Antisense oligonucleotides for the treatment of liver disease

Abstract

The present invention relates to antisense oligonucleotides (AONs) that can mediate RNA editing by binding to a target RNA nucleic acid molecule, preferably an RNA transcript molecule, in a cell and recruiting an endogenous deaminating enzyme in the cell to deaminate a target adenosine in the target RNA molecule to an inosine. The target RNA molecule is a transcript molecule form the SLC10A1 gene that encodes the Na+ / Taurocholate Co-transporting Polypeptide (NTCP). The RNA editing of the target adenosine will result in a loss- of-function of the NTCP protein, which will result in lowered uptake of bile acids from the portal circulation into the liver, thereby lowering the risk of suffering from disorders related to bile accumulation in the liver.

Description

ANTISENSE OLIGONUCLEOTIDES FOR THE TREATMENT OF LIVER DISEASETECHNICAL FIELD

[0001] This disclosure relates to the field of medicine, and in particular to the field of (chronic) liver diseases, like liver cirrhosis, for example caused by non-alcoholic fatty liver disease (NAFLD) and bile accumulation. The disclosure describes antisense oligonucleotides that mediate nucleotide-specific RNA editing in the human SLC10A1 gene transcript to bring about an amino acid change of the encoded Na+ / Taurocholate Co-transporting Polypeptide (NTCP) that influences its activity.BACKGROUND

[0002] Cholestatic disorders are caused by a buildup of bile acids in the liver due to bile duct dysfunction, which causes liver cell damage. The consequences of these disorders can be devastating and significantly impact a person's quality of life. Manifestations include pruritus, dry skin, fatigue, pain, weight loss, and many others. Without treatment, the damage progresses through various stages, from fibrosis to cirrhosis, ultimately leading to liver failure and an increased risk of liver cancer. Liver transplants are often necessary for primary sclerosing cholangitis (PSC) and biliary atresia (BA), two forms of cholestatic disease with high unmet medical needs.

[0003] PSC is a condition that causes inflammation and is typically diagnosed in people aged 30 to 40, more commonly affecting men (66%). It is estimated that approximately 80,000 people in North America and Europe suffer from PSC, with a prevalence of 1 to 9 individuals per 100,000. This condition causes fibrosis and sclerosis of bile ducts, leading to a toxic buildup of bile acids in the liver.

[0004] BA is a paediatric condition that affects newborns, resulting from the absence or defect of bile ducts. This condition causes harmful bile acids to accumulate in the liver, leading to rapid progression to cirrhosis early in life. It is estimated that approximately 20,000 individuals in North America and Europe suffer from BA, with a prevalence of 1 in 10,000 to 15,000 newborns in the western world.

[0005] Currently, there are no approved drugs for treating PSC or BA. For PSC, liver transplantation is the only treatment option with evidence to extend survival. However, PSC can return in 20 to 40% of patients who underwent liver transplantation, and the median survival without a transplant is only approximately 21 years. Surgery in the first weeks of life for BA is the standard treatment. However, most patients who receive this surgery will still require a liver transplant early in life.

[0006] Hepatocytes in the liver predominantly obtain bile acids from the enterohepatic reuptake cycle. The process of taking up bile acids from the portal circulation into hepatocytes is primarily carried out by a transporter protein named Na+ / Taurocholate Cotransporting Polypeptide (NTCP), encoded by the SLC10A 1 gene. The NTCP protein has received significant attention because it has been identified as the main protein involved in recognition and entry of Hepatitis B Virus (HBV) and Hepatitis D Virus (HDV). However, the normal function of NTCP at least relates to taking up conjugated bile acids from the circulation into hepatocytes. As indicated above, an accumulation of bile acids in the liver has been associated with a wide variety of liver diseases and it has been shown that inhibiting NTCP can improve liver function by reducing the intrahepatocellular levels of toxic bile acids, and preventing increased liver damage using certain available markers, such as fibrosis, cholangiocyte proliferation, Alkaline phosphatase (ALP), alanine transaminase (ALT), and inflammation biomarkers like cytokines (Slijepcevic D and Van de Graaf 2017. Dig Dis. 35(3) :251 -258; Slijepcevic D et al. 2018. Hepatology 68(3): 1057-1069).

[0007] Hepatocellular carcinoma (HCC) is the most common primary liver cancer, typically arising from chronic liver disease and cirrhosis. HCC is a global health concern with incidence rates of approximately 500,000 to 1 million new cases globally, and approximately 600,000 deaths per year. The rates tend to be higher in regions where Hepatitis B Virus (HBV) and Hepatitis C Virus (HCV) are endemic, although incidence is also rising in affluent regions due to the increased incidence of fatty liver disease that can develop into Metabolic Dysfunction-Associated Steatohepatitis (MASH), liver fibrosis and ultimately liver cancer. Liver cancer has a poor prognosis due to limited treatment options: surgical resection or liver transplants. Despite the existence of chemotherapies, the survival rates are about 35% at 1 year and 17% at 3 years.

[0008] HCC is refractory to Immune Checkpoint Blockade (ICB) therapies, such as anti-PD-1 treatment, and adoptive T-cell therapies. Combinations of anti-PD-1 and anti-CTLA4 treatment appear to improve overall survival rates, but all in all, there remains a medical need for improved treatment options.

[0009] Liver cancers often exhibit elevated levels of bile acids in the Tumour Microenvironment (TME). The importance of the TME and specific tumour-derived components on response rates of tumour treatment, and on T-cell infiltration, activation and induced senescence or exhaustion, leading to loss of treatment efficacy, is subject to intense research in recent years (Ma F et al. Sci Transl Med. 2025, 17(785):eadm7269).

[0010] Varanasi and colleagues (Varanasi SK et al. Science 2025, 387:192-201) looked closer into organ-specific metabolites to explain why HCC is non-responsive to ICB and adoptive T-cell transfer therapies and found conjugated Bile Acids (cBA) in the TME to beinvolved in T-cell exhaustion. The authors link this effect to cBA induced ER stress, hampering T-cell function. They found that silencing of the BA-conjugating enzyme Bile Acid-CoA:amino acid N-acyltransferase (BAAT) in hepatocellular carcinoma cells rescued T-cell function and induced an anti-PD-1 ICB response in several mouse tumour models.

[0011] Since T-cells do not possess canonical BA-transporters, such as NTCP, it is hypothesized herein that cBA is being transferred from HCC to T-cells by tumour-derived extracellular vesicles (tEV). Regardless of whether tEVs play a role in the process, based on the data presented by Varanasi et al. (2025), including the positive effect of the BAAT silencing in hepatic cells, it is predicted herein that lowering intrahepatic cBA levels is a viable strategy to improve T-cell function in liver cancer.

[0012] As enterohepatic reuptake accounts for about 95% of intrahepatic cBA content, de novo synthesis of BA taking care of the remaining 5%, it was reasoned that disrupting the enterohepatic BA re-uptake cycle to lower overall cBA in hepatocytes and hepatocellular carcinoma cells, is the preferred approach for liver cancer, and especially HCC. It is reasoned that even though increased bile acids may not be the direct cause of the liver cancer, the fact that cBA content is increased in liver cancers such as HCC, lowering such levels would be beneficial in a treatment.

[0013] Previously, International Patent Application Publication No. W02024 / 200472 disclosed methods for inducing loss-of-function mutations in NTCP by targeted RNA editing in hepatocytes in cell culture. Building on that, it has now been shown that this approach works in humanized liver murine models and in non-human primates (NHP). International Patent Application Publication No. W02024 / 200472 proposes this approach to treat cholestatic disease but is silent on the negative effect of elevated bile acids on T-cell function in liver cancer.

[0014] The present disclosure relates to a completely different approach of lowering the activity of NTCP, namely by using antisense oligonucleotide (AONs) and the cell’s own nucleic acid editing machinery in order to specifically target and amend a single nucleotide in the SLC10A1 transcript, thereby providing an NTCP protein with a loss-of-function, which in turn lowers the ability of hepatocytes to take up bile acids from the portal circulation and through this, treat disorders related to bile accumulation in the liver. The technology that the present disclosure relates to is generally referred to as ‘RNA editing’.

[0015] RNA editing is a natural process through which eukaryotic cells alter the sequence of their RNA molecules, often in a site-specific and precise way, thereby increasing the repertoire of genome encoded RNAs by several orders of magnitude. RNA editing enzymes have been described for eukaryotic species throughout the animal and plant kingdoms, and these processes play an important role in managing cellular homeostasis in metazoans fromthe simplest life forms (such as Caenorhabditis elegans) to humans. Examples of RNA editing are adenosine (A) to inosine (I) conversions and cytidine (C) to uridine (II) conversions, which occur through enzymes called Adenosine Deaminases acting on RNA (ADAR) and APOBEC / AID (cytidine deaminases that act on RNA), respectively.

[0016] ADAR is a multi-domain protein, comprising of a catalytic domain and two to three double-stranded RNA recognition domains, depending on the enzyme in question. Each recognition domain recognizes a specific double-stranded RNA (dsRNA) sequence and / or conformation. The catalytic domain does also play a role in recognizing and binding a part of the dsRNA helix, although the key function of the catalytic domain is to convert an A into an I in a nearby, predefined, position in the target RNA, by deamination of the nucleobase. Inosine is read as guanosine by the translational machinery of the cell, meaning that, if an edited adenosine is in a coding region of an mRNA or pre-mRNA, it can recode the protein sequence. A-to-l conversions may also occur in 5’ non-coding sequences of a target mRNA, creating new translational start sites upstream of the original start site, which gives rise to N-terminally extended proteins, or in the 3’ untranslated region (UTR) or other non-coding parts of the transcript, which may affect the processing and / or stability of the RNA. In addition, A-to-l conversions may take place in splice elements in introns or exons in pre- mRNAs, thereby altering the pattern of splicing. As a result, exons may be included or skipped. The enzymes catalysing adenosine deamination are within an enzyme family of ADARs, which include human deaminases hADARI and hADAR2, as well as hADAR3. However, for hADAR3 no deaminase activity has been demonstrated.

[0017] The use of oligonucleotides to edit a target RNA, applying adenosine deaminase, has been described (e.g., Woolf et al. Proc Natl Acad Sci USA 1995, 92:8298-8302; Montiel- Gonzalez et al. Proc Natl Acad Sci USA 2013, 110(45): 18285-18290; Vogel et al. Angewandte Chemie Int 2014, Ed 53:267-271).

[0018] A disadvantage of the method described by Montiel-Gonzalez et al. (2013) is the need for a fusion protein consisting of the boxB recognition domain of bacteriophage Lambda N- protein, fused to the adenosine deaminase domain of a truncated natural ADAR protein. It requires target cells to be either transduced with the fusion protein, which is a major hurdle, or that target cells are transfected with a nucleic acid construct encoding the engineered adenosine deaminase fusion protein for expression.

[0019] The system described by Vogel et al. (2014) suffers from similar drawbacks, in that it is not clear how to apply the system without having to genetically modify the ADAR first and subsequently transfect or transform the cells harboring the target RNA, to provide the cells with this genetically engineered protein. United States Patent No. US 9,650,627 describes a similar system.

[0020] The oligonucleotides of Woolf et al. (1995) were 100% complementary to the target RNA sequences and suffered from severe lack of specificity: nearly all adenosines in the target RNA strand that were complementary to the AON were edited, likely due to a lack of chemical modifications in the oligonucleotides.

[0021] It is known that ADAR may act on any dsRNA. Through a process sometimes referred to as ‘promiscuous editing’, the enzyme will edit multiple A’s in the dsRNA. Hence, there was a need for methods and means that circumvent such promiscuous editing and only target specific adenosines in a target RNA molecule to become therapeutic applicable. Vogel et al. (2014) showed that such off-target editing can be suppressed by using 2’-O-Methyl (2’- OMe) modified nucleosides in the oligonucleotide at positions opposite to adenosines that should not be edited and used a non-modified nucleoside directly opposite to the specifically targeted adenosine on the target RNA. However, the specific editing effect at the target nucleotide has not been shown to take place without the use of recombinant ADAR enzymes having covalent bonds with the AON. Several publications have now shown that the recruitment of endogenous ADAR (hence without the need for an exogenous and / or recombinant source) is feasible while maintaining a specificity in which a single adenosine within a target RNA molecule can be targeted and deaminated to an inosine.

[0022] International Patent Application Publication No. WO2016 / 097212 discloses AONs for the targeted editing of RNA, wherein the AONs are characterized by a sequence that is complementary to a target RNA sequence (therein referred to as the ‘targeting portion’) and by the presence of a stem-loop I hairpin structure (therein referred to as the ‘recruitment portion’), which is preferably non-complementary to the target RNA. Such oligonucleotides are referred to as ‘self-looping AONs’. The recruitment portion acts in recruiting a natural ADAR enzyme present in the cell (endogenously present) to the dsRNA formed by hybridization of the target sequence with the targeting portion. Due to the recruitment portion, there is no need for conjugated entities or presence of modified recombinant ADAR enzymes. International Patent Application Publication No. WO2016 / 097212 describes the recruitment portion as being a stem-loop structure mimicking either a natural substrate (e.g., the GluB receptor) or a Z-DNA structure known to be recognized by the dsRNA binding domains, or Z-DNA binding domains, of ADAR enzymes. A stem-loop structure can be an intermolecular stem-loop structure, formed by two separate nucleic acid strands, or an intramolecular stem loop structure, formed within a single nucleic acid strand. The stem-loop structure of the recruitment portion as described is an intramolecular stem-loop structure, formed within the AON itself, and are thought to attract (endogenous) ADAR. Similar stemloop structure-comprising systems for RNA editing have since then been described in United States Patent No. US11 ,390,865, and in International Patent Application Publication Nos.WO2017 / 050306, W02020 / 001793, WO2017 / 010556, W02020 / 246560, andWO2022 / 078995.

[0023] International Patent Application Publication Nos. WO2017 / 220751 and WO2018 / 041973 describe a next generation type of AONs that do not comprise such a stem-loop structure but that are (almost fully) complementary to the targeted area, and that appeared still capable of attracting endogenous ADAR enzymes. In one embodiment, one or more mismatching nucleotides, wobbles, or bulges exist between the oligonucleotide and the target sequence. A sole mismatch may be at the site of the nucleoside opposite the target adenosine, but in other embodiments AONs (or “RNA editing oligonucleotides” - even though the deamination reaction is carried out by the ADAR enzyme - and often abbreviated to ‘EONs’) were described with multiple bulges and / or wobbles when attached to the target sequence area.

[0024] It appeared possible to achieve in vitro, ex vivo and in vivo RNA editing with AONs lacking a stem-loop structure and with endogenous ADAR enzymes when the sequence of the AON was carefully selected such that it could attract / recruit ADAR. The ‘orphan nucleoside’, which is defined as the nucleoside in the AON that is positioned directly opposite the target adenosine in the target RNA molecule, was a nucleotide with an unmodified cytosine nucleobase and that did not carry a 2’-OMe modification. The orphan nucleoside can be a deoxyribonucleoside (DNA), wherein the remainder of the AON could still carry 2’-O-alkyl modifications at the sugar entity (such as 2’-OMe), or the nucleotides directly surrounding the orphan nucleoside contained chemical modifications (such as DNA in comparison to RNA) that further improved the RNA editing efficiency and / or increased the resistance against nucleases. Such effects could even be further improved by using sense oligonucleotides (SONs) that ‘protected’ the AONs against breakdown upon delivery to the cells (described in WO2018 / 134301 and US11 ,274,300).

[0025] The use of chemical modifications and particular structures in oligonucleotides that could be used in ADAR-mediated editing of specific adenosines in a target RNA have been the subject of numerous disclosures in the field, such as in International Patent Application Publication Nos. WO2019 / 111957, WO2019 / 158475, W02020 / 165077, W02020 / 201406, W02020 / 211780, WO2021 / 008447, WO2021 / 020550, WO2021 / 060527, WO2021 / 117729, WO2021 / 136408, WO2021 / 182474, WO2021 / 216853, WO2021 / 242778, WO2021 / 242870, WO2021 / 242889, W02022 / 007803, W02022 / 018207, WO2022 / 026928, andWO2022 / 124345. The use of specific sugar moieties has been disclosed in for instance International Patent Application Publication Nos. W02020 / 154342, W02020 / 154343, W02020 / 154344, WO2022 / 103839, and WO2022 / 103852, whereas the use of stereodefined linker moieties (in general for oligonucleotides that for instance can be used for exonskipping, in gapmers, in siRNA, or specifically for RNA-editing oligonucleotides, related to a wide variety of target sequences) has been described in International Patent Application Publication Nos. WO2011 / 005761 , W02014 / 010250, WO2014 / 012081 , WO2015 / 107425,WO2017 / 015575 (HTT), WO2017 / 062862, WO2017 / 160741 , WO2017 / 192664,WO2017 / 192679 (DMD), WO2017 / 198775, WO2017 / 210647, WO2018 / 067973,WO201 8 / 098264, WO2018 / 223056 (PNPLA3), WO2018 / 223073 (APOC3),WO2018 / 223081 (PNPLA3), WO2018 / 237194, W02019 / 032607 (C9orf72),WO20 19 / 055951 , WO2019 / 075357 (SMA / ALS), W02019 / 200185 (DM1), WO2019 / 217784(DM1), WO2019 / 219581 , W02020 / 118246 (DM1), W02020 / 160336 (HTT),WO2020 / 191252, W02020 / 196662, WO2020 / 219981 (USH2A), WO2020 / 219983 (RHO), WO2020 / 227691 (C9orf72), WO2021 / 071788 (C9orf72), WO2021 / 071858,WO2021 / 178237 (MAPT), WO2021 / 234459, WO2021 / 237223, WO2022 / 099159, WO202 1 / 030778, WO2022 / 174053, and WO2023 / 278589. Next to these disclosures, an extensive number of publications relate to the targeting of specific RNA target molecules, or specific adenosines within such RNA target molecules, be it to repair a mutation that resulted in a premature stop codon, or other mutation causing disease. Examples of such disclosures in which adenosines are targeted within specified target RNA molecules are International Patent Application Publication Nos. W02020 / 157008 and WO2021 / 136404 (USH2A); WO2021 / 113270 (APP); WO2021 / 113390 (CMT1A); W02021 / 209010 (IDUA, Hurler syndrome); WO2021 / 231673 and WO2021 / 242903 (LRRK2); WO2021 / 231675 (ASS1); WO2021 / 231679 (GJB2); WO2019 / 071274 and WO2021 / 231680 (MECP2);WO2021 / 231685 and WO2021 / 231692 (OTOF, autosomal recessive non-syndromic hearing loss); WO2021 / 231691 (XLRS); WO2021 / 231698 (argininosuccinate lyase deficiency); W02021 / 130313 and WO2021 / 231830 (ABCA4); and WO2021 / 243023 (SERPINA1).

[0026] The present invention aims to provide one or more alternative, and / or improved, compounds or compositions for use in the treatment of liver disease, such as cholestatic disorders caused by bile acid accumulation in the liver.SUMMARY OF THE INVENTION

[0027] Disclosed herein is an antisense oligonucleotide (AON) that is capable of recruiting an endogenous ADAR enzyme in a human 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 nucleotide in the AON that is opposite the target adenosine is the orphan nucleotide, wherein the ADAR enzyme can deaminate the target adenosine into an inosine after binding to the double-stranded complex, and wherein the target RNA nucleic acid molecule is a transcript molecule of the human SLC10A 1 geneencoding the Na7Taurocholate Co-transporting Polypeptide (NTCP). Preferably, the transcript molecule is a pre-mRNA or an mRNA molecule. Preferably, the cell is a liver cell, more preferably a hepatocyte. In one particularly preferred aspect, the target adenosine is the adenosine in the CAG codon coding for glutamine (Q) at position 68 of the NTCP protein, and the deamination of the adenosine changes the amino acid to an arginine (R). Preferably, the deamination of the target adenosine results in an NTCP protein that is impaired in its function to transport bile acids from portal circulation into the cell. In one aspect, the nucleotide numbering in the AON 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. In one aspect, 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. In one preferred aspect, the AON is (non)covalently bound to a Gal N Ac moiety.

[0028] The present disclosure relates to an AON that can bring about the deamination of the adenosine in the CAG codon for Q at position 68 of the NTCP protein, wherein the deamination results in a change of the amino acid to an R at position 68 in the NTCP protein, wherein the AON is according to structural formula (I), and wherein R is a hydrogen atom or an / V-Acetylgalactosamine (GalNAc) moiety that interacts with the asialoglycoprotein receptor on hepatocytes. The AON in its naked form of formula (I), wherein R = H, is herein referred to as RM120215. In a preferred aspect, R is a GalNAc moiety according to structural formula (II) or (III). In a more preferred aspect, R is the GalNAc moiety according to structural formula (III), wherein the resulting RM120215-GalNAc conjugate structure is herein referred to as RM122067. It is noted that the nucleotide sequence of RM120215 and RM122067, as well as all chemical modifications to the sugar moieties (the scaffolds), the nucleobases, and internucleoside linkages are identical for both AONs (SEQ ID NO:50), wherein in RM122067 the GalNAc moiety of formula (III) is linked to the 3’ terminus by a phosphodiester linkage. Hence, RM 122067 and RM 120215 both have the following nucleotide sequence and modifications:5’-Cm!Cm*Gm*Uf*Gf*Ae0Ge0Gf*Um*Ge*Cf!Af*Um*Gf!Am0m5Ue0Gm*Cm*Cf!Ae0m5Ue0Af*Cd*Zd*ldAUmeGf*m5Ce*Cm!Am-3’ (SEQ ID NO:50)

[0029] Disclosed herein is a pharmaceutical composition comprising an AON as disclosed herein, and a pharmaceutically acceptable carrier.

[0030] Disclosed herein is an AON for use in the treatment of a disease caused by bile accumulation in the liver, such as cholestasis, primary sclerosing cholangitis (PSC), biliary atresia (BA), and liver cirrhosis.

[0031] Disclosed herein is also an AON for use in the manufacture of a medicament for the treatment of a disease caused by bile accumulation in the liver, such as cholestasis, PSC, BA, and liver cirrhosis.

[0032] Disclosed herein is a method of editing a human SLC10A 1 polynucleotide in a cell, preferably a hepatocyte, wherein the human SLC10A 1 polynucleotide is a pre-mRNA or mRNA molecule, the method comprising contacting the SLC10A 1 polynucleotide with an AON as disclosed herein, thereby editing the SLC10A1 polynucleotide to encode an NTCP protein with a diminished, lowered, or loss of function in bile acid uptake.

[0033] Disclosed herein is a method of treating, ameliorating, or slowing down the progression of a disease caused by bile accumulation in the liver, such as cholestasis, PSC, BA, and liver cirrhosis, in a human subject in need thereof, the method comprising administering to said subject an AON as disclosed herein, or a pharmaceutical composition as disclosed herein, thereby contacting a SLC10A 1 polynucleotide in a cell of the subject with an AON capable of effecting an ADAR-mediated adenosine to inosine deamination, thereby editing the SLC10A 1 polynucleotide to encode an NTCP protein with a diminished, lowered, or loss of function in bile acid uptake, thereby treating the subject.

[0034] Disclosed herein is a method of treating a human subject suffering from liver cancer, the method comprising: i) administering to said subject an antisense oligonucleotide (AON); ii) allowing the AON to hybridize to a region of a pre-mRNA or mRNA molecule in a liver cancer cell of the subject, wherein the region comprises a target adenosine; iii) allowing the recruitment, by the double-stranded complex formed between the AON and the pre-mRNA or mRNA molecule, of an endogenous ADAR enzyme present in the cell; and iv) allowing the deamination of the target adenosine to an inosine by the ADAR enzyme; wherein the pre-mRNA or mRNA molecule is encoded by the human SLC10A 1 gene encoding NTCP, wherein the deamination of the target adenosine to an inosine results in an NTCP protein with a loss-of-function, wherein the AON can bring about the deamination of the adenosine in the CAG codon for Q at position 68 of the NTCP protein, wherein the deamination results in a change of the amino acid to an R at position 68 in the NTCP protein, wherein the AON is according to structural formula (I), and wherein R is a hydrogen atom or an N- Acetylgalactosamine (GalNAc) moiety, thereby treating the subject. In a preferred embodiment of the method, R is a GalNAc moiety according to structural formula (II) or (III), more preferably the GalNAc moiety of formula (III). In one aspect of the method, the livercancer is hepatocellular carcinoma (HCC). In another aspect of the method, the liver cancer is caused by bile accumulation in the liver.

[0035] Disclosed herein is also a method as disclosed herein, wherein the AON is administered to the human subject in a dose of 1 to 50 mg / kg, preferably 1 to 20 mg / kg, more preferably 1 to 10 mg / kg, even more preferably in a dose of 2, 3, 4, 5, 6, 7, 8, or 9 mg / kg.

[0036] In a preferred aspect, the human subject receives the AON dose in a repeated regimen, preferably once every two weeks, more preferably once a week.

[0037] Disclosed herein is also a kit-of-parts for use in the treatment of a disease caused by bile accumulation in the liver, such as cholestasis, PSC, BA, liver cirrhosis, and / or liver cancer such as HCC, said kit-of-parts comprising: a part (A) comprising an AON as disclosed herein; and a part (B) comprising an AG1856 saponin that is conjugated to a GalNAc moiety according to formula (III), wherein part (A) and part (B) can be mixed to be administered simultaneously, or wherein part (A) and part (B) are administered subsequently with part (A) being administered first, and part (B) being administered thereafter, preferably 24 hours later.BRIEF DESCRIPTION OF THE DRAWINGS

[0038] One or more embodiments of the invention will now be described, by way of example only, with reference to the accompanying drawings, in which:

[0039] Fig. 1 shows part (SEQ ID NO:1) of the 5’ to 3’ sequence of the human SLC10A1 mRNA transcript in which the CAG codon coding for glutamine (Q) at position 68 in the NTCP protein is in bold. The underlined adenosine is the target for RNA editing as disclosed herein, resulting in a codon coding for an arginine (R) residue (CIG / CGG) at this position after editing. Below the target sequence the 5’ to 3’ sequences are provided of a set of AONs that were designed earlier (see International Patent Application Publication No. WO2024 / 200472) to target the target adenosine in SEQ ID NO:1. The chemical modifications are as follows: Ae and Ge are 2’-O-Methoxyethyl (2’-MOE) modified adenosine and guanosine, respectively; m5Ue is a 2’-MOE modified 5-methyl Uridine (identical to Te, which denotes a 2’-MOE modified nucleotide carrying a thymine nucleobase); m5Ce is 2’- MOE modified 5-methyl Cytidine; Cm, Am, Um, and Gm are 2’-OMe modified cytidine, adenosine, uridine, and guanosine, respectively; Gf, Cf, Af, and Uf are 2’-Fluoro modified guanosine, cytidine, adenosine, and uridine, respectively; Zd (orphan nucleotide) is a deoxynucleotide (a DNA nucleotide) carrying a 6-amino-5-nitro-3-yl-2(1 H)-pyridone nucleobase (= Benner’s base), representing a deoxycytidine analog; Id is a DNA nucleotide carrying a hypoxanthine nucleobase (also referred to as deoxyinosine); Cd is deoxycytidine;“!” refers to a (1 ,3-dimethylimidazolidin-2-ylidene) phosphoramidate linkage (PNdmi linkage); “A” refers to a methylphosphonate linkage (MP linkage); “*” refers to a phosphorothioate linkage (PS linkage); “#” refers to a mesylphosphoramidate linkage (PNms linkage); and “e” refers to a phosphodiester linkage (PO linkage).

[0040] Fig. 2 shows the editing percentages in a c.203A>G editing (Q68R) experiment in primary human hepatocytes (PHHs) after transfection using lipofectamine, using twenty- three AONs as indicated (details in Fig. 1) provided from left to right, in comparison to RM 108839 (SEQ ID NO:32). Two unrelated controls (RM4266 and RM4777), a lipofectamine treatment only (MOCK Lipo), and a non-treated (NT) sample served as negative controls.

[0041] Fig. 3 shows the editing percentages in a c.203A>G editing (Q68R) experiment in PHHs, using the nine AONs as indicated (details in Fig. 1) after co-treatment with saponin AG1856 as the transfection (in this case endosomal release) reagent. An unrelated AON (cntrl-ON), a saponin only (AG1856), and a non-treated (NT) sample served as negative controls.

[0042] Fig. 4 shows AONs RM 107385 (left three lanes) and RM 108839 (right four lanes) on a non-denaturing gel after being kept in water at +40°C or -20°C for 6 days, or in PBS pH 7.4 at +5°C, +40°C, or -20°C for 6 days, as indicated.

[0043] Fig. 5 shows the 5’ to 3’ sequences of an additional set of AONs that were designed based on the sequence and chemical modifications of RM 108839 (SEQ ID NO:32). The respective RM names and the SEQ ID NO’s of the AONs are provided. Modifications are as provided in Fig. 1 , wherein (1) represents a 2’-O-Methyl modified a-basic nucleoside, and wherein (2) represents a 2’-Fluoro modified a-basic nucleoside.

[0044] Fig. 6A and Fig. 6B show a variety of AONs that were either kept on water or in PBS at room temperature, before loading on a non-denaturing gel.

[0045] Fig. 7 shows the editing percentage in a c.203A>G editing (Q68R) experiment in PHHs, using the AONs of Fig. 5 as indicated after co-treatment with saponin AG 1856 as the endosomal release agent. Two unrelated AONs (RM4266 and RM4777), a Mock treatment, and a non-treated (NT) sample served as negative controls.

[0046] Fig. 8 shows the editing percentage in a c.203A>G editing (Q68R) experiment in PHHs, using the AONs as indicated either with a RNAiMAX transfection (black bars) or with saponin AG1856 as the endosomal release agent (grey bars). The AONs comprising mismatches and AONs comprising a-basic nucleotides are indicated. An unrelated AON (RM4266), a Mock treatment, and a non-treated (NT) sample served as negative controls.

[0047] Fig. 9 shows the editing percentage in a c.203A>G editing (Q68R) experiment in PHHs, using the four AONs as indicated either with a Lipofectamine RNAiMAX transfectionwith the AONs in a variety of concentrations as follows: for each AON from left to right: 120 nM, 100 nM, 80 nM, 60 nM, 40 nM, 20 nM, 10 nM, and 5 nM. An unrelated AON (RM4266), a Mock treatment (RNAiMAX), and a non-treated (NT) sample served as negative controls.

[0048] Fig. 10 shows the chemical structure of saponin AG1856 as purified from Agrostemma githago L. seeds (Clochard J et al. Int J Pharm 2020. 589:119822) on top. R = either one of the two structures at the bottom left and right that represent the tri-antennary GalNAc structure of formula (III), with two isomeric forms of the linker. It is envisioned that the AG1856-GalNAc compound (formula (X)) is a mixture of structures comprising: i) AG1856 that is attached to the left structure, and ii) AG1856 that is attached to the right structure.

[0049] Fig. 11 shows a sequence alignment of a part of the human SLC10A 1 target sequence (5’ to 3’), comprising the target adenosine as outlined herein, in bold. Underlined is the sequence that is complementary to the sequence of RM 120215 and RM 122067 (except for the orphan position 0, the deoxyinosine at position -1 and the 2’-OMe modified nucleoside carrying an uracil nucleobase at position +15 in the AON, which is opposite a cytidine in the target sequence. Below the human (Homo sapiens) sequence the corresponding sequence of the monkey target sequence is provided (Macaca fascicularis) with the single difference in comparison to the human sequence indicated with an asterisk below the sequence. Also, the corresponding sequence of the mouse target sequence (Mus musculus) with five differences in comparison to the human sequence (in the underlined AON target sequence) indicated with asterisks below the sequence.

[0050] Fig. 12 shows an overview of a repeated-dose study using RM 122067 with or without the co- or subsequent administration of AG1856-GalNAc in PXB-MICE®, followed by an assessment of human SLC10A1 editing in the liver on the day of necropsy. The number of animals in each study group is given between brackets in the first column. Animals in Group 1 were dosed with PBS only, serving as a negative control. Animals in Group 2 were dosed with PBS mixed with 0.3 mg / kg AG1856-GalNAc and served as a negative control. Animals in Group 3 all received a mix of 20 mg / kg RM122067 + 0.3 or 1 mg / kg AG1856-GalNAc (as indicated) on days 1 , 15, 29, and 43 (except for PXB_305). Subsequently, PXB_301 , PXB_302, PXB_307, and PXB_308 again received their same combined dosing on days 57, 71 , and 85 in a mixture. In contrast, PXB_303, PXB_304, PXB_306, PXB_309, PXB_311 , and PXB_312 received the dose of RM122067 on days 57, 71 , and 85, and the respective dose of AG1856-GalNAc on days 58, 72, and 86 (24 hr after dosing of RM122067). The animals in Group 4 all received 6 mg / kg RM 122067 in a mixture with 0.3 mg / kg AG1856- GalNAc on the indicated days and editing was studied on different days thereafter. PXB_402 did not receive the last two doses. The animals in Group 5 all received 20 mg / kg RM 122067 in a mixture with 0.3 mg / kg AG1856-GalNAc on the indicated days and editing was studiedon different days thereafter. PXB_510 did not receive the last two doses. The animals in Group 6 all received 50 mg / kg RM 122067 only, on the indicated days. Here, four animals were dosed 4x, and eight animals were dosed 13x, on the days that are shown. The animals in Groups 7, 8, 9, and 10 all received a single dose with the treatment as indicated on day 1 and editing was studied on day 3 or 13, as indicated. Where applicable (Groups 7, 8, and 9), RM122067 and AG1856-GalNAc were administered as a mixture.DETAILED DESCRIPTION

[0051] An AON as disclosed herein can recruit deaminating enzymes, such as ADARI 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, since the deaminating enzymes are recruited to the double-stranded AON / target RNA molecule complex and subsequently deaminate the target adenosine into an inosine.

[0052] The oligonucleotides are herein abbreviated to “AONs”, but sometimes also referred to as ‘editing oligonucleotides’, or ‘EONs’, even though the RNA editing event is performed by the deamination enzyme and the action of the oligonucleotide only triggers the RNA editing to take place.

[0053] 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 (International Patent Application Publication No. 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 (International Patent Application Publication No. 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 (International Patent Application Publication No. W02020 / 165077). Whereas the properties of phosphonoacetate and UNA modifications were known as such, thecompatibility thereof with engagement of enzymes with nucleotide deamination activity and with the deamination reaction was not known.

[0054] Disclosed herein is an AON 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, generally speaking, may be encoded by a mutated gene, wherein the mutation is the cause of a disease and wherein the editing can reverse the mutation to give rise to a wildtype protein, or a protein with a wildtype function (for instance when the mutated amino acid is changed to an amino acid that does not cause the disease, or that provides an improved phenotype).

[0055] As disclosed in more detail herein, the target transcript molecule may also be encoded by a wildtype gene, such as in an aspect of the present disclosure, wherein the target nucleic acid molecule is a transcript from a wildtype human SLC10A1 gene as shown in the present disclosure, wherein the RNA editing makes that the encoded NTCP protein obtains a loss- of-function, but that improves the disease state of the treated subject.

[0056] Non-limiting examples of transcript molecules 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., International Patent Application Publication Nos. WO2016 / 097212, WO2017 / 220751 , WO2018 / 041973, and WO2021 / 243023), IDUA (for the treatment of Hurler syndrome; see e.g., International Patent Application Publication Nos. WO2017 / 220751 , WO2018 / 041973, and WO2021 / 209010), LRRK2 (for the treatment of Parkinson’s disease; International Patent Application Publication Nos. WO2016 / 097212, WO2017 / 220751 , WO2018 / 041973, WO2021 / 231673 and WO2021 / 242903), ABCA4 (for the treatment of Stargardt disease; see e.g., International Patent Application Publication Nos. W02021 / 130313 and WO2021 / 231830), USH2A (for the treatment of Usher syndrome; e.g., International Patent Application Publication Nos. W02020 / 157008, WO2020 / 219981 and WO2021 / 136404), APP (e.g., International Patent Application Publication No. WO2021 / 113270), CMT1A (e.g., International Patent Application Publication No. WO2021 / 113390), ASS1 (e.g., International Patent Application Publication No.WO2021 / 231675), GJB2 (e.g., International Patent Application Publication No.WO2021 / 231679), MECP2 (for the treatment of Rett syndrome; e.g., International Patent Application Publication Nos. WO2019 / 071274 and WO2021 / 231680), OTOE (for the treatment of autosomal recessive non-syndromic hearing loss; e.g., International Patent Application Publication Nos. WO2021 / 231685 and WO2021 / 231692), XLRS (e.g., International Patent Application Publication No. WO2021 / 231691), and PCSK9 (for the treatment of hypercholesterolemia; e.g., International Patent Application Publication No. WO2023 / 152371).

[0057] The present disclosure relates to AONs that mediate, cause, trigger, induce RNA editing, by recruiting endogenous (naturally present) ADAR enzymes in the host cell, preferably hepatocytes, of the adenosine in the CAG codon coding for glutamine (Q) at position 68 present in the transcript of the SLC10A 1 gene. A two-dimensional structure of the NTCP protein in the cell membrane of hepatocytes is known from the art (Ho RH et al. 2004; J Biol Chem. 279(8): 7213-7222). An AON as disclosed herein aims to reduce the reabsorption of bile acids in the liver by inhibiting NTCP function.

[0058] Several loss-of-function variants of NTCP have been identified in the art. Hence, these mutations naturally occur in a limited number of people, but do not cause any symptoms associated with cholestasis (Vaz et al. 2015; Hepatology. 61 (1):260-267; Schneider et al. 2022; Clin Res Hepatol Gastroenterol. 46(3): 101824), even though bile acid concentrations in the circulation remain relatively high.

[0059] Several amino acid substitutions in the sodium-binding pocket leading to loss-of- function mutants of the NTCP protein have been described in the art leading to the inability of bile acid transport without impairing the expression or localization of the protein (Huan Yan et al. 2014; J Virology. 88(6): 3273-3284). This finding suggests that treating subjects suffering from disorders related to accumulated bile with an AON as disclosed herein will result in the reduction of such accumulation of toxic bile acids in the liver. Moreover, it is anticipated that generating a loss-of-function variant of NTCP will also promote the elimination of bile acids from the body by increasing their excretion in the faeces and urine. This is a process referred to as sulfation of bile acids, which enhances bile acid solubility and reduces their absorption in the intestines.

[0060] The present disclosure relates to an AON that is aimed at deamination of the adenosine in the CAG codon for Q at position 68 in the SLC10A1 transcript, of which it is thought that a change to R can cause a loss-of-function of the NTCP protein and can be used to treat the bile acid accumulation disorders as disclosed herein. However, it is not excluded that two or more adenosines may be targeted for deamination in a single treatment. Without wishing to be bound by theory a synergistic or additive effect may be obtained by combining AONs for targeting a multitude of adenosines, and thereby a multitude of amino acids within a single NTCP protein, to increase the therapeutic effect.

[0061] Cholestasis can cause inflammation and lead to the development of liver fibrosis, liver cirrhosis, hepatocellular carcinoma, and / or liver failure. The AON as disclosed herein is designed to decrease toxic build-up of bile acids to reduce inflammation and lessen or prevent fibrosis and cirrhosis to protect liver cells and function.

[0062] Causes of cholestasis can be chronic or acute. Chronic liver diseases cause fibrosis and cirrhosis (scarring) of liver tissue that reduces liver function. Chronic conditions caninclude infectious diseases such as hepatitis virus (hepatitis B virus, hepatitis C virus, hepatitis D virus); alcohol-induced hepatitis; autoimmune and autoinflammatory conditions including primary sclerosing cholangitis, primary biliary cholangitis, and autoimmune hepatitis; metabolic disorders including non-alcoholic fatty liver disease, metabolic- associated fatty liver disease, and non-alcoholic steatohepatitis; inherited disorders including Wilson's disease, hereditary hemochromatosis, and a1 -antitrypsin (A1AT) disease. Acute hepatitis and cholestasis can result from infections including hepatitis viruses, mononucleosis, HIV, cytomegalovirus, sepsis, gallbladder infection; from alcohol poisoning or toxic hepatitis; from liver cancer or lymphoma; from use of medications (including birth control pills, anabolic steroids, penicillin antibiotics including amoxicillin, azathioprine, imipramine, estradiol, cimetidine, chlorpromazine, prochlorperazine, tolbutamide, terbinafine); and cholestasis of pregnancy.

[0063] Cholestasis can also result from causes outside of the liver from bile duct obstructions or stricture, including gallstones in the common bile duct, cystic duct, or Hartmann’s pouch; pancreatic cysts and pseudocysts; extrahepatic bile duct tumours; chronic pancreatitis; pancreatic cancer; bile duct cancer; cholangitis; biliary atresia; and from prior injury or surgery. Furthermore, cholestasis can also occur in neonates, and the causes can include infectious agents such as viruses, bacteria, spirochetes, parasites; toxins from drugs, endotoxins, total parenteral nutrition-associated cholestasis, or herbal products; metabolic causes including hypothyroidism or panhypopituitarism; immune-related gestational alloimmune liver disease; anatomic obstructions including biliary atresia, choledochal cyst, cholelithiasis, biliary sludge, inspissated bile, spontaneous perforation of common bile duct or tumour; cholestasis from idiopathic neonatal hepatitis (transient neonatal cholestasis), cardiovascular and circulatory disorders, hemophagocytic lymphohistiocytosis, malignancy, or congenital lupus; or genetic and metabolic etiologies including A1AT deficiency, Alagille syndrome, arthrogryposis-renal dysfunction-cholestasis syndrome, Caroli disease, congenital hepatic fibrosis, chromosomal trisomy 21 , Turner syndrome, citrin deficiency, cystic fibrosis, disorders of bile acid synthesis, disorders of bile acid conjugation, fatty acid oxidation defects, galactosaemia, glycogen storage disease type IV, hereditary fructose intolerance, mitochondrial respiratory chain disorders, neonatal ichthyosis-sclerosing cholangitis syndrome, neonatal sclerosing cholangitis, Niemann-Pick disease type C, peroxisomal disorders, progressive familial intrahepatic cholestasis, lipid storage diseases, tyrosinaemia, or urea cycle defects.

[0064] The AON as disclosed herein is designed to be used against all the disorders above, in all pharmaceutical preparations, uses and methods of treatment as disclosed herein, by preventing bile acid build up in the liver and the resulting cholestasis.

[0065] Definitions

[0066] 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 nucleobases adenine (A), guanine (G), cytosine (C), thymine (T); uracil (II), or hypoxantine (I). Even though the nucleosides in an AON may be referred to as ‘RNA’, the 2’ position of the sugar moiety (the scaffold) may be substituted (modified) by the presence of a variety of chemical moieties, such as 2’-OMe, 2’-MOE, 2’-F, or 2’,2’-difluoro modifications. Whenever reference is made to a ‘deoxyoligoribonucleotide’ it may comprise the bases adenine (A), guanine (G), cytosine (C), thymine (T), uracil (II) or hypoxanthine (I). Whenever there is a reference to a DNA nucleotide, the 2’ position of the sugar (scaffold) moiety is not occupied. It is noted that a 5-methyl uridine is the same as a ribonucleoside carrying a thymine nucleobase. A deoxyuridine (lid) is different from a thymidine (DNA nucleotide carrying a thymine nucleobase) because a deoxyuridine is a DNA nucleotide carrying an uracil nucleobase. When a sugar moiety is modified by comprising a 2’-O-Methoxyethyl group, the nucleotide may for instance be referred to as m5Ce when the nucleobase is a 5- methyl cytosine, or as m5Ue when the nucleobase is an uracil with a 5-methyl group. Similarly, m5Ud is just a DNA nucleotide carrying a thymine nucleobase (Td).

[0067] 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), dll (or Ud), or T (or m5Ud) 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. As disclosed herein, the AON may also comprise a-basic nucleotides, with different 2’ substitutions in the sugar moiety, such as 2’-OMe or 2’-F.

[0068] 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 linkercomprising a phosphodiester (PO), phosphonoacetate, phosphotriester, PS, phosphoro(di)thioate, MP (or MeP), methyl thiophosphonate, phosphoramidate linkages, PNdmi, and a linkage according to the structure of formula (IV) as described herein.

[0069] 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.

[0070] Sometimes the terms i) adenine I adenosine I deoxyadenosine; ii) guanine I guanosine I deoxyguanosine; iii) cytosine I cytidine I deoxycytidine; iv) uracil I uridine I deoxyuridine; v) thymine I 5-methyluridine I thymidine; and vi) hypoxanthine I inosine I deoxyinosine, are used interchangeably to refer to the corresponding nucleobase on the one hand, and the (deoxy)nucleoside or (deoxy)nucleotide on the other. The nucleobase thymine (T) is also known as 5-methyluracil (often abbreviated to m5U) and is a uracil (II) 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.

[0071] Whenever reference is made to nucleotides in the oligonucleotide, such as cytosine, 5-methyl-cytosine, 5-hydroxymethylcytosine, 5-formylcytosine, 5-acetylcytosine, 5- hydroxycytosine, and p-D-glucosyl-5-hydroxymethylcytosine are included.

[0072] Whenever reference is made to adenine, N6-methyladenine, 8-oxo-adenine, 2,6- diaminopurine, and 7-methyladenine are included.

[0073] Whenever reference is made to uracil, dihydrouracil, iso-uracil, N3-glycosylated uracil, pseudo-uracil, 5-methyl-uracil (thymine), N1-methylpseudouracil, 4-thiouracil, and 5- hydroxymethyluracil are included.

[0074] Whenever reference is made to guanine, 1-methylguanine, 7-methylguanosine, N2,N2-dimethylguanosine, N2,N2,7-trimethylguanosine, and N2,7-dimethylguanosine are included.

[0075] 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’-OMe), are included, as well as other modifications, including 2’-4’ bridged variants.

[0076] 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, phosphoramidatelinkages, phosphoryl guanidine, thiophosphoryl guanidine, sulfono phosphoramidate, PNdmi and the linkage structure according to formula (IV), further outlined in detail below.

[0077] 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%.

[0078] 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.

[0079] The terms ‘conducive to’, ‘triggering’, ‘inducing’, or ‘mediating’ 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, bring about RNA editing after binding to the target RNA molecule.

[0080] 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. 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. 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 still be defined as a mismatch.

[0081] This aspect is particularly important in relation to the present disclosure in which an AON with very good editing properties, RM 108839 (SEQ ID NO:32), comprises multiple G nucleotides, likely allowing the homodimerization of the AON, thereby potentially lowering its editing efficiency (for further discussion, see below). As outlined in detail herein, replacing one or more guanosine in the GGGG stretch in RM 108839 for a uridine, allows a lower abilityto (homo) dimerize and allows a more efficient editing. Clearly, the introduced II makes a mismatch with its opposite C in the target sequence.

[0082] 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.

[0083] 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.

[0084] 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.

[0085] 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 of the AON as disclosed herein is preferably a deoxynucleotide (DNA nucleotide) carrying a Benner’s base (referred to as Zd), as further outlined in detail below.

[0086] A ‘nucleotide analog’ refers to an analog of a nucleic acid nucleotide. The nucleotide analog is an analog of adenosine, guanosine, cytidine, 5-methyluridine, uridine, inosine, deoxyadenosine, deoxyguanosine, deoxycytidine, thymidine, deoxyuridine, or deoxyinosine.

[0087] 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.

[0088] The nucleotide ‘numbering’ in an AON as disclosed herein is such that the orphan nucleotide opposite the target adenosine, 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.

[0089] 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.

[0090] 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 to 90% sequence identity.

[0091] The term ‘splice mutation’ relates to a mutation in a gene that encodes fora 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 degrade rapidly and do not have any functional activity.

[0092] 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. The naked form of an AON is therefore different from a form in which the AON is encoded (and delivered) by a viral genome or within 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.

[0093] 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.

[0094] Embodiments

[0095] Disclosed herein is an AON that is capable of recruiting an endogenous ADAR enzyme in a human 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 nucleotide in the AON that is opposite the target adenosine is the orphan nucleotide, wherein the ADAR enzyme can deaminate the target adenosine into an inosine after binding to the double-stranded complex, and wherein the target RNA nucleic acid molecule is a transcript molecule of the human SLC10A 1 gene encoding NTCP. Preferably, the transcript molecule is a pre-mRNA or an mRNA molecule. Preferably, the cell is a liver cell, more preferably a hepatocyte.

[0096] In one aspect, the nucleotide numbering within the AON 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.

[0097] In one aspect, 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 a PNms linkage instead of a MP linkage.

[0098] In a preferred aspect, the disclosure relates to an AON capable of recruiting an endogenous ADAR enzyme in a human cell, preferably a liver cell, more preferably anhepatocyte, after the AON has formed a double-stranded complex with a region of a pre- mRNA or mRNA transcript molecule of the human SLC10A 1 gene encoding the Na+ / Taurocholate Co-transporting Polypeptide (NTCP), wherein the region comprises the CAG codon for glutamine (Q) at position 68 of the NTCP protein, and wherein the deamination of the adenosine in the CAG codon, by the recruited ADAR enzyme, results in a change of the amino acid to an arginine (R), and wherein the AON is according to structural formula (I):

[0099] wherein R is a hydrogen atom or an / V-Acetylgalactosamine (GalNAc) moiety that interacts with the asialoglycoprotein receptor on the cell membrane of hepatocytes. When R = H, the AON as disclosed herein is also referred to as RM120215 (SEQ ID NO:50).

[0100] In a preferred aspect, R is a GalNAc moiety according to formula (III):

[0101] When R = the GalNAc moiety of formula (III), the AON as disclosed herein is also referred to as RM 122067 (see SEQ ID NO:50 for its nucleotide sequence and chemical modifications).

[0102] In one aspect, the AON is as disclosed in formula (I), either with or without any of the GalNAc moieties of formula (II) or (III), but wherein one or more of the (1 ,3- dimethylimidazolidin-2-ylidene) phosphoramidate linkages (PNdmi linkages) are replaced by mesylphosphoramidate linkages (PNms linkages).

[0103] In one aspect, the linkage between the most terminal two nucleotides on the 5’ and / or 3’ terminus of the AON is a PNdmi linkage and / or a PNms linkage.

[0104] In one aspect, when R = GalNAc, the AON is covalently or non-covalently, directly or through a linker, bound to the GalNAc moiety. The skilled person can select the right linker and need for covalent or non-covalent binding of a GalNAc moiety, when the AON needs to be delivered to liver cells, especially hepatocytes.

[0105] In another aspect, the disclosure relates to an AON according to formula (I), wherein R is a GalNAc moiety, and the GalNAc moiety is according to structural formula (II):

[0106] In one aspect, the AON is covalently or non-covalently, directly or through a linker, bound to a triterpene glycoside, preferably AG 1856. As described in International Patent Application Publication No. WO2024 / 153801 , it is very efficient to use saponins in in vivo delivery systems to enhance endosomal escape and thereby to increase RNA editing.

[0107] In one aspect, the AON of the present disclosure is connected 1 :1 with a saponin, preferably AG1856. Hence, to increase the endosomal release (intracellularly) of the AON and make it available for RNA target hybridization, in one aspect, the AON is attached (covalently, or non-covalently) to AG 1856 before administration to the cell or the subject to be treated.

[0108] In another aspect, the AON is administered, for in vivo purposes, separately from the saponin. The saponin may be administered before or after the administration of the AON. In one aspect, the saponin is administered together with the AON, but wherein the saponin is not bound to the AON.

[0109] Disclosed herein is therefore also a kit-of-parts comprising at least two parts: i) part (A) comprising an AON as disclosed herein, preferably conjugated to a GalNAc moiety according to formula (II) or (III), preferably to the 3’ terminus of the AON; and ii) part (B) comprising a saponin, preferably AG1856, that is preferably also conjugated to a GalNAc moiety according formula (II) or (III). Part (A) and part (B) may come in a dry (lyophilized) format, or dissolved in a suitable carrier, such as PBS mixed with water-for-injection. The compound of part (A) may be administered before or after or at the same time as the compound of part (B). Further parts in the kit-of-parts as disclosed herein may comprise pharmaceutically acceptable carriers, solvents, and / or diluents that may be used to solubilize the compounds of part (A) and / or (B) to prepare for administration. Suitable solutions, that may act as carriers comprise PBS and water for injection.

[0110] In a preferred aspect, the kit-of-parts as disclosed herein comprises at least two parts, wherein part (A) comprises the AON herein referred to as RM 122067, which is the RM120215 AON (formula (I)) conjugated at the 3’ terminus to the GalNAc moiety of formula (III), and wherein part (B) comprises AG1856 conjugated to the GalNAc moiety of formula (III). Fig. 10 shows the structure of the AG1856 saponin (top structure) with the tri-antennary GalNAc moiety of formula (III) at the bottom left and right, with the two indicated isoforms of the linker part as shown. Hence, the AG1856-GalNAc conjugate comprises a mixture of two isomeric forms. Hence, part (B) in the kit-of-parts as disclosed herein comprises that mixture.

[0111] Without wishing to be bound by theory, in one aspect RM122067 is administered without the preceding or subsequent administration of the AG1856-GalNAc compound, which means that the AON on its own is sufficiently effective in inducing RNA editing of the target adenosine in the SLC10A 1 (pre-) mRNA in the liver of the treated human subject. The effect of RM 122067 with or without the administration of the endosomal escape enhancing agent AG1856-GalNAc is typically investigated during clinical trials with human subjects. Such clinical trials also provide information regarding dosing of the RM 122067 compound as well as the dosing of the AG1856-GalNAc compound, and whether repeated dosing of one or both parts within the kit-of-parts is required to achieve sufficient editing and lowering of bile acids in the circulation of the subjects that are treated. Such clinical trials also provide information regarding toxicity, pharmacodynamics, stability, etc. When administration of RM 122067 is sufficient to provide a healing effect in human subjects suffering from bile accumulation in the liver, and AG1856-GalNAc is not required to achieve that pharmaceutical beneficial effect, the RM 122067 compound (part (A)) may be administered without the administration of part (B). In one aspect, it is envisioned however that the AG1856-GalNAc conjugate - when non-toxic in human subjects when used at pharmaceutically effective levels - can stimulate the endosomal escape of the AON into the cytoplasm for increased efficient targeting of the SLC10A 1 (pre-) mRNA in the liver cells of human subjects, as outlined in detail herein.

[0112] Disclosed herein is an AON that is capable of recruiting an endogenous ADAR enzyme in a human 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 nucleotide in the AON that is opposite the target adenosine is the orphan nucleotide, wherein the ADAR enzyme can deaminate the target adenosine into an inosine after binding to the double-stranded complex, wherein the SLC10A1 gene is wildtype, and wherein the target adenosine is in the CAG codon coding for glutamine at position 68 of the NTCP protein, and wherein the deamination of the adenosine changes the amino acid to anarginine, and wherein the deamination of the target adenosine results in an NTCP protein that is impaired in its function to transport bile acids from portal circulation into the cell.

[0113] The present disclosure also relates to a first pharmaceutical composition comprising an AON as disclosed herein, and a pharmaceutically acceptable carrier. In one aspect, an AON as disclosed herein is in a naked form. 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) that the AON is still regarded as naked as well. In a preferred aspect, the first pharmaceutical composition comprises the RM 122067 conjugate, which is a RM120215 AON conjugated at its 3’ terminus to the tri-antennary GalNAc moiety of formula (III). The first pharmaceutical composition may come in a dry state or dissolved in a suitable carrier, solvent or diluent. Preferably, the RM 122067 compound is stored at -20°C in a dried state. The present disclosure also relates to a second pharmaceutical composition comprising AG1856 conjugated to a tri-antennary GalNAc moiety of formula (III), wherein the AG 1856 moiety acts as an endosomal escape enhancer to free RM 122067 from the endosomes in which it is entrapped upon binding to the liver cells (through the GalNAc moiety interaction with its asialoglycoprotein receptor on hepatocytes). The second pharmaceutical composition therefore is not active when it comes to triggering editing of the target adenosine in the SLC10A1 (pre-) mRNA, but it does enhance the editing effect brought about by the AON that recruits endogenous ADAR to the editing site. Without wishing to be bound by theory, it is envisioned that the AG1856 compound enhances the release of the AON from the endosomes. This effect is likely not sudden and instant but likely proceeds over time allowing a timed-release of AONs from the endosomes and a prolonged effect in respect of deamination of the target adenosine. Preferably, the dosing regimens of the first pharmaceutical composition and the second pharmaceutical composition are such that the editing effect is prolonged while being as beneficial as possible.

[0114] The present disclosure also relates to an AON as disclosed herein, for use in the treatment of a disease caused by bile accumulation in the liver, such as cholestasis, primary sclerosing cholangitis (PSC), biliary atresia (BA), and liver cirrhosis.

[0115] 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 disease caused by bile accumulation in the liver, such as cholestasis, PSC, BA, and liver cirrhosis.

[0116] The present disclosure also relates to a method of editing a human SLC10A 1 pre- mRNA or mRNA molecule in a liver cell, preferably a hepatocyte, the method comprising contacting the SLC10A1 pre-mRNA or mRNA molecule with an AON capable of triggering an ADAR-mediated adenosine to inosine deamination, thereby editing the SLC10A1 pre-mRNA or mRNA molecule to encode an NTCP protein with a diminished, lowered, or loss of function in bile acid uptake, and wherein the AON is as disclosed herein.

[0117] In a preferred aspect, the disclosure relates to an in vitro, ex vivo, or in vivo method of deaminating a target adenosine present in a region of a human SLC10A 1 pre-mRNA or mRNA transcript molecule, to an inosine, in a human cell, preferably a hepatocyte, wherein the target adenosine is in a CAG codon encoding glutamine (Q) at position 68 of the NTCP protein encoded by the SLC10A1 mRNA, the method comprising the step of: administering an AON to the cell, wherein the AON is sufficiently complementary to the region of the human SLC10A 1 pre-mRNA or mRNA to hybridize under physiological conditions to the pre-mRNA or mRNA, wherein the AON forms a double-stranded complex with the region of the SLC10A 1 pre-mRNA or mRNA transcript molecule comprising the target adenosine, wherein the double-stranded complex recruits an endogenous ADAR enzyme, wherein the endogenous ADAR enzyme deaminates the target adenosine in the CAG codon to an inosine, wherein the deamination results in a change of the encoded amino acid to an arginine (R), wherein the AON has a chemical structure according to formula (I), and wherein the AON is conjugated at its 3’ terminus to a tri-antennary GalNAc moiety according to formula (III) to form an AON-GalNAc conjugate compound.

[0118] In a preferred aspect, the disclosure relates to an in vitro, ex vivo, or in vivo method of deaminating a target adenosine present in a region of a human SLC10A 1 pre-mRNA or mRNA transcript molecule, to an inosine, in a human cell, preferably a hepatocyte, wherein the target adenosine is in a CAG codon encoding glutamine (Q) at position 68 of the NTCP protein encoded by the SLC10A1 mRNA, the method comprising the step of: administering an AON to the cell, wherein the AON is sufficiently complementary to the region of the human SLC10A 1 pre-mRNA or mRNA to hybridize under physiological conditions to the pre-mRNA or mRNA, wherein the AON forms a double-stranded complex with the region of the SLC10A 1 pre-mRNA or mRNA transcript molecule comprising the target adenosine, wherein the double-stranded complex recruits an endogenous ADAR enzyme, wherein the endogenous ADAR enzyme deaminates the target adenosine in the CAG codon to an inosine, wherein the deamination results in a change of the encoded amino acid to an arginine (R), wherein the AON has a chemical structure according to formula (I), wherein the AON is conjugated at its 3’ terminus to a tri-antennary GalNAc moiety according to formula (III) to form an AON-GalNAc conjugate compound, the method further comprising the step of simultaneously, before, or after administering the AON-GalNAc conjugate compound, administering to the cell an AG1856-GalNAc conjugate compound, wherein the AG1856-GalNAc conjugate compound has the chemical structure of formula (X), see Fig. 10.

[0119] In a preferred aspect, the disclosure relates to an in vitro, ex vivo, or in vivo method of deaminating a target adenosine present in a region of a human SLC10A 1 pre-mRNA or mRNA transcript molecule, to an inosine, in a human cell, preferably a hepatocyte, wherein the target adenosine is in a CAG codon encoding glutamine (Q) at position 68 of the NTCP protein encoded by the SLC10A1 mRNA, the method comprising the step of: administering an AON to the cell, wherein the AON is sufficiently complementary to the region of the human SLC10A 1 pre-mRNA or mRNA to hybridize under physiological conditions to the pre-mRNA or mRNA, wherein the AON forms a double-stranded complex with the region of the SLC10A 1 pre-mRNA or mRNA transcript molecule comprising the target adenosine, wherein the double-stranded complex recruits an endogenous ADAR enzyme, wherein the endogenous ADAR enzyme deaminates the target adenosine in the CAG codon to an inosine, wherein the deamination results in a change of the encoded amino acid to an arginine (R), wherein the AON has a chemical structure according to formula (I), wherein the AON is conjugated at its 3’ terminus to a tri-antennary GalNAc moiety according to formula (III) to form an AON-GalNAc conjugate compound, the method further comprising the step of administering to the cell an AG1856-GalNAc conjugate compound, wherein the AG1856-GalNAc conjugate compound has the chemical structure of formula (X), see Fig. 10, and wherein the AG1856-GalNAc conjugate compound is pre-mixed with the AON- GalNAc conjugate compound for simultaneous administration.

[0120] In a preferred aspect, the disclosure relates to an in vitro, ex vivo, or in vivo method of deaminating a target adenosine present in a region of a human SLC10A 1 pre-mRNA or mRNA transcript molecule, to an inosine, in a human cell, preferably a hepatocyte, wherein the target adenosine is in a CAG codon encoding glutamine (Q) at position 68 of the NTCP protein encoded by the SLC10A1 mRNA, the method comprising the step of: administering an AON to the cell, wherein the AON is sufficiently complementary to the region of the human SLC10A 1 pre-mRNA or mRNA to hybridize under physiological conditions to the pre-mRNA or mRNA, wherein the AON forms a double-stranded complex with the region of the SLC10A 1 pre-mRNA or mRNA transcript molecule comprising the target adenosine, wherein the double-stranded complex recruits an endogenous ADAR enzyme, wherein the endogenous ADAR enzyme deaminates the target adenosine in the CAG codon to an inosine, wherein the deamination results in a change of the encoded amino acid to an arginine (R), wherein the AON has a chemical structure according to formula (I), wherein the AON is conjugated at its 3’ terminus to a tri-antennary GalNAc moiety according to formula (III) to form an AON-GalNAc conjugate compound, the method further comprising the step of administering to the cell an AG1856-GalNAc conjugate compound, wherein the AG1856-GalNAc conjugate compound has the chemical structure of formula (X), see Fig. 10, and wherein the AG1856-GalNAc conjugate compound is administered after theadministration of the AON-GalNAc conjugate compound, preferably 1 , 2, 3, 4, 5, 6, 7, 8, 9, 10, 11 , 12, 13, or 14 days after administration of the AON-GalNAc conjugate compound. In a preferred regimen, the AON-GalNAc conjugate compound and / or AG1856-GalNAc conjugate compound administration is repeated in a timed interval.

[0121] In a preferred aspect, the present disclosure also relates to a method of treating, ameliorating, or slowing down the progression of a disease caused by bile accumulation in the liver, such as cholestasis, PSC, BA, and / or liver cirrhosis, in a human subject in need thereof, the method comprising the step of administering to said subject an AON that has a chemical structure according to formula (I), wherein the AON is conjugated at its 3’ terminus to a tri-antennary GalNAc moiety according to formula (III), wherein the AON is sufficiently complementary to a region of a human SLC10A 1 pre-mRNA or mRNA transcript molecule to hybridize under physiological conditions to the pre-mRNA or mRNA transcript molecule in a cell, preferably a liver cell, in the human subject, wherein the region comprises a target adenosine that is in a CAG codon encoding glutamine (Q) at position 68 of the NTCP protein encoded by the SLC10A 1 mRNA transcript molecule, wherein after administering the AON, the AON forms a double-stranded complex with the region of the SLC10A 1 pre-mRNA or mRNA transcript molecule in the cell, wherein the double-stranded complex recruits an endogenous ADAR enzyme naturally present in the cell, wherein the endogenous ADAR enzyme deaminates the target adenosine to an inosine, wherein the deamination of the target adenosine to an inosine results in a change of the encoded amino acid to an arginine (R) at position 68 of the NTCP protein, thereby treating, ameliorating, or slowing down the progression of the disease in the subject.

[0122] In a preferred aspect, the present disclosure also relates to a method of treating, ameliorating, or slowing down the progression of a disease caused by bile accumulation in the liver, such as cholestasis, PSC, BA, and / or liver cirrhosis, in a human subject in need thereof, the method comprising the step of administering to said subject an AON that has a chemical structure according to formula (I), wherein the AON is conjugated at its 3’ terminus to a tri-antennary GalNAc moiety according to formula (III), wherein the AON is sufficiently complementary to a region of a human SLC10A 1 pre-mRNA or mRNA transcript molecule to hybridize under physiological conditions to the pre-mRNA or mRNA transcript molecule in a cell, preferably a liver cell, in the human subject, wherein the region comprises a target adenosine that is in a CAG codon encoding glutamine (Q) at position 68 of the NTCP protein encoded by the SLC10A 1 mRNA transcript molecule, wherein after administering the AON, the AON forms a double-stranded complex with the region of the SLC10A 1 pre-mRNA or mRNA transcript molecule in the cell, wherein the double-stranded complex recruits an endogenous ADAR enzyme naturally present in the cell, wherein the endogenous ADAR enzyme deaminates the target adenosine to an inosine, wherein the deamination of thetarget adenosine to an inosine results in a change of the encoded amino acid to an arginine (R) at position 68 of the NTCP protein, the method further comprising the step of simultaneously, before, or after administering the AON-GalNAc conjugate compound, administering to the subject an AG1856-GalNAc conjugate compound, wherein the AG1856- GalNAc conjugate compound has the chemical structure of formula (X), see Fig. 10, thereby treating, ameliorating, or slowing down the progression of the disease in the subject.

[0123] In a preferred aspect, the present disclosure also relates to a method of treating, ameliorating, or slowing down the progression of a disease caused by bile accumulation in the liver, such as cholestasis, PSC, BA, and / or liver cirrhosis, in a human subject in need thereof, the method comprising the step of administering to said subject an AON that has a chemical structure according to formula (I), wherein the AON is conjugated at its 3’ terminus to a tri-antennary GalNAc moiety according to formula (III), wherein the AON is sufficiently complementary to a region of a human SLC10A 1 pre-mRNA or mRNA transcript molecule to hybridize under physiological conditions to the pre-mRNA or mRNA transcript molecule in a cell, preferably a liver cell, in the human subject, wherein the region comprises a target adenosine that is in a CAG codon encoding glutamine (Q) at position 68 of the NTCP protein encoded by the SLC10A 1 mRNA transcript molecule, wherein after administering the AON, the AON forms a double-stranded complex with the region of the SLC10A 1 pre-mRNA or mRNA transcript molecule in the cell, wherein the double-stranded complex recruits an endogenous ADAR enzyme naturally present in the cell, wherein the endogenous ADAR enzyme deaminates the target adenosine to an inosine, wherein the deamination of the target adenosine to an inosine results in a change of the encoded amino acid to an arginine (R) at position 68 of the NTCP protein, the method further comprising the step of administering to the subject an AG1856-GalNAc conjugate compound, wherein the AG1856- GalNAc conjugate compound has the chemical structure of formula (X), see Fig. 10, and wherein the AG1856-GalNAc conjugate compound is pre-mixed with the AON-GalNAc conjugate compound for simultaneous administration, thereby treating, ameliorating, or slowing down the progression of the disease in the subject.

[0124] In a preferred aspect, the present disclosure also relates to a method of treating, ameliorating, or slowing down the progression of a disease caused by bile accumulation in the liver, such as cholestasis, PSC, BA, and / or liver cirrhosis, in a human subject in need thereof, the method comprising the step of administering to said subject an AON that has a chemical structure according to formula (I), wherein the AON is conjugated at its 3’ terminus to a tri-antennary GalNAc moiety according to formula (III), wherein the AON is sufficiently complementary to a region of a human SLC10A 1 pre-mRNA or mRNA transcript molecule to hybridize under physiological conditions to the pre-mRNA or mRNA transcript molecule in a cell, preferably a liver cell, in the human subject, wherein the region comprises a targetadenosine that is in a CAG codon encoding glutamine (Q) at position 68 of the NTCP protein encoded by the SLC10A 1 mRNA transcript molecule, wherein after administering the AON, the AON forms a double-stranded complex with the region of the SLC10A 1 pre-mRNA or mRNA transcript molecule in the cell, wherein the double-stranded complex recruits an endogenous ADAR enzyme naturally present in the cell, wherein the endogenous ADAR enzyme deaminates the target adenosine to an inosine, wherein the deamination of the target adenosine to an inosine results in a change of the encoded amino acid to an arginine (R) at position 68 of the NTCP protein, the method further comprising the step of administering to the subject an AG1856-GalNAc conjugate compound, wherein the AG1856- GalNAc conjugate compound has the chemical structure of formula (X), see Fig. 10, and wherein the AG1856-GalNAc conjugate compound is administered after the administration of the AON-GalNAc conjugate compound, preferably 1 , 2, 3, 4, 5, 6, 7, 8, 9, 10, 11 , 12, 13, or 14 days after administration of the AON-GalNAc conjugate compound, thereby treating, ameliorating, or slowing down the progression of the disease in the subject. In a preferred clinical regimen, the AON-GalNAc conjugate compound and / or AG1856-GalNAc conjugate compound administration is repeated in a timed interval.

[0125] Also disclosed is an in vitro, ex vivo, or in vivo method for the deamination of a target adenosine in a human SLC10A 1 pre-mRNA or mRNA molecule in a cell, preferably a liver cell, more preferably a hepatocyte, 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 SLC10A 1 pre-mRNA or mRNA molecule; (iv) allowing an endogenous ADAR enzyme to deaminate the target adenosine in the SLC10A 1 pre-mRNA or mRNA molecule to an inosine; and optionally (v) using a functional read-out to identify the presence of the inosine in the SLC10A1 pre-mRNA or mRNA molecule.

[0126] In one aspect, a method as disclosed herein comprises the step of administering a triterpene glycoside before, after or simultaneously with administering the AON, wherein in a preferred aspect, the triterpene glycoside is AG1856. However, in one aspect the triterpene glycoside (or ‘saponin’ as it is often referred to) is physically bound to the AON. In another preferred aspect, the AON is conjugated to a GalNAc moiety to form an AON- GalNAc conjugate compound and the AG1856 is conjugated to a GalNAc moiety to form an AG1856-GalNAc conjugate compound. Then the AG1856 saponin and the AON are not physically connected and may be administered separately.

[0127] In one aspect, an AON as disclosed herein comprises a linkage moiety with the structure according to formula (IV)

[0128] wherein: X = O or S; and 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, MsPA or “PNms”. In one embodiment, a PNms linkage is used instead of the one or more, or all MP and / or PNdmi linkages.

[0129] The present disclosure also relates to a method of treating a human subject suffering from liver cancer, the method comprising: i) administering to said subject an antisense oligonucleotide (AON); ii) allowing the AON to hybridize to a region of a pre-mRNA or mRNA molecule in a liver cancer cell of the subject, wherein the region comprises a target adenosine; iii) allowing the recruitment, by the double-stranded complex formed between the AON and the pre-mRNA or mRNA molecule, of an endogenous ADAR enzyme present in the cell; and iv) allowing the deamination of the target adenosine to an inosine by the ADAR enzyme; wherein the pre-mRNA or mRNA molecule is encoded by the human SLC10A 1 gene encoding NTCP, wherein the deamination of the target adenosine to an inosine results in an NTCP protein with a loss-of-function, wherein the AON can bring about the deamination of the adenosine in the CAG codon for Q at position 68 of the NTCP protein, wherein the deamination results in a change of the amino acid to an R at position 68 in the NTCP protein, wherein the AON is according to structural formula (I), and wherein R is a hydrogen atom or a GalNAc moiety, thereby treating the subject. In a preferred embodiment of the method, R is a GalNAc moiety according to structural formula (II) or (III), more preferably formula (III). In a preferred embodiment of the method, the liver cancer is hepatocellular carcinoma (HCC). In another preferred embodiment of the method, the liver cancer is caused by bile accumulation in the liver. In one aspect, the method of treating a human subject suffering from liver cancer further comprises the step of administering to the subject an AG1856-GalNAc conjugate compound, wherein the AG1856-GalNAc conjugate compound has the chemical structure of formula (X), see Fig. 10, and wherein the AG1856- GalNAc conjugate compound is administered after the administration of the AON-GalNAcconjugate compound, preferably 1 , 2, 3, 4, 5, 6, 7, 8, 9, 10, 11 , 12, 13, or 14 days after administration of the AON-GalNAc conjugate compound, thereby treating, ameliorating, or slowing down the progression of the liver cancer in the subject.

[0130] The disclosure also relates to an AON as disclosed herein, for use in the treatment of a liver cancer in a human subject in need thereof, wherein the AON is capable of recruiting an endogenous ADAR enzyme in a liver cancer cell after the AON has formed a doublestranded complex with a region of a target pre-mRNA or mRNA molecule in the cell, wherein the region comprises a target adenosine, wherein the ADAR enzyme can deaminate the target adenosine into an inosine after binding to the double-stranded complex, wherein the target RNA nucleic acid molecule is a transcript molecule of the human SLC10A 1 gene encoding NTCP, wherein the deamination results in a change of the amino acid to an R at position 68 in the NTCP protein, wherein the AON is according to structural formula (I), and wherein R is a hydrogen atom or a GalNAc moiety, thereby treating the subject. In a preferred embodiment of the AON as disclosed herein, for use in the treatment of a liver cancer in a human subject in need thereof, R is a GalNAc moiety according to structural formula (II) or (III), more preferably formula (III).

[0131] The disclosure also relates to the use of an AON as disclosed herein, in the manufacture of a medicament for the treatment of a liver cancer, wherein the AON is capable of recruiting an endogenous ADAR enzyme in a liver cancer cell after the AON has formed a double-stranded complex with a region of a target pre-mRNA or mRNA molecule in the cell, wherein the region comprises a target adenosine, wherein the ADAR enzyme can deaminate the target adenosine into an inosine after binding to the double-stranded complex, wherein the target RNA nucleic acid molecule is a transcript molecule of the human SLC10A 1 gene encoding NTCP, wherein the deamination results in a change of the amino acid to an R at position 68 in the NTCP protein, wherein the AON is according to structural formula (I), and wherein R is a hydrogen atom or a GalNAc moiety, thereby treating the subject. In a preferred embodiment of the use of an AON as disclosed herein, R is a GalNAc moiety according to structural formula (II) or (III), more preferably formula (III).

[0132] Chemical modifications

[0133] 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 International Patent Application Publication No. W02024 / 084048 and as disclosed above, except that the opposite sense strand does nothave 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 (such as GalNAc moieties), that have also been described herein, and in detail in International Patent Application Publication No. W02024 / 084048, which may either be bound to the AON or its opposite strand, or both. GalNAc moieties that are used in combination with the AON as disclosed herein are preferably tri-antennary. Preferred GalNAc moieties that can be used in the context of the AONs as disclosed herein are disclosed in International Patent Application Publication No. WO2022 / 271806.

[0134] 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 International Patent Application Publication Nos. W02020 / 154342, WO2020 / 154343, and W02020 / 154344.

[0135] 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 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.

[0136] 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 delivery vehicle such as a viral vector or plasmid), thebiodistribution 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’-0Me, 2’-F and 2’-MOE modifications of the sugar and the use of PS linkages between nucleosides, as described herein.

[0137] Scaffold modifications (ribose)

[0138] 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’-OMe, 2’-MOE, 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’- OMe or 2’-MOE substitution when the nucleobase is a naturally occurring cytosine, but may carry a 2’-F, a 2’,2’-difluoro (diF), or 2’-ara-F (FANA) substitution or may be DNA. W02024 / 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 diF. 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).

[0139] 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 enzyme with adenosine deaminase activity 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’-fluoro (2’-F), 2’, 2’- difluoro (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. International Patent Application Publication No. 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 linkeror 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 International Patent Application Publication Nos. 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.

[0140] 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 2’,2’-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 with a sugar moiety that comprises a 2’- fluoro (2’-F) modification. A preferred position for the nucleotide that carries a 2’-F modification is nucleotide position -3 in the AON.

[0141] Base modifications

[0142] A base, sometimes called a nucleobase, is generally adenine, cytosine, guanine, thymine, hypoxanthine, 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, hypoxanthine, 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), pseudo-uracil, pseudocytosine, iso-uracil, 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 a-basic 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 chemistry 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.

[0143] 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’-OMe, 2’-O-(2-cyanoethyl), 2’-MOE, 2’-O-(2-thiomethyl)ethyl, 2’-O-butyryl, 2’-O- propargyl, 2’-O-allyl, 2’-O-(2-aminopropyl), 2’-O-(2-(dimethylamino)propyl), 2’-O-(2- amino)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-N- 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’- / V-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-CBBNmonomer, 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.

[0144] The orphan nucleotide

[0145] 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. Nat Struct Mol Biol. 2016. 23(5):426-433). 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 wildtype 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 wild-type 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. International Patent Application Publication No. 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.

[0146] 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’; Yang et al. Nucleic 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 as a 6-amino-5-nitro-3-yl-2(1 H)-pyridone nucleobase. 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.

[0147] Linkage modifications

[0148] 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 bonds. The phosphodiesters 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 phosphodiester 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 phosphorothioate, 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.

[0149] As outlined in detail herein, naked AONs as disclosed herein comprise at least one, preferably multiple linkage modifications. It is generally 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 phosphodiester present in RNA, such as phosphorothioate (PS), chirally pure PS, (R)-PS, (S)-PS, methyl phosphonate (MP or 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 (MM I), 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.

[0150] An AON as disclosed herein may also comprise one or more linkage modifications according to the structure of formulas (IV), (V), (VI), VII), or (VIII).

[0151] In a preferred aspect, the AON as disclosed herein comprises an internucleoside linkage of the structure of formula (IV), wherein X = O and R = CH3, which linkage is generally referred to herein as a PNms linkage. In other preferred aspects, R equals one of the following structures (a), (b), (c), (d), (e), (f), (g), (h), or (i):

[0152] The one or more PN linkages as depicted in formula (IV), present in an AON as disclosed herein, can be independently of each other of R or SP chirality, or stereo random.

[0153] The one or more PN linkages as depicted in formula (IV), in an AON as disclosed herein, can be in tautomeric and / or pH-dependent (de)protonated form, including but not limited to the structures (A), (B), (C), (D), and (E):

[0154] wherein X and R are as indicated above for formula (IV).

[0155] 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 at one and / or both termini with a structure according to formula (V):

[0156] wherein: X = O or S ;

[0157] Y = O’ or S' ; and

[0158] 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. A preferred internucleoside linkage modification that is used in the AON as disclosed herein has the structure of formula (VI):

[0159] which is also referred to as a PNms linkage. PNms linkages and their application in oligonucleotides and as replacement of PS linkages has been described (Chelobanov BP et al. Russ J Bioorganic Chem. 2017. 43(6):664-668; DOI: 10.1134 / S1068162017060024; Klabenkova K et al. Molecules. 2021 . 26(17):5420; Miroshnichenko SK et al. Proc Natl Acad Sci USA. 2019. 116(4): 1229- 1234), and for instance in oligonucleotides that may provide splice switching (Hammond SM et al. Nucleic Acid Ther. 31 (3): 190-200).

[0160] An AON as disclosed herein may comprise a substitution of one of the non-bridging oxygens in the phosphodiester 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 internucleosidic linkages between the nucleotides may be altered by mono- or di-thioation of the phosphodiester bonds to yield PS esters or phosphorodithioate esters, respectively. Other modifications of the internucleosidic 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 stable oligonucleotide compound.

[0161] 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 choiceof 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 (VII):

[0162] 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 a preferred embodiment, this position, in an AON as disclosed herein, comprises a linkage modification according to the structure of formula (IV), more preferably a linkage modification according to the structure of formula (VI), 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 additionalmanufacturing (purification) steps in the coupling and decoupling process. In one aspect, the AON does not comprise an MP linkage.

[0163] 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 that can be used in an AON as disclosed herein has the structure of formula (VIII):PNdmi linkage (VIII)

[0164] In one aspect, one or more of the PNdmi linkages as shown in structural formula (I) herein is replaced by a PNms linkage for manufacturing, RNA editing efficiency, or other reasons. Other internucleoside linkages that may be used in the AONs as disclosed herein are those that are disclosed in WO2023 / 278589.

[0165] Conjugate chemistries

[0166] 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, anester, 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 connects the ligand and the oligonucleotide, and may include for example ethylene glycol, triethylene glycol (TEG), HEG, alkyl chains, propyl, 6-aminohexyl, or dodecyl. In one embodiment, a GalNAc moiety is bound to the AON as disclosed herein via a TEG linker. 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.

[0167] General

[0168] 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 PNms linkage, 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.

[0169] 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., International Patent Application Publication Nos. 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 thedouble-stranded nucleic acid entity that arises. In all aspects of the disclosure, the enzyme with adenosine deaminase activity is preferably ADAR1 , ADAR2, or ADAT.

[0170] 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.

[0171] 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 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 is preferably replaced by an inosine (hence, at the -1 position within the AON), preferably a deoxyinosine (Id).

[0172] 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 double-stranded complexes (see for instance European Patent Application Publication No. EP 3954395 A1), 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 single-stranded 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.

[0173] The AONs as disclosed herein may also be administered in the context of aids that will increase the binding to the cell of interest, and entry of the AON into the cell of interest, and / or its endosomal escape as soon as it is entrapped in endosomes after entry 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 (see Fig. 10), disclosed in International Patent Application Publication No. WO2021 / 122998 and further described for use with RNA editing producing oligonucleotides in International Patent Application Publication No. WO2024 / 153801.

[0174] Disclosed herein is also a pharmaceutical composition comprising the AON as disclosed herein, preferably conjugated to a tri-antennary GalNAc moiety as described 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) that is also preferably conjugated to a tri-antennary GalNAc moiety as disclosed herein,which AG1856-GalNAc conjugate compound in fact may also be administered separately from the AON-GalNAc conjugate compound) and may be dissolved in a pharmaceutically acceptable organic solvent, or the like.

[0175] 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 and the efficacy of the active ingredient.

[0176] 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 to be used to recruit the deaminating enzyme, which therefore allows for a shorter AON and improved cellular delivery and trafficking.

[0177] 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 anoligonucleotide that comprises a mismatch opposite the adenosine to be edited. Following the instructions in the present application, those of skill in the art will be capable of designing the complementary portion of the oligonucleotide according to their needs.

[0178] 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.

[0179] Disclosed herein is the site-specific editing of target adenosine in RNA sequences in eukaryotic, preferably metazoan, more preferably mammalian, more preferably human cells, more preferably human liver cells, and most preferably human hepatocytes. 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 liver 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. In a therapeutic setting they are useful because they can be derived in vitro from a patient’s cells, and the organoids can then be re-introduced to the patient as autologous material which is less likely to be rejected than a normal transplant.

[0180] 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 over-expression 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 localizationsignals, 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.

[0181] 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, the target population, the mode of administration {e.g., intravenously or subcutaneously), 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 (such as plasma levels of bile acids for example). 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.

[0182] 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. As mentioned above, plasma level concentrations of bile acids in a sample from a treated subject is a proper biomarker for assessing the function of certain proteins in the subject, before and after treatment, or with or without treating the subject with an AON as disclosed herein. After this trial has been performed once then the knowledge can be retained, and future delivery can be performed without needing to take biopsy samples. A method as disclosed herein can thus 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.

[0183] After RNA editing has occurred in a cell, the modified mRNA can become diluted over time, for example due to cell division, limited half-life of the edited RNAs, etc. Thus, inpractical 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.

[0184] 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 50 mg / kg, more preferably from about 1 mg / kg to about 10 mg / kg. Preferred dosages of the AON RM 122067 (with or without, and preferably without administering AG1856-GalNAc) in humans are 3, 4, 5, 6, 7, 8, and 9 mg / kg, representing the lower part of an anticipated pharmacological active dose range and held sufficient for a clinically relevant effect.

[0185] Administration may be intranasally, orally, by injection or infusion, intravenously (IV), subcutaneously (SC), intradermally, intramuscularly, intra-tracheally, intra-peritoneally, intrarectally, intrathecally, intra-cisterna magna, parenterally, and the like. Preferably, the AON is administered by IV injection or by SC injection, more preferably by SC injection. 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. Preferably, when administered SC, the AON is dissolved to a solution in a suitable excipient for injection.

[0186] 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; or using a functional read-out, because the target RNA after the deamination should encode a protein with a lower or absent functionality, or on the other hand, an increased or regained functionality. The identification of the deamination into inosine may be a functional read-out using a suitable biomarker. 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 ofliver disease will preferably apply tests to monitor certain biomarkers related to liver function(s).

[0187] As outlined in the accompanying examples, suitable biomarkers are the concentration of total bile acids (TBA) in the serum and / or plasma, after AON administration, which will be indicative for a modulated NTCP activity. Circulating TBA should increase by an inhibited NTCP activity due to the RNA editing of the target adenosine in human SLC10A 1 transcripts as outlined herein. Another suitable biomarker for diminished or reduced NTCP activity is the increased expression of CYP7A1. The upregulation of CYP7A1 and measuring plasma C4 concentrations serve as a sensitive and specific surrogate biomarker for the decrease in intracellular bile acid levels caused by a reduced uptake after editing of the SLC10A 1 mRNA. Across the NHP studies as outlined in the accompanying examples, a positive correlation was observed between increased CYP7A1 expression and editing efficiency. CYP7A1 mRNA fragments can be measured in plasma or in extracellular vesicles and can be used as exploratory biomarkers in clinical trials and beyond.

[0188] In one embodiment, the present disclosure relates to a pharmaceutical composition comprising RM122067 (=RM120215+GalNAc; see SEQ ID NO:50), in which the RM122067 is present such that it can be dosed in a range of 1.0 to 10 mg / kg, preferably at a dose of 3.0 or 6.0 mg / kg.

[0189] In one embodiment, a method as disclosed herein comprises the steps of administering to the subject an AON or 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 the disease.

[0190] 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 interferon-gamma (IFN-y). hADARI is also inducible byTNF-a. This provides an opportunity to develop combination therapy, whereby IFN-y orTNF- 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.

[0191] 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. An AON as disclosed herein, when complexed to ADAR, preferably brings about the deamination of a single target adenosine.EXAMPLES

[0192] Example 1. c.203A>G editing (Q68R) in human SLC1OA1 transcripts in primary human hepatocytes.

[0193] A part of the human SLC10A1 transcript sequence (SEQ ID NO:1) is shown in Fig. 1 , indicating the CAG codon (for glutamine (Q) at position 68 in the NTCP protein) in bold. The underlined adenosine is the target for RNA editing as disclosed herein. Deamination of this adenosine into an inosine will result in a codon for an arginine (R) residue (CIG / CGG) at this position. Below the target sequence a set of 32 AONs is provided that were designed earlier (International Patent Application Publication No. W02024 / 200472) to target the target adenosine in SEQ ID NO:1. These AONs with their respective chemical modifications are represented by SEQ ID NO:2 to 33, respectively. These AONs were tested for editing efficiency in primary human hepatocytes (PHHs) as follows.

[0194] On day 0, PHHs (5.0x104cells / well) were transfected with AONs, in triplicates, using LIPOFECTAMINE® RNAiMAX Reagent at the same time of seeding, following the protocol of the manufacturer. The plates containing cells, medium and AONs were held at 37 °C, 5% CO2 for 72 hrs, during which the medium was refreshed 24 hrs after transfection / plating.

[0195] 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 DNasel (Thermo Fisher Scientific-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 Fisher Scientific-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 for this Q68R target (provided in Table 1) were used with a PCR program that was as follows: 10 min at 95 °C; 40 x 30 sec at 94 °C and 60 sec at 63 °C, 10 min at 98 °C and a hold step at 4°C. Then the plates were placed into a 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 as follows: score = SUM(G) / (SUM(A+G) * 100.

[0196] Table 1 : Primers and probes with their respective sequences for quantitative PCR assays. The SEQ ID NO of each primer or probe is given between brackets following the respective sequence in the middle column. The + indicates a Locked Nucleic Acid (LNA) on the 3’ side of the symbol.

[0197] Fig. 2 shows the result of the experiment described above for RM 117635 to RM 117647 and RM 117837 to RM 117846 in comparison to RM 108839 and the respective negative controls as indicated. Clearly, RM 117635 and RM 108839 outperform the other tested AONs in this experiment.

[0198] Then a similar experiment was performed with a smaller number of AONs and in which cells were treated with 5 pM AON in the presence of 0.5 pM saponin (AG1856) instead of LIPOFECTAMINE® RNAiMAX Reagent, for 72 hr. Fig. 3 shows the result of this additional experiment using AONs RM 108821 , RM 108826, RM 108827, RM 108836, RM 108838, RM 108839, RM 108840, RM 107352 and RM 107368. This experiment shows the reproducible high editing percentages obtained with RM 108821 (SEQ ID NO:27) and RM 108839 (SEQ ID NO:32) with editing levels above 55%.

[0199] Example 2. Optimization.

[0200] International Patent Application Publication No. WQ2024 / 200472 describes the beneficial properties of several AONs regarding the c.203A>G editing (Q68R) in human SLC10A 1 transcripts, including RM108839 in Example 1 above. RM108839 is a 30-mer and has the following sequence and modifications (see Fig. 1 for details; orphan nucleotide in bold): 5’- Cm!Cm*Gm*Uf*Gf*Ae9Ge9Gf*Gm9Ge*Cf!Af9Um*Gf!Am9m5lle9Gm*Cm*Cf!Ae9m5Ue9Af*Cd*Zd*ldA|Jm9Gf*m5Ce*Cm!Am-3’ (SEQ ID NO:32).

[0201] In contrast to multiple other AONs that also appeared to comprise beneficial properties, such as RM 107385, RM 108839 comprises a stretch of four guanosines (underlined in the above sequence) and is relatively high in G / C content. RM 107385 is also described in International Patent Application Publication No. WQ2024 / 200472, and has the following sequence and modifications (a 25-mer; see Fig. 1 for details of the modifications; orphan nucleotide in bold): 5’-m5Ce#Ae9m5Ue9Gm*Am*Um*Gf*Cm*Cf*Am*Uf*Am*m5Ce9Zd* ldA|Jm9Gf*m5Ce*Cm*Af*Cm*Cf*Am*Gf#Gm-3’ (SEQ ID NO:79).

[0202] It was envisioned that a stretch of four or more guanosines in an oligonucleotide and / or the presence of a relatively high number of G / C nucleotides may pose problems in respect of immunogenicity and secondary structures. On top of that, the combination of certain linkages and certain chemical modifications in RM 108839 appeared to prevent an easy scale up process in manufacturing (data not shown). It was investigated whether RM 108839 could indeed form undesired secondary structures and whether it could form homodimers, or unwanted, potentially immunogenic, G-quadruplexes secondary structures (G4 or GQs). Any secondary structures could in principle prevent high editing efficiencies in vivo because it may pose problems in hybridizing to the target sequence.

[0203] In a UV spectrophotometric characterization, it was found that certain monomeric 2D structures and homodimerization of the AON could take place (data not shown). Notably, a negative peak at 295 nm, which is the hallmark of GQ formation, was not observed in isothermal difference spectroscopy, suggesting that such GQ structures did not occur. The hypothetical 2D structure prediction of a homodimer is shown below (formula (IX)) with two identical AONs of SEQ ID NO:32:

[0204] This structure predicts that it may be in competition with intermolecular GQ formation.

[0205] To check for dimerization in vitro, 40 ptM RM 107385 and RM 108839 solutions were freshly prepared, kept in water, or held in 1x PBS pH 7.4 for 6 days at 5°C, 40°C or -20°C. For each sample, 10 |J_ of a 40 ptL AON solution was mixed well with 2.5 ptL 6x TriTrack DNA Loading Dye (Thermo Fisher Scientific, R1161). The samples were then loaded on a 20% TBE gel (Invitrogen, EC63152BOX), resolved, and stained with a toluidine blue solution(0.01 % toluidine blue, 7.5% acetic acid in water) to visualize the oligonucleotides. Fig. 4 shows the results of an experiment with RM 107385 and RM 108839 using a variety of such conditions as indicated. Clearly, RM 107385 shows the same band at the same level under all three tested conditions, which is likely the monomer of that AON. However, RM108839 shows a single band (likely representing its monomer) when held in water but depicts a higher band (likely representing a homodimer) when held in PBS pH 7.4. This suggests that under physiological conditions, RM 108839 can form homodimers as predicted by formula (IX) above, which is disrupted in water. The different temperature conditions did not change this effect. When the RM 108839 AON was kept in water or in PBS and subsequently tested for editing efficiency, the AON when kept in water provided reproducible high editing results as previously observed (approximately 50%), whereas the PBS-contained AON showed significantly lower editing efficiency (approximately 18-20%; data not shown), also suggesting that the secondary structures observed in physiological conditions negatively influences the rate of editing. Moreover, when RM 108839 was conjugated to a GalNAc moiety at the 5’ terminus and held in PBS editing efficiency dropped significantly as well in comparison to samples contained in water (data not shown).

[0206] It was then realized that it may be beneficial to alter the nucleotide content and chemical modifications of RM 108839. For this, a new set of AONs was designed with a variety of replacements in the GGGG stretch present in RM 108839, as well as a variety of PO linkages being replaced by PS linkages. These newly designed AONs and their respective modifications are shown in Fig. 5. RM 119976 to RM 119995 contain a-basic nucleotides distributed within the GGGG stretch with “(1)” representing a 2’-OMe substitution in the sugar moiety and “(2)” representing a 2’-F substitution in the sugar moiety. RM 120203 to RM 120222 are varied in the GGGG stretch by replacing one or more of these guanosine residues with a nucleotide carrying an uracil nucleobase, either substituted at the 2’ position in the sugar by a 2’-OMe (Um), 2’-MOE (m5Ue), or a 2’-F (Uf) group.

[0207] Several of these AONs were first assessed in a non-denaturing gel, following the methods as described above. Fig. 6A and Fig. 6B show the results of these experiments. Like what has been shown in Fig. 4, RM 108839 - when held in water - displays a single band (left lane in both figures, far right lane in Fig. 6B), suggesting the disruption into monomers only, while containment in PBS provides multiple bands (2ndlane in both figures, and 4thlane in Fig. 6B), suggesting the presence of (homo) dimer structures. Fig. 6B (3rdlane) again shows the lower band observed for RM 107385, when held in PBS.

[0208] Interestingly, some newly designed AONs, such as RM119996 (SEQ ID NO:78), RM120213 (SEQ ID NO:48) and RM120214 (SEQ ID NO:49) also suggest dimerizing by displaying multiple bands. Strikingly however, by applying only very minor differences, suchas introduced in RM 120215 (SEQ ID NO:50) and RM 120216 (SEQ ID NO:51) the assessment showed single bands in these non-denaturing conditions after being held in PBS. Both these AONs have a single G>U change in the GGGG stretch discussed above. This suggests that these AONs (that differ from RM 108839 at a few positions only) have a significant lower tendency to dimerize, whereas others with similar minor changes still appear to dimerize.

[0209] Example 3. c.203A>G editing (Q68R) in human SLC10A1 transcripts in PHHs using RM108839-derived AONs.

[0210] All AONs displayed in Fig. 5 were tested for editing SLC10A1 transcripts in PHHs in an in vitro set up using 5 pM AON and 72 hrs incubation with the AG1856 saponin. RNA editing was determined as described above, using the same primers and probes as provided in Table 1. The results are depicted in Fig. 7, that shows that many of the AONs comprising a-basic nucleotides (RM119976 to RM119995) underperformed in comparison to many of the AONs that comprised G>U changes in the GGGG stretch. While RM 108839 scored relatively high in this assay, also RM119996, RM120205, RM120215, RM120216, and RM 120218 gave editing percentages that were like what was observed with RM 108839.

[0211] In a subsequent experiment, some of these AONs, comprising mismatches and / or a- basic nucleotides were again tested in a transfection experiment using LIPOFECTAMINE® RNAiMAX transfection or using AG1856 as the endosomal escape agent, using 100 nM AON in the case of LIPOFECTAMINE® RNAiMAX and 5 pM + 0.5 pM AG1856 in the case of the saponin, in each case for 72 hrs until RNA purification, all following the methods as described above. The results are shown in Fig. 8. This showed that in both transfection methods (LIPOFECTAMINE® RNAiMAX or AG1856) RM 120215 outperformed RM108839, providing editing percentages as high as 65%.

[0212] This experiment was then repeated using RM 120215, RM 108839, RM 118995, and RM 107385 in a variety of concentrations (5, 10, 20, 40, 60, 80, 100, and 120 nM), using LIPOFECTAMINE® RNAiMAX as the transfection agent. RM118995 is as follows (modifications as in Fig. 1): 5’-Cm!Cm*Gm*Uf*Gf*AeeGe*Gf*GfeGe*Cf!AmeUf*Gf!Afem5Ue* Gf*Cf*Uf*Aeem5UeeAf*Cd*Zd*ldAUmeGf*m5Ce*Cm!Am-3’ (SEQ ID NO:80).

[0213] RNA editing was determined again after 72 hrs, as described above. The results are provided in Fig. 9 and show that in all concentrations RM 120215 outperformed the other three AONs in respect of editing efficiency on the target adenosine. This clearly shows that the inventors were able to provide a new AON with lower ability to form (homo) dimers under physiological conditions and with an improved ability to provide RNA editing of the Q68R target site in human SLC10A 1 transcript molecules, even though the introduced chemicalmodifications were a few changes in comparison to RM 108839, with a change from Gm to Um at nucleotide position +15, and a replacement of the PO linkages at nucleotide positions +11 and +14 with PS linkages.

[0214] Example 4. Conjugates ofRM120215 to different GalNAc moieties and activity in vivo.

[0215] In a next set of experiments, RM120215 was conjugated to a GalNAc moiety to target the AON specifically to human hepatocytes in vivo. For this, RM 120215 as depicted in its naked form in structural formula (I) herein, was conjugated to the GalNAc of structural formula (II) at its 3’ terminus, using chemical procedures known to the person skilled in the art, generating an AON referred to as RM 122043. This AON has the same nucleotide sequence and chemical modifications in the nucleotide sequence of SEQ ID NO:50. In another set up, RM120215 as depicted in its naked from in structural formula (I) herein, was conjugated to the GalNAc of structural formula (III) at its 3’ terminus, using chemical procedures known to the person skilled in the art, generating an AON referred to as RM 122067. This AON also has the same nucleotide sequence and chemical modifications in the nucleotide sequence of SEQ ID NO:50. RM 122067 was tested in in vitro and in vivo settings, to determine the cell-targeting efficiency followed by their intracellular RNA editing efficiency that follows delivery to the cells.

[0216] First, editing efficiencies brought about by RM120215 (no GalNAc) was compared to RM122067 (+GalNAc) with and without the addition of AG1856, in PHHs, in triplicates, following the protocol as described in Example 1 and 3. Incubations, concentrations, editing analysis, etc. were as described above. The unrelated RM4266 oligonucleotide, a nontreated sample and a mock treated sample served as negative controls. The results are provided in Table 2 and show that RM120215 outperformed RM122067, which is not unexpected under in vitro circumstances, and it also shows that the GalNAc moiety does not hamper the activity triggered by the AON in these human cells.

[0217] Table 2: Editing percentages of RM120215 and RM122067 in PHHs after co-treatment with AG1856 (non-conjugated to GalNAc).

[0218] A concentration experiment was then performed using 4 different concentrations of RM122067 (0.01 pM, 0.1 pM, 1.0 pM, and 5.0 pM) in a co-treatm ent with AG1856 (also in 3 different concentrations: 0.1 pM, 0.25 pM, and 0.5 pM) in PHHs, in triplicates. The experiment was performed as outlined in Example 1 and 3. The editing results (percentages) are given in Table 3 and show that an increasing concentration of AG 1856 results in an increased editing percentage, with 0.1 pM AG1856 giving relatively low editing, which was still above background levels. The results also show that low concentrations of oligonucleotide (0.01 pM) could still give editing levels of >30%, when AG1856 was coapplied. A mock control (AG1856 only using 0.5 pM saponin) and a non-treated sample served as negative controls. This experiment was also performed using the AG1856-GalNAc conjugate, but then all editing percentages were relatively low, likely because most if not all asialoglycoprotein receptors were occupied with the GalNAc moiety of the AG1856 compound, leaving less room for the oligonucleotide to enter the cells in this in vitro set-up (data not shown). In vivo data should show whether such would also be true in the liver of living organisms.

[0219] Table 3: Editing percentages in PHHs after administration of 0.01 pM, 0.1 pM, 1.0 pM, and 5.0 pM RM122067 in a co-treatment with AG1856, using three different concentrations (0.1 pM, 0.25 pM, and 0.5 pM).

[0220] Example 5. Editing efficiency of RM120215 / RM122067 and surrogate sequences in Primary Mouse and Cynomolgus Monkey hepatocytes in co-incubation with AG1856.

[0221] Editing efficiency of a 3’-GalNAc-conjugated mouse surrogate oligonucleotide (RM122098; SEQ ID NO:81 , see Fig. 5) was evaluated head-to-head with RM122067 and its unconjugated version RM 120215 in freshly isolated primary mouse hepatocytes (PMHs) using: i) transfection (100 nM oligonucleotide); ii) gymnosis (5 pM oligonucleotide); and iii) AG1856 pulse treatment conditions (5 pM oligonucleotide).

[0222] Fig. 11 shows the alignment between the human SLC10A 1 mRNA target sequence and the mouse Slc10a1 mRNA target sequence. It is shown that within the complementary sequence of the AONs (underlined) no less than 5 mismatches (provided by asterisks) are present between the human and the mouse sequence. The mouse sequence also comprises the stretch of four cytidines opposite positions +14 to +17 in the AON. RM 122098 is complementary to the mouse Slc10a1 sequence, comparable to RM120215 and RM122067 that are complementary to the human sequence and include the 2’-OMe modified nucleoside carrying an uracil nucleobase at position +15.

[0223] For transfection, -75,000 cells were seeded in a 24-wells format, and 100 nM AON was transfected using LIPOFECTAMINE®. Cells were harvested after 72 hrs and editing percentages were determined as outlined above. The RNA was isolated using the RELIAPREP™ kit from PROMEGA®.

[0224] For gymnosis, -75,000 cells were seeded in 24-wells format immediately in the presence of 5 pM AON (a so-called ‘seed & treat’ procedure), with fresh medium replacing the original medium at 24 hr after seeding, and then cells were harvested 48 hr after seeding and assessed for RNA editing.

[0225] For AG1856 pulse treatment, -75,000 cells were seeded in 24-wells format immediately in the presence of 5 pM AON (seed & treat) and then 0.5 pM AG1856 was added to the medium 20 hr after seeding (‘pulse’), after which the medium was replaced 4 hr later, and cells were harvested at 72 hr after seeding.

[0226] For the ddPCR, the primers were as provided for the human sequence in Table 1 . The probes were as follows: mSLC10A1_e01_edit-G_FAM: 5’-ATAGTGGCCCGGTACG GTATC-3’ (SEQ ID NO:87) and mSLC10A1_e01_orig-A_HEX: 5’-ATAGTGGCCCAGTA CGGTATCAT-3’ (SEQ ID NO:88).

[0227] RM 122098 showed robust editing using transfection-mediated delivery (28%), with more than 4-fold lower editing (6.1%) with AG 1856 pulse treatment and negligible editing with gymnotic uptake (0.31 %). The AONs targeting the human target mRNA, RM122067 and RM120215, also showed negligible editing in these PMHs (<0.5%) using the differenttreatment regimens. This was not unexpected in view of the lack of complementarity of the human sequence targeting AONs and the mouse target sequence, but it also showed that PMHs were not quite suitable for determining in vitro RNA editing of the target transcript molecules using gymnosis, even though RM 122098 was conjugated to a GalNAc moiety.

[0228] Editing efficiency of two cynomolgus surrogate oligonucleotides (RM122104; SEQ ID NO:82, see Fig. 5; and RM122105; SEQ ID NO:83, see Fig. 5) was assessed head-to-head with RM120215 in primary cynomolgus monkey hepatocytes (PCHs) using the AG1856 pulse treatment as outlined above, except that each of the three triplicates were a pooled sample of three independently treated wells and the number of cells per well was -90,000 PCHs. The RNA was isolated using a zymogen kit. The ddPCR was performed with the same primers and probes as provided for the ddPCR in PHHs, see Table 1.

[0229] Fig. 11 shows the alignment between the human SLC10A 1 mRNA sequence and the monkey (macaque) SLC10A 1 mRNA sequence. It is shown that within the complementary sequence of the AONs (underlined) only one mismatch (provided by an asterisk) is present between the human and the macaque sequence. RM 122105 is fully complementary to the monkey target sequence (except at the orphan position where the orphan nucleotide is a Zd and is a mismatch with the target adenosine, and the position -1 in the AON, that is a deoxyinosine (Id), which is positioned opposite the 5’-C in the target sequence. RM1220104 is identical to RM 122105, except that it comprises a 2’-OMe modified nucleoside comprising an uracil nucleobase at position +15 in the AON (as introduced in RM120215 and RM 122067, discussed above), instead of the 2’-OMe modified nucleoside comprising a guanine at that position in RM122105.

[0230] RM 120215 demonstrated robust editing efficiency in PCHs (42% and 53% in two separate experiments). Similar editing efficiencies (42% in both cases) were observed with the fully homologous surrogate AON RM122105 and RM122104, which is the AON with the 1 additional mismatch with the monkey target sequence in comparison to RM122105 at position +15. Apparently, the presence or the absence of the additional mismatch at position +15 in the AON did not change much in editing efficiency of the monkey target sequence in these PCHs.

[0231] The consistent performance of the SLCIOA I-targeting AONs in human and nonhuman primate cell models indicated the translatability of the editing approach across species. The robust activity of RM120215 (without the GalNAc moiety) in PCH suggested that RM122067 (= RM120215+GalNAc) should demonstrate proof-of-concept in non-human primates, in vivo, see examples below.

[0232] Example 6. Editing efficiency of RM122067 in Primary Human Hepatocytes + / - AG1856-GalNAc.

[0233] In a further in vitro experiment, gymnotic uptake of the RM122067 AON was investigated, since this oligonucleotide should recognize its asialoglycoprotein receptor on the PHHs, and entry followed by editing of the target transcript molecules should potentially also take place in the absence of AG1856. The procedures were as described above, with a seed & treat procedure using 5.0 pM AON, and harvesting cells after 24, 48, and 72 hr, in triplicates (24 hr) and duplicates (48 and 72 hr). An unrelated, r\or\-SLC10A1 targeting AON and a non-treated sample served as negative controls and showed a background level of 0.00-0.12% editing (data not shown). The results are provided in Table 4 and show that the highest editing level was observed after 48 hr gymnotic uptake (>17%) showing that it was feasible to apply a RM120215 AON conjugated to a GalNAc moiety (of formula (III)) leading to the compound RM122067 could enter the PHHs and reach the target transcripts also without the aid of the endosomal escape enhancer AG 1856.

[0234] Table 4: Editing percentages in PHHs after gymnotic uptake of RM122067, with triplicates at the 24 hr timepoint, and duplicates at the 48 and 72 hr timepoints, avg = average between the triplicates / duplicates.

[0235] A similar time path experiment was then conducted using the unconjugated RM120215 AON in comparison to RM122067 in the presence of the AG1856-GalNAc conjugated compound, using a seed & treat procedure followed by a pulse with AG1856-GalNAc, wherein the cells were seeded in the presence of 5.0 pM AON, then 0.5 pM AG1856-GalNAc was added 20 hr after seeding, which was washed away with fresh medium after 4hr. At that stage the 24 hr timepoint cells were harvested, followed by harvesting 48 hr cells and 72 hr cells thereafter. The experiment was conducted in triplicates for all three timepoints. A mock treatment and a non-treated sample served as negative controls and showed a background level of 0.21% editing max (data not shown). The results are provided in Table 5 and show that again highest editing was observed with RM120215 lacking the GalNAc moiety (see also Table 2).

[0236] Table 5: Editing percentages in PHHs after seed & treat with RM122067 (= RM120215-GalNAc) and RM120215, and a 4 hr pulse treatment with AG1856-GalNAc. avg = average between the triplicates.

[0237] Subsequently, RM120215 and RM122067 were tested with two different concentrations (1.0 and 5.0 pM) using a seed & treat procedure, a wash-out after 24 hr and an ongoing culturing up to 96 hr, in triplicates, but without using AG1856 or the AG1856- GalNAc conjugate compound, which is therefore a complete gymnotic uptake procedure. A non-treated sample served as negative control and showed a background level of 0.15% editing max (data not shown). The results are shown in Table 6 and indicate that in the absence of saponin, RM 122067 is significantly more efficient in providing editing than RM 120215, suggesting the beneficial properties of the GalNAc moiety to enter the PHHs. Moreover, editing levels were still robust after 4 days of culturing. Furthermore, the data shows that increasing the concentration of AON five-fold did not provide an increased editing effect, suggesting a very long and active availability of the AON in the cells even with a concentration of 1.0 pM. The percentages observed with RM122067 at the 24, 48, and 72 hr timepoints are in line with the data shown in Table 4.

[0238] Table 6: Editing percentages in PHHs after seed & treat with 1.0 and 5.0 pM RM 122067 and RM 120215, in the absence of saponin, avg = average between the triplicates.

[0239] Yet, in a further in vitro experiment, repeated dosing with 0.5 and 1.0 pM RM120215 and RM 122067 was investigated. For the single dose treatment, PHHs were seeded at -75,000 cells per well in the presence of AON (seed & treat), followed by a wash-out at 24 hr and harvesting at 48, 62, and 72 hr after seeding. For the repeated dosing, PHHs were seeded at -75,000 cells per well in the presence of AON (seed & treat), followed by replacement of the medium at 24 hr, but at that stage again in the presence of AON in thesame concentrations. Cells were harvested at 48, 62, and 72 hr post-seeding, in triplicates. A non-treated sample served as negative control and showed a background level of 0.16% editing max (data not shown). The results are provided in Table 7, showing that RM122067 again outperformed RM 120215 in the absence of saponin, and that robust editing levels were also obtained by using the 0.5 pM concentration. Under these in vitro circumstances, a repeated dosing did not show a significant increase in editing levels likely because of disappearance of the asialoglycoprotein receptor during prolonged (attached) culturing of hepatocytes, albeit that with repeated dosing in most cases the editing percentage was higher than observed in the single dose regimen. Like the results shown above, RM 122067 was more efficient in triggering RNA editing in these PHHs than RM120215, in the absence of AG 1856 or AG1856-GalNAc.

[0240] Table 7: Editing percentages in PHHs after seed & treat with 1.0 and 0.5 pM RM122067 and RM120215, in the absence of saponin, sd = single dose, rd = repeated dose, avg = average between the triplicates.

[0241] Example 7. Editing efficiency of RM122067 in non-human primates.

[0242] RM 122067 powder for solution for injection is a pale yellow to yellow cake or powder in a sterile glass vial closed with a rubber stopper and flip-off cap. The drug product can be administered subcutaneously (SC) or intravenously (IV) following reconstitution in the vial with a 1 :1 (v / v) mixture of sterile saline and water for injection (WFI).

[0243] A first combined proof-of-concept and dose-range finding study in non-human primates (NHP; Cynomolgus monkey) was performed investigating the editing efficiency of RM122067 with and without administration of AG1856-GalNAc and to support dose level selection for a pivotal repeat-dose toxicology study with RM 122067.

[0244] Animals (1 / sex / dose) received repeat SC dosing with 10 or 50 mg / kg RM122067 on day 1 , 3, 5, and 8, followed by an IV 1 hr infusion on day 12 and sacrifice on day 15.

[0245] One control animal received a single dose of 2 mg / kg AG1856-GalNAc, and one control animal received a single dose of 5 mg / kg AG1856-GalNAc.

[0246] The combination group (7 animals in total) received a single dose of RM 122067 (SC) and a single dose of AG1856-GalNAc (SC), as follows: i) 2.5 mg / kg RM122067 + 0.3 mg / kg AG1856-GalNAc; ii) 5.0 mg / kg RM 122067 + 0.3 mg / kg AG1856-GalNAc; iii) 10 mg / kg RM122067 + 0.1 mg / kg AG1856-GalNAc; iv) 10 mg / kg RM122067 + 0.3 mg / kg AG1856- GalNAc; v) 10 mg / kg RM122067 + 1.0 mg / kg AG1856-GalNAc; vi) 10 mg / kg RM122067 + 2.0 mg / kg AG1856-GalNAc; and vii) 50 mg / kg RM122067 + 0.1 mg / kg AG1856-GalNAc.

[0247] Monkey liver samples were homogenized by the QIAGEN TissueLyser II with TRIzol INVITROGEN®). RNA was extracted using chloroform and precipitated with an equal volume of isopropyl alcohol. After incubation at -20°C, RNA pellets were washed twice with 70% ethanol and dissolved in RNase-free-water. The RNA samples were treated with Bio-Gel P- 60 Gel columns to remove oligonucleotides. The concentration of the RNA was measured by NANODROP™ One (Thermo Fisher Scientific) and adjusted to 200 ng / pL. RNA was incubated in DNA-removal reaction mixture (DNase I) at 37°C for 30 min to remove genomic DNA. DNase I was inactivated by heating at 65°C for 10 min in the presence of 1 pL 50 mM EDTA. After the DNA removal, RNAs were reverse transcribed to cDNA as follows: Reverse transcription reaction mixture (Primer Oligo (dT), random hexamer, dNTP mix) was incubated at 75°C for 5 min; MAXIMA™ reverse transcriptase was incubated at 25°C for 10 min and at 50°C for 30 min. The reaction was terminated by heating at 85°C for 5 min.

[0248] The ddPCR was performed with the Digital PCR System (Bio-Rad, QX200™ droplet reader) in 20 pl aliquots of reaction mixtures containing cDNA, appropriate pairs of primers and ddPCR Supermix for Probes (no dllTP) (Bio-Rad). The primers and probes for this Q68R target (provided in Table 1) were used with a PCR program that was as follows: hot start 10 min 95°C, denaturation at 94°C, 40 cycles of annealing / extension for 90 sec at a gradient from 65 °C to 55°C, enzyme deactivation for 10 min at 98°C, and storage at 4°C. The editing percentage was calculated as provided in Example 1.

[0249] Low levels of RNA editing were observed in liver biopsies after repeat-dose administration of RM122067 alone (0.34% at 10 mg / kg and 0.57% at 50 mg / kg). However, upon co-treatment with 0.3 mg / kg AG1856-GalNAc, robust editing was observed, with editing efficiency relatively independent of the RM 122067 dose levels at 2.5 (-27%), 5.0 (-28%) and 10 mg / kg (-40%). Using a fixed dose of RM122067 (10 mg / kg), editing efficiency increased with increasing AG1856-GalNAc dose with maximum editing of 62% at 1 mg / kg. At the lowest dose of AG1856-GalNAc (0.1 mg / kg), editing efficiency remained relatively independent of the dose with 12% editing at a dose of 10 mg / kg RM 122067 and 7% editing at a dose of 50 mg / kg RM 122067.

[0250] The conclusion from this initial study was that SC administration of RM 122067 at 10 or 50 mg / kg on days 1 , 3, 5, and 8, followed by a 1 h IV infusion on day 12 was well-toleratedwithout mortality or clinical signs. No adverse effects were observed up to 50 mg / kg RM122067. Although editing appeared low in the animals that received only RM122067, robust editing was observed in co-administration with AG1856-GalNAc at all doses.

[0251] In a next study, the toxicity profile, toxicokinetic, and safety pharmacology of RM122067 was assessed in NHPs. RM122067 was administered by 1 h IV infusion once weekly for 5 weeks at doses of 0 and 10 mg / kg on days 1 , 8, 15, 22, and 29. The animals were evaluated up to day 36 (main dosing phase). Additional high dose groups received 50 and 250 mg / kg but pharmacodynamics were not performed in these groups. Subgroups in the control and high dose groups were evaluated up to day 113 (12-week recovery phase).

[0252] In a parallel study, RM122067 was combined with AG1856-GalNAc by SC administration as follows: animal A receiving 2 mg / kg RM 122067 on day 1 , 8, 15, 22, and 29, and 0.1 mg / kg AG1856-GalNAc on day 30; animal B receiving 2 mg / kg RM122067 at day 1 , 8, 15, 22, and 29, and 0.3 mg / kg AG1856-GalNAc on day 30; animal C receiving 2 mg / kg RM 122067 at day 1 , 8, 15, 22, and 29, and 1.0 mg / kg AG1856-GalNAc on day 30; animal D receiving 10 mg / kg RM 122067 on day 1 , 8, 15, 22, and 29, and 0.1 mg / kg AG1856- GalNAc on day 30; animal E receiving 10 mg / kg RM122067 on day 1 , 8, 15, 22, and 29, and 0.3 mg / kg AG1856-GalNAc on day 30, and animal F receiving 10 mg / kg RM122067 on day 1 , 8, 15, 22, and 29, and 1.0 mg / kg AG1856-GalNAc on day 30. Animals were sacrificed on day 36 after which RNA editing was assessed in the liver as described above.

[0253] In-life assessments included mortality checks, clinical signs including injection site reactions, body temperature, body weight, food consumption, ophthalmology, safety pharmacology (cardiovascular, respiratory and neurological examination), clinical pathology (hematology, clinical chemistry, urinalysis including kidney biomarkers kidney injury molecule-1 , ALB or beta2-microglobuline), cytokine and complement response. Postmortem evaluations included gross pathology, organ weights, and histopathology. All doses were well-tolerated without any clinical signs. No significant sex differences in systemic exposure were observed at any dose level.

[0254] Like the findings in the initial NHP study, editing levels appeared low (<1 %) in both 10 and 50 mg / kg (RM 122067 alone) groups after IV infusion.

[0255] In the combination study using SC administration of RM122067 and AG1856-GalNAc, editing levels were significantly increased, as follows: animal A 14.8%, animal B 25.4%, animal C 39.9%, animal D 7.2%, animal E 26.2%, and animal F 38.5%, showing that increasing dosing of AG1856-GalNAc provided an increase in editing percentages, whereas not much difference was observed between the three animal receiving 10 mg / kg RM 122067 and the three animals receiving 50 mg / kg RM 122067.

[0256] In a further study, the pharmacodynamics of RM 122067 was assessed in 3 groups of 4 male NHP after weekly SC dose administration of 1 , 3, or 6 mg / kg RM122067 on 4 occasions (day 1 , 8, 15, and 22) followed by single SC injections of AG1856-GalNAc with 0.1 mg / kg one week (day 29) after the last RM 122067 administration and with 0.2 mg / kg at three weeks (day 44) after the last RM 122067 administration (except for the 6.0 mg / kg group). 1 group of 4 male NHPs received PBS instead of RM122067. All animals received 0.1 and 0.2 mg / kg AG1856-GalNAc on day 29 and day 44, respectively. For exploratory purposes this was followed by a third administration of AG1856-GalNAc at different dose levels (0.2 and 0.4 mg / kg) for each group (resulting in n=2 for each subgroup) at day 58, except that one animal in the 6 mg / kg group did not receive the 0.4 mg / kg AG1856-GalNAc on day 58. Two animals in the negative control group received 0.1 mg / kg AG1856-GalNAc on day 58 and two animals in this control group received 0.3 mg / kg AG1856-GalNAc on day 58. Editing was determined in liver biopsies that were taken 2-3 days after each AG1856- GalNAc administration (day 32, 46, 61 , and 75).

[0257] After four weekly administrations of RM 122067 as outlined above, and administration of 0.1 mg / kg AG1856-GalNAc one week later, consistent SLC10A1 editing at the C.203A site was observed across the different RM 122067 dose groups (11.3%, 12.1 %, and 11.1 % at 1 , 3, and 6 mg / kg RM122067, respectively). Additional AG1856-GalNAc administration at 0.2 mg / kg, two weeks later further sustained and increased SLC10A 1 C.203A editing in the 1 and 3 mg / kg RM 122067 groups (13.4% and 19.3%, respectively). A third consecutive injection of AG1856-GalNAc appeared less effective to maintain editing (5%, 8%, and 12% at 1 , 3, and 6 mg / kg RM122067, respectively), which was without additional RM122067 administration. When a higher dose of AG1856-GalNAc (0.4 mg / kg) was used at day 58, 13% and 20% editing was determined in the animals that received 1 and 3 mg / kg RM 122067, respectively. In the 6 mg / kg dose group, for the animals without additional AG1856-GalNAc administration, the editing percentage gradually declined from 11.1% to 7.4% after two weeks and 3% after four weeks. This suggests a PD half-life of approximately 14 days after release of additional RM 122067 from endosomes.

[0258] In vivo, the effects of RM 122067 on RNA editing efficiency and its functional impact through plasma bile acid levels and CYP7A1 expression in the liver can be evaluated. CYP7A1 is a rate-limiting enzyme in bile acid synthesis which is upregulated when intrahepatic bile acid levels decrease due to NTCP inhibition. In the latest NHP study outlined above, with the four weekly administrations of RM 122067, the effects of the RM 122067 compound on bile acid profiles were assessed with the use of taurourosodeoxycholic acid (TLIDCA), a bile acid taken up into hepatocytes primarily via NTCP, as surrogate marker for NTCP activity (Dong Z et al. Int J Pharm. 2015, 478(1):88-95). While circulating bile acid profiles can indicate NTCP modulation by RM122067, their variability due to diet and gallbladder activity limits their reliability. As a more controlled alternative, the clearance of orally administered TLIDCA can therefore serve as a surrogate marker. Reduced NTCP activity would impair TLIDCA uptake, leading to delayed clearance from the bloodstream and increased plasma TLIDCA (and other conjugated bile acids) concentrations. In all 16 animals in the NHP study (one group receiving no RM122067, one group receiving 4x 1 mg / kg RM 122067, one group receiving 4x 3 mg / kg, and one group receiving 4x 6 mg / kg RM 122067), 80 mg / kg TLIDCA was orally dosed on days 25, 36, and 65.

[0259] The CYP7A1 expression was generally higher in treated animals compared to control animals. Noteworthy, CYP7A1 also declined with decreasing editing percentage in the animals of the 6 mg / kg RM 122067 group that did not receive a further administration of AG1856-GalNAc.

[0260] After an oral TLIDCA administration, an increase in the total bile acid (TBA) time profile was observed between 1 and 4.5 hr after administration, although there was considerable variability at different time points. Repeat administration of RM 122067 had no effect on the area under the curve (AUC) or maximal plasma concentration (Cmax) of the TBA time profile after TLIDCA, consistent with the lack of editing in NHP observed in the previous studies when animals were treated with RM122067 alone. After administration of AG1856-GalNAc and assessment on day 29 and 58, the Cmax and AUC TBA exposure after TUDCA increased in the 3 mg / kg RM 122067 group, in which editing between 12% and 19% was observed (3.0 to 3.7-fold compared to the control group without RM 122067).

[0261] In conclusion, repeated 4-weekly RM122067 dosing followed by AG1856-GalNAc dosing resulted in consistent RNA editing of the target adenosine in the SLC10A1 transcript, which could be enhanced and sustained with additional AG1856-GalNAc doses. CYP7A1 expression also increased related to editing. Following TUDCA administration, Cmax and AUC were elevated compared to control, and associated with editing in the range of 12% to 20%. Together, these findings are indicative of reduced hepatic bile acid uptake due to NTCP inhibition.

[0262] Example 8. Editing efficiency of RM122067 in a humanized liver mouse model.

[0263] Since the editing effect of RM 122067 alone without the aid of the endosomal escape enhancer AG1856-GalNAc compound in NHP appeared relatively minor, it was decided to study the editing efficiency of RM 122067, in a pre-clinical set-up, using a humanized liver mouse model. The PXB-MOUSE® (PhoenixBio) is a cDNA-uPA / SCID mouse transplanted and engrafted with primary human hepatocytes (PHHs). The mouse has up to 95% PHHengraftment, displays a normal human liver histology and function, has human-specific metabolism and excretion pathways, and expresses human genes, mRNA and proteins, which makes it an ideal model to study human liver function and in the case of the present disclosure the effect of modulating NTCP activity and bile uptake from the bloodstream. The initial study design, the number of animals per study group, the amount of RM122067 and the amount of AG1856-GalNAc as well as the regimen in which these compounds were provided (SC in all cases) are shown in Fig. 12. The figure also shows the day of necropsy. In the case of an early death, the regimen in that mouse could not be completed. Generally, mice were sacrificed two days after receiving the last dose, unless indicated differently. Livers were dissected and RNA editing was determined in liver samples on the day of necropsy, following the general procedures as outlined above. Human-specific primers and probes were used for the ddPCR, shown in Table 1. All cDNA samples from the mice depicted in Fig. 12 passed the QC test for sufficient target mRNA copy numbers.

[0264] Fig. 12 shows, in the last column, the percentage editing of the human SLC10A1 C.203A target site in the humanized mice.

[0265] Groups 1 and 2 served as negative controls for editing (Group 1 : PBS; Group 2: mix of PBS + AG1856-GalNAc). One early death occurred in Group 1 (PXB_106) and one early death occurred in Group 2 (PXB_202). In each case, the livers were isolated for assessment of RNA editing on the day of necropsy. The editing in Groups 1 and 2 varied between 0.07 and 0.20 %, which is considered background.

[0266] Group 3 received (in most cases) a mix of 20 mg / kg RM 122067 and 0.3 or 1 mg / kg AG1856-GalNAc on the dosing days, except for PXB_303, PXB_304, PXB_306, PXB_309, PXB_311 , and PXB_312 that were dosed with 20 mg / kg RM 122067 on days 57, 71 , and 85 and that received 24 hr (on days 58, 72, and 86) a separate dose of 1 mg / kg AG 1856- GalNAc. Interestingly, this subsequent dosing (first the AON and 24 hr later the saponin conjugate compound) resulted in higher editing two days after the final AG1856-GalNAc administration: compare the average of PXB_303 and PXB_304 (35.73%) with the average of the comparable subjects PXB_301 and PXB_302 (24.62%). The same appears to be the case with the average of PXB_309, PXB_311 , and PXB_312 (32.34%) when compared to the average of PXB_307 and PXB_308 (16.98%), which was almost 50% lower in editing. This is a strong suggestion that it is preferred to administer the AG1856-GalNAc compound after the administration of RM 122067, preferably 24 hr later. It should be noted that more timepoints may provide more information about the exact time that the editing percentage peaks after (co-) administration of RM 122067 + / - AG1856-GalNAc, and that the percentages found in this study are indicative of a bell-shaped curve in the editing percentages. Without wishing to be bound by theory, from the studies in NHPs, discussed above, it appeared thatafter SC administration of RM 122067, the compound distributes very rapidly from the plasma to the liver, generally within 0.5 to 1 day, whereas high levels of the AON remain present in the liver to significant high and datable levels up to at least 126 days (exposure data not shown). This suggests that the AG1856-GalNAc compound, when delivered subsequently to the administration of the AON, can trigger the release of AONs that are entrapped in endosomes after their entry into hepatocytes and that such release preferably follows an earlier administration of the oligonucleotide compound. The exact timing of such administrations and sustained release from endosomes can be investigated in larger cohorts with multiple timepoints. However, such either requires the sacrifice of mice at the time of assessment, or multiple subsequent liver biopsies in larger study subjects.

[0267] Groups 4 and 5 only differ in the amount of RM122067 that was administered; all animals in these two groups received 0.3 mg / kg AG1856-GalNAc in admixture with the AON on the dosing days. No apparent difference in editing between the 6 mg / kg RM122067 (Group 4) and the 20 mg mg / kg (Group 5) subjects was observed, ranging from 5.15% and 10.51% editing on day 87 (PXB_401 and PXB_504, respectively) to 11.74% and 10.34% editing on day 97 (PXB_405 and PXB_505, respectively). The highest percentage editing was observed at early determination after the final dose (day 87 and day 97), which appears to decline at later stages (day 107 and day 117), for both groups.

[0268] Group 6 only received 50 mg / kg RM 122067 without an administration of the saponin conjugate compound. Four animals received 4 doses, whereas eight animals received 13 doses on the indicated dosing days. Necropsy was at different timepoints thereafter. The results clearly indicate that high editing percentages were achieved in the absence of AG1856-GalNAc (albeit not as high as observed in Group 3), with such repeated dosing. Editing was highest at day 97 (2 days after final dosing) in PXB_605, PXB_606, and PXB_607, with an average of 16.52%. This declined somewhat to 10.52% and 12.83% at day 107 (PXB_609) and day 106 (PXB_610), respectively.

[0269] Group 7, 8, and 9 received a single dose of 6 mg / kg RM 122067 in an admixture with 0.15, 0.3, or 0.6 mg / kg AG1856-GalNAc on day 1 as indicated in Fig. 12. Editing was assessed on day 3 or day 13, but editing levels remained relatively low after this single dose, which is indicative that repeated dosing (Groups 3, 4, 5, and 6) is preferred to achieve high editing levels.

[0270] Group 10 received a single dose of 50 mg / kg RM 122067 on day 1 and editing reached an average of 3.00% on day 3 (PXB_1001 , PXB_1002, and PXB_1003) and 3.73% on day 13 (PXB_1004, PXB_1005, and PXB_1006), which suggests that in this time frame a peak editing level is present, which is still significant over background (Group 1 and 2), but that repeated dosing can further boost editing (Group 6).

[0271] In conclusion, the data in the PXB-MICE® clearly show that significant and sustained editing can be achieved in a (humanized) liver using RM 122067, which is a tri-antennary GalNAc moiety (of formula (III)) conjugated to RM 120215 (SEQ ID NO:50) at the 3’ terminus, even in the absence of an endosomal release agent, and suggests a clinically relevant human SLC10A1 C.203A editing to inhibit the encoded NTCP protein and thereby limit bile acid uptake into the liver to treat human disease that are related to bile acid accumulation in the liver, such as cholestasis, primary sclerosing cholangitis (PSC), biliary atresia (BA), and liver cirrhosis, ultimately leading to liver failure, and potentially liver cancer (hepatocellular carcinoma (HCC)) related to bile acid accumulation.

Claims

1. CLAIMS1. An antisense oligonucleotide (AON) capable of recruiting an endogenous ADAR enzyme in a human cell, preferably a liver cell, more preferably an hepatocyte, after the AON has formed a double-stranded complex with a region of a pre-mRNA or mRNA transcript molecule of the human SLC10A1 gene encoding the Na7Taurocholate Co-transporting Polypeptide (NTCP), wherein the region comprises the CAG codon for glutamine (Q) at position 68 of the NTCP protein, and wherein the deamination of the adenosine in the CAG codon by the recruited ADAR enzyme, results in a change of the amino acid to an arginine (R), and wherein the AON is according to structural formula (I):wherein R is a hydrogen atom or an / V-Acetylgalactosamine (GalNAc) moiety.

2. An AON according to claim 1, wherein R is a GalNAc moiety, and the GalNAc moiety is according to structural formula (II):

3. An AON according to claim 1, wherein R is a GalNAc moiety, and the GalNAc moiety is according to structural formula (III):

4. A pharmaceutical composition comprising an AON according to any one of claims 1 to 3, and a pharmaceutically acceptable carrier.

5. An AON according to any one of claims 1 to 3, for use in the treatment of a disease caused by bile accumulation in the liver, such as cholestasis, primary sclerosing cholangitis (PSC), biliary atresia (BA), and liver cirrhosis.

6. An AON according to any one of claims 1 to 3, for use in the treatment of a hepatocellular carcinoma (HOC) in a human subject in need thereof.

767. Use of an AON according to any one of claims 1 to 3 in the manufacture of a medicament for the treatment of a disease caused by bile accumulation in the liver, such as cholestasis, PSC, BA, and / or liver cirrhosis.

8. Use of an AON according to any one of claims 1 to 3 in the manufacture of a medicament for the treatment of HOC.

9. An in vitro, ex vivo, or in vivo method of deaminating a target adenosine present in a region of a human SLC10A 1 pre-mRNA or mRNA transcript molecule, to an inosine, in a human cell, preferably a hepatocyte, wherein the target adenosine is in a CAG codon encoding glutamine (Q) at position 68 of the NTCP protein encoded by the SLC10A 1 mRNA, the method comprising the step of: administering an AON to the cell, wherein the AON is sufficiently complementary to the region of the human SLC10A1 pre-mRNA or mRNA to hybridize under physiological conditions to the pre-mRNA or mRNA, wherein the AON forms a doublestranded complex with the region of the SLC10A1 pre-mRNA or mRNA transcript molecule comprising the target adenosine, wherein the double-stranded complex recruits an endogenous ADAR enzyme, wherein the endogenous ADAR enzyme deaminates the target adenosine in the CAG codon to an inosine, wherein the deamination results in a change of the encoded amino acid to an arginine (R), wherein the AON has a chemical structure according to formula (I), and wherein the AON is conjugated at its 3’ terminus to a tri-antennary GalNAc moiety according to formula (III) to form an AON-GalNAc conjugate compound.

10. A method according to claim 9, further comprising the step of simultaneously, before, or after administering the AON-GalNAc conjugate compound, administering to the cell an AG1856-GalNAc conjugate compound, wherein the AG1856-GalNAc conjugate compound has the chemical structure of formula (X).

11. A method of treating, ameliorating, or slowing down the progression of a disease caused by bile accumulation in the liver, such as cholestasis, PSC, BA, and / or liver cirrhosis, in a human subject in need thereof, the method comprising the step of administering to said subject an AON that has a chemical structure according to formula (I), wherein the AON is conjugated at its 3’ terminus to a tri-antennary GalNAc moiety according to formula (III), wherein the AON is sufficiently complementary to a region of a human SLC10A 1 pre-mRNA or mRNA transcript molecule to hybridize under physiological conditions to the pre-mRNA or mRNA transcript molecule in a cell, preferably a liver cell, in the human subject, wherein the region comprises a target adenosine that is in a CAG codon encoding glutamine (Q) at position 68 of the NTCP protein encoded by the SLC10A 1 mRNA transcript molecule, wherein after administering the AON, the AON forms a double-stranded complex with the region of the77SLC10A 1 pre-mRNA or mRNA transcript molecule in the cell, wherein the double-stranded complex recruits an endogenous ADAR enzyme naturally present in the cell, wherein the endogenous ADAR enzyme deaminates the target adenosine to an inosine, and wherein the deamination of the target adenosine to an inosine results in a change of the encoded amino acid to an arginine (R) at position 68 of the NTCP protein, thereby treating, ameliorating, or slowing down the progression of the disease in the subject.

12. A method of treating, ameliorating, or slowing down the progression of HCC in a human subject in need thereof, the method comprising the step of administering to said subject an AON that has a chemical structure according to formula (I), wherein the AON is conjugated at its 3’ terminus to a tri-antennary GalNAc moiety according to formula (III), wherein the AON is sufficiently complementary to a region of a human SLC10A 1 pre-mRNA or mRNA transcript molecule to hybridize under physiological conditions to the pre-mRNA or mRNA transcript molecule in a cell, preferably a liver cell, in the human subject, wherein the region comprises a target adenosine that is in a CAG codon encoding glutamine (Q) at position 68 of the NTCP protein encoded by the SLC10A 1 mRNA transcript molecule, wherein after administering the AON, the AON forms a double-stranded complex with the region of the SLC10A1 pre-mRNA or mRNA transcript molecule in the cell, wherein the double-stranded complex recruits an endogenous ADAR enzyme naturally present in the cell, wherein the endogenous ADAR enzyme deaminates the target adenosine to an inosine, and wherein the deamination of the target adenosine to an inosine results in a change of the encoded amino acid to an arginine (R) at position 68 of the NTCP protein, thereby treating, ameliorating, or slowing down the progression of the HCC in the subject.

13. A method according to claim 11 or 12, further comprising the step of simultaneously, before, or after administering the AON-GalNAc conjugate compound, administering to the subject an AG1856-GalNAc conjugate compound, wherein the AG1856-GalNAc conjugate compound has the chemical structure of formula (X).

14. A method according to claim 12 or 13, wherein the AON is administered to the human subject in a dose of 1 to 50 mg / kg, preferably 1 to 20 mg / kg, more preferably 1 to 10 mg / kg, even more preferably in a dose of 2, 3, 4, 5, 6, 7, 8, or 9 mg / kg.

15. A method according to any one of claims 12 to 14, wherein the human subject receives the AON dose in a repeated regimen, preferably once every two weeks, more preferably once a week.7816. A kit-of-parts for use in the treatment of a disease caused by bile accumulation in the liver, such as cholestasis, PSC, BA, liver cirrhosis, and / or liver cancer such as HCC, said kit- of-parts comprising: a part (A) comprising an AON according to any one of claims 1 to 3; and a part (B) comprising an AG1856 saponin that is conjugated to a GalNAc moiety according to formula (III), wherein part (A) and part (B) can be mixed to be administered simultaneously, or wherein part (A) and part (B) are administered subsequently with part (A) being administered first and part (B) being administered thereafter, preferably 24 hours later.79

Images