WO2023280175A1

WO2023280175A1 - Methods for treating complex i deficiencies or cancers by modulating gro3p biosynthesis

Info

Publication number: WO2023280175A1
Application number: PCT/CN2022/103982
Authority: WO
Inventors: Hui Jiang; Shanshan Liu; Qinghua Liu
Original assignee: National Institute Of Biological Sciences, Beijing
Priority date: 2021-07-06
Filing date: 2022-07-05
Publication date: 2023-01-12
Also published as: WO2023280175A9

Abstract

A method for modulating Gro3P (glycerol-3-phosphate) biosynthesis, and uses thereof, e.g. therapeutic uses thereof in treating mitochondrial complex I diseases or cancers.

Description

METHODS FOR TREATING COMPLEX I DEFICIENCIES OR CANCERS BY MODULATING GRO3P BIOSYNTHESIS

This application claims the benefit of PCT Application No. PCT/CN2021/104699, filed on July 06, 2021, which is hereby incorporated by reference in its entirety

TECHNICAL FIELD

The present application is directed to methods for treating mitochondrial complex I diseases or cancers by modulating glycerol-3-phosphate (Gro3P) biosynthesis. Also provided are agents and pharmaceutical compositions for use in modulating Gro3P biosynthesis.

BACKGROUND

Mitochondrial respiration oxidizes NADH and transfers electrons to oxygen. ETC (electron transport chain) dysfunction or hypoxia results in toxic NADH reductive stress. Researches in the past decades reveal that NADH reductive stress underlies proliferation defects of ETC-defective human cells. However, how cells manage NADH reductive stress remains elusive.

NADH reductive stress also regulates tumorigenesis under hypoxia. Hypoxia deprives the electron acceptor oxygen to inhibit the ETC and to cause NADH reductive stress. Hypoxic NADH reductive stress restricts the synthesis of aspartate and serine to limit tumor growth (Diehl et al., 2019; Garcia-Bermudez et al., 2018; Sullivan et al., 2018) , and also promotes the synthesis of L-2-hydroxyglutarate from α-ketoglutarate to regulate histone methylation levels and to inhibit glycolysis and electron transport (Intlekofer et al., 2015; Oldham et al., 2015) .

Mitochondrial disease is a group of genetic disorders characterized by defective oxidative phosphorylation, a coupled process of electron transport and ATP synthesis (Gorman et al., 2016; Russell et al., 2020; Thompson et al., 2020) . Given the essential role of ETC in NAD ⁺ regeneration, mitochondrial diseases, especially the severe pediatric forms, are often characterized by metabolic derangements indicative of NADH reductive stress, such as lactic acidosis (Katsyuba et al., 2020; Russell et al., 2020; Thompson Legault et al., 2015) . Transgenic expression of NDI1, the yeast NADH: ubiquinone oxidoreductase, regenerates NAD ⁺ to extend the lifespan of brain NDUFS4-knockout mice (McElroy et al., 2020) , a model of mitochondrial complex I disorder (Kruse et al., 2008; Quintana et al., 2010) . Despite the importance of NADH/NAD ⁺ redox homeostasis in mitochondrial disease and tumorigenesis, our understanding of endogenous NAD ⁺ regeneration pathways remains limited. Identification of therapeutic targets involved in NAD ⁺ regeneration would be highly desirable. Development of agents or methods for modulating such therapeutic targets would also be needed for treating pathological conditions related to ETC dysfunctions.

SUMMARY OF INVENTION

The present inventors revealed Gro3P biosynthesis as an NAD ⁺ regeneration pathway evolutionarily conserved in yeast, C. elegans, mouse, and human. It has been established that enhancing Gro3P synthesis could rescue mitochondrial complex I deficiency in cultured cells. Also, the inventors revealed lacking Gro3P biosynthesis as a metabolic characteristic that sensitizes neurons to complex I deficiency and demonstrated enhancing Gro3P biosynthesis alleviates neuroinflammation and extends lifespan in the Ndufs4 ^-/- mice (an animal model of complex I deficiency) .

Accordingly, in a first aspect, the present disclosure relates to a method for treating complex I deficiency in a subject in need thereof, comprising administering to the subject a therapeutically effective amount of an agent enhancing Gro3P biosynthesis.

In one embodiment, the agent enhancing Gro3P biosynthesis increases the enzymatic activity, e.g. specific activity of one or more of cytosolic Gro3P dehydrogenases (cGPDs) . In one embodiment, the agent is an agonist of one or more of cGPDs.

In one embodiment, the agent enhancing Gro3P biosynthesis increases the level, e.g. mRNA level or protein level, of one or more cGPDs.

In one embodiment, the one or more cGPDs is human GPD1 and/or human GPD1L.

In one embodiment, the agent enhancing Gro3P biosynthesis comprises a nucleic acid molecule comprising a nucleotide sequence encoding the one or more cGPDs, encoding a variant of cGPD having the same or enhanced enzymatic activity or a nucleotide sequence enhancing the expression of one or more cGPDs.

In one embodiment, the agent enhancing Gro3P biosynthesis is a vector comprising the nucleic acid molecule of the above paragraph. In a specific embodiment, the vector further comprises one or more regulatory sequences selected from a promoter, an enhancer, a polyadenylation sequence, and origin of replication. In one embodiment, the promoter is operatively linked to the nucleic acid of the above paragraph. In one embodiment, the promoter is a constitutive promoter. In one embodiment, the promoter is a tissue-specific promoter, e.g. brain-specific promoter or ocular tissue-specific promoter, or cell-type-specific promoter e.g. neuron-specific promoter.

In a specific embodiment, the vector is a viral vector, such as an adenoviral vector, a retroviral vector or a recombinant adeno-associated virus (rAAV) vector. In one embodiment, the rAAV vector comprises the nucleotide sequence encoding one or more cGPDs, encoding a variant of cGPD having the same or enhanced enzymatic activity or a nucleotide sequence enhancing the expression of one or more cGPDs, as well as at least one inverted terminal repeat. In one embodiment, the rAAV vector comprises two ITRs. In a preferred embodiment, the AAV ITR is AAV1, AAV2, AAV3, AAV4, AAV5, AAV6, AAV7, AAV8, AAV9, AAVrh10, AAV. 7m8, AAV2tYF, AAV11, or AAV12 serotype ITR, preferably AAV2, AAV4, AAV5, AAV6, AAV8, AAV9, AAVrh10, AAV. 7m8, or AAV2tYF serotype ITR, more preferably AAV2 ITR. In a specific embodiment, the AAV vector comprises a nucleotide sequence as shown in SEQ ID NO: 31, or a nucleotide sequence having at least 80%, 85%, 90%, 95%, 98%or 99%sequence identity to SEQ ID NO: 31.

In one embodiment, the AAV vector is encapsidated in an AAV particle. In one embodiment, the AAV particle comprises capsid proteins. In a preferred embodiment, at least one, preferably two or three of the capsid proteins (VP1, VP2 or VP3) is of a serotype selected from AAV1, AAV2, AAV3, AAV4, AAV5, AAV6, AAV7, AAV8, AAV9, AAVrh10, AAV. 7m8, AAV2tYF, AAV11, or AAV12 serotype. In a preferred embodiment, the serotype of capsid protein is suitable for brain delivery, e.g. intraparenchymal, intrathecal, intracerebroventricular, subpial and intravenous administration delivery, or eye delivery, e.g. subretinal, intravitreal or intraocular delivery. In a specific embodiment, the AAV vector comprises a capsid of AAV2, AAV4, AAV5, AAV8, AAV9, or AAVrh. 10 serotype.

In one embodiment, the complex I deficiency is selected from a group consisting of Leigh syndrome, Leber’s optical hereditary neuropathy (LHON) (also known as Leber optic atrophy) , Leber optic atrophy and dystonia, mitochondrial complex I deficiency (OMIM entry 252010) , leukodystrophy, MELAS (mitochondrial encephalomyopathy, lactic acidosis, and stroke-like episodes) syndrome, optic atrophy with demyelinating disease of central nervous system (CNS) and other neurodegenerative disorders.

In one embodiment, the subject is mammal, e.g. rodent, or primate, preferably human.

In one embodiment, the subject has at least one mutation in mtDNA encoding a complex I subunit, e.g. ND1, ND2, ND3, ND4, ND5, ND6 or ND4L. For example, the subject has LHON and has at least one mutation selected from Table 2 or Table 3.

In one embodiment, the method of the first aspect has one or more of the following effects: increased C-I-independent OCR (complex I independent oxygen consumption rate) , decreased NADH/NAD ⁺ ratio, increased ATP level, decreased mitochondrial integrative stress response, reduced cell death, deceased neuroinflammation, decreased neurodegeneration, enhanced motor function, alleviated metabolic derangements, decreased glycolysis blockade, decreased αHB level, decreased lactate level, decreased alanine level, decreased β-hydroxybutyrate/acetoacetate (βHB/AcAc) ratio, decreased lactate/pyruvate ratio, increased aspartate level, and increased asparagine level in tissue or plasm.

It has also been found that human cancer cells exploit Gro3P biosynthesis to antagonize ETC inhibition and hypoxia in culture and in xenografts. In xenograft tumors, abolishing Gro3P synthesis significantly suppressed tumor growth, and synergistic effect in inhibiting tumor progression was observed when used with metformin, an inhibitor of complex I.

Accordingly, in second aspect, the present disclosure provides a method for treating tumor in a subject by inhibiting Gro3P biosynthesis of tumor cells. In a specific embodiment, the method comprises administering to the subject a therapeutically effective amount of an agent inhibiting Gro3P biosynthesis of tumor cells.

In one embodiment, the agent inhibiting Gro3P biosynthesis inhibits the activity of one or more cGPDs, e.g. human GPD1 and/or GPD1L.

In one embodiment, the agent inhibiting the activity of one or more cGPDs is a chemical compound, a polypeptide or a polynucleotide.

In a preferred embodiment, the method further comprises inhibiting complex I, e.g. by administering to the subject an inhibitor of complex I.

In specific embodiments, the inhibitor of complex I is selected from a group consisted of metformin, phenformin, BAY84-2243, CAI, ME-344, fenofibrate, and mIBG.

In one embodiment, the agent inhibiting Gro3P biosynthesis is delivered simultaneously with the inhibitor of complex I. In a preferred embodiment, the agent inhibiting Gro3P biosynthesis is delivered after the inhibitor of complex I.

In one embodiment, the inhibition of Gro3P biosynthesis and the inhibition of complex I achieves synergistic anti-tumor effect.

In third aspect, the present disclosure provides a pharmaceutical combination of an agent inhibiting Gro3P biosynthesis and an inhibitor of complex I.

In one embodiment, the agent inhibiting Gro3P biosynthesis is an agent inhibiting or eliminating the activity of one or more cGPDs. In one embodiment, the cGPD is human GPD1 and/or GPD1L.

In one embodiment, the inhibitor of complex I is selected from a group consisted of metformin, phenformin, BAY84-2243, CAI, ME-344, fenofibrate, and mIBG.

In one embodiment, the pharmaceutical combination is used for treating tumor.

In one embodiment, the pharmaceutical combination exhibits synergistic anti-tumor effect.

BRIEF DESCRIPTION OF DRAWINGS

To facilitate a better understanding of the features and advantages of the present invention, the following detailed description and accompanying drawings thereof are provided. However, one skilled in the art will understand that they are provided with a purpose of illustration rather than limitation. The scope of the invention is controlled by the appended claims.

FIG. 1 is a schematic of Gro3P metabolism and chemical inhibitors of the ETC complexes. Human ETC complexes are shown. Complex-II is absent for simplicity. MIM: mitochondrial inner membrane.

FIG. 2 shows the results of immunoblot analysis of WT and cGPD-knockout (KO) 143B, HeLa, SNB-19, and A549 cells. GPD1 is not expressed in 143B cells.

FIG. 3 shows the results of metabolite analysis of dihydroxyacetone phosphate (DHAP) and Gro3P of WT and cGPD-KO 143B cells treated with ETC inhibitors for 2 hours. Pier: piericidin, complex I inhibitor; Anti: antimycin, complex III inhibitor. Data are mean±SD from three biological replicates. Statistics: two-tailed unpaired Student’s t-test, *P<0.05, **P<0.01, ***P<0.001.

FIG. 4 shows the results of metabolite analysis of the NADH/NAD ⁺ ratio of WT and cGPD-KO 143B, HeLa, SNB-19, and A549 cells treated with ETC inhibitors for 2 hours. Data are mean±SD from three biological replicates. Statistics: two-tailed unpaired Student’s t-test, *P<0.05, **P<0.01, ***P<0.001.

FIG. 5 shows the cell number and results of immunoblot analysis of WT and cGPD-KO 143B and HeLa cells complemented with GPD1L (1L) or vector (V) . ETC inhibitors were treated for 4-5 days. Met: metformin, complex I inhibitor. Data are mean±SD from three biological replicates. Statistics: two-tailed unpaired Student’s t-test, *P<0.05, **P<0.01, ***P<0.001.

FIG. 6 shows the metabolite analysis of Gro3P and DHAP of WT and cGPD-KO 143B cells following 0.5%oxygen treatment for 6 hours. Data are mean±SD from three biological replicates. Statistics: two-tailed unpaired Student’s t-test. *P<0.05, **P<0.01, ***P<0.001.

FIG. 7 shows cell number of WT and cGPD-KO 143B cells treated with 0.2%oxygen for 5 days in the presence or absence of 3 mM pyruvate supplementation. Data are mean±SD from three biological replicates. Statistics: two-tailed unpaired Student’s t-test. *P<0.05, **P<0.01, ***P<0.001.

FIGS. 8A-B show the result of metabolite analysis of Gro3P and DHAP (A) and the NADH/NAD ⁺ ratio (B) of WT and cGPD-KO HeLa cells following 0.5%oxygen treatment for 6 hours. Data are mean and mean±SD from three biological replicates. Statistics: two-tailed unpaired Student’s t-test; **P<0.01, ***P<0.001.

FIG. 9 shows the tumor growth curve of WT and cGPD-KO 143B-and HeLa-derived xenograft tumors daily treated with vehicle (water) or 1g/kg metformin. Data are mean±SEM. Statistics: one-way ANOVA with the Tukey-Kramer test, *P<0.05, **P<0.01, ***P<0.001.

FIG. 10 shows the tumor sizes at the end point of WT and cGPD-KO 143B-and HeLa-derived xenograft tumors daily treated with vehicle (water) or 1g/kg metformin. Data are mean±SEM. Statistics: one-way ANOVA with the Tukey-Kramer test, *P<0.05, **P<0.01, ***P<0.001.

FIG. 11 shows the photo images of WT and cGPD-KO HeLa-derived tumors daily treated with vehicle (water) or 1g/kg metformin.

FIG. 12A-B show tumor growth curves (A) and tumor sizes at the end point (B) of the xenograft tumors derived from WT and cGPD-KO143B cells complemented with GPD1L (1L) or vector (V) which were daily treated with vehicle (water) or 1g/kg metformin. Data are mean±SEM. Statistics: one-way ANOVA with the Tukey-Kramer test, *P<0.05, **P<0.01, ***P<0.001.

FIG. 13 shows the results of immunoblot analysis of 143B cells overexpressing WT cGPDs, enzymatically-inactive cGPDs (GPD1L ^G14A, GPD1L ^K206A and GPD1 ^G12A, GPD1 ^K204A) , or the vector control.

FIG. 14 shows the metabolite analysis of Gro3P and DHAP of the 143B cells as in FIG. 13. Cells were treated with ETC inhibitors for 2 hours. Data are mean±SD from three biological replicates. Statistics: two-tailed unpaired Student’s t-test. ***P<0.001. ns: not significant.

FIG. 15 shows the complex I-dependent and complex I-independent oxygen consumption rate (OCRs) of the 143B cells as in FIG. 13. Data are mean±SD from six biological replicates. Statistics: two-tailed unpaired Student’s t-test. ***P<0.001. ns: not significant.

FIG. 16 shows the NADH/NAD ⁺ ratio of WT and GPD2-KO 143B cells overexpressing vector (V) or GPD1L (1L) . Cells were treated with ETC inhibitors for 2 hours. Data are mean±SD from three biological replicates. Statistics: two-tailed unpaired Student’s t-test. ***P<0.001. ns: not significant.

FIGS. 17A-D show that enhancing Gro3P synthesis promotes Gro3P shuttle activity to alleviate metabolic and proliferative defects and to repress mitochondrial integrative stress response in Complex-I impaired Cells. A, Left: heat map representing metabolite levels of WT and GPD2-KO 143B cells overexpressing vector or GPD1L. Cells were treated with DMSO or ETC inhibitors for 2 hours. For each metabolite, the mean log2-transformed fold change is relative to DMSO-treated cells (n = 3) . Arrows indicate major metabolic changes induced by GPD1L-overexpression under piericidin treatment. Right: schematic of GPD1L-overexpression (OE) effects on cells under complex I inhibition. Arrows indicate the direction of glucose carbon flux. B-D, ATP level after ETC inhibition for 2 hours (B) , metabolite analysis of aspartate and asparagine after ETC inhibition for 2 hours (C) , and qPCR analysis of representative ATF4 target genes upregulated by mitochondrial integrative stress response (Quiros et al., 2017) after ETC inhibition for 16 hours (D) in WT and GPD2-KO 143B cells overexpressing vector (V) or GPD1L (1L) .

FIG. 18 shows cell number of WT and GPD2-KO 143B cells treated with ETC inhibitors (Pier: 300 nM; Met: 2.5 mM; Anti: 100 nM) for 4-5 days. Representative images are shown. Data are mean±SD from three biological replicates. Statistics: two-tailed unpaired Student’s t-test. ***P<0.001. ns: not significant.

FIG. 19 shows the results of immunoblot analysis of WT and NDUFS2-KO 143B cells overexpressing WT or enzymatically-inactive (GPD1L ^K206A and GPD1 ^K204A) cGPDs or the vector control.

FIG. 20 shows the cell number of the cells as in FIG. 19 after growing for 5 days. Data are mean±SD from three biological replicates. Statistics: two-tailed unpaired Student’s t-test. ***P<0.001. ns: not significant.

FIG. 21 shows the schematic of the beneficial effects of NAD ⁺ regeneration by the Gro3P shuttle under complex I dysfunction.

FIGS. 22A-D (A) Cell number of 143B cells overexpressing WT cGPDs, enzymatically-inactive cGPDs (GPD1 ^LG14A, GPD1L ^K206A and GPD1 ^G12A, GPD1 ^K204A) , or the vector control. Cells were treated with ETC inhibitors (Pier: 300 nM; Anti: 100 nM) for 5 days. (B) Immunoblot analysis and cell number of HeLa, U2OS and MEF cells overexpressing GPD1L or the vector (V) control. Cells were treated with ETC inhibitors (Pier: 300 nM; Anti: 100 nM) for 5 days. (C) The NADH/NAD ⁺ ratio of 143B cells overexpressing GPD1L or vector. Cells were treated individually or in combination with Piericidin (300 nM) and oxamate (lactate dehydrogenase inhibitor, 40 mM) for 25 minutes. (D) Cell survival of 143B cells treated as in (C) for 10 hours. Representative images are shown. Cell viability was measured by trypan blue staining. Live cell number relative to total cell number in DMSO-treated cells were normalized as 100%. Data are mean±SD from three biological replicates. Statistics: two-tailed unpaired Student’s t-test, ***P<0.001.

FIG. 23 shows the results of immunoblot analysis of Gro3P metabolic enzymes in adult C57BL/6J mouse organs. 6 μg proteins of each sample were loaded for immunoblots.

FIG. 24 shows the immunoblot analysis of cultured cortical neurons infected with AAV-PHP. eB encoding GFP or GPD1 for 10 days.

FIGS. 25A-B show the NADH/NAD ⁺ ratio (A) , and ATP level (B) of neurons as in FIG. 24 treated with ETC inhibitors for 2 hours. Ratio (A) and ATP level (B) in DMSO-treated non-infected neurons were normalized as 1.

FIG. 26 shows the measurement of complex I-dependent and complex I-independent OCRs of neurons as in FIG. 24. DNP (2, 4-dinitrophenol, 200 μM) is an uncoupler to stimulate maximum OCR.

FIG. 27 Left: Schematic of delivering AAV-PHP. eB expressing GFP or GPD1 to the Ndufs4 ^-/- mice via the retro-orbital injection to increase GPD1 level in the brain; Right: Immunoblot analysis of GPD1 in mouse organs one month after AAV delivery.

FIG. 28 shows the immunoblot analysis of Gro3P metabolic enzymes in brain regions collected 3 weeks after AAV transduction to express GPD1 or GPD1 ^K204A. Samples from two mouse brains for each treatment are shown.

FIG. 29 shows the body weight of the indicated mice. KO + X: Ndufs4 ^-/- mice injected with AAV-PHP. eB expressing gene X. Data are mean ± SD.

FIG. 30 shows the lifespan of the indicated mice. KO + X: Ndufs4 ^-/- mice injected with AAV-PHP. eB expressing gene X. Statistics: one-way ANOVA with the tukey-Kramer test. **P < 0.01, ***P < 0.001.

FIGS. 31A-B show the metabolite analysis of brainstem glycolysis intermediates (A) and brainstem and plasma metabolites sensitive to the NADH/NAD ⁺ ratio (B) of the WT, Ndufs4 ^-/-, and Ndufs4 ^-/- mice transduced with AAV expressing GFP or GPD1. KO: Ndufs4 ^-/-.

FIG. 32 shows the representative images of Iba1 immunofluorescence staining at postnatal day 47-50 (P47-50) of the indicated mice. OB: olfactory bulb; CB: cerebellum; IO: inferior olive; VN: vestibular nuclei. Scale bar: 1 mm.

FIG. 33 shows the quantification results of Iba1 fluorescence intensity in different brain areas of the indicated mice.

FIG. 34 shows the representative 30-min locomotor activity traces at P50 (WT: n=6; Ndufs4 ^-/-: n=5; Ndufs4 ^-/- + GFP: n=6; Ndufs4 ^-/- + GPD1: n=7) of the indicated mice.

FIG. 35 shows the body temperature of the indicated mice. Data are mean ± SEM. Statistics: two-way ANOVA with the Tukey-Kramer test; **P < 0.01, ***P < 0.001.

DETAILED DESCRIPTION

All publications, patents, and patent applications mentioned in this specification are herein incorporated by reference to the same extent as if each individual publication, patent, or patent application was specifically and individually indicated to be incorporated by reference.

Unless specifically defined elsewhere in the present disclosure, all other technical and scientific terms used herein have the meaning commonly understood by one of ordinary skill in the art to which this invention belongs.

As used herein, including in the appended claims, the singular forms of words such as “a” , “an” , and “the” , include their corresponding plural references unless the context clearly dictates otherwise.

The term “or” is used to mean, and is used interchangeably with, the term “and/or” unless the context clearly dictates otherwise.

In the context of the present disclosure, unless being otherwise indicated, the wording "comprise" , and variations thereof such as "comprises" and "comprising" will be understood to imply the inclusion of a stated element, e.g. a component, a property, a step or a group thereof, but not the exclusion of any other elements, e.g. components, properties and steps. When used herein the term "comprise" or any variation thereof can be substituted with the term "contain" , “include” or sometimes "have" or equivalent variation thereof. In certain embodiments, the wording “comprise” also includes the scenario of “consisting of” .

The term "treat" , “treating” or “treatment” includes cure or at least alleviate the symptoms.

The term “therapeutically effective amount” as used herein, refers to the amount of an agent that, when administered to a subject for treating a disease, or at least one of the clinical symptoms of a disease or disorder, is sufficient to effect such treatment for the disease, disorder, or symptom. The “therapeutically effective amount” can vary with the agent, the disease, disorder, and/or symptoms of the disease or disorder, severity of the disease, disorder, and/or symptoms of the disease or disorder, the age of the subject to be treated, and/or the weight of the subject to be treated. In the case of combination therapy, the “therapeutically effective amount” refers to the total amount of the combination objects for the effective treatment of a disease, a disorder or a condition.

The term “pharmaceutical composition” refers to a composition suitable for delivering to a subject. The pharmaceutical composition of the present disclosure comprises an agent modulating Gro3P biosynthesis and a pharmaceutically acceptable excipient. For example, in some embodiments of treating complex I deficiency, the pharmaceutical composition comprises a nucleic acid or a vector encoding cGPD. In some embodiments of treating cancer, the pharmaceutical composition comprises an inhibitor of Gro3P biosynthesis. Conventional pharmaceutically acceptable excipients are known in the art.

In the present disclosure, by “subject” it refers to a eukaryote, such as yeast, C. elegans, or an animal, preferably a mammal, e.g., a rodent, such as a mouse or a rat, a primate, preferably a higher primate, such as a human.

The terms “administration” , “administering” , “treating” and “treatment” as used herein, when applied to a subject, e.g. an animal, including human, or to cells, tissue, organ, or biological fluid, mean contact of an exogenous pharmaceutical, therapeutic, diagnostic agent, or composition to the subject, cell, tissue, organ, or biological fluid. Treatment of a cell encompasses contact of a reagent to the cell, as well as contact of a reagent to a fluid, where the fluid is in contact with the cell. The term “administration” and “treatment” also include in vitro and ex vivo treatments, e.g., of a cell, by a reagent, diagnostic, binding compound, or by another cell.

By "isolated nucleic acid" or “isolated polynucleotide” , it means a DNA or RNA which is removed from all or a portion of a polynucleotide in which the isolated polynucleotide is found in nature, or is linked to a polynucleotide to which it is not linked in nature. An isolated nucleic acid molecule "comprising" a specific nucleotide sequence may include, in addition to the specified sequence, operably linked regulatory sequences that control expression of the coding region of the recited nucleic acid sequences. Due to the codon degeneracy, one skilled in the art can understand that a specific amino acid sequence can be coded by different nucleotide sequences.

The term “promoter” is well known in the art and refers to a regulatory sequence directing the transcription of a gene. Promoters can be categorized into constitutive promoters and inducible promoters. A constitutive promoter is active at all circumstances in vivo. An inducible promoter is only active under certain condition, e.g. in the presence of stimulus, within a certain type of tissue (tissue-specific promoter) or cell (cell-type-specific promoter) , or at a certain developmental stage (developmental stage specific promoter) .

The term “operatively linked” refers to the association of two components in a single nucleic acid stretch so that one can affect the function of the other.

The term “transgene” refers to a nucleic acid to be transferred or introduced into a host, for example, a cell or organism. A transgene can be a coding sequence of a functional protein or fragment thereof, e.g. a coding sequence of GPD1 or GPD1L.

NADH/NAD ⁺ REDOX

The cellular NADH/NAD ⁺ redox balance is fundamental to metabolism (Hosios and Vander Heiden, 2018; Katsyuba et al., 2020; Xiao and Loscalzo, 2020) . The oxidized form NAD ⁺ functions as an electron acceptor in diverse metabolic pathways such as glycolysis, fatty acid oxidation, amino acid degradation, citric acid cycle, serine synthesis, and one carbon metabolism (Ducker and Rabinowitz, 2017; Hosios and Vander Heiden, 2018; Katsyuba et al., 2020; Lunt and Vander Heiden, 2011; Xiao and Loscalzo, 2020) . The continuous operation of these metabolic pathways requires NAD ⁺ regeneration by mitochondrial respiration. The ETC oxidizes NADH and transfers electrons to oxygen, the ultimate electron acceptor. The matrix arm of ADH: ubiquinone oxidoreductase or complex I is the entry point for electrons from matrix NADH into the ETC. Because mitochondrial inner membrane is impermeable to NADH, electrons from cytosolic NADH are shuttled to mitochondrial matrix through the malate-aspartate shuttle or directly transferred to the ETC through the Gro3P shuttle. ETC inhibition, resulting from either hypoxia or genetic/pharmacological disruption of ETC function, impairs NADH oxidation and elevates the mitochondrial and cytosolic NADH/NAD ⁺ ratios, a condition termed NADH reductive stress.

Gro3P biosynthesis and cGPDs

Gro3P biosynthesis is a side pathway of glycolysis. It oxidizes NADH by converting glycolysis intermediate dihydroxyacetone phosphate (DHAP) to Gro3P, a reversible reaction catalyzed by NAD ⁺-dependent cytosolic Gro3P dehydrogenases (cGPDs) (Ansell et al., 1997; Kelly et al., 2011) . Gro3P is subsequently metabolized in three ways: 1. Condensation with acyl-CoA to generate glycerolipids (Yu et al., 2018) ; 2. dephosphorylation to generate glycerol (Mugabo et al., 2016; Nevoigt and Stahl, 1997) ; 3. oxidization by FAD ²⁺-dependent mitochondrial Gro3P dehydrogenase (mGPD) to regenerate DHAP; this third metabolic fate for Gro3P feeds electrons to the ETC, forming the Gro3P shuttle (Mracek et al., 2013) (Fig. 1) . Notably, Gro3P biosynthesis supports yeast growth under anoxia (Ansell et al., 1997) . cGPDs are heterogeneously expressed in mouse tissues, with highest expression in brown adipocytes and lowest expression in cerebral cortex (Ratner et al., 1981) . A BALB/c subline of mice lacking cGPD activity exhibited elevated lactate/pyruvate ratio in skeletal muscle (MacDonald and Marshall, 2000) .

Gro3P synthesis was identified as a major NAD ⁺ regeneration pathway in yeast and human cells for the first time by the present inventors, which further explains the findings in previous studies as mentioned above.

The abbreviation “GPD” herein refers to glycerol 3-phosphate dehydrogenase (GPDH) . There are two isoforms of GPD, cytoplasmic GPD (cGPD) located in cytosol and mitochondrial GPD located in mitochondrial inner membrane. cGPD is encoded by two genes, GPD1 and GPD1L, which share around 72%sequence identity with each other. cGPDs catalyze a reversible redox reaction of dihydroxyacetone phosphate to glycerol 3-phosphate. mGPD is encoded by GPD2 and catalyzes an irreversible reaction of glycerol 3-phosphate to dihydroxyacetone phosphate.

In the context of the present disclosure, the phrase “one or more cGPDs” means one or more cytoplasmic GPD, e.g. GPD1 and/or GPD1L of human.

In one aspect of the present disclosure, complex I diseases are treated by increasing the level or activity of cGPD.

In another aspect of the present disclosure, tumors are treated by inhibiting cGPD. The present inventors have surprisingly discovered that inhibition of cGPD produces an enhanced inhibitory effect on tumor growth. Further, when cGPD inhibition is combined with inhibition of complex I, a synergistic inhibitory effect on tumor growth can be observed.

In some embodiment, the present disclosure provides a method of treating, alleviating or preventing tumor in a subject, comprising administering to the subject a pharmaceutically effective amount of an inhibitor of GPD1 and/or GPD1L.

Treatment of Complex I deficiencies

The term “complex I” refers to respiratory complex I, or NADH: ubiquinone oxidoreductase, which couples electron transfer from NADH to ubiquinone with transmembrane proton pumping. Being the largest respiratory complex, complex I is consisted of 44 subunits coded by both mitochondrial DNA (mtDNA) and nuclear DNA (nDNA) .

Genetic defects in the subunits of complex I may cause complex I deficiencies, which is also known as mitochondrial complex I deficiencies. In the context of the present disclosure, the term “complex I deficiency” and “complex I disease” is interchangeable and should be understood in a broad way to include any disorder caused by any defects of complex I, i.e. any mitochondrial complex I linked disease.

Complex I deficiency is the most common cause of pediatric mitochondrial disease (～30%) that typically presents neurological symptoms, and also causes multiple adult-onset diseases, such as Leber’s hereditary optic neuropathy (Fassone and Rahman, 2012) . Table 1 of a review by R. J. Rodenburg (Biochimica et Biophysica Acta 1857 (2016) 938–945) provides a list showing the genetic defects in the subunits of complex I and related mitochondrial diseases.

The present disclosure provides a method for treating complex I deficiency by enhancing Gro3P biosynthesis, specifically by delivering an agent increasing the enzymatic activity or level of one or more cGPDs, e.g. human GPD1 and/or GPD1L.

The increase of enzymatic activity can be indicated by increase of specific activity of cGPD. For example, the activity of cGPDs can be enhanced by an agonist of cGPD.

The increase of cGPD level can be achieved by multiple ways known to those skilled in the art, such as increasing gene copy number or improving protein expression.

In some embodiments, the complex I deficiency which can be treated by the present method has one or more diseases selected from a group consisting of Leigh syndrome, Leber’s optical hereditary neuropathy (LHON) (also known as Leber optic atrophy) , Leber optic atrophy and dystonia, mitochondrial complex I deficiency (OMIM entry 252010) , leukodystrophy, MELAS (mitochondrial encephalomyopathy, lactic acidosis, and stroke-like episodes) syndrome, optic atrophy with demyelinating disease of CNS and other neurodegenerative disorders. One skilled in the art would understand that additional complex I deficiencies may be discovered in the future, which can also be treated by the method of the present disclosure.

Complex I deficiency caused by nDNA can be treated by gene replacement therapy. However, mtDNA mutations cannot be readily remedied in the same way by delivering a transgene encoding complex I component to make compensation. Therefore, the method of the present disclosure is especially suitable for treating complex I deficiencies mainly caused by mtDNA mutations.

In some embodiments, the subject to be treated by the present method has genetic defects in mtDNA encoded complex I subunits, such as ND1, ND2, ND3, ND4, ND4L, ND5, and/or ND6. In some embodiments, the subject has one or more of Leigh syndrome, LHON, or MELAS.

Table 1 below provides an overview of genetic defects in mtDNA encoded complex I subunits and resulting phenotypes (R. J. Rodenburg, 2016, supra) .

Table 1. Overview of genetic defects in mtDNA encoded complex I subunits

It should be noted that mutations in mtDNA encoded complex I subunits MTND1, MTND2, MTND3, MTND4, MTND5, MTND6 can also cause Leigh Syndrome or Leigh- like Syndrome. Mutations MTND3 and MTND5 are the most frequent mtDNA causes of Leigh Syndrome.

For example, mutations leading to LHON are listed in Table 2 and Table 3 below. Data obtained from MITOMAP: www. mitomap. org.

Table 2: 19 primary LHON mutations, the first 3 mutations listed (in boldface) represent approximately 95%of all cases. The remaining mutations are listed in nucleotide order.

No.	Mutation	Subunit	NT Δ	AA Δ	%Patients
1	m. 11778G>A	ND4	G-A	R340H	69
2	m. 3460G>A	ND1	G-A	A52T		13
3	m. 14484T>C	ND6	T-C	M64V	14
4	m. 3376G>A	ND1	G-A	E24K	Rare
5	m. 3635G>A	ND1	G-A	S110N	Rare
6	m. 3697G>A	ND1	G-A	G131S	Rare
7	m. 3700G>A	ND1	G-A	A112T	Rare
8	m. 3733G>A	ND1	G-A	E143K	Rare
9	m. 4171C>A	ND1	C-A	L289M	Rare
10	m. 10197G>A	ND3	G-A	A47T	Rare
11	m. 10663T>C	ND4L	T-C	V65A	Rare
12	m. 13051G>A	ND5	G-A	G239S	Rare
13	m. 13094T>C	ND5	T-C	V253A	Rare
14	m. 14459G>A	ND6	G-A	A72V	Rare
15	m. 14482C>A	ND6	C-A	M64I	Rare
16	m. 14482C>G	ND6	C-G	M64I	Rare
17	m. 14495A>G	ND6	A-G	L60S	Rare
18	m. 14502T>C	ND6	T-C	I58V	Rare
19	m. 14568C>T	ND6	C-T	G36S	Rare

Table 3: Other candidate LHON mutations found as single family or singleton cases

Patient	Mutation	Subunit	NT Δ	AA Δ	#Patients
1	m. 3472T>C	ND1	T-C	F56L	1 case
2	m. 4025C>T	ND1	C-T	T240M	1 family; 3 cases
3	m. 4160T>C	ND1	T-C	L285P	1 family; 9 cases

4	m. 4640C>A	ND2	C-A	I57M		1 family; 4 cases
5	m. 5244G>A	ND2	G-A	G259S		1 case
6	m. 9101T>C	ATP6	T-C	I192T	1 case
7	m. 9804G>A	CO3	G-A	A200T	Multiple unrelated
8	m. 10237T>C	ND3	T-C	I60T	1 family; 2 cases
9	m. 11253T>C	ND4	T-C	I165T	1 case
10	m. 11696G>A	ND4	G-A	V312I		1 family; 11 cases
11	m. 14596A>T	ND6	A-T	I26M	1 family; 11 cases
12	m. 12811T>C	ND5	T-C	Y159H	1 family; 2 cases
13	m. 12848C>T	ND5	C-T	A171V	1 case
14	m. 13637A>G	ND5	A-G	Q434R		1 family; 3 cases
15	m. 13730G>A	ND5	G-A	G465E		1 case
16	m. 14279G>A	ND6	G-A	S132L		1 family; 2 cases
17	m. 14325T>C	ND6	T-C	N117D	1 case
18	m. 14498T>C	ND6	T-C	Y59C	1 case

Accordingly, for example, the subject to be treated by the present method has one or more mtDNA mutations selected from those listed in Table 2 or Table 3.

One skilled in the art would understand that other mutations in mtDNA encoded complex I subunit leading to LHON or other disorders may be revealed in the future. The present disclosure also contemplates application of the present method in a subject with any of these mutations.

The therapeutic effect in treating complex I deficiency achieved by the method of the present disclosure may be shown as or determined by one or more of the followings: an increased C-I-independent OCR (complex I independent oxygen consumption rate) , a decreased NADH/NAD ⁺ ratio, an increased ATP level, decreased mitochondrial integrative stress response, reduced cell death, deceased neuroinflammation, decreased neurodegeneration, enhanced motor function, alleviated metabolic derangements, including decreased glycolysis blockade, decreased αHB level, decreased lactate level, decreased alanine level, decreased β-hydroxybutyrate/acetoacetate (βHB/AcAc) ratio, decreased lactate/pyruvate ratio, increased aspartate level, and increased asparagine level in tissue or plasma. Measurement of therapeutic effects may vary depending on the phenotype of complex I disease and can be determined by one skilled in the art.

Enhancing Gro3P by Increasing cGPDs

In the method of the present disclosure, the complex I deficiency is treated by enhancing Gro3P biosynthesis, specifically by increasing the level of one or more cGPDs, namely GPD1 and GPD1L, or/and by increasing the enzymatic activity of cGPDs.

In some embodiments, the increase of cGPD level can be achieved by delivering a transgene encoding GPD1 and/or a transgene encoding GPD1L.

In some embodiments, the coding sequence of the cGPDs, specifically GPD1 or GPD1L, can be a wild-type nucleotide sequence or can be subjected to codon optimization so as to enhance the expression efficiency in a certain subject, e.g. human.

In some embodiments, the increase of enzymatic activity of cGPD can be achieved by enabling the expression of a mutated version of cGPD with enhanced enzymatic activity or by an agonist of one or more cGPDs.

In some embodiments, the transgene encoding GPD1 or the transgene encoding GPD1L is comprised in a viral vector. For example, the viral vector is a lentiviral vector, an adenoviral vector or an adeno-associated virus (AAV) vector.

In some embodiments, the lentiviral vector includes envelope proteins.

The rAAV vector of the present disclosure can comprise elements of any serotype known in the art or discovered in the future, or variant thereof. Preferably, the vector genome of the rAAV is AAV2, AAV4, AAV5, AAV6, AAV8, AAV9, AAVrh10, AAV. 7m8, AAV2tYF serotype. In one embodiment, the rAAV vector comprises one or more ITRs of AAV2, AAV4, AAV5, AAV6, AAV8, AAV9, AAVrh10, AAV. 7m8, AAV2tYF serotype.

In some embodiments, the AAV vector is self-complementary AAV (scAAV) or single stranded AAV (ssAAV) .

In some embodiments, the AAV vector comprises one or more ITRs from any AAV serotype, including but not limited to AAV1, AAV2, AAV3, AAV4, AAV5, AAV6, AAV7, AAV8, AAV9, AV10, AAVrh10, AAV. 7m8, AAV2tYF, AAV-DJ, AAV-DJ/8, AAV11, AAV12, or variant thereof. For example, the AAV vector comprises two AAV2 ITRs flanking the coding sequence of GPD1 or GPD1L.

In some embodiments, the transgene encoding GPD1 or GPD1L is under the control of a constitutive promoter. For example, the promoter can be EF1-alpha promoter, CAG promoter, PGK promoter or CMV promoter. In some other embodiments, the transgene encoding GPD1 or GPD1L is under the control of a tissue-specific or cell-type-specific promoter. For example, the promoter can be a neuron-specific promoter, e.g. synapsin promoter; a brain-specific promoter; or an ocular tissue-specific promoter, e.g. GRM6 promoter, PCP2 promoter, GNGT2 promoter, PDE6H promoter, PITX3 promoter, or NR2E1 promoter.

In some embodiments, the AAV vector comprises an adenylation signal sequence. For example, the adenylation signal sequence can be bGH polyA, SV40 polyA or WPRE-SV40 poly A, preferably bGH polyA.

In some embodiments, the vector genome is packaged by one or more capsid. The capsid proteins of the present disclosure can be capsid of any AAV serotype known in the art or discovered in the future. Specifically, the capsid protein can be selected from a group consisting of but not limited to AAV1, AAV2, AAV3, AAV4, AAV5, AAV6, AAV7, AAV8, AAV9, AAVrh10, AAV. 7m8, AAV2tYF, AAV11, or AAV12 serotype. The capsid can be derived from the same or different AAV serotype as the ITR. Preferably, the capsid proteins show tissue tropism for cells of the CNS (e.g. brain) or eye. Specifically, the AAV vector comprises a capsid from of AAV2, AAV4, AAV5, AAV8, AAV9, or AAVrh. 10.

In one embodiment, for the assembly of rAAV viral particle, a vector encoding the AAV replication and/or encapsidation protein and a vector encoding helper functions, such as p. Helper vectors, are used along with the vector comprising the nucleotide encoding target sequence flanked by ITR to infect host cells.

In some embodiment, the transgene encoding GPD1 or GPD1L is constructed into closed-ended DNA.

In some embodiment, the transgene encoding GPD1 or GPD1L is made into a circular RNA.

In some embodiments, the transgene encoding GPD1 or GPD1L is comprised in a lipid nano particle (LNP) .

Inhibition of tumor growth

The present disclosure provides a method for treating tumor or prevent tumor progression by inhibiting Gro3P biosynthesis of tumor cells. In a preferred embodiment, the method includes inhibition of both Gro3P biosynthesis and Complex I.

The terms “cancer” or “tumor” herein mean or describe the physiological condition in mammals that is typically characterized by unregulated cell growth. The cancer can be solid tumors, such as sarcomas, carcinomas or lymphomas, or cancers in the blood. Preferably, the cancer expresses one or more of cGPDs.

Since solid tumors often grow in hypoxic microenvironment (Hockel and Vaupel, 2001) , they may produce a more significant response to a treatment by inhibiting Gro3P biosynthesis alone as compared to leukemias. However, the method of the present disclosure is not limited to the use in treating solid tumor, especially when the inhibition of Gro3P biosynthesis is combined with the inhibition of Complex I. Inhibition of Complex I creates an NADH reductive stress in cancer microenvironment. Inhibition of Gro3P biosynthesis in the cells under such NADH reductive stress retards the cell growth and in turn slows down the progression of tumor.

When inhibition of Gro3P biosynthesis is used in combination with inhibition of Complex I, they can be conducted simultaneously or sequentially. In a preferred embodiment, the inhibition of Gro3P can be conducted after the inhibition of Complex I.

In one embodiment, the inhibition of Gro3P biosynthesis is achieved by delivering of an agent, e.g. a compound, a polypeptide, or an isolated nucleotide inhibiting the one or more of cGPDs, e.g. human GPD1 or GPD1L. In one embodiment, the tumor or cancer to be treated expresses one or more of cGPDs.

For example, the agent inhibiting one or more of cGPDs is a chemical compound, including but not limited to (-) -epicatechin, (-) -epicatechin-3-gallate, (-) -epigallocatechin, or (-) -epigallocatechin-3-gallate.

For example, the agent inhibiting one or more of cGPDs is a polypeptide, e.g. an antibody or a fragment thereof specifically binding to one or more of cGPDs.

For example, the agent inhibiting one or more of cGPDs is a nucleotide-based inhibitor, including but not limited to siRNA, shRNA, miRNA, gRNA, or antisense oligonucleotides.

In one embodiment, the inhibition of Gro3P biosynthesis is used in combination with inhibition of complex I to achieve a synergistic anti-tumor effect.

In one embodiment, the inhibition of complex I is achieved by administering to the subject a therapeutically effective amount of complex I inhibitor.

For example, a complex I inhibitor can be a compound known to have inhibitory effect on complex I. Preferably, the complex I inhibitor is a compound suitable for pharmaceutical use, e.g. a compound which has been proven to be safe in a clinical trial or pre-clinical studies. The complex I inhibitor can be selected from a group consisting of metformin, phenformin, BAY84-2243, CAI, ME-344, fenofibrate, and mIBG.

EXAMPLES

Example 1. Identification of Gro3P as NAD ⁺ regeneration pathway

This example describes the process of identifying Gro3P synthesis as candidate NAD ⁺ regeneration pathways under ETC dysfunction.

A total number of 175 genes encoding NAD ⁺-dependent and NAD (P) ⁺ dual-specific dehydrogenases/reductases in the human genome were retrieved from the KEGG database. Among the retrieved genes, 81 genes encode enzymes catalyzing 39 reactions of macronutrient metabolism (glucose, amino acids, fatty acids, nucleic acids, ketone bodies, alcohols and aldehydes) . Under ETC dysfunction, only pathways that are NAD ⁺-regenerating and have continuous substrate supply, as well as continuous product disposal, can substantially contribute to NAD ⁺ regeneration. The inventors thus focused on macronutrient metabolism because macronutrients are the most abundant substrates available in the cell, culture media, and circulating blood. By following metabolic flow starting from macronutrients, eight candidate NAD ⁺ regeneration pathways, including Gro3P synthesis (glucose metabolism) , four amino acid metabolic pathways, transhydrogenation reaction, ketone body metabolism and fatty acids desaturation were identified.

Next, because ETC deficiency and hypoxia are universal stresses for aerobic organisms, the inventors strongly weighted for evolutionary conservation to prioritize amongst candidate pathways. Yeast is an organism well known for its strong resistance to ETC-dysfunction and anoxia when cultured in fermentable medium (Lasserre et al., 2015) , suggesting the presence of effective NAD ⁺ regeneration pathways. Thus, yeast NAD ⁺-dependent and NAD (P) ⁺ dual-specific enzymes were retrieved to examine the conservation of the eight human candidate pathways. It turned out three of them are conserved: Gro3P synthesis, proline synthesis, and fatty acid desaturation. It was recently reported that fatty acid desaturation contributes to NAD ⁺ regeneration under complex I inhibition (Kim et al., 2019) , corroborating the fidelity of the inventors’ analysis. The two conserved and uncharacterized candidate pathways, proline synthesis and Gro3P synthesis, were further investigated.

Experimental results obtained with yeast and human cells do not support proline synthesis as a major NAD ⁺ regeneration pathway. The gene encoding for pyrroline-5-carboxylate reductase (Pro3 in yeast and PYCR1&2 in human) , an enzyme catalyzing the final step of proline synthesis, was deleted in yeast cells with disrupted ETC and in human osteoblastoma cell line 143B treated with ETC inhibitors, respectively, while no effects on the NADH/NAD ⁺ ratio or on the proliferation rates was observed, suggesting the proline synthesis is unlikely a major NAD ⁺ regeneration pathway in yeast and human cells. As a result, the study was continued with Gro3P synthesis as the candidate pathway.

Example 2. ETC Inhibition Promotes Gro3P Synthesis to Alleviate NADH Reductive Stress in Human Cancer Cells

To determine the contribution of Gro3P synthesis to NAD ⁺ regeneration, cGPDs (GPD1L and GPD1) were knocked out in four human cancer cell lines, 143B, Hela, SNB-19 and A549 (Fig. 2) . It was found that cGPD knockout abolished Gro3P synthesis (Fig. 3) , and significantly elevated the NADH/NAD ⁺ ratio under chemical ETC inhibition in all the four cell lines (Fig. 4) .

cGPD knockout sensitized 143B and HeLa cells to the anti-proliferative effects of ETC inhibitors (Fig. 5) . Reconstitution of GPD1L in cGPD-KO cells rescued sensitivity to complex III inhibition by antimycin but surprisingly caused strong resistance to complex I inhibition by piericidin or metformin (Fig. 5) . This is an unexpected result.

Thus, Gro3P synthesis provides an indispensable NAD ⁺ regeneration activity to maintain viability and proliferation of ETC-deficient cells.

Example 3. Gro3P Synthesis Antagonizes Hypoxia and Supports Growth of Xenograft Tumors

This example investigates the correlation between Gro3P synthesis and tumor growth.

Hypoxia is a common feature of most malignant tumors. It was found that hypoxia promoted Gro3P synthesis in 143B cells (Fig. 6) . cGPD-KO 143B cells proliferated significantly slower than WT cells under hypoxia treatment, which was rescued by pyruvate supplementation (Fig. 7) . Hypoxia also promoted Gro3P synthesis in HeLa cells (Fig. 8A) . cGPD-KO HeLa cells exhibited higher NADH/NAD ⁺ ratio as compared to WT cells under hypoxia (Fig. 8B) .

To further examine the function of Gro3P synthesis in tumorigenesis with a physiologically-relevant method, the tumor xenograft model was employed. It was found that cGPD knockout significantly suppressed the growth of 143B-and HeLa-derived xenograft tumors. Metformin treatment, which inhibits tumorigenesis through inhibiting complex I (Wheaton et al., 2014) , synergized with cGPD knockout to further suppress tumor growth (Figs. 9-11) . GPD1L reconstitution in cGPD-KO 143B-derived tumors rescued tumor growth and caused metformin resistance (Fig. 12) , similar to what was previously observed in cell culture (Fig. 5) . Therefore, Gro3P synthesis supports tumor growth in vivo and regulates sensitivity to metformin.

Example 4. Enhancing Gro3P Synthesis Promotes Gro3P Shuttle Activity to Rescue Complex I Dysfunction in Human Cancer Cell Lines

This example investigates whether overexpression of cGPDs rescues complex I dysfunction through the Gro3P shuttle. WT and enzymatically-inactive (Kelly et al., 2011) cGPDs were overexpressed in 143B cells (Fig. 13) . Overexpression of WT but not mutant cGPDs caused twenty-fold increase of Gro3P level (Fig. 14) and fourfold increase of the C-I-independent OCR (complex I independent oxygen consumption rate) (Fig. 15) . GPD2 knockout completely blocked the enhanced C-I-independent OCR (data not shown) . Therefore, enhancing Gro3P synthesis by overexpressing cGPDs can promote Gro3P shuttle activity.

In 143B cells, overexpression of GPD1L offset the elevated NADH/NAD ⁺ ratio under complex I but not complex III inhibition (Fig. 16) . GPD1L-overexpression in complex I inhibited cells relieved glycolysis blockage (Fig. 17A) , rescued the levels of ATP (Fig. 17B) , aspartate, and asparagine (Fig. 17C) , and repressed mitochondrial integrative stress responses (Fig. 17D) , which is also induced by NADH reductive stress (Mick et al., 2020) . All these rescue effects were blocked by GPD2 knockout and by complex III inhibition (Figs. 16 and 17A-D) .

Overexpression of cGPDs rescued the proliferation of 143B cells under complex I but not complex III inhibition; such rescue required the enzymatic activity of cGPDs (Fig. 22A) and was blocked by GPD2 knockout (Fig. 18) . Overexpression of GPD1L also rescued the proliferation of multiple cell lines under complex I inhibition (Figure 22B) . Moreover, beyond chemical inhibition, complex I was also genetically inactivated by deleting the NDUFS2 subunit in 143B cells. NDUFS2-KO 143B cells showed greatly impaired proliferation and were rescued by WT but not by enzymatically-inactive cGPDs (Fig. 19 and Fig. 20) . We further challenged 143B cells by simultaneously inhibiting complex I and lactate dehydrogenase, which led to an extremely high NADH/NAD ⁺ ratio (Fig. 22C) and complete cell death (Fig. 22D) . GPD1L overexpression protected cells from this metabolic crisis (Figs. 22C-D) , demonstrating the capacity of enhanced Gro3P shuttle to regenerate NAD ⁺. Taken together, these results establish proof-of-concept for enhancing Gro3P synthesis as a strategy to rescue complex I dysfunction through the Gro3P shuttle (Fig. 21) .

Example 5. Enhancing Gro3P Synthesis Compensates for Complex I Inhibition in Cultured Neurons

Mouse brain has very low protein levels of cGPDs (Fig. 23) . Examination of the Genotype-Tissue Expression (GTEx) database (Consortium, 2015) also supported the low transcript levels of cGPDs in human brain, especially in brain areas sensitive to complex I deficiency, such as basal ganglia (Fassone and Rahman, 2012) . A question was raised to ask whether low expression level of cGPDs contributes to neuronal vulnerability to ETC dysfunction. This example tries to address this question in cultured neurons.

Consistent with the mouse brain data, cGPDs are almost undetectable in cultured mouse cortical neurons (Fig. 24) . Upon ETC inhibition, these neurons had both a high NADH/NAD ⁺ ratio (> threefold increase) (Fig. 25A) and a low ATP level (< 30%of WT level) (Fig. 25B) , similar to what was previously observed in cGPD-KO cancer cells (Fig. 4) .

Overexpression of GPD1 in cultured neurons (Fig. 24) increased C-I-independent OCR (Fig. 26) , significantly decreased the NADH/NAD ⁺ ratio (from 3.8 to 2, Fig. 25A) , and significantly increased ATP level (from 20%to 60%, Fig. 25B) under chemical inhibition of complex I. GPD1-overexpression had no rescue effects on complex III-inhibited neurons (Figs. 25A-B) . Therefore, weak activity of Gro3P synthesis contributes to neuronal vulnerability to mitochondrial complex I dysfunction.

Example 6. Enhancing Gro3P Synthesis Suppresses Neuroinflammation, Alleviates Neurological Symptoms, and Extends the Lifespan of the Ndufs4 ^-/- Mice

This example determines whether enhancing cGPD expression is applicable to rescuing complex I deficiency in vivo.

The Ndufs4 ^-/- mice was used, which progressively develop fatal neuroinflammation and neurodegeneration, recapitulating the characteristics of human Leigh syndrome (Jain et al., 2016; Johnson et al., 2013; Kruse et al., 2008; Quintana et al., 2010) . The inventors employed the recently described AAV-PHP. eB vector (Chan et al., 2017) that can extensively transduce both neurons and glia to express target genes in the brain. AAV-mediated delivery of NDUFS4 to the Ndufs4 ^-/- mice rescued the lifespan of these mice (Fig. 30) , confirming the validity of our approach. On such basis, AAVs expressing GPD1, GFP, or the GPD1 ^K204A inactive-mutant were constructed and delivered to the Ndufs4 ^-/- mice by retro-orbital injection so that the AAV particles can diffuse in the whole mouse body. The enrichment of AAV virus in mouse brain was achieved by the brain high affinity capsid vector PHP. eB.

In mice delivered with AAV-GPD1 and mice delivered with AAV-GPD1 ^K204A, brain GPD1 was increased to a comparable level as hepatic GPD1 (Fig. 27 and Fig. 28) . GPD1, GFP, and GPD1 ^K204A did not rescue body weight loss of the Ndufs4 ^-/- mice (Fig. 29) . But strikingly, GPD1 not GFP or GPD1 ^K204A significantly extended the lifespan of the Ndufs4 ^-/- mice by 44%, from 58.3± 1.92 to 84.0± 3.65 days (Fig. 30) .

To characterize the rescue effect in detail, the brainstem metabolite levels were examined first. It was found that Ndufs4 ^-/- mice exhibited increased accumulation of glycolysis intermediates upstream of the NADH-producing step (F6P, G3P, and DHAP) (Fig. 31A) , indicating glycolysis blockade. The Ndufs4 ^-/- mice also exhibited increased αHB and lactate levels and decreased aspartate levels in the brainstem, as well as increased plasma αHB levels (Fig. 31B) , indicating elevated brainstem NADH/NAD ⁺ ratio. Overexpression of GPD1 but not GFP significantly alleviated these metabolic derangements (Figs. 31 A-B) .

To characterize the rescue effect in detail, brain pathology and motor function of age-matched mice were examined. Consistent with previous reports (Kruse et al., 2008; Quintana et al., 2010) , Ndufs4 ^-/- mice exhibited strong neuroinflammation manifested by the greatly-increased number and hypertrophic morphology of Iba1-positive microglia within multiple brain regions (Fig. 32 and Fig. 33) . Ndufs4 ^-/- mice progressively developed ataxia and locomotor defects (Fig. 34) and suffered from reduced body temperature (Fig. 35) . Expression of GPD1 but not GFP ameliorated most of the neuroinflammation and partially prevented the decline of motor function and the reduction in body temperature (Figs. 32-35) .

EXPERIMENTAL MODEL AND SUBJECT DETAILS

Cell Lines

143B, HeLa, U2OS, SNB-19, A549 and MEF cells were cultured in DMEM without pyruvate supplemented with 10%fetal bovine serum (FBS) , 1%penicillin-streptomycin and 2 mM L-glutamine. All cells were incubated at 37 ℃with 5%CO2.

Primary Neuron

Cortex from anesthetized P0 mouse was cut into small pieces and digested with 0.25%trypsin at 37 ℃ for 8 minutes and stopped by 3 volumes of DMEM supplemented with 10%FBS. Digested cortical pieces were dissociated by gentle pipetting for 5 times, rested for 3 minutes to precipitate undigested large pieces. Supernatants were transferred through a 40 μm cell strainer to remove cell aggregates. Dissociated neurons were cultured at 0.5-1×10 ⁶ cells per well (24 well plate) in dish or coverslips pre-coated with poly-D-lysine in DMEM supplemented with 10%FBS. Medium was switched to Neurobasal medium supplemented with 1 mM sodium pyruvate, 2%B27 and 1%GlutaMAX 12 hours after attachment. Culture was maintained by changing medium every 3 days.

Mice

Ndufs4 ^-/- mice were purchased from The Jackson Laboratory. BALB/c nude mice were purchased from Beijing Vital River Laboratory Animal Technology. Mice were maintained under a 12-hours light/dark cycle and on a standard chow diet at the specific pathogen-free (SPF) facility at the National Institute of Biological Sciences, Beijing. All mouse experiments were carried out following the national guidelines for housing and care of laboratory animals (Ministry of Health, China) and the protocol was in compliance with institutional regulations after review and approval by the Institutional Animal Care and Use Committee at National Institute of Biological Sciences, Beijing.

METHOD DETAILS

Lentivirus Production and Generation of Cells Stably Expressing cDNAs

Lentiviruses were produced by co-transfection of HEK293T with cDNA containing lentiviral vectors FUIPW, lentiviral packaging vectors psPAX2 and pMD2. G at a ratio of 5: 3: 2. Virus containing supernatants were collected and filtered through 0.45 μm filters 48 hours post-transfection and stored at -80 ℃. For infection, targeted cells were infected with virus-containing supernatants for 48 hours in the presence of 8 μg/ml polybrene. Infected cells were selected by 1-3 μg/ml puromycin. cDNA expression was validated by immunoblotting.

Generation of Knockout Cell Lines

Generation of knockout cell lines by the CRISPR–Cas9 technology was performed as described (Ran et al., 2013) . Two guide RNAs (gRNAs) were designed for targeting each gene. The oligonucleotides were listed below. Annealed gRNAs were cloned into the gRNA-expression plasmid PX458. PX458 was then transfected to targeted cells using the Polyethylenimine (PEI) transfection reagent. Two days after transfection, GFP-positive cells were sorted into single clones and were grown for 2 weeks or longer. The resultant colonies were trypsinized and expanded. Knockout clones were selected after immunoblotting examination.

Proliferation Assay of Human Cancer Cells

All the cell proliferation analyses were performed in growth medium: DMEM supplemented with 10%FBS, 1%P/Sand 50 μg/ml uridine with media changes every 48 hours. Uridine compensates for the pyrimidine synthesis function of mitochondrial Complex-III, which is blocked under Complex-III inhibition and is not related to the NADH/NAD ⁺ ratio (Bajzikova et al., 2019; Martinez-Reyes et al., 2020) . Cell lines growing in log phase were trypsinized, counted, and plated onto 6 well dishes in 2 ml growth medium with respiratory chain inhibitors at indicated concentrations. Initial seeding density was 25,000 cells per well. For all conditions the seeding densities used allowed exponential proliferation for 4-5 days and final cell counts were measured 4-5 days after treatment. Cells were counted using TC20 automated cell counter. Relative cell number was determined by cell number of inhibitor treatment divided by cell number of vehicle treatment.

Immunoblotting of Human Cancer Cells and Mouse Tissue Samples

Lysis buffer contains 20 mM Tris–HCl (pH 7.5) , 1 mM EDTA, 1%Triton X-100, 150 mM NaCl, 0.1%SDS, 2.5 mM sodium pyrophosphate, 1 mM β-glycerophosphate, a protease inhibitor cocktail and a phosphatase inhibitor. Human cells were scraped and pellets were collected. Tissues from anesthetized mice were dissected out and snap-frozen in liquid nitrogen. Tissues were weighted and cryohomogenized on dry ice.

Human cell pellets and homogenized frozen mouse tissue powders were extracted by sonication in the lysis buffer. Cell lysates were clarified by centrifuging at 15,000 rpm for 10 min at 4 ℃. Protein concentrations were determined by Bradford assay kit following the manufacturer instructions. Lysates were resolved by SDS–PAGE at 120 V. Proteins were transferred for 1.5 hours at 400 mA to PVDF membranes. Membranes were incubated with primary antibodies in 5%nonfat milk in PBST overnight. Membranes were washed three times in PBST before incubation in 1: 10,000 species-specific HRP-conjugated antibody. Membranes were washed again three times in PBST before visualization with ECL substrate.

Measurement of the NADH/NAD ⁺ ratio in Human Cancer Cells

The NADH/NAD ⁺ ratio was measured by modification of manufacturer instructions for NAD ⁺/NADH Glo Assay kit. The lysis buffer is 1%Dodecyltrimethylammonium bromide (DTAB) in 0.2 N NaOH.

To measure the NADH/NAD ⁺ ratio in human cell lines, cells were plated in 96-well plates at 16,000 cells per well. The following day cells were incubated with DMSO or ETC inhibitors (piericidin A: 10 μM; antimycin A, 10 μM) in growth medium for 2 hours. Media were aspirated and cells were extracted by adding 125 μl ice cold lysis buffer (diluted 1: 1 with PBS) into the plate and incubated 3 minutes on an orbital shaker to ensure homogenous cell lysis. 50 μl supernatants were used to measure NADH and NAD ⁺respectively.

To measure NADH, 50 μl of samples were moved to empty wells of 96 well plate and incubated at 60 ℃ for 15 minutes to degrade NAD ⁺. To measure NAD ⁺, 50 μl of the samples were moved to empty wells of 96 well plate, mixed with 25 μl 0.4 N HCl and incubated at 60 ℃ for 15 minutes to degrade NADH. Following incubations, samples were allowed to equilibrate for 10 minutes at room temperature and quenched by neutralizing with 25 μl 0.5 M Tris (NAD ⁺) or 50 μl of HCl/Tris (NADH) . 20 μl samples were mixed with 20 μl detection reagents in a 384 well plate. Luminescence signals were recorded after 30 minutes incubation at room temperature.

Metabolite Extraction from Human Cancer Cells and Mouse Tissues

For metabolite extraction from human cells, 143B cells (2 million per sample) cultured in 6-cm dish were incubated with DMSO, 10 μM Piericidin A or 10 μM Antimycin A for 2 hours. After washing with PBS twice, cells were extracted by -80 ℃ pre-chilled 80%methanol for the cellular steady metabolite measurements. The remaining cell pellets were dissolved in 0.1 M KOH by shaking at 4 ℃ overnight. Protein concentration was measured by Bradford assay kit. The metabolic data was normalized to protein concentration.

For metabolite extraction from mouse tissues, mice were anesthetized during tissue collection. Tissues were dissected out and snap-frozen in liquid nitrogen. Tissues were weighed and cryohomogenized on dry ice and the homogenized frozen powders were extracted by ice-cold 4/4/2 acetonitrile/methanol/water according to the weight (100 ul/10 mg) . Samples were incubated on ice for 30 minutes. During incubation, samples were mixed for 1 minute every 10 minutes. After centrifugation at 15,000 rpm for 15 minutes at 4 ℃, 500 μl supernatants from each sample were transferred to a new tube and were lyophilized by Speedvac to dry pellet at 30 ℃. Dried samples were stored at -80 ℃.

LC-MS Analysis of Metabolites

LC-MS analysis was performed by a Thermo Vanquish UHPLC coupled to a Thermo Q Exactive HF-X hybrid quadrupole-Orbitrap mass spectrometer. Authentic reference standard compounds were purchased from Sigma-Aldrich to confirm the retention time of each targeted metabolites. Dried samples were resuspended in 100 μl 50%methanol and filtered through 0.45 μm filters. 6 μl of each sample was injected for analysis. Chromatographic separation was performed on a Merck ZIC-HILIC column (2.1×100 mm, 3.5 μm) at a flow rate of 0.5 ml/min and maintained at 40 ℃, with the mobile phases of 10 mM ammonium acetate in ACN/water 5/95 (A) and 10 mM ammonium acetate in ACN/water 95/5 (B) . The following gradient was applied: 0-5 min, 99%B; 5-20 min, 99-20%B; 20-21 min, 20-99%B; 21-25 min, 99%B. Full-scan mass spectra were acquired in the range of m/z 66.7 to 1,000 with the following ESI source settings: spray voltage: 3.5 kV, auxiliary gas heater temperature: 380 ℃, capillary temperature: 320 ℃, sheath gas flow rate: 35 units, auxiliary gas flow: 10 units in the positive mode; Spray voltage: 2.5 kV, auxiliary gas heater temperature: 380 ℃, capillary temperature: 320 ℃, sheath gas flow rate: 30 units, auxiliary gas flow: 10 units in the negative mode. MS1 scan parameters included resolution 60,000, AGC target 3e6, and maximum injection time 200 ms. Data processing was performed with Thermo Xcalibur software (4.2) . Results were manually validated and compounds with relative standard deviation (RSD) of quality control peak areas greater than 20 %were excluded.

Measurement of Complex-I-Dependent and Complex-I-Independent Oxygen Consumption Rates in Human Cancer Cells and Cultured Mouse Cortical Neurons

Oxygen consumption rates (OCRs) were determined using an XF96 Extracellular Flux Analyzer. Cell lines were plated at 10,000 cells per well in 80 μl DMEM supplemented with 10%FBS and 1%P/Sthe day before OCR measurement.

Neurons were cultured at 50,000 cells per well (pre-coated with poly-D-lysine) with 150 μl Neurobasal medium supplemented with 1 mM sodium pyruvate, 2%B27 and 1%GlutaMAX. Neurons were infected with AAV-PHP. eB virus expressing GFP or GPD1 three days after attachment (virus packaging and infection protocol are in section below) . OCRs of neurons were measured 10 days after viral infection.

Before measurements, cells were washed 2 times in assay media: XF Base Medium supplemented with 25 mM glucose and 4 mM glutamine and adjusted to pH 7.4. Cells were then incubated in 180 μl assay media at 37℃ without CO ₂ for 1 hour. Measurements were performed at consecutive intervals of mixing (3 minutes) and measurement (3 minutes) . Complex-I dependent and Complex-I independent OCRs of cell lines were determined by injection of compounds: piericidin A (2 μM) followed by antimycin A (2 μM) . Neuronal Complex-I dependent and Complex-I independent OCRs were determined by injection of compounds: DNP (2, 4-dinitrophenol, 200 μM) , piericidin A (2 μM) followed by antimycin A (2 μM) . DNP is an uncoupler to stimulate maximum OCR.

Measurement of Cellular ATP Level in Human Cancer Cells and Cultured Mouse Cortical Neurons

To measure ATP level, cell lines were plated in triplicates in 96-well plates at 10,000 cells per well 16 hours before measurement. Neurons were cultured and infected as the OCR experiment above. ATP levels of neurons were measured 10 days after viral infection.

Neurons and Cell lines were incubated with DMSO, 10 μM piericidin A or 10 μM antimycin A in DMEM supplemented with 10%FBS and 1%FBS for 2 hours. ATP level was measured by Cell Titer-Glo Luminescent assay kit following the manufacturer instructions. Briefly, Cells were lysed by adding 20 μl CellTiter-Glo reagent into the media and mixed on an orbital shaker for 5 minutes to lyse the cells. Luminescence signals were recorded after 10 minutes equilibration at room temperature.

Quantitive PCR Analysis of Mitochondrial Integrative Stress Response in 143B Cells

350,000 143B cells were seeded in 6-cm dishes in 4 ml DMEM 24 hours prior to inhibitor treatment. Cells were incubated with DMSO, 10 μM piericidin A or 10 μM antimycin A for 12 hours in the presence of 50 μg/ml uridine. Total RNAs were extracted using a Direct-zol RNA Miniprep Kit. cDNA, generated by reverse transcription by 5×All-In-One RT MasterMix, together with primers were combined with the TB Green Premix Ex Taq and assayed by quantitative PCR (qPCR) . All CT values for genes of interest were normalized against the ACTINB gene (probe/control) . The ratio (probe/control) was set to 1 for DMSO-treated cells. The oligonucleotides for qPCR analysis were listed below.

Xenograft Tumor Growth

Cancer cells were suspended in 125 μl PBS and injected into the flanks of 5 week-old, female BALB/c nude mice at 1×10 ⁶ per injection for 143B cells and 5×10 ⁶ cells per injection for HeLa cells. When tumors became palpable, tumor volume was measured by calipers in two dimensions and the volumes were estimated using the equation V = (π /6) (L×W ²) . For experiments determining the efficacy of metformin treatment in tumors, tumors were permitted to grow to 50 mm ³, after which animals were randomly assigned to the metformin treatment or the vehicle group. Vehicle (water) or 1 g/kg metformin (in water) was dosed via oral gavage daily for indicated time. Tumor volume was measured every two days. Mice were killed at end points that were consistent with our mouse protocol. Primary investigator was not blinded during or after experiments, although results were qualitatively verified by blinded investigators.

Adeno-associated Virus (AAV) Package

To produce AAV particles, three plasmids were required: the AAV vector, the capsid vector PHP. eB, and the p. Helper vector. AAV vectors expressing mouse Ndufs4, human GPD1, GPD1 ^K204A, or GFP-3×HA were driven by the EF1α promoter, added with bGH poly (A) signal at 3’ end of the coding sequence and flanked by AAV2 ITRs at both ends.

Specifically, the nucleotide sequence of the vector construct expressing human GPD1 is shown as SEQ ID NO. 31. The GPD1 cDNA accession number is NM_005276.4. Virus package was conducted as described above.

AAV-pro cells (80%density) were transfected with AAV vectors along with AAV packaging vectors PHP. eB or PHP. B and p. Helper at a ratio of 4: 6: 5 by the PEI transfection reagent. 60 hours later, cell pellets were scraped and collected by centrifuging into tubes containing GB buffer (10 mM Tris (PH 7.6) , 150 mM NaCl and10 mM MgCl ₂) and glass beads. Cells were lysed by freezing in liquid nitrogen for 8 minutes followed by thawing at 37 ℃ for 3 minutes and then shaking for 2 minutes. After repeating this freezing-thawing-shaking procedure for 5 times, benzonase nuclease was used to eliminate cellular DNA after incubation at 37 ℃ for 30 minutes. Virus-containing supernatants were obtained by centrifuging at 15,000 rpm for 45 minutes. Optiprep iodixanol solution was used to prepare different iodixanol gradient: 15%iodixanol fraction (15%iodixanol and 1 M NaCl in GB buffer) , 25%iodixanol fraction (25%iodixanol in GB buffer) , 40%iodixanol fraction (40%iodixanol in GB buffer) and 58%iodixanol fraction (58%iodixanol in GB buffer) . Virus was purified by iodixanol gradient ultracentrifugation at 41,000 rpm for 4 hours (Zolotukhin et al., 1999) . Virus titer was determined by qPCR. All centrifugations were performed at 4 ℃.

Adeno-associated Virus (AAV) Delivery to Ndufs4 ^-/- Mice and Cultured Neurons

At P16-18, Ndufs4 ^-/- mice and control littermates were randomly assigned to receive an intravenous injection of 150 μl of AAV-PHP. eB virus (4×10 ¹⁰ genome copies/μl, diluted in PBS) through the retro-orbital injection (Yardeni et al., 2011) .

Neurons plated in 24 well plate after 3 days of culture were infected with 1 μl of AAV-PHP. eB virus (4×10 ¹⁰ genome copies/μl, diluted in PBS) per well. Culture was maintained by changing virus containing medium every 3 days. Experiments with neurons were performed 10 days after viral infection.

Measurements of Body Temperature

A thermal imaging camera was used to measure mouse body temperature. The pictures were taken at cochlea site of free moved mice at ～P30, P40, and P50 days. Six pictures were taken for one mouse. For a single time point, such as P30 body temperature, mice body temperatures were measured three consecutive days from P29-P31 to confirm the exact body temperature. Pictures were analyzed by Acrobat Reader DC. Temperatures are shown as mean ± SEM.

Histology

Mice were anesthetized during tissue collection. The chest cavity was opened and a catheter was placed in the left ventricle. The whole body was perfused with PBS and then with 4%PFA. The brain was dissected out, stored overnight in 4%paraformaldehyde (PFA) and then placed in 30%sucrose (in PBS) for two days. PFA-perfused whole brains were sectioned parasagittally or coronally by 50 μm. Immunohistochemistry was performed using an antibody recognizing the microglial marker Iba-1 or antibody recognizing HA according to previous described methods (Quintana et al., 2010) .

QUANTIFICATION AND STATISTICAL ANALYSIS

Quantification of Iba1 Fluorescence Intensity

The mean fluorescence intensities of Alexa Fluor 568 in different brain areas were measured with the ImageJ software. Background fluorescence was subtracted from the measured fluorescence. Quantification results are represented as the mean ± SEM.

Statistical Analyses

Data were presented as the mean ± standard deviation (SD) unless otherwise indicated. Sample size (n) indicates biological replicates from a single representative experiment. For analyzing C. elegans lifespan data, statistical significance was determined by the log-rank test. For mouse study, Sample size (n) indicates data from a distinct mouse. The results of all experiments were validated by independent repetitions. For comparing of two groups, statistical significance was determined using a two-tailed unpaired Student’s t-test; For comparing multiple groups, statistical significance was determined by one-way ANOVA using Tukey’s test or two-way ANOVA using Tukey’s test. P values are denoted in figures as: not significant [ns] , *P < 0.05, **P < 0.01, ***P < 0.001.

The materials used in the experiments as described above can be obtained from following sources.

The sequence information is shown in the following table.

Claims

A method for treating mitochondrial complex I deficiency in a subject, comprising administering to the subject a therapeutically effective amount of an agent enhancing Gro3P biosynthesis.
The method of claim 1, wherein the agent enhancing Gro3P biosynthesis increases the level or enzymatic activity of one or more cytosolic Gro3P dehydrogenases (cGPDs) .
The method of claim 2, wherein the one or more cGPDs is human GPD1 and/or human GPD1L.
The method of claim 2 or claim 3, wherein the agent is a nucleic acid molecule comprising a nucleotide sequence encoding the one or more of cGPDs, a nucleotide sequence encoding a variant of cGPD having the same or enhanced enzymatic activity as compared to wild-type cGPD, or a nucleotide sequence enhancing the expression of one or more cGPDs.
The method of claim 2 or claim 3, wherein the agent is a vector comprising the nucleic acid molecule of claim 4.
The method of claim 5, wherein the vector further comprises a promoter which controls the expression of the nucleotide sequence, preferably the promoter is selected from a constitutive promoter, a tissue-specific promoter or a cell-type-specific promoter, more preferably a neuro-specific, a brain-specific promoter or an ocular tissue-specific promoter.
The method of claim 5 or claim 6, wherein the vector is a recombinant adeno-associated virus (rAAV) vector.
The method of claim 7, wherein the rAAV vector comprises at least one, preferably two ITRs.
The method of claim 8, wherein the ITR is AAV1, AAV2, AAV3, AAV4, AAV5, AAV6, AAV7, AAV8, AAV9, AAVrh10, AAV. 7m8, or AAV2tYF, AAV11, AAV12 serotype ITR, preferably AAV2, AAV4, AAV5, AAV6, AAV8, AAV9, AAVrh10, AAV. 7m8, or AAV2tYF serotype ITR, more preferably AAV2 ITR.
The method of any one of claims 7 to 9, wherein the rAAV vector comprises a polyadenylation sequence, preferably bGH polyA, SV40 polyA or WPRE-SV40 poly A, more preferably bGH polyA.
The method of any one of claims 7 to 10, wherein the rAAV vector comprises a nucleotide sequence as shown in SEQ ID NO: 31 or a nucleotide sequence having at least 80%homology to SEQ ID NO: 31.
The method of any one of claims 7 to 11, wherein the rAAV vector is encapsidated in capsid proteins, preferably of AAV1, AAV2, AAV3, AAV4, AAV5, AAV6, AAV7, AAV8, AAV9, AAVrh10, AAV. 7m8, AAV2tYF, AAV11, or AAV12 serotype.
The method of claim 12, wherein at least one of the capsid proteins is of AAV2, AAV4, AAV5, AAV8, AAV9, or AAVrh. 10 serotype.
The method of claim 2 or claim 3, wherein the agent is an agonist of one or more of cGPDs.
The method of any one of claims 1-14, wherein the complex I deficiency is selected from a group consisting of Leigh syndrome, Leber’s optical hereditary neuropathy (LHON) , Leber optic atrophy and dystonia, mitochondrial complex I deficiency (OMIM entry 252010) , leukodystrophy, MELAS (mitochondrial encephalomyopathy, lactic acidosis, and stroke-like episodes) syndrome, optic atrophy with demyelinating disease of CNS and other neurodegenerative diseases caused by complex I deficiency.
The method of any one of claims 1-15, wherein the subject is a mammal, preferably human.
The method of any one of claims 1-16, wherein the subject has at least one mutation in the mtDNA encoding complex I subunit.
A method for treating tumor in a subject, comprising administering to the subject a therapeutically effective amount of an agent inhibiting Gro3P biosynthesis of tumor cells.
The method of claim 18, wherein the agent inhibiting Gro3P biosynthesis is a chemical compound, a polypeptide, or a polynucleotide inhibiting the activity of one or more cGPDs.
The method of claim 19, wherein the cGPD is human GPD1 and/or GPD1L.
The method of any one of claims 18 to 20, further comprising administering to the subject a therapeutically effective amount of an inhibitor of complex I.
The method of claim 21, wherein the inhibitor of complex I is selected from a group consisted of metformin, phenformin, BAY84-2243, CAI, ME-344, fenofibrate, and mIBG.
A pharmaceutical combination of an agent inhibiting Gro3P biosynthesis and an inhibitor of complex I.
The combination of claim 23, wherein the agent inhibiting Gro3P biosynthesis is an agent inhibiting or eliminating the activity of one or more cGPDs.
The combination of claim 24, wherein the cGPD is human GPD1 and/or GPD1L.
The combination of claim 23, wherein the inhibitor of complex I is selected from a group consisted of metformin, phenformin, BAY84-2243, CAI, ME-344, fenofibrate, and mIBG.
The combination of any one of claims 23-26, which is used for treating tumor.
The combination of any one of claims 23-27, which exhibits synergistic anti-tumor effect.