WO2023066881A1 - Inhibition of map3k15 for treating and preventing diabetes - Google Patents

Abstract

Provided herein are methods of inhibiting MAP3K15 and inhibitors of MAP3K15 useful for, e.g., for the treatment and prevention of diabetes.

Description

INHIBITION OF MAP3K15 FOR TREATING AND PREVENTING DIABETES

BACKGROUND

[0001] Diabetes mellitus is a worldwide health concern projected to affect 700 million people by 2045. It is currently the leading cause of micro- and macrovascular disease, including kidney failure, blindness, heart disease, and lower limb amputations. Characterized by elevated levels of blood glucose, diabetes mellitus is generally categorized into type 1 diabetes mellitus (T1DM), type 2 diabetes mellitus (T2DM), and other rarer forms. T1DM is caused by autoimmune destruction of insulin-producing pancreatic P cells, while T2DM is primarily caused by peripheral insulin resistance. Both types of diabetes eventually lead to progressive loss of pancreatic P cells and deficient insulin secretion.

[0002] Genome wide association studies (GWAS) have implicated over 60 loci in T1DM and hundreds of loci in T2DM. With the exception of a few GWAS loci for T1DM or T2DM that map to protein-coding genes (e.g., SLC30A8, PAM), the remaining reside in non-coding regions of the genome, making it challenging to characterize the underlying causal gene. Moreover, the modest risk conferred by most GWAS loci further limits their potential as therapeutic targets. The growing availability of whole-exome sequences in large populationscale biobanks offers unprecedented opportunities to explore the association of proteincoding variants that have demonstrably large effects on human traits. In particular, identifying loss-of-function alleles that protect against disease is of significant interest since they can provide in-vivo identification and validation of therapeutic targets.

[0003] There is a need in the art for an effective treatment of T1DM and T2DM. In addition to the treatment of the symptoms of T1DM and T2DM, there is a need for a treatment that would change the progression of T1DM and T2DM.

BRIEF SUMMARY

[0004] Here, a multi-ethnic exome-sequencing association study for T1DM and T2DM in -412,000 participants from the UK Biobank (UKB) is reported and was used to identify MAP3K15 as a target for the treatment of diabetes. As described herein, complete loss of the X chromosome gene MAP3K15 was found to reduce the odds of developing type 2 diabetes mellitus by 40% and is associated with a significant decrease in glucose and haemoglobin A1C levels. The risk of developing type 1 diabetes is also reduced in subjects having complete loss of the MAP3K15 gene. Notably, the protective effect against development of T2DM is independent of a subjects’ body mass index (BMI). The findings were replicated in two ethnically diverse population-based studies: Mexico City Project (MCP) and FinnGen. Furthermore, the loss of MAP3K15 was not associated with any apparent adverse phenotypes, indicating a lack of clinical safety concerns from therapeutic inhibition of this target. MAP3K15 is thus indicated as a novel target for therapeutic treatment of diabetes. Accordingly, provided herein are methods of inhibiting MAP3K15, e.g., for the treatment and prevention of diabetes.

[0005] MAP3K15 is also known as Apoptosis Signal-regulating Kinase 3 (ASK3). It may well be that the loss of function mutations of MAP3K15 that are observed to be protective against the development of diabetes in the UKBB cohort and elsewhere as decribed herein, and in particular type 2 diabetes, are protective against the development of diabetes due to their ability to promote the survival of pancreatic beta cells through elimination of apoptotic stimuli provided by MAP3K15 that would otherwise trigger beta cell death. As is well known in the art, pancreatic beta cells play a pivotal role on regulating blood glucose levels through production and secretion of insulin and maintenance of beta cell mass will be protective against disease progression. It follows that MAP3K15 inhibitors may be particularly suitable for the treatment of subjects with early stage diabetes. This is because the inhibition of MAP3K15, through promotion of beta cell survival and preservation of beta cell mass, could stop the progression of diabetes over a time course involving gradually reduced beta cell function to an insulin dependent state. MAP3K15 inhibitors, unlike all existing therapies for diabetes, offer the potential to prevent disease progression and to minimize the therapeutic burden on subjects.

[0006] Accordingly, provided herein are methods of treating or preventing diabetes mellitus in a subject comprising administering an inhibitor of MAP3K15 to the subject.

[0007] Also provided herein are methods of decreasing hemoglobin Ale (HbAlc) in a subject in need thereof comprising administering an inhibitor of MAP3K15 to the subject. Also provided herein are methods of preventing pancreatic cell death in a subject in need thereof comprising administering an inhibitor of MAP3K15 to the subject. In some aspects, the subject has diabetes.

[0008] In some aspects of the methods provided herein, the diabetes is type 2 diabetes. In some aspects, the diabetes is type 1 diabetes. In some aspects, the diabetes is type 1 diabetes and the subject has a C-peptide level of at least 0.2 nmol/L prior to the administration.

[0009] Also provided herein are methods of treating or preventing a disease or condition caused or characterized by high blood sugar in a subject, the method comprising administering an inhibitor of MAP3K15 to the subject.

[0010] In some aspects of the methods provided herein, the inhibitor is a small molecule inhibitor, a peptide inhibitor, or an aptamer.

[0011] In some aspects, the inhibitor is a small molecule inhibitor. In some aspects, the small molecule inhibitor is staurosporine, lestaurtinib, TAE-684, ruxolitinib, sunitinib, nintedanib, crizotinib, or fostamatinib.

[0012] In some aspects, the inhibitor comprises a nucleic acid molecule and an enzymatic protein, wherein the nucleic acid molecule is a guide RNA (gRNA) molecule and the enzymatic protein is a Cas protein or Cas ortholog.

[0013] In some aspects, the inhibitor is a polynucleotide.

[0014] In some aspects, the inhibitor is a small interference RNA (siRNA), a short hairpin

RNA (shRNA), a microRNA (miRNA), an antagomiRNA, or an antisense RNA. In some aspects, the inhibitor comprises the nucleotide sequence CTCTCTCTTTGAAGGACATGGTAA (SEQ ID NO:1), the nucleotide sequence CGGAUUUCAGGAUGCCGUAAAUAAA (SEQ ID NOG), or the nucleotide sequence CACCGAAGAGCAGUGCAGUAGAUUU (SEQ ID NOG). In some aspects, the inhibitor comprises at least two nucleotide sequences selected from a group consisting of the nucleotide sequences: GCTCTCTCTTTGAAGGACATGGTAA (SEQ ID NO:1), CGGAUUUCAGGAUGCCGUAAAUAAA (SEQ ID NOG), and CACCGAAGAGCAGUGCAGUAGAUUU (SEQ ID NOG).

[0015] In some aspects, the inhibitor is an enzymatic protein that binds to a gene encoding MAP3K15. In some aspects, the enzymatic protein is a Transcription activator-like effector nuclease (TALEN), a zinc-finger nuclease, or a meganuclease.

[0016] In some aspects, the inhibitor is a proteolysis targeting chimera (PROTAC). [0017] In some aspects, the inhibitor binds to the ATP-binding domain of MAP3K15.

[0018] In some aspects, the inhibitor is a non-ATP competitive inhibitor.

[0019] In some aspects, the inhibitor binds to the metal-binding domain of MAP3K15.

[0020] In some aspects, the inhibitor inhibits the serine-threonine kinase activity of

MAP3K15. In some aspects, the inhibitor inhibits the transferase activity of MAP3K15.

[0021] In some aspects, the inhibitor decreases binding of MAP3K15 to ATP. In some aspects, the inhibitor decreases binding of MAP3K15 to magnesium.

[0022] In some aspects, the inhibitor decreases phosphorylation of MAP3K15 Thr-812. In some aspects, the inhibitor increases phosphorylation of MAP3K15 Ser-924 and/or Ser-994. In some aspects, the inhibitor decreases homooligomerisation of MAP3K15.

[0023] In some aspects, the inhibitor has an IC50 of less than 50 nM, optionally wherein the inhibitor has an IC50 of less than 10 nM. In some aspects, the concentration of the inhibitor required to cause 50% inhibition of MAP3K15 is 10-fold lower than that required to cause 50% inhibition of ASK1. In some aspects, the concentration of the inhibitor required to cause 50% inhibition of MAP3K15 is 10-fold lower than that required to cause 50% inhibition of ASK2.

[0024] In some aspects, the the inhibitor is a reversible inhibitor. In some aspects, the inhibitor is an irreversible inhibitor.

[0025] In some aspects, the administration reduces blood sugar, for example as reflected by measured HbAlc levels, in the subject relative to the level measured prior to treatment. In some aspects, the administration reduces HbAlc by at least 10 mmol/mol. In some aspects, the administration reduces HbAlc by at least 20 mmol/mol.

[0026] In some aspects, the HbAlc level post treatment is 48 mmol/mol or below. In some aspects, the HbAlc level post treatment is between 20 mmol/mol and 42 mmol/mol.

[0027] In some aspects, the measured HbAlc level in the subject prior to treatment is at least 48 mmol/mol.

[0028] In some aspects, the HbAlc in the subject prior to treatment at least 63.8 mmol/mol.

[0029] In some aspects, the administration reduces HbAlc to 48 mmol/mol or below, for example from 20 mmol/mol to 42 mmol/mol.

[0030] In some aspects, the the administration is oral. In some aspects, the administration is intravenous. [0031] In some aspects, the method further comprises administering an additional therapeutic agent. In some aspects, the additional therapeutic agent is a biguanidine, a thiazolidinedione, a sylfonylureas, a meglitinides, a sodium-glucose cotransporter 2 inhibitor (SGLT2i), a dipeptidyl peptidase IV (DPP-IV) inhibitor, and/or a glucagon-like peptide- 1 (GLP-1). In some aspects, the additional therapeutic agent is metformin, a dipeptidyl peptidase IV (DPP- IV) inhibitor, and/or a sodium-glucose cotransporter 2 inhibitor (SGLT2i). In some aspects, the administration is in combination with diet and exercise.

[0032] In some aspects, the subject is male. In some aspects, the subject is female. In some aspects, the female carries a single loss of function mutation in the gene encoding MAP3K15. In some aspects, the mutation is a truncation. In some aspects, the subject does not contain a loss of function mutation in the gene encoding MAP3K15.

[0033] In some aspects, the subject has a BMI of 30 or more prior to the administration. In some aspects, the subject has a BMI of at least 25 and less than 30 prior to the administration. In some aspects, the subject has a BMI of at least 18.5 and less than 25 prior to the administration. In some aspects, the subject has a BMI of less than 25, less than 20, or less than 18.5 prior to the administration.

[0034] In some aspects, the administration reduces MAP3K15 activity in the subject by at least 40%, optionally wherein the administration reduces MAP3K15 activity in the subject by at least 50%, at least 60%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, or at least 95%.

[0035] In some aspects, the subject has not achieved significant HbAlc reduction following treatment with metformin. In some aspects, the the subject has an HbAlc level equal to or greater than 63.9 mmol/mol after treatment with metformin for at least 3 months.

[0036] In some aspects, the subject is a mouse. In some aspects, the subject is human.

[0037] Also provided herein are methods for screening for a substance for treating or prophylaxis of diabetes mellitus comprising contacting MAP3K15 with a test substance and determining if the test substance inhibits MAP3K15.

[0038] Also provided herein are inhibitors of MAP3K15 for use in the treatment or prophylaxis of diabetes mellitus. Also provided herein are inhibitors of MAP3K15 for use in treatment or prophylaxis of type 2 diabetes mellitus. Also provided herein are inhibitors of MAP3K15 inhibitor for use in treatment or prophylaxis of type 1 diabetes mellitus. [0039] In some aspects of the inhibitors for use herein, the inhibitor is an ATP competitive inhibitor. In some aspects, the inhibitor decreases the binding of magnesium to MAP3K15. In some aspects, the inhibitor is a reversible inhibitor of MAP3K15. In some aspects, the inhibitor is an irreversible inhibitor of MAP3K15. In some aspects, the inhibitor is a selective inhibitor of MAP3K15.

[0040] In some aspects, the concentration of the inhibitor required to cause 50% inhibition of MAP3K15 is 10-fold lower than that required to cause 50% inhibition of ASK1. In some aspects, the concentration of the MAP3K15 inhibitor required to cause 50% inhibition of MAP3K15 is 10-fold lower than that required to cause 50% inhibition of ASK2.

[0041] In some aspects, the inhibitor is a proteolysis targeting chimera (PROTAC). In some aspects, the inhibitor is an antisense oligonucleotide (ASO) that reduces expression of MAP3K15.

[0042] In some aspects, the use results in a reduction of a subject’s HbAlc as measured in a sample obtained from the subject during treatment relative to that measured in a sample obtained from the subject prior to treatment with the inhibitor. In some aspects, the use results in a reduction of HbAlc of > 10 mmol/mol, for example > 20 mmol/mmol. In some aspects, the use results in an HbAlc level post treatment of 48 mmol/mol or below. In some aspects, the use results in an HbAlc level post treatment of between 20 mmol/mol and 42 mmol/mol.

[0043] In some aspects, the measured HbAlc level in the subject prior to treatment is at least 48 mmol/mol. In some aspects, the measured HbAlc level in the subject prior to treatment is at least 63.8 mmol/mol.

[0044] In some aspects, the use results in HbAlc of 48 mmol/mol or below post treatment. In some aspects, the use results in HbAlc 20 mmol/mol to 42 mmol/mol post treatment.

[0045] In some aspects, the use is in combination with a further therapeutic agent. In some aspects, the further therapeutic agent is a biguanidine, a thiazolidinedione, a sylfonylureas, a meglitinides, a sodium-glucose cotransporter 2 inhibitor (SGLT2i), a dipeptidyl peptidase IV (DPP-IV) inhibitor, and/or a glucagon-like peptide- 1 (GLP-1). In some aspects, the further therapeutic agent is metformin, a dipeptidyl peptidase IV (DPP-IV) inhibitor, and/or a sodium-glucose cotransporter 2 inhibitor (SGLT2i). In some aspects, the further therapeutic agent is metformin. [0046] In some aspects, the use is for a subject that has not achieved significant HbAlc reduction following treatment with metformin. In some aspects, the use is for a subject with a HbAlc level equal to or greater than 63.9 mmol/mol after treatment with metformin for at least 3 months. In some aspects, the use is for a subject with a BMI of 30 or more prior to the commencement of treatment with the MAP3K15 inhibitor. In some aspects, the use is for a subject with a BMI of less than 25 prior to the commencement of treatment with the MAP3K15 inhibitor. In some aspects, the use is for a subject with a BMI of less than 20 prior to the commencement of treatment with the MAP3K15 inhibitor.

[0047] In some aspects, the use is for a male subject. In some aspects, the use is for a female subject. In some aspects, the use is for a female subject with a single loss of function mutation in the gene encoding MAP3K15

[0048] Also provided herein are methods of identifying a subject that would benefit from treatment with a MAP3K15 inhibitor comprising: i) sequencing a biological sample obtained from a subject; ii) analyzing the sequence to determine whether there is a loss of function mutation of MAP3K15; and iii) based on the result of ii), indicating whether the subject is indicated for treatment with a MAP3K15 inhibitor.

[0049] In some aspects, the methods comprising administering a MAP3K15 inhibitor to a subject as provided herein further comprise identifying the subject using a method of identifying provided herein.

[0050] In some aspects, provided herein are methods of treating a male subject that would benefit from treatment with a MAP3K15 inhibitor, the method comprising: i) sequencing a biological sample obtained from a male subject; ii) analyzing the sequence to determine whether there is a loss of function mutation of MAP3K15; and iii) administering an inhibitor of MAP3K15 to the subject if there is not a loss of function of MAP3K15. In some aspects, the administration is according to any of the methods provided herein.

[0051] In some aspects, provided herein are methods of treating a female subject that would benefit from treatment with a MAP3K15 inhibitor comprising: i) sequencing a biological sample obtained from a female subject; ii) analyzing the sequence to determine whether there is a loss of function mutation of MAP3K15; and iii) administering an inhibitor of MAP3K15 to the subject if there is not a loss of function of MAP3K15 or if there is only a loss of function of one MAP3K15-encoding gene. In some aspects, the administration is according to any of the methods provided herein.

BRIEF DESCRIPTION OF THE DRAWINGS/FIGURES

[0052] Fig. 1 shows the effect sizes of genetic associations with diabetes and related traits.

(A) Odds ratios and allele frequencies of gene-level (collapsing) and variant-level (ExWAS) associations (p<lxl0^-7) with diabetes diagnoses. (B) Effect sizes and allele frequencies of gene-level (collapsing) and variant-level (ExWAS) associations (p<lxl0^-7) with haemoglobin A1C (HbAlC). (C) Odds ratios of partial loss (i.e., heterozygous female carriers) and complete loss (i.e., hemizygous male carriers) of MAP3K15 for diabetes and hypertension diagnoses. (D) Effect sizes of partial loss (i.e., heterozygous female carriers) and complete loss (i.e., hemizygous male carriers) of MAP3K15 on various cardiovascular and metabolic traits. (E) Lollipop plot depicting MAP3K15 PTVs observed in individuals of European ancestry. The two most common PTVs are annotated. The y-axis was capped at 40.

[0053] Fig. 2 shows the expression of MAP3K15. (A) Expression of MAP3K15 in human tissues contained in the GTEx database. TPM = transcripts per million. Only tissues with a median TPM > 0.1 were included. (B) MAP3K15 expression in major classes of human pancreatic cells derived from a previously published single-cell RNA- sequencing dataset. (C) Volcano plot depicting differential expression of mouse insulinoma cell line cell lines stably expressing two MODY-associated variants and a functionally impaired control mutation in Nkx6-1.

[0054] Fig. 3 shows MAP3K15 biomarker and disease signatures. (A) Genes with similar biomarker profiles as MAP3K15 in the UK Biobank, derived from Gene-SCOUT. (B) Linear regression coefficients associating HbAlC and glucose levels with rare variant genetic architectures from genes in panel A (genes are sorted from top to bottom in decreasing order of similarity to MAP3K15). (C) Mantis-ML predictions of MAP3K15 disease associations.

[0055] Fig. 4 shows pre-clinical studies in full body MAP3K15 knock out mice.

[0056] Fig. 5 shows pre-clinical studies in conditional pancreatic specific MAP3K15-

Ins2Cre knock out mice (Map3kl5fl/fl/ins2Cre mice).

[0057] Fig. 6 shows pre-clinical studies in Inducible MAP3K15-CreERT2 knock out mice (MAP3K15 deletion in adult mice). [0058] Fig. 7 shows the distribution of PTVs in the MAP3K15 gene sequence across all ancestries in the UK Biobank. Lollipop plot depicting MAP3K15 PTVs (stop gain and frameshift variants) observed among participants across all ancestries in the UK Biobank. Essential splice variants were not included in the lollipop plot. The two most frequent PTVs have been annotated. Just one individual in the entire cohort carried both PTVs. The y-axis was capped at 40.

[0059] Fig. 8 shows the association signatures with HbAlc and glucose for genes most similar to MAP3K15. Comparison of linear regression coefficients for HbAlc and glucose between MAP3K15 and the genes most “similar” to it, derived from Gene-SCOUT. All collapsing models for which MAP3K15 showed significant associations with HbAlc and glucose have been provided.

DETAILED DESCRIPTION OF THE INVENTION

[0060] The headings provided herein are not limitations of the various aspects described herein, which can be defined by reference to the specification as a whole. It is also to be understood that the terminology used herein is for the purpose of describing particular aspects only, and is not intended to be limiting, since the scope of the present disclosure will be limited only by the appended claims.

Definitions

[0061] For convenience, the meaning of some terms and phrases used in the specification, examples, and appended claims are provided below. Unless stated otherwise, or implicit from context, the following terms and phrases include the meanings provided below. The definitions are provided to aid in describing particular aspects, and are not intended to limit the claimed technology, because the scope of the technology is limited only by the claims. Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this technology belongs. If there is an apparent discrepancy between the usage of a term in the art and its definition provided herein, the definition provided within the specification shall prevail.

[0062] The term “MAP3K15,” also known as apoptosis signal-regulating kinase 3 (ASK3), refers to mitogen-activated protein kinase kinase kinase 15 protein that in humans is encoded by the MAP3K15 gene. The protein encoded by this gene is a member of the mitogen- activated protein kinase (MAPK) family. These family members function in a protein kinase signal transduction cascade, where an activated MAPK kinase kinase (MAP3K) phosphorylates and activates a specific MAPK kinase (MAP2K), which then activates a specific MAPK. This MAP3K protein plays an essential role in apoptotic cell death triggered by cellular stresses. MAP3K15 refers to any native MAP3K15 polypeptide or MAP3K15- encoding polynucleotide. The term “MAP3K15” encompasses “full-length,” unprocessed MAP3K15 polypeptide as well as any forms of MAP3K15 that result from processing within the cell (e.g., removal of the signal peptide). The term also encompasses naturally occurring variants of MAP3K15, e.g., those encoded by splice variants and allelic variants. The MAP3K15 polypeptides described herein can be isolated from a variety of sources, such as from human tissue types or from another source, or prepared by recombinant or synthetic methods. Human MAP3K15 sequences are known and include, for example, the sequences publicly available as UniProt No. Q6ZN16 (including isoforms). As used herein, the term “human MAP3K15 protein” refers to MAP3K15 protein comprising the amino acid sequence set forth in SEQ ID NO:4.

MESGGGNAPAGALGAASESPQCPPPPGVEGAAGPAEPDGAAEGAAGGSGEGESGG GPRRALRAVYVRSESSQGGAAGGPEAGARQCLLRACEAEGAHLTSVPFGELDFGET AVLDAFYDADVAVVDMSDVSRQPSLFYHLGVRESFDMANNVILYHDTDADTALSL KDMVTQKNTASSGNYYFIPYIVTPCADYFCCESDAQRRASEYMQPNWDNILGPLCM PLVDRFISLLKDIHVTSCVYYKETLLNDIRKAREKYQGEELAKELARIKLRMDNTEVL TSDIIINLLLSYRDIQDYDAMVKLVETLEMLPTCDLADQHNIKFHYAFALNRRNSTGD REKALQIMLQVLQSCDHPGPDMFCLCGRIYKDIFLDSDCKDDTSRDSAIEWYRKGFE LQSSLYSGINLAVLLIVAGQQFETSLELRKIGVRLNSLLGRKGSLEKMNNYWDVGQF FSVSMLAHDVGKAVQAAERLFKLKPPVWYLRSLVQNLLLIRRFKKTIIEHSPRQERL NFWLDIIFEATNEVTNGLRFPVLVIEPTKVYQPSYVSINNEAEERTVSLWHVSPTEMK QMHEWNFTASSIKGISLSKFDERCCFLYVHDNSDDFQIYFSTEEQCSRFFSLVKEMIT NTAGSTVELEGETDGDTLEYEYDHDANGERVVLGKGTYGIVYAGRDLSNQVRIAIK EIPERDSRYSQPLHEEIALHKYLKHRNIVQYLGSVSENGYIKIFMEQVPGGSLSALLRS KWGPMKEPTIKFYTKQILEGLKYLHENQIVHRDIKGDNVLVNTYSGVVKISDFGTSK RLAGVNPCTETFTGTLQYMAPEIIDQGPRGYGAPADIWSLGCTIIEMATSKPPFHELG EPQAAMFKVGMFKIHPEIPEALSAEARAFILSCFEPDPHKRATTAELLREGFLRQVNK GKKNRIAFKPSEGPRGVVLALPTQGEPMATSSSEHGSVSPDSDAQPDALFERTRAPR HHLGHLLSVPDESSALEDRGLASSPEDRDQGLFLLRKDSERRAILYKILWEEQNQVA SNLQECVAQSSEELHLSVGHIKQIIGILRDFIRSPEHRVMATTISKLKVDLDFDSSSISQI HLVLFGFQDAVNKILRNHLIRPHWMFAMDNIIRRAVQAAVTILIPELRAHFEPTCETE GVDKDMDEAEEGYPPATGPGQEAQPHQQHLSLQLGELRQETNRLLEHLVEKEREYQ NLLRQTLEQKTQELYHLQLKLKSNCITENPAGPYGQRTDKELIDWLRLQGADAKTIE KIVEEGYTLSDILNEITKEDLRYLRLRGGLLCRLWSAVSQYRRAQEASETKDKA (SEQ ID NO:4)

[0063] The terms “MAP3K” or “MAPKKK” refer to a mitogen activated protein (MAP) kinase kinase kinase, which is a serine/threonine-specific protein kinase which acts upon MAP kinase kinase (MAP2K). Subsequently, MAP2K activates MAP kinase. Several types of MAP3K can exist but are mainly characterized by the MAP kinases they activate.

[0064] The terms “MAP3K15 inhibitor” and “inhibitor of MAP3K15” are used interchangeably and refer to a molecule that reduces the level of MAP3K15 protein and/or inhibits or reduces at least one biological activity of MAP3K15 protein. A MAP3K15 inhibitor can be any class of molecule, including small molecules, inhibitory nucleic acid (e.g., antisense oligonucleotides or siRNA (including shRNA)), a gene -regulating system, peptides, aptamers, proteolysis targeting chimeras (PROTACs), and the like.

[0065] The term “ASK1” refers to apoptosis signal-regulating kinase 1, also known as mitogen-activated protein kinase 5 (MAP3K5). ASK1 is a member of MAP kinase family and is a part of mitogen-activated protein kinase pathway. It activates c-Jun N-terminal kinase (JNK) and p38 mitogen-activated protein kinases in a Raf-independent fashion in response to an array of stresses such as oxidative stress, endoplasmic reticulum stress and calcium influx. ASK1 has been found to be involved in cancer, diabetes, rheumatoid arthritis, cardiovascular and neurodegenerative diseases.

[0066] The term “ASK2” refers to apoptosis signal-regulating kinase 2, also known as mitogen-activated protein kinase 6 (MAP3K6). ASK2 is closely related to ASK1 but can activate the JNK and p38 pathways only by forming a heteromeric complex with ASK1.

[0067] As used herein, the term “gene-regulating system” refers to a protein, nucleic acid, or combination thereof that is capable of modifying an endogenous target nucleotide (e.g., DNA) sequence when introduced into a cell, thereby regulating the expression or function of the encoded gene product. Numerous gene editing systems suitable for use in the methods of the present disclosure are known in the art including, but not limited to, zinc -finger nuclease systems, TALEN systems, and CRISPR/Cas systems.

[0068] The term “antisense oligonucleotide” refers to a single-stranded oligonucleotide comprising 8 to 50 monomeric units and having a nucleobase sequence that permits hybridization to a corresponding segment of a target nucleic acid sequence. An antisense oligonucleotide can comprise natural, non-natural, and/or modified nucleosides and/or internucleoside linkages.

[0069] The term “siRNA” refers to a double-stranded oligonucleotide comprising a first strand comprising 10 to 30 monomeric units and a second strand comprising 10 to 30 monomeric units that is complementary to the first strand, wherein the first strand or second strand has a nucleobase sequence that permits hybridization to a corresponding segment of a target nucleic acid. The first strand and second strand can have 0, 1, 2, or 3 mismatches with respect to one another, and can be linked or contiguous to form a single molecule. In some such aspects when the first strand and second strand are linked or contiguous, the siRNA is referred to as shRNA. A siRNA double-stranded oligonucleotide can comprise natural, nonnatural, and/or modified nucleosides and/or internucleoside linkages.

[0070] The terms “short hairpin RNA” or “shRNA” refer to an artificial RNA molecule with a tight hairpin turn that can be used to silence target gene expression via RNA interference (RNAi).

[0071] The term “RNA interference” or “RNAi” refers to a biological process in which RNA molecules are involved in sequence-specific suppression of gene expression by doublestranded RNA, through translational or transcriptional repression.

[0072] The term “polynucleotide” refers to a biopolymer composed nucleotide monomers covalently bonded in a chain. DNA (deoxyribonucleic acid) and RNA (ribonucleic acid) are examples of polynucleotides with distinct biological function.

[0073] The term “guide RNA” or “gRNA” refers to a piece of RNAs that function as guides for RNA- or DNA-targeting enzymes, which they form complexes with. These enzymes typically delete, insert or otherwise alter the targeted RNA or DNA. [0074] The term “antagomiRNA” as used herein refers to a class of chemically engineered oligonucleotides that prevent other molecules from binding to a desired site on an mRNA molecule. AntagomiRNAs are used to silence endogenous microRNA (miRNA).

[0075] The term “microRNA” or “miRNA” refer to a small single-stranded non-coding RNA molecule (e.g., containing about 22 nucleotides) found in plants, animals, and some viruses, that functions in RNA silencing and post-transcriptional regulation of gene expression. miRNAs resemble the small interfering RNAs (siRNAs) of the RNA interference (RNAi) pathway, except miRNAs derive from regions of RNA transcripts that fold back on themselves to form short hairpins, whereas siRNAs derive from longer regions of doublestranded RNA.

[0076] The term “meganuclease” as used herein refers to endodeoxy-ribonucleases characterized by a large recognition site (e.g., double-stranded DNA sequences of 12 to 40 base pairs). As a result this site generally occurs only once in any given genome. Meganucleases are considered to be the most specific naturally occurring restriction enzymes. Meganucleases are "molecular DNA scissors" that can be used to replace, eliminate or modify sequences in a highly targeted way. By modifying their recognition sequence through protein engineering, the targeted sequence can be changed. Meganucleases are used to modify all genome types, whether bacterial, plant or animal.

[0077] The terms “polypeptide,” “peptide,” and “protein” are used interchangeably to refer to a polymer of amino acid residues, and are not limited to a minimum length. Such polymers of amino acid residues can contain natural and/or non-natural amino acid residues, and include, but are not limited to, peptides, oligopeptides, dimers, trimers, and multimers of amino acid residues. Both full-length proteins and fragments thereof are encompassed by the definition. The terms also include post-expression modifications of the polypeptide, for example, glycosylation, sialylation, acetylation, phosphorylation, and the like.

[0078] The term “specifically binds” to a protein or domain of a protein is a term that is well understood in the art, and methods to determine such specific binding are also well known in the art. A molecule is said to exhibit “specific binding” or “preferential binding” if it reacts or associates more frequently, more rapidly, with greater duration and/or with greater affinity with a particular protein or domain of a protein than it does with alternative proteins or domains. It should be understood that a molecule that specifically or preferentially binds to a first protein or domain may or may not specifically or preferentially bind to a second protein or domain. As such, “specific binding” or “preferential binding” does not necessarily require (although it can include) exclusive binding. Generally, but not necessarily, reference to binding means preferential binding. For example, a MAP3K15 inhibitor that specifically binds to MAP3K15 may not bind to other family members, such as ASK1 protein or ASK2 protein or may bind to other family members, such as ASK1 protein or ASK2 protein, with a reduced affinity as compared to binding to MAP3K15.

[0079] The terms “reduction” or “reduce” or “inhibition” or “inhibit” refer to a decrease or cessation of any phenotypic characteristic or to the decrease or cessation in the incidence, degree, or likelihood of that characteristic. To “reduce” or “inhibit” is to decrease, reduce or arrest an activity, function, and/or amount as compared to a reference. “Decrease” or “reduce” refers to a decrease or a reduction in a particular value of at least 5%, for example, a 5%, 6%, 7%, 8%, 9%, 10%, 15%, 20%, 25%, 30%, 35%, 40%, 45%, 50%, 55%, 60%, 65%, 70%, 75%, 80%, 85%, 90%, 95%, 99%, or 100% decrease as compared to a reference value. A decrease or reduction in a particular value may also be represented as a fold-change in the value compared to a reference value, for example, at least a 1.1, 1.2, 1.3, 1.4, 1.5, 1.6, 1.7, 1.8, 1.9, 2, 3, 4, 5, 6, 7, 8, 9, 10, 15, 20, 30, 40, 50, 60, 70, 80, 90, 100, 200, 500, 1000-fold, or more, decrease as compared to a reference value. In some aspects, by “reduce” or “inhibit” is meant the ability to cause an overall decrease of 20% or greater. In some aspects, by “reduce” or “inhibit” is meant the ability to cause an overall decrease of 50% or greater. In some aspects, by “reduce” or “inhibit” is meant the ability to cause an overall decrease of 75%, 85%, 90%, 95%, or greater. In some aspects, the amount noted above is inhibited or decreased over a period of time, relative to a control over the same period of time. In the context of MAP3K15, inhibition of MAP3K15 includes, but is not limited to, reducing of the odds of developing diabetes in a subject, and/or decreasing glucose and haemoglobin A1C levels.

[0080] The term “hemoglobin Ale,” “HbAlc,” or “A1C” refers to a form of hemoglobin (Hb) that is chemically linked to a sugar. Most monosaccharides, including glucose, galactose and fructose, spontaneously (i.e. non-enzymatically) bond with hemoglobin, when present in the bloodstream of humans. The formation of the sugar -hemoglobin linkage indicates the presence of excessive sugar in the bloodstream, often indicative of diabetes. A1C is of particular interest because it is easy to detect. The process by which sugars attach to hemoglobin is called glycation. HbAlc is a measure of the beta-N-1 -deoxy fructosyl component of hemoglobin. Hemoglobin Ale is measured primarily to determine the three - month average blood sugar level and can be used as a diagnostic test for diabetes mellitus and as an assessment test for glycemic control in people with diabetes. The test is limited to a three-month average because the average lifespan of a red blood cell is four months. Since individual red blood cells have varying lifespans, the test is used as a limited measure of three months. Normal levels of glucose produce a normal amount of glycated hemoglobin. As the average amount of plasma glucose increases, the fraction of glycated hemoglobin increases in a predictable way. In diabetes, higher amounts of glycated hemoglobin, indicating poorer control of blood glucose levels, have been associated with cardiovascular disease, nephropathy, neuropathy, and retinopathy. Typically, a baseline level of HbAlc that indicates T2DM is > 48mmol/mol and a level of HbAlc for prediabetes is 42-28 mmol/mol in patient pretreatment.

[0081] The term “significant HbAlc reduction,” as used herein, refers to a reduction in HbAlc levels by from about 10 mmol/mol to about 20 mmol/mol from the baseline level of HbAlC in a patient pretreatment. In some instances, a significant reduction is the reduction of HbAlc to 48 mmol/mol or less. In some instances, a significant reduction is the reduction of HbAlc to 42 mmol/mol or less. In some instances, a significant reduction is the reduction of HbAlc to 20 to 42 mmol/mol. In some instances, a significant reduction of HbAlc is the reduction of HbAlc to at least to prediabetic levels of 42 to 48 mmol/mol.

[0082] A “reference” as used herein, refers to any sample, standard, level, or value that is used for comparison purposes. A reference can be obtained from a healthy and/or nondiseased sample, or can be a previously determined level or value (e.g., a level or value prior to administration). In some aspects, the reference level or value has been previously determined from a healthy and/or non-diseased and/or untreated sample or collection of healthy and/or non-diseased and/or untreated samples. In some examples, a reference can be obtained from an untreated sample. In some examples, a reference is obtained from a nondiseased or non-treated sample of a subject individual. In some examples, a reference is obtained from one or more healthy individuals who are not the subject or patient. A reference for a cell, in some aspects, can be the same cell type that has not been contacted with a test substance.

[0083] The term “irreversible inhibitor” refers to an enzyme inhibitor that reacts with the enzyme and changes it chemically (e.g. via covalent bond formation). Irreversible inhibitors modify key amino acid residues needed for enzymatic activity.

[0084] The term “reversible inhibitor” refers to an enzyme inhibitor that binds non- covalently and different types of inhibition are produced depending on whether these inhibitors bind to the enzyme, the enzyme-substrate complex, or both.

[0085] The term “body mass index” or “BMI” is a value derived from the mass (weight) and height of a person. The BMI is defined as the body mass divided by the square of the body height, and is expressed in units of kg/m², resulting from mass in kilograms and height in metres. The BMI is used to categorize a person as underweight, normal weight, overweight, or obese based on tissue mass (muscle, fat, and bone) and height. Typically, major adult BMI classifications are underweight (under 18.5 kg/m²), normal weight (18.5 to 24.9 kg/m²), overweight (25 to 29.9 kg/m²), and obese (30 kg/m² or more).

[0086] The term ’’DPP-IV” refers to dipeptidyl peptidase-4 (DPP4), also known as adenosine deaminase complexing protein 2 or CD26 (cluster of differentiation 26) which is a protein that, in humans, is encoded by the DPP4 gene. DPP-IV plays a major role in glucose metabolism.

[0087] The term “competitive inhibitor” refers to an inhibitor that interrupts a chemical pathway owing to one chemical substance inhibiting the effect of another by competing with it for binding or bonding.

[0088] The term “non-competitive inhibitor” refers to an inhibitor that reduces the activity of an enzyme and binds equally well to the enzyme whether or not it has already bound a substrate.

[0089] The term “ATP” refers to adenosine triphosphate which is an organic compound and hydrotrope that provides energy to drive many processes in living cells, such as muscle contraction, nerve impulse propagation, condensate dissolution, and chemical synthesis. ATP consists of three components: a nitrogenous base (adenine), the sugar ribose, and the triphosphate. Found in all known forms of life, ATP is often referred to as the "molecular unit of currency" of intracellular energy transfer. When consumed in metabolic processes, it converts either to adenosine diphosphate (ADP) or to adenosine monophosphate (AMP). Other processes regenerate ATP so that the human body recycles its own body weight equivalent in ATP each day. It is also a precursor to DNA and RNA, and is used as a coenzyme.

[0090] The articles “a,” “an,” and “the” are used herein to refer to one or to more than one (i.e., to at least one) of the grammatical object of the article. By way of example, “an element” means one element or more than one element.

[0091] As used herein, the term “about” means ± 10% of the specified value, unless otherwise indicated. It is understood that wherever aspects are described herein with the language “about” or “approximately” a numeric value or range, otherwise analogous aspects referring to the specific numeric value or range are also provided.

[0092] The term “at least” prior to a number or series of numbers is understood to include the number adjacent to the term “at least,” and all subsequent numbers or integers that could logically be included, as clear from context. When at least is present before a series of numbers or a range, it is understood that “at least” can modify each of the numbers in the series or range.

[0093] The term “less than” prior to a number or series of numbers is understood not to include the number adjacent to the term “less than,” and all preceding numbers or integers that could logically be included, as clear from context. When “less than” is present before a series of numbers or a range, it is understood that “less than” can modify each of the numbers in the series or range.

[0094] As used herein, the terms “comprises,” “comprising,” “having,” “including,” “containing,” and the like are open-ended terms meaning “including, but not limited to.” To the extent a given aspect disclosed herein “comprises” certain elements, it should be understood that present disclosure also specifically contemplates and discloses aspects that “consist essentially of’ those elements and that “consist of’ those elements.

[0095] The terms “treat,” “treating,” and “treatment” refer to any indicia of success in the treatment or amelioration of an injury, disease, or condition, including any objective or subjective parameter such as abatement; remission; diminishing of symptoms or making the injury, disease, or condition more tolerable to the patient; slowing in the rate of degeneration or decline; or improving a patient’s physical or mental well-being. The treatment or amelioration of symptoms can be based on objective or subject parameters, including the results of a physical examination, neuropsychiatric examinations, or psychiatric evaluation.

[0096] In the context of diabetes mellitus, the term “treating” includes, but is not limited to, maintaining normal blood glucose levels, lowering blood glucose levels, lowering glucose production in the liver, improving sensitivity of the body’s tissues to insulin, stimulating the pancreas to secrete more insulin, and inhibiting the return of glucose to the bloodstream.

[0097] Terms such as “decreasing the severity” refer to therapeutic measures that slow down or lessen the symptoms of a diagnosed pathologic condition or disorder.

[0098] As used herein, “delaying” means to defer, hinder, slow, retard, stabilize, suppress and/or postpone development or progression of the disease (such as diabetes). This delay can be of varying lengths of time, depending on the history of the disease and/or individual being treated.

[0099] Prophylactic or preventative measures refer to measures (e.g., administration of an inhibitor of MAP3K15 to a subject) that decrease the likelihood and/or slow the development of a targeted pathologic condition or disorder. Thus, those in need of prophylactic or preventative measures include those prone to have the disorder and those in whom the disorder is to be prevented.

[0100] By an “effective amount” or “a pharmaceutically effective amount” of a drug or pharmacologically active agent is meant a nontoxic but sufficient amount of the drug or agent to provide the desired effect. A “therapeutically effective amount” refers to an amount effective to achieve the desired therapeutic result. A “prophylactically effective amount” refers to an amount effective to achieve the desired prophylactic result.

[0101] The term “subject” used herein refers to an animal; for example, a mammal, such as a human. In some aspects, methods of treating mammals, including, but not limited to, humans, rodents, simians, felines, canines, equines, bovines, porcines, ovines, caprines, mammalian laboratory animals, mammalian farm animals, mammalian sport animals, and mammalian pets, are provided. In some examples, a “subject” refers to a subject in need of treatment for a disease or disorder. In some aspects, the subject to receive the treatment can be a patient, designating the fact that the subject has been identified as having a disorder of relevance to the treatment, or being at particular risk of contracting the disorder. In some aspects, the subject is human. In some aspects, the subject is male. In some aspects, the subject is female. In some aspects, the subject is a mouse.

[0102] A “disease” or “disorder” as used herein refers to a condition where treatment is needed and/or desired.

[0103] As used herein, the term "loss of function" mutation refers to a mutation that that results in the absence of a gene, decreased expression of a gene, or the production of a gene product (e.g. protein) having decreased activity or no activity. Loss of function mutations include for example, missense mutations, nucleotide insertions, nucleotide deletions, and gene deletions. Loss of function mutations also include dominant negative mutations.

[0104] The terms “administer,” “administering,” “administration,” and the like refer to methods that can be used to enable delivery of the therapeutic agent to the desired site of biological action. Administration techniques that can be employed with the agents and methods described herein are found in e.g., Goodman and Gilman, The Pharmacological Basis of Therapeutics, current ed.; Pergam on; and Remington’s, Pharmaceutical Sciences (current edition), Mack Publishing Co., Easton, Pa.

[0105] The terms “pharmaceutical formulation” and “pharmaceutical composition” refer to a preparation which is in such form as to permit the biological activity of the active ingredient(s) to be effective, and which contains no additional components which are unacceptably toxic to a subject to which the formulation would be administered. Such formulations may be sterile.

[0106] A “pharmaceutically acceptable carrier” refers to a non-toxic solid, semisolid, or liquid filler, diluent, encapsulating material, formulation auxiliary, or carrier conventional in the art for use with a therapeutic agent that together comprise a “pharmaceutical composition” for administration to a subject. A pharmaceutically acceptable carrier is nontoxic to recipients at the dosages and concentrations employed and is compatible with other ingredients of the formulation. The pharmaceutically acceptable carrier is appropriate for the formulation employed.

[0107] Administration “in combination with” one or more further therapeutic agents includes simultaneous (concurrent) and consecutive or sequential administration in any order. Where two or more therapeutic agents are administered “in combination” but not concurrently, they are administered in sufficient proximity in time that the therapeutic activities of the two or more therapeutic agents in the subject at least partially overlap.

[0108] The term “concurrently” is used herein to refer to administration of two or more therapeutic agents, where at least part of the administration overlaps in time or where the administration of one therapeutic agent falls within a short period of time relative to administration of the other therapeutic agent. For example, the two or more therapeutic agents are administered with a time separation of no more than about a specified number of minutes or hours, such as within 12 hours, within 6 hours, within 3 hours, within 2 hours, within 1 hour, or within 30 minutes of one another.

[0109] The term “sequentially” is used herein to refer to administration of two or more therapeutic agents where the administration of one or more agent(s) continues after discontinuing the administration of one or more other agent(s), or wherein administration of one or more agent(s) begins before the administration of one or more other agent(s). For example, administration of the two or more therapeutic agents are administered with a time separation of more than about a specified number of hours or days, such as at least 12 hours, at least 1 day, at least 2 days, at least 3 days, at least 1 week, at least 2 weeks, or at least 1 month apart.

[0110] In this application, the use of “or” means “and/or” unless expressly stated or understood by one skilled in the art. In the context of a multiple dependent claim, the use of “or” refers back to more than one preceding independent or dependent claim.

Mitogen- Activated Protein Kinase Kinase Kinase 15 (MAP3K15) Inhibitors

[0111] A MAP3K15 inhibitor is an agent that reduces the level of MAP3K15 protein and/or inhibits or reduces at least one biological activity of MAP3K15 protein. In some aspects, a MAP3K15 inhibitor is a small molecule (including a PROTAC), an inhibitory nucleic acid (e.g., antisense oligonucleotides, an siRNA (including shRNA), a micro RNA, or an antagomiRNA), a gene-regulating system, a protein, or an aptamer.

[0112] In some aspects, a MAP3K15 inhibitor comprises a nucleic acid molecule and an enzymatic protein. In some aspects, the nucleic acid molecule is a guide RNA (gRNA) molecule. In some aspects, the enzymatic protein is a Cas protein or Cas ortholog. [0113] In some aspects, a MAP3K15 inhibitor specifically binds to MAP3K15 protein. In some aspects, a MAP3K15 inhibitor specifically binds to the ATP -binding domain of MAP3K15. In some aspects, a MAP3K15 inhibitor is a non-ATP competitive inhibitor. In some aspects, a MAP3K15 inhibitor specifically binds to the meatal-binding domain of MAP3K15.

[0114] In some aspects, a MAP3K15 inhibitor inhibits the serine-threonine kinase activity of MAP3K15. In some aspects, a MAP3K15 inhibitor inhibits the transferase activity of MAP3K15. In some aspects, a MAP3K15 inhibitor decreases binding of MAP3K15 to ATP. In some aspects, a MAP3K15 inhibitor decreases binding of MAP3K15 to magnesium. In some aspects, a MAP3K15 inhibitor decreases phosphorylation of MAP3K15 Thr-812. In some aspects, a MAP3K15 inhibitor increases phosphorylation of MAP3K15 Ser-924 and/or Ser-994. In some aspects, a MAP3K15 inhibitor decreases homooligomerisation of MAP3K15.

[0115] In some aspects, a MAP3K15 inhibitor has an ICsoof less than 50 nM. In some aspects, a MAP3K15 inhibitor has an IC50 of less than 10 nM. In some aspects, the concentration of the MAP3K15 inhibitor required to cause a 50% inhibition of MAP3K15 is at least 10-fold lower than that required to cause a 50% inhibition of ASK1. In some aspects, the concentration of the MAP3K15 inhibitor required to cause a 50% inhibition of MAP3K15 is at least 10-fold lower than that required to cause a 50% inhibition of ASK2.

[0116] In some aspects, a MAP3K15 inhibitor is a reversible inhibitor. In some aspects, a MAP3K15 inhibitor is an irreversible inhibitor.

Small molecule inhibitors

[0117] In some aspects, a MAP3K15 inhibitor is a small molecule. A “small molecule” is defined herein to have a molecular weight below about 1000 Daltons, for example, below about 900 Daltons, below about 800 Daltons, below about 700 Daltons, below about 600 Daltons, or below about 500 Daltons. In some aspects, a small molecule has a molecular weight between 150 Daltons and 1000 Daltons, or 150 Daltons and 800 Daltons, or 150 Daltons and 700 Daltons, or 150 Daltons and 600 Daltons. Small molecules can be organic or inorganic, and can be isolated from, for example, compound libraries or natural sources, or may be obtained by derivatization of known compounds. [0118] Small molecule MAP3K15 inhibitors include, for example, staurosporine

((9S, 10R, 11R, 13R)-2,3, 10, 11 , 12, 13-hexahydro- 10-methoxy-9-methyl- 11 -(methylamino)- 9,13-epoxy-lH,9H-diindolo[l,2,3-gh:3',2',r-lm]pyrrolo[3,4-j][l,7]benzodiazonin-l-one, available, e.g., from Sigma), lestaurtinib ((5S,6S,8R)-6-hydroxy-6-(hydroxymethyl)-5- methyl-7,8, 14, 15-tetrahydro-5H- 16-oxa-4b,8a, 14-triaza-5,8-methanodibenzo- [b,h]cycloocta[jkl]cyclopenta[e]-as-indacen-13(6H)-one, available, e.g., from Fermentek Ltd.), TAE-684 (5-chloro-N2-[2-methoxy-4-[4-(4-methyl-l-piperazinyl)-l- piperidinyl]phenyl]-N4-[2-[(l-methylethyl)sulfonyl]phenyl]-2,4-pyrimidinediamine, available, e.g., from MedChemExpress), ruxolitinib ((3R)-3-cyclopentyl-3-[4-(7H- pyrrolo[2,3-d]pyrimidin-4-yl)pyrazol-l-yl]propanenitrile, sold under the brand names Jakafi® and Jakavi®), sunitinib (N-(2-diethylaminoethyl)-5-[(Z)-(5-fluoro-2-oxo-lH-indol- 3-ylidene)methyl]-2,4-dimethyl-lH-pyrrole-3-carboxamide, sold under the brand name Sutent®), nintedanib (methyl (3Z)-3-{[(4-{methyl[(4-methylpiperazin-l- yl)acetyl]amino]phenyl)amino](phenyl)methylidene}-2-oxo-2,3-dihydro-lH-indole-6- carboxylate, sold under the brand names Ofev® and Vargatef®), crizotinib (3-[(lR)-l-(2,6- dichloro-3-fluorophenyl)ethoxy]-5-(l-piperidin-4-ylpyrazol-4-yl)pyridin-2-amine, sold under the brand name, e.g., Xalkori®), and fosfamatinib ([6-({5-fluoro-2-[(3,4,5-trimethoxy- phenyl)amino]-4-pyrimidinyl}amino)-2,2-dimethyl-3-oxo-2,3-dihydro-4H-pyrido[3,2- b] [1, 4] oxazin-4-yl] methyl dihydrogen phosphate, sold under the brand names Tavalisse® and Tavlesse®).

[0119] In some aspects, a small molecule inhibitor of MAP3K15 is identified by screening a library of small molecules. The generation and screening of small molecule libraries is well known in the art. See, e.g., Thompson et al., Chem. Rev. 96: 555-600 (1996); and the National Institutes of Health Molecular Libraries Program. A combinatorial chemical library, for example, can be formed by mixing a set of chemical building blocks in various combinations, and can result in millions of chemical compounds. For example, the systematic, combinatorial mixing of 100 interchangeable chemical building blocks theoretically results in the synthesis of 100 million tetrameric compounds or 10 billion pentameric compounds. See, e.g., Gallop et al. 1994, J. Med. Chem. 37: 1233-1250). Various other types of small molecule libraries can also be designed and used, such as, for example, natural product libraries. [0120] In some aspects, a small molecule inhibitor of MAP3K15 binds to MAP3K15 protein. In some aspects, a small molecule MAP3K15 inhibitor does not bind to other mitogen- activated protein kinases (MAPKs, such as, e.g., ASK1 or ASK2) or other MAP3K proteins or binds mitogen-activated protein kinases or other MAP3K proteins with at least 5-fold, at least 10-fold, at least 20-fold, or at least 100-fold reduced affinity compared to the affinity for MAP3K15 (i.e., the KD of the MAP3K15 inhibitor for other mitogen-activated protein kinases or other MAP3K proteins is at least 5-fold, at least 10-fold, at least 20-fold, or at least 100-fold higher than the KD for MAP3K15).

[0121] In some aspects, a small molecule MAP3K15 inhibitor inhibits MAP3K15 activity with an IC50 of less than about 10 nM, less than about 50 nM, between about 50 nM and about 200 nM, between about 200 nM and about 2 pM, or greater than 2 pM. In some aspects, a small molecule MAP3K15 inhibitor does not inhibit the activity of other mitogen- activated protein kinases or other MAP3K proteins, or inhibits the activity of other mitogen- activated protein kinases or other MAP3KP proteins with at least 5-fold, at least 10-fold, at least 20-fold, or at least 100-fold higher IC50 compared to the IC50 for inhibition of MAP3K15 activity.

Proteolysis Targeting Chimeras (PROTAC)

[0122] A proteolysis targeting chimera (PROTAC) is a heterobifunctional small molecule composed of two active domains and a linker, capable of removing specific unwanted proteins. Rather than acting as a conventional enzyme inhibitor, a PROTAC works by inducing selective intracellular proteolysis. PROTACs consist of two covalently linked protein-binding molecules: one capable of engaging an E3 ubiquitin ligase, and another that binds to a target protein meant for degradation. Recruitment of the E3 ligase to the target protein results in ubiquitination and subsequent degradation of the target protein by the proteasome. PROTACs need only to bind their targets with high selectivity (rather than inhibit the target protein's enzymatic activity).

[0123] PROTACs achieve degradation through "hijacking" the cell's ubiquitin-proteosome system (UPS). The UPS consists of an El activating enzyme which conjugates to an E2 enzyme transferring a ubiquitin molecule to the E2. E2 then binds to the E3 ligase in a complex which can then recognize target proteins for subsequent ubiquitin tagging and degradation by the 26S pro teo some.

[0124] PROTAC-like molecules are discussed, for example, in U.S. Patent No. 6,306,663. [0125] As provided herein, a PROTAC can be used to inhibit MAP3K15.

Gene regulating systems for inhibiting MAP3K15

[0126] A gene regulating system for inhibiting MAP3K15 can be a nucleic acid-based gene regulating system or a protein-based gene regulating system. A nucleic acid-based gene regulating system is a system comprising one or more nucleic acid molecules that is capable of regulating the expression of an endogenous target gene without the requirement for an exogenous protein. In some aspects, the nucleic acid-based gene regulating system comprises a small interfering RNA or antisense oligonucleotide. Examples of each of these are described below.

[0127] In some aspects, a nucleic acid-based gene regulating system MAP3K15 inhibitor is a polynucleotide.

[0128] In some aspects, a nucleic acid-based gene regulating system MAP3K15 inhibitor comprises a nucleotide sequence GCTCTCTCTTTGAAGGACATGGTAA (SEQ ID NO:1).

[0129] In some aspects, a nucleic acid-based gene regulating system MAP3K15 inhibitor comprises a nucleotide sequence CGGAUUUCAGGAUGCCGUAAAUAAA (SEQ ID NOG).

[0130] In some aspects, a nucleic acid-based gene regulating system MAP3K15 inhibitor comprises a nucleotide sequence CACCGAAGAGCAGUGCAGUAGAUUU (SEQ ID NOG).

[0131] In some aspects, a nucleic acid-based gene regulating system MAP3K15 inhibitor comprises at least two nucleotide sequences selected from a group consisting of the nucleotide sequences: GCTCTCTCTTTGAAGGACATGGTAA (SEQ ID NO:1), CGGAUUUCAGGAUGCCGUAAAUAAA (SEQ ID NOG), and CACCGAAGAGCAGUGCAGUAGAUUU (SEQ ID NOG).

Antisense oligonucleotides

[0132] In some aspects, a MAP3K15 inhibitor is an antisense oligonucleotide. In some aspects, an antisense oligonucleotide hybridizes to a region of MAP3K15 mRNA and/or MAP3K15 pre-mRNA and directs its degradation through RNase H, which cleaves doublestranded RNA/DNA hybrids. By mediating cleavage of MAP3K15 mRNA and/or MAP3K15 pre-mRNA, an antisense oligonucleotide can reduce the amount of MAP3K15 protein in a cell (i.e., can inhibit expression of MAP3K15). In some aspects, an antisense oligonucleotide does not mediate degradation through RNase H, but rather “blocks” translation of the mRNA, e.g., through interference with translational machinery binding or processivity, or “blocks” proper splicing of the pre-mRNA, e.g., through interference with the splicing machinery and/or accessibility of a splice site. In some aspects, an antisense oligonucleotide can mediate degradation of an mRNA and/or pre-mRNA through a mechanism other than RNase H. Any inhibitory mechanism of an antisense oligonucleotide is contemplated herein. [0133] In some aspects, an antisense oligonucleotide is 10 to 100 nucleotides long, or 15 to 100 nucleotides long, or 10 to 50 nucleotides long, or 15 to 50 nucleotides long, or 15 to 30 nucleotides long. In some aspects, an antisense oligonucleotide hybridizes to a region of the MAP3K15 mRNA and/or pre-mRNA comprising at least 10, at least 15, at least 20, at least 25, at least 30, at least 40, at least 50, at least 60, at least 70, at least 80, at least 90, or at least 100 nucleotides. Further, in some aspects, an antisense oligonucleotide need not be 100% complementary to a region MAP3K15 mRNA and/or a region of MAP3K15 pre-mRNA, but can have one or more mismatches. Thus, in some aspects, an antisense oligonucleotide is at least 80% complementary, at least 85% complementary, at least 90% complementary, at least 95% complementary, or 100% complementary to a region of MAP3K15 mRNA and/or a region of MAP3K15 pre-mRNA.

[0134] Antisense oligonucleotides can comprise modifications to one or more of the internucleoside linkages, sugar moieties, and/or nucleobases. Further, the sequence of nucleotides can be interrupted by non-nucleotide components, and/or non-nucleotide components can be attached at one or both ends of the oligonucleotide.

[0135] Nonlimiting exemplary nucleotide modifications include sugar modifications, for example, analogous forms of ribose or deoxyribose sugars that are generally known in the art, including, for example, 2'-O-methyl-2'-O-allyl, 2'-fluoro- or 2'-azido-ribose, carbocyclic sugar analogs, a-anomeric sugars, epimeric sugars such as arabinose, xyloses or lyxoses, pyranose sugars, furanose sugars, sedoheptuloses, acyclic analogs and abasic nucleoside analogs such as methyl riboside. [0136] In some aspects, one or more phosphodiester linkages can be replaced by modified internucleoside linkages. These modified intemucleoside linkages include, but are not limited to, aspects wherein phosphate is replaced by P(O)S (“thioate”), P(S)S (“dithioate”), (O)NRi (“amidate”), P(O)R, P(O)OR’, CO or CH2 (“formacetal”), in which each R or R’ is independently H or substituted or unsubstituted alkyl (1-20 C) optionally containing an ether (-O-) linkage, aryl, alkenyl, cycloalkyl, cycloalkenyl or araldyl. Not all linkages in a polynucleotide need be identical. The preceding description applies to all oligonucleotides referred to herein, including antisense oligonucleotides and siRNA.

[0137] In some aspects, one or more internucleoside linkages in an antisense oligonucleotide are phosphorothioates. In some aspects, one or more sugar moieties in an antisense oligonucleotide comprise 2’ modifications, such as 2’-O-alkyl (such as 2’-0Me) and 2’- fluoro; or are bicyclic sugar moieties (such as LNA). Nonlimiting exemplary nucleobase modifications include 5-methylcytosine. An antisense oligonucleotide can comprise more than one type of modification within a single oligonucleotide. That is, as a nonlimiting example, an antisense oligonucleotide can comprise 2’-0 alkyl modifications, bicyclic nucleotides, and phosphorothioate linkages in the same oligonucleotide. In some aspects, an antisense oligonucleotide is a “gapmer.” Gapmers comprise a central region of deoxyribonucleotides for mediating RNase H cleavage, and 5’ and 3’ “wings” comprising modified sugar moieties that increase the stability of the duplex.

[0138] Antisense oligonucleotide design and mechanisms are described, e.g., in van Roon- Mom et al., Methods Mol. Biol., 867: 79-96 (20120); Prakash, Chem. Biodivers., 8: 1616- 1641 (2011); Yamamoto et al., Future Med. Chem., 3: 339-365 (2011); Chan et al., Clin. Exper. Pharmacol. Physiol., 33: 533-540 (2006); Kurreck et al., Nucl. Acids Res., 30: 1911- 1918 (2002); Kurreck, Eur. J. Biochem., 270: 1628-1644 (2003); Geary, Expert Opin. Drug Metab. Toxicol., 5: 381-391 (2009); “Designing Antisense Oligonucleotides,” available online from Integrated DNA Technologies (2011).

Small interfering RNAs

[0139] In some aspects, a MAP3K15 inhibitor is an siRNA. As used herein, siRNAs are synonymous with double- stranded RNA (dsRNA) and include double-stranded RNA oligomers with or without hairpin structures at each end (also referred to as small hairpin RNA, or shRNA). Short interfering RNAs are also known as small interfering RNAs, silencing RNAs, short inhibitory RNA, and/or small inhibitory RNAs, and these terms are considered to be equivalent herein.

[0140] The term “short-interfering RNA (siRNA)” refers to small double-stranded RNAs that interfere with gene expression. siRNAs are mediators of RNA interference, the process by which double- stranded RNA silences homologous genes. In some aspects, siRNAs comprise two single- stranded RNAs of about 15-25 nucleotides in length that form a duplex, which can include single- stranded overhang(s). In some aspects, siRNAs comprise a single RNA that forms a hairpin structure that includes a double -stranded portion that can be 15-25 nucleotides in length and can include a single-stranded overhang. Such hairpin siRNAs can be referred to as a short hairpin RNA (shRNA). Processing of the double -stranded RNA by an enzymatic complex, for example, polymerases, can result in cleavage of the doublestranded RNA to produce siRNAs. The antisense strand of the siRNA is used by an RNA interference (RNAi) silencing complex to guide mRNA cleavage, thereby promoting mRNA degradation. To silence a specific gene using siRNAs, for example, in a mammalian cell, a base pairing region is selected to avoid chance complementarity to an unrelated mRNA. RNAi silencing complexes have been identified in the art, such as, for example, by Fire et al., Nature 391:806-811, 1998, and McManus et al., Nat. Rev. Genet. 3(10):737-747 , 2002.

[0141] In some aspects, small interfering RNAs comprise at least about 10 to about 100 nucleotides, including at least about 16 nucleotides, at least about 17 nucleotides, at least about 18 nucleotides, at least about 19 nucleotides, at least about 20 nucleotides, at least about 21 nucleotides, at least about 22 nucleotides, at least about 23 nucleotides, at least about 24 nucleotides, at least about 25 nucleotides, at least about 26 nucleotides, at least about 27 nucleotides, at least about 28 nucleotides, at least about 29 nucleotides, at least about 30 nucleotides, at least about 35 nucleotides, at least about 40 nucleotides, at least about 45 nucleotides, at least about 50 nucleotides, at least about 55 nucleotides, at least about 60 nucleotides, at least about 65 nucleotides, at least about 70 nucleotides, at least about 75 nucleotides, at least about 80 nucleotides, at least about 85 nucleotides, at least about 90 nucleotides, at least about 95 nucleotides, or at least about 100 nucleotides. In some aspects, an siRNA is 15 to 100 nucleotides long, or 10 to 60 nucleotides long, or 15 to 60 nucleotides long, or 10 to 50 nucleotides long, or 15 to 50 nucleotides long, or 10 to 30 nucleotides long, or 15 to 30 nucleotides long. In certain aspects, the siRNA comprises an oligonucleotide from about 21 to about 25 nucleotides in length. In some aspects, the siRNA molecule is a heteroduplex of RNA and DNA. In some aspects, an siRNA comprises a nucleotide sequence that is at least 75%, at least 80%, at least 85%, at least 90%, at least 95%, or 100% identical to a nucleotide sequence found in the MAP3K15 mRNA and/or pre- mRNA.

[0142] As with antisense oligonucleotides, siRNAs can include modifications to the sugar, internucleoside linkages, and/or nucleobases. Nonlimiting exemplary modifications suitable for use in siRNAs are described herein and also, e.g., in Peacock et al., J. Org. Chem., 76: 7295-7300 (2011); Bramsen et al., Methods Mol. Biol., 721: 77-103 (2011); Pasternak et al., Org. Biomol. Chem., 9: 3591-3597 (2011); Gaglione et al., Mini Rev. Med. Chem., 10: 578- 595 (2010); Chernolovskaya et al., Curr. Opin. Mol. Ther., 12: 158-167 (2010). See also siRNA Design: Methods and Protocols, Ed. Debra J. Taxman, Springer- Verlag New York, LLC, 2013.

[0143] In some aspects, an siRNA specific to the MAP3K15 gene is synthesized and introduced directly into a subject. In some aspects, the siRNA can be formulated as part of a targeted delivery system, such as a target specific liposome, which specifically recognizes and delivers the siRNA to an appropriate tissue or cell type. Upon administration of the targeted siRNA to a subject, the siRNA is delivered to the appropriate cell type, thereby increasing the concentration siRNA within the cell type. This process can provide partial or complete loss of MAP3K15 protein expression.

[0144] In some aspects, an appropriate cell or tissue is provided with an expression construct that comprises a nucleic acid encoding one or both strands of an siRNA that is specific to the MAP3K15 gene. In these aspects, the nucleic acid that encodes one or both strands of the siRNA can be placed under the control of either a constitutive or a regulatable promoter. In some aspects, the nucleic acid encodes an siRNA that forms a hairpin structure, e.g., a shRNA.

[0145] Various carriers and drug-delivery systems for siRNAs are described, e.g., in Seth et al., Ther. Deliv., 3: 245-261 (2012); Kanasty et al., Mol. Ther., 20: 513-524 (2012); Methods Enzymol., 502: 91-122 (2012); Vader et al., Curr. Top. Med. Chem., 12: 108-119 (2012); Naeye et al., Curr. Top. Med. Chem., 12: 89-96 (2012); Eoged, Curr. Top. Med. Chem., 12: 97-107 (2012); Chaturvedi et al., Expert Opin. Drug Deliv., 8: 1455-1468 (2011); Gao et al., Int. J. Nanomed., 6: 1017-1025 (2011); Shegokar et al., Pharmazie., 66: 313-318 (2011);

Kumari et al., Expert Opin. Drug Deliv., 11: 1327-1339 (2011).

Protein-based gene regulating systems

[0146] As used herein, a protein-based gene regulating system is a system comprising one or more proteins capable of regulating the expression of an endogenous target gene in a sequence specific manner without the requirement for a nucleic acid guide molecule. In some aspects, a protein-based gene regulating system MAP3K15 inhibitor is an enzymatic protein that binds to a gene encoding MAP3K15. In some aspects, the protein-based gene regulating system comprises a protein comprising one or more zinc -finger binding domains and an enzymatic domain. In some aspects, the protein-based gene regulating system comprises a protein comprising a transcription activator-like effector nuclease (TALEN) domain and an enzymatic domain. Such aspects are referred to herein as “TALENs.” In some aspects, the protein-based gene regulating system comprises a megalonuclease.

Zinc finger-based systems

[0147] Zinc finger-based systems comprise a fusion protein comprising two protein domains: a zinc finger DNA binding domain and an enzymatic domain. A “zinc finger DNA binding domain,” “zinc finger protein,” or “ZFP” is a protein, or a domain within a larger protein, that binds DNA in a sequence-specific manner through one or more zinc fingers, which are regions of amino acid sequence within the binding domain whose structure is stabilized through coordination of a zinc ion. The zinc finger domain, by binding to a target DNA sequence, directs the activity of the enzymatic domain to the vicinity of the sequence and, hence, induces modification of the endogenous target gene in the vicinity of the target sequence. A zinc finger domain can be engineered to bind to virtually any desired sequence. Accordingly, after identifying a target genetic locus containing a target DNA sequence at which cleavage or recombination is desired (e.g., a target locus in the MAP3K15 gene (GenBank Accession No. NG_021184.1 or NM_001001671.4)), one or more zinc finger binding domains can be engineered to bind to one or more target DNA sequences in the target genetic locus. Expression of a fusion protein comprising a zinc finger binding domain and an enzymatic domain in a cell, effects modification in the target genetic locus.

[0148] In some aspects, a zinc finger binding domain comprises one or more zinc fingers. See, e.g., Miller et al. (1985) EMBO J. 4:1609-1614; Rhodes (1993) Scientific American Febuary: 56-65; U.S. Pat. No. 6,453,242. Typically, a single zinc finger domain is about 30 amino acids in length. An individual zinc finger binds to a three -nucleotide (z.e., triplet) sequence (or a four-nucleotide sequence which can overlap, by one nucleotide, with the four- nucleotide binding site of an adjacent zinc finger). Therefore the length of a sequence to which a zinc finger binding domain is engineered to bind (e.g., a target sequence) will determine the number of zinc fingers in an engineered zinc finger binding domain. For example, for ZFPs in which the finger motifs do not bind to overlapping subsites, a six- nucleotide target sequence is bound by a two-finger binding domain; a nine-nucleotide target sequence is bound by a three-finger binding domain, etc. Binding sites for individual zinc fingers (z.e., subsites) in a target site need not be contiguous, but can be separated by one or several nucleotides, depending on the length and nature of the amino acids sequences between the zinc fingers (z.e., the inter-finger linkers) in a multi-finger binding domain. In some aspects, the DNA-binding domains of individual ZFNs comprise between three and six individual zinc finger repeats and can each recognize between 9 and 18 base pairs.

[0149] Zinc finger binding domains can be engineered to bind to a sequence of choice. See, for example, Beerli et al. (2002) Nature Biotechnol. 20:135-141; Pabo et al. (2001) Ann. Rev. Biochem. 70:313-340; Isalan et al. (2001) Nature Biotechnol. 19:656-660; Segal et al. (2001) Curr. Opin. Biotechnol. 12:632-637; Choo et al. (2000) Curr. Opin. Struct. Biol. 10:411-416. An engineered zinc finger binding domain can have a novel binding specificity, compared to a naturally-occurring zinc finger protein. Engineering methods include, but are not limited to, rational design and various types of selection.

[0150] Selection of a target DNA sequence for binding by a zinc finger domain can be accomplished, for example, according to the methods disclosed in U.S. Pat. No. 6,453,242. It will be clear to those skilled in the art that simple visual inspection of a nucleotide sequence can also be used for selection of a target DNA sequence. Accordingly, any means for target DNA sequence selection can be used in the methods described herein. A target site generally has a length of at least 9 nucleotides and, accordingly, is bound by a zinc finger binding domain comprising at least three zinc fingers. However binding of, for example, a 4-finger binding domain to a 12-nucleotide target site, a 5 -finger binding domain to a 15 -nucleotide target site or a 6-finger binding domain to an 18-nucleotide target site, is also possible. As will be apparent, binding of larger binding domains (e.g., 7-, 8-, 9-finger and more) to longer target sites is also possible.

[0151] In some aspects, the zinc finger binding domain binds to a target DNA sequence that is at least 90% identical to a target DNA sequence within a MAP3K15 gene sequence (GenBank Accession No. NG_021184.1 or NM_001001671.4). In some aspects, the zinc finger binding domain binds to a target DNA sequence that is at least 95%, 96%, 97%, 98%, or 99% identical to a target DNA sequence within a MAP3K15 gene sequence (GenBank Accession No. NG_021184.1 or NM_001001671.4). In some aspects, the zinc finger binding domain binds to a target DNA sequence that is 100% identical to a target DNA sequence within a MAP3K15 gene sequence (GenBank Accession No. NG_021184.1 or NM_001001671.4).

TALEN-based systems

[0152] TALEN-based systems comprise a protein comprising a TAL effector DNA binding domain and an enzymatic domain. They are made by fusing a TAL effector DNA-binding domain to a DNA cleavage domain (a nuclease which cuts DNA strands). The FokI restriction enzyme is an exemplary enzymatic domain suitable for use in TALEN-based gene regulating systems.

[0153] TAL effectors are proteins that are secreted by Xanthomonas bacteria via their type III secretion system when they infect plants. The DNA binding domain contains a repeated, highly conserved, 33-34 amino acid sequence with divergent 12th and 13th amino acids. These two positions, referred to as the Repeat Variable Diresidue (RVD), are highly variable and strongly correlated with specific nucleotide recognition. Therefore, the TAL effector domains can be engineered to bind specific target DNA sequences by selecting a combination of repeat segments containing the appropriate RVDs. The nucleic acid specificity for RVD combinations is as follows: HD targets cytosine, NI targets adenenine, NG targets thymine, and NN targets guanine (though, in some aspects, NN can also bind adenenine with lower specificity).

[0154] In some aspects, the TAL effector domains bind to a target DNA sequence that is at least 90% identical to a target DNA sequence within a MAP3K15 gene sequence (GenBank Accession No. NG_021184.1 or NM_001001671.4 ). In some aspects, the TAL effector domains bind to a target DNA sequence that is at least 95%, 96%, 97%, 98%, or 99% identical to a target DNA sequence within a MAP3K15 gene sequence (GenBank Accession No. NG_021184.1 or NM_001001671.4). In some aspects, the TAL effector domains bind to a target DNA sequence that is 100% identical to a target DNA sequence within a MAP3K15 gene sequence (GenBank Accession No. NG_021184.1 or NM_001001671.4).

Combination nucleic acid/protein-based gene regulating systems

[0155] Combination gene-regulating systems comprise a site-directed modifying polypeptide and a nucleic acid guide molecule. Herein, a “site-directed modifying polypeptide” refers to a polypeptide that binds to a nucleic acid guide molecule, is targeted to a target nucleic acid sequence, such as, for example, a DNA sequence, by the nucleic acid guide molecule to which it is bound, and modifies the target DNA sequence (e.g.. cleavage, mutation, or methylation of target DNA). A site-directed modifying polypeptide comprises two portions, a portion that binds the nucleic acid guide and an activity portion. In some aspects, a site- directed modifying polypeptide comprises an activity portion that exhibits site-directed enzymatic activity (e.g.. DNA methylation, DNA cleavage, histone acetylation, histone methylation, etc.), wherein the site of enzymatic activity is determined by the guide nucleic acid. In some cases, a site-directed modifying polypeptide has enzymatic activity that modifies target DNA (e.g.. nuclease activity, methyltransferase activity, demethylase activity, DNA repair activity, DNA damage activity, deamination activity, dismutase activity, alkylation activity, depurination activity, oxidation activity, pyrimidine dimer forming activity, integrase activity, transposase activity, recombinase activity, polymerase activity, ligase activity, helicase activity, photolyase activity or glycosylase activity). In some aspects, the activity portion modulates transcription of the target DNA sequence (e.g.. to increase or decrease transcription).

[0156] The nucleic acid guide comprises two portions: a first portion that is complementary to, and capable of binding with, an endogenous target DNA sequence (referred to herein as a “DNA-binding segment”), and a second portion that is capable of interacting with the site- directed modifying polypeptide (referred to herein as a “protein-binding segment”). In some aspects, the DNA-binding segment and protein-binding segment of a nucleic acid guide are comprised within a single polynucleotide molecule. In some aspects, the DNA-binding segment and protein-binding segment of a nucleic acid guide are each comprised within separate polynucleotide molecules, such that the nucleic acid guide comprises two polynucleotide molecules that associate with each other to form the functional guide.

[0157] The nucleic acid guide mediates the target specificity of the combined protein/nucleic gene regulating systems by specifically hybridizing with a target DNA sequence comprised within the DNA sequence of a target gene. Reference herein to a target gene encompasses the full-length DNA sequence for that particular gene and a full-length DNA sequence for a particular target gene will comprise a plurality of target genetic loci, which refer to portions of a particular target gene sequence (e.g., an exon or an intron). Within each target genetic loci are shorter stretches of DNA sequences referred to herein as “target DNA sequences” or “target sequences” that can be modified by the gene -regulating systems described herein. Further, each target genetic loci comprises a “target modification site,” which refers to the precise location of the modification induced by the gene -regulating system (e.g., the location of an insertion, a deletion, or mutation, the location of a DNA break, or the location of an epigenetic modification). The gene -regulating systems described herein can comprise a single nucleic acid guide, or may comprise a plurality of nucleic acid guides (e.g., 2, 3, 4, 5, 6, 7, 8, 9, 10, or more nucleic acid guides).

[0158] The CRISPR/Cas systems described below are examples of a combination protein/nucleic acid systems.

CRISPR/Cas gene regulating systems

[0159] In some aspects, the gene editing systems described herein are CRISPR (Clustered Regularly Interspaced Short Palindromic Repeats)/Cas (CRISPR Associated) nuclease systems. In such aspects, the site-directed modifying polypeptide is a CRISPR-associated endonuclease (a “Cas” endonuclease) and the nucleic acid guide molecule is a guide RNA (gRNA).

[0160] A “Cas polypeptide” and “Cas protein” are used interchangeably to refer to a polypeptide that can interact with a gRNA molecule and, in concert with the gRNA molecule, homes or localizes to a target DNA sequence and includes naturally occurring Cas proteins and engineered, altered, or otherwise modified Cas proteins that differ by one or more amino acid residues from a naturally -occurring Cas sequence. In some aspects, the Cas protein is a Cas9 protein. [0161] A guide RNA (gRNA) comprises two segments, a DNA-binding segment and a protein-binding segment. In some aspects, the protein -binding segment of a gRNA is comprised in one RNA molecule and the DNA-binding segment is comprised in another separate RNA molecule. Such aspects are referred to herein as “double-molecule gRNAs” or “two-molecule gRNA” or “dual gRNAs.” In some aspects, the gRNA is a single RNA molecule and is referred to herein as a “single-guide RNA” or an “sgRNA.” The term “guide RNA” or “gRNA” is inclusive, referring both to two-molecule guide RNAs and sgRNAs.

[0162] The protein-binding segment of a gRNA comprises, in part, two complementary stretches of nucleotides that hybridize to one another to form a double stranded RNA duplex (dsRNA duplex), which facilitates binding to the Cas protein.

[0163] The DNA-binding segment (or “DNA-binding sequence”) of a gRNA comprises a nucleotide sequence that is complementary to and capable of binding to a specific sequence target DNA sequence. The protein-binding segment of the gRNA interacts with a Cas polypeptide and the interaction of the gRNA molecule and site-directed modifying polypeptide results in Cas binding to the endogenous DNA and produces one or more modifications within or around the target DNA sequence. The precise location of the target modification site is determined by both (i) base-pairing complementarity between the gRNA and the target DNA sequence; and (ii) the location of a short motif, referred to as the protospacer adjacent motif (PAM), in the target DNA sequence. The PAM sequence is required for Cas binding to the target DNA sequence. A variety of PAM sequences are known in the art and are suitable for use with a particular Cas endonuclease (e.g., a Cas9 endonuclease) are known in the art (See e.g., Nat Methods. 2013 Nov; 10(11): 1116-1121 and Sci Rep. 2014; 4: 5405). In some aspects, the PAM sequence is located within 50 base pairs of the target modification site. In some aspects, the PAM sequence is located within 10 base pairs of the target modification site. The DNA sequences that can be targeted by this method are limited only by the relative distance of the PAM sequence to the target modification site and the presence of a unique 20 base pair sequence to mediate sequencespecific, gRNA-mediated Cas binding. In some aspects, the target modification site is located at the 5’ terminus of the target locus. In some aspects, the target modification site is located at the 3’ end of the target locus. In some aspects, the target modification site is located within an intron or an exon of the target locus. [0164] In some aspects, the present disclosure provides a polynucleotide encoding a gRNA. In some aspects, a gRNA-encoding nucleic acid is comprised in an expression vector, e.g., a recombinant expression vector. In some aspects, the present disclosure provides a polynucleotide encoding a site-directed modifying polypeptide. In some aspects, the polynucleotide encoding a site-directed modifying polypeptide is comprised in an expression vector, e.g., a recombinant expression vector.

[0165] A gRNA comprises a DNA-targeting segment and protein-binding segment. The DNA-targeting segment of a gRNA comprises a nucleotide sequence that is complementary to a sequence in the target DNA sequence. As such, the DNA-targeting segment of a gRNA interacts with a target DNA in a sequence-specific manner via hybridization (z.e., base pairing), and the nucleotide sequence of the DNA-targeting segment determines the location within the target DNA that the gRNA will bind. The DNA-targeting segment of a gRNA can be modified (e.g., by genetic engineering) to hybridize to any desired sequence within a target DNA sequence.

[0166] The protein-binding segment of a guide RNA interacts with a site-directed modifying polypeptide (e.g. a Cas9 protein) to form a complex. The guide RNA guides the bound polypeptide to a specific nucleotide sequence within target DNA via the above-described DNA-targeting segment. The protein-binding segment of a guide RNA comprises two stretches of nucleotides that are complementary to one another and which form a double stranded RNA duplex.

[0167] In some aspects, a gRNA comprises two separate RNA molecules. In such aspects, each of the two RNA molecules comprises a stretch of nucleotides that are complementary to one another such that the complementary nucleotides of the two RNA molecules hybridize to form the double- stranded RNA duplex of the protein-binding segment. In some aspects, a gRNA comprises a single RNA molecule (sgRNA).

[0168] The specificity of a gRNA for a target loci is mediated by the sequence of the DNA- binding segment, which comprises about 20 nucleotides that are complementary to a target DNA sequence within a MAP3K15 nucleic acid sequence (GenBank Accession No. NG_021184.1 or NM_001001671.4). In some aspects, the corresponding target DNA sequence is approximately 20 nucleotides in length. In some aspects, the DNA-binding segments of the gRNA sequences provided herein are at least 90% complementary to a target DNA sequence within a MAP3K15 nucleic acid sequence (GenBank Accession No.

NG_021184.1 or NM_001001671.4). In some aspects, the DNA-binding segments of the gRNA sequences provided herein are at least 95%, 96%, 97%, 98%, or 99% complementary to a target DNA sequence within a MAP3K15 nucleic acid sequence. In some aspects, the DNA-binding segments of the gRNA sequences provided herein are 100% complementary to a target DNA sequence within a MAP3K15 nucleic acid sequence (GenBank Accession No. NG_021184.1 or NM_001001671.4).

Peptides

[0169] In some aspects, a MAP3K15 inhibitor is a peptide. A peptide is a sequence of amino acids of made up of a single chain of D- or L-amino acids or a mixture of D- and L-amino acids joined by peptide bonds. The amino acid subunits of the peptide may be naturally - occurring amino acids or may be non-naturally occurring amino acids. Many non-naturally occurring amino acids are known in the art and are available commercially. Further, the peptide bonds joining the amino acid subunits may be modified. See, e.g., Sigma-Aldrich; Gentilucci et al., Curr. Pharm. Des. 16: 3185-3203 (2010); US 2008/0318838. Generally, peptides contain at least two amino acid residues and are less than about 50 amino acids in length. In some aspects, peptide inhibitors may comprise or consist of between 3 and 50, between 5 and 50, between 10 and 50, between 10 and 40, between 10 and 35, between 10 and 30, or between 10 and 25 amino acids. In some aspects, peptide inhibitors may comprise 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, or 25 amino acids. In some aspects, peptide inhibitors may consist of 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, or 25 amino acids.

[0170] Methods of developing peptides that specifically bind a target molecule are known in the art, including phage display methods. See, e.g., US. Patent No. 5,010,175; WO 1996/023899; WO 1998/015833; Bratkovic, Cell. Mol. Life Sci., 67: 749-767 (2010); Pande et al., Biotech. Adv. 28: 849-858 (2010). In some aspects, following selection of a peptide, the peptide may be modified, e.g., by incorporating non-natural amino acids and/or peptide bonds.

[0171] Amino acids that are important for peptide inhibition can be determined, in some aspects, by alanine scanning mutagenesis. Each residue is replaced in turn with a single amino acid, typically alanine, and the effect on MAP3K15 inhibition is assessed. Truncation analyses may also be used to determine not only the importance of the amino acids at the ends of a peptide, but also the importance of the length of the peptide, on inhibitory activity. In some aspects, truncation analysis may reveal a shorter peptide that binds more tightly than the parent peptide. The results of various mutational analyses, such as alanine scanning mutagenesis and truncation analyses, may be used to inform further modifications of an inhibitor peptide.

[0172] One skilled in the art will appreciate that, in some aspects, one or more peptide inhibitors of MAP3K15 identified, for example, in an initial screen, can be modified in order to generate further peptide inhibitors with desirable properties, such as improved specificity for MAP3K15, stronger binding to MAP3K15, improved solubility, reduced toxicity, and/or improved cell membrane permeability.

Aptamers

[0173] In some aspects, a MAP3K15 inhibitor is an aptamer. The term “aptamer” as used herein refers to a nucleic acid molecule that specifically binds to a target molecule, such as MAP3K15. Aptamers can be selected to be highly specific, relatively small in size, and/or non-immunogenic. See, e.g., Ni, et al., Curr. Med. Chem. 18: 4206 (2011). In some aspects, a aptamer is a small RNA, DNA, or mixed RNA/DNA molecule that forms a secondary and/or tertiary structure capable of specifically binding and inhibiting MAP3K15.

[0174] In some aspects, an aptamer includes one or more modified nucleosides (e.g., nucleosides with modified sugars, modified nucleobases, and/or modified internucleoside linkages), for example, that increase stability in vivo, increase target affinity, increase solubility, increase serum half-life, increase resistance to degradation, and/or increase membrane permeability, etc. In some aspects, aptamers comprise one or more modified or inverted nucleotides at their termini to prevent terminal degradation, e.g., by an exonuclease.

[0175] The generation and therapeutic use of aptamers are well established in the art. See, e.g., U.S. Pat. No. 5,475,096. In some aspects, aptamers are produced by systematic evolution of ligands by exponential enrichment (SELEX), e.g., as described in Ellington et al., Nature 346: 818 (1990); and Tuerk et al., Science 249: 505 (1990). In some aspects, aptamers are produced by an AptaBid method, e.g., as described in Berezovski et al., J. Am. Chem. Soc. 130: 913 (2008). Slow off-rate aptamers and methods of selecting such aptamers are described, e.g., in Brody et al., Expert Rev. Mol. Diagn., 10: 1013-22 (2010); and U.S. Patent No. 7,964,356.

Exemplary assays

[0176] Any suitable assay in the art can be used to determine an activity, detect an outcome or effect, determine efficacy, etc. Certain non-limiting exemplary assays that can be used in the methods provided herein are described.

[0177] In some aspects, a method of determining whether an agent inhibits MAP3K15 activity comprises contacting MAP3K15 with a test substance and determining if the test substance inhibits MAP3K15.

Methods of Medical Treatment with MAP3K15 Inhibitors, Methods of Identifying Subjects that Would Benefit from Treatment with MAP3K15 Inhibitors, and Methods of Treating such Subjects

Methods of Medical Treatment with MAP3K15 Inhibitors

[0178] Various methods of treating or preventing diabetes mellitus in a subject with MAP3K15 inhibitors are provided herein. Provided herein are also methods of decreasing hemoglobin Ale (HbAlc) in a subject in need thereof (such as a subject with diabetes) with MAP3K15 inhibitors. Provided herein are also methods of preventing pancreatic cell death in a subject in need thereof with MAP3K15 inhibitors. In addition, provided herein are methods of treating or preventing a disease or condition (such as diabetes) in a subject with MAP3K15 inhibitors, wherein the disease or condition is caused or characterized by high blood sugar.

[0179] In some aspects, such methods comprise administering to a subject in need thereof a therapeutically or prophylactically effective amount of an inhibitor of MAP3K15.

[0180] In some aspects, the diabetes is type 2 diabetes. In some aspects, the diabetes is type 1 diabetes. In some aspects, a subject with diabetes type 1 has a C-peptide level of at least 0.2 nmol/L prior to administration of a MAP3K15 inhibitor. C-peptide levels, methods of determining C-peptide levels, and their relevance to diabetes are discussed, for example, in Leighton, E., et al, Diabetes Ther 5:475-487 (2017) and Rao, P, et al., J Diabetes Sci Technol doi: 10.1177/1932296821995557 (2021), each of which is herein incorporated by reference in its entirety.

[0181] In some aspects, provided herein is a method of treating diabetes in a subject comprising administration of a MAP3K15 inhibitor to a subject in need thereof, where the subject presents with diabetes, and in particular early stage diabetes wherein the protective effect on beta cells can prevent or substantially reduce disease progression.

[0182] In some aspects, a subject in need with a MAP3K15 inhibitor is a patient diagnosed as being in the prediabetic state or having a high risk of developing T2DM on the basis of having fasted glucose levels of 5.5 mmol/L to 6.9 mmol/L, or HbAlC level of 42 to 47 mmol/mol or 6.0 to 6.4% or an oral glucose tolerance test reading of 7.8 to 11.1 mmol/L (Type 2 diabetes: prevention in people at high risk | NICE Public Health Guideline 38 - NICE. Published July 12, 2012). In such aspects, the use of a MAP3K15 inhibitor can prevent the development of type 2 diabetes.

[0183] In some aspects, a subject in need of treatment with a MAP3K15 inhibitor can be a Type 1 diabetes patient presenting with a C-peptide level of at least 0.2 nmol/L. In such cases the c-peptide level indicates that the patient is still producing insulin, and the use of a MAP3K15 inhibitor can substantially prevent, or significantly reduce the extent of, beta cell death over time and the accompanying reduction of pancreatic insulin production capacity. In such aspects, the use of a MAP3K15 inhibitor can decrease (minimize) or even eliminate the need for insulin use.

[0184] In some aspects, a subject in need of treatment with a MAP3K15 inhibitor is be a Type 2 diabetes patient presenting with fasting plasma glucose levels of > 7 mmol/L or a HblAc level of greater than 47 mmol/L or > 6.4% or an oral glucose tolerance test reading >11.1 mmol/L.

[0185] Subjects presenting with diabetes often present with high BMI, for example recent research suggests that obese people, i.e. those with a BMI >30, are up to 80 times more likely to develop type 2 diabetes than those with a BMI of less than 22. Nonetheless significant populations of type 2 diabetes present with a low BMI, and it may be that such subjects have a greater tendency to have abnormal pancreatic function than subjects presenting with high BMI. Accordingly, in some aspects provided herein, a method for the treatment of T2DM comprises administering an inhibitor of MAP3K15 to a subject characteristically having a BMI of 25 or less, for example 20 or less or 18 or less.

[0186] In some aspects, the administration of an inhibitor of MAP3K15 reduces blood sugar in the subject. In some aspects, the administration reduces HbAlc by at least 10 mmol/mol, at least 15 mmol/mol, or at least 20 mmol/mol.

[0187] In some aspects, the HbAlc level post treatment is 48 mmol/mol or below. In some aspects, the HbAlc level post treatment is 42 mmol/mol or below. In some aspects, the HbAlc level post treatment is between 20 mmol/mol and 42 mmol/mol.

[0188] In some aspects, the HbAlc level in the subject prior to treatment is at least 48 mmol/mol. In some aspects, the HbAlc in the subject prior to treatment at least 63.8 mmol/mol.

[0189] In some aspects, administration of a MAP3K15 inhibitor reduces HbAlc in a subject to 48 mmol/mol or below, for example from 20 mmol/mol to 42 mmol/mol.

[0190] The subject in the methods of the present disclosure can be a male or female subject. In some aspects, the subject is male. In some aspects, the subject is female. In some aspects, the subject is a mouse. In some aspects, the subject is human.

[0191] In some aspects, when the subject is female, the subject can carry a single loss of function mutation in the gene encoding MAP3K15. In some aspects, this mutation is a truncation.

[0192] In some aspects, the subject being administered an inhibitor of MAP3K15 does not contain a loss of function mutation in the gene encoding MAP3K15.

[0193] The subject can be obese or non-obese based on a BMI value prior to the administration of an inhibitor of MAP3K15. In some aspects, the subject has a BMI of 30 or more prior to the administration. In some aspects, the subject has a BMI of at least 25 and less than 30 prior to the administration. In some aspects, the subject has a BMI of at least 18.5 and less than 25 prior to the administration. In some aspects, the subject has a BMI of less than 25 prior to the administration. In some aspects, the subject has a BMI of less than 20 prior to the administration. In some aspects, the subject has a BMI of less than 18.5 prior to the administration. [0194] In some aspects, the subject does not have a loss of function mutation in a MAP 3 KI 5 gene. In some aspects, the female subject does not have a loss of function mutation in a MAP 3 KI 5 gene or only has one MAP 3 KI 5 gene with a loss of function mutation.

[0195] In some aspects, the administration of an inhibitor of MAP3K15 reduces MAP3K15 activity in the subject by at least 40%, optionally wherein the administration reduces MAP3K15 activity in the subject by at least 50%, at least 60%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, or at least 95%.

[0196] In some aspects, an inhibitor of MAP3K15 is administered to the subject in combination with metformin. In some aspects, the subject has not achieved significant HbAlc reduction following treatment with metformin. In some aspects, the subject has an HbAlc level equal to or greater than 63.9 mmol/mol after treatment with metformin for at least 3 months.

Methods of Identifying a Subject for Treatment with MAP3K15 Inhibitors and Methods of Treating such Subjects

[0197] Provided herein are method of identifying subjects that that would benefit from the treatment with MAP3K15 inhibitors, and methods of treating such identified subjects.

[0198] Accordingly, a method of identifying a subject that would benefit from treatment with a MAP3K15 inhibitor comprises: i) sequencing a biological sample obtained from a subject; ii) analyzing the sequence to determine whether there is a loss of function mutation of MAP3K15; and iii) based on the result of ii), indicating whether the subject is indicated for treatment with a MAP3K15 inhibitor.

[0199] In some aspects, the subject is identified by using this method prior to the administration of an inhibitor of MAP3K15.

[0200] The subject identified to benefit from treatment with a MAP3K15 inhibitor can be a male subject or a female subject. The subject identified to benefit from the treatment with an inhibitor of MAP3K15 can be subsequently treated.

[0201] In some aspects, a method of treating a male subject that would benefit from treatment with a MAP3K15 inhibitor comprises: i) sequencing a biological sample obtained from a male subject; ii) analyzing the sequence to determine whether there is a loss of function mutation of MAP3K15; and iii) administering an inhibitor of MAP3K15 to the subject if there is not a loss of function of MAP3K15.

[0202] In some aspects, a method of treating a female subject that would benefit from treatment with a MAP3K15 inhibitor comprises: i) sequencing a biological sample obtained from a female subject; ii) analyzing the sequence to determine whether there is a loss of function mutation of MAP3K15; and iii) administering an inhibitor of MAP3K15 to the subject if there is not a loss of function of MAP3K15 or if there is only a loss of function of one MAP3K15-encoding gene.

[0203] In some aspects, the step of administering the inhibitor of MAP3K15 is performed as described herein. The sequencing of a biological sample and the analyzing of the sample can be performed by any suitable method known in the art.

[0204] In some aspects, a subject that would benefit from receiving a MAP3K15 inhibitor does not have a loss of function mutation in a MAP3K15 gene. In some aspects, a female subject a subject that would benefit from receiving a MAP3K15 inhibitor does not have a loss of function mutation in a MAP 3 KI 5 gene or only has one MAP 3 KI 5 gene with a loss of function mutation.

[0205] In some aspects, a subject that would benefit from receiving a MAP3K15 inhibitor has a BMI of less than 25 prior to the administration. In some aspects, a subject that would benefit from receiving a MAP3K15 inhibitor has a BMI of less than 20 prior to the administration. In some aspects, a subject that would benefit from receiving a MAP3K15 inhibitor has a BMI of at least 18.5 and less than 25 prior to the administration. In some aspects, a subject that would benefit from receiving a MAP3K15 inhibitor has a BMI of less than 18.5 prior to the administration

[0206] In some aspects, a subject that would benefit from receiving a MAP3K15 inhibitor has a BMI of 30 or more prior to the administration. In some aspects, a subject that would benefit from receiving a MAP3K15 inhibitor has a BMI of 25 or more prior to the administration. In some aspects, a subject that would benefit from receiving a MAP3K15 inhibitor has a BMI of at least 25 and less than 30 prior to the administration.

[0207] In some aspects, a subject that would benefit from receiving a MAP3K15 inhibitor has an HbAlc level of at least 42 mmol/mol. In some aspects, a subject that would benefit from receiving a MAP3K15 inhibitor has an HbAlc level of at least 48 mmol/mol. In some aspects, a subject that would benefit from receiving a MAP3K15 inhibitor has an HbAlc level of at least 63.8 mmol/mol.

[0208] In some aspects, a subject that would benefit from receiving a MAP3K15 inhibitor has type 2 diabetes. In some aspects, a subject that would benefit from receiving a MAP3K15 inhibitor has type 1 diabetes. In some aspects, a subject that would benefit from receiving a MAP3K15 inhibitor has type 1 diabetes and a C-peptide level of at least 0.2 nmol/L.

Pharmaceutical Compositions and. Routes of Administration

[0209] In some aspects, pharmaceutical compositions comprising one or more of MAP3K15 inhibitors are provided. Such pharmaceutical compositions may be formulated with a wide variety of pharmaceutically acceptable carriers (see, for example, Gennaro, Remington: The Science and Practice of Pharmacy with Facts and Comparisons: Drugfacts Plus, 20th ed. (2003); Ansel et al., Pharmaceutical Dosage Forms and Drug Delivery Systems, 7th ed., Lippencott Williams and Wilkins (2004); Kibbe et al., Handbook of Pharmaceutical Excipients, 3rd ed., Pharmaceutical Press (2000)). Various pharmaceutically acceptable carriers, which include vehicles, adjuvants, and diluents, are available. Moreover, various pharmaceutically acceptable auxiliary substances, such as pH adjusting and buffering agents, tonicity adjusting agents, stabilizers, wetting agents and the like, are also available. Nonlimiting exemplary carriers include saline, buffered saline, dextrose, water, glycerol, ethanol, and combinations thereof.

[0210] In some aspects, MAP3K15 inhibitors can be administered in vivo by various routes, including, but not limited to, oral, intravenous, intra-arterial, parenteral, intraperitoneal, or subcutaneous. In some aspects, the administration is oral. In some aspects, the administration is intravenous.

Combination Therapy

[0211] In some aspects, methods of the disclosure comprise administering a MAP3K15 inhibitor in combination with an additional therapeutic agent. In some aspects, the additional therapeutic agent is an anti-diabetic drug. In some aspects, the anti-diabetic drug is approved for hyperglycemic treatment in type 2 diabetes mellitus. In some aspects, the anti-diabetic drug is metformin. [0212] Such administration can be concurrent or sequential. In some aspects, concurrent administration is administration that is close enough in time that the activity of the two or more therapeutic agents overlap for some period of time. In other words, concurrent administration need not be at the same time, but can be separated in time so long as there is an overlap in therapeutic activity of the concurrently administered agents. In some aspects, a MAP3K15 inhibitor that specifically binds to MAP3K15 is administered in combination with an anti-diabetic drug.

[0213] There are different classes of anti-diabetic drugs, and their selection depends on the nature of the diabetes, age and situation of the person, as well as other factors. In some aspects, the anti-diabetic drug is selected from biguanidines (such as, e.g., metformin), thiazolidinediones (such as, e.g., rosiglitazone, pioglitazone, and troglitazone), sylfonylureas (such as, e.g., tolbutamide, acetohexamide, tolazamide, chlorpropamide, glipizide, glyburide, glimepiride, gliclazide, glyclopyramide, and gliquidone), meglitinides (such as, e.g., repaglinide and nateglinide), alpha-glucosidase inhibitors (such as, e.g., miglitol, acarbose, and voglibose), sodium-glucose cotransporter 2 (SGLT-2) inhibitors (such as, e.g., dapagliflozin, canagliflozin, empagliflozin, and remogliflozin), dipeptidyl peptidase IV (DPP-IV) inhibitors (such as, e.g., sitagliptin, vildagliptin, saxagliptin, and linagliptin, and glucagon-like peptide- 1 (GLP-1).

[0214] In some aspects, the anti-diabetic drug is metformin which is an oral anti-diabetic drug in the biguanide class approved for the treatment of type 2 diabetes mellitus. Metformin is suitable for use in overweight and obese people and those with normal kidney function. Metformin is sold under several trade names in the U.S., such as Glucophage®, Riomet®, Fortamet®, and Glumetza®.

[0215] In some aspects, methods of the disclosure comprise administering a MAP3K15 inhibitor in combination with in combination with diet and exercise. Examples

Example 1: Materials and Methods

Cohorts

[0216] Discovery genetic association studies were performed using the 454,796 exomes available for the UK Biobank (UKB) cohort. The UKB is a prospective study of approximately 500,000 participants aged 40-69 years at time of recruitment. Participants were recruited in the UK between 2006 and 2010 and are continuously followed. The average age at recruitment for sequenced individuals was 56.5 years, and 54% of the sequences were obtained from females. Participant data include health records that are periodically updated by the UKB, self -reported survey information, linkage to death and cancer registries, collection of urine and blood biomarkers, imaging data, accelerometer data and various other phenotypic end points.

[0217] Replication studies were performed in the Mexico City Prospective Study (MCPS) cohort and the FinnGen cohort. MCPS cohort study includes people at least 35 years of age recruited from two contiguous urban districts in Mexico City between 1998 and 2004. Exome sequencing data and clinical and quantitative traits information was available in -141,000 participants. The FinnGen cohort (release 7) includes -356,000 individuals from Finland with genotype and health registry data (www.finngen.fi/en/about). Phenotypes have been derived from nationwide health registries. Patients and control subjects in FinnGen provided informed consent for biobank research, based on the Finnish Biobank Act. Alternatively, older research cohorts, collected prior the start of FinnGen (in August 2017), were collected based on study-specific consents and later transferred to the Finnish biobanks.

Phenotypes

[0218] The UKB phenotype data was harmonized as previously described. Briefly, PEACOK and union mapping were used to parse binary and quantitative traits included in the February 2020 UKB release (accessed March 27, 2020; UKB application 26041). 92 binary and 5 quantitative traits related to diabetes were considered. The association analyses for systolic blood pressure (SBP) and diastolic blood pressure (DBP) were adjusted for commonly prescribed blood pressure medications in the UKB. Genetic Data

[0219] Whole-exome sequencing data for 454,988 UKB and additional MCPS participants were generated at the Regeneron Genetics Center (RGC). Genomic DNA underwent paired- end 75-bp whole-exome sequencing at Regeneron Pharmaceuticals using the IDT xGen vl capture kit on the NovaSeq6000 platform. Conversion of sequencing data in BCL format to FASTQ format and the assignments of paired-end sequence reads to samples were based on 10-base barcodes, using bcl2fastq v2.19.0. Initial quality control included sex discordance, contamination, unresolved duplicate sequences, and discordance with microarray genotyping data checks.

[0220] In FinnGen, genotyping of the samples was done using a ThermoFisher Axiom custom array. In addition to the core GWAS markers (about 500,000), it contains 116,402 coding variants enriched in Finland, 10,800 specific markers for the HLA/KIR region, 14,900 ClinVar variants, 4,600 pharmacogenomic variants, and 57,000 additional markers. AstraZeneca Centre for Genomics Research (CGR) Bioinformatics Pipeline

[0221] The UKB exome sequences and MCPS exome sequences were processed at AstraZeneca from their unaligned FASTQ state. A custom-built Amazon Web Services (AWS) cloud compute platform running Illumina DRAGEN Bio-IT Platform Germline Pipeline v3.0.7 was used to align the reads to the GRCh38 genome reference and perform single-nucleotide variant (SNV) and insertion and deletion (indel) calling. SNVs and indels were annotated using SnpEFF v4.3 (Cingolani, P. et al. A program for annotating and predicting the effects of single nucleotide polymorphisms, SnpEff: SNPs in the genome of Drosophila melanogaster strain wl l l8; iso-2; iso-3. Fly 6, 80-92 (2012) against Ensembl Build 38.92. All variants were further annotated with their genome Aggregation Database (gnomAD) MAFs (gnomAD v2.1.1 mapped to GRCh38). Missense variants were also annotated with MTR (Traynelis, J. et al. Optimizing genomic medicine in epilepsy through a gene-customized approach to missense variant interpretation. Genome research 27, 1715— 1729 (2017)) and REVEL scores (loannidis, N. M. et al. REVEL: An Ensemble Method for Predicting the Pathogenicity of Rare Missense Variants. American journal of human genetics 99, 877-885 (2016)).

Additional Quality Control [0222] To complement the quality control, the UKB and MCPS exome sequences were passed through the AstraZeneca CGR bioinformatics pipeline as previously described. Briefly, the analyses excluded sequences that achieved a VerifyBAMID freemix (contamination) level of more than 4%, and additional sequences where less than 94.5% of the consensus coding sequence (CCDS release 22) achieved a minimum of tenfold read depth. The cohort was screened with KING v2.2.3 to remove related individuals (less than third-degree).

Genetic Ancestry

[0223] The primary discovery analysis was performed in UKB participants of European ancestry. Genetic ancestries were predicted from the exome data using PEDDY vO.4.2 with the ancestry labelled 1,000 Genomes Project as reference (Auton, A. et al. A global reference for human genetic variation. Nature vol. 526 (2015)). Some of the 454,796 UKB sequences had a probability of European ancestry prediction less than 0.99. Focusing on the remaining UKB participants, the European ancestry cohort was further restricted to those within +4SD across the top four principal component means. This resulted in the exclusion of additional outlier participants. In total, there were 394,695 predominantly unrelated participants of European ancestry who were included in the European case-control analyses. PEDDY- derived ancestry predictions were also used to perform case-control collapsing analyses within non-European populations where there were at least 1,000 exome -sequenced individuals available (see ‘Collapsing analyses’ description below). Through this step, 7,412 (Pr( African) > 0.95), 2,209 (Pr(East Asian) > 0.95), and 8,078 (Pr(South Asian) > 0.95) UKB participants were identified for ancestry-independent collapsing analyses.

Discovery Associations

Collapsing Analyses

[0224] A previously described gene-level collapsing analysis framework was performed for 92 binary and 5 quantitative phenotypes related to diabetes. Ten (10) non-synonymous collapsing models were used, including 9 dominant and one recessive model, plus an additional synonymous variant model as an empirical negative control. For the dominant collapsing models, the carriers of at least one qualifying variant (QV) in a gene were compared to non-carriers. In the recessive model, an indiviual must have two qualifying alleles either in homozygous or putatively compound heterozygous form. Hemizygous genotypes for X chromosome genes also qualified for the recessive model.

[0225] Using SnpEff annotations, synonymous variants were defined as those annotated as ‘synonymous variant’. PT Vs were defined as variants annotated as exon loss variant, frameshift_variant, start_lost, stop_gained, stop_lost, splice_acceptor_variant, splice_donor_variant, gene_fusion, bidirectional_gene_fusion, rare_amino_acid_variant, and transcript_ablation. Missense was defined as: missense_variant_splice_region_variant, and missense_variant. Non-synonymous variants included: exon_loss_variant, frameshift_variant, start_lost, stop_gained, stop_lost, splice_acceptor_variant, splice_donor_variant, gene_fusion, bidirectional_gene_fusion, rare_amino_acid_variant, transcript_ablation, conservative_inframe_deletion, conservative_inframe_insertion, disruptive_inframe_insertion, disruptive_inframe_deletion, missense_variant_splice_region_variant, missense_variant, and protein_altering_variant.

[0226] For binary traits, the difference in the proportion of cases and controls carrying QVs in a gene was tested using a Fisher’s exact two-sided test. For quantitative traits, the difference in mean between the carriers and non-carriers of QVs was determined by fitting a linear regression, correcting for age, sex, and median intake (for SBP and DBP). For all models, the following quality control filters were applied: minimum coverage 10X; annotation in CCDS transcripts (release 22; approximately 34 Mb); at most 80% alternate reads in homozygous genotypes; percent of alternate reads in heterozygous variants > 0.25 and < 0.8; binomial test of alternate allele proportion departure from 50% in heterozygous state P > 1 x 10-6; GQ > 20; FS < 200 (indels) < 60 (SNVs); MQ > 40; QUAL > 30; read position rank sum score > -2; MQRS > -8; DRAGEN variant status = PASS; the variant site achieved tenfold coverage in > 25% of gnomAD exomes, and if the variant was observed in gnomAD exomes, the variant achieved exome z-score > -2.0 and exome MQ > 30. Genes previously found associated with batch effects (Wang, Q. el al. Rare variant contribution to human disease in 281,104 UK Biobank exomes. Nature (2021) doi:10.1038/s41586-021- 03855-y) were excluded. Pan-ancestry Collapsing Analyses

[0227] Additional collapsing analysis in each individual non-European ancestral population were performed as described above. Using a previously introduced approach (see Wang, Q, 2021 ibid), a pan-ancestry analysis was performed using a Cochran-Mantel-Haenszel test to generate combined 2x2xN stratified P-values, with N representing up to all four genetic ancestry groups. For the quantitative traits, a linear regression model was used that included the following covariates: categorical ancestry (European, African, East Asian or South Asian), the top five ancestry principal components, age, and sex.

ExWAS Analysis

[0228] Variant-level association tests were performed in addition to the gene-level collapsing analyses for the 92 binary and 5 quantitative traits related to diabetes. Variants identified in at least six individuals from the predominantly unrelated European ancestry UKB exomes as previously described were tested. In summary, variants were required to pass the following quality control criteria: minimum coverage 10X; percent of alternate reads in heterozygous variants > 0.2; binomial test of alternate allele proportion departure from 50% in heterozygous state P > 1 x 10’⁶; genotype quality score (GQ) > 20; Fisher’s strand bias score (FS) < 200 (indels) < 60 (SNVs); mapping quality score (MQ) > 40; quality score (QUAL) > 30; read position rank sum score (RPRS) > -2; mapping quality rank sum score (MQRS) > -8; DRAGEN variant status = PASS; variant site is not missing (that is, less than 10X coverage) in 10% or more of sequences; the variant did not fail any of the aforementioned quality control in 5% or more of sequences; the variant site achieved tenfold coverage in 30% or more of gnomAD exomes, and if the variant was observed in gnomAD exomes, 50% or more of the time those variant calls passed the gnomAD quality control filters (gnomAD exome AC/AC_raw > 50%). P values were generated adopting a Fisher’s exact two-sided test. Three distinct genetic models were studied for binary traits: allelic (A versus B allele), dominant (AA + AB versus BB) and recessive (AA versus AB + BB), where A denotes the alternative allele and B denotes the reference allele. For quantitative traits, we adopted a linear regression (correcting for age, sex and age x sex) and replaced the allelic model with a genotypic (AA versus AB versus BB) test. Phenome-wide analysis for MAP 3 KI 5 Protein-Truncating Variants (PTVs)

[0229] A phenome-wide association study (PheWAS) of PTVs in MAP3K15 was performed with binary and quantitative traits in UKB participants of European ancestry. This was done for the “ptv,” “ptv5pcnt,” and the “rec” collapsing models. These phenotype data were harmonized and union mapped as previously described. All 11 collapsing models were included in the PheWAS as described above. The methodology used here was the previously published PheWAS on -300,000 UKB participants.

P -Value Threshold

[0230] A study-wide significance threshold was defined as p<lxl0^-8.

Replication Analyses

[0231] The MCPS and FinnGen cohorts were used to replicate the association between MAP3K15 and diabetes and HbAlC levels. For the FinnGen cohort, association statistics for phenome-wide associations of non-synonymous variants within MAP3K15 through the FinnGen portal were accessed. For MCPS, a collapsing analysis was performed using the recessive model for HbAlc and available diabetes-related clinical phenotypes.

Secondary Association Analysis

[0232] A total of 44 unique PTVs in MAP3K15 were observed in hemizygous male carriers. Two of these PTVs (Argl l22* and Argl l36*) were relatively more frequent. Carriers of these two alleles excluded, and the following analyses were re -performed: Fisher’s exact test for diabetes (‘ 20002# 1220#diabetes’) and linear regression for HbAlc.

[0233] To determine whether the effect of complete loss of MAP3K15 on diabetes is mediated via adiposity or the insulin resistance pathway, additional analyses were performed in which HbAlc and the diabetes phenotype (‘ 20002# 1220#diabetes’) were regressed on MAP3K15 PTV carrier status in males, with age and BMI as covariates.

[0234] To investigate the joint effects of complete loss of MAP3K15 and a significantly associated indel in PDHA1 (X-19360844-AAC-A), a gene that overlaps the 3’-UTR of MAP3K15, HbAlc and the diabetes phenotype (‘ 20002# 1220#diabetes’) were regressed on carrier status for the MAP3K15 PTVs (Argl 122* and Argl 136*) and the PDHA1 indel in males, with age as a covariate.

Expression Analysis

[0235] Previously published bulk RNA-sequencing data available from a mouse insulinoma cell line (P-TC-6) transfected with three different clones carrying MODY -associated variants in NKX6-1 were leveraged. The DESeq2-derived log fold changes, p-values, and FDR values were extracted from the supplementary data. Tissue expression using the GTEx portal (http://gtexportal.org/home/) was determined. For single-cell RNA-sequencing analysis, a collection of eight previously published datasets using tissue from human pancreatic islets spanning 27 healthy donors, five technologies, and four laboratories was examined. Data was integrated using Seurat as previously described (Stuart et al., 2019, Cell 177, 1888-1902).

Gene-SCOUT

[0236] The tool Gene-SCOUT estimates similarity between genes by leveraging association statistics from the collapsing analysis across -1,500 quantitative traits available in the UKB. This tool was utilized to identify genes that were most similar to the ‘seed gene’ MAP3K15.

Mantis-ML

[0237] Mantis-ML (Vitsios, D. & Petrovski, S. Mantis-ml: Disease-Agnostic Gene Prioritization from High-Throughput Genomic Screens by Stochastic Semi-supervised Learning. American journal of human genetics 106, 659-678 (2020)) is a gene prioritization machine learning framework, integrating a diverse set of annotations, including intolerance to variation, tissue expression, and animal models. This tool was used to obtain the top disease predictions for MAP3K15 across 985 diseases parsed from the Human Phenotype Ontology (HPO).

Example 2: Cohort Characterization and Study Design

[0238] Exome sequences from 454,796 UKB participants were processed through our previously described cloud-based pipeline. Through stringent quality control, samples with low sequencing quality, low depth of coverage, and from closely related individuals were removed (see Example 1). Association tests were performed on 92 binary clinical phenotypes related to T1DM and T2DM. There was a total of 56,801 cases of European ancestry for the diabetes-related clinical phenotypes. Additionally, 4,724 individuals of South Asian, East Asian, and African ancestries had a diabetes-related diagnosis. Association tests were also performed for quantitative traits related to diabetes, including blood glucose, glycosylated haemoglobin (HbAlc), body mass index (BMI), systolic blood pressure (SBP) and diastolic blood pressure (DBP).

Example 3: Collapsing Analysis Results

[0239] A gene-level collapsing framework was employed to test the aggregate effect of rare functional variants in each gene (N=18,762) with diabetes-related clinical phenotypes (Example 1). Each gene-phenotype combination was tested under 10 non-synonymous collapsing models (nine dominant and one recessive model) to evaluate a range of genetic architectures .

[0240] Six genes significantly associated (p<lxl0^-8) with at least one diabetes-related clinical phenotype in individuals of European ancestry (N=394,695), amongst which MAP3K15) was associated with a protective signal.

Table 1: Association of MAP3K15 with diabetes-related clinical phenotypes in the gene-level collapsing analysis

Genes across the exome were tested under 10 different non-synonymous collapsing models with diabetes-related clinical phenotypes available in the UK Biobank. The most significant phenotypic term, the most significant collapsing model, and the corresponding association statistics have been provided for the six genes that were significantly associated (p<lxl0^-8) with at least one diabetes-related clinical phenotype. (Chr=Chromosome; QV=Qualifying Variant; OR=Odds Ratio, CI=Confidence Intervals) [0241] An intronic variant in MAP3K15 was associated with T2DM in a large trans-ethnic GWAS4, but functional variation in this gene had not been previously linked to diabetes. A recessive collapsing model detected associations for homozygous or potential compound heterozygous carriers of rare (minor allele frequency (MAF) < 0.5%) nonsynonymous variants in MAP3K15 with diabetes-related phenotypes. The 5,118 carriers observed for the 353 distinct qualifying variants (QVs) in MAP3K15 for this model had a significantly lower risk of diabetes (phenotype: ‘Union#E14#E14 Unspecified diabetes mellitus’ : OR=0.70 [0.62,0.79], P=5.0xl0^-9). Consistent with this observation, the MAP3K15 QV carriers also had significantly lower HbAlc levels (Figure lb, beta=-0.14 [-0.16,-0.11], P=3.1xl0^-23) and blood glucose levels (beta=-0.13 [-0.16, -0.10], P=2.5xl0 ¹⁷). This reduction in HbAlc and blood glucose levels is relative to non-diabetic controls - i.e. subjects with levels of HbAlc and blood glucose in the normal range.

[0242] To perform a pan-ancestry analysis, individual gene-level collapsing analysis from each major ancestral group (South Asian, East Asian, African, European) were performed and combined via a Cochran-Mantel-Haenszel test. In this analysis, associations between MAP3K15 and diabetes (‘Union#E14#E14 Unspecified diabetes mellitus: OR=0.70, [0.62,0.79], P=5.7x10-10), HbAlc (beta=-0.14 [-0.16,-0.11], P=l.1x10-24) and blood glucose (beta=-0.13 [-0.15, -0.10], P=5.5x10-18) remained significant for the recessive collapsing model.

Table 2

[0243] Next, the effect of complete loss of MAP3K15 on diabetes was determined. Since MAP3K15 resides on chromosome X, hemizygous male PTV carriers (i.e., complete loss) were compared to male non-carriers. The 1,216 hemizygous male MAP3K15 PTV carriers of European ancestry had 40% reduced odds of developing diabetes (‘ 20002# 1220#diabetes’ : OR=0.60 [0.45,0.81], P=0.0007) and 0.21 standard deviation (SD) units lower HbAlc levels (beta=-0.21 [-0.26, -0.15], P=1.2xl0^-11) (Table 3). The decrease in HbAlc levels and diabetes risk observed with complete loss of MAP3K15 in hemizygous male carriers was significantly stronger than with partial (50%) loss of MAP3K15 observed in heterozygous female carriers (Figures 1C and ID). This indicates an additive effect of MAP3K15 loss on HbAlc levels and diabetes risk.

Table 3

[0244] Two PTVs in MAP3K15 were relatively more frequent among the hemizygous male carriers: Argl l22* (MAF=0.1%) and Argl l36* (MAF=0.4%) (Figure IE; Table 4).

Table 4

[0245] Both variants were independently associated with reduced risk of diabetes (Argl l22*: OR = 0.33,95% CI: [0.13,0.80], p = 0.015; Argl l36*: OR = 0.60, 95% CI: [0.41,0.88], p = 0.010) (Table 5). Both were also associated with reduced Hbalc levels (Argl l22*: beta = -0.30, 95% CI: [-0.44, -0.15], p = 4.32x10-5; Argl l36*: beta = -0.20, 95% CI: [-0.27, -0.12], p = 1.26xl0^-6). When excluding these two common PTVs from the collapsing test, the association between the remaining recessive PTVs in MAP3K15 remained significantly associated with reduced HbAlC (beta = -0.16, 95% CI: [-0.28, -0.05], p = 0.005) and nominally associated with reduced diabetes risk, likely due to reduced power (beta = 0.84, 95% CI: [0.50,1.39], p = 0.49) (Table 5)). Altogether, these results further support loss of MAP3K15 function as the mechanism for the protective effect against diabetes.

Table 6

[0246] MAP3K15 overlaps with the 3’-UTR of PDHA1, a gene that encodes a subunit of the enzyme pyruvate dehydrogenase, which catalyzes a step in the glycolysis pathway. Moreover, an indel (X-19360844-AAC-A) in the 3’-UTR of PDHA1 is significantly associated with HbAlc levels (beta=-0.06, P=l.0x10-17). The joint effects of the PDHA1 indel and the two MAP3K15 PTVs on HbAlc levels were examined. The associations for all three variants with HbAlc remained significant in the joint analysis (Table 5), suggesting that PTVs in MAP3K15 influence HbAlc levels and diabetes risk independent of the nearby PDHA1 locus.

Example 4: Replication Studies

[0247] Replication studies of the MAP3K15 associations were performed using data from two genetically diverse populations: Mexico City Project (MCP) (N=139,511) and FinnGen (N=309,154). In MCP, MAP3K15 was associated with HbAlc in the recessive collapsing model (beta=-0.08, P=0.004). The association between the recessive model and diabetes did not achieve significance, though the effect was in the protective direction (P = 0.1438, OR = 0.87, 95% CI: 0.72-1.04). In FinnGen (release 7), a missense variant (rs56381411) and a PTV (rsl40104197) were independently associated with decreased risk of T2DM (OR=0.89, P=2.1x10-5) and T1DM (OR=0.32, P=4.4x10-5), respectively. This indicates that loss of MAP3K15 is protective against both T1DM and T2DM. Collectively, these replications strongly support the association between variation in MAP3K15 with reduced blood glucose, HbAlc, and risk of diabetes.

Example 5: Association of MAP3K15 with Other Phenotypes

[0248] Obesity, which can lead to increased insulin resistance, is a strong risk factor for T2DM. To investigate whether the effect of MAP3K15 on diabetes is mediated via adiposity or the insulin resistance pathway, the effect of complete loss of MAP3K15 was tested adjusting for BMI. The associations between MAP3K15 PTVs and both HbAlc (unadjusted: beta=-0.21, P=1.2xl0^-11; BMI-adjusted: beta=-0.20, P=l.lxl0^-11) and diabetes (unadjusted: OR=0.60, P=0.0007; BMI-adjusted: OR=0.59, P=0.0005) remained essentially unchanged on adjusting for BMI, indicating that the protective effect of loss of MAP3K15 on diabetes is likely not mediated via the insulin resistance pathway. [0249] To explore possible adverse phenotypic effects of MAP3K15 loss, associations between variation in MAP3K15 with roughly 15,000 clinical phenotypes in UKB participants were surveyed. This includes 0.6% European ancestry males with lifetime loss of MAP3K15. No significant adverse phenotypic associations (P<lxl(F⁸) were observed in individuals with MAP3K15 loss (“ptv”, “ptv5pcnf ’ and “rec” collapsing models) in either the individual ancestry or the pan-ancestry analyses. A number of phenotypic associations observed with MAP3K15 loss in a less conservative p-value range of lxlO’⁸<P<lxlO’⁴. This included an association with a hypertension phenotype observed in the South Asian ancestry participants for the recessive collapsing model (‘41202#BlockI10-I15#H0-I15 Hypertensive diseases’ : OR=6.3, P=4.8xl0^-5). This association appears to be preferentially driven by missense variation, with 7 missense carriers and 1 PTV carrier. Intriguingly, one of the MAP3K15 missense variants that protects against diabetes in FinnGen (rs56381411) was also associated with hypertension, but no single PTV in FinnGen was associated with hypertension. A hypertensive phenotype has been previously reported in Map3kl5 knockout mice. However, in European UKB participants, which represents the most well-powered cohort in this study, complete loss of MAP3K15 was not significantly associated with quantitative blood pressure measurements (SBP or DBP), and the odds ratios were in fact in the protective direction (Figure 1C). Complete loss was also not associated with a diagnosis of hypertension (‘Union#I10#I10 Essential (primary) hypertension’ : OR=0.90, P=0.10), or the hypertensive phenotype with a suggestive association in the South Asian ancestry participants (‘41202#BlockI10-I15#I10-I15 Hypertensive diseases”: OR=, P=) (Figures 1C and ID).

Example 6: Supporting Evidence

[0250] Tissue expression profile from the GTEx database demonstrates high expression of the MAP3K15 transcript in the adrenal glands. The transcript is also expressed at lower levels in spleen, kidney, pancreas, and pituitary glands (Figure 2A). Additionally, single-cell expression data from pancreatic endocrine cells indicate that MAP3K15 is expressed in several islet cell subpopulations, including a, b and d cells (Figure 2B). The expression profile for MAP3K15 indicates that its effect on diabetes could be mediated through an involvement in pancreatic islet cell function and/or stress response. The latter has been reported as an important physiological function for other ASK (apoptosis signal-regulated kinase) genes, the family to which MAP3K15 belongs.

[0251] A previous study performed RNA-sequencing on a mouse insulinoma cell line that stably expressed two variants associated with Maturity Onset Diabetes of the Young (MODY), autosomal dominant form of noninsulin-dependent diabetes diagnosed at a young age. This study specifically focused on MODY -associated mutations, and one functionally impaired control variant in the gene NKX6-1. MAP3K15 was the most strongly upregulated gene across all three perturbations.

[0252] Orthogonal evidence derived from Gene-SCOUT and phenome-wide mantis-ml provide further support for the role of MAP3K15 in diabetes. Gene-SCOUT enables the clustering of genes based on the effects estimated in collapsing analyses performed across around 1,500 quantitative traits using UKB exome sequences. This tool suggests that variation in SLC30A8, a zinc transporter gene (ZnT8) expressed in pancreatic islet P- and a- cells, results in a biomarker profile most similar to MAP3K15 (Figures 3A and 3B). Similar to MAP3K15, loss of SLC30A8 is known to protect against T2DM, potentially via increased glucose responsiveness. Mantis-ml, an automated machine-learning framework designed to identify novel gene-phenotype relationships based on publicly available disease-specific features (such as tissue specific expression, genic intolerance, and others), suggests MAP3K15 may have a role in disorders related to impaired glucose homeostasis, including “diazoxide-resistant diffuse hyperinsulinism” and “hyperinsulinemic hypoglycaemia” (Figure 3C.

Example 7: Inhibition of MAP3K15 (ASK3) for Improvement of Pancreatic Beta-Cell Survival and Function in Diabetes

[0253] MAP3K15, also known as Apoptosis signal-regulating kinase 3 (ASK3), is a member of the MAP3K family. Proteins in this family are activated in response to an array of stressors controlling diverse cellular functions such as proliferation, differentiation, and apoptosis. As demonstrated above, loss of function (LoF) of MAP3K15 reduces clinical diabetes risk by 40%, and there is no association between MAP3K15 LoF and body mass index (BMI), indicating that MAP3K15 is unlikely to be involved in the regulation of glucose-sensitive insulin secretion under basal conditions. Type 2 diabetes is characterised by pancreatic beta-cell dysfunction and death and is associated with low grade inflammation. Emerging evidence suggest that MAP3K signaling (MAP3K5 and MAP3K8) is implicated in beta-cell death in response to pro -inflammatory cytokines and glucolipotoxicity. MAP3K15 may act as an upstream regulator of the activation of p38 MAPK and c-Jun N-terminal kinase (JNK), effector MAPKs that are known to control beta-cell function and apoptosis. Accordingly, inhibition of MAP3K15 could improve beta-cell survival and function in (Type 2) diabetes. To demonstrate the role of MAP3K15 in beta-cell survival, the following studies were designed.

[0254] To demonstrate the role of MAP3K15 in basic metabolic function, its physiological role in development and full body metabolism (including pancreatic function), mouse models are generated for both total MAP3K15 gene deletion as well as tissue specific/inducible MAP3K15 gene deletion.

[0255] In the total gene deletion case, MAP3K15 knock out (KO) mice are generated by using the mRNA guided Crispr Cas9 technology, using Streptococcus pyogenes Cas9 nuclease (SpCas9). Genetically modified mice were generated by introducing a deletion in exon 5 of the Map3K15 gene. The deletion generated a preterm stop and was introduced using two, synthetic guide, sgRNA guides (Sigma Aldrich), (Crl binding site CTTTCGCCAGTTCATCACCT (SEQ ID NO: 5); Cr2 binding site TCATACCTGGATATCACGGT (SEQ ID NO:6)) and SpCas9 (IDT, Integrated DNA technologies, Alt-R® S.p. Cas9 Nuclease V3, 500 pg cat no: 1081059). The Ribonuclear particle (RNP) was made up by mixing the sgRNAs and spCas9 in 100 mM KC1; 20 mM Hepes buffer, and electroporated into Zona intact C57bl/6N mouse zygotes, Janvier Laboratories, using the BioRAD Gene Pulser X-cell TM electroporation system (BioRad Square wave protocol; Voltage 30 V, Pulse length 3 ms, 10 pulses, pulse interval 100 ms, 1mm Cuvette). After electroporation, the embryos were kept in a water jacket CO2 incubator in M2 medium overnight. Two (2) cell stage embryos were implanted into pseudo pregnant B6D2F1/Crl females. After birth, offspring were analysed for deletions using fragment analysis and subsequent Sanger sequencing (Applied Biosystems, ABI 3730 sequencer) to confirm the correct targeting. These mice, that have total MAP3K15 KO, are also referred to as hemizygous or “Hem” males, whereas their “normal” counterparts with no MAP3K15 KO are referred to as wild type (WT). [0256] The MAP3K15 KO mice on chow diet (R70, Lantmannen) are used to demonstrate that loss of MAP3K15 function does not have severe adverse effects in vivo (see schematic in Figure 4, Study 1). As noted above, no adverse effect in the human population was associated with MAP3K15 KO. General health, morphologic appearance, normal cage behaviour, and gross neurological functions of the MAP3K15 KO mice and their wild type littermates are studied at 4-7 weeks of age as previously described (Gerdin AK, et al., Biochem Biophys Res Commun. 349(2):825-32 (2006)). At 8-9 weeks of age, several behavioural tests are performed to evaluate if there are any neurological off-target effects when knocking out MAP3K15. Open field tests are performed to assess activity; a zero maze test is performed to assess anxiety and passive avoidance test to assess memory and learning (id). Electrocardiography (ECG) is performed on awake unstrained mice that are standing on a platform with electrodes during the measurement (Monroe LL, et al., Shock 46(6):723-30 (2016)). Metabolic phenotyping includes body composition measured by Dual Energy X-ray Absorptiometry (DEXA) and energy expenditure measured by indirect calorimetry (Gerdin AK, et al., Biochem Biophys Res Commun. 349(2):825-32 (2006)). Furthermore, glucose metabolism is measured by an oral glucose tolerance test (Monroe LL, et al., Shock 46(6):723-30 (2016)) and insulin tolerance test (Khatib Shahidi R, et al., PLoS One 7<5(2 :e0247300 (2021)) at 13-14 weeks of age, followed by termination and tissue collection. These tests demonstrate the ability of loss of MAP3K15 function to protect against diabetes associated with a high fat and high carbohydrate diet.

[0257] The study described above is repeated for MAP3K15 KO and wild type mice on a high fat high sugar (HFHS) diet (D 12492, Research Diets Inc) where the HFHS diet is given from 8 weeks of age and the tests in adult age start at 16 weeks of age (see schematic in Figure 4, Study 2). The same test battery as for the mice on chow diet is used to demonstrate that the MAP3K15 KO mice on HFHS diet are protected against diabetes and have a better glucose tolerance than their wild type littermates.

[0258] In a further related study (see schematic in Figure 4, Study 3) the blood pressures of the MAP3K15 KO and wild type mice are assessed by tail-cuff measurements (Bjursell M, et al., Diabetes 56(3): 583-93 (2007)) before and after a period of diet with increased salt content (Custom made 4% NaCl, Research Diets Inc) thus establishing that MAP3K15 KO mice have an altered blood pressure compared to the WT mice when challenged with a high salt diet as previously reported (id).

[0259] For the generation of the Map3kl5fl/fl line, a conditional knock-out strategy flanking critical Map3kl5 exons with LoxP sites, was used to target the Map3kl5 locus. The targeting vector was built using homologous recombination in bacteria BHR (Wanner, K.A.D.a.B.L., PNAS 97(12): 6640-6645 (June 6, 2000.)). A C57 mouse BAC, served as template for the extraction of homology arms of the targeting vector. The targeting vector contained an frt flanked neomycin phosphotransferase, Neo, selectable marker cassette. After linearization, the targeting construct was electroporated into Primogenix C57B16/N ES derived ES cell line. PCR screens and targeted locus amplification analyses revealed clones that had undergone the desired homologous recombination event. Several of these clones was expanded and injected into Balb/c (Janvier laboratories) blastocysts to generate chimeric males which were then bred to C57B1/6N Crl females, and black-coated offspring were genotyped on both sides of the homology arms for correct integration into the Map3kl5 locus. The neomycin phosphotransferase selectable marker cassette, which was flanked by frt sites, was deleted after subsequent breeding to mice expressing flp recombinase under the CAG promoter.

[0260] Beta-cell specific Map3kl5 KO mice (Map3kl5fl/fl/ins2Cre mice) are produced by breeding AstraZeneca generated Map3kl5 floxed (Map3kl5fl/fl) mice with the AstraZeneca generated Ins2Cre transgenic mice which have a Cre recombinase activated by the Insulin 2 promotor. Wild type litter mates are used as comparator arms for one study on chow diet and one study on HFHS diet (see schematic in Figure 5, Studies 1-2). These studies are performed to confirm that Map3kl5fl/fl/ins2Cre mice are better protected against type 2 diabetes, using metabolic tests, DEXA, indirect calorimetry, oral glucose tolerance tests, and i.p. insulin tolerance tests.

[0261] Inducible models are used for tissue specific (including pancreatic cells, kidney, liver and other relevant tissue) knock out of MAP3K15 as well as a Cre-loxP based inducible CreERT2 system to delete MAP3K15 in adult mice.

[0262] Treatment of Map3kl5fl/fl/CreERT2 mice (Jost Seibier, et al., Nucleic Acids Research 31(4): el2 DOI: 10.1093/nar/gng012 (2003)) (CreERT2 mice used under license from Taconic/ Artemis) with tamoxifen, the cre -recombination knocking out the Map3kl5 gene, is induced in adult mice. The effects of knocking out Map3kl5 in adult age after a normal developmental phase are evaluated by comparison with vehicle treated Map3kl5fl/fl/CreERT2 mice, where the knock out is not induced. All three diets (Chow, HFHS diet and High sodium diet - see above) are used for these mice, and the study design is similar to the one for the full body Map3kl5 KO mice (see schematic in Figure 6.)

[0263] At termination of all preclinical studies outlined in Figures 4-6, the serum and tissues are collected. Changes of clinical chemistry components, hormonal changes etc. in serum are examined. Tissues are interrogated by both by histology and gene expression to establish any changes due to deletion of MAP3K15. In addition, cells are extracted from tissues for ex vivo investigations. The results are analysed to demonstrate that inhibition of MAP3K15 can be protective against the development of diabetes, for example as induced by diet and provide insights on how this protective effect is expressed in terms of clinical chemistry biomarkers.

Example 8: Human genetic evidence supports MAP3K15 inhibition as a therapeutic strategy for diabetes

Abstract

Diabetes mellitus is a chronic health condition that can result in significant end-organ complications and is estimated to impact at least 8.5% of the global adult population. Here, we performed gene-level collapsing analysis on exome sequences from 454,796 multi-ancestry UK Biobank participants to detect genetic associations with diabetes. Rare non -synonymous variants in GCK, GIGYF1, HNF1A, and HNF4A were significantly associated (P<1 x10'⁸) with increased risk of diabetes, whereas rare non-synonymous variants in MAP3K15 were significantly associated with reduced risk of diabetes. Recessive carriers of rare non-synonymous variants in the X chromosome gene MAP3K15 had a 30% reduced risk of diabetes (OR=0.70, 95% Cl: [0.62,0.79], P=5.7x10'¹°), along with reduced blood glucose (beta=-0.13, 95% Cl: [-0.15,-0.10], P=5.5x10'¹⁸) and reduced glycosylated haemoglobin levels (beta=-0.14, 95% Cl: [-0.16,-0.11 ], P=1 .1 x1 O'²⁴). Hemizygous males carrying protein -truncating variants (PTVs) in MAP3K15 demonstrated a 40% reduced risk of diabetes (OR=0.60, 95% Cl: [0.45,0.81], P=0.0007). These findings were independently replicated in FinnGen, with a MAP3K15 PTV associating with decreased risk of both type 1 diabetes (T1 DM) and type 2 diabetes (T2DM) (p<0.05). The effect of MAP3K15 loss on diabetes was independent of body mass index, suggesting its protective effect is unlikely to be mediated via the insulin resistance pathway. Tissue expression profile of MAP3K15 indicates a possible involvement of pancreatic islet cell or stress response pathways. No safety concerns were identified among heterozygous or recessive MAP3K15 PTV carriers across over 15,719 studied endpoints in the UK Biobank. Human population genetic evidence supports MAP3K15 inhibition as a novel therapeutic target for diabetes.

Introduction

Diabetes mellitus is a worldwide health concern projected to affect 700 million people by 2045 It is currently the leading cause of micro- and macrovascular disease, including kidney failure, blindness, heart disease, and lower limb amputations². Characterized by elevated levels of blood glucose, diabetes mellitus is generally categorised into type 1 diabetes mellitus (T1 DM), type 2 diabetes mellitus (T2DM), and other rarer forms. T 1 DM is caused by autoimmune destruction of insulin-producing pancreatic p-cells, while T2DM is primarily caused by peripheral insulin resistance. Both types of diabetes eventually lead to progressive loss of pancreatic p-cells and deficient insulin secretion.

Genome wide association studies (GWAS) have implicated over 60 loci in T1 DM³ and many hundreds of loci in T2DM⁴’⁵. Except for a few loci that map to protein-coding regions (e.g., PAM)⁴, the majority reside in non-coding regions of the genome, making it challenging to map the candidate gene and characterise the underlying causal biology. The growing availability of whole- exome sequences in large population-scale biobanks offers unprecedented opportunities to identify protein-coding variants that have demonstrably large effects on human traits and thus potentially constitute more clinically efficacious target opportunities⁶. Identifying loss-of-function variants that protect against disease is of particular interest since these discoveries can provide direct human-validated therapeutic targets^7-9.

Here, we report a multi-ancestry exome-sequencing association study for diabetes in 412,394 exomes from the UK Biobank (UKB). Using our previously described gene-level collapsing framework¹⁰, we identified that recessive loss of the X chromosome gene MAP3K15 was associated with 40% reduced risk of developing diabetes and decreased circulating glucose and haemoglobin A1 c levels. The findings were replicated in the FinnGen study, with a PTV in MAP3K15 associating with decreased risk of both T1 DM and T2DM. Furthermore, the loss of MAP3K15 was not associated with any apparent on-target adverse phenotypes in a phenome- wide assessment of 15,719 clinical endpoints, supporting MAP3K15 as a potentially safe target for selective therapeutic inhibition. Results

Cohort characteristics and study design

We processed exome sequences from 454,796 UKB participants through our previously described cloud-based pipeline¹⁰. Through stringent quality control, we removed samples with low sequencing quality, low depth of coverage, and from closely related individuals (Methods). For this study, we focused on 90 binary clinical phenotypes related to T1 DM and T2DM available in the UKB (Table 9A). In total, there were 39,044 cases that mapped to at least one of the diabetes-related clinical phenotypes, including 35,035 of European ancestry, 2,262 of South Asian ancestry, 249 of East Asian ancestry, and 1 ,498 of African ancestry. Measurements for quantitative traits related to diabetes, including blood glucose, glycosylated haemoglobin (HbA1 c), and body mass index (BMI), were also available for participants (Table 9B).

We employed a gene-level collapsing framework to test the aggregate effect of rare non- synonymous variants in each gene (N=18,762) against each of the diabetes-related clinical phenotypes (Methods). Each gene-phenotype combination was tested under 10 non- synonymous collapsing models (including one recessive model) to evaluate a range of genetic architectures, as previously described¹⁰ (Table 10). We performed two versions of the collapsing analysis: one restricted to individuals of European ancestry and the other a pan -ancestry analysis, as previously described¹⁰ (Methods). No inflation of test statistics was observed in the gene-level collapsing analysis for the 90 diabetes-related clinical phenotypes that were tested (median genomic inflation lambda across all models = 1 .01 ).

Rare-variant collapsing analysis

We identified four protein-coding genes significantly associated (P<1 x1 O'⁸) with at least one diabetes-related clinical phenotype in the European -only analysis (Figure 1 A; Table 8; Table 11). Rare non-synonymous variants in GCK, GIGYF1, and HNF1A were associated with increased risk of diabetes, whereas rare non-synonymous variants in MAP3K15 were associated with reduced risk of diabetes. In the pan-ancestry analysis, all four of these genes maintained statistical significance; one additional gene, HNF4A, that was associated with increased risk of diabetes, achieved significance in the pan-ancestry analysis (Table 12).

In our phenome-wide association study of 269,171 European UKB participants¹⁰, the recessive collapsing model, which includes homozygous, hemizygous, and putative compound heterozygous carriers of rare non-synonymous variants, identified unequivocal associations (P<1 x10^-8) between MAP3K15 and reduced HbA1 c and glucose levels¹⁰, accompanied by a suggestive association between MAP3K15 and decreased risk of diabetes. Consistent with this, a more recent study of 454,787 UKB participants¹¹ also found suggestive association between MAP3K15 and T2DM (OR = 0.85, P = 2.8x1 O'⁶) under an additive genetic model. Among common variant literature, an intronic variant in MAP3K15 was one of over 300 novel loci reported in a large trans-ethnic GWAS of T2DM⁵ (OR = 1.14, P = 1.4x10'⁸). In this current study, with an increased sample size of 394,695 European participants, the association between MAP3K15ard diabetes reached study-wide significance (p<1 x10^-8) in the recessive model (OR = 0.70; 95% Cl: [0.62,0.79], P = 5.0x1 O'⁹). Consistent with our prior findings, the M A P3K15 qualify ing variant (QV) carriers had significantly lower HbA1 c levels (beta = -0.14, 95% Cl: [-0.16,-0.1 1], P = 3.1 x1 O'²³) (Figure 1 B) and blood glucose levels (beta = -0.13, 95% Cl: [-0.16,-0.10], P = 2.5x10'¹⁷). In our pan-ancestry analysis, the associations became more significant between MAP3K15 recessive variants and diabetes (OR = 0.70, 95% Cl: [0.62,0.79], P = 5.7x10'¹°), HbA1c (beta = -0.14, 95% Cl: [-0.16, -0.1 1], P= 1 .1x10²⁴), and blood glucose (beta = -0.13, 95% Cl: [-0.15, -0.10], P= 5.5x1 O' ¹⁸). Collectively, our results implicate loss of MAP3K15 as a protective factor for diabetes.

Most of the protective MAP3K15 signals emerged for T2DM phenotypes (Table 9A). To determine if MAP3K15 loss also protects from T 1 DM, we defined a T 1 DM-specific phenotype in the UKB (N = 881 cases) using available diagnostic information (Methods). The effect of MAP3K15 recessive variants on T1 DM remained in the protective direction, but the association did not achieve study-wide significance at the current T1 DM sample size (OR = 0.52, 95% Cl: [0.25,1.09], P = 0.09).

Complete versus partial loss of MAP3K15

Given that the MAP3K15 associations emerged strongest among the recessive model, we next tested whether the effect of MAP3K15 loss on diabetes could be dose dependent. Since MAP3K15 resides on chromosome X, hemizygous male PTV carriers are expected to have complete loss of the protein, and heterozygous female carriers are expected to have a 50% loss. Consistent with a dose-dependent effect, we found that hemizygous male carriers of European ancestry (N=1 ,216) demonstrated a 40% decreased risk of developing diabetes compared to male non-carriers (OR = 0.60, 95% Cl: [0.45,0.81 ], P = 7.2x1 O'⁴) (Table 13). In comparison, heterozygous female carriers (N = 2,604) had a 23% reduced risk of diabetes compared to female non-carriers (OR = 0.77; 95% Cl: [0.61 , 0.99], P = 0.04) (Figure 1 C). Decrease in HbA1 c levels were also three times greater in hemizygous male carriers (beta = -0.21 , 95% Cl: [-0.26,-0.15], P = 1 ,2x10'¹¹) (Table 13) than in heterozygous female carriers (beta = -0.07, 95% Cl: [-0.11 ,-0.04], P = 5.3x10⁵) (Figure 1 D). The significantly stronger effects observed with complete loss of MAP3K15 as compared to partial (50%) loss (Figures 1 C and 1 D) suggests an additive protective effect of MAP3K15 loss on diabetes-related traits.

We also found that the contributing PTVs occurred throughout the MAP3K15 gene sequence (Figure 1 E; Figure 7). Two MAP3K15 PTVs were relatively more frequent and accounted for 74% of the European ancestry hemizygous male carriers: Arg1122* (MAF = 0.11 %) and Arg 1136* (MAF = 0.35%) (Figure 1 E; Table 14). Although proximally close, none of the European ancestry males carried both PTVs. Consistent with this, the two PTVs were found to be independently associated with reduced risk of diabetes (Arg1 122*: OR = 0.33, 95% Cl: [0.13,0.80], P= 0.02; Arg1 136*: OR = 0.60, 95% Cl: [0.41 ,0.88], P= 0.01 ) and lower HbA1 c levels (Arg 1 122*: beta = -0.30, 95% Cl: [-0.44,-0.15], P = 4.3x1 O'⁵; Arg 1136*: beta = -0.20, 95% Cl: [- 0.27,-0.12], P = 1.3x1 O'⁶) (Table 15). When excluding these two PTVs from the collapsing test, carriers of the remaining 38 ultra-rare MAP3K15 PTVs observed among the European ancestry males also had significantly reduced HbA1 c levels (beta = -0.16, 95% Cl: [-0.28,-0.05], P= 5.2x1 O' ³). Due to reduced sample and, thus, statistical power, the association with reduced diabetes risk (OR = 0.84, 95% Cl: [0.50,1 .39], P= 0.49) did not achieve significance among the remaining ultra- rare PTV carriers (Table 15).

Replication analysis

Using summary statistics from the FinnGen study release 5 [N=218,792], we next aimed to replicate the MAP3K15 findings. The two more frequent MAP3K15 PTVs (Arg1 122* and Arg1136*) were both well-imputed (INFO scores: 0.98 and 0.84, respectively) in the FinnGen dataset. The Arg1 122* PTV (rs140104197), which is three times more common in individuals of Finnish descent (MAF = 0.33%) than in UKB Europeans (MAF = 0.11 %), was significantly associated with protection from both T 1 DM (OR = 0.58, P = 4.9x1 O'⁴) and T2DM (OR = 0.82, P = 0.04) (Table 16). Arg1136* (rs148312150) was less frequent in individuals of Finnish descent compared to Europeans in UKB (MAF: 0.16% versus 0.35%), and it individually did not reach statistical significance with any diabetes-related phenotype in FinnGen. There were no other MAP3K15 PTVs detected in FinnGen.

MAP3K15 protective PTV signal is not associated with changes in body mass index or metabolic derangements Obesity, which can lead to increased insulin resistance, is a strong risk factor for T2DM. To investigate whether the effect of MAP3K15 on diabetes is mediated via adiposity, we further tested the effect of complete loss of MAP3K15 adjusting for BMI. The associations between hemizygous MAP3K15 PTV carrier status and both HbA1c (BMI-unadjusted: beta = -0.21 , 95% Cl: [-0.15,-0.26], P = 1.2x10'¹¹; BMI-adjusted: beta = -0.20, 95% Cl: [-0.14,-0.26], P = 1.1x10'¹¹) and diabetes (BMI-unadjusted: OR = 0.60, 95% Cl: [0.45,0.81], P = 7.2x1 O'⁴; BMI-adjusted: OR = 0.59, 95% Cl: [0.43,0.79], P = 5.3x1 O'⁴) remained consistent after adjusting for BMI, suggesting the protective effect of MAP3K15 loss on diabetes is unlikely to be mediated via insulin resistance and is likely to benefit individuals irrespective of BMI.

Certain genes that influence diabetes risk can also impact other clinically relevant biomarkers. For example, although PTVs in GIGYF1 are associated with increased risk of diabetes, they are also associated with reduced low-density lipoprotein cholesterol¹². We thus tested whether collapsing analyses of MAP3K15 rare non-synonymous variants associated with any of 168 NMR-based blood metabolite measurements available for approximately 120,000 of the UKB participants. Among the studied metabolites, MAP3K15 was only associated with reduced glucose (“ptv5pcnt” model, beta = -0.16, 95% Cl: [-0.23,-0.10], P= 4.4x1 O'⁷).

Potential MAP3K15 inhibition safety liabilities

We observed that approximately 1 in every 150 (0.6%) European ancestry male participants in the UKB has a lifetime systemic absence of functional MAP3K15. Given these individuals are participants in a generally healthy cohort such as the UKB provides considerable support to the tolerability of MAP3K15 inhibition in humans. This is further supported by this gene’s pLI score of 0.0, which is a measure of the tolerance of a given gene to protein-truncating variants¹³.

However, we sought to systematically evaluate whether therapeutically inhibiting MAP3K15 could associate with on-target adverse phenotypes. To achieve this, we surveyed associations between non-synonymous variants in MAP3K15 and 15,719 clinical phenotypes in the UKB, as described previously¹⁰. We did not observe any significant adverse phenotypic associations (P<1x10'⁸) in individuals with MAP3K15 loss (“ptv”, “ptv5pcnt”, and “rec” collapsing models) in either the individual ancestry or the pan -ancestry analysis. We next tested whether there were any other non-diabetes MAP3K15 associations at a less conservative p-value threshold (P<1x10'⁴). In the European ancestry participants, there were no associations in the “ptv” or “ptv5pcnt” model even at this more liberal p-value cut-off. There were two associations observed in models that included missense variants: hepatomegaly with splenomegaly and diseases of the tongue (Table 17A). Finally, as previous animal model studies have highlighted that knockout of Map3k15 in mice models introduces a hypertensive phenotype¹⁴, we sought to look at this specific phenotype in greater detail. Among the large European sample, there was no evidence of increased risk to hypertension. Instead, the effect of MAP3K15 PTVs on human blood pressure-related traits seems to be in the protective direction. The hemizygous MAP3K15 PTV carriers (i.e., complete loss of MAP3K15) showed a modest effect in the protective direction for both hypertension (‘Union#l10#l10 Essential (primary) hypertension’-. OR = 0.90, 95% Cl: [0.80,1 .02], P = 0.10) and systolic blood pressure (beta = -0.07, 95% Cl: [-0.12,-0.01], P= 0.01 ) (Figures 1C and 1 D; Table 18). Similarly, among the independent FinnGen cohort, the Finnish-enriched MAP3K15 PTV (Arg1 122* (rs140104197)) that associated strongly with T1 DM and T2DM, replicated a modest protective effect on hypertension (OR = 0.84, P = 8.5x1 O'³). Expanding our assessment to MAP3K15 missense variants, the strongest signal for hypertension arises among participants of South Asian ancestry (N=8,078), where the recessive collapsing model showed an association between MAP3K15 and a hypertension phenotype (‘41202#Blockl10-l15#l10-115 Hypertensive diseases’: OR = 6.33, 95% Cl: [3.02,13.28], P = 4.8x1 O'⁵) (Table 17A). Although not study-wide significant (P<1 x10'⁸), more importantly, this signal was driven by MAP3K15 missense variants, with seven hemizygous missense carriers and one hemizygous PTV carrier among affected males of South Asian descent (Table 17B). A more detailed screen of MAP3K15 missense variants identified rs5638141 1 associating with an increased risk of hypertension in FinnGen (MAF = 1 .5%, OR = 1 .17, P = 2.4x1 O'⁷). Evaluation of the effect of MAP3K15 missense variants in recessive form in European ancestry participants revealed a nominally significant association with a blood pressure phenotype in the UK Biobank too ( ‘Union#R030#R03.0 Elevated bloodpressure reading! without diagnosis of hypertension’ : OR = 1 .44, 95% Cl: [1 .02,2.03], P = 0.05).

Whereas certain missense variants in MAP3K15 may increase the risk of hypertension, there is no evidence for PTVs in this gene conferring risk. Collectively, these findings suggest a potential MAP3K15 allelic series, whereby putative gain-of-function missense variants might increase the risk of hypertension while PTV (putative loss-of-function) alleles protect against hypertension. Further functional characterisation of variation in MAP3K15 is required to better understand their variable effects on blood pressure- and diabetes-related traits. Altogether, the lack of association between PTVs in MAP3K15 and any adverse phenotypes suggest that there is a low human safety risk for selective therapeutic inhibition of MAP3K15.

Orthogonal evidence Because MAP3K15 appears to be associated with reduced risk of T1 DM and T2DM and is not associated with BMI, the data suggests that the protective effect is unlikely to be operating through insulin sensitisation. Physiologically, MAP3K15 encodes a mitogen -activated protein kinase that is known to play a role in regulating cell stress and apoptotic cell-death¹⁵. To gain more insight into potential protective mechanisms, we examined the tissue expression profiles of MAP3K15 n GTEx¹⁶. MAP3K15 is most strongly expressed in the adrenal glands and is also expressed at relatively lower levels in the spleen, kidney, pancreas, and pituitary glands (Figure 2A). Singlecell expression data from human pancreatic endocrine cells indicate that MAP3K15 is most strongly expressed in islet cell subpopulations, including a-, p- and 8— cells^17-21 (Figure 2B). Its ubiquitous expression profile in the adrenals could suggest a role in mediating catecholamine biosynthesis or glucocorticoid response²². Alternatively, its effect could be mediated through maintenance of endogenous pancreatic islet cells.

To further explore whether MAP3K15 contributes to the pathophysiology of diabetes in pancreatic cells, we assessed differential gene expression data from a prior study focused on Maturity Onset Diabetes of the Young (MODY)²³, an early-onset, autosomal dominant form of noninsulin-dependent diabetes. In this study, expression profiling was performed on a mouse insulinoma cell line carrying mutations in the MODY-associated gene Nkx6-1. In all three mutations tested, including a positive control mutant known to impair Nkx6-1 as well as two MODY-associated genetic variants, MAP3K15 was found to be the most significantly upregulated gene (Figure 2C). This finding suggests that increased MAP3K15 activity may mediate the pathophysiology of diabetes, potentially through an increased rate of beta cell loss.

Orthogonal evidence derived from two in silico tools, Gene-SCOUT²⁴ and phenome-wide Mantis- ML²⁵ provide further support for a role of MAP3K15 n diabetes. Gene-SCOUT provides biomarker fingerprint similarity between any pair of human genes based on UKB exome sequencing cohort statistics using 1 ,419 quantitative traits²⁴. Entering MAP3K15 as the seed gene in this tool highlights SLC30A8 as having the most similar human biomarker profile to what is observed for MAP3K15 (Figures 3A and 3B; Figure 8). SLC30A8, a zinc transporter gene (ZnT8) expressed in pancreatic islet a- and 0— cells, is reported to have a protective effect against T2DM potentially via increased glucose responsiveness²⁶²⁷. Mantis-ML, an automated machine-learning framework designed to identify gene-phenotype relationships based on compendium of publicly available disease-specific features (such as tissue expression, preclinical models, genic intolerance, among others), suggests disorders related to impaired glucose homeostasis, including “diazoxide-resistant diffuse hyperinsulinism” and “hyperinsulinemic hypoglycaemia” (Figure 3C), among the top 1% of human phenotypes that MAP3K15 may have a role in (Table 19). While mantis-ml does not indicate whether a gene may have a causal versus protective role for a given phenotype, these results converge on MAP3K15’s involvement in diabetes-related biology. Both these tools provide diverse and independent support of a biological role for MAP3K15 '\r human diabetes.

Discussion

This exome-sequencing study of 456,796 UKB participants increases our understanding of high - effect size genetic factors involved in both propensity for and protection from diabetes in humans. We found that recessive loss of MAP3K15 reduces the risk of developing diabetes by approximately 40%. This was supported by the association between recessive loss of MAP3K15 and decreased HbA1 c and blood glucose. Although the protective signal was strongest for T2DM, the effect of MAP3K15 PTVs was also in the protective direction for T1 DM risk in both the UKB and FinnGen. This supports the notion that despite being defined as distinct clinical entities, T 1 DM and T2DM share some common pathophysiological pathways such as 0-cell dysfunction²⁸²⁹.

Crucially, loss-of-function mutations that protect against human disease can act as direct in vivo validation of therapeutic targets; thus, MAP3K15 inhibition could have therapeutic value in both T 1 DM and T2DM. Given the relatively modest decrease in glucose (0.25 mmol/L) and HbA1 c levels (1 .36 mmol/mol) that associate with MAP3K15 loss, one possibility is that the observed protective effect on clinical diabetes is not primarily mediated through glucose / HbA1 c reduction. We also found that the protective effect of MAP3K15 loss is independent of BMI, which might suggest that its effect on diabetes is unrelated to insulin sensitivity, though we note that the relationship between BMI and insulin resistance is correlational. While not currently available for UK Biobank participants, quantitative measures of insulin resistance in MAP3K15 PT\/ carriers in future studies could illuminate whether the protective mechanism is indeed unrelated to the insulin resistance pathway. Nonetheless, our results suggest that therapeutic selective inhibition of MAP3K15 could also benefit patients living with diabetes who are in the low-to-normal BMI range.

Through a phenome-wide association study in the 454,796 human participants, we showed that MAP3K15 loss is not significantly associated with any phenotypes that would suggest safety concerns due to therapeutic inhibition of this target. Prior work observed that knocking out Map3k15 in mice led to hypertension¹⁴. Interestingly, we found that in humans, PTVs appeared to provide a protective effect on hypertension, whereas certain missense variants in MAP3K15 appeared to increase the risk of hypertension. Collectively, these results not only highlight possible species-specific differences upon loss of MAP3K15, but also suggest an allelic series in humans, in which missense variants exert a spectrum of loss- to gain-of-function effects. Future functional characterisation of clinically associated missense variants and PTVs could provide further insights into this potential allelic series.

In our previously published work, MAP3K15 was one of 15 genes that had unequivocal associations with glucose and/or HbA1 c¹⁰. What sets MAP3K15 apart from these other genes is that, with the addition of 150,000 more exomes, we also observe a statistically significant reduced risk for diabetes diagnosis, in addition to the biomarker associations. This finding has important implications for the interpretation of genetic biomarker associations. Crucially, not all genetic associations with clinically relevant biomarkers will be related to the pathophysiology of the underlying disease. Here, anchoring biomarker genetic signals with relevant clinical endpoints can help identify those that are more likely to modify the underlying disease. Therapeutically, this suggests that inhibiting MAP3K15 may target the core pathophysiology of the disease process rather than targeting reduced blood glucose.

The tissue expression profile of MAP3K15 demonstrates predominant expression in adrenal glands and several islet cell subpopulations, suggesting that MAP3K15 might be involved in pancreatic islet cell functional maintenance and / or stress response pathways. Dysregulation of stress response in diabetes³⁰³¹ and the role of ASK MAP kinase) family of genes in regulating stress response (e.g., apoptosis, inflammation) to external stimuli^{14 15} offer further support to these mechanisms. These provide important clues regarding the otherwise unknown pathways that mediate the protective effect of MAP3K15 natural inhibition on diabetes.

Methods

Cohorts

Discovery genetic association studies were performed using the 454,796 exomes available in the UK Biobank (UKB) cohort³². The UKB is a prospective study of approximately 500,000 participants aged 40-69 years at time of recruitment. Participants were recruited in the UK between 2006 and 2010 and are continuously followed. The average age at recruitment for sequenced individuals was 56.5 years and 54% of the sequenced cohort is of female genetic sex. Participant data include health records that are periodically updated by the UKB, self-reported survey information, linkage to death and cancer registries, collection of urine and blood biomarkers, imaging data, accelerometer data and various other phenotypic end points. All study participants provided informed consent and the UK Biobank has approval from the North-West Multi-centre Research Ethics Committee (MREC; 1 1/NW/0382).

Replication of the findings in the UKB was performed using the summary statistics from the FinnGen study. The FinnGen cohort (release 5) includes 218,792 individuals from Finland with genotype and health registry data. Phenotypes have been derived from nationwide health registries. Patients and control subjects in FinnGen provided informed consent for biobank research, based on the Finnish Biobank Act. Alternatively, older research cohorts, collected prior the start of FinnGen (in August 2017), were collected based on study-specific consents and later transferred to the Finnish biobanks after approval by Fimea, the National Supervisory Authority for Welfare and Health. Recruitment protocols followed the biobank protocols approved by Fimea. The Coordinating Ethics Committee of the Hospital District of Helsinki and Uusimaa (HUS) approved the FinnGen study protocol Nr HUS/990/2017. The FinnGen study is approved by Finnish Institute for Health and Welfare.

Phenotypes

We harmonized the UKB phenotype data as previously described¹⁰. Briefly, we used PEACOK and union mapping to parse binary and quantitative traits included in the February 2020 UKB release (accessed March 27, 2020; UKB application 26041 ). Here, we considered 90 binary (clinical) phenotypes related to diabetes available in the UKB (Tables 8A), three quantitative traits related to diabetes (blood glucose, glycosylated haemoglobin and body mass index), and two quantitative traits related to hypertension (systolic blood pressure (SBP) and diastolic blood pressure (DBP)) (Tables 8B). Additionally, for a type 1 diabetes (T 1 DM)-specific analysis, we defined the case population using ICD-9 and ICD-10 codes that captured T 1 DM diagnoses and the control population by excluding participants with any diabetes diagnoses.

For analyses involving SBP and DBP, we adjusted for commonly prescribed blood pressure medications (Table 20).

Genetic data

Exome sequencing data for 454,988 UKB participants were generated at the Regeneron Genetics Center (RGC) as part of a pre-competitive data generation collaboration between AbbVie, Alnylam Pharmaceuticals, AstraZeneca, Biogen, Bristol-Myers Squibb, Pfizer, Regeneron and Takeda with the UKB. Genomic DNA underwent paired-end 75-bp whole-exome sequencing at Regeneron Pharmaceuticals using the IDT xGen v1 capture kit on the NovaSeq6000 platform. Conversion of sequencing data in BCL format to FASTQ format and the assignments of paired- end sequence reads to samples were based on 10-base barcodes, using bcl2fastq v2.19.0. Initial quality control was performed by Regeneron and included sex discordance, contamination, unresolved duplicate sequences and discordance with microarray genotyping data checks.

In FinnGen, genotyping of the samples was done using a ThermoFisher Axiom custom array. In addition to the core GWAS markers (about 500,000), it contains 116,402 coding variants enriched in Finland, 10,800 specific markers for the HLA/KIR region, 14,900 ClinVar variants, 4,600 pharmacogenomic variants and 57,000 selected markers.

AstraZeneca Centre for Genomics Research (CGR) bioinformatics pipeline

The 454,796 UKB exome sequences were re-processed at AstraZeneca from their unaligned FASTQ state. A custom-built Amazon Web Services (AWS) cloud compute platform running Illumina DRAGEN Bio-IT Platform Germline Pipeline v3.0.7 was used to align the reads to the GRCh38 genome reference and perform single-nucleotide variant (SNV) and insertion and deletion (indel) calling. SNVs and indels were annotated using SnpEFF v4.3³³ against Ensembl Build 38.92. We further annotated all variants with their genome Agg regation Database (gnomAD) MAFs (gnomAD v2.1 .1 mapped to GRCh38)¹³. We also annotated variants using MTR score³⁴ to identify if they mapped to genic regions under constraint for missense variants and REVEL scores³⁵ for their predicted deleteriousness.

Additional quality control

To complement the quality control performed by Regeneron Genomics Centre, we passed the UKB exome sequences through our internal bioinformatics pipeline as previously described ¹⁰. Briefly, for UKB, we excluded from our analyses an additional 122 sequences that achieved a VerifyBAMID freemix (measure of DNA contamination) of more than 4%, and an additional 5 sequences where less than 94.5% of the consensus coding sequence (CCDS release 22) achieved a minimum of ten-fold read depth. The cohort was also screened to remove participants that were second-degree relatives or closer (equivalent to kinship coefficient > 0.0884), as determined using the -kinship function in KING v2.2.3³⁶. After the above quality control steps, there remained 412,394 unrelated UKB sequences of any genetic ancestry that were available for analyses presented in this study.

Genetic ancestry

The primary discovery analysis was performed in UKB participants of European ancestry. We used the available exome sequencing data to perform genetic ancestry prediction in PEDDY vO.4.2. We leveraged sequences from the 1 ,000 Genomes Project as population references³⁷ for ancestry estimation. 394,695 (93%) of the 422,488 unrelated UKB participants - that had European ancestry prediction >0.99 and were within 4 SD of the means for the top four principal components - were selected for the European ancestry case-control analyses. We also used the PEDDY-derived ancestry predictions to identify non-European ancestry populations that had at least 1 ,000 individuals with exome sequences to perform pan -ancestry collapsing analyses (see the section ‘Collapsing analyses’). This identified 7,412 African, 2,209 East Asian and 8,078 South Asian UKB participants based on predicted ancestry >0.95 for the respective ancestries.

Discovery analyses

Collapsing analyses

We performed our previously described gene-level collapsing analysis framework¹⁰ for 90 binary and 5 quantitative traits related to diabetes. We included 10 non -synonymous collapsing models, including 9 dominant and one recessive model, plus an additional synonymous variant model as an empirical negative control (Table 10). For the dominant collapsing models, the carriers of at least one qualifying variant (QV) in a gene were compared to the non -carriers. In the recessive model, individuals with two copies of QVs either in homozygous or putatively compound heterozygous form were compared to the non-carriers. Hemizygous genotypes for X chromosome genes also qualified for the recessive model.

Using SnpEff annotations, we defined synonymous variants as those annotated as ‘synonymous_variant’. We defined PTVs as variants annotated as exon_loss_variant, frameshift_variant, startjost, stop_gained, stopjost, splice_acceptor_variant, splice_donor_variant, gene_fusion, bidirectional_gene_fusion, rare_amino_acid_variant, and transcript_ablation. We defined missense as: missense_variant_splice_region_variant, and missense_variant. Non-synonymous variants included: exon_loss_variant, frameshift_variant, startjost, stop_gained, stopjost, splice_acceptor_variant, splice_donor_variant, gene_fusion, bidirectional_gene_fusion, rare_amino_acid_variant, transcript_ablation, conservative_inframe_deletion, conservativejnframejnsertion, disruptivejnframejnsertion, disruptive_inframe_deletion, missense_variant_splice_region_variant, missense_variant, and protein altering variant.

For binary traits, the difference in the proportion of cases and controls carrying QVs in a gene was tested using a Fisher’s exact two-sided test. For quantitative traits, the difference in mean between the carriers and non-carriers of QVs was determined by fitting a linear regression model, correcting for age, sex and medication intake (for SBP and DBP).

For all models, we applied the following quality control filters: minimum coverage 10X; annotation in CCDS transcripts (release 22; approximately 34 Mb); at most 80% alternate reads in homozygous genotypes; percent of alternate reads in heterozygous variants > 0.25 and < 0.8; binomial test of alternate allele proportion departure from 50% in heterozygous state P > 1 x 10"⁶; GQ > 20; FS < 200 (indels) < 60 (SNVs); MQ > 40; QLIAL > 30; read position rank sum score > -2; MQRS > -8; DRAGEN variant status = PASS; the variant site achieved ten-fold coverage in > 25% of gnomAD exomes, and if the variant was observed in gnomAD exomes, the variant achieved exome z-score > -2.0 and exome MQ > 30. We excluded 46 genes that we previously found associated with batch effects¹⁰.

Pan-ancestry collapsing analyses

We performed additional collapsing analysis in each individual non -European ancestral population as described above. For binary traits, we then performed a pan-ancestry analysis using our previously introduced approach¹⁰ of applying a Cochran-Mantel-Haenszel test to generate combined 2x2xN stratified P-values, with N representing up to all four genetic ancestry groups. For quantitative traits, the pan-ancestry analysis was performed using a linear regression model that included the following covariates: age, sex, categorical ancestry (European, African, East Asian or South Asian), and top five ancestry principal components.

Variant-level (ExWAS) analyses

We performed variant-level association tests in addition to the gene-level collapsing analyses for the 90 binary and 5 quantitative traits related to diabetes. We tested 3.3 million variants identified in at least six individuals from the 394,695 predominantly unrelated European ancestry UKB exomes as previously described¹⁰. In summary, variants were required to pass the following quality control criteria: minimum coverage 10X; percent of alternate reads in heterozygous variants > 0.2; binomial test of alternate allele proportion departure from 50% in heterozygous state P > 1 x 10^-6; genotype quality score (GQ) > 20; Fisher’s strand bias score (FS) < 200 (indels) < 60 (SNVs); mapping quality score (MQ) > 40; quality score (QLIAL) > 30; read position rank sum score (RPRS) > -2; mapping quality rank sum score (MQRS) > -8; DRAGEN variant status = PASS; variant site is not missing (that is, less than 10X coverage) in 10% or more of sequences; the variant did not fail any of the aforementioned quality control in 5% or more of sequences; the variant site achieved tenfold coverage in 30% or more of gnomAD exomes, and if the variant was observed in gnomAD exomes, 50% or more of the time those variant calls passed the gnomAD quality control filters (gnomAD exome AC/AC_raw > 50%). P values were generated adopting a Fisher’s exact two-sided test. Three distinct genetic models were studied for binary traits: allelic (A versus B allele), dominant (AA + AB versus BB) and recessive (AA versus AB + BB), where A denotes the alternative allele and B denotes the reference allele. For quantitative traits, we adopted a linear regression (correcting for age and sex) and replaced the allelic model with a genotypic (AA versus AB versus BB) test.

Phenome-wide analysis for MAP3K15

\Ne performed a phenome-wide collapsing analysis for MAP3K15w\t 15,719 binary phenotypes for each individual ancestry in the UKB. We harmonized and union mapped these phenotype data as previously described¹⁰. We included all 11 collapsing models in the PheWAS, as described above. The methodology used here was identical to our previously published PheWAS on 281 ,104 UKB participants¹⁰.

P-value threshold

We defined the study-wide significance threshold as p<1 x10^-8. We have previously shown using an n-of-1 permutation approach and the empirical null synonymous model that this threshold corresponds to a false positive rate of 9 and 2, respectively, out of -346.5 million tests for binary traits in the setting of collapsing analysis PheWAS¹⁰.

Replication analyses

We performed replication analysis of the association between MAP3K15 and diabetes using the publicly available results from the GWAS in the FinnGen cohort. We accessed the association statistics for the phenome-wide analysis of non-synonymous variants within MAP3K15 through the FinnGen portal (release 5). Secondary association analyses

A total of 40 unique PTVs in MAP3K15were observed among the hemizygous male carriers. Two of these PTVs (Arg 1 122* and Arg 1136*) were relatively more frequent. We excluded carriers of these two alleles and re-performed the collapsing analyses for the remaining MAP3K15 PTVs: Fisher’s exact test for diabetes (‘ 20002#1220#diabetes’) and linear regression for HbA1c.

To determine whether the effect of complete loss of MAP3K15 on diabetes is mediated via adiposity or the insulin resistance pathway, we performed additional analyses in which we regressed HbA1 c and the diabetes phenotype (‘ 20002#1220#diabetes’) on MAP3K15 PTV carrier status in males, with BMI as the covariate.

To investigate the joint effects of complete loss of MAP3K15 and a nearby significantly associated indel in PDHA 1 (X-19360844-AAC-A), a gene that overlaps the 3’-UTR of MAP3K15, we regressed HbA1c and the diabetes phenotype (‘ 20002#1220#diabetes’) on the carrier status for the two frequent MAP3K15 PTVs (Arg1 122* and Arg1 136*) and the PDHA 1 indel in males.

Expression analyses

We studied previously published bulk RNA-sequencing data available from a mouse insulinoma cell line (P-TC-6) transfected with three different clones carrying MODY-associated variants in NKX6-125. We extracted the DESeq2-derived log fold changes, p-values, and FDR values from the supplementary data (Tables 9-20 and Figures 7 and 8). We determined tissue expression using the GTEx portal (httD://qtexDortal.org/home/). For single-cell RNA-sequencing analysis, we examined eight previously published datasets using tissue from human pancreatic islets spanning 27 healthy donors, five technologies, and four laboratories^17-21. Data was integrated using Seurat, as previously described³⁸.

Gene-SCOUT

The tool Gene-SCOUT²⁴ estimates similarity between genes by leveraging association statistics from the collapsing analysis across 1 ,419 quantitative traits available in the UKB. We utilised this tool to identify genes that were most similar to the ‘seed gene’ MAP3K15.

Mantis-ML

Mantis-ML²⁵ is a gene prioritisation machine learning framework, integrating a diverse set of annotations, including intolerance to variation, tissue expression and animal models. We used this tool to obtain the top disease predictions for MAP3K15 across 2,536 diseases parsed from Open Targets. Supplemental Methods

Joint analysis of the MAP3K15 and PDHA1 loci

MAP3K15 overlaps with the 3’-UTR of PDHA1, a gene that encodes a subunit of the enzyme pyruvate dehydrogenase and catalyzes a step in the glycolysis pathway. Moreover, an indel (X- 19360844-AAC-A) in the 3’-UTR of PDHA1 is significantly associated with HbA1c levels in the UKB (beta = -0.13, 95% Cl: [-0.17, -0.09], P = 2.1x10^-11). We performed a conditional analysis to ensure that the observed effect of the two more common MAP3K15 PTVs on HbA1 C levels was independent of the PDHA1 indel. Indeed, the associations for all three variants with HbA1c remained significant in the joint analysis (Table 15).

References

1. Saeedi, P. et al. Global and regional diabetes prevalence estimates for 2019 and projections for 2030 and 2045: Results from the International Diabetes Federation Diabetes Atlas, 9(th) edition. Diabetes research and clinical practice 157, 107843 (2019).

2. Deshpande, A. D., Harris-Hayes, M. & Schootman, M. Epidemiology of diabetes and diabetes-related complications. Physical therapy 88, 1254-1264 (2008).

3. Pociot, F. Type 1 diabetes genome-wide association studies: not to be lost in translation. Clinical & translational immunology 6, e162 (2017).

4. Mahajan, A. etal. Fine-mapping type 2 diabetes loci to single-variant resolution using high- density imputation and islet-specific epigenome maps. Nature genetics 50, 1505-1513 (2018).

5. Vujkovic, M. et al. Discovery of 318 new risk loci for type 2 diabetes and related vascular outcomes among 1 .4 million participants in a multi-ancestry meta-analysis. Nature genetics 52, 680-691 (2020).

6. Walter, K. et al. The UK10K project identifies rare variants in health and disease. Nature 526, 82-90 (2015).

7. Cohen, J. C., Boerwinkle, E., Mosley, T. H. & Hobbs, H. H. Sequence variations in PCSK9, low LDL, and protection against coronary heart disease. The New England journal of medicine 354, (2006).

8. Akbari, P. etal. Sequencing of 640,000 exomes identifies GPR75 variants associated with protection from obesity. Science (New York, N.Y.) 373, (2021 ).

9. Abul-Husn, N. S. et al. A Protein-Truncating HSD17B13 Variant and Protection from Chronic Liver Disease. New England Journal of Medicine 378, (2018).

10. Wang, Q. etal. Rare variant contribution to human disease in 281 ,104 UK Biobank exomes. Nature (2021 ) doi : 10.1038/s41586-021 -03855-y.

11 . Backman, J. D. etal. Exome sequencing and analysis of 454,787 UK Biobank participants. Nature (2021 ) doi : 10.1038/s41586-021 -04103-z.

12. Aimee M. Deaton et al. Gene-level analysis of rare variants in 379,066 whole exome sequences identifies an association of GIGYF1 loss of function with type 2 diabetes. Scientific Reports 11 , (2021 ).

13. Karczewski, K. J. et al. The mutational constraint spectrum quantified from variation in 141 ,456 humans. Nature 581 , (2020).

14. Naguro, I. et al. ASK3 responds to osmotic stress and regulates blood pressure by suppressing WNK1 -SPAK/OSR1 signaling in the kidney. Nature communications 3, 1285 (2012).

15. Hattori, K., Naguro, I., Runchel, C. & Ichijo, H. The roles of ASK family proteins in stress responses and diseases. Cell communication and signaling : CCS 7, 9 (2009).

16. Carithers, L. J. & Moore, H. M. The Genotype-Tissue Expression (GTEx) Project. Biopreservation and biobanking 13, (2015).

17. Lawlor, N. et al. Single-cell transcriptomes identify human islet cell signatures and reveal cell-type-specific expression changes in type 2 diabetes. Genome research 27, 208-222 (2017).

18. Grun, D. et al. De Novo Prediction of Stem Cell Identity using Single-Cell Transcriptome Data. Cell stem cell 19, 266-277 (2016).

19. Muraro, M. J. etal. A Single-Cell Transcriptome Atlas of the Human Pancreas. Cell systems 3, 385-394. e3 (2016).

20. Baron, M. et al. A Single-Cell Transcriptomic Map of the Human and Mouse Pancreas Reveals Inter- and Intra-cell Population Structure. Cell systems 3, 346-360. e4 (2016). Segerstolpe, A. et al. Single-Cell Transcriptome Profiling of Human Pancreatic Islets in Health and Type 2 Diabetes. Cell metabolism 24, 593-607 (2016). Uhlen, M. et al. Proteomics. Tissue-based map of the human proteome. Science (New York, N.Y.) 347, (2015). Mohan, V. et al. Comprehensive genomic analysis identifies pathogenic variants in maturity-onset diabetes of the young (MODY) patients in South India. BMC medical genetics 19, 22 (2018). Lawrence Middleton et al. Gene-SCOUT: identifying genes with similar continuous trait fingerprints from phenome-wide association analyses. Nucleic Acids Res (in submission). Vitsios, D. & Petrovski, S. Mantis-ml: Disease-Agnostic Gene Prioritization from High- Throughput Genomic Screens by Stochastic Semi-supervised Learning. American journal of human genetics 106, 659-678 (2020). Dwivedi, O. P. et al. Loss of ZnT8 function protects against diabetes by enhanced insulin secretion. Nature genetics 51 , 1596-1606 (2019). Kleiner, S. etal. Mice harboring the human SLC30A8 R138X loss-of-function mutation have increased insulin secretory capacity. Proceedings of the National Academy of Sciences of the United States of America 115, E7642-E7649 (2018). Ahlqvist, E. et al. Novel subgroups of adult-onset diabetes and their association with outcomes: a data-driven cluster analysis of six variables. The lancet. Diabetes & endocrinology 6, 361-369 (2018). Cnop, M. et al. Mechanisms of pancreatic beta-cell death in type 1 and type 2 diabetes: many differences, few similarities. Diabetes 54 Suppl 2, S97-107 (2005). Surwit, R. S., Schneider, M. S. & Feinglos, M. N. Stress and diabetes mellitus. Diabetes care 15, 1413-1422 (1992). Chiodini, I. etal. Cortisol secretion in patients with type 2 diabetes: relationship with chronic complications. Diabetes care 30, 83-88 (2007). Bycroft, C. et al. The UK Biobank resource with deep phenotyping and genomic data. Nature 562, 203-209 (2018). Cingolani, P. etal. A program for annotating and predicting the effects of single nucleotide polymorphisms, SnpEff: SNPs in the genome of Drosophila melanogaster strain w1118; iso-2; iso-3. Fly 6, 80-92 (2012). Traynelis, J. et al. Optimizing genomic medicine in epilepsy through a gene-customized approach to missense variant interpretation. Genome research 27, 1715-1729 (2017). loannidis, N. M. et al. REVEL: An Ensemble Method for Predicting the Pathogenicity of Rare Missense Variants. American journal of human genetics 99, 877-885 (2016). Manichaikul, A. et al. Robust relationship inference in genome-wide association studies. Bioinformatics 26, (2010). Auton, A. et al. A global reference for human genetic variation. Nature vol. 526 (2015). Stuart, T. et al. Comprehensive Integration of Single-Cell Data. Cell 177, (2019).

Table 8: Genes significantly associated with at least one diabetes-related clinical phenotype in the gene-level collapsing analysis among European ancestry participants in the UK Biobank

Genes across the exome were tested under 10 different non-synonymous collapsing models with 90 diabetes-related clinical phenotypes available in the UK Biobank. The most significant diabetes-related clinical phenotype and the corresponding association statistics have been provided for the four genes that were significantly associated (p<1x10^-8) with at least one diabetes-related clinical phenotype among European ancestry participants.

(Chr=Chromosome; QV=Qualifying Variant; OR=Odds Ratio, CI=Confidence Intervals)

201151-WO-PCT

Table 9A: The diabetes-related clinical phenotypes and quantitative traits from the UK Biobank analysed in this study.

The diabetes-related clinical phenotypes from the UK Biobank were captured using the ICD-10 codes E10-E14 and any additional phenotypic terms containing the strings ‘diabetes’ or ‘Diabetes’. Phenotypic terms related to diabetes insipidus, diabetes medications, or other endocrine conditions were excluded. We additionally performed manual inspection of the 90 diabetes -related clinical phenotypes that were selected for this study. Phenotypes with less than 5 cases in a particular ancestry were not analysed and the corresponding case/control numbers have been labelled as “NA”.

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

Table 9B: The diabetes-related clinical phenotypes and quantitative traits from the UK Biobank analysed in this study.

Five quantitative traits related to diabetes were analysed for genes that were significantly associated with one of the diabetes-related clinical phenotypes.

201151-WO-PCT

Table 10: Summary of the different models implemented in the gene-level collapsing analysis.

Criteria used for determining qualifying variants (QVs) for the 10 different non-synonymous models used in the gene-level collapsing analysis. In addition, a synonymous collapsing model was used for the purpose of establishing an empirical negative control.

201151-WO-PCT

201151-WO-PCT

(MAF = Minor Allele Frequency; Q.C = Quality Control; MTR = Missense Tolerance Ratio)

■"reflects the gnomAD global_raw MAF unless otherwise specified.

^Areflects the maximum proportion of UKB exome sequences permitted to either have < 10-fold coverage at variant site or carry a low-confidence variant that did not meet one of the quality-control thresholds applied to collapsing analyses (see methods).

Synonymous: synonymous_variant

PTV: exon_loss_variant, frameshift_variant, startjost, stop_gained, stopjost, splice_acceptor_variant, splice_donor_variant, gene_fusion, bidirectional_gene_fusion, rare_amino_acid_variant, transcript_ablation Missense: missense_variant_splice_region_variant, missense_variant

Nonsynonymous:exon loss_variant, frameshift_variant, startjost, stop_gained, stopjost, splice_acceptor_variant,splice_donor_variant, gene_fusion, bidirectional_g ene_fusion, rare_amino_acid_variant, transcript_ablation, conservativejnframe_deletion, conservativejnframejnsertion, disruptivejnframejnsertion, disruptivejnframe_deletion, missense_variant_splice_region_variant, missense_variant, protein_altering_variant

201151-WO-PCT

Table 11: Distinct gene-phenotype relationships identified for the four genes that were significantly associated with at least one diabetes-related clinical phenotypes in the gene-level collapsing analysis among European ancestry participants in the UK Biobank.

The diabetes-related clinical phenotypes that were associated (p< 1x10-7) with each of the four genes significantly associated (p<lxl0- 8) with at least one diabetes-related clinical phenotype in the gene-level collapsing analysis among European ancestry participants in the UK Biobank. The most significant collapsing model and the corresponding association statistics have been provided for each genephenotype relationship.

(Chr=Chromosome, QV=Qualifying Variant, OR=Odds Ratio, CI=Confidence Intervals)

201151-WO-PCT

201151-WO-PCT

Table 12: Genes significantly associated with diabetes-related clinical phenotypes in the pan-ancestry gene-level collapsing analysis in the UK Biobank.

A pan-ancestry was performed by combining results from the gene-level collapsing analyses for the four major ancestral groups in the UKB (Europeans, Africans, South Asians and East Asians) via a Cochran-Mantel-Haenszel test. The most significant diabetes-related clinical phenotype and the corresponding association statistics have been provided for the seven genes that were significantly associated (p< 1x10-8) with at least one diabetes-related clinical phenotype.

(Chr=Chromosome, QV=Qualifying Variant, OR=Odds Ratio, CI=Confidence Intervals, CMH= Cochran-Mantel-Haenszel test)

201151-WO-PCT

Table 13: Effect of complete loss otMAP3K15 on diabetes-related traits.

The effect of complete loss of MAP3K15 on HbAlc and diabetes-related clinical phenotypes was evaluated by comparing hemizygous male PTV carriers to male non-carriers of European ancestry. The diabetes-related clinical phenotypes that were associated with MAP3K15 in the initial collapsing analysis were selected.

(OR=Odds Ratio, CI= Confidence Intervals)

201151-WO-PCT

Table 14: Protein-truncating variants in MAP3K15 observed among the European ancestry males in the UK Biobank.

Hemizygous carriers for 40 unique protein-truncating variants in MAP3K15 were observed among the European ancestry males in the UK Biobank. Two particular protein-truncating variants (Argl 122* and Argl 136*), that were relatively more frequent among the carriers, have been highlighted.

(MAF=Minor Allele Frequency)

201151-WO-PCT

201151-WO-PCT

Table 15: Effects of the MAP3K15 PTVs and the PDHA1 indel on diabetes-related traits.

The effects of the two relatively more frequent MAP3K15 PTVs [Argl 122* and Argl 136*] and the collection of remaining PTVs on HbAlc and diabetes were each tested separately. An indel (X-19360844-AAC-A) in the 3’-UTR of PDHA1, which overlaps with MAP3K15, is also significantly associated with HbAlc. The joint effects of the two MAP3K15 PTVs and the PDHA1 indel were tested.

(OR=Odds Ratio, CI=Confidence Intervals)

201151-WO-PCT

Table 16: MAP3K15 associations with diabetes-related clinical phenotypes in the FinnGen cohort.

Association statistics for variants in MAP3K15 that were associated with type 1 or type 2 diabetes in the FinnGen cohort have been provided.

(OR=Odds Ratio)

201151-WO-PCT

Table 17A: Non-diabetes-related clinical phenotypes associated with MAP3K15 in the collapsing analysis across all ancestries in the UK Biobank. A total of 10 non-diabetes-related traits were suggestively associated (p< 1x10-4) with MAP3K15 in the collapsing analysis across all non- synonymous models and ancestries in the UK Biobank.

(QV=Qualifying Variants, OR=Odds Ratio, CI=Confidence Intervals)

*None achieve study-wide signidicance (p<lx!0-8)

201151-WO-PCT

Table 17B: Non-diabetes-related clinical phenotypes associated with MAP3K15 in the collapsing analysis across all ancestries in the UK Biobank.

Association of hypertension-related traits (having at least one case QV carrier) with recessive form of MAP3K15 missense variants (’’recmissense” model) and PTVs (“recptv” model) in the UK Biobank.

(QV=Qualifying Variants, OR=Odds Ratio, CI=Confidence Intervals)

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

Table 18: Association of complete loss otMAP3K15 with blood pressure-related traits among European ancestry participants in the UK Biobank.

The effect of complete loss of MAP3K15 in hemizygous male PTV carriers on blood pressure -related traits was tested. Quantitative traits that were analysed included automated measurements of systolic and diastolic blood pressure (UK Field IDs: 4080 and 4079, respectively). Binary traits that were analysed included a diagnosis of hypertension (Union#I10#I10 Essential (primary) hypertension) and a specific hypertensive phenotype that showed suggestive association in the South Asian ancestry participants (41202#BlockI10- II 5#I 10-115 Hypertensive diseases).

(OR=Odds Ratio, CI=Confidence Intervals)

201151-WO-PCT

Table 19: Mantis-ML predictions of MAP3K15 disease associations.

The list provides the Mantis-ML predictions (‘Disease rank’ and ‘Gene rank percentile’) for MAP3K15 for the entire set of 2,536 diseases that were parsed from Open Targets.

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

201151-WO-PCT

Table 20: Commonly prescribed blood pressure-lowering medications in the UK Biobank.

The list provides the number of participants taking each of the commonly prescribed blood pressure-lowering medications in the UK Biobank. Analyses involving blood pressure measurements (systolic and diastolic blood pressure) were adjusted for the intake of one of these blood pressure-lowering medications.

201151-WO-PCT

201151-WO-PCT

Claims

WHAT IS CLAIMED IS: A method of treating or preventing diabetes mellitus in a subject comprising administering an inhibitor of Mitogen-Activated Protein Kinase Kinase Kinase 15 (MAP3K15) to the subject. A method of decreasing hemoglobin Ale (HbAlc) in a subject in need thereof comprising administering an inhibitor of MAP3K15 to the subject. A method of preventing pancreatic cell death in a subject in need thereof comprising administering an inhibitor of MAP3K15 to the subject. The method of claim 2 or 3, wherein the subject has diabetes. The method of claim 1 or 4, wherein the diabetes is type 2 diabetes. The method of claim 1 or 4, wherein the diabetes is type 1 diabetes. The method of claim 6, wherein the subject has a C -peptide level of at least 0.2 nmol/L prior to the administration. A method of treating or preventing a disease or condition caused or characterized by high blood sugar in a subject, the method comprising administering an inhibitor of MAP3K15 to the subject. The method of any one of claims 1-8, wherein the inhibitor is a small molecule inhibitor, a peptide inhibitor, or an aptamer. The method of claim 9, wherein the small molecule inhibitor is staurosporine, lestaurtinib, TAE-684, ruxolitinib, sunitinib, nintedanib, crizotinib, or fostamatinib. The method of any one of claims 1-8, wherein the inhibitor comprises a nucleic acid molecule and an enzymatic protein, wherein the nucleic acid molecule is a guide RNA (gRNA) molecule and the enzymatic protein is a Cas protein or Cas ortholog. The method of any one of claims 1-8, wherein the inhibitor is a polynucleotide. The method of any one of claims 1-8, wherein the inhibitor is a small interference RNA (siRNA), a short hairpin RNA (shRNA), a microRNA (miRNA), an antagomiRNA, or an antisense RNA. The method of claim 13, wherein the inhibitor comprises the nucleotide sequence GCTCTCTCTTTGAAGGACATGGTAA (SEQ ID NO:1), the nucleotide sequence CGGAUUUCAGGAUGCCGUAAAUAAA (SEQ ID NOG), or the nucleotide sequence CACCGAAGAGCAGUGCAGUAGAUUU (SEQ ID NOG). The method of claim 13, wherein the inhibitor comprises at least two nucleotide sequences selected from a group consisting of the nucleotide sequences: GCTCTCTCTTTGAAGGACATGGTAA (SEQ ID NO:1), CGGAUUUCAGGAUGCCGUAAAUAAA (SEQ ID NOG), and CACCGAAGAGCAGUGCAGUAGAUUU (SEQ ID NOG). The method of any one of claims 1-8, wherein the inhibitor is an enzymatic protein that binds to a gene encoding MAP3K15. The method of claim 16, wherein the enzymatic protein is a Transcription activator-like effector nuclease (TALEN), a zinc-finger nuclease, or a meganuclease. The method of any one of claims 1-8, wherein the inhibitor is a proteolysis targeting chimera (PROTAC). The method of any one of claims 1-9, wherein the inhibitor binds to the ATP-binding domain of MAP3K15. The method any one of claims 1-9, wherein the inhibitor is a non-ATP competitive inhibitor. The method of any one of claims 1-9, wherein the inhibitor binds to the metal -binding domain of MAP3K15. The method of any one of claims 1-21, wherein the inhibitor inhibits the serine-threonine kinase activity of MAP3K15. The method of any one of claims 1-22, wherein the inhibitor inhibits the transferase activity of MAP3K15. The method of any one of claims 1-23, wherein the inhibitor decreases binding of MAP3K15 to ATP. The method of any one of claims 1-24, wherein the inhibitor decreases binding of MAP3K15 to magnesium. The method of any one of claims 1-25, wherein the inhibitor decreases phosphorylation of MAP3K15 Thr-812. The method of any one of claims 1-26, wherein the inhibitor increases phosphorylation of MAP3K15 Ser-924 and/or Ser-994. The method of any one of claims 1-27, wherein the inhibitor decreases homooligomerisation of MAP3K15. The method of any one of claims 1-28, wherein the inhibitor has an IC50 of less than 50 nM, optionally wherein the inhibitor has an IC50 of less than 10 nM. The method of any one of claims 1-29, wherein the concentration of the inhibitor required to cause 50% inhibition of MAP3K15 is 10-fold lower than that required to cause 50% inhibition of ASK1. The method of any one of claims 1-30, wherein the concentration of the inhibitor required to cause 50% inhibition of MAP3K15 is 10-fold lower than that required to cause 50% inhibition of ASK2. The method of any one of claims 1-31, wherein the inhibitor is a reversible inhibitor. The method of any one of claims 1-31, wherein the inhibitor is an irreversible inhibitor. The method of any one of claims 1-33, wherein the administration reduces blood sugar, for example as reflected by measured HbAlc levels, in the subject relative to the level measured prior to treatment. - 218 - The method of claim 34, wherein the administration reduces HbAlc by at least 10 mmol/mol. The method of claim 35, wherein the administration reduces HbAlc by at least 20 mmol/mol. The method of any one of claims 34-36 wherein the HbAlc level post treatment is 48 mmol/mol or below. The method of any one of claims 34-36 wherein the HbAlc level post treatment is between 20 mmol/mol and 42 mmol/mol. The method of any one of claims 34-36, wherein the measured HbAlc level in the subject prior to treatment is at least 48 mmol/mol. The method of any one of claims 1-39, wherein the HbAlc in the subject prior to treatment at least 63.8 mmol/mol. The method any one of claims 1-40, wherein the administration reduces HbAlc to 48 mmol/mol or below, for example from 20 mmol/mol to 42 mmol/mol. The method of any one of claims 1-41, wherein the administration is oral. The method of any one of claims 1-41, wherein the administration is intravenous. The method of any one of claims 1-41, further comprising administering an additional therapeutic agent. The method of claim 44, wherein the additional therapeutic agent is a biguanidine, a thiazolidinedione, a sylfonylureas, a meglitinides, a sodium-glucose cotransporter 2 inhibitor (SGLT2i), a dipeptidyl peptidase IV (DPP-IV) inhibitor, and/or a glucagon-like peptide- 1 (GLP-1). The method of claim 44, wherein the additional therapeutic agent is metformin, a dipeptidyl peptidase IV (DPP-IV) inhibitor, and/or a sodium-glucose cotransporter 2 inhibitor (SGLT2i). - 219 - The method of any one of claims 1-46, wherein the administration is in combination with diet and exercise. The method of any one of claims 1-47, wherein the subject is male. The method of any one of claims 1-47, wherein the subject is female. The method of claim 49, wherein the female carries a single loss of function mutation in the gene encoding MAP3K15. The method of claim 50, wherein the mutation is a truncation. The method of any one of claims 1-50, wherein the subject does not contain a loss of function mutation in the gene encoding MAP3K15. The method of any one of claims 1-52, wherein the subject has a BMI of 30 or more prior to the administration. The method of any one of claims 1-52, wherein the subject has a BMI of at least 25 and less than 30 prior to the administration. The method of any one of claims 1-52, wherein the subject has a BMI of at least 18.5 and less than 25 prior to the administration. The method of any one of claims 1-52, wherein the subject has a BMI of less than 25, less than 20, or less than 18.5 prior to the administration. The method of any one of claims 1-56, wherein the administration reduces MAP3K15 activity in the subject by at least 40%, optionally wherein the administration reduces MAP3K15 activity in the subject by at least 50%, at least 60%, at least 70%, at least 75%, at least 80%, at least 85%, at least 90%, or at least 95%. The method of any one of claims 1-57, wherein the subject has not achieved significant HbAlc reduction following treatment with metformin. - 220 - The method of any one of claims 1-58, wherein the subject has an HbAlc level equal to or greater than 63.9 mmol/mol after treatment with metformin for at least 3 months. The method of any one of claims 1-59, wherein the subject is a mouse. The method of any one of claims 1-59, wherein the subject is human. A method for screening for a substance for treating or prophylaxis of diabetes mellitus comprising contacting MAP3K15 with a test substance and determining if the test substance inhibits MAP3K15. An inhibitor of MAP3K15 for use in the treatment or prophylaxis of diabetes mellitus. An inhibitor of MAP3K15 for use in treatment or prophylaxis of type 2 diabetes mellitus. An inhibitor of MAP3K15 inhibitor for use in treatment or prophylaxis of type 1 diabetes mellitus. The inhibitor for use of any one of claims 63-65, wherein the inhibitor is an ATP competitive inhibitor. The inhibitor for use of any one of claims 63-66, wherein the inhibitor decreases the binding of magnesium to MAP3K15. The inhibitor for use of any one of claims 63-67, wherein the inhibitor is a reversible inhibitor of MAP3K15. The inhibitor for use of any one of claims 63-67, wherein the inhibitor is an irreversible inhibitor of MAP3K15. The inhibitor for use of any one of claims 63-69, wherein the inhibitor is a selective inhibitor of MAP3K15. The inhibitor for use of any one of claims 63-70, wherein the concentration of the inhibitor required to cause 50% inhibition of MAP3K15 is 10-fold lower than that required to cause 50% inhibition of ASK1. - 221 - The inhibitor for use of any one of claims 63-71, wherein the concentration of the MAP3K15 inhibitor required to cause 50% inhibition of MAP3K15 is 10-fold lower than that required to cause 50% inhibition of ASK2. The inhibitor for use of any one of claims 63-72, wherein the inhibitor is a proteolysis targeting chimera (PROTAC). The inhibitor for use of any one of claims 63-72, wherein the inhibitor is an antisense oligonucleotide (ASO) that reduces expression of MAP3K15. The inhibitor for use of any one of claims 63-74, wherein the use results in a reduction of a subject’s HbAlc as measured in a sample obtained from the subject during treatment relative to that measured in a sample obtained from the subject prior to treatment with the inhibitor. The inhibitor for use of claim 75, wherein the use results in a reduction of HbAlc of > 10 mmol/mol, for example > 20 mmol/mmol. The inhibitor for use of claim 75 or 76, wherein the use results in an HbAlc level post treatment of 48 mmol/mol or below. The inhibitor for use of claim 75 or 76, wherein the use results in an HbAlc level post treatment of between 20 mmol/mol and 42 mmol/mol. The inhibitor for use of any one of claims 63-78, wherein the measured HbAlc level in the subject prior to treatment is at least 48 mmol/mol, for example at least 63.8 mmol/mol. The inhibitor for use of any one of claims 63-79, wherein the use results in HbAlc of 48 mmol/mol or below post treatment, optionally wherein the use results in HbAlc 20 mmol/mol to 42 mmol/mol post treatment. The inhibitor for use of any one of claims 63-80, wherein the use is in combination with a further therapeutic agent. The inhibitor for use of claim 81, wherein the further therapeutic agent is a biguanidine, a thiazolidinedione, a sylfonylureas, a meglitinides, a sodium-glucose cotransporter 2 inhibitor - 222 - (SGLT2i), a dipeptidyl peptidase IV (DPP-IV) inhibitor, and/or a glucagon-like peptide- 1

(GLP-1). The inhibitor for use of claim 81, wherein the further therapeutic agent is metformin, a dipeptidyl peptidase IV (DPP-IV) inhibitor, and/or a sodium-glucose cotransporter 2 inhibitor (SGLT2i). The inhibitor for use of claim 81, wherein the further therapeutic agent is metformin. The inhibitor for use of any one of claims 63-84, wherein the use is for a subject that has not achieved significant HbAlc reduction following treatment with metformin. The inhibitor for use of any one of claims 63-85, wherein the use is for a subject with a HbAlc level equal to or greater than 63.9 mmol/mol after treatment with metformin for at least 3 months. The inhibitor for use of any one of claims 63-86, wherein the use is for a subject with a BMI of 30 or more prior to the commencement of treatment with the MAP3K15 inhibitor or wherein the use is for a subject with a BMI of less than 25 or less than 20 prior to the commencement of treatment with the MAP3K15 inhibitor. The inhibitor for use of any one of claims 63-87, wherein the use is for a male subject. The inhibitor for use of any one of claims 63-87, wherein the use is for a female subject. The inhibitor for use of any one of claims 63-87, wherein the use is for a female subject with a single loss of function mutation in the gene encoding MAP3K15 A method of identifying a subject that would benefit from treatment with a MAP3K15 inhibitor comprising: i) sequencing a biological sample obtained from a subject; ii) analyzing the sequence to determine whether there is a loss of function mutation of MAP3K15; and - 223 - iii) based on the result of ii), indicating whether the subject is indicated for treatment with a MAP3K15 inhibitor. The method of any one of claims 1-61, wherein the method comprises identifying the subject using the method of claim 91, prior to the administration. A method of treating a male subject that would benefit from treatment with a MAP3K15 inhibitor, the method comprising: i) sequencing a biological sample obtained from a male subject; ii) analyzing the sequence to determine whether there is a loss of function mutation of MAP3K15; and iii) administering an inhibitor of MAP3K15 to the subject if there is not a loss of function of MAP3K15. A method of treating a female subject that would benefit from treatment with a MAP3K15 inhibitor comprising: i) sequencing a biological sample obtained from a female subject; ii) analyzing the sequence to determine whether there is a loss of function mutation of MAP3K15; and iii) administering an inhibitor of MAP3K15 to the subject if there is not a loss of function of MAP3K15 or if there is only a loss of function of one MAP3K 15 -encoding gene. The method of claim 93 or 94, wherein the administering is according to the method of any one of claims 1-61.