Antisense oligonucleotides for the treatment of poly-q disease

Abstract

The disclosure relates to the field of poly-Q diseases such as Huntington's Disease and Spinocerebellar ataxias (e.g., SCA3). The disclosure involves antisense oligonucleotides and their use in RNA editing methods, using endogenous ADAR enzymes, in targeting an adenosine in a (pre-) mRNA for human PIAS1, wherein the adenosine is in a codon encoding a serine residue in a region actively involved in the SUMOylation activity of the encoded PIAS1 protein. Through the editing, the serine is replaced by a glycine (= S510G variant), which decreases the SUMOylating activity of the PIAS1 protein. Human PIAS1 is involved in the breakdown of poly-Q proteins, such as human Huntingtin. Changing wild-type human PIAS1 to a S510G variant diminishes, decreases, or prevents the occurrence of toxic fragments from the poly-Q protein, its aggregation into inclusion bodies and thereby prevents the sequestration of essential cellular proteins, thereby treating the poly-Q disease.

Description

[0001] ANTISENSE OLIGONUCLEOTIDES FOR THE TREATMENT OF POLY-Q DISEASE

[0002] TECHNICAL FIELD

[0003] This disclosure relates to the field of medicine, and in particular to the field of neurological disorders. The disclosure describes antisense oligonucleotides that mediate nucleotide-specific RNA editing in the human PIAS1 gene transcript to bring about amino acid changes of the encoded protein that influence its activity in SUMOylation, thereby providing a treatment for certain poly-Q disorders.

[0004] BACKGROUND

[0005] Human polyglutamine (poly-Q) disorders are a group of neurodegenerative disorders caused by the atypical expansion of tandemly arranged cytosine-adenine-guanine (CAG) trinucleotide repeats (TN Rs) within the coding region of the causative gene, resulting in an extended stretch of glutamine (Q) residues in the mutated protein. There are nine human poly-Q disorders: Huntington’s Disease (HD), Spinal and bulbar muscular atrophy (SBMA, or Kennedy’s disease), six of the Spinocerebellar ataxias: SCA1 , SCA2, SCA3 / Machado-Joseph Disease (MJD), SCA6, SCA7 and SCA17, and Dentatorubral-pallidoluysian atrophy (DRPLA, or Haw River syndrome). All poly-Q disorders share some salient features, such as: i) they are autosomal dominant, except SBMA which is X-linked recessive; ii) they predominantly affect the central nervous system (CNS); iii) they are generally adult onset and progressive in nature; iv) they display genetic anticipation; v) they show high or complete penetrance; vi) they show neuronal accumulation of the toxic protein aggregates; and viii) they culminate in death and have no cure till date.

[0006] Extensive genotype-phenotype correlation studies show that the aberrantly expanded poly-Q stretch is sufficient to induce disease phenotype and neurodegeneration. The length of CAG repeat plays a decisive role in determining the age of onset, severity of disease and the rate of progression: the longer the poly-Q repeat, the earlier the disease onset with more severe symptoms and accelerated progression. Besides repeat length, the homozygous or heterozygous state of an individual also influences the severity and age of disease onset. The expanded poly- Q stretches in the coding regions of the proteins like Huntingtin (HTT), Ataxin-1 , -2, -3, etc., contribute to the disease pathogenesis and cytotoxicity via diverse mechanisms: 1) loss-of- function that affects the normal interactions and functions of the proteins and 2) toxic gain-of- function of the poly-Q stretch which increases the interaction with other proteins and hampers their inherent functioning (Lieberman AP et al. Annu Rev Pathol. 2019, 14:1-27). Notably, most lines of evidence indicate that the toxic gain-of-function is the major cause of pathogenesis which is common to all poly-Q disorders and can be ascribed to the non-conventional protein folding, perturbed post-translational modification and aggregation into inclusion bodies as well as sequestration of other essential proteins necessary for maintaining cellular vitality (Laundos TL et al. Front Cell Dev Biol. 2023, 11 :1252521). Said differently, the toxic gain-of-function by poly- Q proteins can be attributed to their non-specific interaction and sequestration of other protein that results in indirect deficiency of normally functioning proteins in the cellular pool. The sequestration process is the main feature of the inclusion bodies which in a location-specific (cytoplasm / nucleus) manner traps enormous numbers of essential proteins. Thus, rather than outcome(s) of a single protein mutation, bulk sequestration of other proteins exert wider ranges of deleterious effects. The formation of poly-Q inclusion bodies is a well-established multi-step process. The mutant poly-Q containing proteins are highly susceptible to misfolding and such misfolded proteins readily undergo proteolytic cleavage to generate sticky and toxic oligopeptides with heightened tendency to aggregate together (Havel LS et al. Mol Brain. 2009, 2:21 ; Jimenez- Sanchez M et al. Cold Spring Harb Perspective Med, 2017, 2:a024240). The ‘toxic-fragment hypothesis’ postulates that the release of the expanded poly-Q fragment from the protective sequence of the full-length protein is crucial to initiate the aggregation process. The proteolytic cleavage of HTT, Ataxin-3 and Atrophin-1 is also essential for their nuclear localization while TBP and Ataxin-7 (which are already located in the nucleus) may or may not require proteolytic cleavage.

[0007] One way to address poly-Q disorders is by looking for ways to lower the ability of the mutant forms of the protein to cleave, either at a genetic level, or at a protein level, or to lower the expression of the mutant form. For instance, WQ2012 / 018257 and WQ2019 / 043027 revealed the use of antisense oligonucleotides to induce skipping of a part of exon 12 of the human HTT transcript to remove the caspase-6 cleavage site from the encoded HTT protein, thereby preventing cleavage at that position (Evers MM et al. Nucl Acid Ther. 2014, 24:4-12; Casaca- Carreira J et al. Biomed Pharmacother. 2016, 84:93-96). Strategies using / - / TT-silencing siRNA, miRNA, or zinc-finger proteins (sometimes applying adeno-associated virus (AVV) vectors) targeting the expanded CAG repeats or (CAG)? antisense oligonucleotides to lower the expression of the HTT transcript or the encoded protein have also been reported {e.g., Datson NA et al. PLOSone 2017, 12(2):e0171127). A variety of drug-based interventions against poly-Q disorders have been developed (reviewed in Tandon S et al. Life Sciences. 2024, 344:122562).

[0008] Modulating effects of other post-translational modifications such as SUMOylation have been implicated in conditions such as HD, SCA1 , SCA3, and DRPLA (lleda H et al. Biochem Biophys Res Common. 2002, 293:307-313). SUMOylation is a form of post-translational modification of proteins, whereby a string of Small Ubiquitin-related Modifier (SUMO) proteins / peptides are enzymatically ligated to specific lysine residues of the substrate protein. SUMOylation may affect the fate of the protein in terms of localization (e.g., cytoplasmic, or nuclear), degradation and / or function, and is thereby involved in a wide variety of cellular processes, such as cell cycle, transcription, DNA repair, etc. Unlike in the ubiquitin proteosome system, SUMOylation is a process in which target proteins are not directly degraded. To date, four SUMO isoforms (SUMO1 to SUMO4) have been isolated in mammalian cells. Like ubiquitination, SUMOylation occurs through a series of enzymatic reactions. In the first step, SUMO precursors are converted to the active form of SUMO protein, exposing a glycine (Gly)- Gly motif through cleavage of the carboxyl-terminal tails of the SUMO precursors via the hydrolase activity of sentrin-specific proteases (SENPs). The Gly-Gly motif of the mature SUMO protein is covalently conjugated with a conserved catalytic cysteine in the heterodimeric SUMO- activating enzyme E1 (SAE1 / SAE2) via a thioester bond in an ATP-dependent reaction. Ubiquitin- conjugating enzyme 9 (UBC9), the only SUMO-conjugating E2 enzyme identified in mammalian cells, then attaches the SUMO protein directly to a lysine located within the consensus sequence: ^P-KX-E / D, in which is a hydrophobic amino acid and X is any amino acid, in the substrate. SUMO E3 ligases, such as the Protein Inhibitor of Activated STAT (PIAS) family of proteins (PIAS1 , PIAS3, PIASx and PIASy), ring finger protein 4 (RNF4), Ran binding protein 2 (RANBP2), and the Polycomb protein 2 (Pc2), stimulate protein SUMOylation by associating with both UBC9 and substrates, thereby promoting poly-SUMO chain formation. It has been demonstrated that enhanced SUMOylation of the N-terminal fragment of HTT at K6, K9, and K15 positions facilitates the accumulation of neurotoxic poly-Q oligomers (Steffan JS et al. Science. 2004, 304:100-104), indicating that SUMOylation of HTT adds to the severity of HD. PIAS1 , as indicated above, is an E3 SUMO ligase and is, among other things, responsible for SUMOylation of a variety of substrate (also called ‘client’ or ‘target’) proteins, including HTT (Ochaba J et al. Neuron. 2016, 90:507- 520), indicating that PIAS1 regulates mutant HTT toxic-fragment accumulation and HD- associated phenotypes. SUMOylation is important in HD because both wild-type and mutant HTT can be modified by SUMO-2 / 3 at its N-terminus, which subsequently modulates the homeostasis of HTT proteins.

[0009] Recently, a variant of PIAS1 was identified that affects the severity and onset of poly-Q disorders (Lee YH et al. Movement Disorders 2022, 37:767-777). This variant, in which an A>G mutation present in the encoding gene (referred to as position A1528G, or c.1528A>G), following the sequence as referenced by NCBI reference sequence No. NM_016166, gives rise to a serine (S) to glycine (G) mutation at amino acid position 510 in the mature PIAS1 protein (S510G, or p.S510G), was shown to delay onset and to mitigate severity of patients with HD caused by expanded poly-Q chains. Similarly, patients suffering from SCA3 were partially protected from disease by the presence of this PIAS1 variant. In a mouse model recapitulating the poly-Q disease (in this case HD), the knock-in of the PIAS1 S510G variant improved several aspects of HD, such as the formation of HTT aggregates, neuronal pathology (inclusions) and grip strength. Without wishing to be bound by theory, the S510G variant is suggested to bind less well to mutant HTT protein in comparison to wild-type HTT protein. Hence, this lowered affinity for the mutant HTT protein results in reduced SUMOylation of the mutant poly-Q protein, thereby changing the degradation fate of the mutant poly-Q protein. Because degradation of HTT and aggregation into inclusion bodies and the effect of sequestration of other essential proteins appear to contribute to the severity of HD, as outlined above, the PIAS1 S510G variant may be a protective variant for individuals that normally would be prone to develop and / or that may suffer from a poly-Q disorder, such as HD and SCA3.

[0010] BRIEF SUMMARY

[0011] Disclosed herein is an antisense oligonucleotide (AON) capable of forming a doublestranded complex with a region of an endogenous (human) wild-type PIAS1 transcript molecule in a cell, wherein the transcript molecule is a pre-mRNA or an mRNA molecule, wherein the region of the PIAS1 transcript molecule comprises a target adenosine, wherein the nucleotide in the AON that is directly opposite the target adenosine is the orphan nucleotide, and wherein the double-stranded complex can recruit an endogenous ADAR enzyme to deaminate the target adenosine into an inosine, thereby editing the PIAS1 transcript molecule. Preferably, the deamination of the target adenosine results in an encoded PIAS1 protein with a decreased ability (or loss-of-function) to SUMOylate a substrate (‘target’ or ‘client’) protein. A preferred client protein that is less efficiently targeted for SUMOylation by the human PIAS1 protein (after it has been amended through the action of the AON as disclosed herein) is encoded by the human HTT gene that comprises at least 36 CAG repeats in a CAG trinucleotide repeat (TNR) expansion, more preferably 40 or more CAG repeats. Hence, the client protein of the mutant variant PIAS1 protein is then a human HTT protein comprising 36, 37, 38, 39, 40, or more glutamine residues in a poly- Q tract (stretch). Another preferred client protein that is less efficiently targeted for SUMOylation by the human PIAS1 protein (after it has been amended through the action of the AON as disclosed herein) is encoded by the human ATXN3 gene that comprises at least 45 CAG repeats in a CAG TNR expansion, more preferably 55 or more CAG repeats. Hence, the client protein of the mutant variant PIAS1 protein is then a human Ataxin-3 protein comprising 45, 46, 47, 48, 49, 50, 51 , 52, 53, 54, 55, or more glutamine residues in a poly-Q tract (stretch).

[0012] In one aspect the deamination of the target adenosine into an inosine in the PIAS1 transcript molecule results in an amino acid change from serine to glycine at position 510 of the encoded PIAS1 protein (mutant variant S510G). In one aspect, the cell in which the deamination occurs is a neuron, preferably a cell of the central nervous system (CNS), more preferably a brain cell, even more preferably a cell involved in the rise, and progression of the poly-Q disorder.

[0013] In another aspect, disclosed is an AON as disclosed herein for use in the treatment of a poly-Q disorder, preferably wherein the poly-Q disorder is Huntington’s disease (HD) or Spinocerebellar Ataxia 3 (SCA3).

[0014] In another aspect, disclosed is the use of an AON as disclosed herein in the manufacture of a medicament for the treatment of a poly-Q disorder, preferably HD and / or SCA3.

[0015] In another aspect, disclosed is a method of treating a poly-Q disorder, preferably HD and / or SCA3, in an individual in need thereof, the method comprising contacting a wild-type PIAS1 polynucleotide in a cell of the subject with an AON as disclosed herein, thereby treating the individual.

[0016] In another aspect, disclosed is a method of deaminating a target adenosine in a PIAS1 pre-mRNA or mRNA molecule in a cell, the method comprising the steps of: i) providing the cell with an AON as disclosed herein; ii) allowing uptake by the cell of the AON; iii) allowing annealing of the AON to the PIAS1 pre-mRNA or mRNA molecule; iv) allowing an endogenous ADAR enzyme to deaminate the target adenosine in the target RNA molecule to an inosine; and optionally v) identifying the presence of the inosine in the target RNA molecule, preferably wherein the target adenosine is in the AGO codon encoding serine at position 510 of the human PIAS1 protein, and wherein the deamination of the target adenosine into an inosine results in a glycine at position 510.

[0017] BRIEF DESCRIPTION OF THE DRAWINGS

[0018] One or more embodiments will now be described, by way of example only, with reference to the accompanying drawings, in which:

[0019] FIG. 1 shows on top part of the human PIAS1 (pre-)mRNA target transcript sequence (5’ to 3’; SEQ ID NO:51) including the target adenosine (A) in capitals and in bold face and the serine encoding AGO codon underlined. Below the target sequence, the sequences (also 5’ to 3’) are given of an initial set of 50 antisense oligonucleotides (AONs RM107489 to RM107538; SEQ ID NO:1 to 50, respectively) that were designed to bring about editing of the target adenosine. All 50 AONs comprise a tri-antennary GalNAc moiety (OP-042; Hongene Biotech) linked through a phosphodiester (PO) linkage on the 5’ terminus (L001e; not depicted) for targeting to hepatocytes. The chemical modifications in the AONs are as follows: Gm, Am, Um, and Cm are 2’-OMe modified guanosine, adenosine, uridine, and cytidine, respectively; m5Ce is 2’-MOE modified 5- methylcytidine; Ge is 2’-MOE modified guanosine; Ae is 2’-MOE modified adenosine; m5Ue is 2’- MOE modified 5-methyluridine (also sometimes named “Te”; 2’-MOE modified thymidine); Af, Uf, Gf, and Cf are 2’-F modified adenosine, uridine, guanosine, and cytosine, respectively; Zd is a cytidine analog that is also referred to as a nucleoside carrying a Benner’s base (as further outlined herein), with a deoxy moiety (= DNA) at the 2’ ribose position; m5Ud is a deoxynucleotide with a 5-methyluridine; Gl is a Locked Nucleic Acid (LNA) with a guanine nucleobase; Al is an LNA with an adenine nucleobase; * refers to a phosphorothioate (PS) linkage; I refers to a (1 ,3- dimethylimidazolidin-2-ylidene) phosphoramidate (PNdmi) linkage;Arefers to a methylphosphonate (MP) linkage; and0refers to a PO linkage.

[0020] FIG. 2 shows the editing percentages obtained after transfection of 100nM of the indicated AONs in Primary Human Hepatocytes (PHH’s), using ddPCR.

[0021] FIG. 3 shows the editing percentages obtained after transfection of 200nM of the indicated AONs in PHH’s, using ddPCR. DETAILED DESCRIPTION

[0022] Poly-Q diseases (or poly-Q disorders, which is an alternative term that may be used throughout the description) are a group of disorders that find their cause in genes that comprise a CAG-repeat expansion in the coding region of the transcript that - upon translation - gives rise to expanded poly-glutamine (or poly-Q) tracts (or ‘stretches’). The poly-Q disorders have been assumed to arise because of protein gain-of-function, whereby the CAG repeat mutations give rise to poly-Q-rich proteins that form brain-specific inclusions. These inclusions may contain cellular components such as ubiquitin, HSP70, the proteasome and transcription factors. As such, the poly-Q disorders are part of a wider category of diseases mediated by TNR expansions, of which the majority is in non-coding regions of the gene or transcript. The latter category of TNR disorders is typically mediated by RNA gain-of-function mechanisms, in contrast to the TNR disorders that are poly-Q disorders, which are mediated (primarily) by protein gain-of function mechanisms.

[0023] Huntington’s Disease (HD), perhaps the most well-known poly-Q disease, is a severe genetic neurodegenerative disorder caused by mutations in the HTT gene. The HTT gene codes for a protein of about 348 kDa called Huntingtin, also abbreviated to HTT (or sometimes as Htt). HTT is involved in microtubular transport and function of vesicles in the cell. Its polypeptide sequence is characterized by the poly-Q tracts referred to above. In the non-pathogenic state, human HTT contains stretches of less than 36, typically around 30, glutamine residues. In the diseased state, these poly-Q stretches are significantly elongated, containing a minimum of 36 glutamines, with full penetrance if the poly-Q stretch is 40+. The severity and age of onset of HD is determined by the degree of elongation of the poly-Q stretches. In some patients, the poly-Q stretches may be increased to hundreds of glutamines. These poly-Q stretch elongations are caused by DNA instability, leading to expansion of the CAG codon stretches.

[0024] Expanded poly-Q HTT shows aberrant behaviour in terms of vesicular transport, function and localization and they have a higher propensity to aggregate, forming toxic cell inclusions leading to neuronal dysfunction and associated disorders. HD is a progressive disease that typically starts in the 3rdor 4thdecade of life, but onset and severity may vary and is dependent on the degree of the poly-Q expansion. Symptoms are involuntary movements (chorea), general motor dysfunction, psychiatric disorders, and dementia.

[0025] SCA3 is another well-known examples of a poly-Q disease. For instance, SCA3 is the most common spinocerebellar ataxia (SCA) in Taiwan. Other SCAs caused by poly-Q tract expansion are SCA1 , SCA2, SCA6, SCA7 and SCA17. Further examples of poly-Q disorders are spinobulbar muscular atrophy (SBMA) and dentatorubral-pallidoluysian atrophy (DRPLA), as outlined above.

[0026] It is an object of the present invention to create the protective PIAS1 protein variant S510G by targeted RNA editing, preferably ADAR-mediated targeted RNA editing, in individuals suffering from a poly-Q disorder, preferably HD or SCA3. It depends on the poly-Q disease which PIAS1 target (client, or substrate) protein is affected. For HD, HTT is the client protein. For SCA3, Ataxin- 3 is the client protein. None of this matters from a therapeutic perspective, as the target protein that is to be changed by targeted RNA editing remains the same, viz. PIAS1. It is yet another preferred embodiment to create this PIAS1 S510G variant in the brain, preferably in the cells affected in poly-Q disorders. These may be the neurons of the caudate and / or putamen in the case of HD. According to a preferred aspect of the invention, targeted RNA editing is mediated by ADAR that is endogenous (= naturally present) to the cell. According to yet another preferred embodiment, the ADAR endogenous to the cell is being recruited to the target RNA in conjunction with a guide RNA (an AON as disclosed herein) that is at least partially complementary to the target RNA in the region of the proposed change. According to yet another preferred embodiment, the AON is designed to bind the target RNA in the region of the intended edit to form a (partially) double stranded RNA sequence allowing the endogenous ADAR to bind and exerts an enzymatic deamination of a target adenosine in the target RNA, changing the adenosine into inosine. Preferably, the target adenosine is the adenosine in the triplet coding for serine 510 of PIAS1 , effectively changing the triplet so that - upon canonical translation - the triplet codes for glycine. Further aspects of the invention and how to put it into practice are set forth in the detailed description herein.

[0027] The AONs as disclosed herein can recruit deaminating enzymes, such as ADAR1 and / or ADAR2 that are endogenously present in a cell. An AON as disclosed herein can mediate RNA editing of a target adenosine present in a target RNA molecule after it is bound to the target RNA molecule, because the deaminating enzymes are recruited to the double-stranded AON / target RNA molecule complex and subsequently deaminate the target adenosine into an inosine.

[0028] The oligonucleotides are herein abbreviated to “AONs”, but sometimes also referred to as ‘editing oligonucleotides’, or ‘EONs’, even though the RNA editing event itself is performed by the deamination enzyme and the action of the oligonucleotide only triggers the RNA editing to take place. There is a constant need for improving the pharmacokinetic properties of the AONs without negatively affecting the efficiency in which the target adenosine is edited in the target RNA, and / or without negatively affecting the stability of the AON itself, which is constantly prone to breakdown because of nucleases present in a natural cell. Many chemical modifications are available for the generation of AONs (and many have been applied in the art). However, many of these properties are not always compatible with the desire of achieving efficient RNA editing. In the search for better pharmacokinetic properties, it was found earlier that a 2’-O-methoxyethyl (or 2’- methoxyethoxy, or 2’-MOE) modification of the ribose of some, but not all, nucleotides surprisingly appeared compatible with efficient ADAR engagement and editing (WO2019 / 158475). In a similar fashion, it was found earlier that a PS linkage at some, but not all, internucleoside linkages surprisingly appeared compatible with efficient ADAR engagement and editing (WO2019 / 219581). Also, it was found earlier that phosphonoacetate linkage modifications and / or unlocked nucleic acid (UNA) ribose modifications of some, but not all, positions in the AON appeared compatible with efficient engagement of an enzyme with nucleotide deamination activity and with subsequent deamination (W02020 / 165077). Whereas the properties of phosphonoacetate and UNA modifications were known as such, the compatibility thereof with engagement of enzymes with nucleotide deamination activity and with the deamination reaction was not known. Also, it was found earlier that methyl phosphonate (MP) linkage modifications of some, but not all, positions in the AON appeared compatible with efficient engagement of an enzyme with nucleotide deamination activity and with subsequent deamination (W02020 / 201406). Also, it was found earlier that mesyl phosphoramidate linkages (PNms) at some positions in the AON appeared compatible with efficient engagement of an enzyme with nucleotide deamination activity and with subsequent deamination (WO2024 / 200278).

[0029] Disclosed herein are AONs that can provide (mediate, cause, or trigger) RNA editing of a target adenosine in a target transcript molecule, such as pre-mRNA and / or mRNA. The target transcript molecule as disclosed herein (human PIAS1 (pre-) mRNA) is preferably wildtype and the editing results in a transcript that encodes a protein with a loss-of-function, which in this case means that the mutant PIAS1 protein variant (S510G) is no longer able, or less active, or has a lowered or diminished ability to actively contribute in the SUMOylation of the mutant forms of human HTT and / or the mutant forms of human Ataxin-3. The loss-of-function subsequently results in a lowered, diminished, or most preferably even absent breakdown of the mutant proteins which would normally result in the formation of inclusion bodies and sequestration of bound proteins, and eventually disease (HD and SCA3, respectively).

[0030] RNA editing is often applied to correct G>A mutations that cause a disease, such as a change from TGG (encoding tryptophan; W) to TAG or TGA, each being a premature translation stop codon. This is not the case in the present disclosure. Here, RNA editing is applied to introduce a mutation that causes the resulting protein to act less effectively, or preferably with a complete loss in functionality in respect of part of the target proteins. Particularly, the variant that is disclosed herein (S510G) may keep its SUMOylation functionality towards normal, wildtype proteins, but as outlined herein, preferably is less capable of SUMOylating the mutant forms of poly-Q causing proteins, such as mutant HTT and mutant Ataxin-3.

[0031] Non-limiting examples of transcript molecules (as disclosed in the art) that are targeted using RNA editing for a variety of treatments are SERPINA 1 (for the treatment of alphal- antitrypsin (A1AT) deficiency; see e.g., WO2016 / 097212, WO2017 / 220751 , WO2018 / 041973, and WO2021 / 243023), IDUA (for the treatment of Hurler syndrome; see e.g., WO2017 / 220751 , WO2018 / 041973, and WO2021 / 209010), LRRK2 (for the treatment of Parkinson’s disease; see e.g., WO2016 / 097212, WO2017 / 220751 , WO2018 / 041973, WO2021 / 231673 and WO2021 / 242903), ABCA4 (for the treatment of Stargardt disease; see e.g., W02021 / 130313 and WO2021 / 231830), USH2A (for the treatment of Usher syndrome; see e.g., W02020 / 157008, WO2020 / 219981 and WO2021 / 136404), APP (see e.g., WO2021 / 113270), CMT1A (see e.g., WO2021 / 113390), ASS1 (see e.g., WO2021 / 231675), GJB2 (see e.g., WO2021 / 231679), MECP2 (for the treatment of Rett syndrome; see e.g., WO2019 / 071274 and WO2021 / 231680), OTOF (for the treatment of autosomal recessive non-syndromic hearing loss; see e.g., WO2021 / 231685 and WO2021 / 231692), XLRS (see e.g., WO2021 / 231691), and PCSK9 (for the treatment of hypercholesterolemia; see e.g., WO2023 / 152371).

[0032] The present disclosure relates to AONs that mediate RNA editing, using endogenous (naturally present) ADAR enzymes in the host cell (preferably neurons, more preferably cells of the CNS) of an adenosine present in the transcript of the human PIAS1 gene. As described, the PIAS1 protein is involved in SUMOylation processes. An AON as disclosed herein aims to lower the activity of the PIAS1 protein and its ability to SUMOylate mutant HTT and mutant Ataxin-3, wherein ‘mutant’ means carrying a disease-giving number of glutamine residues in poly-Q stretches. Targeting the adenosine in the codon for serine at position 510 in the PIAS1 protein amino acid sequence and changing it to an inosine resulting in a glycine at that position (S510G) is a preferred example of a wildtype PIAS1 transcript that is turned into a mutant variant. It cannot be excluded that other PIAS1 variants are equally deficient in SUMOylating mutant forms of proteins involved in poly-Q diseases, while maintaining their normal functionality towards nonmutant forms of such proteins and proteins not related to poly-Q disorders. Similarly, it may be that other variants of PIAS1 have greater loss-of-function in comparison to the S510G variant. Therefore, the S510G variant serves as an example of PIAS1 variants that may be generated using the teaching of the present disclosure.

[0033] Definitions

[0034] Whenever reference is made to an oligonucleotide, oligo, ON, ASO, oligonucleotide composition, antisense oligonucleotide, AON, (RNA) editing oligonucleotide, EON, and RNA (antisense) oligonucleotide, both oligoribonucleotides and deoxyoligoribonucleotides are meant unless the context dictates otherwise. Potentially the oligonucleotide may completely lack RNA and DNA nucleotides (as they appear in nature) and may consist completely of modified nucleotides. Whenever reference is made to an ‘oligoribonucleotide’ it may comprise the bases A, G, C, U, or I. Whenever reference is made to a ‘deoxyoligoribonucleotide’ it may comprise the bases A, G, C, T, or I. However, an AON as disclosed herein may comprise a mix of ribonucleotides and deoxyribonucleotides. When a deoxyribonucleotide is used, hence without a modification at the 2’ position of the sugar, the nucleotide is often abbreviated to dA (or Ad), dC (or Cd), dG (or Gd), dl (or Id), m5Ud (or T) in which the ‘d’ represents the deoxy nature of the nucleoside, while a ribonucleoside that is either normal RNA or modified at the 2’ position is often abbreviated without the ‘d’, and often abbreviated with their respective modifications and as explained herein. The term ‘nucleoside’ refers to the nucleobase linked to the (deoxy)ribosyl sugar, without phosphate groups. A ‘nucleotide’ is composed of a nucleoside and one or more phosphate groups. The term ‘nucleotide’ thus refers to the respective nucleobase-(deoxy)ribosyl- phospholinker, as well as any chemical modifications of the ribose moiety or the phospho group. Thus, the term would include a nucleotide including a locked ribosyl moiety (comprising a 2’-4’ bridge, comprising a methylene group or any other group), an unlocked nucleic acid (UNA), a threose nucleic acid (TNA), a nucleotide including a linker comprising a phosphodiester, phosphonoacetate, phosphotriester, PS, phosphoro(di)thioate, MP (or sometimes abbreviated to MeP), methyl thiophosphonate, phosphoramidate linkages, PNdmi according to the structure of formula (IV) as described herein, and a linkage according to the structure of formula (I) as described herein, preferably PNms. Sometimes the terms nucleobase, nucleoside and nucleotide are used interchangeably, unless the context clearly requires differently, for instance when a nucleoside is linked to a neighbouring nucleoside and the linkage between these nucleosides is modified. As stated herein, a nucleotide is a nucleoside plus one or more phosphate groups. The terms ‘ribonucleoside’ and ‘deoxyribonucleoside’, or ‘ribose’ and ‘deoxyribose’ are as used in the art.

[0035] Sometimes the terms adenosine and adenine, guanosine and guanine, cytidine and cytosine, uracil and uridine, thymine and thymidine / uridine, inosine, and hypoxanthine, are used interchangeably to refer to the corresponding nucleobase on the one hand, and the nucleoside or nucleotide on the other. The nucleobase thymine (T) is also known as 5-methyluracil (sometimes abbreviated to m5U) and is an uracil (U) derivative; thymine and 5-methyluracil can be interchanged throughout the document text. Likewise, the nucleotide thymidine is also known as 5-methyluridine and is a uridine derivative; thymidine and 5-methyluridine can be interchanged throughout the document text.

[0036] Whenever reference is made to nucleotides in the oligonucleotide, such as cytosine, 5- methylcytosine, 5-hydroxymethylcytosine, 5-formylcytosine, 5-acetylcytosine, 5-hydroxycytosine, and p-D-glucosyl-5-hydroxymethylcytosine are included. Whenever reference is made to adenine, N6-methyladenine, 8-oxo-adenine, 2,6-diaminopurine and 7-methyladenine are included. Whenever reference is made to uracil, dihydrouracil, isouracil, N3-glycosylated uracil, pseudouracil, 5-methyluracil, N1-methylpseudouracil, 4-thiouracil and 5-hydroxymethyluracil are included. Whenever reference is made to guanine, 1-methylguanine, 7-methylguanosine, N2,N2- dimethylguanosine, N2,N2,7-trimethylguanosine and N2,7-dimethylguanosine are included. Whenever reference is made to nucleosides or nucleotides, ribofuranose derivatives, such as 2’- deoxy, 2’-hydroxy, and 2’-O-substituted variants, such as 2’-O-methyl (2’-0Me), are included, as well as other modifications, including 2’-4’ bridged variants. Whenever reference is made to oligonucleotides, one or more linkages may be a naturally occurring PO linkage, whereas the remaining linkages between two mononucleotides may be a modified linkage. Examples of such modified linkages are phosphonoacetate, phosphotriester, PS, phosphoro(di)thioate, MP, phosphoramidate linkages, phosphoryl guanidine, thiophosphoryl guanidine, sulfono phosphoramidate, PNdmi according to the formula (IV) as further outlined below, and the linkage structure according to formula (I) as further outlined in detail below, preferably PNms.

[0037] 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’ in relation to a numerical value x is optional and means, e.g., x+10%.

[0038] The word ‘substantially’ does not exclude ‘completely’, e.g., a composition which is ‘substantially free from Y’ may be completely free from Y. Where relevant, the word ‘substantially’ may be omitted from the definition of the invention.

[0039] The term ‘conducive to’ or ‘mediate’ can be used interchangeably with ‘capable of facilitating’. When used in the context of an AON that is conducive to ADAR editing (or can mediate ADAR editing), this means that the AON, after entry into the cell, interacts with the target RNA sequence, thereby forming a double stranded structure which is recognized by the ADAR enzyme, which can then deaminate the target adenosine into an inosine. Hence, the AON itself does not have the enzymatic function (the ADAR enzyme has), but it can trigger, induce, cause, organize, mediate, provide, give, produce, facilitate, result in RNA editing after binding to the target RNA molecule.

[0040] The term ‘mismatch’ is used herein to refer to opposing nucleotides in a double stranded RNA complex which do not form perfect base pairs according to the Watson-Crick base pairing rules. In the historical sense, mismatched nucleotides are G-A, C-A, ll-C, A-A, G-G, C-C, Il-Il pairs. A nucleotide mismatches also with the opposing nucleotide when that opposite nucleotide is a-basic (lacks a nucleobase). In some embodiments AONs as disclosed herein comprise fewer than four mismatches with the target sequence, for example 0, 1 or 2 mismatches. ‘Wobble’ base pairs are G-ll, l-ll, l-A, and l-C base pairs. When a II is placed opposite the target A, there is no mismatch, and the AON may be 100% complementary, although an iso-uridine (iso-ll) opposite the target adenosine qualifies as a mismatch, since it does not pair according to the Watson-Crick rules of base pairing. When a C is placed opposite the target A, there is at least 1 mismatch between the AON and the target sequence. Although a G:G pairing would be considered a mismatch, that does not necessarily mean that the interaction is unstable, which means that the term ‘mismatch’ may be somewhat outdated based on the current disclosure where a Hoogsteen base-pairing may be seen as a mismatch based on the origin of the nucleotide but still be relatively stable. An isolated G:G pairing in duplex RNA can for instance be quite stable but will still be defined as a mismatch. Analysis of natural targets of ADAR enzymes has indicated that these generally include mismatches between the two strands that form the RNA helix edited by ADAR1 or 2. It has been suggested that these mismatches 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 optimal patterns of paired / mismatched nucleotides between the AONs and the target RNA also appears important to the development of efficient ADAR-based AON therapy.

[0041] The term ‘complementary’ as used herein refers to the fact that the AON hybridizes under physiological conditions to a second nucleic acid strand. Examples are (i) when the AON as a first nucleic acid strand (= guide oligonucleotide) forms a heteroduplex RNA editing oligonucleotide complex with second complementary nucleic acid strand {in vitro), or (ii) when it forms a double stranded complex with the target RNA molecule. The term does not necessarily mean that each nucleotide in a nucleic acid strand has a perfect pairing with its opposite nucleotide in the opposite sequence. In other words, while an AON may be complementary to a target sequence, there may be mismatches, wobbles and / or bulges between the AON and the target sequence, while under physiological conditions that AON still hybridizes to the target sequence such that the cellular RNA editing enzymes can deaminate the target adenosine to an inosine. The term ‘substantially complementary’ therefore also means that despite the presence of the mismatches, wobbles, and / or bulges, the AON has enough matching nucleotides with the target sequence that under physiological conditions the AON hybridizes to the target RNA molecule. As shown herein, an AON may be complementary, but may also comprise one or more mismatches, wobbles and / or bulges with the target sequence, if under physiological conditions the AON is able to hybridize to its target.

[0042] ADARs are multidomain proteins with N-terminal double stranded RNA binding domains (dsRBDs) and C-terminal deaminase domains. Two ADAR genes encode catalytically active ADARs in humans {ADAR encoding ADAR1 proteins and ADARB1 encoding the ADAR2 protein). ADAR1 is expressed in two protein isoforms (p110 and p150) that differ in their N-terminal structures. Since the substrate for ADARs is an RNA duplex, the enzymes access the reactive adenosine using a base flipping mechanism (Stephens OM et al. Biochemistry. 2000. 39(40): 12243-12251). Also, because ADARs require duplex RNA for activity, their reaction can be directed to specific adenosines in different transcripts using complementary guide strands for duplex formation at the target sites. This approach is currently being pursued to develop therapeutic guide strands that recruit ADARs to correct disease-causing mutations in RNA (Qu L. et al. Nat. Biotechnol. 2019. 37(9): 1059-1069; Merkle T. et al. Nat. Biotechnol. 2019. 37(2):133- 138; Katrekar D. et al. Nat. Methods 2019. 16(3):239-242; Monian P. et al. Nat. Biotechnol. 2022. 40(7): 1093-1102). While this approach is promising, ADARs have sequence preferences that make certain adenosines disfavored for reaction, limiting the current scope of this approach. For instance, the nearest neighbor nucleotide preferences for ADARs show a strong bias against reaction at adenosines in 5’-GA sites (Eggington JM et al. Nat. Commun. 2011. 2(319): 1-9). This preference is explained by structural studies of ADAR2 bound to transition state analogcontaining RNA that suggest a clash between the 2-amino group of the 5’-G and G489 of the ADAR2 loop involved in stabilizing the flipped-out conformation required for the adenosine deamination reaction (Matthews et al. Nat Struct Mol Biol 2016. 23(5):426-433). Earlier work with fusion proteins bearing ADAR deaminase domains indicated that editing efficiency at 5’-GA sites could be improved with a G-A or G-G pair at the 5’ nearest neighbor (Schneider MF et al. Nucleic Acids Res. 2014. 42(10):p.e87). However, the basis for this effect has not been reported nor has this effect been established for full length ADARs bearing native dsRBD RNA binding domains. In WO2024 / 013361 it is shown that G-A and G-G pairs on the 5’ side of an editing site improve editing efficiency compared to a 5’ G-C pair for full length ADAR2 and ADAR1 p110. Using X-ray crystallography, the structure of an active fragment of human ADAR2 bound to duplex RNA bearing a G:G pair adjacent to an editing site was determined. WO2024 / 013361 discloses that the ADAR enzyme prefers the Gsyn:Ganti pairing since the 2-amino group of the 5’-G in the syn conformation does not clash with the minor groove of the enzyme. It was shown that the use of nucleosides capable of stable pairing with the 5’-G in the syn conformation enables more efficient editing within 5’-GA target sites, providing a solution to the problem in deaminating these unfavored editing sites. In the present disclosure related to editing of the adenosine in the AGO codon at position 1528 no 5’-G is present next to the target adenosine in the target sequence. There is an adenosine in that position.

[0043] The term ‘orphan nucleotide’ relates to the nucleotide in the AON that is directly opposite the target adenosine, which is the adenosine that is deaminated by the deaminating enzyme. The orphan nucleotide may be a natural cytidine, a deoxycytidine, a uridine, or a deoxyuridine. It may also be a chemically modified nucleotide, as further described in detail below, or a known or chemically modified analog of a natural (deoxy)cytidine, such as a nucleotide carrying a Benner’s base, or a known or chemically modified analog of a natural (deoxy)uridine, such as iso-uridine, as further outlined in detail below.

[0044] A ‘nucleotide analog’ refers to an analog of a nucleic acid nucleotide. The nucleotide analog is an analog of adenosine, guanosine, cytidine, thymidine, uridine, deoxyadenosine, deoxyguanosine, deoxycytidine, deoxythymidine or deoxyuridine.

[0045] The term ‘downstream’ in relation to a nucleic acid sequence means further along the sequence in the 3' direction; the term ‘upstream’ means the converse. Thus, in any sequence encoding a polypeptide, the start codon is upstream of the stop codon in the sense strand but is downstream of the stop codon in the antisense strand. The same holds true for the AONs as disclosed herein. Nucleotides that are upstream of the orphan nucleotide in the antisense oligonucleotide are located towards the 5’ terminus, and nucleotides that are downstream of the orphan nucleotide are located towards the 3’ terminus.

[0046] The nucleotide ‘numbering’ in an AON as disclosed herein is such that the orphan nucleotide is number 0 and the nucleotide 5’ from the orphan nucleotide is number +1. Counting is further positively (+) incremented towards the 5’ end and negatively (-) incremented towards the 3’ end, wherein the first nucleotide 3’ from the orphan nucleotide is number -1. The internucleoside linkage numbering in the AON is such that linkage number 0 is the linkage 5’ from the orphan nucleotide, and the linkage positions in the oligonucleotide are positively (+) incremented towards the 5’ end and negatively (-) incremented towards the 3’ end.

[0047] References to ‘hybridisation’ typically refer to specific hybridisation and exclude non-specific hybridisation. Specific hybridisation can occur under experimental conditions chosen, using techniques well known in the art, to ensure that most stable interactions between probe and target are where the probe and target have at least 70%, preferably at least 80%, more preferably at least 90% sequence identity.

[0048] The term ‘splice mutation’ relates to a mutation in a gene that encodes for a pre-mRNA, wherein the splicing machinery is dysfunctional in the sense that splicing of introns from exons is disturbed and due to the aberrant splicing, the subsequent translation is out of frame resulting in premature termination of the encoded protein. Often such shortened proteins are degraded rapidly and do not have any functional activity.

[0049] Whenever a ‘naked’ form in relation to the AON as disclosed herein is referred to, it means that the AON is manufactured in a laboratory or manufacturing facility, through which it is generally chemically modified to prevent it from rapid degradation after it enters the mammalian body or a tissue, or cell, upon administration. A naked form of an AON is therefore different from a form in which the AON is encoded (and delivered) by a viral genome or from a plasmid vector. When such viral vectors or plasmid vectors are administered, the encoded AON is expressed from the viral vector genome or from the plasmid in the cell to which the viral vector or plasmid vector is delivered. Consequently, the AON is then not chemically modified, and comprises solely naturally occurring RNA nucleotides.

[0050] The length of the AON as disclosed herein, and when delivered in a naked form is preferably 15, 16, 17, 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, 50, 51 , 52, 53, 54, 55, 56, 57, 58, 59, or 60 nucleotides in length. However, when the AON as disclosed herein is to be delivered through the expression of a viral vector, then the AON may be longer, such as 70, 80, 90, 100, 150, or 200 or more nucleotides in length.

[0051] The term ‘HEON’ refers to a heteroduplex double-stranded complex molecule wherein an AON as disclosed herein is hybridized to a partially or fully complementary, partially of fully overlapping sense oligonucleotide. Because the AON as disclosed herein often has specified chemical modifications that are different from the chemical modifications in the sense strand, the two strands form such a heteroduplex RNA editing oligonucleotide complex. The sense strand may be chemically modified almost in its entirety, similar or different to what is performed in the AON as disclosed herein, for example by providing nucleotides with a ribose sugar moiety carrying a 2’-OMe substitution, a 2’-F substitution, or a 2’-MOE substitution. It is to be understood that the sense strand present in the HEON is a different entity in comparison to the target RNA molecule in the cell. The sense strand in an HEON is preferably 15, 16, 17, 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, 50, 51 , 52, 53, 54, 55, 56, 57, 58, 59, or 60 nucleotides in length. The HEON is often generated in vitro and used as a delivery tool to protect the AON from degradation when administered to the cell. In other words, the HEON is preferably formed before the AON is administered to the cell. Preferred aspects of HEONs that may be used for AONs as disclosed herein are discussed in PCT / EP2023 / 079290 (not published).

[0052] Embodiments

[0053] Disclosed herein is an AON capable of forming a double-stranded complex with a region of an endogenous human wild-type PIAS1 transcript molecule in a cell, wherein the transcript molecule is a pre-mRNA or an mRNA molecule, wherein the region of the PIAS1 transcript molecule comprises a target adenosine, wherein the nucleotide in the AON that is directly opposite the target adenosine is the orphan nucleotide, and wherein the double-stranded complex can recruit an endogenous ADAR enzyme to deaminate the target adenosine into an inosine, thereby editing the PIAS1 transcript molecule. In one aspect, the deamination of the target adenosine results in an encoded PIAS1 protein with a decreased ability or loss-of-function to SUMOylate a target protein. As outlined above, PIAS1 is an E3 SUMO ligase and is, among other things, responsible for SUMOylation of a variety of substrate (also referred to as ‘client’ or ‘target’) proteins, including HTT (Ochaba J et al. 2016). This indicates that PIAS1 regulates mutant HTT toxic-fragment accumulation and HD-associated phenotypes. SUMOylation is important in HD because both wild-type and mutant HTT can be modified by SUMO-2 / 3 at its N-terminus, which subsequently modulates the homeostasis of HTT proteins. As discussed infra, a variant of PIAS1 was identified that affects the severity and onset of poly-Q disorders (Lee YH et al. 2022). This variant is referred to as S510G, following the transcript and amino acid sequence referenced by NCBI reference sequence No. NM_016166, and is caused by an A>G mutation in the PIAS1 gene, wherein the serine residue at position 510 has changed to a glycine residue in that position. Hence, in a preferred aspect herein, the deamination of the target adenosine into an inosine results in an amino acid change from serine to glycine at position 510 of the encoded PIAS1 protein. The ‘target protein’ that is SUMOylated through the activity of PIAS1 is preferably a protein involved in the occurrence of a poly-Q disorder, more preferably HD or SCA3. In a preferred aspect, the deamination leads to a loss-of-function of the PIAS1 protein in the sense that it has a diminished, lowered or even absent ability to SUMOylate the mutant form of HTT and / or Ataxin-3. It is known in the art that a normal length of the poly-Q stretch in HTT is 6 to 35, and that a stretch of 36 to 121 or longer is considered pathogenic. Hence, for the sake of the present disclosure, the ‘target protein’ for which the S510G PIAS1 variant has a lowered affinity (or activity) regarding SUMOylation is an HTT protein with 36 or more glutamine residues in the poly-Q stretch. The same holds true for Ataxin-3, where the art reveals that 12 to 44 glutamines in the poly-Q stretch is considered non-pathogenic (or wildtype) and 55 to 86, or more glutamines in the poly-Q stretch are considered pathogenic (or mutant). Hence, for the sake of the present disclosure, the ‘target protein’ for which the S510G PIAS1 variant has a lowered affinity (or activity) regarding SUMOylation is an Ataxin-3 protein with 55 or more glutamine residues in the poly-Q stretch. It does not exclude that the S510G variant also has a lowered affinity or activity towards the non-pathogenic versions of HTT and / or Ataxin-3, but that would be less relevant as it would not be therapeutically relevant in the treatment of HD and / or SCA3, respectively. In one aspect, the cell in which the PIAS1 transcript molecule targeting takes place is a neuron, preferably a cell of the central nervous system (CNS), more preferably a brain cell. In one aspect, the cell is a cell in which inclusion bodies are detected, which is related to the occurrence of HD and / or SCA3. SCA3 affected tissues can - in humans - often be found in the substantia nigra, the caudate nucleus, the putamen, the globus pallidus, and the pons. HD affected tissues - in humans - often involve the cerebral cortex, the caudate nucleus, the putamen, the globus pallidus, and the red nucleus. Cells within these areas are preferred target cells when it comes to deaminating the adenosine in the AGC codon, encoding serine in position 510 of the human PIAS1 protein.

[0054] The present disclosure also relates to AONs as disclosed herein, that comprise at least one non-naturally occurring chemical modification, and / or comprising one or more additional non- naturally occurring chemical modifications in the ribose, linkage, or base moiety. Preferably, the orphan nucleotide is a deoxynucleotide comprising a 6-amino-5-nitro-3-yl-2(1 H)-pyridone nucleobase (Benner’s base; Z). In one preferred aspect, the AON as disclosed herein comprises one or more modifications in the linkage moiety, which is each independently selected from the group consisting of: PS, phosphonoacetate, phosphorodithioate, MP, sulfonylphosphoramidate, PNdmi, and PNms. In a preferred aspect, linkage position -2 is an MP linkage or a PNms linkage, more preferably a PNms linkage. In another preferred aspect, the linkage between the most terminal two nucleotides on the 5’ and / or 3’ terminus of the AON is a PNdmi linkage or a PNms linkage, preferably both terminal linkages are PNms linkages. If another moiety is attached, or conjugated to the AON, that moiety may be connected to the oligonucleotide by a PO linkage for instance. If that is the case, then the linkage connecting the two most terminal nucleosides is considered as the ‘terminal’ linkage (5’ and 3’) in an AON, even though a further linkage may be present connecting another moiety to the AON. In one aspect, the AON as disclosed herein comprises one or more nucleotides comprising a mono- or di-substitution at the 2', 3' and / or 5' position of the ribose, each independently selected from the group consisting of: -OH; -F; substituted or unsubstituted, linear or branched lower (C1-C10) alkyl, alkenyl, alkynyl, alkaryl, allyl, or aralkyl, that may be interrupted by one or more heteroatoms; -O-, S-, or N-alkyl; -O-, S-, or N- alkenyl; -0-, S-, or N-alkynyl; -O-, S-, or N-allyl; -O-alkyl-O-alkyl; -methoxy; -aminopropoxy; - methoxyethoxy; -dimethylamino oxyethoxy; and -dimethylaminoethoxyethoxy.

[0055] In a preferred aspect, the AON as disclosed herein is selected from the group consisting of SEQ ID NO:1 to 50, preferably from the group consisting of SEQ ID NO:27, 28, 30, 31 , 33, 34, 35, 36, 37, 38, and 39. In a preferred aspect, the AON is covalently or non-covalently, directly or through a linker, bound to a triterpene glycoside, preferably AG1856. AG1856 is also referred to as a ‘saponin’ (see WO2024 / 153801 and WO2021 / 122998). The saponin allows for improved endosomal release of the AON once taken up by the target cell and may provide more efficient editing during treatment. The saponin, preferably AG1856, may be co-administered (or administered before / after administration of the AON), but is preferably conjugated to the AON, either directly or indirectly through one or more linking moieties. The skilled person is, based on these teachings, capable of to find the best format of such conjugates to serve the purpose of targeting the PIAS1 pre-mRNA and / or mRNA and to direct the AON to the target cell of choice.

[0056] The present disclosure also relates to a vector, preferably a viral vector, more preferably an adeno-associated virus (AAV) vector, comprising a nucleic acid molecule encoding an AON as disclosed herein, wherein the orphan nucleotide is a cytidine or a uridine. As outlined elsewhere herein, when the AON is encoded by another nucleic acid and transcribed therefrom, it cannot comprise the chemical modifications as outlined herein. Hence, the orphan nucleotide is not one that carries a Benner’s base but is then preferably cytidine or uridine.

[0057] Disclosed herein is also a pharmaceutical composition comprising an AON, or a vector as disclosed herein, and a pharmaceutically acceptable carrier. Pharmaceutically acceptable carriers are well known in the art.

[0058] The present disclosure also relates to an AON as disclosed herein for use in the treatment of a poly-Q disorder. In a preferred aspect, the poly-Q disorder that is treated is Huntington’s disease (HD) or Spinocerebellar Ataxia 3 (SCA3), although other poly-Q disorders are not particularly excluded from benefitting from the AONs of the present disclosure, if PIAS1 also is involved in the SUMOylation of Ataxin-1 , Ataxin-2, etc.

[0059] The present disclosure also relates to a use of an AON as disclosed herein in the manufacture of a medicament for the treatment of a poly-Q disorder, preferably HD and / or SCA3.

[0060] The present disclosure also relates to a method of treating a poly-Q disorder, preferably HD and / or SCA3, in an individual in need thereof, the method comprising contacting a wild-type PIAS1 polynucleotide in a cell of the subject with an AON as disclosed herein, thereby treating the individual. The present disclosure also relates to an in vitro, ex vivo, or in vivo method of deaminating a target adenosine in a PIAS1 pre-mRNA or mRNA molecule in a cell, the method comprising the steps of: i) providing the cell with an AON as disclosed herein; ii) allowing uptake by the cell of the AON; iii) allowing annealing of the AON to the PIAS1 pre-mRNA or mRNA molecule; iv) allowing an endogenous ADAR enzyme to deaminate the target adenosine in the target RNA molecule to an inosine; and optionally v) identifying the presence of the inosine in the target RNA molecule. In a preferred method, the target adenosine is in the AGC codon encoding serine at position 510 of the human PIAS1 protein, and the deamination of the target adenosine into an inosine results in a glycine at position 510.

[0061] The present disclosure also relates to a nucleic acid molecule for editing a target adenosine in a human PIAS1 pre-mRNA or mRNA molecule, wherein the target region is SEQ ID NO:51 , and wherein the target adenosine is in the AGC codon encoding serine at position 510 of the human PIAS1 protein. Preferably, the nucleic acid molecule is selected from the group consisting of SEQ ID NO:1 to 50, preferably from the group consisting of SEQ ID NO:27, 28, 30, 31 , 33, 34, 35, 36, 37, 38, and 39.

[0062] In one aspect, the AON as disclosed herein is 15, 16, 17, 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, 50, 51 , 52, 53, 54, 55, 56, 57, 58, 59, or 60 nucleotides in length. In one aspect, an AON as disclosed herein is in a naked form. In one aspect, an AON as disclosed herein is in a circular format. In one aspect, an AON as disclosed herein is not in a naked form but is expressed from the genome of a viral vector. In one aspect, an AON as disclosed herein is not in a naked form but is expressed from an expression vector such as a plasmid. In one aspect, when the AON as disclosed herein is not in a naked form, the AON is 15 to 60 nucleotides in length as indicated above, or in another embodiment, from 61 to 300 nucleotides in length.

[0063] The present disclosure relates to an AON capable of forming a double-stranded complex with a region of an endogenous human wild-type PIAS1 transcript molecule in a cell, wherein the transcript molecule is a pre-mRNA or an mRNA molecule, wherein the region of the PIAS1 transcript molecule comprises a target adenosine, wherein the nucleotide in the AON that is directly opposite the target adenosine is the orphan nucleotide, wherein the double-stranded complex can recruit an endogenous ADAR enzyme to deaminate the target adenosine into an inosine, thereby editing the PIAS1 transcript molecule, and wherein the orphan nucleotide is a deoxynucleotide comprising a 6-amino-5-nitro-3-yl-2(1 H)-pyridone nucleobase (Zd), wherein the nucleotide at position +1 is a deoxycytidine (Cd) or a 2’-MOE modified 5-methylcytidine (m5Ce), wherein the nucleotide at position -1 is a deoxynucleotide carrying a 5-methyluracil nucleobase (m5Ud), and wherein the internucleoside linkage at linkage position -2 is an MP or a PNms linkage.

[0064] Chemical modifications

[0065] Various chemistries and modifications are known in the field of oligonucleotides that can be readily used in accordance with the invention. All chemical modifications listed herein that may be used in the AON as disclosed herein may also be used for a sense strand that is complementary to the AON, when the AON and the complementary strand form a HEON complex, such as described in W02024 / 084048 and as disclosed above, except that the opposite sense strand does not have an orphan nucleotide. Hence, the modification related to the orphan nucleotide relate only to the AON as disclosed herein, but all other modifications relate to the AON as disclosed herein and any (protecting) sense oligonucleotide that may be used together with the AON in a pharmaceutical product. This includes the use of hydrophobic moieties (such as tocopherol and cholesterol) and cell-specific ligands, that have also been described herein, and in detail in W02024 / 084048, which may either be bound to the AON or its opposite strand, or both.

[0066] The skilled person knows that an oligonucleotide, such as an AON as outlined herein, generally consists of repeating monomers. Such a monomer is most often a nucleotide or a chemically modified nucleotide. The most common naturally occurring nucleotides in RNA are adenosine monophosphate (A), cytidine monophosphate (C), guanosine monophosphate (G), and uridine monophosphate (II). These consist of a pentose sugar, a ribose, a 5’-linked phosphate group which is linked via a phosphate ester, and a T-linked base. The sugar connects the base and the phosphate and is therefore often referred to as the “scaffold” of the nucleotide. A modification in the pentose sugar is therefore often referred to as a ‘scaffold modification’. The original pentose sugar may be replaced in its entirety by another moiety that similarly connects the base and the phosphate. It is therefore understood that while a pentose sugar is often a scaffold, a scaffold is not necessarily a pentose sugar. Examples of scaffold modifications that may be applied in the monomers of the AON as disclosed herein are disclosed in W02020 / 154342, WO2020 / 154343, and W02020 / 154344.

[0067] A nucleoside in the AON as disclosed herein may be a natural nucleoside (deoxyribonucleoside or ribonucleoside) or a non-natural nucleoside. It is noted that for RNA editing, in which double-stranded RNA is generally the substrate for enzymes with deamination activity (such as ADARs), ribonucleosides are considered ‘natural’, while deoxyribonucleosides may then be, for the sake of argument, considered as non-natural, or considered as modified, simply because DNA is not present in the RNA-RNA double stranded (natural) substrate configurations. The skilled person appreciates that when the nucleotide has a natural ribose moiety, it may still be non-naturally modified in the base and / or the linkage.

[0068] It is recognized in the art that common limiting factors in oligonucleotide-based therapies are the oligonucleotide’s ability to be taken up by the cell, when delivered per se, or ‘naked’ (= without applying a viral vector or plasmid), the biodistribution and the resistance to nuclease- mediated breakdown. The skilled person is aware, and it has been described in detail in the art, that a variety of chemical modifications can assist in overcoming such limitations. Examples of such now commonly used chemical modifications are the 2’-OMe, 2’-F, 2’,2’-diF, and 2’-MOE modifications of the sugar and the use of PS linkages between nucleosides, as described herein. Scaffold modifications (ribose)

[0069] The ribose 2’ groups in all nucleotides of the AON as disclosed herein, except for the ribose sugar moiety of the orphan nucleotide that has certain limitations in respect of compatibility with RNA editing, can be independently selected from 2’-H (i.e., DNA), 2’-OH (i.e., RNA), 2’-0Me, 2’-M0E, 2’-F, or 2’-4’-linked (for instance a locked nucleic acid (LNA)), or other ribosyl T- substitutions, 2’ substitutions, 3’ substitutions, 4’ substitutions or 5’ substitutions. The orphan nucleotide in the AON that comprises no other chemical modifications to the ribose sugar, the base, or the linkage preferably does not carry a 2’-0Me or 2’-M0E substitution when the nucleobase is a naturally occurring cytosine, but may carry a 2’-F, a 2’,2’-difluoro (di F) , or 2’-ara- F (FANA) substitution or may be DNA. WO2024 / 013360 discloses the modification of the 2’ position of the ribose sugar moiety of the orphan nucleotide by a 2’,2’-disubstituted substitution such as di F, which is also applicable to what is disclosed herein. The 2’-4’ linkage can be selected from many linkers known in the art such as a methylene linker, amide linker, or constrained ethyl linker (cEt).

[0070] An AON as disclosed herein may comprise one or more nucleotides carrying a 2’-MOE ribose modification. Also, an AON as disclosed herein may comprise one or more nucleotides not carrying a 2’-MOE ribose modification, or wherein the 2’-MOE ribose modifications are at positions that do not prevent the ADAR enzyme from deaminating the target adenosine. An AON as disclosed herein may comprise a 2’-OMe ribose modification at a position that does not comprise a 2’-MOE ribose modification. An AON as disclosed herein may comprise deoxynucleotides at positions that do not comprise a 2’-MOE or a 2’-OMe ribose modification, or other 2’ ribose substitution. An AON as disclosed herein may comprise one or more nucleotides comprising a 2’ substitution comprising a 2’-MOE, 2’-OMe, 2’-OH, 2’-deoxy, TNA, 2’-F, diF modification, 2’-fluoro- 2’-C-methyl modification, or a 2’-4’-linkage (i.e., a bridged nucleic acid such as a locked nucleic acid (LNA or examples mentioned in e.g. WO2018 / 007475)). Other nucleic acid monomers that may be used in an AON as disclosed herein are arabinonucleic acids and 2’-deoxy-2’- fluoroarabinonucleic acid (FANA), for instance for improved affinity purposes. The 2’-4’ linkage can be selected from linkers known in the art, such as a methylene linker or constrained ethyl linker. A wide variety of 2’ modifications that may present in an AON as disclosed herein are known in the art, including but not limited to the modifications outlined in detail in WO2016 / 097212, WO2017 / 220751 , WO2018 / 041973, WO2018 / 134301 , WO2019 / 219581 , WO2019 / 158475, and WO2022 / 099159. In all cases, the modifications should be compatible with RNA editing such that the AON fulfils its role as an oligonucleotide that can form a double stranded complex with the target RNA and by generating this double-stranded nucleic acid complex, recruit a deaminating enzyme, which can subsequently deaminate the target adenosine. Where a monomer in an AON as disclosed herein comprises an unlocked nucleic acid (UNA) ribose modification, that monomer can have a 2’ position comprising the same modifications discussed above, such as a 2’-MOE, a 2’-OMe, a 2’-OH, a 2’-deoxy, a 2’-F, a diF, a 2’-fluoro-2’-C-methyl, an arabinonucleic acid, a FANA, or a 2’-4’-linkage (i.e. , a bridged nucleic acids such as a locked nucleic acid (LNA)). In one aspect, the AON as disclosed herein comprises at least one nucleotide comprising a threose nucleic acid (TNA) ribose modification. In one aspect, the AON as disclosed herein comprises at least one nucleotide with a sugar moiety that comprises a 2’-F modification. A preferred position for the nucleotide that carries a 2’-F modification is position -3 in the AON.

[0071] Base modifications

[0072] A base, sometimes called a nucleobase, is generally adenine, cytosine, guanine, thymine or uracil, or a derivative thereof. A nucleobase is defined as a moiety that can bond to another nucleobase through H-bonds, polarized bonds (such as through OF moieties) or aromatic electronic interactions. Cytosine, thymine, and uracil are pyrimidine bases, and are generally linked to the scaffold through their 1 -nitrogen. Adenine and guanine are purine bases and are generally linked to the scaffold through their 9-nitrogen. The terms ‘adenine’, ‘guanine’, ‘cytosine’, ‘thymine’, ‘uracil’ and ‘hypoxanthine’ as used herein refer to the nucleobases as such. The terms ‘adenosine’, ‘guanosine’, ‘cytidine’, ‘thymidine’, ‘uridine’ and ‘inosine’ refer to the nucleobases linked to the (deoxy)ribosyl sugar. The nucleobases in an AON as disclosed herein can be adenine, cytosine, guanine, thymine, or uracil or any other moiety able to interact with another nucleobase through H-bonds, polarized bonds (such as CF) or aromatic electronic interactions. The nucleobases at any position in the AON as disclosed herein can be a modified form of adenine, cytosine, guanine, or uracil, such as hypoxanthine (the nucleobase in inosine), pseudouracil, pseudocytosine, isouracil, N3-glycosylated uracil, 1 -methylpseudouracil, orotic acid, agmatidine, lysidine, 2-thiouracil, 2-thiothymine, 5-substituted pyrimidine (e.g., 5-halouracil, 5-halomethyluracil, 5-trifluoromethyluracil, 5-propynyluracil, 5-propynylcytosine, 5- aminomethyluracil, 5-hydroxymethyluracil, 5-formyl uracil, 5-aminomethylcytosine, 5- formylcytosine), 5-hydroxymethylcytosine, 7-deazaguanine, 7-deazaadenine, 7-deaza-2,6- diaminopurine, 8-aza-7-deazaguanine, 8-aza-7-deazaadenine, 8-aza-7-deaza-2,6- diaminopurine, 8-oxo-adenine, 3-deazapurine (such as a 3-deaza-adenosine), pseudoisocytosine, N4-ethylcytosine, N2-cyclopentylguanine, N2-cyclopentyl-2-aminopurine, N2-propyl-2-aminopurine, 2,6-diaminopurine, 2-aminopurine, G-clamp and its derivatives, Super A, Super T, Super G, amino-modified nucleobases or derivatives thereof; and degenerate or universal bases, like 2,6-difluorotoluene, or absent like abasic sites (e.g. 1 -deoxyribose, 1 ,2- dideoxyribose, 1-deoxy-2-O-methylribose, azaribose). Modified bases comprise synthetic and natural bases such as inosine, xanthine, hypoxanthine and other -aza, deaza, -hydroxy, -halo, - thio, thiol, -alkyl, -alkenyl, -alkynyl, thioalkyl derivatives of pyrimidine and purine bases that are or will be known in the art. Purine nucleobases and / or pyrimidine nucleobases may be modified to alter their properties, for example by amination or deamination of the heterocyclic rings. The exact chemistries and formats may vary from oligonucleotide construct to oligonucleotide construct and from application to application, and may be worked out in accordance with the wishes and preferences of those of skill in the art.

[0073] A scaffold modification indicates the presence of a modified version of the ribosyl moiety as naturally occurring in RNA (i.e., the pentose moiety), such as bicyclic sugars, tetrahydropyrans, hexoses, morpholinos, 2’-modified sugars, 4’-modified sugar, 5’-modified sugars and 4’- substituted sugars. Examples of suitable modifications include, but are not limited to 2’-O- modified RNA monomers, such as 2’-O-alkyl or 2’-O-(substituted)alkyl such as 2’-0Me, 2’-O-(2- cyanoethyl), 2’-M0E, 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)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-alkoxycarbonyl such as 2’-O-[2-(methoxycarbonyl)ethyl] (MOCE), 2’-O-[2- / V-methylcarbamoyl)ethyl] (MCE), 2’- O-[2-( / V, / V-dimethylcarbamoyl)ethyl] (DCME); 2’-halo e.g. 2’-F, FANA; 2'-O-[2-(methylamino)-2- oxoethyl] (NMA); a bicyclic or bridged nucleic acid (BNA) scaffold modification such as a conformationally restricted nucleotide (CRN) monomer, a locked nucleic acid (LNA) monomer, a xy / o-LNA monomer, an a-LNA monomer, an a-l-LNA monomer, a p-d-LNA monomer, a 2’-amino- LNA monomer, a 2’-(alkylamino)-LNA monomer, a 2’-(acylamino)-LNA monomer, a 2’-N- substituted 2’-amino-LNA monomer, a 2’-thio-LNA monomer, a (2’-O,4’-C) constrained ethyl (cEt) BNA monomer, a (2’-O,4’-C) constrained methoxyethyl (cMOE) BNA monomer, a 2’,4’- BNANC(NH) monomer, a 2’,4’-BNANC(NMe) monomer, a 2’,4’-BNANC(NBn) monomer, an ethylene- bridged nucleic acid (ENA) monomer, a carba-LNA (cLNA) monomer, a 3,4-dihydro-2 / 7-pyran nucleic acid (DpNA) monomer, a 2’-C-bridged bicyclic nucleotide (CBBN) monomer, an oxo- CBBN monomer, a heterocyclic-bridged BNA monomer (such as triazolyl or tetrazolyl-linked), an amido-bridged BNA monomer (such as AmNA), an urea-bridged BNA monomer, a sulfonamide- bridged BNA monomer, a bicyclic carbocyclic nucleotide monomer, a TriNA monomer, an a-l- TriNA monomer, a bicyclo DNA (bcDNA) monomer, an F-bcDNA monomer, a tricyclo DNA (tcDNA) monomer, an F-tcDNA monomer, an alpha anomeric bicyclo DNA (abcDNA) monomer, an oxetane nucleotide monomer, a locked PMO monomer derived from 2’-amino LNA, a guanidine-bridged nucleic acid (GuNA) monomer, a spirocyclopropylene-bridged nucleic acid (scpBNA) monomer, and derivatives thereof; cyclohexenyl nucleic acid (CeNA) monomer, altriol nucleic acid (ANA) monomer, hexitol nucleic acid (HNA) monomer, fluorinated HNA (F-HNA) monomer, pyranosyl-RNA (p-RNA) monomer, 3’-deoxypyranosyl DNA (p-DNA), unlocked nucleic acid UNA); an inverted version of any of the monomers above. All these modifications are known to the person skilled in the art.

[0074] The orphan nucleotide Mutagenesis studies of human ADAR2 revealed that a single mutation at residue 488 from glutamate to glutamine (E488Q), gave an increase in the rate constant of deamination by 60-fold when compared to the wild-type enzyme (Kuttan and Bass. Proc Natl Acad Sci USA 2012. 109(48): 3295-3304). During the deamination reaction, ADAR flips the edited base out of its RNA duplex, and into the enzyme active site (Matthews et al. 2016). When ADAR2 edits adenosines in the preferred context (an A:C mismatch) the nucleotide opposite the target adenosine is often referred to as the ‘orphan nucleotide’ (or ‘orphan cytidine’ as the case may be), as indicated above. The crystal structure of ADAR2 E488Q bound to double stranded RNA (dsRNA) revealed that the glutamine (Gin; Q) side chain at position 488 can donate an H-bond to the N3 position of the orphan cytidine, which leads to the increased catalytic rate of ADAR2 E488Q. In the wild-type enzyme, wherein a glutamate (or glutamic acid; Glu; E) is present at position 488 instead of a glutamine (Gin) the amide group of the glutamine is absent and is instead a carboxylic acid. To obtain the same contact of the orphan cytidine with the E488Q mutant would then, for the wildtype situation, require protonation for this contact to occur. To make use of endogenously expressed ADAR2 to correct disease relevant mutations, it is essential to maximize the editing efficiency of the wild type ADAR2 enzyme present in the cell. WO2020 / 252376 discloses the use of AONs with modified RNA bases, especially at the position of the orphan cytidine to mimic the hydrogen-bonding pattern observed by the E488Q ADAR2 mutant. By replacing the nucleotide opposite the target adenosine in the AON with cytidine analogs that serve as H-bond donors at N3, it was envisioned that it would be possible to stabilize the same contact that is believed to provide the increase in catalytic rate for the mutant enzyme. Two cytidine analogs were of particular interest: pseudoisocytidine (also referred to as ‘piC’; Lu et al. J Org Chem 2009. 74(21):8021-8030; Burchenal et al. (1976) Cancer Res 36:1520-1523) and Benner’s base Z (also referred to as ‘dZ’ or ‘Zd’ when the ribose comprises a 2’-H group (DNA); Yang et al. Nucl Acid Res 2006. 34(21):6095-6101) that were initially selected because they offer hydrogen-bond donation at N3 with minimal perturbation to the shape of the nucleobase. Benner’s base is also referred to with its chemical name 6-amino-5-nitro-3-yl-2(1 H)-pyridone. The presence of the cytidine analog in the AON may exist in addition to modifications to the ribose 2’ group. The ribose 2’ groups in the orphan nucleotide can be independently selected from 2’-H (i.e., DNA), 2’-OH (i.e., RNA), 2’-OMe, 2’-MOE, 2’-F, or 2’-4’-linked (i.e., a bridged nucleic acid such as a locked nucleic acid (LNA)), or other 2’ substitutions. The 2’-4’ linkage can be selected from linkers known in the art, such as a methylene linker or constrained ethyl linker.

[0075] The orphan nucleotide in the AON as disclosed herein is preferably a cytidine or analog thereof (such as a nucleotide carrying a Benner’s base) or a uridine or analog thereof (such as iso-uridine). The orphan nucleotide, whether it is a cytidine or analog thereof, or a uridine or analog thereof, preferably comprises a deoxyribose (2’-H; = DNA) but may also comprise a diF modification at the 2’ position of the sugar. In one aspect at least one and in another aspect both the neighbouring (directly adjacent) nucleotides flanking the orphan nucleotide do not comprise a 2’-0Me modification. Complete modification wherein all nucleotides of the oligonucleotide hold a 2’-0Me modification (including the orphan nucleotide), with natural bases, results in a nonfunctional oligonucleotide as far as RNA editing goes (known in the art), presumably because it hinders the ADAR activity at the targeted position. In general, an adenosine in a target RNA can be protected from editing by providing an opposing nucleotide with a 2'-0Me group (at least when there are no other chemical substitutions or modifications within the nucleotide), or by providing a guanine or adenine as opposing base, as these two nucleobases are also able to reduce editing of the opposing adenosine.

[0076] Linkage modifications

[0077] A nucleoside is generally connected to neighboring nucleosides through condensation of its 5’-phosphate moiety to the 3’-hydroxyl moiety of the neighboring nucleotide monomer. Similarly, its 3’-hydroxyl moiety is generally connected to the 5’-phosphate of a neighboring nucleotide monomer. This forms phosphodiester (PO) bonds. The PO’s and the scaffold form an alternating copolymer. The bases are grafted on this copolymer, namely to the scaffold moieties. Because of this characteristic, the alternating copolymer formed by linked scaffolds of an oligonucleotide is often called the ‘backbone’ of the oligonucleotide. Because PO bonds connect neighboring monomers together, they are often referred to as “backbone linkages”. It is understood that when a phosphate group is modified so that it is instead an analogous moiety such as a PS, such a moiety is still referred to as the backbone linkage of the monomer. This is referred to as a “backbone linkage modification”. In general terms, the backbone of an oligonucleotide comprises alternating scaffolds and backbone linkages.

[0078] As outlined in detail herein, naked AONs as disclosed herein comprise at least one, preferably multiple linkage modifications. It is generally more preferred that the AON as disclosed herein comprises linkage modifications at most, and potentially all positions if the AON is capable of mediating RNA editing through the deamination enzyme when the AON is bound to the target RNA nucleic acid molecule. A linkage modification can be, but is not limited to, a modified version of the PO present in RNA, such as PS, chirally pure PS, (R)-PS, (S)-PS, MP (also referred to as MeP), chirally pure MP, (R)-MP, (S)-MP, phosphoryl guanidine (such as PNdmi), chirally pure phosphoryl guanidine, (R)-phosphoryl guanidine, (S)-phosphoryl guanidine, phosphorodithioate (PS2), phosphonacetate (PACE), phosphonoacetamide (PACA), thiophosphonoacetate, thiophosphonoacetamide, methyl phosphorohioate, methyl thiophosphonate, PS prodrug, alkylated PS, H-phosphonate, ethyl phosphate, ethyl PS, boranophosphate, borano PS, metyl boranophosphate, methyl borano PS, methyl boranophosphonate, methyl boranophosphothioate, phosphate, phosphotriester, aminoalkylphosphotriester, and their derivatives. Another modification includes phosphoramidite, phosphoramidate, N3’->P5’ phosphoramidate, phosphorodiamidate, phosphorothiodiamidate, sulfamate, diethylenesulfoxide, amide, sulfonate, siloxane, sulfide, sulfone, formacetyl, alkenyl, methylenehydrazino, sulfonamide, triazole, oxalyl, carbamate, methyleneimino (MMI), and thioacetamide nucleic acid (TANA); and their derivatives. Various salts, mixed salts, deprotonated, protonated, tautomeric, and free acid forms are also included, as well as 3’->3’ and 2’->5’ linkages.

[0079] An AON as disclosed herein may also comprise one or more linkage modifications according to the structure of formula (I) wherein: X = O or S; and

[0080] R = an aryl, a substituted aryl, a heterocycle, a substituted heterocycle, an aromatic heterocycle, a substituted aromatic heterocycle, a Ci-Ce alkoxy, a substituted Ci-Ce alkoxy, a C1-C20 alkyl, a substituted C1-C20 alkyl, a Ci-Ce alkenyl, a Ci-Ce substituted alkenyl, a Ci-Ce alkynyl, a substituted Ci-Ce alkynyl, or a conjugate group. In a preferred embodiment, X = O and R = methyl and the linkage modification is referred to as “mesyl phosphoramidate” (abbreviated to MsPA or PNms).

[0081] In other preferred aspects, R equals one of the following structures (a), (b), (c), (d), (e), (f), (g), (h), or (i):

[0082]

[0083] Disclosed herein is also an AON that is able to mediate adenosine deamination by recruitment of a deaminating enzyme in a cell after the AON has formed a double-stranded complex with a region of a target RNA nucleic acid molecule in a cell, wherein the region comprises a target adenosine, wherein the deaminating enzyme can deaminate the target adenosine into an inosine, and wherein the AON comprises a moiety with a structure according to formula (II): wherein: X = O or S;

[0084] Y = O' or S'; and

[0085] R = an aryl, a substituted aryl, a heterocycle, a substituted heterocycle, an aromatic heterocycle, a substituted aromatic heterocycle, a Ci-Ce alkoxy, a substituted Ci-Ce alkoxy, a C1-C20 alkyl, a substituted C1-C20 alkyl, a Ci-Ce alkenyl, a Ci-Ce substituted alkenyl, a Ci-Ce alkynyl, a substituted Ci-Ce alkynyl, or a conjugate group. In a preferred embodiment, X = O and R = methyl. An AON as disclosed herein may comprise a substitution of one of the non-bridging oxygens in the PO linkage. This modification slightly destabilizes base pairing but adds significant resistance to nuclease degradation. A preferred nucleotide analogue or equivalent comprises PS, phosphonoacetate, phosphorodithioate, phosphotriester, aminoalkylphosphotriester, H- phosphonate, methyl and other alkyl phosphonate including 3'-alkylene phosphonate, 5'-alkylene phosphonate and chiral phosphonate, phosphinate, phosphoramidate including 3'-amino phosphoramidate and aminoalkylphosphoramidate, thionophosphoramidate, thionoalkylphosphonate, thionoalkylphosphotriester, selenophosphate or boranophosphate. Particularly preferred are internucleoside linkages that are modified to contain a PS. Particularly preferred are internucleoside linkages that are modified to contain a PNms. Particularly preferred are internucleoside linkages that are modified to contain a PNdmi. The regular internucleoside linkages between the nucleosides may be altered by mono- or di-thioation of the PO bonds to yield PS esters or phosphorodithioate esters, respectively. Other modifications of the internucleoside linkages are possible, including amidation and peptide linkers. The skilled person can determine for what target RNA nucleic acid molecule the AON comprises a certain linkage modification at each linkage position of the AON as disclosed herein to generate the most effective and most stable oligonucleotide compound.

[0086] Many of the non-naturally occurring modifications of the linkage, such as PS, are chiral. This means that there are Rp and Sp configurations, known to the person skilled in the art. In one embodiment, the chirality of the PS linkages is controlled, which means that each of the linkages is either in the Rp or in the Sp configuration, whichever is preferred. The choice of an Rp or Sp configuration at a specified linkage position may depend on the target sequence and the efficiency of binding and induction of causing RNA editing of the target adenosine. However, if such is not specifically desired, a composition may comprise AONs as active compounds with both Rp and Sp configurations at a certain specified linkage position. Mixtures of such AONs are also feasible, wherein certain positions preferably have either one of the configurations, while for other positions such does not matter. In one aspect, the AON as disclosed herein comprises one or more (chirally pure or chirally mixed) PS linkages. In one aspect, the AON as disclosed herein comprises one of more (chirally pure or chirally mixed) phosphoramidate (PN) linkages. In one aspect, the AON as disclosed herein comprises one or more (chirally pure or chirally mixed) PNms linkages. In one aspect, a PN linkage connects the terminal two nucleotides on each end of the AON. AONs as disclosed herein may also comprise linkage modifications at all positions that are not chirally controlled. The AON as disclosed herein may also comprise one or more naturally occurring internucleoside linkages. The choice and number of modified linkages may depend on the specific target, the sequence, the length, and the stability of the AON observed in a particular cell type of interest, which can be assessed by methods known to the person skilled in the art. In one aspect, at least one, at least two, at least three, or at least four internucleoside linkages between the 5’ and / or the 3’ terminal two, three, four, or five nucleosides respectively of the AON as disclosed herein are modified internucleoside linkages. In one aspect, the AON as disclosed herein comprises at least one MP internucleoside linkage according to the structure of formula (III):

[0087] As was noted in the art, a preferred position for an MP linkage in an AON is linkage position -2, thereby connecting the nucleoside at position -1 with the nucleoside at position -2. In one aspect, this position, in an AON as disclosed herein, comprises a linkage modification according to the structure of formula (I), preferably PNms, instead of an MP linkage. W02020 / 201406 discloses the use of MP linkage modifications at certain positions surrounding the orphan nucleotide in the first nucleic acid strand. Although the presence of MP linkages is compatible with RNA editing by human ADAR enzymes, introducing MP linkages during the manufacturing of oligonucleotides is challenging in view of additional manufacturing (purification) steps in the coupling and decoupling process. In one aspect, the AON does not comprise an MP linkage.

[0088] In one aspect, the AON as disclosed herein comprises at least one PNdmi linkage, preferably linking the most terminal two nucleosides at the 5’ and / or 3’ end of the AON. A PNdmi linkage as preferably used in the AONs as disclosed herein has the structure of formula (IV)

[0089] PNdmi linkage (|V)

[0090] Other internucleoside linkages that may be used in the AONs as disclosed herein are those that are disclosed in WO2023 / 278589. In one aspect, the AON as disclosed herein comprises at least one phosphonoacetate and / or at least one phosphonoacetamide internucleoside linkage.

[0091] Conjugate chemistries

[0092] In one aspect, the AON as disclosed herein, or the sense strand to which it may be annealed before entering a target cell (in an HEON as disclosed herein), is bound to a hydrophobic moiety, such as palmityl or an analog thereof, cholesterol or analog thereof, or tocopherol or analog thereof. It is preferably bound to the 5’ terminus. In case a hydrophobic moiety is bound to the 5’ terminus as well as to the 3’ terminus, such hydrophobic moieties may the same or different. The hydrophobic moiety bound to the oligonucleotide may be bound directly, or indirectly mediated by another substance. When the hydrophobic moiety is bound directly, it is sufficient if the moiety is bound via a covalent bond, an ionic bond, a hydrogen bond, or the like. When the hydrophobic moiety is bound indirectly, it may be bound via a linking group (a linker). The linker may be a cleavable or an uncleavable linker. A cleavable linker refers to a linker that can be cleaved under physiological conditions, for example, in a cell or an animal body (e.g., a human body). A cleavable linker is selectively cleaved by an endogenous enzyme such as a nuclease, or by physiological circumstances specific to parts of the body or cell, such as pH or reducing environment (such as glutathione concentrations). Examples of a cleavable linker comprise, but is not limited to, an amide, an ester, one or both esters of a phosphodiester, a phosphoester, a carbamate, and a disulfide bond, as well as a natural DNA linker. Cleavable linkers also include self-immolative linkers. An uncleavable linker refers to a linker that is not cleaved under physiological conditions, or very slowly compared to a cleavable linker, for example, in a PS linkage, modified or unmodified deoxyribonucleosides linked by a PS linkage, a spacer connected through a PS bond and a linker consisting of modified or unmodified ribonucleosides. There is no restriction on the chain length, when a linker is a nucleic acid such as DNA, or an oligonucleotide. However, it may be usually from 2 to 20 bases in length, from 3 to 10 bases in length, or from 4 to 6 bases in length. There is no restriction on the length or composition of a spacer that is connects the ligand and the oligonucleotide, and may include for example ethylene glycol, triethylene glycol (TEG), HEG, alkyl chains, propyl, 6-aminohexyl, or dodecyl. One or more other types of molecules may be bound to the AON through one or more linkers, including peptides, sugars, vitamins, polymers, aptamers, (fragments of) antibodies, small molecules, and the like.

[0093] General

[0094] In addition to the specific preferred chemical modifications at certain positions in compounds as disclosed herein, AONs as disclosed herein may comprise one or more (additional) modifications to the nucleobase, scaffold and / or backbone linkage, which may or may not be present in the same monomer, for instance at the 3’ and / or 5’ position. In one aspect, the AON as disclosed herein comprises at least one internucleoside linkage according to the structure of formula (I), (II), (III), and / or (IV), and / or the AON further comprises at least one nucleotide with a sugar moiety that comprises a 2’-OMe modification, and / or the AON comprises at least one nucleotide with a sugar moiety that comprises a 2’-MOE modification, and / or the AON comprises at least one nucleotide with a sugar moiety that comprises a 2’-F modification, and / or the AON comprises an orphan nucleotide that carries a 2’-H in the sugar moiety and is therefore referred to as a DNA nucleotide, even though additional modifications may exist in its base and / or linkage to its neighbouring nucleosides. In one aspect, the orphan nucleotide carries a 2’-F in the sugar moiety. In one aspect, the orphan nucleotide carries a diF substitution in the sugar moiety. In one aspect, the orphan nucleotide carries a 2’-F and a 2’-C-methyl in the sugar moiety. In one aspect, the orphan nucleotide comprises a 2’-F in the arabinose configuration (FANA) in the sugar moiety.

[0095] In one aspect, the AON is an antisense oligonucleotide that can form a double stranded nucleic acid complex with a target RNA molecule, wherein the double stranded nucleic acid complex can recruit an adenosine deaminating enzyme for deamination of a target adenosine in the target RNA molecule, wherein the nucleotide in the AON that is opposite the target adenosine is the orphan nucleotide, and wherein the orphan nucleotide has the structure of formula (V): wherein: X is O, NH, OCH2, CH2, Se, or S; B is a nitrogenous base selected from the group consisting of: cytosine, uracil, isouracil, N3-glycosylated uracil, pseudoisocytosine, 8-oxo- adenine, and 6-amino-5-nitro-3-yl-2(1 H)-pyridone; R1 and R2 are both selected, independently, from H, OH, F or CH3; R3 is the part of the AON that is 5’ of the orphan nucleotide, consisting of 7 to 30 nucleotides; and R4 is the part of the AON that is 3’ of the orphan nucleotide, consisting of 4 to 25 nucleotides. The nucleotide 3’ and / or 5’ from the orphan nucleotide may be DNA, more preferably the nucleotide at the 3’ (position -1).

[0096] Other chemical modifications of the AON as disclosed herein include the substitution of one or more than one of any of the hydrogen atoms with deuterium or tritium, examples of which can be found in e.g., WO2014 / 022566 or WO2015 / 011694. Again, in all cases, the modifications should be compatible with editing such that the AON fulfils its role as an oligonucleotide that can, after binding to its target sequence, recruit an adenosine deaminase enzyme because of the double-stranded nucleic acid entity that arises. In all aspects of the disclosure, the enzyme with adenosine deaminase activity is preferably ADAR1 , ADAR2, or ADAT.

[0097] AONs as disclosed herein preferably do not include a 5’-terminal O6-benzylguanosine or a 5’-terminal amino modification and preferably are not covalently linked to a SNAP-tag domain (an engineered O6-alkylguanosine-DNA-alkyl transferase). An AON as disclosed herein preferably does not comprise a boxB RNA hairpin sequence. In one aspect, an AON as disclosed herein comprises 0, 1 , 2 or 3 wobble base pairs with the target sequence, and / or 0, 1 , 2, 3, 4, 5, 6, 7, or 8 mismatching base pairs with the target RNA sequence. No mismatch exists when the orphan nucleotide is uridine, which may be defined differently when the orphan nucleotide is a uridine analog or derivative. One alternative for uridine is positioning an iso-uridine opposite the target adenosine, which likely does not pair like G pairs with II. Preferably, the target adenosine in the target sequence forms a mismatch base pair with the nucleoside in the AON that is directly opposite the target adenosine.

[0098] As outlined above, an AON as disclosed herein makes use of specific nucleotide modifications at predefined spots to ensure stability as well as proper ADAR binding and activity. These changes may vary and may include modifications in the backbone of the AON, in the sugar moiety of the nucleotides as well as in the nucleobases or the phosphodiester linkages, as outlined in detail herein. They may also be variably distributed throughout the sequence of the AON. Specific modifications may be needed to support interactions of different amino acid residues within the RNA-binding domains of ADAR enzymes, as well as those in the deaminase domain. For example, PS linkages between nucleotides or 2’-OMe or 2’-MOE modifications may be tolerated in some parts of the AON, while in other parts they should be avoided so as not to disrupt crucial interactions of the enzyme with the phosphate and 2’-OH groups. Specific nucleotide modifications may also be necessary to enhance the editing activity on substrate RNAs where the target sequence is not optimal for ADAR editing. Previous work has established that certain sequence contexts are more amenable to editing. For example, a target sequence 5’- UAG-3’ (with the target A in the middle) contains the most preferred nearest-neighbor nucleotides for ADAR2, whereas a 5’-CAA-3’ target sequence is disfavored (Schneider et al. 2014. Nucleic Acids Res 42(10):e87). The structural analysis of ADAR2 deaminase domain hints at the possibility of enhancing editing by careful selection of the nucleotides that are opposite to the target trinucleotide. For example, the 5’-CAA-3’ target sequence, paired to a 3’-GCU-5’ sequence on the opposing strand (with the A-C mismatch formed in the middle), is disfavored because the guanosine base sterically clashes with an amino acid side chain of ADAR2. The guanosine opposite the C in such circumstances (= at the -1 position within the AON) is preferably replaced by an inosine, more preferably a deoxyinosine.

[0099] The AON as disclosed herein, in contrast to what has been described for siRNA, or gapmers and their relation towards RNase breakdown and the use of such gapmers in doublestranded complexes (see for instance EP3954395A1), does not comprise a stretch of DNA nucleotides which would make a target sequence (or a sense nucleic acid strand) a target for RNase-mediated breakdown. It is not desired that the target transcript molecule is degraded through the binding of the AON to the transcript molecule. In one embodiment, the AON does not comprise four or more consecutive DNA nucleotides anywhere within its sequence. In an embodiment, the AON is composed of as much (chemically) modified nucleotides as possible to enhance the resistance towards RNase-mediated breakdown, while at the same time being as efficient as possible in producing an RNA editing effect. This means that the orphan nucleotide and several other nucleotides within the AON may be DNA, but also that there is no stretch of four or more consecutive DNA nucleotides within the AON. Hence, the AON as disclosed herein is not a gapmer. A gapmer reduces the expression of a target transcript but does not produce RNA editing of a specified adenosine within the target transcript. A gapmer is in principle a singlestranded nucleic acid consisting of a central region (DNA gap region with at least four consecutive deoxyribonucleotides) and wing regions positioned directly at the 5’ end (5’ wing region) and the 3’ end (3’ wing region) thereof. In contrast, the AON as disclosed herein may be any oligonucleotide that produces an RNA editing effect in which a target adenosine in a target RNA molecule is deaminated to an inosine, and accordingly is resistant to RNase-mediated breakdown as much as possible to yield this effect and to allow the mRNA transcript being translated into a protein.

[0100] The AONs as disclosed herein may also be administered in the context of aids that will increase the entry of the AON into the target cell and / or its endosomal escape as soon as it is in the cell. Moieties that can be applied for such applications are for example a set of chemical compounds (generally purified from nature) referred to as “saponins” or “triterpene glycosides”. A preferred saponin that can be used in the methods as disclosed herein is AG1856, disclosed in WO2021 / 122998 and further described for use with RNA editing producing oligonucleotides in WO2024 / 153801. In a preferred embodiment, the saponin is conjugated to the AON as disclosed herein, providing a 1 :1 ratio of saponin versus AON for optimal endosomal release of the AON, once it has reached the target cell in which it should act.

[0101] Disclosed herein is also a pharmaceutical composition comprising the AON as disclosed herein, and further comprising a pharmaceutically acceptable carrier, solvent, diluent, and / or other additive (such as a saponin or triterpene glycoside like AG1856 (as discussed above), which in fact may also be administered separately from the AON) and may be dissolved in a pharmaceutically acceptable organic solvent, or the like. Dosage forms in which the AON or the pharmaceutical composition are administered may depend on the disorder to be treated and the tissue that needs to be targeted and can be selected according to common procedures in the art. The pharmaceutical compositions may be administered by a single-dose administration or by multiple dose administration. It may be administered daily or at appropriate time intervals, which may be determined using common general knowledge in the field and may be adjusted based on the disorder, the severity thereof, and the efficacy of the active ingredient.

[0102] Although in a preferred embodiment, the AON as disclosed herein is a single-stranded oligonucleotide comprising an orphan nucleotide opposite the target adenosine, wherein the orphan nucleotide is chemically modified as disclosed herein, and wherein the remainder of the oligonucleotide is chemically modified to prevent it from nuclease breakdown also as disclosed herein, in another embodiment, disclosed is any kind of oligonucleotide or heteroduplex oligonucleotide complex, that may or may not be bound to hairpin structures (internally or at the terminal end(s)), that may be bound to ADAR or catalytic domains thereof, or wherein the oligonucleotide is in a circular format. In a preferred aspect, the AON as disclosed herein is a ‘naked’ oligonucleotide, comprising a variety of chemical modifications in the ribose sugar and / or the base of one or more of the nucleotides within the sequence, that preferably comprises at least one linkage according to the structure of formula (I) as disclosed herein, that can hybridize to the target transcript or a part thereof that includes the target adenosine, and can recruit endogenous (naturally present) ADAR in the target cell for the deamination of the target adenosine. In another aspect, the AON as disclosed herein, that is delivered in a ‘naked’ form, does not comprise a stem-loop structure for recruitment of the deaminating enzyme, which allows for a shorter AON and improved cellular delivery and trafficking.

[0103] Notably, when the AON comprises chemical modifications, as detailed herein, it may still be delivered through the means of a delivery vehicle. Suitable delivery vehicles are nanoparticle delivery vehicles such as polymeric nanoparticles, dendrimers, inorganic nanoparticles and nanocrystals, organic nanocrystals, and liposomes. Preferred nanoparticles are Lipid Nanoparticles (LN P’s) that are nano-sized lipid vesicles that carry the AON of the present invention and aid to the delivery of target cells. If an LNP is applied or any other similar type of carrier, the AON is still considered naked because it is not transcribed from an encoding polynucleotide (such as in the case of a plasmid or a vector, in which the AON is not regarded as ‘naked’). So, even though a chemically modified AON is encapsulated by a carrier, preferably an LNP, it is still seen as naked, as it has been manufactured as such in a laboratory setting and encapsulated thereafter in the carrier using methods known to the person skilled in the art. The disclosure also relates to a delivery vehicle, preferably an LNP, which comprises a ‘naked’ and chemically modified AON as disclosed herein, even more preferably as disclosed in any one of SEQ ID NO:1 to 50. The person skilled in the art understands that when a delivery moiety, or attachment to the AON is used (such a GalNAc moiety to target hepatocytes in the liver, and / or when attached / conjugated to a saponin as discussed above) that the AON is still seen as naked as well, also when a GalNAc-AON, or saponin-AON, or even a saponin-AON-GalNAc, or saponin- GalNAc-AON (or in other orders) is encapsulated in a delivery vehicle such as an LNP. In other words, a variety of non-limiting administration methods is feasible: i) a naked AON as is; ii) a naked AON encapsulated in a delivery vehicle, preferably an LNP; iii) a naked AON administered together or separately from (but not bound to) a saponin such as AG1856; iv) a naked AON conjugated to a saponin such as AG 1856; v) a naked AON conjugated to a saponin such as AG1856, and wherein the saponin-AON conjugate is encapsulated in a delivery vehicle, preferably an LNP; or vi) through an encoding vector, such as a plasmid or a viral vector from which the AON is transcribed. Depending on the disease target and the cells that need to be targeted an administration method is being selected, although such is preferably an administration in which the AON is in a naked form, either or not conjugated to a delivery moiety (or endosomal release agent), or either or not encapsulated in a delivery vehicle such as an LNP. It is known in the art that RNA editing entities (such as human ADAR enzymes) edit dsRNA structures with varying specificity, depending on several factors. One important factor is the degree of complementarity of the two strands making up the dsRNA sequence. Perfect complementarity of the two strands usually causes the catalytic domain of human ADAR to deaminate adenosines in a non-discriminative manner, reacting with any adenosine it encounters. The specificity of hADARI and 2 can be increased by introducing chemical modifications and / or ensuring several mismatches in the dsRNA, which presumably helps to position the dsRNA binding domains in a way that has not been clearly defined yet. Additionally, the deamination reaction itself can be enhanced by providing an oligonucleotide that comprises a mismatch opposite the adenosine to be edited. Following the instructions in the present application, those skilled in the art will be capable of designing the complementary portion of the oligonucleotide according to their needs.

[0104] It will be understood by a person having ordinary skill in the art that the extent to which the editing enzymes inside the cell are redirected to other target sites may be regulated by varying the affinity of the first nucleic acid strand for the recognition domain of the editing enzyme. The exact modification may be determined through some trial and error and / or through computational methods based on structural interactions between the AON and the recognition domain of the editing enzyme. In addition, or alternatively, the degree of recruiting and redirecting the editing enzyme resident in the cell may be regulated by the dosing and the dosing regimen of the AON. This is something to be determined by the experimenter in vitro) or the clinician, usually in phase I and / or II clinical trials.

[0105] Disclosed herein is the site-specific editing of target adenosines in RNA sequences in eukaryotic, preferably metazoan, more preferably mammalian, more preferably human cells, more preferably human cells from the central nervous system, more preferably brain cells, even more preferably cells involved in the existence and progression of poly-Q disorders, especially HD and SCA3. The target cell can be located in vitro, ex vivo or in vivo. One advantage of the AON as disclosed herein is that it can be used with cells in situ in a living organism, but it can also be used with cells in culture. In some embodiments cells are treated ex vivo and are then introduced into a living organism (e.g., re-introduced into an organism from whom they were originally derived). The AON as disclosed herein can also be used to edit target RNA sequences in cells from a transplant or within a so-called organoid, e.g., a brain tissue organoid. Organoids can be thought of as three-dimensional in v / tro-derived tissues but are driven using specific conditions to generate individual, isolated tissues.

[0106] Without wishing to be bound by theory, the RNA editing through human ADAR2 for example is thought to take place on primary transcripts in the nucleus, during transcription or splicing, or in the cytoplasm, where e.g., mature mRNA, miRNA or ncRNA can be edited. Generally spoken, RNA editing may be used to create RNA sequences with different properties. Such properties may be coding properties (creating proteins with different sequences or length, leading to altered protein properties or functions), or binding properties (causing inhibition or overexpression of the RNA itself or a target or binding partner; entire expression pathways may be altered by recoding miRNAs or their cognate sequences on target RNAs). Protein function or localization may be changed at will, by functional domains or recognition motifs, including but not limited to signal sequences, targeting or localization signals, recognition sites for proteolytic cleavage or co- or post-translational modification, catalytic sites of enzymes, binding sites for binding partners, signals for degradation or activation and so on. These and other forms of RNA and protein “engineering”, whether to prevent, delay or treat disease or for any other purpose, in medicine or biotechnology, as diagnostic, prophylactic, therapeutic, research tool or otherwise, are encompassed by the present disclosure.

[0107] The amount of AON to be administered, the dosage and the dosing regimen can vary from cell type to cell type, the disease to be treated, age, weight, gender, the target population, the mode of administration {e.g., systemic versus local), the severity of disease and the acceptable level of side activity, but these can and should be assessed by trial and error during in vitro research, in pre-clinical and clinical trials. The trials are particularly straightforward when the modified sequence leads to an easily detected phenotypic change, or a change in (the level of, or activity of) a specified biomarker. It is possible that higher doses of AONs could compete for binding to an ADAR enzyme within a cell, thereby depleting the amount of the enzyme, which is free to take part in RNA editing, but routine dosing trials will reveal any such effects for a given AON and a given target.

[0108] One suitable trial technique involves delivering the AON to cell lines, or a test organism and then taking biopsy samples at various time points thereafter. The sequence of the target RNA can be assessed in the biopsy sample and the proportion of cells having the modification can easily be followed. A method as disclosed herein can include a step of identifying the presence of the desired change in the cell’s target RNA sequence, thereby verifying that the target RNA sequence has been modified. This step will typically involve sequencing of the relevant part of the target RNA, or a cDNA copy thereof (or a cDNA copy of a splicing product thereof, in case the target RNA is a pre-mRNA), as discussed above, and the sequence change can thus be easily verified. Alternatively, as indicated above, the change may be assessed on the function of the protein before, during, and / or after treatment or assessing any other potential marker, which measurements are preferably performed in vitro on samples obtained from the treated subject.

[0109] After RNA editing has occurred in a cell, the modified RNA can become diluted over time, for example due to cell division, limited half-life of the edited RNAs, etc. Thus, in practical therapeutic terms a method as disclosed herein may involve repeated delivery of an AON until enough target RNAs have been modified to provide a tangible benefit to the patient and / or to maintain the benefits over time. AONs as disclosed herein are particularly suitable for therapeutic use, and so disclosed is also a pharmaceutical composition comprising an AON as disclosed herein and a pharmaceutically acceptable carrier, solvent, or diluent. In some embodiments the pharmaceutically acceptable carrier can simply be a saline solution. This can usefully be isotonic or hypotonic, particularly for pulmonary delivery. The AON as disclosed herein is suitably administrated in aqueous solution, e.g. saline, or in suspension, optionally comprising additives, excipients and other ingredients, compatible with pharmaceutical use, at concentrations ranging from 1 ng / ml to 1 g / ml, preferably from 10 ng / ml to 500 mg / ml, more preferably from 100 ng / ml to 100 mg / ml. Dosage may suitably range from between about 1 pg / kg to about 100 mg / kg, preferably from about 10 pg / kg to about 10 mg / kg, more preferably from about 100 pg / kg to about 1 mg / kg. As outlined above, the AONs of the present disclosure may also be delivered through a delivery vehicle such as an LNP. Amounts of LNP carrying AONs as disclosed herein can and will also be determined in (pre-) clinical phases. Administration may be by injection or infusion, intracranially, intrathecally, intranasally, orally, intravenously, subcutaneously, intradermally, intramuscularly, intra-tracheally, intra-peritoneally, intrarectally, intra-cisterna magna, parenterally, and the like. Administration may be in solid form, in the form of a powder, a pill, a gel, a solution, a slow-release formulation, or in any other form compatible with pharmaceutical use in humans.

[0110] In one embodiment, depending on the ultimate deamination effect of A to I conversion, the identification step of whether the editing has taken place, comprises the following steps: sequencing the target RNA; assessing the presence or absence of a non-, or less-functional protein; assessing whether splicing of the pre-mRNA was altered by the deamination; assessing the concentration and / or presence / absence of a biomarker; or using a functional read-out. A functional assessment will generally be according to methods known to the skilled person. A suitable manner to identify the presence of an inosine after deamination of the target adenosine is of course dPCR or even sequencing, using methods that are well-known to the person skilled in the art. However, the person skilled in the art of diseases such as poly-Q disorders will preferably apply tests to monitor the deposit of inclusion bodies in the different brain regions and whether such has increased, diminished, or stabilized during or after treatment with the AON as disclosed herein.

[0111] In one embodiment, a method as disclosed herein comprises the steps of administering to the subject an AON, a nanoparticle delivery vehicle formulation as disclosed herein, or a pharmaceutical composition as disclosed herein, allowing the formation of a double stranded nucleic acid complex of the AON with its specific complementary target nucleic acid molecule in a cell in the subject; allowing the engagement of an endogenous present adenosine deaminating enzyme, such as ADAR 1 or ADAR2; and allowing the enzyme to deaminate the target adenosine in the target nucleic target molecule to an inosine, thereby alleviating, treating, ameliorating, or slowing down progression of a poly-Q disorder, such as HD and SCA3.

[0112] RNA editing molecules present in the cell will usually be proteinaceous in nature, such as the ADAR enzymes found in metazoans, including mammals. The ones of most interest are the human ADARs, hADARI and hADAR2, including any isoforms thereof. RNA editing enzymes known in the art, for which oligonucleotide constructs as disclosed herein may conveniently be designed, include the adenosine deaminases acting on RNA (ADARs), such as hADARI and hADAR2 in humans or human cells and cytidine deaminases. It is known that hADARI exists in two isoforms; a long 150 kDa interferon inducible version and a shorter, 110 kDa version, that is produced through alternative splicing from a common pre-mRNA. Consequently, the level of the 150 kDa isoform available in the cell may be influenced by interferon, particularly interferongamma (IFN-y). hADARI is also inducible by TNF-a. This provides an opportunity to develop combination therapy, whereby IFN-y or TNF-a and AONs as disclosed herein are administered to a patient either as a combination product, or as separate products, either simultaneously or subsequently, in any order. Certain disease conditions may already coincide with increased IFN- y or TNF-a levels in certain tissues of a patient, creating further opportunities to make editing more specific for diseased tissues. It will be understood by a person having ordinary skill in the art that the extent to which the editing entities inside the cell are redirected to other target sites may be regulated by varying the affinity of the first nucleic acid strand for the recognition domain of the editing molecule.

[0113] An AON as disclosed herein can utilise endogenous cellular pathways and naturally available ADAR enzymes to specifically edit a target adenosine in the target RNA sequence. An AON as disclosed herein is capable of recruiting ADAR and complex with it and then facilitates the deamination of a (single) specific target adenosine nucleotide in a target RNA sequence to which it is bound. Ideally, only one adenosine is deaminated. An AON as disclosed herein, when complexed to ADAR, preferably brings about the deamination of a single target adenosine.

[0114] An AON as disclosed herein, especially when it is in a naked form, is normally equal to, or longer than 13 nucleotides, preferably longer than 14, 15, 16 nucleotides, still more preferably longer than than 17 nucleotides. In one aspect the AON as disclosed herein is longer than 20 nucleotides. The AON as disclosed herein is preferably shorter than 100 nucleotides, still more preferably shorter than 60 nucleotides, still more preferably shorter than 50 nucleotides. In a preferred aspect, the AON as disclosed herein comprises 18 to 70 nucleotides, more preferably comprises 18 to 60 nucleotides, and even more preferably comprises 18 to 50 nucleotides. Hence, in a particularly preferred aspect, the AON as disclosed herein comprises 15, 16, 17, 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, 50, 51 , 52, 53, 54, 55, 56, 57, 58, 59, or 60 nucleotides. In one embodiment, the AON is 27, 28, 29, or 30 nucleotides in length. EXAMPLE

[0115] RNA editing of the wildtype human PIAS1 transcript using a variety of AONs.

[0116] An initial set of 50 AONs was designed to target the adenosine in the AGC codon encoding serine at position 510 in the human PIAS1 protein, to allow an endogenous ADAR enzyme to deaminate the adenosine to an inosine, thereby rendering a change to glycine at this position (S510G). The design and chemical modifications of these AONs are provided in FIG. 1 , the sequences, and chemical modifications of these AONs are provided as SEQ ID NO:1 to 50, respectively. FIG. 1 also shows part of the (pre-) mRNA of human PIAS1, which is SEQ ID NO:51. Because these AONs were tested for their ability to cause RNA editing in hepatocytes (liver cells, which are not the preferred target cells in vivo), a GalNAc moiety was linked to the 5’ terminus. The skilled person understands that such GalNAc linkages are not required when CNS cells need to be targeted.

[0117] For the initial screen of the AONs, the following was performed. On day 0, Primary Human Hepatocytes (PHH’s; 5.0x104cells / well) were transfected in two different setups: i) with 100 nM AONs; and ii) with 200 nM AONs; each setup in triplicates, using Lipofectamine® RNAiMAX Reagent at the same time of seeding, following the protocol of the manufacturer. The plates containing cells, medium and AON were held at 37 °C, 5% CO2, and the medium was refreshed 24 hrs after transfection / plating. On day 3 (72 hrs post transfection / plating) the supernatants were discarded, and subsequent analysis was performed as follows. Cells were collected and used for RNA isolation using a RNeasy 96 Kit (Qiagen-74182) according to the manufacturer’s instructions. Extracted RNA was treated with DNase I (ThermoFisher-EN0521) according to manufacturer’s protocol. Samples were incubated at 37 °C for 30 min and then 1 pL 50 mM EDTA was added and further incubated at 60 °C for 2 min. The total RNAs were then reverse-transcribed using the Maxima Reverse Transcriptase (Thermo-EP0742) kit with oligo-dT primer, random Hexamer Primer, and dNTP Mix (10 mM each). A quantitative PCR was then performed with the Digital PCR System (Bio-Rad, QX200) in 22 pl aliquots of reaction mixtures containing cDNA, appropriate pairs of primers and ddPCR Supermix for Probes (no dllTP) (Bio-Rad-1863024). The primers given in Table 1 were used with a PCR program that was as follows: 10 min at 95°C; 40 cycles for 30 sec at 94°C and 60 sec at 61 °C, 10 min at 98 °C and a hold step at 4°C. Then the plate was placed into the droplet QX200 reader to measure the number of positive droplets. The editing percentage was calculated by pooling the three replicates for each transfection for all A and G counts and then scored according to the formula: score = SUM(G) / (SUM(A+G) * 100

[0118] The results of the 100 nM transfections are given in FIG. 2 and the results of the 200 nM transfections are given in FIG. 3. Since in some samples the number of target RNA molecules turned out to be relatively low, meaning that the number of edited transcripts could not validly be compared to the number of non-edited transcripts, only the editing percentages are shown from the AON transfections that could validly be calculated. This shows that RM107515 (SEQ ID NO:27), RM107516 (SEQ ID NO:28), RM107518 (SEQ ID NQ:30), RM107519 (SEQ ID NO:31), RM 107521 (SEQ ID NO:33), RM 107522 (SEQ ID NO:34), RM 107523 (SEQ ID NO:35), RM107524 (SEQ ID NO:36), RM107525 (SEQ ID NO:37), RM107526 (SEQ ID NO:38), and RM 107527 (SEQ ID NO:39) performed best, with RM 107524 reaching editing percentages above 20%.

[0119] A subsequent experiment is performed, following the same protocol as described above, in either PHHs or in other human (-derived) cells, wherein the target transcripts, edited and nonedited, are present such that all 50 AONs are validly assessed for editing of the wildtype human PIAS1 (pre-) mRNA at the 1528 position.

[0120] Table 1. Primers and probes for the ddPCR of S510G editing in human PIAS1 transcripts. The “+” symbol represents a Locked Nucleic Acid (LNA) at the 3’ side of the symbol.

Claims

CLAIMS1. An antisense oligonucleotide (AON) capable of forming a double-stranded complex with a region of an endogenous human wild-type PIAS1 transcript molecule in a cell, wherein the cell is a neuron, preferably a cell of the central nervous system (CNS), more preferably a brain cell, wherein the transcript molecule is a pre-mRNA or an mRNA molecule, wherein the region of the PIAS1 transcript molecule comprises a target adenosine, wherein the nucleotide in the AON that is directly opposite the target adenosine is the orphan nucleotide, wherein the nucleotide numbering is such that the orphan nucleotide is number 0 and nucleotides are further positively (+) incremented towards the 5’ end and negatively (-) incremented towards the 3’ end, wherein the double-stranded complex can recruit an endogenous ADAR enzyme to deaminate the target adenosine into an inosine, thereby editing the PIAS1 transcript molecule, and wherein the deamination of the target adenosine results in an encoded PIAS1 protein with a decreased ability or loss-of-function to SUMOylate a target protein.

2. An AON according to claim 1 , wherein the deamination of the target adenosine into an inosine results in an amino acid change from serine to glycine at position 510 of the encoded PIAS1 protein.

3. An AON according to claim 1 or 2, wherein the orphan nucleotide is a deoxynucleotide comprising a cytosine, a cytosine analogue, an uracil, or an iso-uracil.

4. An AON according to claim 3, wherein the cytosine analogue comprises a 6-amino-5-nitro-3- yl-2(1 H)-pyridone nucleobase.

5. An AON according to any one of claims 1 to 4, wherein the AON comprises one or more modifications in the linkage moiety, which is each independently selected from the group consisting of: phosphorothioate (PS), phosphonoacetate, phosphorodithioate, methylphosphonate (MP), sulfonylphosphoramidate, (1 ,3-dimethylimidazolidin-2-ylidene) phosphoramidate (PNdmi), and mesyl phosphoramidate (PNms).

6. An AON according to claim 5, wherein the internucleoside linkage numbering in the AON is such that linkage number 0 is the linkage 5’ from the orphan nucleotide, and the linkage positions in the oligonucleotide are positively (+) incremented towards the 5’ end and negatively (-) incremented towards the 3’ end, and wherein linkage position -2 is an MP linkage or a PNms linkage.

7. An AON according to claim 5 or 6, wherein the linkage between the most terminal two nucleotides on the 5’ and / or 3’ terminus of the AON is a PNdmi linkage or a PNms linkage.

8. An AON according to any one of claims 1 to 7, wherein the AON comprises one or more nucleotides comprising a mono- or di-substitution at the 2', 3' and / or 5' position of the ribose, each independently selected from the group consisting of: -OH; -F; substituted or unsubstituted, linear or branched lower (C1-C10) alkyl, alkenyl, alkynyl, alkaryl, allyl, or aralkyl, that may be 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-allyl; -O-alkyl-O-alkyl; -methoxy; -aminopropoxy; -methoxyethoxy; -dimethylamino oxyethoxy; and -dimethylaminoethoxyethoxy.

9. An AON according to any one of claims 1 to 8, wherein the AON is selected from the group consisting of SEQ ID NO:1 to 50, preferably from the group consisting of SEQ ID NO:36, 27, 28, 30, 31 , 33, 34, 35, 37, 38, and 39.

10. A vector, preferably a viral vector, more preferably an adeno-associated virus (AAV) vector, comprising a nucleic acid molecule encoding an AON according to claim 1 or 2, wherein the orphan nucleotide is a cytidine or a uridine.

11. A pharmaceutical composition comprising an AON according to any one of claims 1 to 9, or a vector according to claim 10, and a pharmaceutically acceptable carrier.

12. An AON according to any one of claims 1 to 9 for use in the treatment of a poly-Q disorder, preferably wherein the poly-Q disorder is Huntington’s disease (HD) or Spinocerebellar Ataxia 3 (SC A3).

13. Use of an AON according to any one of claims 1 to 9 in the manufacture of a medicament for the treatment of a poly-Q disorder, preferably wherein the poly-Q disorder is HD or SCA3.

14. A method of treating a poly-Q disorder, preferably HD and / or SCA3, in an individual in need thereof, the method comprising contacting a PIAS1 polynucleotide in a cell of the subject with an AON according to anyone of claims 1 to 9, wherein the cell is a neuron, preferably a cell of the CNS, more preferably a brain cell, thereby treating the individual.

15. A method of deaminating a target adenosine in a PIAS1 pre-mRNA or mRNA molecule in a cell, the method comprising the steps of:(i) providing the cell with an AON according to any one of claims 1 to 9;(ii) allowing uptake by the cell of the AON;(iii) allowing annealing of the AON to the PIAS1 pre-mRNA or mRNA molecule;(iv) allowing an endogenous ADAR enzyme to deaminate the target adenosine in the target RNA molecule to an inosine; and optionally(v) identifying the presence of the inosine in the target RNA molecule.

16. The method according to claim 15, wherein the target adenosine is in the AGO codon encoding serine at position 510 of the human PIAS1 protein, and wherein the deamination of the target adenosine into an inosine results in a glycine at position 510.

17. A guide nucleic acid molecule for editing a target adenosine in a human PIAS1 pre-mRNA or mRNA molecule, wherein the pre-mRNA or mRNA molecule comprises the target region of SEQ ID NO:51 , and wherein the target adenosine is in the AGC codon encoding serine at position 510 of the human PIAS1 protein.

18. The nucleic acid molecule of claim 17, wherein the guide nucleic acid molecule is selected from the group consisting of SEQ ID NO:1 to 50, preferably from the group consisting of SEQ ID NO:36, 27, 28, 30, 31 , 33, 34, 35, 37, 38, and 39.

Images