CREB3 for the treatment or the prevention of amyotrophic lateral sclerosis

Abstract

The present invention relates to a composition comprising a c-AMP Response Element-Binding protein 3 (CREB3), in particular a variant of CREB3 protein, or a nucleic acid molecule encoding said CREB3 protein or said variant of CREB3 protein, for use as a medicament, in particular for treating or preventing diseases associated with a degeneration of motor neurons. The present invention further relates to CREB3 protein, in particular said variant of CREB3 protein, as a biomarker for the stratification or the prognosis of patients suffering, or susceptible of suffering, from a disease associated with a degeneration of motor neurons.

Description

[0001] CREB3 FOR THE TREATMENT OR THE PREVENTION OF A DISEASE ASSOCIATED WITH A DEGENERATION OF MOTOR NEURONS

[0002] The present invention relates to a composition comprising a c-AMP Response Element-Binding protein 3 (CREB3), in particular a variant of CREB3 protein, or a nucleic acid molecule encoding said CREB3 protein or said variant of CREB3 protein, for use as a medicament, in particular for treating or preventing diseases associated with a degeneration of motor neurons. The present invention further relates to CREB3 protein, in particular said variant of CREB3 protein, as a biomarker for the stratification or the prognosis of patients suffering, or susceptible of suffering, from a disease associated with a degeneration of motor neurons.

[0003] Motor neurons control skeletal muscle activity such as walking, breathing, speaking, and swallowing. The premature degeneration of motor neurons is associated with numerous specific disorders characterized by a progressive muscular weakness. Amyotrophic lateral sclerosis (ALS or Charcot's disease, or Lou Gherig's disease) is the most common disease associated with a degeneration of motor neurons in adults. Management of this disease is difficult and its psychological, societal and economic impacts are major. ALS is a severely debilitating neurodegenerative disease that manifests between the 6thand 8thdecade of life as a progressive muscular weakness and paralysis leading to death when respiratory muscles become impaired, within only two to three years upon symptom onset. The vast majority of cases are sporadic (sALS), but 5 to 10% of patients have a familial history (fALS). Causative mutations have been identified in about two thirds of fALS cases, involving at least 40 genes, the most represented being C90RF72, SOD1, TARDBP and FUS.

[0004] No treatment currently exists to treat ALS. Riluzole, an anti-glutamatergic, is the only treatment approved worldwide, but its effects are extremely modest, since it is estimated to delay the death of patients by a few months. Tofersen, an antisense oligonucleotide indicated in patients with SOD1 gene mutations, has very promising effects and recently obtained compassionate use authorization in France. However, it is only indicated in around 10% of family cases, or around 1% of total cases. Other gene therapies are being developed for familial cases which have to date proven to be promising (FUS gene, 4% of familial cases, 0.4% of total cases) or deleterious (C9ORF72 gene, 30% of familial cases, 3% of total cases).

[0005] In this context, there remains a need for new means for improving the early prognosis and management of patients suffering from diseases associated with a degeneration of motor neurons, in particular ALS. Surprisingly, the Inventors have identified the CREB3 transcription factor as a resilience marker of neuronal dysfunction in ALS and demonstrated that a missense rare variant of CREB3 protein confers a significant reduction in the risk of developing ALS and is associated with a slower disease progression rate in ALS patients, through gain of function mechanisms. They also showed that the overexpression of CREB3 protein confers a significant reduction in the risk of developing ALS and is associated with a slower disease progression rate in ALS patients. While numerous genetic disease modifiers, such as ATXN2, C9ORF72, FUS or UNC13A have been previously reported, it is noteworthy that identified variants were systematically associated with a negative outcome, with increased susceptibility or decreased survival. In this context, it is striking to note that the present CREB3 variant stands out in an impressive way, with a unique ability to both reduce the risk of developing ALS by ~40%, and to increase the lifespan of ALS patients by 12 months. CREB3 stands out as one of the first genetic modifiers with positive effects reported to date in the field of ALS. In addition, the Inventors also provide the use of CREB3 as a valuable tool to prevent disease onset or slow-down disease progression.

[0006] In a main aspect, the invention thus relates to a composition comprising a c-AMP Response Element-Binding protein 3 (CREB3) comprising or consisting of an amino acid sequence having at least 80% sequence identity with SEQ. ID NO: 1 , or a nucleic acid molecule encoding said CREB3 protein, for use as a medicament.

[0007] The various features of the present invention referred to any individual aspects below apply, as appropriate, to other aspects mutatis mutandis. Consequently, features specified in a given aspect of the invention may be combined with features specified in other aspects as appropriate.

[0008] In the present invention, the expression "c-AMP Response Element-Binding protein 3 " or "CREB3" designates a transcription factor which is a member of the leucine zipper family of DNA binding proteins. This protein binds to the cAMP-response element and regulates cell proliferation. The CREB3 protein is also known as "LUMAN", "LZIP" or "sLZIP".

[0009] In the present invention, the term "CREB3" designates not only the full-length CREB3 protein but also any protein comprising at least the first 267 amino acids of the full-length CREB3 protein. This region corresponds to the N-terminal domain of the protein, which is released by proteases and acts as a transcription factor.

[0010] Some reference sequences are provided in Table 1.

[0011] Table 1. Reference sequences of the CREB3 protein and corresponding nucleic acid molecules. Sequence of the cleavage site is underlined.

[0012] In an embodiment, the CREB3 protein comprises or consists of an amino acid sequence having at least 90%, preferably 95%, more preferably at least 99% amino acid sequence identity to the amino acid sequence SEQ ID NO: 1.

[0013] In an embodiment, the CREB3 protein comprises or consists of an amino acid sequence having at least 80% amino acid sequence identity to the amino acid sequence SEQ ID NO: 2.

[0014] In an embodiment, the CREB3 protein comprises or consists of an amino acid sequence having at least 90%, preferably 95%, more preferably at least 99% amino acid sequence identity to the amino acid sequence SEQ ID NO: 2.

[0015] In an embodiment, the CREB3 protein comprises or consists of an amino acid sequence having at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% amino acid sequence identity to the amino acid sequence SEQ ID NO: 1.

[0016] In an embodiment, the CREB3 protein comprises or consists of the amino acid sequence SEQ ID NO: 1.

[0017] In an embodiment, the CREB3 protein comprises or consists of an amino acid sequence having at least 80%, 81%, 82%, 83%, 84%, 85%, 86%, 87%, 88%, 89%, 90%, 91%, 92%, 93%, 94%, 95%, 96%, 97%, 98% or 99% amino acid sequence identity to the amino acid sequence SEQ ID NO: 2.

[0018] In an embodiment, the CREB3 protein comprises or consists of the amino acid sequence SEQ ID NO: 2.

[0019] In an embodiment, the nucleic acid molecule encoding the CREB3 protein comprises or consists of a nucleic acid sequence having at least 80% identity with the nucleic acid sequence SEQ ID NO: 3. In an embodiment, the nucleic acid molecule encoding the CREB3 protein comprises or consists of the nucleic acid sequence SEQ ID NO: 3.

[0020] For the purpose of comparing two closely-related sequences, the "% identity" between a first sequence and a second sequence may be calculated using an alignment program, such as BLAST® (available at blast.ncbi.nlm.nih.gov) using standard settings. The % identity is the number of identical residues divided by the number of residues in the reference sequence, multiplied by 100. The % identity referred to above and in the claims are percentages calculated by this methodology.

[0021] In an embodiment, the invention relates to a composition comprising a c-AMP Response Element- Binding protein 3 (CREB3) comprising or consisting of an amino acid sequence having at least 80% sequence identity with SEQ ID NO: 1 and having at least one mutation with respect to SEQ ID NO: 1, or a nucleic acid molecule encoding said CREB3 protein, for use as a medicament.

[0022] In an embodiment, the invention relates to a composition for use as defined above, wherein said mutation is located in position 119 with respect to SEQ ID NO: 1.

[0023] In an embodiment, the invention relates to a composition for use as defined above, wherein said mutation is located in position 119 with respect to SEQ ID NO: 2.

[0024] In an embodiment, the invention relates to a composition for use as defined above, wherein said mutation is a substitution, more preferably a substitution R119G with respect to SEQ ID NO: 1 or SEQ ID NO: 2.

[0025] As used herein, the term "R119G" indicates that the Arginine (R) residue located in position 119 of SEQ ID NO: 1 or SEQ ID NO: 2 has been substituted by a Glycine (G) residue.

[0026] The amino acid sequence of the R119G CREB3 variant is provided in Table 2.

[0027] Table 2. Reference sequences of the R119G CREB3 variant. The R119G position of the mutation and the sequence of the cleavage site are underlined.

[0028] In an embodiment, the invention relates to a composition for use as defined above, wherein said CREB3 protein comprises or consists of the amino acid sequence SEQ ID NO: 4. In an embodiment, the invention relates to a composition for use as defined above, wherein said CREB3 protein comprises or consists of the amino acid sequence SEQ ID NO: 5.

[0029] In an embodiment, the invention relates to a composition for use as defined above, comprising a nucleic acid molecule encoding the mutated CREB3 protein as defined above. In an embodiment, the invention relates to a composition for use as defined above, wherein the nucleic acid molecule encoding the CREB3 protein comprises or consists of a nucleic acid sequence having at least 80% identity with the nucleic acid sequence SEQ ID NO: 6.

[0030] In an embodiment, the invention relates to a composition for use as defined above, wherein the nucleic acid molecule encoding the CREB3 protein comprises or consists of the nucleic acid sequence SEQ ID NO: 6.

[0031] In an embodiment, the invention relates to a composition for use as defined above, wherein said mutation is located in position 120 with respect to SEQ ID NO: 1.

[0032] In an embodiment, the invention relates to a composition for use as defined above, wherein said mutation is located in position 120 with respect to SEQ ID NO: 2.

[0033] In an embodiment, the invention relates to a composition for use as defined above, wherein said mutation is a substitution, more preferably a substitution L120P with respect to SEQ ID NO: 1 or SEQ ID NO: 2.

[0034] As used herein, the term "L120P" indicates that the Leucine (L) residue located in position 120 of SEQ ID NO: 1 or SEQ ID NO: 2 has been substituted by a Proline (P) residue.

[0035] In an embodiment, the invention relates to a composition for use as defined above, wherein said mutation is located in position 132 with respect to SEQ ID NO: 1.

[0036] In an embodiment, the invention relates to a composition for use as defined above, wherein said mutation is located in position 132 with respect to SEQ ID NO: 2.

[0037] In an embodiment, the invention relates to a composition for use as defined above, wherein said mutation is a substitution, more preferably a substitution K132E with respect to SEQ ID NO: 1 or SEQ ID NO: 2.

[0038] As used herein, the term "K132E" indicates that the Lysine (K) residue located in position 132 of SEQ ID NO: 1 or SEQ ID NO: 2 has been substituted by a Glutamic Acid (E) residue.

[0039] In an embodiment, the invention relates to a composition for use as defined above, wherein said mutation is located in position 155 with respect to SEQ ID NO: 1.

[0040] In an embodiment, the invention relates to a composition for use as defined above, wherein said mutation is located in position 155 with respect to SEQ ID NO: 2. In an embodiment, the invention relates to a composition for use as defined above, wherein said mutation is a substitution, more preferably a substitution R155W with respect to SEQ ID NO: 1 or SEQ ID NO: 2.

[0041] As used herein, the term "R155W" indicates that the Arginine (R) residue located in position 155 of SEQ ID NO: 1 or SEQ ID NO: 2 has been substituted by a Tryptophan (W) residue.

[0042] In an embodiment, the invention relates to a composition for use as defined above, wherein said mutation is located in position 168 with respect to SEQ ID NO: 1.

[0043] In an embodiment, the invention relates to a composition for use as defined above, wherein said mutation is located in position 168 with respect to SEQ ID NO: 2.

[0044] In an embodiment, the invention relates to a composition for use as defined above, wherein said mutation is a substitution, more preferably a substitution R168H with respect to SEQ ID NO: 1 or SEQ ID NO: 2.

[0045] As used herein, the term "R168H" indicates that the Arginine (R) residue located in position 168 of SEQ ID NO: 1 or SEQ ID NO: 2 has been substituted by a Histidine (H) residue.

[0046] In an embodiment, the invention relates to a composition for use as defined above, wherein said mutation is located in position 172 with respect to SEQ ID NO: 1.

[0047] In an embodiment, the invention relates to a composition for use as defined above, wherein said mutation is located in position 172 with respect to SEQ ID NO: 2.

[0048] In an embodiment, the invention relates to a composition for use as defined above, wherein said mutation is a substitution, more preferably a substitution K172E with respect to SEQ ID NO: 1 or SEQ ID NO: 2.

[0049] As used herein, the term "K172E" indicates that the Lysine (K) residue located in position 172 of SEQ ID NO: 1 or SEQ ID NO: 2 has been substituted by a Glutamic Acid (E) residue.

[0050] In an embodiment, the invention relates to a composition for use as defined above, wherein said mutation is located in position 346 with respect to SEQ ID NO: 1.

[0051] In an embodiment, the invention relates to a composition for use as defined above, wherein said mutation is located in position 346 with respect to SEQ ID NO: 2.

[0052] In an embodiment, the invention relates to a composition for use as defined above, wherein said mutation is a substitution, more preferably a substitution Q346G with respect to SEQ ID NO: 1 or SEQ ID NO: 2. As used herein, the term "Q346G" indicates that the Glutamine (Q) residue located in position 346 of SEQ ID NO: 1 or SEQ ID NO: 2 has been substituted by a Glycine (G) residue.

[0053] In an embodiment, the composition for use as defined above is a pharmaceutical composition which further comprises at least one compound selected from the group consisting of a pharmaceutically acceptable carrier, diluent and excipient.

[0054] In another aspect, the invention relates to a composition as defined above for use in the prevention or the treatment of a disease associated with a degeneration of motor neurons.

[0055] In an embodiment, the invention relates to a composition comprising a CREB3 protein comprising or consisting of an amino acid sequence having at least 80% sequence identity with SEQ ID NO: 1, or a nucleic acid molecule encoding said CREB3 protein, for use in the prevention or the treatment of a disease associated with a degeneration of motor neurons.

[0056] In an embodiment, the invention relates to a composition comprising a CREB3 protein comprising or consisting of an amino acid sequence having at least 80% sequence identity with SEQ ID NO: 1 and having at least one mutation with respect to SEQ ID NO: 1, or a nucleic acid molecule encoding said CREB3 protein, for use in the prevention or the treatment of a disease associated with a degeneration of motor neurons.

[0057] In another aspect, the invention relates to a method of preventing or treating a disease associated with a degeneration of motor neurons by means of administration, to a patient in need thereof, of an effective amount of a composition as defined above.

[0058] In an embodiment, the invention relates to a method of preventing or treating a disease associated with a degeneration of motor neurons by means of administration, to a patient in need thereof, of an effective amount of a composition comprising a CREB3 protein comprising or consisting of an amino acid sequence having at least 80% sequence identity with SEQ ID NO: 1, or a nucleic acid molecule encoding said CREB3 protein.

[0059] In an embodiment, the invention relates to a method of preventing or treating a disease associated with a degeneration of motor neurons by means of administration, to a patient in need thereof, of an effective amount of a composition comprising a CREB3 protein comprising or consisting of an amino acid sequence having at least 80% sequence identity with SEQ ID NO: 1 and having at least one mutation with respect to SEQ ID NO: 1, or a nucleic acid molecule encoding said CREB3 protein.

[0060] In another aspect, the invention relates to the use of a composition as defined above, in the manufacture of a medicament for the prevention or the treatment of a disease associated with a degeneration of motor neurons. In an embodiment, the invention relates to the use of a composition comprising a CREB3 protein comprising or consisting of an amino acid sequence having at least 80% sequence identity with SEQ ID NO: 1, or a nucleic acid molecule encoding said CREB3 protein, in the manufacture of a medicament for the prevention or the treatment of a disease associated with a degeneration of motor neurons.

[0061] In an embodiment, the invention relates to the use of a composition comprising a CREB3 protein comprising or consisting of an amino acid sequence having at least 80% sequence identity with SEQ ID NO: 1 and having at least one mutation with respect to SEQ ID NO: 1, or a nucleic acid molecule encoding said CREB3 protein, in the manufacture of a medicament for the prevention or the treatment of a disease associated with a degeneration of motor neurons.

[0062] In the present invention, the expression "a disease associated with a degeneration of motor neurons" designates any disease that involves a degeneration of motor neurons in the neocortex, brain stem and / or spinal cord.

[0063] As used herein, the expression "a disease associated with a degeneration of motor neurons" includes, but is not limited to, amyotrophic lateral sclerosis (ALS), primary lateral sclerosis (PLS), hereditary spastic paraplegia (HSP), progressive bulbar palsy (PBP), spinal muscular atrophy (SMA), X-linked spinobulbar muscular atrophy (SBMA; Kennedy disease), spinal muscular atrophy with respiratory distress type 1 (SMARD1), postpolio syndrome (PPS), congenital SMA with arthrogryposis, progressive muscular atrophy (PMA), Alzheimer's disease, Parkinson's disease, Huntington's disease and Spinocerebellar Ataxia (SCA).

[0064] In an embodiment, the invention relates to a composition for use as defined above, wherein the disease associated with a degeneration of motor neurons is selected from the group consisting of ALS, PLS, HSP, PBP, SMA, SBMA, SMARD1, PPS, congenital SMA with arthrogryposis, PMA, Alzheimer's disease, Parkinson's disease, Huntington's disease and SCA.

[0065] In an embodiment, the invention relates to a composition for use as defined above, wherein the disease associated with a degeneration of motor neurons is selected from the group consisting of ALS, PLS, HSP, PBP, SMA, SBMA, SMARD1, PPS, congenital SMA with arthrogryposis, PMA, Parkinson's disease, Huntington's disease and SCA.

[0066] In an embodiment, the invention relates to a composition for use as defined above, wherein the disease associated with a degeneration of motor neurons is selected from the group consisting of ALS, PLS, HSP, PBP, SMA, SBMA, SMARD1, PPS, congenital SMA with arthrogryposis and PMA. In an embodiment, the invention relates to a composition for use as defined above, wherein the disease associated with a degeneration of motor neurons is a motor neurons disease (MND).

[0067] As used herein, the expression "motor neurons disease" (or "MND") refers to a commonly accepted group of rare neurodegenerative disorders that selectively affect motor neurons. ALS, PLS, HSP, PBP, SMA, SBMA, SMARD1, PPS, congenital SMA with arthrogryposis and PMA are all subtypes of motor neurons diseases.

[0068] In an embodiment, the invention relates to a composition for use as defined above, wherein the MND is selected from the group consisting of ALS, PLS, HSP, PBP, SMA, SBMA, SMARD1, PPS, congenital SMA with arthrogryposis and PMA, preferably ALS.

[0069] In some embodiments, a nucleic acid molecule encoding said CREB3 protein is particularly appropriate for gene therapy, wherein the objective is to inject the organism with a genetic construct that provides transcription and further translation of the therapeutic protein encoded by such genetic construct.

[0070] In an embodiment, the invention relates to a composition for use as defined above, wherein said nucleic acid molecule is a genetic vector, preferably a viral vector.

[0071] In an embodiment, said viral vector is selected from the group comprising an adeno-associated virus (AAV), an adenovirus vector, a lentivirus vector and a retrovirus vector, preferably an AAV vector.

[0072] In an embodiment, said nucleic acid molecule comprises or consists of an expression cassette comprising a promoter, preferably a neuron specific promoter.

[0073] In an embodiment, said vector can efficiently transduce the central and / or peripheral nervous systems.

[0074] In an embodiment, the invention relates to a composition for use as defined above, wherein said nucleic acid molecule comprises or consists of a complementary DNA (cDNA), a nucleoside- modified messenger RNA (modRNA) or a stabilized mRNA.

[0075] In an embodiment, the nucleic acid molecule is capable of constituting a CREB3 polypeptide messenger RNA in a target cell, preferably a neuron, of a human subject to whom the composition is administered.

[0076] In an embodiment, the nucleic acid molecule is present in a genetically engineered cell encoding said CREB3 protein. In an embodiment, the invention relates to a composition for use as defined above, wherein said composition is administered by intracerebroventricular injection, intrathecal injection or intravenous injection.

[0077] In an embodiment, the invention concerns a nucleic acid according to the invention, an expression cassette according to the invention, or an expression vector according to the invention, for use as a drug.

[0078] The invention also concerns, in another aspect, a pharmaceutical composition comprising a nucleic acid according to the invention, an expression cassette according to the invention, or an expression vector according to the invention.

[0079] In an aspect, the invention also concerns a nucleic acid according to the invention, an expression cassette according to the invention, an expression vector according to the invention, or a pharmaceutical composition according to the invention for use in the treatment of neurodegenerative diseases in a patient in needs thereof.

[0080] In an aspect, the invention also concerns a nucleic acid according to the invention, an expression cassette according to the invention, an expression vector according to the invention, or a pharmaceutical composition according to the invention for use in the treatment of ALS patient.

[0081] In another aspect, the invention relates to the use of a CREB3 protein comprising or consisting of an amino acid sequence having at least 80% sequence identity with SEQ ID NO: 1 as a biomarker for the in vitro / ex vivo stratification of patients suffering, or susceptible of suffering, from a disease associated with a degeneration of motor neurons, wherein the presence of the expression of said CREB3 protein in biological samples of said patients indicates that said patients belong to a group of patients with a reduced risk of developing said disease and / or a reduced rate of progression of said disease, and the absence of the expression of said CREB3 protein in biological samples of said patients indicates that said patients belong to a group of patients with an increased risk of developing said disease and / or an increased rate of progression of said disease.

[0082] In an embodiment, the invention relates to the use of a CREB3 protein comprising or consisting of an amino acid sequence having at least 80% sequence identity with SEQ ID NO: 1 and having at least one mutation with respect to SEQ ID NO: 1 as a biomarker for the in vitro / ex vivo stratification of patients suffering, or susceptible of suffering, from a disease associated with a degeneration of motor neurons, wherein the presence of the expression of said CREB3 protein in biological samples of said patients indicates that said patients belong to a group of patients with a reduced risk of developing said disease and / or a reduced rate of progression of said disease, and the absence of the expression of said CREB3 protein in biological samples of said patients indicates that said patients belong to a group of patients with an increased risk of developing said disease and / or an increased rate of progression of said disease

[0083] As used herein, the expression "biological sample" designates any sample containing proteins or nucleic acids, preferably a sample which contains cells derived from a subject. In particular, the expression "biological sample" may also refer to any sample containing free circulating nucleic acids. The sample may be treated prior to its use, e.g. extracted, purified, concentrated, frozen, etc.

[0084] In an embodiment, the biological sample is selected from the group consisting of blood, serum, plasma, cerebrospinal fluid, lymph, biopsies, organs, tissues, cell samples, preferably blood and cerebrospinal fluid.

[0085] In an embodiment, the level of expression of CREB3 is measured. The level of expression can be measured by any method allowing the quantification of the levels of protein or mRNA. Such methods are well known by the skilled person in the art, e.g. ELISA, RT-qPCR, NGS.

[0086] As used herein, the expression "a reduced risk of developing a disease associated with a degeneration of motor neurons" means that the patient has at least 10%, preferably at least 20%, more preferably at least 40% reduction in the risk of developing a disease associated with a degeneration of motor neurons in comparison with a patient that does not express CREB3 protein.

[0087] As used herein, the expression "an increased reduced risk of developing a disease associated with a degeneration of motor neurons" means that the patient has at least 10%, preferably at least 20%, more preferably at least 40% increase in the risk of developing a disease associated with a degeneration of motor neurons in comparison with a patient that expresses CREB3 protein.

[0088] As used herein, the expression "a reduced rate of progression of a disease associated with a degeneration of motor neurons" means that the patient has an average lifespan extended by at least 6 months, preferably at least 9 months, more preferably at least 12 months in comparison with a patient that does not express CREB3 protein.

[0089] As used herein, the expression "an increased rate of progression of a disease associated with a degeneration of motor neurons" means that the patient has an average lifespan reduced by at least 6 months, preferably at least 9 months, more preferably at least 12 months in comparison with a patient that expresses CREB3 protein. In another aspect, the invention relates to an in vitro / ex vivo method for the prognosis of a patient suffering from a disease associated with a degeneration of motor neurons comprising a step of detecting the presence or the absence of the expression of a CREB3 protein comprising or consisting of an amino acid sequence having at least 80% sequence identity with SEQ ID NO: 1 in a biological sample from said patient, wherein the presence of the expression of said CREB3 protein in the biological sample of said patient indicates that said patient has a reduced rate of progression of said disease, and the absence of the expression of said CREB3 protein in the biological sample of said patient indicates that said patient has an increased rate of progression of said disease.

[0090] In an embodiment, the invention relates to an in vitro / ex vivo method for the prognosis of a patient suffering from a disease associated with a degeneration of motor neurons comprising a step of detecting the presence or the absence of the expression of a CREB3 protein comprising or consisting of an amino acid sequence having at least 80% sequence identity with SEQ ID NO: 1 and having at least one mutation with respect to SEQ ID NO: 1 in a biological sample from said patient, wherein the presence of the expression of said CREB3 protein in the biological sample of said patient indicates that said patient has a reduced rate of progression of said disease, and the absence of the expression of said CREB3 protein in the biological sample of said patient indicates that said patient has an increased rate of progression of said disease.

[0091] In an aspect, the present invention concerns a nucleic acid molecule encoding a CREB3 protein as defined above.

[0092] In an embodiment, the present invention concerns a nucleic acid molecule encoding a CREB3 protein comprising or consisting of an amino acid sequence having at least 80% sequence identity with SEQ ID NO: 1.

[0093] In an embodiment, the invention relates to a nucleic acid molecule encoding a CREB3 protein comprising or consisting of an amino acid sequence having at least 80% sequence identity with SEQ ID NO: 2.

[0094] In an embodiment, the nucleic acid molecule comprises or consists of a nucleic acid sequence having at least 80 % identity with SEQ ID NO: 3.

[0095] In an embodiment, the nucleic acid molecule comprises or consists of the nucleic acid sequence SEQ ID NO: 3. In an embodiment, the present invention concerns a nucleic acid molecule encoding a CREB3 protein comprising or consisting of an amino acid sequence having at least 80% sequence identity with SEQ ID NO: 1 and having at least one mutation with respect to SEQ ID NO: 1.

[0096] In an embodiment, the invention relates to a nucleic acid molecule encoding a CREB3 protein comprising or consisting of an amino acid sequence having at least 80% sequence identity with SEQ ID NO: 2 and having at least one mutation with respect to SEQ ID NO: 2.

[0097] In an embodiment, said mutation is located in position 119 with respect to SEQ ID NO: 1.

[0098] In an embodiment, said mutation is located in position 119 with respect to SEQ ID NO: 2.

[0099] In an embodiment, the invention relates to a nucleic acid molecule encoding a CREB3 protein comprising or consisting of an amino acid sequence having at least 80% sequence identity with SEQ ID NO: 4.

[0100] In an embodiment, the invention relates to a nucleic acid molecule encoding a CREB3 protein comprising or consisting of an amino acid sequence having at least 80% sequence identity with SEQ ID NO: 5.

[0101] In an embodiment, the nucleic acid molecule comprises or consists of a nucleic acid sequence having at least 80 % identity with SEQ ID NO: 6.

[0102] In an embodiment, the nucleic acid molecule comprises or consists of the nucleic acid sequence SEQ ID NO: 6.

[0103] In an embodiment, the nucleic acid molecule comprises or consists of an expression cassette.

[0104] In an embodiment, the nucleic acid molecule comprises or consists of an expression cassette comprising a neuron specific promoter.

[0105] In an embodiment, the nucleic acid molecule is a genetic vector, preferably a viral vector.

[0106] In an embodiment, said viral vector is selected from the group comprising an adeno-associated virus (AAV), an adenovirus vector, a lentivirus vector and a retrovirus vector, preferably an AAV vector.

[0107] In an embodiment, the nucleic acid molecule is a genetic vector, preferably an AAV, comprising the nucleic acid sequence SEQ ID NO: 3.

[0108] In an embodiment, the nucleic acid molecule is an AAV comprising a neuron specific promoter and the nucleic acid sequence SEQ ID NO: 3. In an embodiment, the nucleic acid molecule is a genetic vector, preferably an AAV, comprising the nucleic acid sequence SEQ ID NO: 6.

[0109] In an embodiment, the nucleic acid molecule is an AAV comprising a neuron specific promoter and the nucleic acid sequence SEQ ID NO: 6. The following Examples are put forth so as to provide those of ordinary skill in the art with a complete disclosure and description of how to make and use the present invention and are not intended to limit the scope of what the Inventors regard as their invention nor are they intended to represent that the experiments below are all or the only experiments performed. While the present invention has been described with reference to the specific embodiments thereof, it should be understood by those skilled in the art that various changes may be made and equivalents may be substituted without departing from the true spirit and scope of the invention. In addition, many modifications may be made to adapt a particular situation, material, composition of matter, process, process step or steps, to the objective, spirit and scope of the present invention. All such modifications are intended to be within the scope of the claims appended hereto.

[0110] FIGURES

[0111] Figure 1. Cross-species transcriptomic analysis prioritizes vulnerable cell-types in ALS a. Experimental design displaying integration of single nuclei RNAseq from ALS patients and healthy controls, the human motor cortex and the database of mouse retrogradelly labelled CNS and control cells (mCSN and mCtl cells) from WT and SodlG86Rmice. b. T-SNE plot of the 116 clusters identified in the human Ml motor cortex (left) and expression of the mCNS and mCtl cells enriched genes in each cluster (right), c. Experimental design of the cross-species differentially expressed gene (DEG) enrichment analysis wherein DEG identified in mCSN were intersected with each of the 30 DEG cell populations in human ALS, FTD and healthy controls, d. Cluster dendrogram and heatmaps showing enrichment of mCSN and mCtl cells geneset in each of the 30 DE cell populations as well as enrichment of mCSN DEG in all DE cell populations. A significant enrichment of the mCSN DEG was observed in human excitatory neurons (Hypergeometric test: Bonferroni's adjusted *p < 4e 04). Based on mCSN geneset enrichment and cross-species DE analysis, a cumulative score was calculated for each DE group and ranked to prioritize the most affected cell populations in ALS patients.

[0112] Figure 2. Cell-type specific consensus WGCNA identifies conserved regulatory networks a. Experimental design showing the cross-species consensus WGCNA approach used on the prioritized cell types identified in Figure 2. b. Histogram showing conserved signed correlation of twenty WGCNA modules to ALS in mCSN and human L2 / 3-IT1 and L5-ET. Correlation p-value for each module were corrected using the Benjamini-Hochberg method. Dashed line represents WGCNA module with signed correlation FDR < 0.05. c. UMAP gene network of the lightyellow, turquoise, pink, blue and darkgrey modules and 2 hub genes per module. Network shows intramodule connectivity with each dot representing individual genes. The size of the dot is proportional to the importance of each gene (signed kME) in the network, d. Dot plot showing cross-species conservation of significant modules, with Zscore > 10 (turquoise and blue modules) considered to have strong preservation and 5 < Zscore < 10 to have moderate preservation (darkred, lightyellow and pink modules), e. GO-term enrichment analysis in the significant turquoise module with genes having signed kME (>0.6) in the network. Histograms represent significant terms above the dashed line (FDR < 0.05). f. Violin plot showing module eigengenes expression of the turquoise module in human L2 / 3-IT1 and L5-ET and mCNS (ANOVA: Finteraction group*cell =4.88, **p<0.01) with a significantly increased expression in patients' L2-3-IT1 neurons (Tukey-multiple comparisons: *p- adjusted=0.037) and SodlG86RmCNS neurons (Tukey-multiple comparisons : ***p-adjusted<0.001) compared to healthy controls or WT animals respectively, and a trend to an increase in patients' L5-ET (Tukey-multiple comparisons: p>0.05). g. Violin plot showing module eigengenes expression of the turquoise module in each of the seven major cell classes. The turquoise module is only significantly increased in excitatory neurons (ANOVA: FgrOup= 26.67, p<0.0001, Tukey post-hoc: *p- adjusted = 0.035). h. GO-term enrichment analysis in the significant lightyellow module with genes having signed kME (>0.6) in the network. Histograms represent significant terms above the red dashed line (FDR < 0.05). i. Violin plot showing module eigengenes expression of the lightyellow module in all three cell types (ANOVA: Finteraction group*cell=5.007, **p<0.01) with significantly increased expression in patients' L2-3-IT1 neurons (Tukey-multiple comparisons: ***p- adjusted<0.001) and SodlG86RmCSN (Tukey-multiple comparisons: ***p-adjusted < 0.001) compared to healthy controls and WT mice respectively, and a trend to an increase in patients' L5- ET (Tukey-multiple comparisons : p>0.05). j. Violin plot showing module eigengenes expression of the lightyellow module in each of the seven major cell classes. The lightyellow module is significantly increased in both the excitatory neurons (ANOVA: FgrOup= 26.67, p<0.0001, Tukey post- hoc: ****p-adjusted < 0.0001), and the inhibitory neurons (ANOVA: FgrOup= 26.67, p<0.0001, Tukey post-hoc: ***p-adjusted < 0.001). k. Expression of the turquoise and lightyellow modules in SodlG86RmCSN at presymptomatic (30 and 60 days) and symptomatic (90 and 105 days) stages (ANOVA: Fgroup = 45.25, p<0.001; Tukey post-hoc: WT vs SodlG86Rat 30d p=0.075; 60d p=0.27; 90d ***p<0.001; 105d p=0.089). I. Representative images of puromycine incorporation in mCSN (CTIP2- positive neurons in layer 5 motor cortex) from 90-day-old WT and SodlG86Ranimals, upon intracerebroventricular injection. Scale bar=20pm. m. Dot plot showing the quantification of the puromycin incorporation in mCSN (n=5 WT and 6 SodlG86R; two-tailed nested t-test; *p=0.0207).

[0113] Figure 3. Integrative genetic analysis identifies the master regulator CREB3 as protective factor in ALS. a. Experimental design used to identify TFs upstream of the highly conserved and significant WGCNA turquoise module. Genes in the turquoise module were intersected with a list of transcription factors (TFs) from the ENCODE Consortium to identify potential master regulators. Each TF candidate was then further investigated using an integrative genetic and transcriptomic approach combining WGCNA-based connectivity, expression profile of the TF-target genes and burden of rare variants among identified TFs. b. Cluster dendrogram and heatmap showing the expression of the TF-target genes in each of the 30 DE human cell populations in ALS patients compared to healthy controls. Row clustering of TFs based on their target gene expression identifies two main clusters of TFs. The TFs cluster indicated by the dashed line rectangle was further prioritized based on a higher expression level of their target genes in deep layer excitatory neurons (L5-ET). Right-sided heatmap showing WGCNA-based TF connectivity in prioritized cell populations (Figure 2), association of missense rare variants in each TF with ALS risk, and expression profile of the target genes in mCSN at a presymptomatic (combined 30 and 60 days) and symptomatic (combined 90 and 105 days) stages. Bottom heatmap shows the expression profile of the prioritized TF CREB3 in ALS and FTD patients compared to healthy controls, c. Quantile-quantile plot of the meta-analyzed gene burden of rare missense variants in a cohort of 1,018 ALS patients and 3,850 healthy controls showing SOD1 gene as the top association, followed by other ALS known genes such as TBK1. d. Locus zoom plot showing the SNP (+ / - 500KB) rsll538707 (R119G) association with ALS in the discovery cohort and the replication of 1,018 ALS cases and 3,850 healthy controls. Dashed line shows the genome-wide significant SNP at a p-value < 5e 08and colored dots represent LD with the lead variant (lozenge). Forest plot of the gene burden association for known ALS genes such as SOD1 (OR=5.23; 95%CI: 2.41-11.35; p=2.21e 05), TBK1 (OR=1.79; 95%CI: 1.28-2.5; p=4.71e 04) and TARDBP (OR=1.97; 95%CI:1.26-3.07; p=2.1e 03) as well as the association of CREB3 gene. Aggregation of rare missense variant in CREB3 confers a reduced risk of ALS (OR=0.66 95%CI 0.51- 0.87; p=2.9e 03). f. Forest plot showing four previously identified ALS associations on the KIF5A, UNC13A, C9ORF72 and CFAP410 genes, as well as the novel association of the missense variant rsll538707 (R119G) on CREB3 with ALS risk.

[0114] Figure 4. CREB3 target genes expression as a resilience marker in mouse and human neurons a. Violin plot showing increased CREB3 mRNA expression in mCSN of presymptomatic and symptomatic SodlG86Rmice (ANOVA: FgenotyPes = 4.78, *p < 0.05, Tukey post-hoc: *p-adjusted = 0.029) b. Violin plot showing the expression of CREB3 target genes in motoneurons compared to other cell types in the spinal cord of SOD1G98A(Wilcoxon rank sum test: *p < 0.05, motoneurons vs all cell types), c. Heatmaps showing increased CREB3 target genes expression in the motor and frontal cortices of ALS patients versus FTDL patients, d. Representative images of the Creb3 and Fezf2 mRNA expression revealed by RNAscope in layer 5 of the motor cortex of 90-day-old WT and SodlG86Rmice (left, scale bar=20pm), and dot plot showing the quantification of Creb3 in probe integrated intensity in Fezf2+ neurons (n=l female and 4 male WT mice, and n=2 female and 3 male SodlG86Rmice; two-tailed nested t-test; *p=0.0255). e. Representative images of CREB3 and CTIP2 immunoreactivity layer 5 of the motor cortex of an end-stage SodlG86Rmouse and WT littermate (left, scale bar=20pm), and dot plot showing the quantification of CREB3 immunoreactivity in CTIP2+ neurons (n=l female and 4 male WT mice, and n=l female and 3 male SodlG86Rmice; averaged age at perfusion = 102 days for WT, 101 days for SodlG86R; two-tailed nested t-test; *p=0.0207). Figure 5. Increased CREB3 regulon activity is associated with a slower disease progression rate in ALS a. Schematic identification of CREB3 regulon from RNAseq brain and blood data which encompasses CREB3 target genes identified through Chip-seq screening, and co-regulated genes identified through weighted gene co-expression analysis, b. Scatter plot showing significant positive correlation between CREB3 regulon activity in the motor cortex of ALS patients and overall survival (Pearson R2= 0.28; **p=0.008). c. Survival curve shows that ALS patients with higher brain CREB3 regulon activity have an overall longer disease duration (~46 months) compared to ALS patients with lower CREB3 regulon activity (~29 months) (Kaplan-Meier survival: Chi2= 8.1; df = 1; **p = 0.004). d. Scatter plot showing significant positive correlation between CREB3 regulon activity in the blood of ALS patients and overall survival (Pearson R2= 0.22; *p=0.022). e. Survival curve shows that ALS patients with higher blood CREB3 regulon activity have an overall longer disease duration (~77 months) compared to ALS patients with lower CREB3 regulon activity (~62 months) (Kaplan- Meier survival: Chi2= 5.5; df = 1; *p = 0.02).

[0115] Figure 6. R119G variant slows ALS progression through gain of function mechanisms a. ALS functional rating scale (ALSFRS) progression in rsll538707 (R119G) carriers shows slower disease progression (mean = -0.69 pts / months) compared to non-carriers (-0.93 pts / months), b. ALS patient's carriers of the rsll538707-G (R119G) variants shows a significant slower disease progression characterized by an attenuation of the functional decline compared to non-carriers (Wilcoxon rank test ALS-FRS slope: Bonferroni's adjusted **p = 0.00183). c. ALS patients rsll538707 (R119G) carriers shows a significantly extended disease duration (mean = "'50.1 months compared to non-carriers (mean= ~ 38.5 months) (Kaplan Meier: Chi-squared = 4.1, *p = 0.04). d. Genome browser Chip-seq track showing CREB3 binding on the canonical target gene promoter of SEC24A gene. e. Design of CREB3 and CREB3-R119G expression vectors and SEC24A-GLuc-ON reporter plasmid transfected in HEK293 cells (left), and plots (right) showing increased SEC24A promoter activity in CREB3-transfected cells compared to an empty vector (Nested ANOVA: F2-15 = 45, p < 0.0001; Bonferroni's test: ***p < 0.0001), further enhanced by the R119G missense mutation (grey violins) (Nested ANOVA: F2-15 = 45, p < 0.0001; Bonferroni's test: **p = 0.0068).

[0116] Figure 7. CREB3 rare variants confer protection against 4HNE-induced cellular toxicity onto HEK 293 cells

[0117] Cellular mortality is represented as cumulative percentage of dead or apoptotic cells in each condition. The common variant (WT) and three rare variants (R119G, R155W and R168H) were tested in this experiment. DMSO : Control vehicle ; 4HNE : 4-hydroxynonenal. Figure 8. CREB3 rare variants allow a greater cleavage of CREB3 protein

[0118] The common variant (WT) and three rare variants (R119G, R155W and R168H) were tested in this experiment. DMSO: Control vehicle ; BFA : Brefeldin-A. a. Western Blot using antibody against GFP (A11122 Invitrogen). b. Ratio of cleaved form (lower band) on full-length form (upper band) quantified from the western blot membrane for each variants and conditions. Dark: DMSO; Light: BFA.

[0119] Figure 9. Specific blood proteomic signature for CREB3R119Gcarriers a. Volcano Plot representing b-coefficient versus -loglO(p value) of each blood proteins calculated by linear regression for CREB3R119Gcarriers, b. Forest plot representing b-coefficient of ELAVL4 protein calculated by the linear regression for each CREB3 carrier status.

[0120] EXAMPLES

[0121] Mouse CSN RNAseq

[0122] Animals

[0123] All animal experiments were performed under the supervision of authorized investigators and approved by the local ethical committee of Strasbourg University (CREMEAS, agreements #00766, #1534 and #28521). BAC transgenic male mice with the G86R murine Sodl missense mutation were obtained from the animal facility of the Faculty of Medicine, University of Strasbourg. Non- transgenic age-matched male littermates served as controls. Mice received water and regular rodent chow ad libitum. SodlG86Ranimals were followed daily, and disease progression was rated according to a clinical scale going from score 4 to 0.

[0124] Retrograde labelling ofmCSN and FACS purification ofmCSN and mCtl cells

[0125] 25-day-old WT and SodlG86Rmice were deeply anesthetized with an i.p. injection of Ketamine (Imalgene 1000®, Merial; 120 mg / kg body weight) and Xylazine (Rompun 2%®, Bayer; 16 mg / kg body weight) and placed on a heating pad. A laminectomy was performed on C3-C4 cervical vertebrae and the animals were positioned below an injector (Nanoject II, Drummond Scientific, PA) mounted on a micromanipulator. A pulled glass capillary loaded with Green IX Retrobeads (Lumafluor) was used to puncture the dura and lowered to the dorsal funiculus. Five pressure microinjections of 23 nl were performed on each side of the dorsal funiculus. 30, 60, 90 105-day- old injected mice were deeply anesthetized with by an i.p. injection of Ketamine / Xylazine (120 mg / kg; 16 mg / kg) before decapitation. The brains were sectioned in a stainless-steel coronal brain matrix (Harvard Apparatus, MA), and 1 mm-thick sections were transferred under a fluorescence SMZ18 microscope (Nikon). Cortical layer 5 and layers 2-3 from same animals were microdissected from 4 coronal sections and collected in separate tubes filled with iced HABG (Hibernate A (BrainBits UK), B27, Glutamax (Gibco) and 0.1N NaOH). The microdissected tissues were enzymatically digested with 34 U / ml papain at 37°C for 30 min. Cells were mechanically dissociated by gentle trituration in iced HABG, filtered through 70 pm cell strainer (BD Falcon) and subjected to density centrifugation through a three-density step gradient of Percoll (Sigma, MO). Upon centrifugation, the cell pellets were resuspended in cold 0.01 PBS and fixed with 70% EtOH for 30 min at 4°C. Fixed cells were centrifugated to eliminate EtOH and resuspended in 0.01 M PBS complemented with RNAse inhibitors (Promega). Microsphere-labelled mCSN were purified using the FACS Aria II (BD Biosciences), based on their fluorescence, size and granularity. Unlabelled control cells were purified by FACS, using the same size and granularity as for the mCSN. Approximately 2,000 CSN and exactly 2,000 control cells were collected from each adult mouse brain and used as individual biological replicates for RNA sequencing.

[0126] RNA sequencing

[0127] Full length cDNA was obtained using the SMART-Seq v4 Ultra Low Input RNA kit for Sequencing (Clontech, CA) according to the manufacturer's instructions. 11 cycles of cDNA amplification were performed using the Seq-Amp polymerase. 600ng of pre-amplified cDNA were then used as input for Tn5 transposon tagmentation by the Nextera XT kit (Illumina) followed by 12 cycles of library amplification. Following purification using Agencourt AMPure XP beads (Beckman Coulter), the size and concentration of the libraries were assessed on an Agilent 2100 Bioanalyzer. The libraries were then loaded in the flow cell at a concentration of 3nM, clusters were generated by using the Cbot and sequenced on the Illumina HiSeq 4000 system as paired- end 2x50-base reads, following Illumina's instructions. Raw reads were mapped to the mouse reference genome GRCm38 with STAR version 2.7.0 and default parameters using Ensembl gene annotations (version 87). Gene-level abundance estimates were estimated using the option- quantMode geneCount in STAR. We filtered the lowly expressed genes wherein each gene was required to have at least 15 counts across all samples and used both exonic and intronic reads. The filtered set of genes was used for the PCA plot and differential expression analysis. Differential gene expression analysis was performed with the ARMOR workflow and a cut off FDR value of 0.05 was set in both datasets.

[0128] Human single nucleus RNAseq (snRNAseq) data analysis and cluster annotation

[0129] SMART-seq v4

[0130] Raw read (fastq) files from the Allen Brain Atlas were aligned to the GRCh38 human genome sequence (Genome Reference Consortium, 2011) with the RefSeq transcriptome version GRCh38.p2 (RefSeq, RRID SCR_003496, current as of 13 April 2015) and updated by removing duplicate Entrez gene entries from the gtf reference file for STAR processing.

[0131] 10x Chromium RNA sequencing (Cv3)

[0132] Raw FASTQfiles were downloaded from Pineda et al., 2024 and the Allen Brain Atlas and aligned to the pre-mRNA annotated human reference genome GRCh38 using Cell Ranger v4.0 (lOx Genomics, Pleasanton CA) except for substituting of the curated genome annotation used for SMART-seq v4 quantification. Introns were annotated as 'mRNA', and intronic reads were included to quantify expression. Quality control criteria were used, so that for Cv3, criteria were: more than 500 (nonneuronal nuclei) or more than 1,000 (neuronal nuclei) genes were detected and doublet score was less than 0.3.

[0133] Clustering of snRNA-seq data

[0134] Nuclei were grouped into transcriptomic cell types using an iterative clustering procedure. Read counts were summed, and logj-transformed expression was centered and scaled across nuclei. Clusters were identified with Louvain community and pairs of clusters were merged if either cluster lacked marker genes. Clustering was applied iteratively to each subcluster until clusters could not be further split and robustness was assessed by repeating iterative clustering 100 times for random subsets of 80% of nuclei. Consensus clusters were defined by iteratively splitting the co-clustering matrix. The clustering pipeline is implemented in the R package scrattch.hicat v0.0.22 with marker genes defined using the limma package; the clustering method is provided by the 'run_consensus_clust' function. Clusters were curated based on quality-control criteria or the expression of markers genes.

[0135] To establish a set of human consensus cell types across all three datasets, we performed a separate integration of snRNA-seq technologies on the major cell classes (glutamatergic, GABAergic, and non-neuronal).

[0136] Each expression matrix was log2(CPM + 1) transformed then placed into a Seurat object and variable genes were determined by down sampling each expression matrix to a maximum of 300 nuclei per scrattch.hicat-defined with n set to 20, to generate a list of up to 20 marker genes per cluster. The union of the Cv3, Cv3-ALS and SSv4 gene lists were then used as input for anchor finding, dimensionality reduction, and Louvain clustering of the full expression matrices. Louvain clustering was performed to over cluster the dataset to identify more integrated clusters than the number of scrattch.hicat-defined clusters. For instance, glutamatergic neurons had 30, 47 and 48 scrattch.hicat-defined clusters, 110 overclustered integrated clusters, and 38 final human consensus clusters after merging for Cv3 and SSv4 datasets, respectively. To merge the over clustered integrated clusters, up to 20 marker genes were found for each cluster to establish the neighborhoods of the integrated dataset. Clusters were then merged with their nearest neighbor if there were not a minimum of ten Cv3 and two SSv4 nuclei in a cluster, and a minimum of 4 DEGs that distinguished the query cluster from the nearest neighbor.

[0137] Cell-type specific differential expression

[0138] Differentially expressed genes (DEGs) for a given species were identified by using Seurat's FindAIIMarkers function with a Wilcox test and comparing each cluster with every other cluster under the same subclass, with logfc.threshold set to 0.7 and min. pct set to 0.5. The union of up to 100 genes per cluster with the highest avgJogFC was used. The average Iog2 expression of the DEGs was then used as input for the build_dend function from scrattch.hicat to create the dendrograms.

[0139] Cell type-specific pseudo-bulk differential gene expression (DGE) groups were built based on hierarchical clustering and Euclidean distance between each cluster leading to 30 DGE cell groups. For each of the DGE cell group, differential expression analysis was performed as described in the ARMOR workflow for sufficiently abundant cell types using age, sex, and disease group as design covariates and gene-wise single-cell-level variance as weights for the linear model.

[0140] CSN gene signature in human snRNAseq

[0141] Ageneset of GFP-positive cells was built based on DGE analysis of CSN vs and control cells in mouse. Our final set of ~50 genes was used as an input to AUcell to evaluate the distribution of AUC scores across all the cells and explore the relative expression of the signature. The function AUCell_exploreThresholds() was used to determine the minimum AUC values where cells are considered to express the CSN geneset. For each cluster, the proportion of cells expressing the CSN geneset was calculated.

[0142] To identify overlap between mouse and human snRNAseq gene signatures, cell-type specific DGE genes were intersected with mCSN DEG genes, and significant overlap was calculated using an hypergeometric test with the phyper() function in R.

[0143] Cross-species consensus weighted-gene coexpression network analysis (WGCNA)

[0144] To facilitate comparison across species, mouse gene identifiers were re-annotated with human Ensembl gene orthologs using biomaRt, an R interface with the Biomart database . Only identifiers that were common to both human and mouse meta-sets were retained. A consensus network represents a single network arising from multiple sources of data constructed from the weighted average of correlation matrices from both the human and mouse in this study. By definition, consensus modules are the branches of a clustering tree developed from a consensus gene dissimilarity, comparable to the single-network approach; consensus modules contain genes that are closely related in both networks, i.e., the modules are present in both networks. After scaling the network (consensus scaling quantile = 0.2), a threshold power of 14 was chosen (as the smallest threshold resulting in a scale-free R2fit of 0.9) and the consensus topological matrix was created as follows : consTOM <- consensusTOM(multiExpr,checkMissingData = TRUE,maxBlockSize = Inf, randomSeed = 12345, corType = "pearson", maxPOutliers = 0.05, quickCor = 0, pearsonFallback = "individual", power = 14, networkType = "signed", TOMType = "signed", networkcalibration = "full quantile", calibrationQuantile = 0.95,sampleForCalibration = TRUE, sampleForCalibrationFactor = 5000). The consensus tree was then built and the modules identified with the function cutTreeDynamic() with the default parameters and the minModuleSize (=30). This approach identified 36 modules for which correlation to phenotype (ALS) and adjusted-pvalue (FDR) were calculated and intersected between mouse and human to identify conserved directionality and association. This approach led to the identification of a final set of 20 modules. Module eigengenes (M Es) were used for module-trait association analysis, differential eigengene network analysis, and for differential gene expression analysis. Difference in expression across trait groups was tested using a Kruskall- Wallis one-way analysis of variance. A gene's module membership (k ME) is defined as the Pearson correlation between each gene and each ME; genes with high k E values were considered "hub" genes and were highly co-expressed within a subnetwork. Module preservation statistical tests were used to assess how well network properties of a module in one reference data set were preserved in a comparator data set (modulepreservation function in WGCNA). Preservation statistics are influenced by a number of variables (module size, network size, etc). A composite preservation Z-score (Z summary) was used to define preservation relative to a module of randomly assigned genes where values 5 < Z < 10 represent moderate preservation, while Z > 10 indicated high preservation. Genes in each network module were characterized using EnrichR (version 1.2.5), and we considered a term to be significant for FDR < 0.05. Genes network modules were constructed using metacells transformed data aggregated by cell class. For glutamatergic metacells matrices, we calculated signed kMEs for each gene and each module identified in the WGCNA analysis. The topological matrix was filtered to contain selected hub genes (2 gene / module) and was used an input to UMAP dimensionality reduction. Finally, each dot size (genes) was scaled to the signed kMEs for the corresponding module.

[0145] Violin plot of eigengenes expression

[0146] For each module and corresponding cell type, module eigengenes was calculated in control and ALS conditions using the moduleEigengenes function. Comparison across conditions was calculated using a Kruskall— Wallis one-way analysis of variance and considered significant when FDR < 0.05.

[0147] ENCODE Chip-seq transcription factors

[0148] A list of 281 transcription factors and their bed files were downloaded from the ENCODE consortium. Bed files were intersected with TSS annotations from the GRCh38 version and peaks were annotated to the closest TSS genes.

[0149] A final list of 278 TFs was intersected with genes identified in the WGCNA turquoise module, which lead to a final set of 18 TFs candidates. For each of TFs candidate, target gene Z-scores were average 1 for each cell type and cluster using the hclust() function in R. CREB3 target gene Z-scores were averaged across cell types and compared between FTD and ALS patients.

[0150] Genetic data processing and association

[0151] Association testing and meta-analysis

[0152] Whole genome sequencing from 1,345 ALS patients and 3,860 controls were processed as described above. Duplicate individuals were removed (king-cutoff = 0.084). Population structure was assessed by projecting l,000G principal components (PCs) and outliers from the European ancestry' population were removed (> 4 SD on PCI-4). Finally, samples in common between the individual genotype data and van Rheenen's study were identified using the checksum program id_geno_checksum and were removed from our analyses. In total, 1,018 ALS cases and 3,850 controls pass quality check analysis and were used for rare variant burden and genome-wide association analyses (GWAS). After quality control, a null logistic mixed model was fitted using SAIGE with principal component (PC)l-PClO as covariates. The model was fit on a set of high-quality (INFO > 0.9) SNPs pruned with PLINK 2.0 ('-indep-pairwise 50 25 0.2') in a leave-one-chromosome- out scheme. Subsequently, a SNP-wise logistic mixed model including the saddle point approximation test was performed using genotype dosages with SAIGE. To assess any residual confounding due to population stratification and artificial structure in the data, we calculated the LDSC intercept using SNP LD scores calculated in the HapMap3 CEU population.

[0153] Genic burden association analyses

[0154] To aggregate rare variants in a genic burden test framework we used the method described in the ALS 2021 GWAS. In short, a variety of variant filters was applied to allow for different genetic architectures of ALS associated variants per gene. In summary, variants were annotated according to allele-frequency threshold (MAF < 0.01 or MAF < 0.005) and predicted variant impact ("missense", "damaging", "disruptive"). "Disruptive" variants were those variants classified as frameshift, splice-site, exon loss, stop gained, start loss and transcription ablation. "Damaging" variants were missense variants predicted to be damaging by seven prediction algorithms (SIFT, Polyphen-2, LRT, MutationTaster2, Mutations Assessor, and PROVEAN). "Missense" variants are those missense variants that did not meet the "damaging" criteria. All combinations of allele frequency threshold and variant annotations were used to test the genic burden on a transcript level in a Firth logistic regression framework where burden was defined as the number of variants per individual. Sex and the first 20 principal components were included as covariates. All ENSEMBL protein coding transcripts for which at least five individuals had a non-zero burden were included in the analysis. Meta-analysis was performed using an inverse variance weighted method.

[0155] Validations on mouse tissues

[0156] Puromycin administration

[0157] Six 90-day-old female SodlG86Rmice and five WT female littermates were anaesthetized with 2% isoflurane / 98% air, and two intracerebroventricular injections (one per hemisphere) were performed in the lateral ventricles to deliver 2.07pl of puromycin (Sigma, P7255) diluted in saline at 25pg / pl. Animals were allowed to recover from surgery and sacrificed and perfused one hour later. Their brains were collected and processed for immunofluorescence.

[0158] Immunofluorescence

[0159] Mice were euthanized with an overdose of pentobarbital sodium and phenytoin sodium (120 mg / kg) and transcardially perfused with cold 0.01 M PBS, followed by cold 4% PFA in 0.01 M PBS. Brains were dissected and post-fixed overnight in 4% PFA. Brains were cut coronally in 40pm-thick sections on a vibratome (Leica Biosystems). To reveal CREB3 expression, brain sections were first heated at 80°C in lOmM citrate buffer for 30 min. This step was not used to reveal puromycin. Brain sections were incubated lh in blocking solution (8% goat serum, 0.3% BSA, 0.3% Triton in PBS), and 48 to 72h at 4°C with primary antibodies. Sections were rinsed, incubated 2h with the secondary antibodies, rinsed and mounted in Prolong Diamond mounting medium (Invitrogen, # P36970). The primary antibodies used were: mouse@puromycine (DSHB, #PMY-2H4; 1 / 100), rat@CTIP2 (Abeam, #Abl8465; 1 / 100) and rabbit@CREB3 (Aviva Systems Biology #OAAN03577, 1 / 50). Secondary antibodies were from the Alexa series (Life Technologies; 1 / 1000).

[0160] RNAScope

[0161] Upon fixation as described above, brains were cryoptotected and cut coronally in 14pm-thick sections on a cryostat (Leica). RNAscope® Multiplex Fluorescent Reagent Kit v2 (Advanced Cell Diagnostics, #323100) was employed according to the manufacturer's instructions, using probes targeting Creb3 and Fezf2 (Advanced Cell Diagnostics #1263611-C3 and #313301-C2). Briefly, sections were incubated with IX PBS for 5min to wash out OCT, baked for 30min at 60°C, post-fixed with 4% fresh PFA for 90min, sequentially dehydrated with increasing concentrations of ethanol, baked again for 30min at 60 °C, and incubated for lOmin in RNAscope™ Hydrogen Peroxide Reagent at RT. Target retrieval was performed for 5min in a steamer, and RNAscope™ Protease III Reagent was applied for 15min at 40°. Slides were then hybridized with target probes for 2h at 40°C in a hybridization oven (Boekel). Slides were stored with 5X Saline Sodium Citrate overnight at RT and signals were amplified using amplifiers and horse radish peroxidases from the reagent kit and TSA fluorophores (TOCRIS, # 7526 and # 7527). Samples were mounted in Superfrost® Plus slides (VWR, #631-0108) using Prolong™ Diamond Antifade Mountant medium (Invitrogen, #P36965).

[0162] Image acquisition and quantification

[0163] Images were captured at 63X using an Axiolmager.M2 microscope equipped with a structured illumination system (Zeiss) and a high-resolution B / W camera (Hamamatsu), and run by the ZEN 2 software (Zeiss). Image analyses were performed with ImageJ (NIH). Modal Grey Value of puromycin and CREB3 signals was measured in CTIP2+ neurons in layer 5 of the mouse motor cortex, and Creb3 integrated density was measures in Fezf2+ cells.

[0164] Dual luciferase assays

[0165] HEK-293 cells were cultured in Dulbecco's modified Eagle's Medium (DMEM), 10% foetal bovine serum and 1% penicillin-streptomycin at 37°C in 5% CO2. Cells were plated in a 6-well plate and transfected 24 h after plating in DMEM + 0.1% foetal bovine serum and 1% penicillin-streptomycin using Lipofectamine 2000 (Invitrogen). 800 ng of dual luciferase reporter containing CREB3 response element was co-transfected with 2400 ng of pCMV6, pCMV6-CREB3 or pCMV6-CREB3- R119G plasmid. Dual luciferase assays were performed 24h after transfection as described by the manufacturer (Promega Secrete-Pair Dual Luminescence Assay Kit # LF033).

[0166] Statistics

[0167] Immunofluorescence and RNAScope data are represented as nested scatter dot plots, and were analysed using two-tailed nested t tests performed on Prism 6 (GraphPad). Results were considered significant when p<0.05. For the luciferase assay, transfected cells were run as duplicate for a total of 6 independent experiment. A linear mixed-model was fitted to integrate the technical replicate as random effect followed by a Bonferroni's test for multiple comparison. Results were considered significant when adjusted-p < 0.05.

[0168] Evaluation of cell survival

[0169] Plasmids - pcDNA3.1 (+) plasmids encoding a fusion protein between common or rare variants of CREB3 and a fluorescent reporter (EGFP) were designed. The EGFP coding sequence was fused to the N-terminal of CREB3 or variants, separated by a flexible glycine-serine linker.

[0170] Day 1 & 2 - 7.0- 105HEK 293 cells were seeded on 6-wells plate in technical triplicate. Cells were transfected the next day with 2 pg of plasmid using Lipofectamine™ 2000 (Invitrogen) in low serum medium following the manufacturer's recommendations. Day 3 - Cells were treated with 50 pM 4-hydroxynonenal diluted in DMSO for 4 hours. After treatment, CellROX® Oxidative Stress Reagent [Orange - C10443 Molecular Probes] were added directly into the medium at 5 pM final concentration for 30 minutes. Cells were detached using Trypsin + ETDA 0,05% (Gibco) 5 min at 37°C and centrifuged 5 min at 500 g. Cells pellets were washed with PBS 0.1M and resuspended in Annexin Binding Buffer [Invitrogen], 1-105cells were transferred into a new tube and stained with 5 pL Annexin V Conjugate [Pacific Blue - A35122 Invitrogen] for 15 min in a dark environment. After incubation, the staining reaction was stopped by addition of 4V of Annexin Binding Buffer. Then, TO-PRO3 iodide [Red - T3605 Invitrogen] was added at a final concentration of 35 nM 15 min before acquisition.

[0171] Acquisition - Tubes were passed through a Flow Cytometer (BD FACSymphony™) with recording of 4 channels (EGFP, Pacific Blue, PE, Alexa-647). 10 000 EGFP(+)cells were recorded in each tube and analysis was performed using the BD FACSDiva 9.0.2 software. Apoptosis determination was performed by plotting Pacific Blue (Annexin) signal versus Alexa-647 (TO-PRO-3) signal.

[0172] Western blot and detection of cleaved forms

[0173] Day 1 & 2 - 7.0- 105HEK 293 cells were seeded on 6-wells plate. Cells were transfected the next day with 2 pg of plasmids encoding a fusion protein between common or rare variants of CREB3 and a fluorescent reporter (EGFP), using Lipofectamine 2000 (Invitrogen) in low serum medium following the manufacturer's recommendations.

[0174] Day 3 - Cells were treated with 2,5 pg / mL brefeldin-A diluted in DMSO for 8 hours. After treatment, cells were detached using Trypsin + ETDA 0,05% (Gibco) 5 minutes at 37°C and centrifuged.

[0175] Western Blot - Cells pellets were washed with PBS 0.1M and resuspended in RIPA Buffer (Tris-HCI pH 8 50mM, sodium chloride 150mM, sodium deoxycholate 0.5 %, SDS 0.1 %, Triton-XlOO 1 %). Then cells were mechanically lysed by up and down pipette movements and incubated for 30 minutes on ice. Cells were sonicated 10 minutes at 20 kHz and centrifugated 10 minutes at 10,000 g at 4°C. Proteins in the supernatant were dosed using the Pierce™ BCA kit (Thermo Fischer). Laemmli Sample Buffer (Bio-Rad) containing 10% p-mercaptoethanol (Sigma) was added and 8 pg of proteins were denatured for 5 min at 95°C. Proteins were separated on a 4-20% gradient stain- free gel (Bio-Rad) and let for migration at 90V for 30 min followed by 180-200V for lh. Proteins were transferred on a 0.2 pm nitrocellulose membrane using the Trans-blot® TurboTM (Bio-Rad), 7 min at 30V in a Transfer buffer. Membranes were blocked in PBS 0.1M containing 10% BSA for lh. The primary antibody was diluted at the appropriate concentration in PBS 0.1M containing 5% BSA and membrane were incubated O / N at 4°C. After three washes of 10 min with washing buffer (Tris pH 7.4 1 M, NaCI 5 M, Tween 20 0.1 %), the corresponding secondary antibody coupled with HRP was added at 1 / 5,000 in PBS 0.1M. Three last washing steps of 10 min with washing buffer and one with PBS 0.1M were performed and membranes were revealed with ECL LuminataTM Forte (BioRad). Signals were acquired with the ChemiDoc XRS+ System (BioRad). Band intensity was analyzed using ImageLab (Bio-Rad) by normalizing the band intensity on total protein from stain-free gel. qPCR - Cells pellets were washed with PBS 0.1M and resuspended in TRIzolTM (ThermoFisher). RNA was purified by a chloroform extraction, followed by precipitation with 2.5 V / V of ethanol 100% and 0.1 V / V of sodium acetate 3M pH 5.5. RNA was resuspended in RNase-free water. Reverse transcription was performed using 1 pg of RNA and PrimeScript RT Master Mix (Takara). cDNA was amplified using SsoAdvanced SYBR Green (Bio-Rad) reagent and using a thermocycler (Bio-Rad) with a usual program (denaturation of 30 sec at 95°C, followed by 40 cycles of 4 sec at 95°C and 4 sec at 60°C). The amplification step was completed with the melting curve analysis (3 sec at a temperature from 65°C to 90°C with 0,5°C increments) to ensure amplification specificity. Gene expression was determined using GeNorm based on the housekeeping genes TBP and P0L2.

[0176] Proteomic signature

[0177] Olink proteomic profiling was conducted on blood plasma samples collected from 39 739 UK Biobank participants using the Olink Explore 3072 platform. This platform has measured 2,923 protein analytes, reflecting 2,941 unique proteins across the Olink panels that comprise the 3072 panel. For whole-exome sequencing-based proteogenomic analyses, the samples were analyzed with available paired-exome sequence data. Clinical, biochemical and proteomic data were associated for each 39 739 individuals based on their CREB3 carrier status (CREB3V T, CREB3R119G, CREB3R155W, CREB3R168H, CREB3L120P, CREB3K132E, CREB3K172E, CREB3G346R). Groupwise outlier removal of protein measurements (±1.5 x IQR) was performed before regression analysis. Linear regression adjusted for sex was performed for each protein and carrier status.

[0178] Example 1 - Cross-species RNAseq integration prioritizes disease-vulnerable human neuronal populations

[0179] The purpose was to identify and investigate ALS-vulnerable human neuronal populations. While snRNAseq datasets from ALS motor cortex (Ml) are available and allow accurate identification of cell populations, they do not allow unambiguous corticospinal neurons (CSN) identification. However, in mice, projection-based anatomical definition of CSN is accessible. Therefore, crossspecies transcriptomics were performed in order to identify disease vulnerable CSN in ALS (Figure la).

[0180] To accurately identify cell populations in the human motor cortex, independent snRNAseq datasets obtained from 23 sporadic and familial ALS patients, 26 FTLD patients and 17 aged-matched control, and healthy human individuals were integrated (Figure la). The analysis yielded a total of 245,143 nuclei. High-quality nuclei were normalized using SCTransform and integrated across the independent datasets allowing the identification of a total of 116 clusters (Figure lb). These were further hierarchically organized based on transcriptomic similarities and associated with their positioning across the cortical layers, providing a full representation of glutamatergic excitatory neurons, GABAergic inhibitory neurons, and non-neuronal cells. In the context of the present purpose, it is noteworthy that the population of extra-telencephalic neurons from the cortical layer 5, L5-ET, suspected to represent ALS-vulnerable human CSN, yielded the smallest number of neurons ("'1000 neurons, corresponding to 0.6% of the total nuclei and 1.7% of all excitatory neurons). Indeed, after proper filtering, it was observed that human L5-ET have the lowest number of expressed genes compared to other excitatory neurons. More importantly, it was observed that the gene-level dispersion is much higher for low to medium expressed genes and negatively correlates with the number of cells sequenced per individual, suggesting that higher gene expression variability across individuals arises from lower sampling of rare cell types.

[0181] To identify presumable CSN amongst the different human neuronal populations, mouse models were used to generate a new dataset of anatomically-identified, disease-vulnerable mouse corticospinal neurons (mCSN) and compare mCSN dataset with the human datasets. mCSN were retrogradelly labelled from the cervical dorsal funiculus of SodlG86Rmice, a transgenic mouse model of ALS, and their control littermates at two presymptomatic ages (30 and 60 days), and two symptomatic ages (90 and 105 days) (Figure la). Labelled mCSN were then purified by FACS. Nonlabelled cells from layers 2 / 3 were also purified from the same animals to serve as a control cellular population (mCtl cells) (Figure la). 64 samples were initially sequenced, 7 were excluded based on quality or purity criteria, and 57 samples were further processed. Unsupervised hierarchical clustering clearly separated the mCSN from mCtl cells and suggested that a larger part of the variance observed between WT and SodlG86Rmice is attributable to mCSN dysfunction.

[0182] It was reasoned that cross-species RNAseq integration could be employed i) to identify the human neuronal populations that more closely ressemble mCSN, independently of the disease condition or genotype ( / .e. 'cellular identity', Figure lb, lc), ii) to identify the human neuronal populations whose differential gene expression (DGE) in ALS or FTLD conditions mimicks mCSN DGE between SodlG86Rand control mice, ( / .e. 'cellular vulnerability', Figure lb, lc), and Hi) to prioritize human neuronal populations based on the combination of 'cellular identity' and 'cellular vulnerability' ( / .e. 'ranking', Figure lb, lc).

[0183] First, the top differentially expressed genes (DEG) between mCSN and mCtl cells were selected to create a mCSN geneset that represents a set of human homologous genes highly enriched in mCSN genes. Using this geneset, human and mouse datasets were integrated and revealed that most of the L5-ET neurons expressed the mCSN geneset (Figure lb). Strikingly, the mCSN geneset was also enriched in marker genes from three additional neuronal populations: L5 / 6 NP, L5-IT and L2 / 3-IT. mCtl cells genes instead were enriched in markers of non-neuronal populations: endothelial cells, microglia, and oligodendrocytes. To maximize power with differential expression analysis, aggregated hierarchical clustering was used to regroup the initially identified 116 cell clusters into 30 cell populations. This confirmed that the mCSN geneset is expressed in a large proportion of L5- ET (~60%) and a significant proportion (~40%) of L2 / 3 IT-1 neurons (Figure Id). Importantly, these neurons express a high amount of NEFH, encoding the heavy subunit of neurofilament, which is typical of long-range projection neurons. This also confirmed the identity of mCtl cells as non- neuronal populations (Figure Id). Together, this shows that several sub-populations of excitatory neurons from the human motor cortex express the gene signature of mCSN, including L5 / 6 NP, L5- IT and L2 / 3-IT populations, in addition to the expected L5-ET neurons.

[0184] Second, DEG were intersected between WT and SodlG86RmCSN with DEG between controls and ALS patients, or DEG between controls and FTLD patients for each of the 30 cell populations (Figure Id). This analysis confirmed a significant enrichment of mCSN DEG in human excitatory neuron populations as opposed to inhibitory neuron and non-neuronal cell populations (Bonferroni's adjusted *p < 4e 04). More precisely in ALS, this identified L5 / 6 NP, L6b, L6 IT, L 2 / 3 IT 1, 2 and 4, and L5 IT 1, 3 and 5 (Figure Id). The same populations were identified in FTLD, together with L5 ET, L2 / 3 IT 3, and L5 IT 2 a,d 4 (Figure Id).

[0185] Third, each human cell population was ranked based on the combination of the 'cellular identity' and 'cellular vulnerability' results which prioritized L5-ET and L2 / 3 IT-1 populations as the human populations that more closely recapitulate the transcriptomic signature of diseased mCSN (Figure Id). In whole, cross-species RNAseq integration enabled to prioritize L5-ET and L2 / 3 IT-1 populations for further analyses.

[0186] Example 2 - WGCNA consensus on prioritized cell populations reveals conserved gene regulatory network in ALS

[0187] Next, it was aimed to unravel relevant disease-associated pathways in affected cell types in both species. To this aim, it was sought to identify gene modules commonly affected in both species and used weighted-gene co-expression network analysis (WGCNA) consensus across species and prioritized cell types (Figure 2a). WGCNA analysis identified four mRNA modules significantly correlated with the genotype (mice) or disease condition (humans) in the motor cortex and labelled as lightyellow, darkgreen, turquoise and darkgrey modules according to the WGCNA conventions (Benjamini-Hochberg p < 0.05; Figure 2b). Then, a graphical network of significantly associated modules was constructed to highlight hub genes, related weight of the genes within the network and connectivity (Figure 2c), that illustrates independent and relative proportions and nonoverlapping networks, and validates the quality of the WGCNA analysis. WGCNA modules based on their levels of conservation across species were further prioritized. Among the significantly associated modules, the turquoise module showed a strong preservation (Zscorepreservation> 10. Figure 2d), while the lightyellow module showed a moderate conservation (5 < Zscorepreservation< 10, Figure 2d) and the darkgreen and darkgrey modules showed poor conservation (Zscorepreservation< 5). Interestingly, the blue module also showed a high level of conservation (Zscorepreservation> 10. Figure 2d) but was only suggestively associated with ALS phenotype (Benjamini-Hochberg p=0.061, signed correlation = 0.36, Figure 2b). Based on these results, the turquoise and lightyellow modules were selected for further analyses. GO analysis on the turquoise module revealed a strong enrichment in RNA-binding genes such as EIF4A2, VCP, HNRNPA0, SF3A3 and TUBA1B (Figure 2e), along with a strong enrichment of terms associated with mRNA translation, ribosome, and mitochondrial translation (FDR < 0.05). The average expression of the turquoise module eigengenes revealed a significant increase in human L2 / 3 IT-1 and in mCSN, while it was only observed a trend to an increase in human L5-ET (Figure 2f). GO analysis on the lightyellow module highlighted terms such as ubiquitin activity such as HECTD1, UHRF2, UBE2K, as well as term associated with translation initiation factor activity such as EIF2B4, EIF3L, EIF2D, EIF2S1, the latter encoding the El F2-alpha protein, a master regulator of the endoplasmic reticulum stress response (ER stress) (Figure 2h). Additionally, it was observed a significantly increased expression of the lightyellow module eigengenes in human L2 / 3-IT-1 and mCSN neurons and a trend in human L5-ET (Figure 2i).

[0188] Since a strong enrichment of mCSN DEG was identified in a large population of excitatory neurons (Figure Id), it was tested whether the increase of the turquoise and lightyellow module eigengenes would be conserved in human cell types other than L2 / 3-IT-1 and L5-ET. To do so, all neuronal and non-neuronal human subtypes were aggregated into seven major classes and the turquoise and lightyellow module eigengenes expression across species and genotypes / conditions was assessed. Human excitatory neurons showed a significantly increased expression of the turquoise (Figure 2g) and lightyellow modules (Figure 2j) while inhibitory neurons display a marked increase of the lightyellow module (Figure 2j). Other cell types instead showed non-significant trends (Figures 2g, 2j). The data suggest that the molecular mechanisms associated with vulnerable mCSN and L5-ET affect all human excitatory neuronal populations.

[0189] To test the dynamics of the turquoise and lightyellow module eigengene regulation over time, it was taken advantage of the longitudinal mCSN neurons dataset and the expression profile of both modules at pre-symptomatic (30 and 60 days) and symptomatic (90 and 105 days) stages were interrogated. It was observed an increased expression of both modules eigengenes in SodlG86Rcompared to WT mCSN (Figure 2k), that was non-significant at presymptomatic ages, significant at 90 days, and had a non-significant trend at 105 days (adjusted-p = 0.089). Lack of significance at 105 days could be due to lower sample sizes at end stages of the disease.

[0190] Together, the data indirectly suggest that mutant mCSN and ALS patient L2 / 3-IT-1 and L5-ET populations may undergo dynamically altered gene expression related to mRNA translation following ER stress. To provide more direct evidence of altered mRNA translation and ER stress in mCSN, an in vivo puromycin assay was performed on early symptomatic (90 days) SodlG86Rmice and their WT littermates. Puromycin levels quantified in CTIP2-potitive L5 neurons of the motor cortex revealed a significant decrease of puromycin incorporation indicative of a slower rate of protein synthesis (Figures 21, 2m). Thus, the results identified a cross species transcriptomic footprint of disease vulnerable neurons associated to ER stress and related altered mRNA translation. neuron function and a novel factor from ALS Since the turquoise module was strongly associated with ALS and preserved across species, it was then sought to identify potential upstream regulators. To do so, the turquoise module genes were intersected with a list of TFs from the ENCODE project (Figure 3a), which led to prioritize 18 TFs (Figure 3b). Among these, hierarchical clustering based on TFs target gene expression across cell types identified a cluster of 5 TFs with a relatively high expression in L5-ET neurons (Figure 3b). These TFs include 2 members of the CREB family (CREB1, CREB3) and others such as BRF2 or SP2.

[0191] To further prioritized potential candidates, TFs-based WGCNA connectivity was used. This approach reflects the connection strength of each TFs with the neighbouring gene in the network, and revealed a strong connectivity of CREB3 with the genes of the turquoise module (Figure 3b).

[0192] To reinforce the role of CREB3 as a master regulator in human ALS, an orthogonal approach was used to test whether the CREB3 gene accumulates rare variants in ALS and / or healthy individuals. (Figure 3c). To do so, a rare missense variant burden analysis was performed on a cohort of 1,018 ALS patients and 3,850 healthy controls which was then meta-analyzed to the ALS Project Mine cohort (http: / / databrowser.projectmine.com / ). This revealed association in all known ALS genes such as SOD1 (OR=6.93 [95%CI: 3.19-15.06]; Pmeta= 5.02e07), TBK1 (OR = 1.83 [95%CI:1.30-2.56]; Pmeta=2.21e04) or TARDBP (OR = 2.11 [95%CI:1.35-3.29]; Pmeta=5.1e’04) which are conserved across cohorts, and, importantly, highlighted CREB3 and unraveled a protective effect of rare missense variants which replicated across independent cohorts (OR = 0.61 (95%CI:0.46-0.81); Pmeta=2.79e’04) (Figures 3c-e). Instead of aggregating rare variants over genes, it was tested whether individual genetic variants could be associated with ALS risk through a sequenced-based genome-wide association study (seqGWAS). A null logistic model was fitted on the cohort of 1,018 ALS cases and 3,850 controls. It was observed moderate inflation of the test statistics (XGC = 1.05), and linkage disequilibrium (LD) score regression yielded an intercept of 0.983 (s.e. = 0.0068), indicating that most of the inflation was due to the polygenic signal in ALS (LD score regression [LDSC]: / r2lhl2 = 0.69, s.e. = 0.0932). Most of the previously identified ALS GWAS variants were replicated, such as C9ORF72 (rs2484319, Pmeta= 1.50.10’45), UNC13A (rsl2608932, Pmeta= 7.93.10’26), KIF5A (rsll3247976, Pmeta= 6.01.10’13) and CFAP410 (rs75087725, Pmeta= 4.62.10’13) (Figure 3f). In addition, three novel loci were identified through inverse-variance meta-analysis on chromosome 15 (rsl2907456, Pmeta= 2.9.1O’08), chromosome 17 (rsl2907456, Pmeta= 2.75. IO’08) and chromosome 9 (rsll538707, Pmeta= 3.96. IO08) (Figures 3d-f). Most importantly, mapping rsll538707 to the human genome revealed a missense variant p.Argll9Gly (R119G) in the CREB3 gene which confers a 40% reduction in the risk of developing ALS (OR=0.61, 95%CI 0.51-0.73, Figure 3f). Overall, our cross-species transcriptomic analysis yielded a specific gene network associated with ALS from which CREB3 appears as a major master regulator, and genetic data integration further identified CREB3 as a protective factor in ALS.

[0193] 4 - CREB3 regulatory network as a resilience marker in ALS

[0194] To determine if CREB3 is actually causing altered gene expression in vulnerable neuronal populations, a series of orthogonal approaches was performed. First, the expression of CREB3 was investigated in RNAseq data from mCSN. Interestingly, CREB3 mRNA itself (Figure 4a FDR < *p < 0.05), as well as CREB3 target genes (Figure 4b) were significantly upregulated in SodlG86RmCSN. Importantly, it was also observed a strong up-regulation of CREB3 target genes in human excitatory and inhibitory neurons in the motor and frontal cortex from ALS, but not FTLD patients (Figure 4c), suggesting that CREB3 hyperactivity might be a marker of neuronal resilience in ALS. To confirm changes of CREB3 expression at the cellular level, it was performed fluorescent in situ hybridization and immunostaining on WT and SodlG86Rtissues at early symptomatic or disease end-stage, respectively, which confirmed a significant increase of CREB3 mRNA in layer 5 excitatory Fezf2+ neurons (Figure 4d) as well as a significant increase of CREB3 protein in layer 5 excitatory CTIP2+ neurons of the motor cortex (Figure 4e). Recently, disease vulnerable and disease-resistant subpopulations, respectively expressing GprinB and Colgalt2, were identified among motor cortex layer 5 neurons of the SOD1G93Amouse model of ALS (Moya et al., Cell Reports 2022). Taking advantage of these published TRAPseq data, it was observed a significant increase of the turquoise module and of CREB3 targets genes in disease vulnerable Gpr / n3-positive neurons compared to Co / go / t2-positive neurons and whole motor cortex. Finally, to test the extent of the present findings in other ALS-vulnerable neurons, the recently published snRNAseq database of SOD1G93Aspinal cord was investigated, and it was observed a significant upregulation of the turquoise module and of CREB3 target genes selectively in motoneurons, compared to astrocytes, microglia, endothelial cells, and other neuronal populations. Overall, this data confirms that mCSN and human excitatory neurons display increased CREB3 mRNA and protein expression, associated with an activation of the CREB3 regulatory networks in ALS, and shows that this signature characterizes disease- vulnerable neuronal populations.

[0195] To further test whether increased transcriptional activity of CREB3-regulated genes would confer neuroprotection in ALS patients, it was first identified, using CREB3-targeted chromatin immunoprecipitation and co-expression analysis, the CREB3 "regulon" which contains genes that would reflect CREB3 activity in tissues (Figure 5a). It was shown that CREB3 regulon activity in the motor cortex (Figure 5b, 5c) and in the blood of ALS patients (Figures 5d, 5e) positively correlates with overall survival. Noticeably, it was possible to replicate the neuroprotective effect of higher CREB3 regulon activity using two different tissues (blood and brain) from two independent datasets, demonstrating an increased survival of ~17 and ~15 months respectively (Figures 5c, e). The data highlight CREB3 regulatory network as a resilience marker in ALS.

[0196] 5 - The R119G variant slows disease progression through gain of function

[0197] Finally, it was investigated whether the R119G variant would also act as a disease modifier. Since R119G mutation acts as a protective factor, it was observed a higher allele frequency in control (minor allele frequency = 0.47%) compared to patients (minor allele frequency = 0.21%), rendering difficult phenotyping characterization of rsll538707 carriers in ALS patients. However, leveraging a large cohort of whole-genome sequenced datasets, it was shown that carriers of the rare minor allele rsll538707-G have a slower progression rate characterized by a slower decline in ALS-FRS score over time (mean ALS-FRS slope = -0.69 pts / months) compared to rsll538707-A allele (mean ALS-FRS slope = -0.93 pts / months) (Figures 6a, 6b, **p<0.001). Importantly, the carriers of the rsll538707-G also show an extended lifespan with a mean disease duration of ~ 50.1 months compared to the rsll538707-A carriers whose disease duration is 1 year shorter on average (mean disease duration ~ 38.5 month) (Figure 6c, *p<0.05). Overall, this data confirms that the R119G mutation has a protective role in ALS and demonstrates that ALS patients that do carry the variant have a significant slower rate of progression.

[0198] To determine whether R119G mutation has an effect on CREB3 transcriptional activity, a reporter based assay was designed. To this aim, SEC24A was identified as a canonical CREB3 target gene with multiple CRE-responsive elements in its promoter region (Figure 6d), and a reporter plasmid placing luciferase downstream the SEC24A promoter was generated (Figure 6e). Co-transfection of a wild type CREB3 expression plasmid increased SEC24A promoter activity by a factor of 2, confirming SEC2A4 as a bona fide CREB3 target gene (Figure 6e). Importantly, co-transfection of a CREB3- R119G expression plasmid further enhanced the SEC24A promoter activity by 48% (Figure 6e). Thus, the ALS protective CREB3 R119G variant confers a higher transcriptional activity on its target genes compared to the common CREB variant.

[0199] 6 - Evaluation of the effects of CREB3 :ic variants on cell survival

[0200] This experiment was conducted to test whether CREB3 and its rare variants can protect cells against an oxidative stress. HEK 293 cells were chosen as cellular model and an oxidative stress was induced by 4- hydroxynonenal (4HNE). The common variant (WT) and three rare variants (R119G, R155W and R168H) were tested in this experiment.

[0201] These results indicate that CREB3 rare variants confer protection against 4HNE-induced cellular toxicity onto HEK 293 cells (Figure 7).

[0202] It was shown that for all tested CREB3 variants, the level of cellular mortality observed after 4HNE treatment was lower than for the CREB3 common variant. This confirms that expression of CREB3 rare variants is associated with cellular survival in the context of oxidative stress, that is a relevant mechanism in the context of degeneration in ALS and other neurodegenerative diseases. Comparing the mean intensity of CellROX* fluorescence between each condition and assess the type of cellular death (apoptosis, necrosis) implicated in cellular mortality through the cytometry gating strategy described beforehand was also performed.

[0203] Example 7 - Evaluation of the effects of CREB3 genetic variants on the proteolytic activation of CREB3 protein and on downstream expression of CREB3 target genes

[0204] To exert its transcriptional activity, CREB3 needs to be transported from the endoplasmic reticulum (ER) to the Golgi apparatus where it is cut by proteases, allowing the release of the N-terminal fragment that migrates to the nucleus. The level of CREB3 transcriptional activity might therefore be modulated either by its ability to be cleaved and released, and / or its ability to bind the chromatin and recruit the transcriptional machinery. This experiment was designed to test the effect of CREB3 variants on the cleavage of CREB3 protein, using HEK293 cells treated or not with brefeldin-A (BFA) that induced ER stress and apoptosis.

[0205] This experiment shows that CREB3 rare variants allow a greater cleavage of CREB3 protein, that must result in increasing downstream gene expression (Figure 8). In most conditions (3 variants out of 4), BFA further enhances CREB3 cleavage in accordance with a proteolytic activation of CREB3 upon ER stress.

[0206] Western blot experiments and testing proteolytic cleavage inhibitors (PF-429242, 10 pM, added to cell culture media 24h before BFA-induced stress) are planned to verify whether CREB3 cleavage effect is mediated by SIP Golgi-membrane protease as reported or not. In qPCR experiments, it is expected to observe an increased expression of CREB3 target genes (such as SEC23A, SEC24A, USO1, ATG5, ARF4, COPB1). 8 - Identification of the clinical and bi of CREB3 rare variants carriers

[0207] Clinical, biochemical and proteomic data of human participants were integrated in order to identify the clinical and biological signature for CREB3 rare variants towards a better understanding of the downstream CREB3 mechanisms and evaluation of its general contribution to human health.

[0208] It was observed a specific blood proteomic signature for CREB3R119Gcarriers compared to CREB3V Tcarriers that highlights the decreased expression of four proteins such as MANSC1, ELAVL4, LAYN and CFHR4, among others (Figure 9). Linear regression analysis projected to all variant carriers revealed that ELAVL4 expression seems to be decreased in most of other genetic status except CREB3G346Rand CREB3R155Wcarriers.

[0209] It is expected to identify a specific blood proteomic signature associated with individual CREB3 carrier status that could help to understand the mechanisms underlying the neuroprotective effects of the variant expression. Using WGCNA, it was sought to find common regulated pathways that are more / less activated in CREB3 rare variant carriers. Extending this analysis to all CREB3 rare variant carriers is planned to see if this effect is conserved across other variants.

[0210] Using the same database, Odd Ratio of developing different neurological diseases (DEM, PD, AD, VaD, FTD) will also be calculated depending on the CREB3 carrier status to determine whether the protective effect observed in ALS could also be projected to other neurological diseases. Subclustering the blood proteomic analysis to each neurological diseases available in the database is also planned to build a disease specific proteogenomic signature and find common implicated actors or pathways.

Claims

CLAIMS1. A composition comprising a c-AMP Response Element-Binding 3 (CREB3) protein comprising or consisting of an amino acid sequence having at least 80% sequence identity with SEQ ID NO: 1, or a nucleic acid molecule encoding said CREB3 protein, for use as a medicament.

2. A composition for use according to claim 1, wherein said CREB3 protein comprises or consists of an amino acid sequence having at least 80% sequence identity with SEQ ID NO: 1 and having at least one mutation with respect to SEQ ID NO: 1.

3. A composition comprising a CREB3 protein as defined in claim 1 or 2 or a nucleic acid molecule encoding said CREB3 protein, for use in the prevention or the treatment of a disease associated with a degeneration of motor neurons.

4. A composition for use according to claim 3, wherein the disease associated with a degeneration of motor neurons is selected from the group consisting of amyotrophic lateral sclerosis (ALS), primary lateral sclerosis (PLS), hereditary spastic paraplegia (HSP), progressive bulbar palsy (PBP), spinal muscular atrophy (SMA), X-linked spinobulbar muscular atrophy (SBMA; Kennedy disease), spinal muscular atrophy with respiratory distress type 1 (SMARD1), postpolio syndrome (PPS), congenital SMA with arthrogryposis, progressive muscular atrophy (PMA), Alzheimer's disease, Parkinson's disease, Huntington's disease and Spinocerebellar Ataxia (SCA).

5. A composition for use according to any one of claims 3 to 4, wherein the disease associated with a degeneration of motor neurons is a motor neurons disease (MND).

6. A composition for use according to claim 5, wherein the MND is selected from the group consisting of ALS, PLS, HSP, PBP, SMA, SBMA, SMARD1, PPS, congenital SMA with arthrogryposis and PMA, preferably ALS.

7. A composition for use according to any one of claims 2 to 6, wherein said mutation is located in position 119 with respect to SEQ ID NO: 1.

8. A composition for use according to any one of claims 2 to 7, wherein said mutation is a substitution, more preferably a substitution R119G with respect to SEQ ID NO: 1.

9. A composition for use according to any one of claims 2 to 8, wherein said CREB3 protein comprises or consists of the amino acid sequence SEQ ID NO: 4.

10. A composition for use according to any one of claims 1 to 9, wherein said nucleic acid molecule is a genetic vector, preferably a viral vector.

11. A composition for use according to any one of claims 1 to 10, wherein said nucleic acid molecule comprises or consists of a complementary DNA (cDNA), a nucleoside-modified messenger RNA (modRNA) or a stabilized mRNA, encoding said CREB3 protein.

12. Use of a CREB3 protein as defined in claim 1 or 2 as a biomarker for the in vitro / ex vivo stratification of patients suffering, or susceptible of suffering, from a disease associated with a degeneration of motor neurons, wherein the presence of the expression of said CREB3 protein in biological samples of said patients indicates that said patients belong to a group of patients with a reduced risk of developing said disease and / or a reduced rate of progression of said disease, and the absence of the expression of said CREB3 protein in biological samples of said patients indicates that said patients belong to a group of patients with an increased risk of developing said disease and / or an increased rate of progression of said disease.

13. An in vitro / ex vivo method for the prognosis of a patient suffering from a disease associated with a degeneration of motor neurons comprising a step of detecting the presence or the absence of the expression of a CREB3 protein as defined in claim 1 or 2 in a biological sample from said patient, wherein the presence of the expression of said CREB3 protein in the biological sample of said patient indicates that said patient has a reduced rate of progression of said disease, and the absence of the expression of said CREB3 protein in the biological sample of said patient indicates that said patient has an increased rate of progression of said disease.

14. A nucleic acid molecule encoding a CREB3 protein comprising or consisting of an amino acid sequence having at least 80% sequence identity with SEQ ID NO: 1.

15. A nucleic acid molecule according to claim 14, wherein said nucleic acid molecule comprises or consists of the nucleic acid sequence SEQ ID NO: 3.

16. A nucleic acid molecule according to claim 14, wherein said nucleic acid molecule comprises or consists of the nucleic acid sequence SEQ ID NO: 6.

17. A nucleic acid molecule according to any one of claims 14 to 16, wherein said nucleic acid molecules is a genetic vector, preferably a viral vector.

Images