phosphate, mannose-6-phosphate, and maltotriose. The carbohydrate may also be contained in a carbohydrate cluster moiety, such as a GalNAc cluster moiety. The carbohydrate cluster moiety may comprise a targeting moiety and, optionally, a conjugate linker. In some embodiments, the carbohydrate cluster contains 1, 2, 3, 4, 5, or 6 or more GalNAc groups. As used herein, "carbohydrate cluster" refers to a compound having one or more carbohydrate residues attached to a scaffold or linker group (Maier et al. Bioconj Chem 2003, 14: 18). In this context, "modified carbohydrate" refers to any carbohydrate having one or more chemical modifications relative to a naturally occurring carbohydrate. As used herein, "carbohydrate derivative" refers to any compound that can be synthesized using a carbohydrate as a starting material or intermediate. As used herein, "carbohydrate" refers to a naturally occurring carbohydrate, a modified carbohydrate, or a carbohydrate derivative. As specified herein, both types of excipients can be combined into a single composition. Examples of trivalent N-acetylglucosamine clusters are described in WO 2017 / 062862, which also describes clusters of sulfonamide small molecules.An example of a single conjugate of the small molecule sertraline (Ferres-Coy et al. Mol. Psych. 2016, 21: 328) has also been described, as well as examples of conjugates of protein-binding small molecules, including ibuprofen (U.S. Pat. No. 6,656,730), spermine (Noir et al. J. Am. Chem. Soc. 2008, 130: 13500), anisamide (Nakagawa J. Am. Chem. Soc. 2010, 132, 8848), and folic acid (Dahmen Mol. Ther. Nucl. Acids 2012, 1, e7). Examples of lipid conjugates include fatty acids and derivatives thereof, such as palmityl, palmitoyl, stearyl, stearoyl, myristyl, myristoyl, lauryl, lauroyl, arachidonyl, arachidonoyl, behenyl, behenoyl, lignoceryl, lignoceroyl, sapienyl, sapienoyl, oleyl, oleoyl, elaidyl, elaidoyl, basenyl, basenoyl, linoleyl, linoleoyl, EPA, DHA, cholesteryl, steroids, omega-3 fatty acids, and omega-6 fatty acids, and one or more lipids or mixtures of lipids are conjugated to the oligonucleotides of the present invention. Examples of lipid conjugates of oligonucleotides have been described (WO 2019 / 232255; Biscans J. Control. Rel. 2019, 302, 116; Biscans Nucl. Acids Res. 2019, 47, 1082; Wang Nucl. Acid Ther. 2019, 29, 245). Examples of vitamins used for conjugation are known (Winkler Ther. Deliv. 2013, 4, 791; US ​​6,127,533). Conjugates of oligonucleotides and aptamers are also known in the art (Zhao Biomaterials 2015, 67, 42).

[0056] Antibodies and antibody fragments can also be conjugated to the oligonucleotide of the present invention.In a preferred embodiment, antibodies or fragments thereof that target specific target tissues, particularly retinal and / or corneal tissues, are conjugated to the oligonucleotide of the present invention.Examples of such antibodies and / or fragments include those that target CD71 (transferrin receptor; WO 2016 / 179257; Sugo J. Control. Rel. 2016, 237, 1).Other oligonucleotide conjugates are known to those skilled in the art and have been reviewed by Winkler et al. (Ther. Deliv. 2013, 4, 791, Manoharan Antisense Nucl. Acid Dev 2004, 12, 103) and Ming et al. (Adv. Drug Deliv. Rev. 2015, 87, 81).

[0057] One skilled in the art can select, combine, and / or adapt one or more of these or other alternative excipients and delivery systems to formulate and deliver EONs for use in the present invention.

[0058] The compounds contained in the composition according to the present invention can also be provided separately, for example, to allow for sequential administration of the active ingredients of the composition according to the present invention. In such a case, the composition according to the present invention is a combination of a compound comprising a conjugated ligand, at least one excipient, and at least one EON according to the present invention, optionally with or without the above-mentioned CAC.

[0059] Preferably, the pharmaceutical composition is for intravitreal administration, preferably by direct injection into the vitreous, and is administered in a total EON amount ranging from 0.05 mg to 5 mg per eye. The present invention also relates to pharmaceutical compositions according to the present invention, wherein the pharmaceutical composition is for direct intraocular administration by injection and is administered in a total EON amount ranging from 0.1 to 1 mg per eye, for example, about 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, or 1.0 mg per eye. Preferred amounts are 80, 160, and 320 μg of oligonucleotide per eye, with the first dose preferably being twice the amount of the second, third, and subsequent doses. Thus, the administration regimen may involve administering 160 μg of oligonucleotide (total nucleotides if more than one type is administered) in an initial injection, followed by 80 μg injections, or an initial injection containing 320 μg of oligonucleotide, followed by maintenance injections of 160 μg of oligonucleotide per eye. Such dosages may be adjusted depending on clinical outcome. The present invention also relates to a viral vector expressing an EON according to the present invention. In yet another aspect, the present invention relates to an EON according to the present invention, a pharmaceutical composition according to the present invention, or a viral vector according to the present invention for use as a medicament. In yet another aspect, the present invention relates to an EON according to the present invention, a pharmaceutical composition according to the present invention, or a viral vector according to the present invention for use in treating, preventing, or delaying Stargardt's disease. In yet another embodiment, the present invention relates to the use of an EON according to the present invention, a pharmaceutical composition according to the present invention, or a viral vector according to the present invention for treating, preventing, or delaying Stargardt's disease.

[0060] The term "pre-mRNA" refers to unprocessed or partially processed precursor mRNA that is synthesized from a cellular DNA template by transcription, such as in the nucleus.

[0061] Given the progress made to date, improvements in the means for delivering EONs of the present invention to an individual or the cells, tissues, or organs of the individual are expected. Such future improvements can, of course, be incorporated to achieve the aforementioned effects on mRNA remodeling using the methods of the present invention. EONs of the present invention can be delivered directly to an individual or the cells, tissues, or organs of the individual. When administering EONs of the present invention, the EONs are preferably dissolved in a solution compatible with the delivery method. EON expression plasmids can be delivered to retinal cells by providing the plasmid in an aqueous solution. Alternatively, preferred delivery methods for EONs or EON expression plasmids are viral vectors or nanoparticles, as outlined above. Preferably, viral vectors or nanoparticles are delivered to retinal cells, more preferably photoreceptor cells or RPE cells, where dysfunctional ABCA4 protein is present and causes disease. Such delivery to retinal cells or other relevant cells can be in vivo, in vitro, or ex vivo.

[0062] Those skilled in the art can select and adapt any of these or other commercially available alternative excipients and delivery systems to package and deliver EONs for use in the present invention and deliver EONs for the prevention, treatment, or delay of diseases or conditions associated with ABCA4-variants. "Prevention, treatment, or delay of diseases or conditions associated with ABCA4-variants" is preferably defined herein as preventing, halting, halting the progression, or reversing partial or complete visual impairment or blindness caused by a genetic defect in the ABCA4 gene.

[0063] When multiple separate EONs according to the present invention are used, the concentration or dosage defined herein can refer to the total concentration or dosage of all EONs used, or the concentration or dosage of each EON used or added.Therefore, in one embodiment, a composition is provided in which each or the total amount of EONs according to the present invention used is administered in an amount ranging from 0.01 to 20 mg / kg, preferably from 0.05 to 20 mg / kg.Suitable intravitreal dosage is between 0.05 mg and 5 mg per eye, preferably between about 0.1 and 1 mg, for example, about 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9 or 1.0 mg per eye.

[0064] A preferred AON according to the present invention is for the treatment of a disease or condition associated with an ABCA4-variant in a human individual. Preferably, the ABCA4 variant is selected from the group of variants listed in Table 1, more preferably the c.5882G>A mutation in exon 42 of the human ABCA4 gene, although it cannot be excluded that multiple AONs can be used to target other ABCA4 mutations in the same treatment.

[0065] In all embodiments of the present invention, the term "treatment" is understood to include prevention and / or delay of a disease or condition associated with an ABCA4 variant. An individual who can be treated using an EON according to the present invention may already have been diagnosed with a disease or condition associated with an ABCA4 variant. Alternatively, an individual who can be treated using an EON according to the present invention may not have yet been diagnosed with a disease or condition associated with an ABCA4 variant, such as Stargardt disease, but may be at high risk of developing such a disease or condition in the future, given their genetic background. Preferred individuals are human. In a preferred embodiment, the disease or condition associated with an ABCA4 variant is Stargardt disease. Thus, the present invention further provides an EON according to the present invention, a viral vector according to the present invention, or a composition according to the present invention for use as a pharmaceutical for treating a disease or condition associated with an ABCA4 variant that requires RNA editing of ABCA4, and for use as a pharmaceutical for preventing, treating, or delaying a disease or condition associated with an ABCA4 variant. Each of the features of the use is defined hereinabove.

[0066] The present invention further provides the use of an EON according to the present invention, or a viral vector according to the present invention, or a (pharmaceutical) composition according to the present invention, for the treatment of a disease or condition associated with an ABCA4-variant that requires RNA editing (or deamination of specific adenosines) of ABCA4 (pre-)mRNA. In a preferred embodiment, for all aspects of the present invention, the ABCA4-variant associated disorder, disease, or condition is caused by the c.5882G>A mutation in the human ABCA4 gene.

[0067] The present invention further provides the use of an EON according to the present invention, or a viral vector according to the present invention, or a composition according to the present invention, for the preparation of a medicament for treating a disease or condition associated with an ABCA4 variant that requires RNA editing of a specific adenosine in ABCA4(pre-)mRNA, and for the preparation of a medicament for preventing, treating, or delaying a disease or condition associated with an ABCA4 variant. Thus, in a further aspect, there is provided the use of an EON, viral vector, or composition as defined herein for the preparation of a medicament for treating a condition that requires RNA editing of a specific target adenosine in ABCA4(pre-)mRNA, and for the preparation of a medicament for preventing, treating, or delaying a disease or condition associated with an ABCA4 variant.

[0068] The treatment in the uses or methods according to the present invention may be at least once, at least one week, one month, several months, one, two, three, four, five, six years, or longer, for example, lifelong. It should be understood that the treatments disclosed herein do not edit cellular DNA, and that mutant (pre-)mRNA is continually produced by cells, and that mutant (pre-)mRNA may require continued editing to reverse the disease. Each EON or its equivalent as defined herein for use in accordance with the present invention may be suitable for direct administration to cells, tissues, and / or organs in vivo of an individual already suffering from or at risk of developing an ABCA4-variant-associated disease or condition, and may be administered directly in vivo, ex vivo, or in vitro. The frequency of administration of the EON, composition, compound, or supplemental compound of the present invention may depend on several parameters, such as the severity of the disease, the patient's age, the patient's mutation, the number of EONs (i.e., dose), the formulation of the EON, and the route of administration. The frequency can vary between daily, weekly, at least once every two weeks, once every three weeks, once every four weeks, once every five weeks, or once every longer. The dose range of the EON according to the present invention is preferably designed based on dose-escalation studies in clinical trials (in vivo use) with strict protocol requirements. The concentration of the EON as defined herein is used in a preferred embodiment in the range of 0.1 nM to 1 μM. Preferably, this range is for in vitro use in cell models such as retinal cells or tissues, preferably photoreceptor cells and / or RPE. More preferably, the concentration used is in the range of 1 to 400 nM, even more preferably 10 to 200 nM, and even more preferably 50 to 100 nM. When several EONs are used, this concentration or dose can refer to the total concentration or dose of the EONs or the concentration or dose of each EON added. In a preferred embodiment, a viral vector, preferably an AAV vector as described previously herein, as a delivery vehicle for the molecules according to the present invention is administered in a dose of 1 x 10 per injection. 9 ~1×10 17 viral particles, more preferably 1 x 10 per injection 10~1×10 12 It is administered at a dose of about 1000 viral particles.The above-mentioned EON concentration or dose range is the preferred concentration or dose for in vivo, in vitro or ex vivo use.Those skilled in the art will understand that depending on the EON used, the target cell treated, gene target and its expression level, the medium used and transfection and incubation conditions, the EON concentration or dose used may further vary and may need to be further optimized.

[0069] The present invention further provides a method for RNA editing of ABCA4 pre-mRNA in cells, preferably retinal cells, more preferably photoreceptor cells and / or RPE cells, comprising contacting the cells with an EON according to the present invention, or a viral vector according to the present invention, or a composition according to the present invention. The features of this embodiment are preferably as defined herein above. Contacting the cells with an EON according to the present invention, or a viral vector according to the present invention, or a composition according to the present invention can be carried out by any method known to those skilled in the art. Included are the use of methods for delivery of EONs, viral vectors, and compositions described herein. Contacting can be direct or indirect, and can be in vivo, ex vivo, or in vitro.

[0070] The present invention further provides a method for treating a disease or condition associated with an ABCA4 variant that requires RNA editing of a specific target adenosine in ABCA4 pre-mRNA in an individual in need thereof (e.g., a patient suffering from Stargardt disease), comprising contacting cells of the individual, preferably retinal cells, more preferably photoreceptor cells and / or cells in the RPE, with an EON according to the present invention, a viral vector according to the present invention, or a composition according to the present invention, to deaminate the specific adenosine in the pre-mRNA. Features of this aspect are preferably as defined herein above. Contacting cells, preferably retinal cells, more preferably photoreceptor cells and / or cells in the RPE, with an EON according to the present invention, a viral vector according to the present invention, or a composition according to the present invention can be carried out by any method known to those skilled in the art. This includes the use of methods for delivery of EONs, viral vectors, and compositions described herein. Contacting can be direct or indirect, and can be in vivo, ex vivo, or in vitro. Unless otherwise specified, each embodiment described herein may be combined with another embodiment described herein.

[0071] The sequence information provided herein should not be so narrowly construed as to require the inclusion of erroneously identified bases, as those skilled in the art will be able to identify such erroneously identified bases and will know how to correct such errors.

[0072] All patents and references cited herein are incorporated by reference in their entirety. [Example]

[0073] [Example 1] RNA editing of the c.5882G>A mutation in exon 42 of human ABCA4(pre-)mRNA using four EONs in a biochemical assay We aimed to determine whether the relatively common mutation causing Stargardt disease (c.5882G>A mutation in exon 42 of the human ABCA4 gene) could be repaired by an oligonucleotide capable of recruiting ADAR2 (ex vivo, in vitro, and in vivo) without the need for prior physical binding of the oligonucleotide to the ADAR2 enzyme. This mutation changes the wild-type GGA codon (encoding glycine) to GAA (encoding glutamic acid). Deamination of the first adenosine (A), the central nucleotide, to inosine (I) results in a reading frame seen as GGA, reverting to the wild-type state. Deamination of both adenosines to inosines generates a GGG codon. This codon also encodes glycine at the protein level, thus also resulting in the wild-type state. Therefore, deamination of the first adenosine, or deamination of the first adenosine plus the second adenosine in the same reaction, would be beneficial. Notably, if only the second adenosine were deaminated (yielding GAG, which also encodes glutamic acid), there would be no change in pathology. Thus, while deamination of only the first adenosine is preferred in RNA editing reactions, deamination of both adenosines is not potentially harmful, while deamination of the second adenosine alone would leave the amino acid (glutamic acid) derived from the already mutated mRNA unchanged.

[0074] The first biochemical assay, using four different EONs (ABCA4-1, -2, -3, and -4, respectively; see Figure 1 for sequences and chemical modifications), was performed by saturating full-length hADAR2 protein (recombinantly produced; Genscript) with double-stranded RNA at 12 nM and 1 nM concentrations (target RNA:EON ratio 1:6). The target RNA / EON mixture was heated to 95°C and slowly cooled to room temperature (30 min) to promote intermolecular interactions. Duplex formation was performed in annealing buffer (10 mM Tris-Cl pH 7.4; 1 mM EDTA; 100 mM NaCl). The resulting double-stranded RNA was diluted two-fold with water and then mixed with 10 ng / μL yeast tRNA (Invitrogen), 20 ng / μL poly(A) RNA (Qiagen), protease (EDTA-free protease inhibitor cocktail), and RNase inhibitor (RNasin ribonuclease inhibitor) in editing reaction buffer (15 mM Tris-Cl pH 7.4; 1.5 mM EDTA; 60 mM KCl; 3 mM MgSO4; 0.5 mM DTT; 3% glycerol; 0.003% NP-40; 40 mM K-glutamate). The editing reaction was initiated by adding hADAR2 protein to the reaction mixture (final volume 50 μL). The reaction was carried out at 37°C for up to 50 minutes. 5 μL samples were taken at eight different time points. Reactions at each time point were stopped by adding 5 μL of sample to 95 μL of boiling 3 mM EDTA and incubating at 95°C to denature the protein. Six μL of the stopped reaction mixture was then used as a template for cDNA synthesis using the Maxima reverse transcriptase kit with primer gBlock PCR REV: 5'-TGG ACC GAC TGG AAA CGT AG-3' (SEQ ID NO: 13) according to the manufacturer's instructions. The total reaction volume was 20 μL, and an extension temperature of 62°C was used.

[0075] For pyrosequencing analysis, PCR products were amplified using 1 μl of cDNA as template using the Amplitaq Gold 360 DNA Polymerase kit (Applied Biosystems) according to the manufacturer's instructions. The following primers were used at a concentration of 10 μM: Pyroseq Fwd ABCA4: 5′-ATG ATG ATG TGG CTG AAG AAA GA-3′ (SEQ ID NO: 14), and Pyroseq Rev ABCA4 Biotin: 5′-CCC AGT GAG CAT CTT GAA TGT-3′ (SEQ ID NO: 15). The latter primer also contains a biotin conjugated to its 5′ end, necessary for automated processing during the pyrosequencing reaction.

[0076] Because inosine pairs with cytidine during cDNA synthesis in the reverse transcription reaction, the nucleotide incorporated into edited positions during PCR is guanosine. The percentage of guanosine (edited) and adenosine (unedited) was defined by pyrosequencing. Pyrosequencing of PCR products and data analysis were performed using a PyroMark Q48 Autoprep instrument (Qiagen) according to the manufacturer's instructions, using 10 μl of PCR product input and 4 μM sequencing primer ABCA4-Seq2: 5'-CTC CAG CCC AGC AGT-3' (SEQ ID NO: 16). The setup specifically defined for this target RNA strand contained two sets of sequence information. The first set defined the sequence that the instrument would analyze, while the second set defined the order in which the sequencing reagents corresponding to each nucleotide were dispensed. It also included a blank control (i.e., a nucleotide that should not be incorporated at that specific position) used by the instrument to define the background signal. The analysis performed by the instrument provides results for selected nucleotides as the percentage of adenosines and guanosines detected at that position. Therefore, the degree of A-to-I editing at a selected position is measured by the percentage of guanosines at that position. Pyrosequencing revealed that editing efficiencies reached up to 140% (data not shown). If only one adenosine is edited in the target codon, efficiency cannot exceed 100%. This suggests that not only was the first adenosine in the target GAA codon deaminated to inosine, but the second adenosine was also edited. It cannot be excluded that in some cases, the target sequence was edited only at the second adenosine, resulting in GAG and no change in pathology (as outlined above). Nevertheless, the percentage of editing in each case indicated that the first and / or second adenosine of the targeted GAA codon was edited, indicating that we were able to target and edit this ABCA4 mutation very efficiently.

[0077] The results of biochemical experiments using the ABCA4-1 to 4 EONs are shown in Figure 2A, clearly demonstrating that all four EONs tested were able to relatively rapidly edit the target adenosines to inosines. Three EONs (ABCA4-1, -2, and -3) outperformed the ABCA4-4 EON, which was almost completely phosphorothioated. This demonstrates that we are now able to target the G>A mutation in ABCA4, enabling further development of pharmaceutical compounds useful for the treatment of Stargardt disease, in which patients suffer from a G>A mutation in the ABCA4 gene.

[0078] [Example 2] RNA editing of the c.5882G>A mutation in exon 42 of human ABCA4(pre-)mRNA using additional EONs in biochemical assays In addition to the four originally designed EONs, we designed ABCA4-5 through 29 (see Figure 1 for sequences and modifications). ABCA4-5 contains an intramolecular loop structure, while ABCA4-29 contains various additional mismatches to the target sequence. ABCA4-17 through 27 were designed to be significantly shorter than the original EONs and those previously used in the art, while ABCA4-21 through 27 each form a set of split EONs, accompanied by the ABCA4-28 oligonucleotide, which does not target the region of the mutation. The use of short and split EONs working in concert is described in UK Patent Application No. 2011428.6 (unpublished).

[0079] An additional set of EONs was also tested in a biochemical assay with minor adjustments compared to the assay used for the first four EONs disclosed in Example 1 (see Figure 1 for sequences and chemical modifications). Saturated full-length hADAR2 protein (recombinantly produced; Genscript) was used at double-stranded RNA concentrations of 24 nM and 2 nM (target RNA:EON ratio 1:3). The target RNA / EON mixture was heated to 95°C and slowly cooled to room temperature (1 hour 15 minutes) to promote intermolecular interactions. Duplex formation was performed in annealing buffer (10 mM Tris-Cl pH 7.4; 1 mM EDTA; 100 mM NaCl). The resulting double-stranded RNA was diluted 2-fold with water and 5-fold with 0.5x annealing buffer (5 mM Tris-HCl pH 7.4, 0.5 mM EDTA, 50 mM NaCl), and then mixed with 10 ng / μL yeast tRNA (Invitrogen), 20 ng / μL poly(A) RNA (Qiagen), protease (EDTA-free protease inhibitor cocktail), and RNase inhibitor (RNasin ribonuclease inhibitor) in editing reaction buffer (15 mM Tris-Cl pH 7.4; 1.5 mM EDTA; 60 mM KCl; 3 mM MgSO; 0.5 mM DTT; 3% glycerol; 0.003% NP-40; 40 mM K-glutamate). The editing reaction was initiated by adding hADAR2 protein to the reaction mixture (final volume 50 μL). The reaction was allowed to proceed for up to 60 minutes at 37°C. Five-microliter samples were withdrawn at six different intermediate time points. Each time point was terminated by adding the 5-microliter sample to 95°C of boiling 3 mM EDTA and incubating at 95°C to denature the protein. Ten microliters of the terminated reaction mixture was then used as a template for cDNA synthesis using the Maxima reverse transcriptase kit with primer gBlock PCR REV according to the manufacturer's instructions. The total reaction volume was 20 μl, and an extension temperature of 62°C was used.

[0080] For pyrosequencing analysis, the product was amplified using 1 μl of cDNA as template using the Amplitaq Gold 360 DNA Polymerase kit (Applied Biosystems) according to the manufacturer's instructions. The following primers were used at a concentration of 10 μM: Pyroseq Fwd2 ABCA4: 5'-ACT AAC CAA GAT TTA TCC AGG C-3' (SEQ ID NO: 17) and Pyroseq Rev ABCA4 Biotin (see above). The latter primer also contains a biotin conjugated to its 5' end, which is necessary for automated processing during the pyrosequencing reaction.

[0081] The results of the biochemical assays are shown in Figure 2B–G, which show the results for three or four EONs, respectively. Clearly, some EONs perform better than others, but the shorter versions, ABCA4-17, 18, 19, 20, 21, 22, 23, 24, 25, and 27, are less efficient than the longer versions. However, another short EON, ABCA4-26, performs very well. ABCA4-29, which has many mismatches with the target sequence, performs poorly compared to the other EONs, at least in this assay. Biochemical assays were not performed on the split EONs (ABCA4-21–27, respectively, along with ABCA4-28), which were tested in cell-based assays (see below).

[0082] [Example 3] RNA editing determined in cellular assays Cell transfection experiments were performed using a variety of different EONs (see Figure 1 for sequences and chemical modifications). Mouse RPE cells were transfected in DMEM (10% PBS + P / S) onto collagen-coated 12-well plates at 4 × 10 4 cells / cm 2After incubation at 5% CO2 and 37°C for 2-3 hours, cells were transfected with 500 ng of midigene and 500 ng of a plasmid expressing human ADAR2 in a total volume of 1 mL using Dharmafect duo (Dharmacon) as the transfection reagent (3 µL). The midigene is a construct containing a portion of the human ABCA4 genomic sequence, i.e., intron 40 to intron 43, including the middle exon flanked by exons 3 and 5 of rhodopsin regulated by T7 and preceded by a CMV promoter. Two versions of the midigene were generated: one containing the wild-type ABCA4 sequence and the other containing the c.5882G>A mutation in exon 42 ("mutant version"). The mutant midigene was cotransfected with EON as outlined herein.

[0083] After 24 hours of incubation at 37°C with 5% CO2, the medium was removed and the cells were washed with PBS. The cells were supplemented with fresh medium (900 μL) and transfected with 100 nM EON using Lipofectamine 2000 (Invitrogen) as the transfection reagent. When transfecting EON and helper AON (ABCA4-28), both oligonucleotides were transfected at 100 nM, forming a composition containing split EON. Transfection was performed in a total volume of 1 mL, with 1 μg EON per 2 μL of Lipofectamine 2000. After 6 hours of incubation at 37°C with 5% CO2, the medium was removed, the cells were washed, and fresh medium was added to the cells.

[0084] After a total incubation time of 24 h at 5% CO and 37°C, RNA was isolated from the cells using the RNeasy plus mini kit (Qiagen), and the RNA was dissolved in 50 μL of RNase-free water.

[0085] Next, random hexamer cDNA synthesis was performed using the Verso cDNA Synthesis Kit (Thermo Scientific) with 250 ng of RNA as template, according to the manufacturer's instructions. The total reaction volume was 20 μL. The samples were then used for Pyromark sequencing and ddPCR analysis.

[0086] For pyrosequencing analysis, products were amplified by PCR using 5 μl of cDNA as template using the Amplitaq Gold 360 DNA Polymerase kit (Applied Biosystems) according to the manufacturer's instructions. The following primers were used at a final concentration of 0.4 μM: Pyroseq Fwd2 ABCA4 and Pyroseq Rev ABCA4 Biotin (see above). Because inosine pairs with cytidine during cDNA synthesis in the reverse transcription reaction, the nucleotide incorporated at the edited position during PCR is guanosine. The percentage of guanosine (edited) versus adenosine (unedited) is determined by pyrosequencing. Pyrosequencing of PCR products and data analysis were performed using a PyroMark Q48 Autoprep instrument (Qiagen) according to the manufacturer's instructions, using 10 μl of PCR product input and 4 μM ABCA4-Seq2 primer (SEQ ID NO: 16). The setup specifically defined for this target RNA strand contained two sets of sequence information. The first set defined the sequence for the instrument to analyze the difference between guanosine and adenosine in the variable region, while the second set defined the order in which the corresponding determination reagents were dispensed to each nucleotide and also included a blank control (i.e., a nucleotide that should not be incorporated at that particular position) used by the instrument to define the background signal. The analysis performed by the instrument provided the results for the selected nucleotide as the percentage of adenosine and guanosine detected at that position. Therefore, the degree of A-to-I editing at a selected position was measured by the percentage of guanosine at that position.

[0087] The RNA editing results determined by Pyromark sequencing are shown in Figure 3 , which reveals that many of the tested EONs were indeed able to edit the targeted adenosine to inosine in exon 42 of human ABCA4 (pre-) mRNA in a cell-based assay using ADAR2 overexpression.

[0088] Next, we used duplex ddPCR assays to measure the copy number of wild-type (WT) and reference (ref) genes in the samples. The WT assay was specifically designed to detect G in ABCA4 exon 42 (from ABCA4 exon 41 to exon 43) and results in FAM-labeled positive droplets. The ref assay was designed to detect ABCA4 exon 41 from the reference exon Rho3 and results in HEX-labeled positive droplets.

[0089] Each ddPCR sample contained 1x ddPCR probe supermix (without dUTP) (Biorad), 0.6 μM of each primer (forward ref primer sequence: 5'-TTC TGC TAT GGG CAG CTC-3' (SEQ ID NO: 18); reverse ref primer sequence: 5'-TGT CTT TCT TCA GCC ACA TC-3' (SEQ ID NO: 19); forward WT primer sequence: 5'-CTA ACC AAG ATT TAT CCA GGC-3' (SEQ ID NO: 20); reverse WT primer sequence: 5'-CCT GAG GTC ACT GTG GT-3' (SEQ ID NO: 21)), and 0.6 μM of each double quench probe (ref probe sequence: 5'-HEX-CAC CGT CAA-ZEN-GGA GGA TTG CCG-IABkFQ-3' (SEQ ID NO: 22); WT probe sequence: 5'-6-FAM-TGT GTC GGA -ZEN- GTT CGC CCT GGA The PCR mixture contained GA-IABkFQ-3' (SEQ ID NO: 23) and 25 ng of cDNA in a total volume of 21 μL. Droplets were generated from the PCR mix using a QX200 droplet generator (Biorad). Droplet PCR was then performed in a T100 thermal cycler (Biorad) with a lid heated to 105°C and a ramp temperature of 2°C / sec. The polymerase was heat activated at 95°C for 10 min. In each cycle, denaturation was performed at 94°C for 30 s and annealing / extension was performed at 60°C for 30 s. This was repeated for a total of 40 cycles. The enzyme was inactivated at 98°C for 10 min, and the reaction was held at 4°C. Fluorescence signals from the droplets were measured using a QX200 droplet reader (Biorad). Absolute quantification was performed using QuantaSoft software (Bio-Rad). To determine the number of positive droplets, the fluorescence threshold was set at 3534 for the FAM signal and 2000 for the HEX signal. The WT / ref ratio was calculated using the copy number per reaction. Each sample was measured in duplicate. The average value was used for calculations.

[0090] Figure 4 shows the WT / ref ratio for each sample. It shows that the highest WT / ref ratio was measured in samples treated with smaller EONs. Because the shorter EONs were all complementary to the internal sequence of exon 42, whereas the longer versions overlapped the exon / intron boundary, we speculated that the use of longer EONs might have caused exon skipping of exon 42, thereby reducing the ddPCR signal and, therefore, the observed level of RNA editing. Therefore, we confirmed whether such exon skipping actually occurred.

[0091] The exon 42 skipping assay was specifically designed for the deletion of ABCA4 exon 42. The same ddPCR method as above was performed using the forward exon skipping primer: 5'-GAT GTG GCT GAA GAA AGA CA-3' (SEQ ID NO: 24), the same reverse primer as in the WT assay, and a double-quenched exon 42 skipping probe: 5'-6-FAM- AAC TAA CCA -ZEN- AGT GCT TTG GCC TCC-IABkFQ-3' (SEQ ID NO: 25). To determine the number of positive droplets, the fluorescence threshold was set at 4923 for the FAM signal and 2000 for the HEX signal. The ratio of exon 42 skipping to reference was calculated using the copy number per reaction.

[0092] Figure 5 shows the exon 42 skip / ref ratio for each sample, indicating that the highest skip ratio was measured in samples treated with EONs 13-16. Figure 4 shows that these EONs 13-16 exhibited relatively low RNA editing of the target adenosine. In contrast, relatively high and efficient RNA editing was observed with smaller EONs complementary to internal exon 42 sequences, and such exon skipping was barely detectable. Importantly, when ABCA4-21-27 (without ABCA4-28) was used, no skipping was observed at all. Therefore, it is preferable to use EONs that do not overlap with the 3' end of exon 42 and downstream introns, such as sequences in SEQ ID NOs: 8 or 9, but rather are fully complementary to sequences located entirely within exon 42.

[0093] These results clearly demonstrate that we have provided an EON capable of editing the target adenosine to inosine in exon 42 of human ABCA4 (pre-)mRNA in cells and in vitro biochemical settings. Interestingly, experiments using the split EONs ABCA4-24 and ABCA4-28 together resulted in a higher percentage of RNA editing than that observed with transfection of individual (single) EONs, indicating that such a configuration may be preferable. Furthermore, the use of shorter EONs (ABCA4-17 to 27, more preferably ABCA4-21 to 27) is preferred because they do not induce detectable levels of exon skipping. It should be understood that exon skipping of exon 42 is undesirable. Rather, exon skipping should not occur, and only specific RNA editing of the target adenosine in exon 42 should occur. Here, we demonstrate that this can be achieved. The preference for using smaller (segmented) EONs is supported by results obtained as described in UK Patent Application No. 2011428.6 (unpublished).

[0094] These experiments demonstrate that the inventors were able to target the G>A mutation in exon 42 of the human ABCA4 pre-mRNA, enabling the further development of pharmaceutical compounds useful for the treatment of Stargardt disease, in which patients suffer from a G>A mutation in the ABCA4 gene. The inventions described in the original claims of this application are listed below. [1] An RNA-editing oligonucleotide (EON) capable of forming a double-stranded complex with a target RNA molecule, wherein the EON, upon forming a complex with the target RNA molecule, is capable of recruiting an "RNA-acting adenosine deaminase" (ADAR) enzyme and forming a complex with the ADAR, thereby enabling the ADAR enzyme to deaminate a target adenosine in the target RNA molecule, wherein the EON does not form an intramolecular loop structure, and the target RNA molecule is a pre-mRNA or mRNA of human ABCA4, or a portion thereof. [2] The EON described in [1] above, wherein the EON comprises a central triplet of three consecutive nucleotides, and the nucleotide directly opposite the target adenosine is the central nucleotide of the central triplet and is cytidine. [3] The EON described in [1] or [2] above, wherein the ADAR enzyme is ADAR2. [4] An EON described in any one of [1] to [3] above, wherein one, two, or three nucleotides in the central triplet contain modifications, provided that the central nucleotide does not have a 2'-O-methyl (2'-OMe) modification or a 2'-methoxyethoxy (2'-MOE) modification in the sugar moiety. [5] The EON according to [4] above, wherein the modification is selected from the group consisting of deoxyribose (DNA), unlocked nucleic acid (UNA), and 2'-fluororibose. [6] The EON is selected from the group consisting of phosphorothioate, chirally pure phosphorothioate, Rp phosphorothioate, Sp phosphorothioate, phosphorodithioate, phosphonoacetate, tphosphonoacetate, phosphonacetamide, thiophosphonacetamide, phosphorothioate prodrug, S-alkylated phosphorothioate, H-phosphonate, methyl phosphonate, methyl phosphonothioate, methyl phosphate, methyl phosphorothioate, ethyl phosphate, ethyl phosphorothioate, boranophosphate, boranophosphorothioate, methylboranophosphate, methylboranophosphorothioate, methyl The EON according to any one of [1] to [5] above, comprising at least one non-naturally occurring internucleoside linkage modification selected from the group consisting of methylboranophosphonate, methylboranophosphonothioate, phosphorylguanidine, methylsulfonylphosphoramidate, phosphoramidite, phosphonamidite, N3'→P5' phosphoramidate, N3'→P5' thiophosphoramidate, phosphorodiamidate, phosphorothiodiamidate, sulfamate, dimethylenesulfoxide, sulfonate, triazole, oxalyl, carbamate, methyleneimino, thioacetamide, and derivatives thereof. [7] An EON according to [6] above, wherein 2, 3, 4, 5 or 6 terminal nucleotides at the 5' and 3' ends of the EON are linked by phosphorothioate linkages, preferably 5 terminal nucleotides at the 5' and 3' ends are linked by phosphorothioate linkages. [8] One or more nucleotides in the EON outside the central triplet may be substituted or unsubstituted, linear or branched lower (C 1 ~C 10 ) The EON according to any one of [1] to [7] above, comprising mono- or di-substitutions at the 2', 3' and / or 5' positions of the sugar selected from the group consisting of alkyl, alkenyl, alkynyl, alkaryl, allyl, or aralkyl; O-, S-, or N-alkyl; O-, S-, or N-alkenyl; O-, S-, or N-alkynyl; O-, S-, or N-allyl; O-alkyl-O-alkyl; -methoxy; -aminopropoxy; -methoxyethoxy; -dimethylaminooxyethoxy; and -dimethylaminoethoxyethoxy. [9] The EON according to any one of [1] to [8] above, wherein the target adenosine is part of a premature stop codon in pre-mRNA or mRNA of human ABCA4.

[10] The EON according to any one of [1] to [8] above, wherein the target adenosine is any one of the G>A mutations provided in Table 1, preferably the c.5882G>A mutation in exon 42 of the human ABCA4 gene.

[11] The EON described in any one of [1] to

[10] above, wherein the EON comprises or consists of a sequence selected from the group consisting of SEQ ID NOs: 1, 2, 3, 4, 5, 6, 7, 8, 9, and 11, preferably SEQ ID NOs: 8 and 9, more preferably SEQ ID NO: 9.

[12] A pharmaceutical composition comprising the EON according to any one of [1] to

[11] above and a pharmaceutically acceptable carrier.

[13] A composition comprising a set of two single-stranded antisense oligonucleotides (AONs) for use in deaminating a target adenosine in human ABCA4 pre-mRNA or mRNA, or a portion thereof, wherein one AON is an EON described in any one of [1] to

[11] above, and the other AON is a "helper AON", which is complementary to a stretch of nucleotides in human ABCA4 pre-mRNA or mRNA other than the stretch of nucleotides complementary to the EON, and which has a length of 16 to 22 nucleotides.

[14] The composition described in

[13] above, wherein the EON comprises or consists of the sequence of SEQ ID NO: 9, and the helper AON comprises or consists of the sequence set forth in SEQ ID NO: 10.

[15] The oligonucleotide according to any one of [1] to

[11] above, or the composition according to any one of

[12] to

[14] above, for use in treating Stargardt's disease.

[16] A method for deaminating at least one specific target adenosine present in a target RNA molecule in a cell, comprising: the target RNA molecule is human ABCA4 pre-mRNA or mRNA, or a portion thereof, and the method comprises: (i) providing the cells with the EON described in any one of [1] to

[11] above; (ii) allowing the EON to be taken up by the cells; (iii) annealing the EON to the target RNA molecule; (iv) allowing a mammalian ADAR enzyme containing a naturally occurring dsRNA-binding domain as found in a wild-type enzyme to deaminate the target adenosine in the target RNA molecule to inosine; and (v) optionally confirming the presence of said inosine in said target RNA molecule.

[17] Step (v) is a) determining the sequence of the target RNA molecule; b) assessing the presence of functional, extended, full-length and / or wild-type ABCA4 protein; c) assessing whether splicing of said pre-mRNA is modulated by said deamination; or d) using a functional readout, wherein the target RNA after deamination encodes a functional, full-length, extended and / or wild-type ABCA4 protein. The method according to

[14] above, comprising:

Claims

1. An RNA-editing oligonucleotide (EON) capable of forming a double-stranded complex with a target RNA molecule, wherein the EON, upon complexing with the target RNA molecule, is capable of recruiting an "RNA-acting adenosine deaminase" (ADAR) enzyme and forming a complex with the ADAR, thereby enabling the ADAR enzyme to deaminate a target adenosine in the target RNA molecule, wherein: The EON does not form an intramolecular loop structure, the EON comprises a central triplet of three consecutive nucleotides, the nucleotide immediately opposite the target adenosine is the central nucleotide of the central triplet and is a cytidine, the nucleotide 5' from the central nucleotide of the central triplet is a thymidine, and the nucleotide 3' from the central nucleotide of the central triplet is a thymidine or an adenosine; The target adenosine is the c.5882G>A mutation in exon 42 in the pre-mRNA or mRNA of human ABCA4. EON.

2. 2. The EON of claim 1, wherein one, two, or three nucleotides in the central triplet comprise a modification, with the proviso that the central nucleotide in the central triplet does not have a 2'-O-methyl (2'-OMe) modification or a 2'-methoxyethoxy (2'-MOE) modification in its sugar moiety.

3. 3. The EON of claim 2, wherein the modification is selected from the group consisting of deoxyribose (DNA), unlocked nucleic acid (UNA), and 2'-fluororibose.

4. The EON may be selected from the group consisting of phosphorothioates, chirally pure phosphorothioates, Rp phosphorothioates, Sp phosphorothioates, phosphorodithioates, phosphonoacetates, phosphonoacetates, phosphonacetamides, thiophosphonacetamides, phosphorothioate prodrugs, S-alkylated phosphorothioates, H-phosphonates, methyl phosphonates, methyl phosphonothioates, methyl phosphates, methyl phosphorothioates, ethyl phosphates, ethyl phosphorothioates, boranophosphates, boranophosphorothioates, methylboranophosphates, methylboranophosphorothioates, methyl 4. The EON of any one of claims 1 to 3, comprising at least one non-naturally occurring internucleoside linkage modification selected from the group consisting of methylboranophosphonate, methylboranophosphonothioate, phosphorylguanidine, methylsulfonylphosphoramidate, phosphoramidite, phosphonamidite, N3'→P5' phosphoramidate, N3'→P5' thiophosphoramidate, phosphorodiamidate, phosphorothiodiamidate, sulfamate, dimethylenesulfoxide, sulfonate, triazole, oxalyl, carbamate, methyleneimino, thioacetamide, and derivatives thereof.

5. The EON of claim 4, wherein 2, 3, 4, 5 or 6 terminal nucleotides at the 5' and 3' ends of the EON are linked by phosphorothioate linkages.

6. The EON of claim 5, wherein the five terminal nucleotides at the 5' and 3' ends are linked by phosphorothioate linkages.

7. One or more nucleotides in the EON outside the central triplet may be substituted or unsubstituted, linear or branched lower (C １ ~C １０ 7. The EON of any one of claims 1 to 6, comprising mono- or di-substitutions at the 2', 3' and / or 5' positions of the sugar selected from the group consisting of alkyl, alkenyl, alkynyl, alkaryl, allyl, or aralkyl; O-, S-, or N-alkyl; O-, S-, or N-alkenyl; O-, S-, or N-alkynyl; O-, S-, or N-allyl; O-alkyl-O-alkyl; -methoxy; -aminopropoxy; -methoxyethoxy; -dimethylaminooxyethoxy; and -dimethylaminoethoxyethoxy.

8. The EON according to any one of claims 1 to 7, wherein the EON comprises or consists of a sequence selected from the group consisting of SEQ ID NOs: 1, 2, 3, 4, 6, 7, 8, 9, and 11.

9. The EON of claim 8, wherein the EON comprises or consists of the sequence of SEQ ID NO: 8 or 9.

10. A pharmaceutical composition comprising the EON of any one of claims 1 to 9 and a pharmaceutically acceptable carrier.

11. 1. A method for in vitro or ex vivo deamination of at least one specific target adenosine present in a target RNA molecule in a cell, comprising: the target RNA molecule is human ABCA4 pre-mRNA or mRNA, or a portion thereof, and the method comprises: (i) providing the cells with an EON according to any one of claims 1 to 9; (ii) allowing the EON to be incorporated into the cells; (iii) annealing the EON to the target RNA molecule; and (iv) allowing a mammalian ADAR enzyme containing a naturally occurring dsRNA binding domain as found in a wild-type enzyme to deaminate the target adenosine to inosine in the target RNA molecule. A method comprising:

12. (v) The method described in claim 11, further comprising a step of confirming the presence of the inosine in the target RNA molecule.

13. Step (v) a) determining the sequence of the target RNA molecule; b) assessing the presence of functional, expanded, full-length and / or wild-type ABCA4 protein; c) assessing whether splicing of said pre-mRNA is modulated by said deamination; or d) using a functional readout, wherein the target RNA after deamination encodes a functional, full-length, extended and / or wild-type ABCA4 protein.

13. The method of claim 12, comprising:

Images