Direct transdifferentiation treatment of nervous system diseases

Abstract

The present disclosure relates to the use of RE1 / NRSE elements as targets for transdifferentiation of non-neuronal cells into neuronal cells; and the use of the RE1 / NRSE blocker in the prevention and / or treatment of diseases associated with neuronal dysfunction or death. The present disclosure provides a method of blocking a RE1 / NRSE element to modulate the expression of a neuronal-related gene in a non-neuronal cell, comprising reducing the binding of REST to the RE1 / NRSE element, or reducing the amount or activity of REST. The present disclosure also provides RE1 / NRSE element blockers, in particular binding domains of endogenous RE1 / NRSE binding proteins and variants thereof, which are useful in the prevention and / or treatment of diseases associated with neuronal dysfunction or death.

Description

The present application is a divisional application of case number 202111158620.4, with the application date of September 30, 2021, and the invention title of Direct transdifferentiation treatment of nervous system diseases. TECHNICAL FIELD

[0001] The present disclosure relates to the field of biological medicine. More specifically, the present disclosure relates to RE1 / NRSE elements as a target for transdifferentiating non-neuronal cells into neuronal cells; and the use of RE1 / NRSE blockers in preventing and / or treating diseases associated with loss or death of neuronal function. BACKGROUND

[0002] The repressor element 1 / neuron-restrictive silencer element (RE1 / NRSE) is a specific DNA sequence of about 21 bp (20-23 bp varies) that mainly binds with REST (RE1-silencing transcription factor, also known as neuron-restrictive silencer factor (NRSF)) to regulate the expression of genes related to neuronal development and maturation. RE1 is a negative regulatory element related to neuronal maturation, which was first found in the 5' end of the promoters of NaV1.2 and SCG10 to regulate the expression of these genes. In non-neuronal cells, the RE1 site is bound by a silencing complex composed of histone deacetylases and methylases, etc., to inhibit the expression of neuronal-related genes. However, there are more than 1800 RE1 elements in mice and humans, making it difficult to regulate them with some existing technologies. For example, CRISPR-mediated gene regulation and epigenetic modification technologies have very high precision and can accurately regulate the expression of specific genes, but it is difficult to regulate the expression of genes regulated by RE1 in this way.

[0003] Parkinson's disease (PD) is a disease associated with loss or death of neurons, characterized by loss of dopaminergic neurons in the substantia nigra of the midbrain. Currently, the treatment for Parkinson's disease is mainly small molecule drugs represented by levodopa and other dopamine analogs. Surgical treatment developed in recent years can also improve the symptoms of the disease to some extent, such as deep brain electrode stimulation. However, these methods can only partially alleviate the disease and cannot stop the progression of the disease or slow down the death of dopamine neurons. The concept of transdifferentiation therapy brings hope for the regeneration of dopamine neurons. By overexpressing some genes in glial cells or editing genes in glial cells, glial cells are transdifferentiated into dopamine neurons to supplement the missing or dead dopamine neurons. Müller glia (MG) is the main glial cell in the retina. Retinal ganglion cells (RGCs) are the innermost layer of the retina, and their dendrites mainly establish synaptic connections with bipolar cells. Their axons extend to the optic nerve head to form the optic nerve and extend to the brain. Retinal ganglion cell (RGC) damage or degenerative disease is the main cause of permanent blindness. RGC is the only output neuron in the retina, so RGC damage or degenerative disease will lead to permanent blindness. Rebuilding functional RGCs is currently the only way to help blind patients regain vision, and RGC cells are very difficult to regenerate, so exploring how to regenerate RGC cells will bring hope to the vast number of blind patients caused by RGC cell death. Therefore, there is an urgent need in the art to develop methods that can regenerate dopamine neurons, retinal ganglion cells, or other functional neurons. SUMMARY

[0004] In one embodiment, the disclosure provides a method of blocking RE1 / NRSE elements to modulate expression of neuron-associated genes in non-neuronal cells, comprising reducing the amount or activity of REST, or reducing the binding of REST to RE1 / NRSE elements.

[0005] In a preferred embodiment, the amount of REST is reduced by a method such as gene editing, small RNA interference, or protein accelerated degradation.

[0006] In a preferred embodiment, the amount of REST is reduced by a method such as gene editing, antisense oligonucleotides (ASOs), small RNA interference, miRNA technology, small molecule compounds, or protein accelerated degradation.

[0007] In a preferred embodiment, the binding of REST to RE1 / NRSE elements is blocked by binding of a REST-binding agent, such as a REST antibody, to REST.

[0008] In a preferred embodiment, the binding of REST to the RE1 / NRSE element is blocked by the binding of a RE1 / NRSE element blocker to the RE1 / NRSE element.

[0009] In a preferred embodiment, the RE1 / NRSE element blocker is a competitive binding protein, a short peptide or a gene editing protein of REST or a coding nucleic acid thereof, or a nucleic acid and nucleic acid analog, or a small molecule RE1 / NRSE element blocker.

[0010] In a preferred embodiment, the RE1 / NRSE element blocker is a REST variant or a coding nucleic acid thereof.

[0011] In a preferred embodiment, the REST variant is a DNA binding domain of REST which lacks the N-terminal and C-terminal repression domains of REST, preferably the amino acids 155-420 of REST.

[0012] In a preferred embodiment, the DNA binding domain of REST is fused to an activation domain.

[0013] In a preferred embodiment, the activation domain is selected from the group consisting of an epigenetic modifier protein or a gene activation regulatory element, such as VP64, P65-HSF1, VP16, RTA, Suntag, P300, CBP or a combination thereof, preferably VP64 or P65-HSF1.

[0014] In a preferred embodiment, the REST variant has the amino acid sequence of SEQ ID NO: 1, 3, 5 and 9 or the nucleotide sequence of SEQ ID NO: 2, 4, 6 and 10, or a sequence having at least 70%, 60%, 50% percent identity thereto.

[0015] In a preferred embodiment, the non-neuronal cell comprises a glial cell, a fibroblast cell, a stem cell, a neural precursor cell, a neural stem cell, wherein the glial cell is selected from the group consisting of an astrocyte, an oligodendrocyte, an ependymal cell, a Schwann cell, an NG2 cell, a satellite cell, a Muller glial cell, an inner ear glial cell or a combination thereof, preferably an astrocyte, a Muller glial cell and a cochlear glial cell.

[0016] In a preferred embodiment, the glial cell is derived from the brain, the spinal cord, the eye or the ear, wherein the brain glial cell is derived from the striatum, the substantia nigra, the ventral midbrain tegmentum, the spinal cord, the hypothalamus, the dorsal midbrain or the cerebral cortex, preferably the striatum and the substantia nigra.

[0017] In a preferred embodiment, the expression of neuron-related genes in the non-neuronal cells is regulated so that the non-neuronal cells are transdifferentiated into neuronal cells, wherein the neuronal cells are mammalian neurons, preferably dopamine neurons, GABA neurons, 5-HT neurons, glutamatergic neurons, ChAT neurons, NE neurons, motor neurons, spinal neurons, spinal motor neurons, spinal sensory neurons, photoreceptor cells (rod cells and cone cells), bipolar cells, horizontal cells, amacrine cells, retinal ganglion cells (RGC), cochlear nerve cells (cochlear spiral ganglion cells and vestibular neurons), pyramidal neurons, interneurons, medium spiny neurons (MSN), Purkinje cells, granule cells, olfactory sensory neurons, periglomerular cells or a combination thereof, more preferably dopamine neurons, retinal ganglion cells, photoreceptor cells and cochlear spiral ganglion cells.

[0018] In a preferred embodiment, the non-neuronal cells and / or neuronal cells are from, for example, humans, non-human primates, rats and mice, preferably humans.

[0019] In another embodiment, the present disclosure provides the use of an RE1 / NRSE element blocker for the preparation of a medicament for preventing and / or treating a disease associated with neuronal dysfunction or death, wherein the RE1 / NRSE element blocker reduces the binding of an endogenous RE1 / NRSE binding factor to the RE1 / NRSE element, wherein the endogenous RE1 / NRSE binding factor includes a zinc finger protein such as REST.

[0020] In a preferred embodiment, the RE1 / NRSE element blocker binds to the RE1 / NRSE element, thereby blocking the binding of the RE1 / NRSE endogenous binding factor to the RE1 / NRSE element.

[0021] In a preferred embodiment, the RE1 / NRSE element blocker is a competitive binding protein, short peptide or gene editing protein of REST or its encoding nucleic acid, or nucleic acid and nucleic acid analogs, or a small molecule RE1 / NRSE element blocker.

[0022] In a preferred embodiment, the RE1 / NRSE element blocker is a REST variant or a nucleic acid encoding it.

[0023] In a preferred embodiment, the REST variant is the DNA binding domain of REST, which lacks the N-terminal and C-terminal inhibitory domains of REST, preferably amino acids 155-420 of REST.

[0024] In a preferred embodiment, the DNA binding domain of REST is fused to the activation domain.

[0025] In a preferred embodiment, the activation domain is selected from the group consisting of: epigenetic modifier proteins or gene activation regulatory elements, such as VP64, P65-HSF1, VP16, RTA, Suntag, P300, CBP, or combinations thereof, preferably VP64 or P65-HSF1.

[0026] In a preferred embodiment, the REST variant has the amino acid sequence of SEQ ID NO: 1, 3, 5 and 9 or the nucleotide sequence of SEQ ID NO: 2, 4, 6 and 10, or a sequence having at least 70%, 60%, 50% percent identity thereto.

[0027] In a preferred embodiment, the disease associated with loss or death of neurons is selected from the group consisting of: Parkinson’s disease, Alzheimer’s disease, stroke, schizophrenia, Huntington’s disease, depression, motor neuron disease, amyotrophic lateral sclerosis, spinal muscular atrophy, Pick’s disease, sleep disorders, epilepsy, ataxia, vision impairment due to RGC cell death, glaucoma, age-related RGC pathology, optic nerve injury, retinal ischemia or hemorrhage, Leber’s hereditary optic neuropathy, photoreceptor degeneration or death due to injury or degenerative disease, macular degeneration, retinitis pigmentosa, diabetes-related blindness, night blindness, color blindness, inherited blindness, congenital blindness, hearing loss or hypoacusia due to spiral ganglion cell death, or combinations thereof.

[0028] In another embodiment, the present disclosure provides a RE1 / NRSE element blocker that is a REST variant or a nucleic acid encoding thereof, or an artificially designed REST DNA binding domain analog.

[0029] In a preferred embodiment, the REST variant is a REST DNA binding domain that lacks the N-terminal and C-terminal repression domains of REST, preferably amino acids 155-420 of REST.

[0030] In a preferred embodiment, the REST DNA binding domain is fused to an activation domain.

[0031] In a preferred embodiment, the activation domain is selected from the group consisting of: epigenetic modifier proteins or gene activation regulatory elements, such as VP64, P65-HSF1, VP16, RTA, Suntag, P300, CBP, or combinations thereof, preferably VP64 or P65-HSF1.

[0032] In a preferred embodiment, the REST variant has the amino acid sequence of SEQ ID NO: 1, 3, 5 and 9 or the nucleotide sequence of SEQ ID NO: 2, 4, 6 and 10, or a sequence having at least 70%, 60%, 50% percent identity thereto.

[0033] In a preferred embodiment, the REST variant or the DNA binding domain of REST is from, for example, human, non-human primates, rats and mice, preferably human.

[0034] In another embodiment, the present disclosure provides a pharmaceutical composition or a kit or a reagent comprising the above blocking agent.

[0035] In a preferred embodiment, the pharmaceutical composition or the kit or the reagent is formulated for injection, intracranial administration, intraocular administration, intra-aural administration, inhalation, parenteral administration, intravenous administration, intramuscular administration, intradermal administration, topical administration or oral administration.

[0036] In a preferred embodiment, the pharmaceutical composition or the kit or the reagent further comprises a vector or carrier for delivering the RE1 / NRSE element blocking agent, wherein the vector or carrier is a viral vector, a liposome, a nanoparticle, an exosome, a virus-like particle, wherein the viral vector comprises a recombinant adeno-associated viral vector (rAAV), an adeno-associated viral (AAV) vector, an adenoviral vector, a lentiviral vector, a retroviral vector, a poxviral vector, a herpes virus, an SV40 viral vector, and combinations thereof, wherein AAV and rAAV are preferred.

[0037] In a preferred embodiment, the pharmaceutical composition or the kit or the reagent comprises an expression vector for expressing the REST variant, wherein the expression vector comprises a nucleotide sequence encoding the REST variant operably linked to a promoter that causes its expression.

[0038] In a preferred embodiment, the pharmaceutical composition or the kit or the reagent is for local administration to at least one of: i) glial cells in the striatum; ii) glial cells in the substantia nigra of the brain; iii) glial cells in the retina; iv) glial cells in the inner ear; v) glial cells in the spinal cord; vi) glial cells in the prefrontal cortex; vii) glial cells in the motor cortex; viii) glial cells in the hypothalamus; and ix) glial cells in the ventral tegmental area (VTA).

[0039] In a preferred embodiment, the pharmaceutical composition or the kit or the reagent further comprises i) one or more dopamine neuron related factors, or ii) one or more retinal ganglion cell related factors for expression in Muller glial cells, 1) wherein the one or more dopamine neuron related factors are selected from the group consisting of FoxA2, Lmxla, Lmxlb, Nurr1, Pbxla, Pitx3, Gata2, Gata3, FGF8, BMP, Enl, En2, PET1, Pax family proteins (Pax3, Pax6, etc.), SHH, Wnt family proteins, and TGF-beta family proteins, or a combination thereof; 2) wherein the one or more retinal ganglion cell related factors include factors such as beta-catenin, Oct4, Sox2, Klf4, Crx, aCamKII, Brn3a, Brn3b, Brn3C, Math5, Otx2, Ngn2, Ngnl, AscLl, miRNA9, miRNA-124, Nr2e3, and Nrl.

[0040] In a preferred embodiment, the promoter is a glial cell specific promoter or a Muller glial cell (MG) cell specific promoter selected from the group consisting of GFAP promoter, ALDH1L1 promoter, EAAT1 / GLAST promoter, glutamine synthetase promoter, S100β promoter EAAT2 / GLT-1 promoter, and Rlbp1 promoter, preferably the GFAP promoter.

[0041] In a preferred embodiment, the transdifferentiation efficiency of the glial cells is at least 1%, or at least 10%, 20%, 30%, 40%, or 50%. BRIEF DESCRIPTION OF DRAWINGS

[0042] Figure 1Figure 1. Schematic diagram of the design of endogenous zinc finger protein. (A) Schematic diagram of REST protein structure. REST protein contains N-terminal inhibitory domain and middle DNA binding domain responsible for binding RE1, and C-terminal transcriptional inhibitory domain. RZFD-V1 represents the first design of RZFD, which contains eight zinc finger domains in the middle of human REST protein responsible for binding RE1 (RZFD, REST Zinc Figure domain). (B) Two designs of RZFD-V2 and RZFD-V3. In order to activate the neuron-related genes that are inhibited by REST through binding RE1 / NRSE, we fused RZFD with two different activators, respectively named RZFD-V2 and RZFD-V3. RZFD-V2 is expressed by fusion of RZFD with VP64, and RZFD-V3 is composed of a P65 and a HSF1 transcriptional activation domain fused to the C-terminus of RZFD. (C) In glial cells, REST binds to RE1, and the C- and N-termini of REST recruit some transcriptional inhibitory factors, respectively, so that the neuron-related genes regulated by RE1 cannot be expressed. (D) We envisage that after expressing RZFD-V1 in glial cells, RZFD binds to RE1, preventing REST from binding to RE1, so that the REST silencing complex cannot inhibit the genes regulated by RE1, and the neuron-related genes regulated by RE1 are expressed. (E) RZFD-V2 binds to RE1 through the RZFD domain, preventing REST / NRSF from binding to RE1, while its VP64 domain can recruit transcriptional activators, enhancing the expression of neuron-related genes regulated by RE1. (F) During the process of glial cell transdifferentiation into neurons, the RZFD domain of RZFD-V3 binds to RE1, preventing REST from binding to RE1, relieving REST from the expression of genes regulated by RE1, while the P65-HSF1 activation domain recruits some transcriptional activators, promoting the expression of neuron-related genes regulated by RE1.

[0043] Figure 2RZFD-mediated glial-to-neuronal transdifferentiation. (A) Schematic diagram of AAV vector design. Vector 1 is a schematic diagram of the vector that expresses mCherry driven by GFAP, a glial-specific promoter. mCherry is a red fluorescent protein used to label glial cells. Vector 2 is a schematic diagram of the human RZFD expression vector, in which the astrocyte-specific promoter GFAP drives RZFD expression. (B) Schematic diagram of injection and sample collection and analysis. The day of AAV injection is designated as day 0. Samples are collected 2 weeks after injection for analysis of glial-to-neuron transdifferentiation. 1.5 months after injection, analysis of glial-to-neuron and dopamine-neuron transdifferentiation is performed. (C) Schematic diagram of AAV viral injection and transdifferentiation. GFAP-mCherry alone or a mixture of GFAP-mCherry and GFAP-RZFD is injected into the striatum or substantia nigra of mice. GFAP-mCherry labels glial cells red, while GFAP-RZFD induces glial-to-neuron transdifferentiation. (D) After injection of GFAP-mCherry alone into the striatum of wild-type C57 mice, astrocytes are labeled red. The mCherry channel indicates GFAP-mCherry-labeled astrocytes, DAPI stains the nuclei, and NeuN is a neuron-specific marker. The Merge image shows that GFAP-mCherry specifically labels astrocytes but not neurons. (E) Two weeks after injection of a mixed AAV of GFAP-mCherry and GFAP-RZFD into the striatum, samples were collected and analyzed. Most glial cells have begun to deform, and a small number of cells have begun to express NeuN, but no cells express TH, a neuron-specific marker and a dopamine neuron-specific marker. (F) 1.5 months after injection of a mixed AAV of GFAP-mCherry and GFAP-RZFD into the striatum, samples were collected and analyzed. Most mCherry-positive cells express neuron-specific markers, and a small number of cells express the dopamine-specific marker TH. Arrows indicate neurons that express both mCherry and TH.

[0044] Figure 3RZFD-VP64 or RZFD-P65-HSF1-mediated glial-to-neuronal transdifferentiation. (A) Schematic diagram of the design of the GFAP-RZFD-V2 and GFAP-RZFD-V3 expression vectors. RZFD-V2 is a fusion protein of RZFD and VP64, driven by the glial cell-specific promoter GFAP. RZFD-V3 consists of RZFD and the activation domain of P65-HSF1. (B) Schematic diagram of the mouse striatum or substantia nigra injected with a mixture of GFAP-RZFD-V2 and GFAP-mCherry via AAV. Samples were collected and analyzed 1.5 months after injection. (C) Representative images of GFAP-RZFD-V2 transdifferentiating glial cells into neurons and dopamine neurons in the striatum. mCherry indicates cells labeled with GFAP-mCherry, TH is a dopamine neuron-specific marker, and NeuN is a neuron-specific marker. Arrows indicate neurons expressing both mCherry and TH. (D) Schematic diagram of GFAP-RZFD-V3 injection into DAT-Cre:Ai9 mice. GFAP-RZFD-V3 was injected into the striatum or substantia nigra of DAT-Cre:Ai9 mice, and brain tissue was harvested and analyzed 1.5 months later. Dat-Cre mice are constructed by inserting Cre behind the endogenous Dat promoter, and Ai9 mice are constructed by Rosa26-CAG-LSL-tdTomato-WPRE mice. Only mature dopamine neurons in the brains of DAT-Cre:Ai9 mice are labeled with the tdTomato red fluorescent signal. (E) Representative images of brain tissue analysis results 1.5 months after GFAP-RZFD-V3 injection into the striatum. The red fluorescent signal (tdTomato) represents converted mature dopamine neurons. The green signal represents staining for the dopamine neuron-specific marker TH, and the white signal represents staining for the neuron-specific marker NeuN. The merged image shows the overlap of the tdTomato red and TH green signals. Arrows indicate neurons that express both tdTomato and TH. DETAILED DESCRIPTION

[0045] Although previous studies have shown that RE1 is present in the promoter regions of many neural-related genes, whether regulating RE1 can promote glial-to-neuronal transdifferentiation remains unknown. In this study, we genetically engineered an endogenous RE1-binding protein and exploited the zinc finger domain (ZFD) of the endogenous RE1-binding protein REST to regulate RE1 and thereby modulate the expression of related genes. We first analyzed the REST protein and found that its eight zinc finger domains at positions 159-412 are likely involved in binding to RE1. Positions 1-83 at the N-terminus contain an N-terminal inhibitory region that primarily binds to proteins such as Sin3a and Sin3b. Positions 1008-1097 contain a C-terminal inhibitory domain and a zinc finger domain that primarily binds to proteins such as RCOR1. Positions 84-158 and 413-1007 lack distinct protein domains, and their functions remain unclear, but they may be involved in regulating REST binding to RE1. Previous studies have shown that deleting positions 1-83 and 1008-1097 of the REST protein does not affect the binding of REST to RE1, but it cannot function normally.

[0046] This study utilized protein structure prediction and genetic engineering techniques to analyze and modify the endogenous zinc finger domain of REST. First, through truncation experiments, we demonstrated that overexpression of amino acids 155–420 of REST (containing only the eight zinc finger domains, designated RZFD (REST Zinc Finger Domain)) blocked REST binding to RE1. Using AAV-mediated gene delivery, we demonstrated that overexpression of RZFD in mouse striatal astrocytes resulted in neuronal transdifferentiation of glial cells. We further engineered RZFD by fusing VP64 to either the N-terminus or the C-terminus of RZFD, creating RZFD-VP64, and successfully achieved AAV-mediated glial-to-neuronal transdifferentiation in vivo. We further fused the transcriptional activation domains of P65 and HSF1 to RZFD, creating RZFD-P65-HSF1. Using AAV-mediated transdifferentiation in vivo, we demonstrated that RZFD-P65-HSF1 also transdifferentiated glial cells into neurons in the striatum.

[0047] Previous studies have shown that inhibiting REST can induce neuronal transdifferentiation of glial cells. REST binds to a DNA sequence in the genome called RE1. RE1 is a class of sequences that cannot be targeted using CRISPR technology. In this study, an endogenous human zinc finger construct (REST zinc finger domain (RZFD)) cleverly targets the RE1 sequence, thereby blocking the binding of the REST silencing complex to RE1. Expression of RZFD (RZFD-V1) in non-neuronal cells such as glial cells relieved the REST silencing complex's repression of neuronal gene expression. Furthermore, by fusing RZFD with activation domains such as VP64 (RZFD-V2) or P65-HSF1 (RZFD-V3), we further promoted the expression of neuronal genes and promoted glial-to-neuronal transdifferentiation. Using immunofluorescence staining and the DAT-Cre:Ai9 labeling system, we also demonstrated that RZFD-V1, RZFD-V2, and RZFD-V3 can transdifferentiate glial cells into dopamine neurons.

[0048] By injecting AAV into the inferior ventricular cavity of Ai9 retinas, we found that RZFD, RZFD-VP64, and RZFD-P65-HSF1 can transdifferentiate Müller cells into retinal ganglion cells, and some photoreceptors were observed. Retinal ganglion cells are the only cells in the visual pathway that transmit visual signals to the brain; their loss or death leads to permanent blindness. Studies in nonhuman primates, similar to the results in mice, showed that RZFD, RZFD-VP64, and RZFD-P65-HSF1 can transdifferentiate glial cells into dopamine neurons in the brain and Müller cells into retinal ganglion cells and photoreceptors in the retina. Diseases associated with loss of neuronal function or death

[0049] In the present disclosure, diseases associated with neuronal dysfunction or death mainly include diseases associated with dopamine neuron dysfunction or death, and visual impairment associated with optic ganglion or photoreceptor cell loss or death. Diseases associated with neuronal dysfunction or death include, but are not limited to: Parkinson's disease, schizophrenia, depression, visual impairment caused by RGC cell death, glaucoma, age-related RGC lesions, optic nerve damage, retinal ischemia or hemorrhage, Leber hereditary optic neuropathy, photoreceptor cell degeneration or death caused by damage or degenerative lesions, macular degeneration, retinitis pigmentosa, diabetic-related blindness, night blindness, color blindness, hereditary blindness, congenital amaurosis, deafness or hearing loss caused by spiral ganglion cell death. dopamine neurons

[0050] Dopaminergic neurons are neurons that contain and release dopamine (DA) as a neurotransmitter. Dopamine is a catecholamine neurotransmitter that plays an important biological role in the central nervous system. Dopaminergic neurons in the brain are mainly concentrated in the substantia nigra pars compacta (SNc), ventral tegmental area (VTA), hypothalamus and periventricular area. Many experiments have shown that dopaminergic neurons are closely related to many diseases of the human body, the most typical of which is Parkinson's disease. General procedure Animal ethics:

[0051] The animal feeding and use in this study were completed under the guidance of the principles of the Biomedical Research Ethics Committee of the Brain Science and Intelligence Technology Excellence Innovation Center of the Chinese Academy of Sciences. Plasmid construction:

[0052] The plasmids in this study were constructed by the laboratory itself. The AAV backbone vector was digested with restriction enzymes and subjected to agarose gel electrophoresis, and the backbone vector was recovered. The inserted DNA fragment was PCR with cell cDNA as template, and the PCR fragment was recovered after agarose gel electrophoresis. ClonExpress MultiS One Step Cloning Kit (Vazyme, C113-02) of Nuoyuan Biotechnology Co., Ltd. was used to connect the backbone vector and the fragment. After connection, it was transformed into DH5a E. coli and plated. The next day, single colonies were picked for identification, positive clones were sequenced, and clones with completely correct sequencing were cultured and plasmid extracted. Mouse brain AAV injection:

[0053] The AAV serotype used in this study was AAV8. Stereotactic injection was performed using the Sterotaxic Injection System (C57BL / 6 or Dat-Cre:Ai9 mice, more than two months old). The titers of AAV-GFAP-RZFD-V1, AAV-GFAP-RZFD-V2 and AAV-GFAP-RZFD-V3 were more than 5×10 12 vg / ml (1-3 μl per injection). AAV was injected into the striatum (AP+0.8 mm, ML±1.6 mm and DV-2.8 mm) or substantia nigra (AP-3.0 mm, ML±1.25 mm and DV-4.5 mm). Mouse tissue immunofluorescence staining:

[0054] Samples were collected, sliced ​​and immunofluorescence stained 2 weeks and 1.5-2 months after injection. The brains were removed after perfusion of mice with normal saline and 4% PFA, and fixed with 4% paraformaldehyde (PFA) overnight, then dehydrated in 30% sucrose for at least 12 hours until the tissue sank to the bottom of the solution. After embedding in OCT, frozen sections were made with a thickness of 30μm or 40μm. Before immunofluorescence staining, brain slices were washed three times with 0.1M phosphate buffered saline (PBS) for 5-10 minutes each time. After incubation with the primary antibody at 4°C overnight, the slices were washed 3-4 times with PBS for 10-15 minutes each time. Then, the secondary antibody diluted in antibody diluent was added for incubation, incubated at room temperature for 2-3 hours, and washed 3-4 times with PBS for 10-15 minutes each time. Finally, the slices were sealed and preserved with anti-fluorescence quenching sealing medium (Life Technology). Antibody:

[0055] The primary antibodies used in this study included guinea pig anti-NeuN (1:500, ABN90, Millipore), rabbit anti-TH (1:500, AB152, Millipore), rat anti-DAT (1:100, MAB369, Millipore), rabbit anti-RBPMS (Proteintech, Cat#15187-1-AP), and mouse anti-Flag (1:2000, F3165, Sigma). The secondary antibodies used in this study were: Cy5-AffiniPure Donkey Anti-Guinea Pig IgG(H+L) (1:500,706-175-148,Jackson ImmunoResearch), Alexa Fluora-488 AffiniPure Donkey Anti-Rabbit IgG(H+L) (1:500,711-545-152,Jackson ImmunoResearch), Alexa Fluora-488 AffiniPure Donkey Anti-Mouse IgG(H+L) (1:500,715-545-150,Jackson ImmunoResearch), and Cy5 AffiniPure Donkey Anti-Rabbit IgG(H+L) (1:500,711-175-152,Jackson ImmunoResearch). Electrophysiological recordings:

[0056] Electrophysiological recordings were performed 1-3 months after AAV injection. First, mice were anesthetized and perfused with pre-chilled NMDA, and then the brain was placed in NMDG artificial cerebrospinal fluid (ACSF) [NMDG ACSF (mM): NMDG 92, potassium chloride 2.5, sodium phosphate monobasic 1.25, sodium bicarbonate 30, HEPES 20, glucose 25, sodium thio sulfate 2, sodium ascorbate 5) at room temperature, sodium pyruvate 3, calcium chloride 0.5, magnesium sulfate 10] bubbled with carbon dioxide. After perfusion, the brain was extracted and placed in ice-cold NMDG ACSF solution for 30 seconds. Thick brain slices of 250-350 pm in thickness were cut with a vibratome at a speed of 0.04-0.05 mm / s. The brain slices were moved into a culture dish filled with carbon dioxide-bubbled NMDG ACSF and kept at 32-34°C for < 12 minutes. The slices were transferred to a new culture dish of carbon dioxide-bubbled HEPES ACSF [HEPES, containing ACSF (mM): sodium chloride 92, potassium chloride 2.5, sodium phosphate monobasic 1.25, sodium bicarbonate 30, HEPES 20, glucose 25, sodium thio sulfate 2, sodium ascorbate 5, sodium pyruvate 3, calcium chloride 2, magnesium sulfate 2] at room temperature. After 1 hour, the slices were transferred to a recording dish containing recording buffer [recording ACSF (mM): sodium chloride 119, potassium chloride 2.5, sodium phosphate monobasic 1.25, sodium bicarbonate 24, glucose 12.5, calcium chloride 2, magnesium sulfate 2]. Neuron-like cells with red fluorescent signals were recorded under a microscope (Olympus BX51WI) and the electrical activity of the transdifferentiated cells was captured using Clampex 10. 6-OHDA PD mouse model

[0057] The mice used in this experiment were adult C57BL / 6 mice (7-10 weeks). One half hour before anesthesia, 25 mg / kg of desipramine hydrochloride (D3900, Sigma-Aldrich) was injected intraperitoneally. After anesthesia, 3 pg of 6-OHDA (H116, Sigma-Aldrich) or saline was injected into the right medial forebrain bundle of the mice: anteroposterior (A / P) = -1.2 mm, mediolateral (M / L) = -1.1 mm, dorsoventral (D / V) = -5 mm. One hour after the surgery, 1 ml of 4% glucose-saline solution was injected subcutaneously into the mice. Apomorphine-induced rotation test

[0058] Mice were injected i.p. with 0.5 mg / kg of apomorphine (A4393, Sigma-Aldrich) 10 minutes before testing. During testing, mice were placed individually in an opaque cylinder (30 cm in diameter) and their behavior was recorded for 20 minutes above them by a video camera. A rotation was defined as a full body turn with one hind paw as a center and without a change in the head orientation. The number of rotations on the injected side and the contralateral side to the injection was counted and the data was quantified as the number of contralateral reversals in 20 minutes. Cylinder test

[0059] Mice were gently placed in a glass jar (1000 ml) and their behavior was recorded for 10 minutes in front of them by a video camera. The number of forepaw wall touches on the injected side and the contralateral side to the injection was counted separately and the data was quantified as the ratio of ipsilateral wall touches to the total wall touches. Rotarod test

[0060] Mice were first trained for 2 days and on the 3rd day, the behavioral test was performed. On the 1st day, mice were trained 4 times for 300 seconds at a constant speed of 4 revolutions per minute. On the 2nd day, mice were trained 4 times with an acceleration from 4 to 40 revolutions per minute. On the 3rd day, the behavioral test was performed with an acceleration from 4 to 40 revolutions per minute for 4 times. The time mice spent on the rod before falling off was recorded as the latency period and the mean of the 3 longest latency periods was used for analysis. NMDA modeling:

[0061] To investigate whether RGCs can be regenerated in damaged retina, 1.5 μΐ of NMDA solution at 200 mM prepared in PBS was injected into the eyes of 4-8 week old Ai9 mice or 5-6 week old C57BL / 6 mice (for VEP and black and white scene preference test) via intravitreal injection. GFAP-GFP-Cre was co-delivered with GFAP-CasRx-REST or GFAP-CasRx to the retina via subretinal injection 2-3 weeks after NMDA injection. To assess the functional rescue of damaged retina (VEP and light-dark box shuttle test), 5 weeks to 12 months old mice (C57BL / 6) were injected with NMDA to induce retinal damage and GFAP-mCherry (0.1 μΐ) mixed with PBS (0.9 μΐ) or GFAP-RZFD-V1 (0.9 μΐ) / GFAP-RZFD-V2 (0.9 μΐ) / GFAP-RZFD-V3 (0.9 μΐ) was injected subretinally 2-3 weeks after injection. Subretinal AAV injection

[0062] High titer (>1 x 1013vg / ml) AAV2 / 5 virus was injected into the eyes of 4-8 week old Ai9 mice or 5-6 week old C57BL / 6 mice (for VEP and black and white scene preference test) via intravitreal injection under an Olympus microscope (Olympus, Tokyo, Japan) with a Hamilton syringe (32G needle). 13AAV. To determine reprogramming in the whole retina, a total of 1 pl of GFAP-GFP-Cre (0.1 ul) + pbs (0.9 ul), or GFAP-GFP-Cre (0.1 pl) and GFAP-RZFD-V1 (0.9 pl) / GFAP-RZFD-V2 (0.9 pl) / GFAP-RZFD-V3 (0.9 pl) were injected subretinally (Ai9 and C57BL / 6 mice, 4 weeks to 12 months old). Retinal dissection and sectioning:

[0063] After 1-3 months of AAV injection, eyes, optic nerves and brain tissues were taken, fixed with 4% paraformaldehyde (PFA) for 2 hours (eyes and optic nerves) or 24 hours (brain), then dehydrated in 30% sucrose solution for 2 hours (eyes) or 24 hours (brain), and the optic nerves were directly washed with PBS and then mounted for observation. Then the eyes and brain were embedded with OCT and sectioned with a thickness of 30 pm. Retinal immunofluorescence staining:

[0064] Primary antibodies for immunofluorescence staining: rabbit anti-RBPMS (1:500, 15187-1-AP, Proteintech), mouse anti-Brn3a (1:100, MAB1585, Millipore), rabbit anti-Sox9 (1:500, AB5535, Millipore), rabbit anti-Proxl (1:500, AB5475, Millipore), rabbit anti-Pax6 (1:500, 901301, Biolegent). Secondary antibodies: Cy TM 5AffiniPure Donkey Mouse Anti-IgG (H+L) (1:500, 715-175-150, Jackson ImmunoResearch) and Cy TM5AffiniPure Donkey Rabbit Anti-IgG (H+L) (1:500, 711-175-152, Jackson ImmunoResearch). Primary antibodies were incubated at 4°C overnight, then washed with PBS for 3 times, 10 minutes each time. Secondary antibodies were incubated at room temperature for 2-3 hours, then washed with PBS for 3 times, 10 minutes each time. Finally, the slides were mounted with an anti-fluorescence quencher (Life technology), and imaging was performed using an Olympus FV3000 microscope. Retinal electrophysiological recording

[0065] Mice were dark-adapted overnight in a dark room the day before. Retinal dissections were performed under an infrared microscope in oxygenated (95% 02 / 5% C02) artificial cerebral spinal fluid (ACSF) containing 126 mM NaCl, 2.5 mM KCl, 1.25 mM NaH2P04, 2 mM CaCl2, 2 mM NaHC03, and 10 mM glucose. The RGC side of the retina was placed on the stage of an upright microscope facing the cell recording chamber. tdTomato-positive cells in the ganglion cell layer were identified using two-photon (λ = 1030 nm) microscopy and cell-attached recordings were made from them under infrared light. The electrodes used for recording (4-7 MΩ) were filled with ACSF and 0.25 mM Alexa 488 fluorescent dye. Recordings were made using a Multiclamp 700A amplifier and pClamp 10 software suite (Molecular Devices). Signals were low-pass filtered at 1 kHz and digitized at 10 kHz. Whole-field light stimuli were delivered with white LED light. After recording, fluorescent dye was injected into the cells with current pulses to mark the recorded cells. Mouse visual evoked potentials

[0066] Mice were anesthetized with an intraperitoneal injection of a mixture of fentanyl (0.05 mg / kg), midazolam (5 mg / kg), and medetomidine (0.5 mg / kg). The mice's heads were fixed in a stereotaxic position, and their body temperature was maintained at 37°C using a heating pad. A craniotomy (approximately 1 mm in diameter) was performed bilaterally above the visual cortex (V1) (AP -3.6 to -3.9 mm, ML 2.2 mm), and the dura was carefully removed. Visual stimuli were presented on a 17-inch LCD monitor (Dell P170S, maximum brightness 69 cd / m2), which was positioned 8 cm from the recording eye. The lateral side of the eye ipsilateral to the recording site was shielded from visual stimulation. One hundred 2-second repetitive flashes (full field of view, 100% contrast) were presented, separated by 2-second intervals. Recordings were made in V1 (AP -3.6 to -3.9 mm, ML 2.2 mm) using a multisite silicon probe (A1×16-5mm-50-177, NeuroNexus Technologies), with the electrode tip reaching a cortical depth of approximately 900 μm for each recording. Both the reference and ground wires were placed at least 3 mm from the recording site. Signals were amplified and neural responses were filtered using a Cerebus 32-channel system (Blackrock microsystems). Local field potential (LFP) signals were sampled at 2 kHz or 10 kHz using a broadband front-end filter (0.3–500 Hz). LFP responses to full-screen flash stimulation were used for current source density (CSD) analysis to determine the location of cortical layer 43. To generate CSD distribution profiles, the second-order spatial derivative of the LFP was calculated using the following equation:

[0067] in is the LFP, z is the coordinate of the recording tip, Δz is the distance between adjacent recording tips, and nΔz is the differentiation grid (n = 2). Layer 4 (stratum granulosum) was defined as the recording site at the initial current receptor. We used the layer 4 channel showing the largest mean amplitude to analyze the visually evoked response for each mouse. Black and white box preference test

[0068] The black-white preference test was conducted in a two-compartment chamber: one-third dark and two-thirds light (550-600 lumens). Mice were allowed to move freely between the two compartments, and a camera was used to record the time spent in the light and dark compartments for 10 minutes. The time spent in the light and dark compartments was counted and analyzed using Ethovision XT software. After each trial, the chamber was wiped with 70% ethanol to prevent olfactory interference. PD model monkey production

[0069] The cynomolgus monkeys used in this study were adult (7-10 years old) male cynomolgus monkeys. In order to obtain a stable animal model of Parkinson's disease, MPTP (1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine) was injected intravenously to establish the model. After one week of continuous injection, the animals began to show phenotypes, and the injection volume was adjusted according to the animal phenotype. As a result, a large number of dopamine neurons in the substantia nigra region of their brains apoptosis, and the drug administration was stopped after the disease phenotypes such as tremor, bradykinesia and impaired balance continued to appear. The pathological phenotype was continuously monitored, and AAV injection was performed after the pathological phenotype was stably shown for five weeks. Positron emission tomography experiment

[0070] To monitor the characteristics and function of newly converted dopamine neurons in the monkey brain, we performed PET imaging (using radionuclides including 18F-FP-CIT, 18F-DTBZ, and 18F-DOPA) one month before and one to six months after AAV administration. Prior to scanning, cynomolgus monkeys were induced anesthetized and then placed in a PET scanner, where all physiological parameters were monitored. Raw data from the scans were analyzed using standard methods. NMDA-induced retinal damage in cynomolgus monkeys

[0071] The cynomolgus monkeys were anesthetized and then a 200 mM NMDA solution was injected into the vitreous body using a 1 ml syringe to remove most of the RGCs. Within one week after the injection, the monkeys were given an appropriate amount of Diclofenac eye ointment daily to prevent infection or eye redness and swelling. Subretinal injection in cynomolgus monkeys:

[0072] First, the modeling monkeys were anesthetized and dilated 2-3 weeks after NMDA modeling, and then AAV injection was performed. A 1ml syringe was used to make a hole at the edge of the junction of the sclera and cornea, and then a 100ul Halilton flat-tip needle syringe was used to inject AAV. The virus mixture (100ul) of the experimental group or control group was slowly injected into the retina. After the injection, the needle was pulled out and eye ointment was applied. Diclofenac eye ointment was applied to the animals for 7 consecutive days after the injection to prevent eye infection or redness and swelling. Behavioral experiments and video analysis in cynomolgus monkeys

[0073] For video analysis, monkeys were behaviorally videotaped before and 1 to 12 months after drug administration. The videotapes were analyzed after collection. PD symptoms in the monkeys were analyzed using the following sub-items: head movement, facial expressions, voluntary movements, post-stimulus movements, tremors, body posture, and gait. Cynomolgus monkey tissue sampling and sectioning

[0074] For cynomolgus macaque brain tissue sectioning and staining, the monkeys were euthanized 1-6 months after AAV injection, and then perfused with saline and 4% PFA. The brain tissue was removed and placed in 4% PFA for fixation for 1 week. After fixation, the tissue was removed and placed in 30% sucrose solution for dehydration for 1-2 weeks. Then, the brain tissue was cut into appropriate sizes, embedded in OCT embedding medium, and sectioned on a freezing microtome with a thickness of 30-40 μm. For retinal sampling, similar to brain tissue sampling, the monkeys were euthanized approximately 1-6 months after AAV injection, and then perfused with saline and 4% PFA. The eyes, optic nerves, and brains were removed and fixed with 4% paraformaldehyde (PFA) for 1-2 weeks, and then dehydrated in 30% sucrose solution. After embedding, the eyes and brain were cut into 30-40 μm slices. Immunofluorescence staining of cynomolgus monkey tissues:

[0075] The staining procedure for cynomolgus macaque brain tissue is similar to that for mouse. After adding the primary antibody, incubate at 4°C for 24 hours. Wash with PBS 3-4 times for 15-20 minutes each. After washing, add the secondary antibody and incubate at room temperature for 6-8 hours. Wash with PBS 3-4 times for 15-20 minutes each. Then, mount the slides with anti-quenching mounting medium and store. The staining procedure for cynomolgus macaque retinal tissue is similar to that for mouse retinal tissue. First, wash with PBS three times for 10-15 minutes each, then stain in a humidified chamber. First, dilute the primary antibody in antibody diluent and add it to the slide, completely covering the slide. Stain at 4°C for 12-24 hours, then wash with PBS 3-4 times for 10-15 minutes each. Then, apply the secondary antibody, fully covering the tissue sample, and incubate at room temperature for 6-8 hours. Wash with PBS 3-4 times for 10-15 minutes each, and finally mount the slides with anti-quenching mounting medium. Statistical analysis:

[0076] Error bars are given by the mean square (SEM), and statistical significance (p < 0.05) was calculated using unpaired two-tailed t-tests or one-way analysis of variance. All experiments were randomized, and sample size was not predetermined using statistical methods. Normal distribution of the data was assumed but not formally tested. Data collection and analysis were not performed under blinded conditions. Examples Example 1. Construction of REST variants

[0077] In order to achieve the regulation of RE1, we took advantage of the fact that REST can bind to RE1 and explored the RE1 binding domain of its endogenous zinc finger protein (ZF) to regulate RE1, which is called RZFD (REST Zinc Finger Domain). Through domain prediction and protein structure modeling, we found that the 159-412 positions of human REST protein contain 8 zinc finger protein domains ( Figure 1 A). To further regulate the expression of neural-related proteins controlled by RE1, we fused a gene expression regulatory factor VP64 and P65-HSF1 to the C-terminus of RZFD to construct RZFD-VP64 and RZFD-P65-HSF ( Figure 1 B). VP64 and P65-HSF can recruit transcription factors and histone acetylation proteins to the vicinity of RE1, regulating the chromosome structure near RE1 and enabling the expression of genes it controls. In non-neurons (such as glial cells), the REST complex binds to RE1. Under the action of histone deacetylases and methyltransferases in the complex, the chromatin near RE1 is transformed into a dense state, and the expression of neuron-related genes is turned off ( Figure 1 C). We expressed RZFD, RZFD-VP64, or RZFD-P65-HSF1 in non-neuronal cells (e.g., glial cells). RZFD, RZFD-VP64, or RZFD-P65-HSF1 competitively bound to RE1, preventing the binding of the REST silencing complex to RE1 ( Figure 1 D, E, and F). The mechanism of RZFD is through competitive binding with RE1. RZFD-VP64 and RZFD-P65-HSF1 not only competitively bind to RE1 but also further alter the chromosomal state of RE1-adjacent regions, promoting the expression of neural-related genes regulated by it. Example 2. RZFD transdifferentiates astrocytes into neurons

[0078] To achieve the transdifferentiation of glial cells into neurons in mice, we constructed an AAV expression vector. To specifically label astrocytes, we used the astrocyte-specific promoter GFAP to drive mCherry expression. Human RZFD was also driven by the GFAP promoter and specifically expressed in astrocytes ( Figure 2 A). To test whether glial cells could be induced to transdifferentiate into neurons or dopamine neurons 2 weeks or more after injection, we collected samples 2 weeks or 1.5 months after AAV injection and performed NeuN and TH staining ( Figure 2B). Injection of GFAP-mCherry or GFAP-mCherry + GFAP-RZFD in the striatum or substantia nigra of 8-week-old C57 mice, where GFAP-mCherry alone is the control group Figure 2 C). Our previous studies have shown that there is no neuron production in the control group from 1 week after injection to 3 months after injection. Consistent with previous research results, we did not observe neuron production in the injection control group AAV2 weeks after injection Figure 2 D). While in the experimental group, we found that 2 weeks after injection, some cells have begun to deform and a small number of cells have begun to express neuron-specific marker NeuN Figure 2 E). This indicates that expressing RZFD in astrocytes can transdifferentiate astrocytes into neurons. After 1.5 months of analysis of injection of mixed AAV virus GFAP-mCherry + GFAP-RZFD, we found a large number of mCherry + NeuN + Double positive cells, which indicates that RZFD can efficiently transdifferentiate astrocytes into neurons. At the same time, we found that some mCherry positive cells express TH, which indicates that RZFD can transdifferentiate astrocytes into dopamine neurons Figure 2 F). Example 3. RZFD-VP64 and RZFD-P65-HSF1 transdifferentiate astrocytes into dopamine neurons

[0079] To further study the regulation of RE1 neuron-related gene expression, we constructed AAV expression vectors for RZFD-VP64 and RZFD-P65-HSF1 Figure 3 A). Express RZFD-VP64 and RZFD-P65-HSF1 under the control of the glial cell-specific promoter GFAP. We injected mixed AAV of GFAP-mCherry + GFAP-RZFD-VP64 into the striatum or substantia nigra of C57 mice, and analyzed 1.5 months after injection Figure 3 B). We found that most of the red fluorescently labeled cells in the GFAP-RZFD-VP64 injection group expressed the neuron-specific marker NeuN, and some cells also expressed the dopamine neuron-specific marker TH Figure 3 C). We performed similar AAV injections in the striatum or substantia nigra of DAT-Cre: Ai9 mice, injecting GFAP-RZFD-P65-HSF1 into the striatum or substantia nigra of mice, and analyzing 1.5 months later Figure 3D). In DAT-Cre:Ai9 mice, only dopamine neurons in Chengdu can be labeled, while our study found that there were no red fluorescent cells in the striatum of control mice. However, red cells were produced in the striatum of DAT-Cre:Ai9 mice injected with GFAP-RZFD-P65-HSF1. After NeuN and TH staining, we found that these cells with red fluorescent signals expressed the neuron-specific marker NeuN, and also expressed the dopamine neuron-specific marker TH ( Figure 3 E). This indicates that expressing RZFD-P65-HSF1 in glial cells can transdifferentiate astrocytes into dopamine neurons.

[0080] Sequence information Human RZFD amino acid sequence (SEQ ID NO.: 1) The amino acid sequence similarity between human and mouse RZFD is 95.49%. Human RZFD coding sequence (SEQ ID NO.: 2) NLS-RZFD-V1 amino acid sequence (SEQ ID NO.: 3) NLS-RZFD-V1 nucleotide sequence (SEQ ID NO.: 4) Amino acid sequence of NLS-RZFD-V2 (SEQ ID NO.: 5) Nucleotide sequence of NLS-RZFD-V2 (SEQ ID NO.: 6) VP64 amino acid sequence (SEQ ID NO.: 7) VP64 nucleic acid sequence (SEQ ID NO.: 8) NLS-RZFD-V3 amino acid sequence (SEQ ID NO.: 9) NLS-RZFD-V3 nucleotide sequence (SEQ ID NO.: 10) P65-HSF1 amino acid sequence (SEQ ID NO.: 11) P65-HSF1 nucleotide sequence (SEQ ID NO.: 12)

Claims

1. Use of a RE1 / NRSE element blocker for preparing a medicament for regulating neuron-related gene expression in non-neuronal cells, the RE1 / NRSE element blocker comprising a REST variant or its encoding nucleic acid, the REST variant comprising a human REST DNA binding domain and lacking the N-terminal and C-terminal inhibitory domains of human REST, the REST DNA binding domain being amino acids 159-412 of human REST.

2. Use of a RE1 / NRSE element blocker for the preparation of a medicament for preventing and / or treating diseases associated with neuronal dysfunction or death, the RE1 / NRSE element blocker comprising a REST variant or its encoding nucleic acid, the REST variant comprising the DNA binding domain of human REST and lacking the N-terminal and C-terminal inhibitory domains of human REST, the DNA binding domain of REST being amino acids 159-412 of human REST.

3. The use according to claim 1 or 2, wherein The DNA binding domain of REST is amino acids 155-420 of human REST; preferably, the DNA binding domain of REST is shown in SEQ ID NO.:

1.

4. The use according to any one of claims 1 to 3, wherein The RE1 / NRSE element blocker can reduce the binding of REST to the RE1 / NRSE element.

5. The use according to any one of claims 1 to 4, wherein the REST variant further comprises an activation domain, preferably, wherein the DNA binding domain of REST is fused to the activation domain.

6. The use according to claim 5, wherein the activation domain is selected from epigenetic modification proteins or gene activation regulatory elements, preferably, the activation domain is selected from VP64, P65-HSF1, VP16, RTA, Suntag, P300, CBP or a combination thereof.

7. The use according to claim 6, wherein the VP64 comprises an amino acid sequence that is at least 90% identical to SEQ ID NO.: 7, or comprises a nucleotide sequence that is at least 90% identical to SEQ ID NO.: 8; and the P65-HSF1 comprises an amino acid sequence that is at least 90% identical to SEQ ID NO.: 11, or comprises a nucleotide sequence that is at least 90% identical to SEQ ID NO.:

12.

8. The method according to any one of claims 1 to 7, wherein the REST variant comprises a sequence having at least 70%, 60%, or 50% identity with the amino acid sequence shown in SEQ ID NO.: 1, 3, 5, or 9 or the nucleotide sequence shown in SEQ ID NO.: 2, 4, 6, or 10.

9. The method of claim 1 , wherein the non-neuronal cells comprise glial cells, fibroblasts, stem cells, neural progenitor cells, neural stem cells, wherein the glial cells are selected from astrocytes, oligodendrocytes, ependymal cells, Schwann cells, NG2 cells, satellite cells, Muller glial cells, inner ear glial cells, or a combination thereof; preferably astrocytes, Muller glial cells, and cochlear glial cells.

10. purposes according to claim 9, wherein said glial cell is derived from brain, spinal cord, eye or ear, and wherein the glial cell of brain is derived from striatum, substantia nigra, ventral tegmental area of ​​midbrain, spinal cord, hypothalamus, dorsal midbrain or cerebral cortex.

Images