WO2021133777A1 - Monoamine oxidase inhibitors as modifiers of beta cell vulnerability in type 1 diabetes - Google Patents

Abstract

Compositions for use in methods of lowering blood glucose, increasing insulin secretion in response to glucose, preventing the death of pancreatic beta cells or beta-like cells, and preventing the development of type 1 diabetes are provided.

Description

MONOAMINE OXIDASE INHIBITORS AS MODIFIERS OF BETA CELL VULNERABILITY IN TYPE 1 DIABETES

DESCRIPTION

[001] This application claims the benefit of priority to United States Provisional Application No. 62/952,538, which was filed on December 23, 2019, and which is incorporated by reference in its entirety.

SEQUENCE LISTING

[002] The present application is filed with a Sequence Listing in electronic format. The Sequence Listing is provided as a file entitled “2020-11-04_01123-0011- 00PCT_Sequence_Listing_ST25.txt” created on November 4, 2020, which is 12,288 bytes in size. The information in the electronic format of the sequence listing is incorporated herein by reference in its entirety.

FIELD

[003] This application relates to monoamine oxidase inhibitors (MAOIs) for treating diabetes, including autoimmune diabetes.

BACKGROUND

[004] Type 1 diabetes (T1D) is caused by the immune-mediated killing of beta cells in the pancreas (1). Several groups have developed effective differentiation protocols to generate insulin-producing beta-like cells from human embryonic or induced pluripotent stem cells (2). These advances have raised the prospect of replacing lost beta cells in T1D patients using autologous stem cell-derived beta cells, a strategy with the potential to provide an unlimited supply of cells while also circumventing issues of transplant rejection. However, key hurdles persist. In the absence of immune suppression, recurrent autoimmunity rapidly destroys transplanted beta cells. Immune therapies that would induce tolerance to beta cells in T1D patients have not yet been successfully translated from animal models into human (1).

To overcome this critical issue, genetic modifications were investigated that render transplanted beta cells resistant to autoimmune killing. Others have attempted to produce hypoimmunogenic cells by targeting a series of rationally chosen genes related to immune recognition including antigen-presenting HLA molecules (3, 4). Although this approach was reported to be partially effective, it requires the complete abrogation of immune surveillance that protects against infection and tumor formation. [005] This disclosure describes monoamine oxidase inhibitors (MAO inhibitors) as novel modifiers of beta cell vulnerability in autoimmune diabetes, such as T1D.

SUMMARY

[006] In accordance with the description, this application describes a method of lowering blood glucose or increasing insulin secretion in response to glucose in a subject comprising administering a monoamine oxidase inhibitor (MAOI), wherein the monoamine oxidase inhibitor binds renalase, binds flavin adenine dinucleotide (FAD), and/or produces an active agent that binds renalase or FAD.

[007] In some embodiments, the subject has autoimmune diabetes. In some embodiments, the subject has type 1 diabetes. In some embodiments, the subject has autoimmune diabetes induced by an immunotherapy. In some embodiments, the monoamine oxidase inhibitor is administered in combination with an additional treatment. In some embodiments, the additional treatment is insulin. In some embodiments, the insulin is a rapid acting, intermediate-acting, or long-acting insulin. In some embodiments, the additional treatment is a glucagon-like peptide analog or agonist, dipeptidyl peptidase-4 inhibitor, amylin analog, biguanide, thiazobdinedione, sulfonylurea, meglitinide, alpha-glucosidase inhibitor, or sodium/glucose transporter 2 inhibitor. In some embodiments, the subject has a blood sugar level higher than 11.1 mmol/liter or 200 mg/dl.

[008] In some embodiments, a method of preventing the death of pancreatic beta cells or beta-like cells comprises administering a monoamine oxidase inhibitor (MAOI), wherein the monoamine oxidase inhibitor binds renalase, binds flavin adenine dinucleotide (FAD), and/or produces an active agent that binds renalase or FAD.

[009] In some embodiments, the beta cells are those of the subject. In some embodiments, the beta cells are not those of the subject. In some embodiments, the subject is being treated with an immunotherapy. In some embodiments, the immunotherapy is a checkpoint antibody.

[0010] In some embodiments, the checkpoint antibody is an anti -PD- 1 antibody, anti- PD -LI antibody, or anti-CTLA-4 antibody. In some embodiments, the beta cells or beta-like cells are transplanted. In some embodiments, the beta cells or beta-like cells are transplanted into a patient with autoimmune diabetes. In some embodiments, the beta cells or beta-like cells are administered by transplant into the pancreas, liver, or fat pads via surgery, injection, or infusion. [0011] In some embodiments, a method of preventing the development of type 1 diabetes comprises screening a subject for risk factors for type 1 diabetes; determining if the subject has increased risk of developing type 1 diabetes; and administering a monoamine oxidase inhibitor if the subject has an increased risk of type 1 diabetes.

[0012] In some embodiments, screening a subject for risk factors comprises obtaining data on a genetic risk score that is based on the known type 1 diabetes-associated gene variants, a family history of type 1 diabetes, the presence of one or more autoantibodies against beta cell antigens that are known to predict disease risk, and/or abnormal glucose tolerance.

[0013] In some embodiments, the subject is a mammal. In some embodiments, the mammal is a human.

[0014] In some embodiments, the monoamine oxidase inhibitor is a propargylamine, hydrazine, propylamine, or oxazolidinone derivative. In some embodiments, the monoamine oxidase inhibitor is clorgyline, pargyline, rasagiline, selegiline, ladostigil, ASS234, isocarboxazid, toloxatone, or tranylcypromine.

[0015] Additional objects and advantages will be set forth in part in the description which follows, and in part will be obvious from the description, or may be learned by practice. The objects and advantages will be realized and attained by means of the elements and combinations particularly pointed out in the appended claims.

[0016] It is to be understood that both the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the claims.

[0017] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate one (several) embodiment(s) and together with the description, serve to explain the principles described herein.

BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 shows that a genome-scale CRISPR/Cas9 screen identifies Rnls as a modifier of beta cell survival in the NOD mouse model. NIT-1 cells (10⁷) transduced with the mouse GECKO A CRISPR lentiviral library (MOI=0.3) and selected with puromycin were implanted subcutaneously (SubQ) into NOD.sc/ri mice, with or without intravenous injection of 10⁷ splenocytes from diabetic NOD mice. After 8 weeks, NIT-1 grafts were retrieved from recipients with (autoimmune) and without (non-autoimmune) splenocyte co-injection. Next- generation sequencing of gRNAs present in surviving grafts identified Rnls gRNA (MGLibA_46009, 5’-CTACTCCTCTCGCTATGCTC-3’ (SEQ ID NO: 1)) as one of only 11 gRNAs detected at high frequency in mice with beta cell autoimmunity.

[0019] Figures 2A-2L show that Rnls mutation protects NIT-1 and primary NOD beta cells against autoimmune destruction in vivo. (A) Experimental approach used to test NIT-1 beta cell survival after transplantation and induction of autoimmunity. Control (NT) and Rnls^mut NIT-1 cells (10⁷) carrying a luciferase reporter were implanted on opposite flanks of NOD. sad mice. Autoimmunity was induced by injection of 10⁷ splenocytes from diabetic (DM) NOD mice. (B) Representative images of graft luminescence at days 0, 3, 7, 14 and 18 post-transplantation. (C, D) Relative luminescence of paired Rnls^mut and control grafts over time (C) and at day 18 (D), normalized to the ratio on day 0. Data show mean ± SEM of n=5 (+splenocytes) and n=3 (-splenocytes) mice. (E) Experimental approach used to test NIT-1 cell survival transplanted into diabetic NOD mice. Control (NT) and Rnls"^m NIT-1 cells (10⁷) were implanted on opposing flanks of overtly diabetic NOD mice. (F) Representative images of graft luminescence at day 0, 5, 8, 10, 14 and 18 post-transplantation (G) Proportion of remaining luminescence relative to day 0 (100%). (H) Relative luminescence of paired Rnls^mut and control grafts. Data show mean ± SEM of n=5 mice. (I) Experimental approach used to test autoimmune killing of primary islet beta cells. NOD. sc id islet cells transduced with lentivirus encoding a non-targeting (NT) control or Rnls- targeting gRNA and rat insulin promoter (RJP)-driven Cas9 endonuclease were transplanted under the left and right kidney capsule, respectively, of the same NOD. sad recipients. Autoimmunity was induced as in (A). (J) Representative images of transplanted islets on the explanted kidney at day 39. (K) Quantification of insulin mRNA relative to glucagon ( Gcg ) mRNA in paired grafts from non- autoimmune (- splenocytes, n=5) and autoimmune (+ splenocytes, n=6) mice at day 39. (L) Relative insulin expression in paired Rnls^mM and control (NT) grafts. Data represent mean ± SEM. * P < 0.05, ** P < 0.01, calculated by nonparametric unpaired Mann- Whitney test (C- D, G-H, I), and paired Wilcoxon test (K). All data are representative of 2 or more similar experiments.

[0020] Figures 3A-3H shows that Rnls deficiency diminishes immune recognition of beta cells in vitro. (A-D) Representative flow cytometry data and summary data for MHC-I (A, B, MFI: mean fluorescent intensity) and MHC-II (C, D, expressed as % MHC-lE cells) expression in control and Rnls^mut cells treated with thapsigargin (TG) or vehicle (DMSO). Data are representative of three independent experiments. (E and F) BDC2.5-TCR transgenic CD4⁺ T cells were co-cultured with NIT-1 cells and irradiated splenocytes from NOD. scid mice. IFN-g expression in CD4⁺ T cells was measured at 24 h by flow cytometry. Representative (E) and combined data (F) from technical triplicates are shown. Data are representative of 5 independent experiments. (G and H) ELISPOT measurement for the activation of polyclonal CD8⁺ T cells from a diabetic NOD mouse following stimulation with control or Rnls^mut NIT-1 cells. Wells without NIT-1 cells or with PMA and ionomycin (Iono) were used as negative and positive controls, respectively. Data are representative of three independent experiments. Data were compared by ANOVA with Tukey's multiple comparison test ns, not significant, *** P=0.0001. All data show mean ± SEM.

[0021] Figures 4A-4E show that Rnls deficiency confers ER stress resistance in vitro. (A and B) Cell viability measurement 24 h after thapsigargin (TG) (A) and tunicamycin (TC) (B) treatment. n=4 technical replicates per condition, representative of 2 independent experiments. (C-E) Measurement of the unfolded protein response (UPR) in response to TG challenge. ER stress pathway protein phosphorylation (PERK, eIF2a and IRE la), expression (ATF4, Txnip, NRF2) and cleavage (ATF6, Caspase3) (C), Xbpl splicing (D) and Chop and Txnip mRNA levels (E) were measured in control (NT) and /rii/.v"’^ul NIT- 1 cells treated with or without 1 mM TG for 5 h. Data represent mean ± SEM, *# P < 0.05, **## P < 0.01,

***### P < 0.001, calculated by ANOVA with Sidak’s multiple comparisons test. * Control vs. Rnls^mut cells in non-treatment group; # Control vs. Rnls^mut cells in TG-treatment group (D and E).

[0022] Figures 5A-5H show that pargyline binds RNLS and protects beta cells against autoimmunity. (A) Schematic representation of the predicted interaction of pargyline with the FAD co-factor within the RNLS active site. (B) Human recombinant RNLS protein denaturation profile in the presence and absence of pargyline (PG) by a SYPRO orange protein-dye-based thermal shift assay. (C) RNLS melting temperature (Tm) change in response to pargyline. Data were fit to a variable slope four-parameter sigmoid curve. n=3 technical replicates per condition, representative of 2 independent experiments. (D) Experimental approach used to test pargyline for the protection of NIT-1 cells transplanted into diabetic NOD mice. (E) Representative images of graft luminescence at days 0, 3, 5, 7,

12 and 19 post- transplantation. (F) Proportion of remaining graft luminescence relative to day 0 (100%). (G) Blood glucose levels in the control and pargyline treated mice over time. (H) Plasma insulin levels in the control and pargyline treated mice at day 20 (D20). Data show mean ± SEM of n=5 mice per group and are representative of 2 similar experiments, *

P < 0.05, calculated by ANOVA with Tukey’s multiple comparisons test. [0023] Figures 6A-6B show autoimmune killing of NIT-1 cells in NOD mice can be visualized by bioluminescence imaging. (A and B) Bioluminescence imaging of 10⁷ NIT-1 cells transplanted subcutaneously into NOD.sc/r/ mice. Transplanted cells were engineered to carry a CMV-luciferase2 (Luc2) reporter. Some recipient mice were also injected intravenously with 10⁷ splenocytes isolated from spontaneously diabetic (DM) NOD mice to cause beta cell killing. Images were taken at day 1 (A) and 15 (B) post-injection.

[0024] Figures 7A and 7B show validation of Rnls mutation by CRISPR-Cas9 targeting. (A) T7 endonuclease I assay. Genomic DNA from NIT-1 wild-type (WT) and Rnls^mut cells was tested for CRISPR-Cas9 gene editing events. Cleavage at heteroduplex mismatch sites by T7 endonuclease I digestion was analyzed by agarose gel electrophoresis. DNA from Rnls"^m cells segregated into multiple digested fragments, indicating efficient mutation of the targeted region in the Rnls gene. (B) Genomic DNA from Rnls""" cells was sequenced to identify individual mutations (SEQ ID NOs: 17-23). The Rnls gRNA targeting site in the wildtype (SEQ ID NO: 16) is labelled with underlining. The mutations with the highest frequencies in Rnls^mut cells are shown.

[0025] Figure 8 shows that Rnls mutation does not impair insulin secretion. Islets («1700) were purified from 8-week old CD1 mice, dispersed and transduced with lentivirus encoding a non-targeting (NT) or Rn I -targeting gRNA together with the Cas9 endonuclease driven by the rat insulin promoter. After 72 hours, islets were stimulated sequentially with 2.8 mM glucose, 16.8 mM glucose and finally 30 mM KC1 to induce insulin secretion. Islet genomic DNA was quantified for normalization of ELISA insulin measurements to DNA content. n=5 technical replicates per condition and genotype. Data show mean ± SEM. Note that islet dispersion necessary for lentiviral transduction decreased the overall responsiveness of purified islets compared to intact islets. Insulin secretion by Rnls mutant islet cells was not significantly different from that of control (NT) islets.

[0026] Figures 9A-9B show that Rnls mutation does not prevent allo-rejection of NIT-1 beta cells. Control and Rnls^mM NIT-1 cells (10⁷) carrying aluciferase reporter were implanted on opposing flanks of C57BL/6 mice (n=3). Graft bioluminescence was measured on days 0, 4 and 7 after transplantation. Representative bioluminescence images (A) and relative luminescence of grafts over time (B) are shown. Data represent mean ± SEM. Both control and mutant grafts were destroyed by allo-rejection within a week.

[0027] Figures 10A-10B show that Rnls^mut cells are not resistant to non-ER stress induced cell death. (A) Relative cell viability at 6 hours and 24 hours after treatment with 40 mM streptozotocin (STZ) (n=3 technical replicates). (B) Relative cell viability 48 hours after mitomycin C (MMC) at the indicated concentration (n=3 technical replicates). Data represent mean ± SEM, *** P < 0.001, calculated by ANOVA with Sidak’s multiple comparisons test.

[0028] Figures 11 A-l ID show that Rnls knockout replicates the ER stress-resistant phenotype of Rnls^mut cells. Rnls knockout NIT-1 cell lines were generated by deleting either exons 2-4 or exon 5. Deletion efficiency was confirmed by qPCR of genomic DNA. Rnls AEx2/4 cells showed -60% deletion of exons 2-4 genomic DNA qPCR (A) while Rnls AEx5 cells showed -87% deletion of exon 5 (B). (C and D) Cell viability of Rnls deficient cells was measured 72 h after thapsigargin (TG, C) and tunicamycin (TC, D) treatment. Data show mean ± SEM, *** P < 0.001, calculated by unpaired t-test (A and B) and ANOVA with Sidak’s multiple comparisons test (C and D).

[0029] Figures 12A-12B show that Rnls overexpression sensitizes NIT-1 cells to ER stress-induced death and reverses the resistant phenotype of Rnls^mut cells. (A and B) NIT-1 cell viability was measured 24 h after treatment with thapsigargin (TG) at the indicated concentrations. Overexpression of Rnls in WT NIT-1 cells increased sensitivity to low dose- TG-induced killing (A). CRISPR-immune Rnls (CiRnls) expressed in Rnls^mut cells restored sensitivity to TG-induced killing (B). n=4 technical replicates per group. Data represent mean ± SEM, *# P < 0.05, ***### P < 0.001, calculated by ANOVA plus Sidak’s multiple comparisons test. *Comparison of control vs. Rnls""" cells: //comparison of Rnls""" vs. Rnls^mut + CiRnls cells.

[0030] Figures 13A-13C sho that Rnls overexpression increases sensitivity to autoimmune killing in vivo. Control (WT) and Rnls overexpressing (Rnls^0E) NIT-1 cells carrying a luciferase reporter were implanted on opposing flanks of NOD.sc/<i mice. Some graft recipients were also injected intravenously with splenocytes from diabetic NOD mice (DM NOD splenocytes). Graft bioluminescence was imaged on days 0, 2, 3 and 7 (A). The relative luminescence of Rnls^0E and control grafts over time, normalized to day 0, is shown in (B). Data for all mice analyzed on day 3 is shown in (C). Rnls^0E graft were more sensitive to autoimmune killing as evidenced by more rapid loss of luminescence. By day 7, both control and Rnls^0E grafts were killed to -90% (data not shown), resulting in a similar relative luminescence level. n=6 mice (each with two grafts). Data represent mean ± SEM, * P <

0.05, **P < 0.01, calculated by unpaired t-test.

[0031] Figures 14A-14B show that Rnls overexpression restores the sensitivity of Rnls^mut cells to autoimmune killing in vivo. Rnls""" NIT- 1 cells and Rnls""" cells expressing the CRISPR-insensitive Rnls transgene (CiRnls), all carrying a luciferase reporter, were implanted on opposing flanks of NOD.sc/ri mice. Graft recipients were also injected intravenously with splenocytes from diabetic (DM) NOD mice. Graft bioluminescence was imaged on days 0, 2, 3 and 5 post-injection (A). Relative luminescence of paired grafts over time normalized to day 0 is shown in (B). n=5 mice. Data represent mean ± SEM.

[0032] Figure 15 shows that islet Rnls expression is elevated in the diabetes-prone NOD mouse strain. Rnls mRNA expression levels in islets isolated from C57BL/6, NOD and NOD..Y676/ mice was determined by TaqMan assays using Rnls (Mm04178677_ml) mdHprtl (Mm0302475_ml) probes from Thermo Fisher Scientific. n=8 mice per group. Data show mean ± SEM, * P < 0.05, **P < 0.01, calculated by nonparametric Kruskal-Wallis with Dunn’s multiple comparisons test.

[0033] Figures 16A-16H show that Rnls deficiency diminishes the UPR following ER stress using a variety of different markers. Data represent quantification of Western blot data shown in Figure 4C. Images were obtained and quantified using a C-DiGit scanner and the Image Studio software (LI-COR Biosciences). n=3 per group. Data show mean ± SEM, *# P < 0.05, **## P < 0.01, ***### P < 0.001, calculated by ANOVA with Dunnetf s multiple comparisons test. *Comparison to control cells without TG treatment; Comparison to control cells with 5-hour TG treatment.

[0034] Figure 17 shows Rnls deficiency confers resistance to oxidative stress. Control (Ctrl) and Rnls^mut NIT-1 cells were cultured overnight with or without hydrogen peroxide (H2O2) at the indicated concentrations. Cell viability was assessed using the CellTiter-Glo luminescence Cell Viability Assay. Data show mean ± SEM of triplicate cultures and are representative of three independent experiments. **** P0.0001.

[0035] Figure 18 shows that pargyline treatment preserves insulin expression in NOD mice with long-duration diabetes. Pancreases were isolated from control and pargyline- treated diabetic NOD mice described in Figure 5 that were euthanised at day 20 post-beta cell-transplantation. Pancreatic sections were stained with anti -insulin (DAKO, #A0564), anti-CD3 (Bio-rad, #MCA500), and DNA dye Hoechst 33342 (Invitrogen, #H3570). Goat anti -guinea pig Alexa Flour 488 and donkey anti -rat Alexa Flour 594 secondary antibodies (Thermo Fisher Scientific, #A11073 and #A21209) were used to detect insulin and CD3 antibodies, respectively. All images were taken using a Zeiss LSM710NLO confocal microscope. [0036] Figures 19A-19B show that pargyline treatment does not prevent beta cell destruction after allo-transplantation. Wild-type NIT- 1 cells (10⁷) carrying aluciferase reporter were implanted into C57BL/6 mice that were treated or not with oral pargyline via addition to the drinking water. Graft bioluminescence was measured on days 1, 2, 3 and 4 after transplantation. Representative bioluminescence images (A) and relative luminescence of grafts over time (B) are shown. Data show mean ± SEM for n=3 mice per group.

[0037] Figures 20A-20F show thermal shift assay results for human recombinant renalase in the presence of pargyline (A), rasagiline (B), selegiline (C), tranylcypromine (D), isocarboxazid (E), or toloxatone (F). All drugs were used at lOOmM, and caused a destabilization of the enzyme, as evidence by a decrease in the unfolding temperature (left- shift of the curves).

[0038] Figures 21A and 21B show in vivo effects of pargyline in diabetes models. (A) Pargyline treatment decreases diabetes onset after cyclophosphamide injection. Groups of 10- week-old male NOD mice were fed pargyline via the drinking water (5pg/ml). Diabetes was induced by intraperitoneal injection of cyclophosphamide (200mg/kg). (B) Pargyline treatment decreases diabetes onset after PD-1 blockade. Groups of 10-week-old female NOD mice were fed pargyline via the drinking water (10pg/ml). Diabetes was induced by intravenous injection of anti-PD-1 antibody (250pg/mouse).

[0039] Figures 22A-22C show in vivo effects of pargyline treatment (25pg/ml pargyline in the drinking water) in mouse models of type 1 diabetes. Pargyline treatment improved survival when diabetes was induced by adoptive transfer (A) or when diabetes was induced by low doses of streptozotocin (STZ, B). Pargyline also delayed the onset of diabetes following STZ treatment (C).

DESCRIPTION OF THE SEQUENCES

[0040] Table 1 provides a listing of certain sequences referenced herein.

DESCRIPTION OF THE EMBODIMENTS

I. Definitions

[0041] In addition to definitions included in this sub-section, further definitions of terms are interspersed throughout the text.

[0042] In this invention, “a” or “an” means “at least one” or “one or more,” etc., unless clearly indicated otherwise by context. The term “or” means “and/or” unless stated otherwise. In the case of a multiple-dependent claim, however, use of the term “or” refers back to more than one preceding claim in the alternative only.

[0043] “Autoimmune” or “autoimmune attack,” as used herein, refers to an attack by the subject’s immune system against cells that are part of the subject. As such, an autoimmune disease is an abnormal immune response to a normal body part. In the case of type 1 diabetes, the autoimmune attack is predominantly against the beta cells of the pancreas that normally secrete insulin in a glucose-dependent manner. As used herein, “autoimmune diabetes” relates to any diabetes induced an autoimmune attack, such as type 1 diabetes or diabetes induced by an immunotherapy.

[0044] As used herein, “beta-like cell” refers to any cell that secretes insulin in response to glucose. Thus, a pancreatic beta cell is a “beta-like cell.” Beta-like cells may be derived from cells that do not normally produce insulin in response to glucose. For example, a beta-like cell may be a stem cell that is induced to differentiate into a “beta-like cell” that produces insulin in a glucose-responsive manner (see FW Pagliuca et ak, Cell 159:428-439 (2014); E Kroon et ak, Nature Biotech 26(4):443-452 (2008); and A Rezania et ak, Nature Biotech 32(11): 1121-1133 (2014). Likewise, a “beta-like cell” may also be a pancreatic exocrine cell (see Q Zhou et ak, Nature 455:627-633 (2008)), pancreatic alpha cell (see Li et al, Cell 168:86-100 (2017), or gut cell (see Ariyachet C et ak, Cell Stem Cell 18(3):410-21 (2016)) that is induced to produce insulin in response to glucose. The term “beta-like cells” also includes cells that become glucose responsive insulin secretors after transplantation into a subject.

[0045] The term “treatment,” as used herein, covers any administration or application of a therapeutic for disease in a subject, and includes inhibiting the disease, arresting its development, relieving one or more symptoms of the disease, or preventing reoccurrence of one or more symptoms of the disease. For example, treatment of diabetes type 1 subjects may comprise alleviating hyperglycemia as compared to a time point prior to administration or reducing the subject’s need for exogenous insulin administration.

II. MAOIs that bind renalase or FAD

[0046] This application relates to lowering blood glucose or increasing insulin secretion in response to glucose in a subject comprising administering a monoamine oxidase inhibitor (MAOI) wherein the monoamine oxidase inhibitor binds renalase (RNLS), binds flavin adenine dinucleotide (FAD), and/or produces an active agent that binds RNLS or FAD.

[0047] Flavin adenine dinucleotide (FAD) binding sites are conserved across flavoprotein oxidases, such as RNLS and MAO (See Gaweska and Fitzpatrick Biomol Concepts. 2(5): 365-377 (2011)). These flavoprotein oxidases may have structural diversity in other regions, while retaining homology in the FAD binding site. FAD is comprised of an adenine nucleotide (adenosine monophosphate) and a flavin mononucleotide bridged together through their phosphate groups.

[0048] Human RNLS (hRNLS) comprises a flavin adenine dinucleotide (FAD) binding site and uses FAD as a co-factor for catalysis. MAO (including MAO A and MAO B isoforms) also interacts with FAD and can complex with inhibitors that bind FAD.

[0049] A wide variety of MAOI have been described, some of which can bind to FAD (See Ramsay and Albreht Journal of Neural Transmission 125:1659-1683 (2018)).

[0050] In some embodiments, the MAOI binds renalase (RNLS). MAOIs that bind RNLS can be determined using standard binding assays, such as radioligand binding, functional assays, or thermal shift assays. This application describes a variety of MAOIs that can bind FAD and can also bind to RNLS. In some embodiments, a thermal shift assay can be used to determine binding of an MAOI to RNLS, such as data shown in Figures 20A-20F.

[0051] In some embodiments, the MAOI binds RNLS via binding to the FAD bound to RNLS. In some embodiments, binding of the MAOI to RNLS blocks or inhibits function of RNLS.

[0052] In some embodiments, the MAOI binds FAD. In some embodiments, the MAOI binds to the flavin of FAD. In some embodiments, the MAOI binds to the N5 atom of flavin.

[0053] In some embodiments, the MAOI produces an active species that can bind RNLS. In some embodiments, the MAOI produces an active species that can bind FAD. In some embodiments, the MAOI is modified by RNLS to an active form, which can bind to FAD or RNLS. In some embodiments, the MAOI or its active species forms a covalent adduct with FAD or RNLS. [0054] In some embodiments, the MAOI binds reversibly to FAD or RNLS. In some embodiments, the MAOI binds irreversibly to FAD or RNLS.

[0055] In some embodiments, the MAOI binds covalently to FAD or RNLS. In some embodiments, the MAOI binds non-covalently to FAD or RNLS.

[0056] In some embodiments, the monoamine oxidase inhibitor is a propargylamine, hydrazine, propylamine, or oxazolidinone derivative.

[0057] In some embodiments, the monoamine oxidase inhibitor is clorgyline, pargyline, rasagiline, selegiline, ladostigil, ASS234, isocarboxazid, toloxatone, or tranylcypromine.

III. Methods of treatment

[0058] A method of treating diabetes mellitus comprising administering a MAOI is encompassed. This method may be for treating diabetes, including autoimmune diabetes. In some embodiments, the subject has type 1 diabetes or autoimmune diabetes induced by an immunotherapy.

[0059] A method of lowering blood glucose levels comprising administering a MAOI is also encompassed. This method may be for treating subjects with diabetes, including autoimmune diabetes. In some embodiments, the subject has type 1 diabetes or autoimmune diabetes induced by an immunotherapy.

[0060] A method increasing insulin secretion in response to glucose comprising administering a MAOI is also encompassed. This method may be for treating subjects with diabetes, including autoimmune diabetes. In some embodiments, the subject has type 1 diabetes or autoimmune diabetes induced by an immunotherapy.

[0061] A method of preventing or slowing the death of pancreatic beta cells or beta like cells comprising administering a MAOI is also encompassed. This method may be for treating subjects with diabetes, including autoimmune diabetes. In some embodiments, the subject has type 1 diabetes or autoimmune diabetes induced by an immunotherapy.

[0062] Glucose levels in the blood are normally tightly regulated to maintain an appropriate source of energy for cells of the body. Insulin and glucagon are principal hormones that regulate blood glucose levels. In response to an increase in blood glucose, such as after a meal, insulin is released from beta cells of the pancreas. Insulin regulates the metabolism of carbohydrates and fats by promoting uptake of glucose from the blood into fat and skeletal muscle. Insulin also promotes fat storage and inhibits the release of glucose by the liver. Regulation of insulin levels is a primary means for the body to regulate glucose in the blood. [0063] When glucose levels in the blood are decreased, insulin is no longer released and instead glucagon is released from the alpha cells of the pancreas. Glucagon causes the liver to convert stored glycogen into glucose and to release this glucose into the bloodstream. Thus, insulin and glucagon work in concert to regulate blood glucose levels.

[0064] In one embodiment, treatment of diabetes mellitus is to administer a MAOI to a subject to lower blood glucose.

[0065] Hyperglycemia refers to an increased level of glucose in the blood. Hyperglycemia can be associated with high levels of sugar in the urine, frequent urination, and increased thirst. Diabetes mellitus refers to a medical state of hyperglycemia.

[0066] The American Diabetes Association (ADA) suggests that fasting plasma glucose (FPG) levels of 100 mg/dL to 125 mg/dL or HbAlc levels of 5.7% to 6.4% may be considered hyperglycemia and may indicate that a subject is at high risk of developing diabetes mellitus (i.e. prediabetes, see ADA Guidelines 2015).

[0067] The ADA states that a diagnosis of diabetes mellitus may be made in a number of ways. A diagnosis of diabetes mellitus can be made in a subject displaying an HbAlc level of >6.5%, an FPG levels of >126 mg/dL, a 2-hour plasma glucose of >200 mg/dL during an OGTT, or a random plasma glucose level >200 mg/dL in a subject with classic symptoms of hyperglycemia. In some embodiments, the subject has a blood sugar level higher than 11.1 mmol/liter or 200 mg/dl.

[0068] Diabetes mellitus can be broken into Type 1 and Type 2. Type 1 diabetes mellitus (previously known as insulin-dependent diabetes or juvenile diabetes) is an autoimmune disease characterized by destruction of the insulin-producing beta cells of the pancreas. Classic symptoms of Type 1 diabetes mellitus are frequent urination, increased thirst, increased hunger, and weight loss. Subjects with Type 1 diabetes mellitus are dependent on administration of insulin for survival.

[0069] In the absence of regulation of glucose levels in subjects with diabetes, a range of serious complications may be seen. These include atherosclerosis, kidney disease, stroke, nerve damage, and blindness.

[0070] In some embodiments, the subject treated has diabetes mellitus based on diagnosis criteria of the American Diabetes Association. In some embodiments, the subject with diabetes mellitus has an HbAlc level of >6.5%. In some embodiments, the subject with diabetes mellitus has an FPG levels of >126 mg/dL. In some embodiments, the subject with diabetes mellitus has a 2-hour plasma glucose of >200 mg/dL during an OGTT. In some embodiments, the subject with diabetes mellitus has a random plasma glucose level >200 mg/dL or 11.1 mmol/L. In some embodiments, the subject with diabetes mellitus has a random plasma glucose level >200 mg/dL or 11.1 mmol/L with classic symptoms of hyperglycemia.

[0071] In some embodiments of the invention, the subject treated is a mammal. In some embodiments, the mammal is a human, non-human primate, cow, horse, pig, sheep, goat, dog, cat, or rodent. In some embodiments, the subject is a human subject.

[0072] In some embodiments, the subject has autoimmune diabetes. In some embodiments, the subject treated has Type 1 diabetes mellitus.

[0073] In some embodiments, the subject treated has a relative decrease in insulin levels. In some embodiments, the subject treated has decreased beta cell mass. In some embodiments, the decrease in beta cell mass in a subject is due to an autoimmune disease.

[0074] Treatment of patients with therapeutics targeted to increase the body’s immune response to cancers, termed immunotherapies, has also been associated with the development of autoimmune diabetes (See Alrifai T et al., Case Reports in Oncological Medicine 2019: Article ID 8781347). For example, immune checkpoint antibodies have been reported to cause immune-mediated damage of islet cells leading to induction of autoimmune diabetes similar to type 1 diabetes.

[0075] In some embodiments, the subject has autoimmune diabetes induced by an immunotherapy. In some embodiments, the immunotherapy is a checkpoint antibody. In some embodiments, the checkpoint antibody is an anti-PD-1 antibody, anti-PD -LI antibody, or anti-CTLA-4 antibody.

[0076] In one embodiment, the method comprises lowering blood glucose levels in the diabetic subject to below about 200 mg/dL, 150 mg/dL, 100 mg/dL, or about 125 mg/dL.

[0077] In some embodiments, treatment of diabetes is lowering blood glucose in the subject after administering a MAOI. In some embodiments, treatment of diabetes is increasing insulin levels in the subject after administering a MAOI. In some embodiments, treatment of diabetes is increasing insulin secretion in the subject after administering a MAOI.

[0078] In some embodiments, administering a MAOI causes a decrease in blood glucose levels such that levels are less than 200 mg/dL.

[0079] In some embodiments, the MAOI is administered in combination with an additional treatment. In some embodiments, the additional treatment is insulin. In some embodiments, the insulin is a rapid-acting, intermediate-acting, or long-acting insulin. In some embodiments, the additional treatment is a glucagon-like peptide analog or agonist, dipeptidyl peptidase-4 inhibitor, amylin analog, biguanide, thiazobdinedione, sulfonylurea, meglitinide, alpha-glucosidase inhibitor, or sodium/glucose transporter 2 inhibitor.

[0080] This application also encompasses methods of preventing the death of pancreatic beta cells or beta-like cells comprising administering a monoamine oxidase inhibitor. Exemplary beta cells or beta-like cells include those that are transplanted in a subject as a means to treat diabetes. Currently, transplanted beta or beta-like cells are prone to cell death due to autoimmune atack on the transplanted cells.

[0081] In some embodiments, the beta cells or beta-like cells are transplanted. In some embodiments, the beta cells or beta-like cells are transplanted into a subject with autoimmune diabetes. In some embodiments, the beta cells or beta-like cells are administered by transplant into the pancreas, liver, or fat pads via surgery, injection, or infusion.

[0082] Other exemplary beta cells or beta-like cells that can be protected by a MAOI include those in subjects who are at risk of developing autoimmune diabetes, such as those of a subject being treated with an immunotherapy. In some embodiments, the beta or beta-like cells are those of a subject who is being treated with an immunotherapy. In some embodiments, the immunotherapy is a checkpoint antibody. In some embodiments, the checkpoint antibody is an anti-PD-1 antibody, anti-PD -LI antibody, or anti-CTLA-4 antibody.

A. Transplanted beta cells or beta-like cells

[0083] Beta cells of pancreas are the cells that normally can secrete insulin. These beta cells of the pancreas are located in pancreatic islets, also known as the islets of Langerhans. A transplanted beta or beta-like cell refers to a cell that is placed in an individual, wherein the cell is from a different individual or is from the same individual but from a different original source in the body than the pancreatic islets.

[0084] In some embodiments, the beta or beta-like cell has been reintroduced into the same or different individual from which it was isolated. In some embodiments, the beta or beta-like cell are those of the subject. When introduced into the same subject from which it was isolated it is an autologous beta or beta-like cell. When introduced into a different subject from which it was isolated it is a heterologous beta or beta-like cell.

[0085] In some embodiments, the beta-like cell is a cell that does not normally produce insulin in response to glucose, but is induced or designed to have a phenotype of a beta-like cell, i.e., induced or designed to produce insulin in response to glucose. Beta-like cells include “designer beta cells,” which have been described as using synthetic pathways to produce insulin ( see M Xie et ak, Science 354(6317): 1296-1301 (2016)). [0086] A variety of beta-like cells have been described. a) Stem cells

[0087] Any stem cell capable of differentiating into a beta-like cell may be a beta-like cell according to the invention. In some embodiments, the beta-like cell may be differentiated from a hematopoietic stem cell, bone marrow stromal stem cell, or mesenchymal stem cell.

[0088] Beta-like cells capable of secreting insulin in response to glucose can be generated from pluripotent stem cells (PSCs) (see FW Pagliuca et ak, Cell 159:428-439 (2014)) or embryonic stem cells (ESCs) (see E Kroon et ak, Nature Biotech 26(4):443-452 (2008) and A Rezania et ak, Nature Biotech 32(11): 1121-1133 (2014)).

[0089] In some embodiments, the stem cell may be an embryonic stem cell. In some embodiments, the embryonic stem cell is taken from a blastocyst. In some embodiments, the embryonic stem cell may be derived from an embryo fertilized in vitro and donated. In some embodiments, the embryonic stem cell undergoes directed differentiation.

[0090] In some embodiments, the stem cell may be an adult stem cell. An adult stem cells may also be referred to as a “somatic” stem cell. In some embodiments, the adult stem cell is an undifferentiated cell found among differentiated cells in a tissue or organ.

[0091] In some embodiments, the stem cell is an induced pluripotent stem cell (iPSC).

[0092] In some embodiments, the stem cells may be from bone marrow, adipose tissue, or blood. In some embodiments, the cells may be from umbilical cord blood.

[0093] In some embodiments, stem cells undergo directed differentiation into beta like cells. In some embodiments, the directed differentiation is based upon treatment of stem cells with modulators. In some embodiments, the directed differentiation is based on culture conditions.

[0094] In some embodiments, beta-like cells are generated from human PSCs (hPSCs) in vitro. In some embodiments, beta-like cells are generated from hPSCs using directed differentiation. In some embodiments, beta-like cells are generated from hPSCs using a multi-step protocol. In some embodiments, beta-like cells are generated from hPSCs using sequential modulation of multiple signaling pathways. In some embodiments, beta-like cells are generated from hPSCs using a three-dimensional cell culture system.

[0095] In some embodiments, beta-like cells are generated from human ESCs (hESCs) in vitro. In some embodiments, beta-like cells are generated from hESCs using directed differentiation. In some embodiments, beta-like cells are generated from hPSCs using a multi-step protocol. In some embodiments, beta-like cells are generated from hESCs using sequential modulation of multiple signaling pathways. In some embodiments, beta-like cells are generated from hESCs using a planar cell culture and air-liquid interface at different stages of differentiation. b) Non-stem cells

[0096] In some embodiments, beta-like cells are produced from non-stem cells. In some embodiments, beta-like cells are produced from differentiated non-beta cells. In some embodiments, beta-like cells are produced from reprogramming or transdifferentiation of differentiated non-beta cells.

[0097] In some embodiments, the beta-like cell is a reprogrammed non-beta cell. In some embodiments, the beta-like cell is a transdifferentiated non-beta cell.

[0098] As all cells of the body contain the full genome, any type of cell could be induced into a beta-like cell based on principles of reprogramming and transdifferentiation. Thus, the invention is not limited by the original phenotype of the beta-like cell.

[0099] Pancreatic exocrine cells can be reprogrammed into beta-like cells that secrete insulin (see Q Zhou et ak, Nature 455:627-633 (2008)).

[00100] In some embodiments, a pancreatic exocrine cell is reprogrammed into a beta-like cell. In some embodiments, the pancreatic exocrine cell is differentiated into a beta-like cell based on re-expression of transcription factors. In some embodiments, these transcription factors are NgnJ Pdxl, mdMafa.

[00101] Pancreatic alpha cells can be transdifferentiated into beta-like cells.

The anti-malarial drug, artemisin, inhibits the master regulatory transcription factor Arx (Aristaless related homeobox) and enhances gamma-amino butyric acid (GABA) receptor signaling, leading to impaired pancreatic alpha cell identity and transdifferentiation of alpha cells into a beta-like cell phenotype (see Li et al, Cell 168:86-100 (2017) and Ben-OthmanN et al, Cell 168(l-2):73-85 (2017)).

[00102] In some embodiments, the beta-like cell is a transdifferentiated cell. In some embodiments, an alpha cell is transdifferentiated into a beta-like cell. In some embodiments, the transdifferentiation into a beta-like cell is due to inhibition of Arx. In some embodiments, the transdifferentiation into a beta-like cell is due to enhancement of GABA receptor signaling.

[00103] Stomach tissue can be reprogrammed into beta-like cells (see Ariyachet C et al, Cell Stem Cell 18(3):410-21 (2016)). In some embodiments, a gut or stomach cell is reprogrammed into a beta-like cell. In some embodiments, the reprogramming is based on expression of beta cell reprogramming factors. In some embodiments, cells of the antral stomach are reprogrammed into beta-like cells. In some embodiments, these cells of the antral stomach are antral endocrine cells. In some embodiments, reprogrammed antral endocrine cells can be assembled into a mini-organ of beta-like cells.

[00104] Also encompassed is a method of preventing the development of type 1 diabetes by administering a MAOI. Certain individuals can be predicted to have a high risk for developing type 1 diabetes based on one or more risk factors, such as family history.

[00105] In some embodiments, a method of preventing the development of type 1 diabetes comprises screening a subject for risk factors for type 1 diabetes; determining if the subject has increased risk of developing type 1 diabetes; and administering a MAOI if the subject has an increased risk of type 1 diabetes.

[00106] In some embodiments, screening a subject for risk factors comprises obtaining data on a genetic risk score that is based on the known type 1 diabetes-associated gene variants, a family history of type 1 diabetes, the presence of one or more autoantibodies against beta cell antigens that are known to predict disease risk, and/or abnormal glucose tolerance.

[00107] Many individuals with type 1 diabetes have a genetic susceptibility because their genome comprises one or more type 1 diabetes-associated gene variant. The presence of one or more of these variants leads to an increased risk of type 1 diabetes. As these type 1 diabetes-associated gene variants can be inherited, a subject with a positive family history of type 1 diabetes may have an increased risk of developing the disease.

[00108] A wide variety of type 1 diabetes-associated gene variants have been described (See, for example, Watkins RA et ak, Transl Res. 164(2): 110-21 (2014)). In some embodiments, the one or more type 1 diabetes-associated gene variant are comprised in one or more HLA gene. In some embodiments, the one or more type 1 diabetes-associated gene variant are HLA polymorphisms conferring greater risk for type 1 diabetes. In some embodiments, the one or more type 1 diabetes-associated gene variant are comprised in one or more non-HLA gene.

[00109] In some embodiments, a family history of type 1 diabetes is determined by patient history or a questionnaire. In some embodiments, a family history of type 1 diabetes is based on one or more sibling, parent, or grandparent having type 1 diabetes.

[00110] In some embodiments, autoantibody levels against beta cell antigens are measured to determine increased risk of developing type 1 diabetes. A wide variety of autoantibodies against beta cell antigens have described in the literature (See. for example, Watkins 2014). Autoantibody panels are commercially available to identify individuals at risk of developing type 1 diabetes. Inclusion of certain antibodies, such as anti-ZnT8, in autoantibody levels can predict individuals at risk of developing type 1 diabetes. In some embodiments, the presence of one or more autoantibodies is used to determine an increased risk of developing type 1 diabetes. In some embodiments, the number of autoantibodies or the titer of a specific autoantibody is used to determine an increased risk of developing type 1 diabetes.

[00111] In some embodiments, an abnormal glucose tolerance is used to determine an increased risk of developing type 1 diabetes. In some embodiments, a subject with increased risk of developing type 1 diabetes shows abnormal glucose tolerance results without presently meeting criteria for type 1 diabetes.

[00112] In some embodiments, a subject is determined to have an increased risk of developing type 1 diabetes based on the presence of more than one risk factor. For example, a subject with a positive family history for type 1 diabetes may be determined to also have an abnormal glucose tolerance. Multiple risk factors for type 1 diabetes can assessed to determine a subject’s risk of developing type 1 diabetes. In some embodiments, a subject’s risk of developing type 1 diabetes is determined using an algorithm based on multiple risk factors (See. for example, Watkins 2014).

[00113] In some embodiments, a subject having an increased risk of type 1 diabetes is administered a MAOI. In some embodiments, administration of a MAOI prevents the development of type 1 diabetes in a subject with increased risk. In some embodiments, administration of a MAOI slows the time period until development of type 1 diabetes in a subject with increased risk.

EXAMPLES

Example 1. CRISPR screen for beta cell protective mutations identifies the T1D GWAS candidate gene Rnls

[00114] The selective pressure of autoimmunity was used to screen for protective gene mutations in beta cells on a genome-wide scale. To allow for efficient genome editing and experimental reproducibility, we employed the NIT-1 beta cell line, originally derived from a nonobese diabetic (NOD) mouse insulinoma (5). These cells are suitable for autologous transplantation into NOD mice, the most extensively studied animal model for type 1 diabetes (6). Of importance, NIT-1 cells transplanted into diabetic NOD mice are rapidly destroyed by autoimmunity (Figs. 6A-6B). NIT-1 cells were transduced with the mouse lentiviral GeCKO A CRISPR library that comprises ~ 60,000 gRNAs targeting a total of approximately 19,050 genes (7). Use of a low multiplicity of infection (MOI) ensured that cells carried at most one mutation. A total of 10⁷ mutant NIT-1 cells were transplanted into immuno-deficient NOO.scicI mice and injected splenocytes from diabetic NOD mice into transplant recipients to elicit beta cell killing (Fig. 1). Despite almost total beta cell destruction, a small population of NIT- 1 cells were retrieved after 8 weeks that survived the onslaught of autoimmunity. Targeted genes were identified by sequencing the gRNAs present in surviving beta cells. Only 11 unique gRNA sequences were detected, corresponding to 11 target genes, at significant frequencies in NIT-1 cells that survived autoimmune killing (Fig. 1). Notably, one of these genes was Rnls, the candidate gene for a region in the human genome associated both with the overall risk of T1D (8) and with the age of diabetes onset (9) by GWAS. Based on its prior association with human autoimmune diabetes, Rnls was prioritized for validation.

Example 2. Rnls deletion protects beta cells against autoimmune killing

[00115] A Rnls mutant NIT- 1 cell line (Rnls^mM) was generated using the Rnls gRNA identified in the screen (Figs. 7A-7B). NIT-1 cells were also engineered to carry a luciferase reporter for longitudinal non-invasive imaging of beta cells after transplantation (Figs. 6A-6B). A validation experiments was performed using an approach similar to the original genome-wide screen. As illustrated in Fig. 2A, Rnls^mut cells and control NIT-1 cells transduced with a non-targeting (NT) gRNA were co-transplanted on opposing flanks of NOD. sad mice. Transplant recipients were then injected with splenocytes from diabetic NOD mice. To control for beta cell survival and proliferation in the absence of autoimmunity, monitored beta cell transplants were monitored in NOD. sc id mice that did not receive diabetogenic immune cells. Control NIT-1 cells were killed by autoimmunity within one to two weeks after transplantation, as measured by loss of graft luminescence and analysis of grafts explanted from euthanized mice. In contrast, Rnls^mut NIT-1 cells persisted for up to 2 months in the same recipient mice (Figs. 2B-2D and data not shown). Next, the protective capacity of Rnls deletion was validated by implanting NIT-1 cells directly into overtly diabetic NOD mice (Fig. 2E). Again, control NIT-1 cells were rapidly eliminated while Rnls^mut cells survived significantly longer in diabetic NOD mice with ongoing autoimmunity (Figs. 2F-2H).

[00116] Disruption of Rnls was tested to see if it would similarly protect primary mouse beta cells. Pancreatic islets were isolated from immuno-deficient NODxc/ri mice that are devoid of autoimmune infiltrates in the pancreas. Dispersed islet cells were transduced with lentivirus encoding rat insulin promoter-driven Cas9 endonuclease and either the Rnls -targeting (Rnls"'"¹) gRNA or a non-targeting (NT) control gRNA (Fig. 21). Gene edited and control islets cells were transplanted each under one kidney capsule of the same NOD. sad mice. Two weeks later, graft recipients were injected with splenocytes from diabetic NOD mice to induce autoimmune beta cell killing. To control for the effects of gene disruption in the absence of autoimmunity, islet grafts were also followed in NOD. sad recipients that did not receive splenocytes from diabetic mice. As anticipated, autoimmunity decreased the size and insulin expression in control grafts (Figs. 2J-2L). In contrast, Rnls^mut islets survived autoimmunity and maintained insulin expression. These results show that targeting Rnls in primary beta cells was protective in a pathophysiologically relevant setting of autoimmune diabetes. Of note, Rnls targeting did not affect the insulin secretory capacity of islet cells in vitro (Fig. 8).

Example 3. Rnls mutation diminishes immune recognition of beta cells

[00117] The effect of Rnls deficiency was investigated for a direct effect on immune recognition. The expression MHC class I and class II molecules on the surface of /ri?/.v"^u'¹ NIT- 1 cells was comparable to that of control cells (Figs. 3A and 3B). Rnls did not significantly affect the response of beta cell-reactive (BCD2.5 TCR transgenic {10, 11))

CD4⁺ T cells co-cultured with antigen presenting cells and NIT-1 beta cells (12) (Fig. 3C). However, Rnls^mut NIT-1 cells elicited a significantly weaker response from polyclonal beta cell-reactive CD8⁺ T cells isolated from diabetic NOD mice (Fig. 3D). Because Rnls deficiency diminished the response of autoreactive cytotoxic T cells, Rnls^mut NIT-1 cells were tested for protection against T cell allo-reactivity. To test this, Rnls""" and control NIT-1 cells were transplanted into opposite flanks of MHC-mismatched C57BL/6 mice. Both beta cell grafts were rapidly destroyed by the strong allogenic response of host immune cells (Figs. 9A-9B), showing that Rnls deficiency did not affect allo-rejection. These data suggest that Rnls^mut beta cells are not impervious to immune detection or killing, but rather that they only fail to fully stimulate autoreactive CD8⁺ T cells.

Example 4. Rnls mutation confers ER stress resistance

[00118] A growing body of evidence supports a role for ER stress in the demise of beta cells in diabetes. The unfolded protein response (UPR) that is triggered by ER stress has been implicated in beta cell apoptosis in both T1D and type 2 diabetes (13-15). Significantly, ER stress was proposed to contribute not only directly but also indirectly to beta cell death in T1D owing its ability to increase the presentation of auto- and neoantigens, for example by affecting post-translational modifications (12, 16 18). The Rnls mutation could affect the cellular response to ER stress and thereby diminish the stimulation of diabetogenic CD8+ T cells. To test this notion, NIT-1 cells were challenged with the ER stressor thapsigargin (TG). Control cells were highly sensitive to TG treatment, with concentrations greater than 50 nM killing a majority of cells. Remarkably, Rnls mutant cells withstood even a 20-fold greater concentration of TG (Fig. 4A). Similar results were seen with the alternative ER stressor tunicamycin (TC) (Fig. 4B). Of note, Rnls^mut NIT-1 cells remained sensitive to mitomycin C and streptozotocin that cause ER stress-independent cell death (Figs. 10A-10B). These data indicate that Rnls deficiency does not prevent all forms of cell death and that its protective effect may be limited to specific sources of cellular stress. To ascertain that ER stress resistance was a direct effect of Rnls mutation and not caused by an off-target effect of the Rnls gRNA, additional cell lines were generated in which either exons 2 to 4 or exon 5 of the Rnls gene were deleted using different sets of gRNAs. These alternative /rii/.v-deficient beta cell lines were again protected against ER stress-induced cell death (Figs. 11 A-l ID), confirming that ER stress resistance was a direct result of Rnls deletion.

Example 5. Rnls overexpression sensitizes beta cells to ER stress and autoimmunity

[00119] To further evaluate the role of Rnls in modifying the sensitivity of beta cells to ER stress and autoimmunity, Rnls was overexpressed in NIT-1 beta cells using a lentiviral transgene. While Rnls overexpression alone only marginally increased sensitivity to TG-induced killing (Fig. 12A-12B), it appeared to significantly accelerate the autoimmune killing of beta cells implanted into diabetic mice (Fig. 13A-13C). Re-introduction of Rnls into Rnls^mut cells was done using a transgene that carried a synonymous mutation at the gRNA target site to prevent CRISPR-Cas9 targeting. Rnls re-expression restored the sensitivity of Rnls^mut cells to ER stress (Fig. 12A-12B). Moreover, Rnls re-expression also accelerated the autoimmune destruction of Rnls^mut cells in diabetic NOD mice (Figs. 14A-14B). Collectively, the data show that Rnls expression modulates the vulnerability of beta cells to ER stress and autoimmunity. Of interest, Rnls expression was 10-15 fold higher in pancreatic islets of diabetes-prone NOD mice than in diabetes-resistant C57BL/6 mice (Fig. 15). Elevated Rnls expression was not merely a result of pancreas inflammation, because similar Rnls mRNA levels were measured in the islets of both NOD and immuno-deficient NODxc/ri mice. This intriguing observation suggests the Rnls may be a genetically encoded modifier of beta cell vulnerability in both mouse and human, though exactly how Rnls expression is regulated in both species remains to be elucidated.

Example 6. Rnls modifies the cellular response to ER stress

[00120] To understand how Rnls deficiency increases ER stress resistance, the UPR that mediates the cellular adaptation to ER stress was measured. Activation of critical ER stress sensors IREla (19), PERK (20) and ATF6 (21) was diminished in Rnls"'¹'¹ cells following TG treatment (Figs. 4C and Figs. 16A-16H). Downstream of these UPR triggers, the phosphorylation of eIF2a, protein levels of ATF4 and splicing of XBP1 were markedly reduced (Figs. 4C and 4D, and Figs. 16A-16H). The expression of Chop and Txnip, both implicated in ER stress-induced apoptosis (22-24), was also diminished (Fig. 4E and Figs. 16A-16H). The data suggest that Rnls deficiency increased the threshold for ER stress that triggers the UPR, which could explain how Rnls mutation inhibits the pro-apoptotic effect of stimuli that cause cellular stress. The protective effect of Rnls deletion was not limited to ER stress, because Rnls^mut cells also better withstood oxidative stress compared to control NIT-1 cells (Fig. 17). Consistent with this finding, Rnls deficiency increased the expression of a key regulator of the oxidative stress response, NRF2 (25) (Fig. 4C and Fig. 16H). Rnls deficiency appears to increase the ability of beta cells to withstand cellular stress involved in their destruction during type 1 diabetes.

Example 7. Pargyline phenocopies the protective effects of Rnls deletion

[00121] Rnls is a flavoprotein oxidase whose cellular function has not yet been elucidated (26). Its proposed substrates are 2- and 6-dihydroNAD(P) (27), isoforms of b- NAD(P)H, though whether these are physiologically relevant is unknown. However, the crystal structure of human RNLS was solved several years ago (28). The enzyme utilizes an FAD co-factor for catalysis, resembling other oxidases including monoamine-oxidase B (MAO-B). Based on structural similarities, the FDA-approved MAO-B inhibitor pargyline (29) was predicted bind to RNLS (Fig. 5A). To test this prediction, the thermal stability of RNLS was measured in the presence or absence of pargyline (Figs. 5B and 5C). Pargyline decreased the thermal stability of RNLS in a dose-dependent manner, suggesting a direct interaction between the drug and the enzyme. Pargyline is an oral drug and is water soluble, lending itself to treating mice via their drinking water. Pargyline’ s efficacy was evaluated in a stringent beta cell transplantation model. Recently diabetic NOD mice with severe hyperglycemia (blood glucose > 600 mg/dL) were transplanted with NIT-1 beta cells with or without continuous drug feeding (Fig. 5D). Graft survival was again monitored longitudinally by non-invasive bioluminescence imaging (Fig. 5E). Untreated mice remained hyperglycemic and rapidly lost their beta cell graft (Fig. 5E-5G). Remarkably, pargyline treatment allowed transplanted beta cells to survive in diabetic mice, produce insulin and reverse hyperglycemia (Figs. 5E-5H). Upon histological analysis of the pancreas 3 weeks after diabetes onset and transplantation, pargy line-treated mice still harbored a significant number of insulin-rich islets (Fig. 18). In contrast, the pancreas of untreated diabetic mice was entirely devoid of insulin staining. These observations suggest that pargyline not only protected grafted NIT-1 cells but also endogenous beta cells against autoimmunity, recapitulating the protective effect of Rnls deletion. Of note, pargyline did not prevent the allo-rejection of NIT- 1 cells transplanted into C57BL/6 mice, indicating that the drug is not immunosuppressive (Fig. 19A19B). This observation again replicates the effects of Rnls deletion that confered protection against autoimmunity but not allo-reactivity.

[00122] These data show that RNLS is a modifier of beta cell vulnerability in T1D. This finding may explain why genome variants in the RNLS locus impact the overall risk ( 8 ) and the age of onset (9) of T1D. How disease-associated variants modify RNLS function or expression is unknown. In light of our results, exploring how this candidate T1D risk gene is regulated seems warranted. Rnls was differentially expressed in islets of diabetes- prone NOD mice and diabetes-free C57BL/6 mice, lending further support to the notion that RNLS may be a genetic risk factor for beta cell autoimmunity.

[00123] These data also underscore the central role of beta cell ER stress in promoting islet autoimmunity. The ER and oxidative stress resistance afforded by Rnls deficiency was correlated with protection against autoimmunity, consistent with a growing body of literature that implicates ER stress in T1D.

[00124] Significantly, the discovery that RNLS deficiency endows beta cells with the ability to resist autoimmunity suggests a genetic engineering solution to beta cell replacement in T1D that would interfere neither with the identity of the beta cell nor with immunity and immune surveillance. RNLS deletion could be a safe and effective modification in SC-beta cells to overcome autoimmunity in patients with T1D. Finally, an FDA-approved drug that replicates the protective effect of RNLS deletion. Its apparent efficacy in protecting beta cells, together with its favorable safety profile, should make pargyline, and other MAO- B inhibitors predicted to target RNLS, worthy of further evaluation for the prevention or treatment of type 1 diabetes. Example 9. Evaluation of MAO inhibitors as RNLS inhibitors

[00125] To test the predicted binding of monoamine oxidase-inhibitors (MAOI’s) to human recombinant renalase (rRNLS), thermal shift assays were performed, whereby the unfolding of the protein at increasing temperature was measured using a hydrophobic fluorescent probe. With increasing temperature, fluorescence increases as the protein unfolds and exposes hydrophobic residues to which the probe can bind.

[00126] Human recombinant RNLS protein was generated by GenScript USA Inc., using the E. Coli expression vector pET28a-MBP. RNLS protein was obtained from the supernatant of cell lysates, followed by purification via Ni Bio-rad column. 2 mM RNLS dissolved in PBS was incubated with pargyline (Sigma-Aldrich, #P8013), rasagiline (Tocris, cat# 4308), or selegiline (Tocris, cat# 1095) at 100 mM for 20 min at 4 °C before addition of SYPRO Orange dye (Invitrogen, #S6650) for the measurement of thermal denaturation. The thermal shift assay was performed using the QuantStudio 6 Flex Real-Time PCR system (Applied Biosystems) with an initial temperature hold at 25 °C for 2 min, followed by a temperature ramp up to 95°C at a rate of 1°C / s, and a final temperature hold at 95°C for 2 min. Results were collected at 0.25 °C increments.

[00127] Three MAO-Ts in the propargylamine class of drugs (pargyline (Figure 20 A), rasagiline (Figure 20B), and selegiline (Figure 20C)) shifted the thermal unfolding curve of rRNLS, demonstrating that these compounds directly interact with the enzyme, as predicted by structure-based modeling.

Example 10. In vivo effects of pargyline treatment in diabetes prevention models

[00128] To test the ability of the MAO-I pargyline to prevent the onset of autoimmune diabetes in the nonobese diabetic (NOD) mouse model, the best characterized model for type 1 diabetes, animals were administered oral pargyline prior to disease induction. Diabetes was induced using either cyclophosphamide (Figure 21A) or anti-PD-1 antibody (Figure 2 IB). Cyclophosphamide is a compound known to accelerate diabetes onset in the NOD model (See Harada M & Makino S, Diabetologia 1984, 3: 429). PD-1 blockade has also been shown to cause rapid onset of diabetes in NOD mice (See Ansari MJ et al J Exp Med 2003, 198: 63), and immune checkpoint inhibition (including PD-1 inhibition) for the treatment of various cancers can cause autoimmune diabetes in some patients (reviewed in Clotman K et al J Clin Endocrinol Metab 2018, 103: 3144).

[00129] 10-week-old male or female NOD mice were injected intraperitoneally with cyclophosphamide (200 pg/g of body weight, Sigma-Aldrich, #C0768) or anti-PD-1 (250 mg/mouse i.v., BioXcell, #BE0146) for diabetes induction. Pargyline treatment was started one week prior disease induction. Blood glucose was monitored every 1-2 days to test for diabetes onset.

[00130] Pargyline treatment protected against diabetes induction and significantly decreased the proportion of diabetic animals after either cyclophosphamide or anti-PD-1 treatment (Figs. 21A-21B).

[00131] The protective effects of pargyline treatment were also evaluated in additional mouse models for type 1 diabetes.

[00132] In one type 1 diabetes model, diabetes was induced by adoptive transfer (transplantation by intravenous injection) of 10 million splenocytes from diabetic NOD mice into immuno-deficient NOD.scid mice treated or not with pargyline (25pg/ml pargyline in the drinking water). Recipient mice were tested for diabetes for 110 days following cell transfer. Kaplan-Meier survival curves were compared by log-rank test, and treatment with pargyline significantly increased survival (P=0.027, Figure 22A).

[00133] In another type 1 diabetes model, diabetes was induced in C57BL/6 mice by repeated injection of low doses of streptozotocin (STZ, 5 consecutive daily intraperitoneal injections of STZ at 50mg/kg/day). Mice were treated with pargyline (25pg/ml pargyline in the drinking water) starting one week before induction of diabetes. Kaplan-Meier survival curves were compared by log-rank test (P=0.01, Figure 22B). The day of diabetes onset in diseased mice was compared using the Mann- Whitney test (P=0.004, Figure 22C). Pargyline treatment significantly increased survival and delayed the onset of diabetes following STZ treatment.

Example 11. Screen for additional MAOI that bind to RNLS

[00134] Additional known MAOIs were tested for rRNLS binding by thermal shift assay. The cyclopropylamine tranylcypromine (Figure 20D) and the hydrazine isocarboxazid (Figure 20E), both irreversible MAOIs, and the reversible MAO-A inhibitor toloxatone (Figure 20F), shifted the thermal unfolding curve of rRNLS in a manner comparable to pargyline, demonstrating that these compounds also directly interact with the enzyme.

Example 12. Materials and Methods

[00135] The following methods were used in the experiments.

Mice [00136] Nonobese diabetic (NOD) mice, NOD.scid (NOD. CB Y1 -Prkdc^scld! J) and C57BL/6J were purchased from the Jackson Laboratory (Bar Harbor, ME). Animals were housed in pathogen-free facilities at the Joslin Diabetes Center and all experimental procedures were approved and performed in accordance with institutional guidelines and regulations.

CRISPR GeCKOA library screen

[00137] The mouse GeCKO-v2 (Genome-Scale CRISPR Knock-Out) A lentiviral pooled library was obtained from Addgene (Addgene, #1000000052), targeting 19050 with 3 gRNAs/gene (30) and was prepared as previously described (31). Wild type NIT-1 cells (ATCC #CRL-2055) were infected with GeCKO A CRISPR lentiviral library at MOI=0.3, and then selected by puromycin (2 pg/ml) at day 3 post lentivirus infection. 10⁷ mutant NIT-1 cells were transplanted subcutaneously into 8-week-old female NOD.sc/ri mice, and 10⁷ of diabetic NOD splenocytes in 200 mΐ sterile PBS were injected intravenously at the same time to induce autoimmunity. NOD.sc/ri mice with subcutaneously transplanted mutant NIT-1 cells but without diabetic NOD splenocytes injection were used as control (non-autoimmune group). Diabetic NOD splenocytes were isolated from spontaneously diabetic female NOD mice as described previously (32). The screen was terminated at 8 weeks post-injection and the remaining grafts were retrieved from both the autoimmune group and the non-autoimmune group of mice. Genomic DNA was extracted from the grafts (Quick-gDNA midiprep kit, Zymo Research), the NGS (Next Generation Sequencing) libraries were prepared as previously described (33), and subjected to NGS sequencing analysis (Novogene, CA). The gRNA sequences from the NGS sequencing data were extracted using standard bioinformatics methods, and the distribution of gRNAs were calculated as Count Per Million (CPM).

Cell line

[00138] NIT-1 (#CRL-2055) and 293FT (#R7007) cell lines were obtained from ATCC and Thermo Fisher Scientific respectively. Cells were maintained in DMEM (Gibco, 10313039), supplemented with 10% fetal bovine serum (FBS, Gibco), glutagro and penicillin/streptomycin (Coming), in a 37° C incubator with 5% CCh. To generate control and Rnls^mut NIT-1 cells, non-targeting (NT) gRNA (5’ TAAAAACGCTGGCGGCCTAG 3’, MGLibA_67395 (SEQ ID NO: 2)) mdRnls gRNA (5’ CTACTCCTCTCGCTATGCTC 3’, MGLibA_46009 (SEQ ID NO: 1)) were cloned into LentiCRISPR-v2 vector, and the NT or Rnls gRNA containing lentivirus was used to establish these cell lines, respectively. Rnls mutation in Rnls^mut cells was confirmed by deep sequencing analysis (MGH DNA Core Facility, Cambridge, MA). The Rnls overexpressing NIT-1 cell line was generated by lentiviral infection of wild-type (WT) NIT-1 cells with EFla promoter-driven full-length mouse Rnls (Of note, the full-length mouse Rnls that we cloned and used is based on the annotation from the NCBI database in late 2017 that included 300aa, the Rnls annotation in NCBI was updated in March 2019 and now encodes a protein with 42 additional amino acids at the N-terminus). For generation of (V/rii/.v-expressing Rnls^mut cells, Rnls mutant NIT-1 cells were transduced with lentivirus carrying a CRISPR-immune EFla promoter-driven full- length mouse Rnls ( CiRnls ) carrying a synonymous mutation in the Rnls gRNA target site. The modified gRNA targeting site sequence used in CiRnls was 5’

TTATAGTAGCCGGTACGCA 3’(SEQ ID NO: 3). The /rii/.v-deficient NIT-1 cell lines (Rnls AEx2/4 and Rnls AEx5) were generated following previously published protocols (36). Two gRNAs were designed to target the 5’- and 3 ’-end of Rnls exon 2-4 or exon 5 genomic DNA sequences. gRNA sequences for exon 2-4 were 5’CGTCTGGGAAGTCTTGGTCG 3’ and 5’ CGGGACTCATCCCATTGTCG 3’ (SEQ ID NOs: 4-5); gRNA sequences for exon 5 were 5 ’ GGGGAGTGAGGATAGGATAG 3’ and 5’ TCCGTAGTGGTTTTAGAGTG 3’ (SEQ ID NOs: 6-7). The lenti-multi-CRISPR plasmid (Addgene, #85402) was used to express two single gRNAs cassettes for the deletion of exons 2-4 or exon 5 of Rnls. The two gRNAs cassettes were amplified by Phusion High-Fidelity PCR kit: 40 cycles: 98°C, 15 sec; 60°C,

15 sec; 72°C, 30sec. The PCR products were digested with Bbsl (Invitrogen) and sub-cloned into the pSpCas9(BB)-2A-Puro (PX459) v2.0 vector (Addgene, #62988). NIT-1 cells were then transfected with these plasmids by polyethylenimine (Fisher Scientific), followed by puromycin selection. All plasmid sequences were verified by Sanger sequencing before transduction and transfection.

Preparation and transplantation of primary islets

[00139] Islets were prepared and purified as described previously (35) from 8 week old female NOD.sc/ri mice. Purified NOD.sc/ri islets were immediately cultured in a low-attachment plate in RMPI 1640 medium (Gibco), supplemented with 10% FBS and penicillin/streptomycin. Lentivirus encoding a NT or Rnls gRNA together with Cas9 endonuclease under the control of the rat insulin promoter (RIP) was added to the culture media for overnight infection. The next day, islets were washed with culture media twice and -300 islets were transplanted under each kidney capsule of 8 week old of female NOD.sc/ri mice. Graft recipients were left to recover from surgery for two weeks, then mice were randomly assigned to non-autoimmune and autoimmune groups. Mice in the autoimmune group were injected intravenously with 10⁷ splenocytes purified from spontaneously diabetic female NOD mice. Splenocytes were prepared as described previously (32). At day 25 post splenocyte injection, islet grafts were retrieved for gene expression analysis by quantitative real-time PCR (qPCR).

Quantitative real-time PCR (gPCR)

[00140] Cells or islet grafts were treated with TRIzol (Thermo Fisher Scientific) for RNA extraction following the manufacturer’s protocol. Purified RNA was reverse-transcribed into cDNA using the Superscript IV first-strand synthesis kit (Invitrogen). Insulin 1 (MmO 1259683_g 1 ), Glucagon (Mm01269055_ml) mdHprtl (Mm0302475_ml) probes for TaqMan assays were purchased from Thermo Fisher Scientific. Gene expression levels of Chop and Txnip were analyzed by SYBR green PowerUp qPCR assays (Applied Biosystems). Primer sequences used for Chop· forward - 5’ CCACCACACCTGAAAGCAGAA 3’ (SEQ ID NO: 8); reverse - 5 ’ AGGTGAAAGGC AGGGACTC A 3’(SEQ ID NO: 9); Txnip·. forward - 5’ TCAAGGGCCCTGGGAACATC 3’ (SEQ ID NO: 10); reverse - 5’

GACACTGGTGCCATTAAGTCAG 3’ (SEQ ID NO: 11). All qPCR assays were performed using a QuantStudio 6 Flex Real-Time PCR system (Applied Biosystems).

Cell viability assay

[00141] Cells were seeded in 96-well white plate (50,000 cells / well) for overnight culture with or without thapsigargin, tunicamycin, streptozotocin (Sigma- Aldrich), mitomycin C (Fisher Scientific) or hydrogen peroxide (H2O2, Fisher Scientific) at the indicated concentrations. Cell viability was assessed after 24 h using the CellTiter-Glo luminescence Cell Viability Assay (Promega).

In vivo bioluminescence imaging

[00142] NIT-1 cell lines were engineered to constitutively express the firefly luciferase gene (Luc2) driven by the EFla promoter via lentiviral delivery, except in Fig. 7, where a CMV-Luc2 construct was used. Mice transplanted with luciferase-expressing cells were injected with D-luciferin intraperitoneally at a dose of 150 mg/kg for bioluminescence imaging. D-luciferin (Gold Biotechnology, Cat# LUCK) solution was prepared in sterile DPBS (without calcium or magnesium) at a concentration of 15 mg/ml and filtered (0.22 pm) prior to injection. Luminescence was measured using an IVIS Spectrum imaging system (PerkinElmer). Xbpl splicing assay

[00143] Cells were treated either with DMSO or thapsigargin at 1 mM for 5 h. RNA was extracted by TRIzol and reverse transcribed into cDNA as described for qPCR. Spliced (s) and unspbced (u) Xbpl cDNA were amplified by PCR using the Phusion High- Fidelity DNA polymerase (Invitrogen) for 35 cycles: 94°C, 10 sec; 64°C, 30 sec; 72°C,

30sec. The PCR products oiXbpl were sized at 473bp iXbplu) and 447bp ( Xbpls ) and segregated by electrophoresis using a 3% agarose gel. Ratio of Xbpls /Xbplu was measured by Adobe Photoshop CC 2019. Primer sequences used iorXbpl were forward - 5’ AAACAGAGTAGCAGCGCAGACTGC 3’ (SEQ ID NO: 12) and reverse - 5’ TCCTTCTGGGTAGACCTCTGGGAG 3’ (SEQ ID NO: 13).

Western blotting

[00144] Cell lysates were collected on ice in RIPA lysis buffer containing proteinase and phosphatase inhibitors (complete proteinase inhibitor cocktail, Sigma- Aldrich; Pierce phosphatase inhibitor, Thermo Scientific Fisher). Protein concentrations were measured by Pierce BCA protein assay (Thermo Scientific Fisher). 40 pg denatured cell lysate protein were used for SDS-PAGE electrophoresis (4-20% TGX gel, Bio-Rad). The following primary antibodies were used: PERK (Cell signaling, #3192), phospho-PERK (Thr980, Cell signaling, #3179), ATF4 (Cell signaling, #11815), ATF6 (Novus Biologicals, #NBP1-40256SS), eIF2a (Cell signaling, #2103), phosphor-eIF2a (Ser51, Cell signaling, #3597), IREla (Cell signaling, #3294), phosphor-IREla (Ser724, Novus Biologicals, #NB100-2323SS), Txnip (MBL, K0205-3), cleaved Caspase 3 (Cell signaling, #9664S), NRF2 (Santa Cruz, #SC365949) and actin (ABclonal, #AC004). All images were obtained and quantified using a C-DiGit blot scanner and the Image Studio software (LI-COR Biosciences).

T7 endonuclease I assay

[00145] CRISPR/Cas9 editing in Rnls""" cells was detected by T7 endonuclease I mismatch cleavage assay. Genomic DNA (gDNA) was purified from control and Rnls^mut NIT-1 cells using Quick-gDNA miniprep kit (Zymo Research). The Rnls gRNA targeting site was amplified using the Phusion high-fidelity PCR kit (Thermo Fisher Scientific). Primers for Rnls gRNA site PCR were: forward 5’ TGCTATAGACAGTTGGGACTTGTTT 3’ (SEQ ID NO: 14); reverse 5’ AT ATT GC GTT C T ATT AT C A AT GGAGAT GA AGC 3’ (SEQ ID NO: 15). The PCR products (-200 ng) were used to form heteroduplexes by denaturing at 95 °C for 5 min and then re-annealing the products in a thermocycler using the following protocol: ramp down to 85°C at -2°C/sec; ramp down to 25°C at -0.1°C/sec; hold at 4°C. 10 units T7 endonuclease I was added to the annealed PCR products and the reaction was incubated at 37°C for 15 min. The digestion reaction was stopped by 1 mΐ 0.5M EDTA and immediately applied to a 1.5% agarose gel to visualize digested and undigested products by electrophoresis.

CD4⁺ T cell stimulation assay

[00146] CD25 CD4⁺ T cells were isolated from BDC2.5-TCR transgenic (Tg)

NOD mice using a CD25⁺ regulatory T cell isolation kit (Miltenyi, 130-091-041). Purified CD4⁺ T cells were maintained in culture for three weeks prior to being cultured with NIT-1 cells by weekly stimulation with 1 mM BDC2.5 mimotope in the presence of irradiated splenocytes from NOD.sc/ri mice and 20 U/mL IL-2 (Peprotech, 212-12-20UG). The day prior to T cell and NIT-1 co-culture, NIT-1 cell lines were incubated with or without 1 mM Thapsigargin for 5 h. NIT-1 cells were washed extensively with complete DMEM after incubation. 5xl0⁴ NIT-1 cells were then seeded in each well of a 96-well plate in 100 pL culture medium. The next day, 10⁵ BDC2.5-Tg CD4⁺ T cells and 5 x 10⁵ NOD.sc/ri splenocytes re-suspended in 100pL RPMI medium were added to NIT-1 cultures. Cells were co-cultured in a 37° C incubator with 5% CCh for 24 hours. Cells were treated with BD Golgi plug (diluted 1 in 1,000; BD bioscience, #555029) for the last 5 h of culture. After the incubation, cells were collected and stained with the following antibodies, all of which were purchased from BioLegend; BV785 CD3 (clone 17A2, #100231), PE-Cy7 CD4 (clone GK1.5, #100421), PE anti-TNF-a (clone MP6-XT22, #506306), PE rat IgGl kappa isotype control (clone RTK2071, #400407), APC IFN-g (clone XMG1.2, #505810), and APC rat IgGl kappa isotype control (clone RTK2071, #400411). Zombie Violet Fixable Viability Kit (BioLegend, #423114) was used for dead cell staining, and BD Cytofix/Cytoperm Plus (BD bioscience, #554714) was used for intra-cellular cytokine staining, following the manufacturer’s instructions. Flow cytometry was performed on a LSR fortessa instrument (BD Biosciences), and data were analyzed using Flow Jo vl0.6 (Flow Jo, LLC).

MHC class I and class II expression analyses

[00147] 5 x 10⁵ NIT-1 cells were seeded in each well of a 24-well plate in 500 pL complete DMEM medium.

[00148] Cells were treated with or without 1.25 mM Thapsigargin in a 37° C incubator with 5% CCh for 24 hours, collected and stained with the following antibodies, all of which were purchased from BioLegend; APC anti-mouse H-2Kd (clone SFl-1.1, #116620), APC mouse IgG2a kappa isotype control (clone MOPC-173, #400219), PE anti mouse IA^k (Ap^k) (clone 10-3.6, #109908) which cross-reacts with mouse I-Ag⁷, PE mouse IgG2a kappa isotype control (clone MOPC-173, #400213). Flow cytometry was performed as described above.

CDS T cell ELISPOT assay

[00149] 5 pg/mL capture antibody (BD NA/LETM Purified Anti-mouse IFN- g, BD Biosciences, #51-2525KA) was coated on a 96-well ELISPOT plate (Millipore Sigma, #MAIPS4510) overnight at 4⁰ C. The following day, the plate was blocked with 200 pL complete RPMI medium for 2 hours at room temperature. NIT-1 cells were treated with lOOU/mL IFN-g for 72 hours prior to use in the assay, washed and suspended at 10⁵ cells / 100 pL in complete DMEM medium. CD8⁺ T cells were isolated and purified from a female diabetic NOD mouse using mouse CD8a⁺ Isolation Kit (Miltenyi, #130-104-075). 10⁵ CD8⁺

T cells were re-suspended in 100 pL complete RPMI medium. NIT-1 cells and CD8⁺ T cells were then mixed at a 1 : 1 ratio in the antibody-coated 96-well plate and co-cultured for 24 hours in a 37° C incubator with 5% CO2. Cells were discarded and the plate was washed with PBS-0.1% Tween20 washing buffer three times. 2 pg/mL detection antibody (Biotinylated Anti-mouse IFN- g, BD Biosciences, #51-1818KA) was added to each well and the plate was incubated for 2 hours at room temperature. Plates were washed three times with washing buffer and HRP-conjugated streptavidin (BD Biosciences, #557630) was added to each well for one hour. After washing four times with washing buffer and three more times with PBS, substrate solution (R&D, #DY999) was added to each well and incubated for 15-30 min. The reaction was stopped by adding deionized water to each well and the plate was further washed with deionized water. Spots were counted and analyzed on the Immunospot S6 Universal-V instrument (Cellular Technology Limited).

RNLS structure analyses and drug-binding modeling

[00150] hRNLS uses flavin adenine dinucleotide (FAD) as a co-factor for catalysis. Therefore, to find compounds that could potentially inhibit hRNLS, the Protein Data Bank was searched for protein-inhibitor complex structures that had FAD. MAO-B in complex with inhibitors that covalently attached to FAD were identified through the search. Structural alignments and analysis of the hRNLS crystal structure with these complex structures based on FAD suggested that these inhibitors, for instance pargyline, may inhibit hRNLS as well. The model of full-length hRNLS in complex with pargyline was built based on the crystal structures of human renalase (PDB: 3QJ4) (36) and MAO-B rasagiline complex (PDB: 2C65) (37). The model was first optimized using the Protein Preparation Wizard (38) from Schrodinger at pH 7.0 and energy minimized with gradually reduced restraints (1000, 5, 0 force constant) on backbone and solute heavy atoms. A multi-stage 100 ns molecular dynamics (MD) simulation using Desmond (39) was performed afterwards. The final frame of the MD simulation was used as the final model in Figure 6 A.

RNLS thermal shift assay

[00151] Human recombinant RNLS protein was generated by GenScript USA Inc., using the E. Coli expression vector pET28a-MBP. RNLS protein was obtained from the supernatant of cell lysates, followed by purification via Ni Bio-rad column. 2 mM RNLS dissolved in PBS was incubated with pargyline (Sigma-Aldrich, #P8013) at concentrations of 0, 0.1, 1, 10, 25, 50, 100 mM for 20 min at 4 °C before addition of SYPRO Orange dye (Invitrogen, #S6650) for the measurement of thermal denaturation. The thermal shift assay was performed using the QuantStudio 6 Flex Real-Time PCR system (Applied Biosystems) with an initial temperature hold at 25°C for 2 min, followed by a temperature ramp up to 95°C at a rate of 1°C / s, and a final temperature hold at 95°C for 2 min. Results were collected at 0.25 °C increments. The melting temperature (Tm) of RNLS in the presence and absence of pargyline was calculated by the first derivative of the fluorescence emission as a function of temperature (- dF/dT).

Oral mrsyline treatment study

[00152] 9-week-old female NOD mice were injected intraperitoneally with cyclophosphamide (200 pg/g of body weight, Sigma-Aldrich, #C0768) for diabetes induction. Diabetic NOD mice (blood glucose >450 mg/dL) identified 10-14 days later were randomly assigned to the control group (normal water) or pargyline treatment group (5 pg/ml pargyline hydrochloride in the drinking water, Sigma-Aldrich #P8013). Treatment was started one week prior to beta cell transplantation. NIT-1 beta cells carrying a luciferase reporter were pre-treated with 5 mM pargyline for 24 hours before transplantation. 10⁷ NIT-1 cells were transplanted subcutaneously into each diabetic NOD mouse. Blood glucose was monitored every 1-2 days, and graft bioluminescence was imaged every 2-3 days. Glucose-stimulated insulin secretion and insulin ELISA

[00153] Primary islets were isolated from 8-week old CD 1 (ICR) mice (Envigo) and immediately cultured in a 24-well low-attachment plate. Islets were transduced with lentivirus encoding a non-targeting (NT) or Rnls gRNA together with rat insulin promoter- driven Cas9 endonuclease. After 72 hours, islets were washed twice with 1 ml Krebs Ringer Bicarbonate HEPES (KRB) buffer containing 2.8 mM glucose, followed by 1 hour incubation at 37 °C in 2.8 mM glucose KRB buffer. 0.8 ml KRB was taken out, saved for insulin measurement and replaced with 0.8 ml of 20.2 mM glucose KRB buffer for a final glucose concentration of 16.8 mM for another 1 hour incubation at 37 °C. The KRB buffer was again sampled for insulin, then islets were incubated with 30 mM KC1 along with 16.8 mM glucose for 1 hour at 37 °C before the final insulin sampling. Genomic DNA was purified from islets for normalization of insulin levels to DNA content. Insulin levels were assessed by ultra sensitive mouse insulin ELISA kit (Crystal Chem, #90080).

Statistical analyses

[00154] Statistical analyses were performed by unpaired or paired tests as indicated using the Prism software version 8.0.2. All data are presented as mean ± SEM. P < 0.05 was considered statistically significant. Sufficient sample size was estimated without the use of a power calculation. No samples were excluded from the analysis. No randomization was used for animal experiments. Data analysis was not blinded. All data are representative of two or more similar experiments.

[00155] References

1. M. A. Atkinson, B. O. Roep, A. Posgai, D. C. S. Wheeler, M. Peakman, The challenge of modulating b-cell autoimmunity in type 1 diabetes. Lancet Diabetes Endocrinol. 7, 52-64 (2019).

2. J. Odorico, J. Markmann, D. Melton, J. Greenstein, A. Hwa, C. Nostro, A. Rezania, J. Oberholzer, D. Pipeleers, L. Yang, C. Cowan, D. Huangfu, D. Egli, U. Ben-David, L. Vallier, S. T. Grey, Q. Tang, B. Roep, C. Ricordi, A. Naji, G. Orlando, D. G. Anderson, M. Poznansky, B. Ludwig, A. Tomei, D. L. Greiner, M. Graham, M. Carpenter, G. Migliaccio,

K. D’Amour, B. Hering, L. Piemonti, T. Bemey, M. Rickels, T. Kay, A. Adams, Report of the Key Opinion Leaders Meeting on Stem Cell-derived Beta Cells. Transplantation. 102, 1223-1229 (2018).

3. X. Han, M. Wang, S. Duan, P. J. Franco, J. H.-R. Kenty, P. Hedrick, Y. Xia, A. Allen,

L. M. R. Ferreira, J. L. Strominger, D. A. Melton, T. B. Meissner, C. A. Cowan, Generation of hypoimmunogenic human pluripotent stem cells. Proc. Natl. Acad. Sci. 116, 10441-10446 (2019).

4. T. Deuse, X. Hu, A. Gravina, D. Wang, G. Tediashvili, C. De, W. O. Thayer, A.

Wahl, J. V. Garcia, H. Reichenspumer, M. M. Davis, L. L. Lanier, S. Schrepfer, Hypoimmunogenic derivatives of induced pluripotent stem cells evade immune rejection in fully immunocompetent allogeneic recipients. Nat. Biotechnol. 37, 252-258 (2019).

5. K. Hamaguchi, H. R. Gaskins, E. H. Leiter, NIT-1, a pancreatic beta-cell line established from a transgenic NOD/Lt mouse. Diabetes. 40, 842-9 (1991).

6. J. A. Pearson, F. S. Wong, L. Wen, The importance of the Non Obese Diabetic (NOD) mouse model in autoimmune diabetes. J. Autoimmun. 66, 76-88 (2016).

7. N. E. Sanjana, O. Shalem, F. Zhang, Improved vectors and genome-wide libraries for CRISPR screening. Nat. Methods. 11, 783-784 (2014).

8. J. C. Barrett, D. G. Clayton, P. Concannon, B. Akolkar, J. D. Cooper, H. A. Erlich, C. Julier, G. Morahan, J. Nerup, C. Nierras, V. Plagnol, F. Pociot, H. Schuilenburg, D. J. Smyth, H. Stevens, J. A. Todd, N. M. Walker, S. S. Rich, Genome-wide association study and meta analysis find that over 40 loci affect risk of type 1 diabetes. Nat. Genet. 41, 703-7 (2009).

9. J. M. M. Howson, J. D. Cooper, D. J. Smyth, N. M. Walker, H. Stevens, J.-X. She, G. S. Eisenbarth, M. Rewers, J. A. Todd, B. Akolkar, P. Concannon, H. A. Erlich, C. Julier, G. Morahan, J. Nerup, C. Nierras, F. Pociot, S. S. Rich, Type 1 Diabetes Genetics Consortium, Evidence of gene-gene interaction and age-at-diagnosis effects in type 1 diabetes. Diabetes. 61, 3012-7 (2012).

10. K. Haskins, M. Portas, B. Bergman, K. Lafferty, B. Bradley, Pancreatic islet-specific T-cell clones fromnonobese diabetic mice. Proc. Natl. Acad. Sci. 86, 8000-8004 (1989).

11. J. D. Katz, B. Wang, K. Haskins, C. Benoist, D. Mathis, Following a diabetogenic T cell from genesis through pathogenesis. Cell. 74, 1089-100 (1993).

12. M. L. Marre, J. L. Profozich, J. T. Coneybeer, X. Geng, S. Bertera, M. J. Ford, M. Trucco, J. D. Piganelli, Inherent ER stress in pancreatic islet b cells causes self-recognition by autoreactive T cells in type 1 diabetes. J. Autoimmun. 72, 33-46 (2016).

13. A. L. Clark, F. Urano, Endoplasmic reticulum stress in beta cells and autoimmune diabetes. Curr. Opin. Immunol. 43, 60-66 (2016).

14. U. Ozcan, Q. Cao, E. Yilmaz, A.-H. Lee, N. N. Iwakoshi, E. Ozdelen, G. Tuncman,

C. Gorgiin, L. H. Glimcher, G. S. Hotamisligil, Endoplasmic Reticulum Stress Links Obesity, Insulin Action, and Type 2 Diabetes. Science. 306, 457-461 (2004).

15. T. Izumi, H. Yokota-Hashimoto, S. Zhao, J. Wang, P. A. Halban, T. Takeuchi, Dominant negative pathogenesis by mutant proinsulin in the Akita diabetic mouse. Diabetes. 52, 409-16 (2003).

16. M. J. L. Kracht, M. van Lummel, T. Nikolic, A. M. Joosten, S. Laban, A. R. van der Slik, P. A. van Veelen, F. Carlotti, E. J. P. de Koning, R. C. Hoeben, A. Zaldumbide, B. O. Roep, Autoimmunity against a defective ribosomal insulin gene product in type 1 diabetes. Nat. Med. 23, 501-507 (2017).

17. M. L. Marre, J. W. McGinty, I.-T. Chow, M. E. DeNicola, N. W. Beck, S. C. Kent, A. C. Powers, R. Bottino, D. M. Harlan, C. J. Greenbaum, W. W. Kwok, J. D. Piganelli, E. A. James, Modifying Enzymes Are Elicited by ER Stress, Generating Epitopes That Are Selectively Recognized by CD4 ⁺ T Cells in Patients With Type 1 Diabetes. Diabetes. 67, 1356-1368 (2018).

18. J. Sidney, J. L. Vela, D. Friedrich, R. Kolia, M. von Herrath, J. D. Wesley, A. Sette, Low HLA binding of diabetes-associated CD8+ T-cell epitopes is increased by post translational modifications. BMC Immunol. 19, 12 (2018).

19. J. S. Cox, C. E. Shamu, P. Walter, Transcriptional induction of genes encoding endoplasmic reticulum resident proteins requires a transmembrane protein kinase. Cell. 73, 1197-206 (1993).

20. H. P. Harding, Y. Zhang, D. Ron, Protein translation and folding are coupled by an endoplasmic-reticulum-resident kinase. Nature. 397, 271-274 (1999).

21. H. Yoshida, K. Haze, H. Yanagi, T. Yura, K. Mori, Identification of the cis -Acting Endoplasmic Reticulum Stress Response Element Responsible for Transcriptional Induction of Mammalian Glucose-regulated Proteins. J. Biol. Chem. 273, 33741-33749 (1998).

22. S. Oyadomari, A. Koizumi, K. Takeda, T. Gotoh, S. Akira, E. Araki, M. Mori, Targeted disruption of the Chop gene delays endoplasmic reticulum stress-mediated diabetes. J. Clin. Invest. 109, 525-532 (2002).

23. A. H. Minn, C. Hafele, A. Shalev, Thioredoxin-Interacting Protein Is Stimulated by Glucose through a Carbohydrate Response Element and Induces b-Cell Apoptosis. Endocrinology. 146, 2397-2405 (2005).

24. C. M. Oslowski, T. Hara, B. O’Sullivan-Murphy, K. Kanekura, S. Lu, M. Hara, S. Ishigaki, L. J. Zhu, E. Hayashi, S. T. Hui, D. Greiner, R. J. Kaufman, R. Bortell, F. Urano, Thioredoxin-Interacting Protein Mediates ER Stress-Induced b Cell Death through Initiation of the Inflammasome. CellMetab. 16, 265-273 (2012).

25. T. W. Kensler, N. Wakabayashi, S. Biswal, Cell survival responses to environmental stresses via the Keapl-Nrf2-ARE pathway. Annu. Rev. Pharmacol. Toxicol. 47, 89-116 (2007).

26. G. R. Moran, The catalytic function of renalase: A decade of phantoms. Biochim. Biophys. Acta - Proteins Proteomics . 1864, 177-186 (2016). 27. B. A. Beaupre, M. R. Hoag, J. Roman, F. H. Forsterling, G. R. Moran, Metabolic Function for Human Renalase: Oxidation of Isomeric Forms of b-NAϋ(R)H that Are Inhibitory to Primary Metabolism. Biochemistry. 54, 795-806 (2015).

28. M. Milani, F. Ciriello, S. Baroni, V. Pandini, G. Canevari, M. Bolognesi, A. Aliverti, FAD-Binding Site and NADP Reactivity in Human Renalase: A New Enzyme Involved in Blood Pressure Regulation. J. Mol. Biol. 411, 463-473 (2011).

29. J. D. Taylor, A. A. Wykes, Y. C. Gladish, W. B. Martin, New Inhibitor of Monoamine Oxidase. Nature. 187, 941-942 (1960).

30. Sanjana, N. E., Shalem, O. & Zhang, F. Improved vectors and genome-wide libraries for CRISPR screening. Nat Methods 11, 783-784, (2014).

31. Joung, J. et al. Genome-scale CRISPR-Cas9 knockout and transcriptional activation screening. Nat Protoc 12, 828-863, (2017).

32. Schuster, C., Jonas, F., Zhao, F. & Kissler, S. Peripherally induced regulatory T cells contribute to the control of autoimmune diabetes in the NOD mouse model. Eur J Immunol 48, 1211-1216, (2018).

33. Shalem, O. et al. Genome-scale CRISPR-Cas9 knockout screening in human cells. Science 343, 84-87, (2014).

34. Cao, J. et al. An easy and efficient inducible CRISPR/Cas9 platform with improved specificity for multiple gene targeting. Nucleic Acids Res 44, el49, (2016).

35. Kuznetsova, A. et al. Trimeprazine increases IRS2 in human islets and promotes pancreatic beta cell growth and function in mice. JCI Insight 1, pii: e80749 (2016).

36. M. Milani, F. Ciriello, S. Baroni, V. Pandini, G. Canevari, M. Bolognesi, A. Aliverti. FAD-binding site and NADP reactivity in human renalase: a new enzyme involved in blood pressure regulation. J. Mol. Biol. 411, 463-473 (2011).

37. C. Binda, F. Hubalek, M. Li, Y. Herzig, J. Sterling, D. E. Edmonson, A. Mattevi. Binding of rasagiline-related inhibitors to human monoamine oxidases: a kinetic and crystallographic analysis. J. Med. Chem. 48, 8148-8154 (2005).

38. G. M. Sastry, M. Adzhigirey, T. Day, R. Annabhimoju, W. Sherman. Protein and ligand preparation: Parameters, protocols, and influence on virtual screening enrichments. J. Comput. Aided Mol. Des., 27, 221-234 (2013).

39. K. J. Bowers, D. E. Chow, H. Xu, R. O. Dror, M. P. Eastwood, B. A. Gregersen, J. L. Kelpeis, I. Kolossvary, M. A. Moraes, F. D. Sacerdoti, J. K. Salmon, Y. Shan, D. E. Shaw. Scalable Algorithms for Molecular Dynamics Simulations on Commodity Clusters. Proceedings of the 2006 ACM/IEEE Conference on Supercomputing (SC06), p43 (2006). EQUIVALENTS

[00156] The foregoing written specification is considered to be sufficient to enable one skilled in the art to practice the embodiments. The foregoing description and Examples detail certain embodiments and describes the best mode contemplated by the inventors. It will be appreciated, however, that no mater how detailed the foregoing may appear in text, the embodiment may be practiced in many ways and should be construed in accordance with the appended claims and any equivalents thereof.

[00157] As used herein, the term about refers to a numeric value, including, for example, whole numbers, fractions, and percentages, whether or not explicitly indicated. The term about generally refers to a range of numerical values (e.g., +/-5-10% of the recited range) that one of ordinary skill in the art would consider equivalent to the recited value (e.g., having the same function or result). When terms such as at least and about precede a list of numerical values or ranges, the terms modify all of the values or ranges provided in the list.

In some instances, the term about may include numerical values that are rounded to the nearest significant figure.

Claims

What is Claimed is:

1. A method of lowering blood glucose, increasing insulin secretion in response to glucose, and preventing or slowing the death of pancreatic beta cells or beta-like cells in a subject comprising administering a monoamine oxidase inhibitor (MAOI), wherein the MAOI: a. binds renalase; and/or b. binds flavin adenine dinucleotide (FAD); and/or c. produces an active agent that binds renalase or FAD.

2. The method of claim 1, wherein the subject has autoimmune diabetes, optionally wherein the autoimmune diabetes is induced by an immunotherapy.

3. The method of claim 1 or claim 2, wherein the subject has type 1 diabetes.

4. The method of claim 2, wherein the autoimmune diabetes is induced by an immunotherapy, optionally wherein the immunotherapy is an immune checkpoint modulator.

5. The method of claim 4, wherein the immune checkpoint modulator is an inhibitor of PD-1, PD-L1, or CTLA-4, optionally wherein the inhibitor is an antibody.

6. The method of any one of the preceding claims, wherein the MAOI is administered in combination with an additional treatment.

7. The method of claim 6, wherein the additional treatment is insulin, optionally wherein the insulin is a rapid-acting, intermediate-acting, or long-acting insulin.

8. The method of claim 6, wherein the additional treatment is a glucagon-like peptide analog or agonist, dipeptidyl peptidase-4 inhibitor, amylin analog, biguanide, thiazolidinedione, sulfonylurea, meglitinide, alpha-glucosidase inhibitor, or sodium/glucose transporter 2 inhibitor.

9. The method of any one of the preceding claims, wherein the subject has a blood sugar level higher than 11.1 mmol/liter or 200 mg/dl.

10. The method of claim 1, wherein the beta or beta-like cells are native to the subject.

11. The method of claim 10, wherein the beta cells or beta-like cells have been removed from the subject, manipulated ex-vivo, and re-implanted into the subject.

12. The method of claim 1, wherein the beta or beta-like cells are non-native to the subject.

13. The method of claim 11 or 12, wherein the beta cells or beta-like cells are transplanted into a subject with autoimmune diabetes.

14. The method of claim 13, wherein the beta cells or beta-like cells are administered by transplant into the pancreas, liver, or fat pads via surgery, injection, or infusion.

15. A method of preventing the development of type 1 diabetes comprising: a. screening a subject for risk factors for type 1 diabetes; b. determining if the subject has increased risk of developing type 1 diabetes; and c. administering a monoamine oxidase inhibitor (MAOI) if the subject has an increased risk of type 1 diabetes, wherein the MAOI: i. binds renalase; and/or ii. binds flavin adenine dinucleotide (FAD); and/or iii. produces an active agent that binds renalase or FAD.

16. The method of claim 15, wherein screening a subject for risk factors comprises obtaining data on a genetic risk score that is based on the known type 1 diabetes-associated gene variants, a family history of type 1 diabetes, the presence of one or more autoantibodies against beta cell antigens that are known to predict disease risk, and/or abnormal glucose tolerance.

17. The method of any one of the preceding claims, wherein the subject is a mammal.

18. The method of claim 17, wherein the mammal is a human.

19. The method of any of the preceding claims, wherein the monoamine oxidase inhibitor is a propargylamine, hydrazine, propylamine, or oxazolidinone derivative.

20. The method of any one of the preceding claims, wherein the monoamine oxidase inhibitor is clorgyline, pargyline, rasagiline, selegiline, ladostigil, ASS234, isocarboxazid, toloxatone, or tranylcypromine.