RELATED APPLICATIONS

This application claims the benefit of U. S. Provisional Application No.

62/517,567, filed on lune 9, 2017. The entire teachings of the above applications are incorporated herein by reference.

GOVERNMENT SUPPORT

This invention was made with government support under Grant Nos.

CA009216 and NS089742 awarded by the National Institutes of Health. The government has certain rights in the invention.

BACKGROUND OF THE INVENTION

Amyotrophic lateral sclerosis (ALS) is an adult onset neurodegenerative disease characterized by the loss of upper and lower motor neurons and muscle atrophy. Patients become progressively paralyzed and develop difficulty speaking, swallowing, and eventually breathing. Survival is typically limited to 2-5 years from the time of onset, and current treatment options remain limited. About 90% of cases are seemingly "sporadic" without a family history of disease, and about 10% are familial. Hundreds of distinct variants in more than a dozen genes, many of which act with high penetrance, can increase a person's risk of developing ALS [1, 2].

The most common genetic contributor to ALS is a hexanucleotide (GGGGCC) repeat expansion within the first intron of C90RF72 [3, 4]. Carriers of the C90RF72 expansion can also present with frontotemporal dementia (FTD), which is characterized by frontotemporal lobar degeneration (FTLD) of the brain. In many cases, these initially diverse diagnoses can progress toward the inclusion of neurological features from each condition, leading many to believe they are spectrums of the same disorder [5]. In addition, both diseases can be characterized by the presence of TDP -43 -positive inclusions [6].

Three distinct mechanisms have been proposed for how the C90RF72 expansion contributes to the development of ALS and FTD. First, C90RF72-ALS brains display reduced abundance of C90RF72 transcripts, suggesting that a loss-of- function mechanism may contribute to disease [3], Although complete loss of C90RF72 in mice leads to fatal autoimmunity and changes in microglia, no obvious signs of neurodegeneration or neural dysfunction have yet been reported in these animals [7-9]. Second, mutant transcripts containing the GGGGCC repeats form intranuclear RNA foci that may sequester RNA binding proteins and lead to nucleolar stress [3, 10]. Finally, dipeptide repeat proteins (DPRs) were unexpectedly found to be translated from both sense and antisense transcripts containing these repeats [1 1]. Several DPRs are toxic when overexpressed in model systems [12-15], and have been shown to affect diverse cellular pathways, including RNA processing and

nucleocytoplasmic transport [1, 5, 16].

The transcriptional response that occurs in various brain regions in ALS and FTD patients has the potential to provide useful insights into whether genetic subgroups of patients display common or divergent mechanisms, and for validating proposed mechanisms through which mutations act.

SUMMARY OF THE INVENTION

The present invention relates to the identification of a C90RF72 gene signature for ALS and methods of inducing the C90RF72 gene signature for ALS in neurons. Disclosed herein are methods of using the C90RF72 gene signature for screening for neurodegenerative disorders, diagnosing neurodegenerative disorders, and identifying agents for treating neurodegenerative disorders.

The invention disclosed herein relates, in some embodiments, to methods of inducing a C90RF72 gene signature (e.g., a C90RF72-ALS gene signature) in a neuron. The method comprises contacting a neuron with an agent, thereby inducing the C90RF72 gene signature in the neuron.

In some embodiments the agent is a dipeptide repeat protein (DPR) (e.g., a synthetic DPR). In some embodiments the DPR is selected from the group consisting of poly-GA, poly-GP, poly-GR, and combinations thereof. In certain embodiments the DPR is poly-GR. In some embodiments the agent is selected from the group consisting of a proteasome inhibitor (e.g., MG-262 and/or thiostreptin), an HSP90 inhibitor (e.g., geldanamycin, monorden, alvespimycin, trichostatin A, tanespimycin, and/or parthenolide), and a translation inhibitor (e.g., puromycin).

In some embodiments the neuron is a human neuron (e.g., a stem cell-derived neuron). In some embodiments, the neuron is a non-human neuron.

Also disclosed herein are methods of screening one or more test agents to identify candidate agents which modulate HSF l activation. The methods comprise

(a) contacting a neuron exhibiting a C90RF72- ALS gene signature with a test agent;

(b) measuring level or activity of HSFl or one or more HSFl target genes in the contacted neuron; and (c) identifying the test agent as a candidate agent that modulates HSFl activation if the level or activity of HSFl or the one or more HSFl target genes in the neuron is decreased as compared with the level or activity of HSF l or the one or more HSFl target genes in the neuron that has not been contacted with the test agent.

In some embodiments the one or more HSFl target genes are selected from the group consisting of BAG2, CHORDC 1, CRYAB, DEDD2, DNAJB 1, DNAJB4, FKBP4, HSPA1A, HSPA1B, HSPB1, SERPINH1, and STIP1.

Also disclosed herein are methods of identifying a candidate agent that inhibits HSFl pathway activation. The methods comprise contacting a neuron with a proteasome inhibitor; contacting the proteasome inhibitor contacted neuron with a test agent; measuring activation of the HSFl pathway in the presence of the test agent; and identifying a candidate agent that inhibits HSF l pathway activation, wherein the test agent is the candidate agent for inhibiting HSF l pathway activation if the test agent decreases activation of the HSFl pathway in the proteasome inhibitor contacted neuron.

Also disclosed herein are methods of identifying a candidate agent that inhibits HSFl pathway activation. The methods comprise contacting a neuron with a poly-GR DPR; contacting the poly-GR DPR contacted neuron with a test agent; measuring activation of the HSF l pathway in the presence of the test agent; and identifying a candidate agent that inhibits HSFl pathway activation, wherein the test agent is the candidate agent for inhibiting HSF 1 pathway activation if the test agent decreases activation of the HSF 1 pathway in the poly-GR DPR contacted neuron.

Also disclosed herein are methods of detecting a C90RF72 gene signature (e.g., a C90RF72-ALS gene signature) in a subject, comprising obtaining a sample from the subject; and detecting whether the C90RF72 gene signature is present in the sample.

In some embodiments the C90RF72 gene signature comprises increased expression of one or more genes selected from the group consisting of BAG2, CHORDC1, CRY B, DEDD2, DNAJB1, DNAJB4, FKBP4, HSPA1A, HSPA1B, HSPB1, SERPINH1, and STIP1.

Also disclosed herein are methods of screening for ALS in a subject. The methods comprise screening a sample from the subject to detect a C90RF72 -ALS gene signature, wherein the presence of the C90RF72-ALS gene signature is indicative of the subject having ALS.

Also disclosed herein are methods of treating a neurodegenerative disorder in a subject in need thereof. The methods comprise administering an effective amount of an agent that inhibits HSF1 pathway activation in the subject, thereby treating the neurodegenerative disorder.

In some embodiments inhibition of HSF 1 pathway activation comprises inhibition of one or more genes selected from the group consisting of BAG2, CHORDC1, CRYAB, DEDD2, DNAJB1, DNAJB4, FKBP4, HSPA1A, HSPA1B, HSPB1, SERPINH1, and STIP1.

In some embodiments the neurodegenerative disorder is a disorder that exhibits increased HSF1 pathway activation. In some aspects the neurodegenerative disorder is a disorder that exhibits a C90RF72 gene signature (e.g., is a C90RF72- associated neurodegenerative disorder). In some embodiments the neurodegenerative disorder is selected from the group consisting of amyotrophic lateral sclerosis (ALS), frontotemporal dementia (FTD), frontotemporal lobar degeneration (FTLD), and combinations thereof. In certain embodiments the neurodegenerative disorder is amyotrophic lateral sclerosis (ALS) or amyotrophic lateral sclerosis (ALS) combined with frontotemporal dementia (FTD) and/or frontotemporal lobar degeneration (FTLD). Also disclosed herein are methods of treating a neurodegenerative disorder in a subject in need thereof. The methods comprise administering an effective amount of an agent that inhibits dipeptide repeat protein activation of the HSF1 pathway, thereby treating the neurodegenerative disorder.

In some embodiments the agent inhibits the activity of one or more DPRs selected from the group consisting of poly-GA, poly-GP, and poly-GR. In certain embodiments the agent inhibits the activity of poly-GR DPR.

In some embodiments the neurodegenerative disorder is a disorder that exhibits increased HSF1 pathway activation. In some embodiments the

neurodegenerative disorder is a disorder that exhibits a C90RF72 gene signature (e.g., is a C9(9i? 72-associated neurodegenerative disorder). In some embodiments the neurodegenerative disorder is selected from the group consisting of amyotrophic lateral sclerosis (ALS), frontotemporal dementia (FTD), frontotemporal lobar degeneration (FTLD), and combinations thereof. In certain embodiments the neurodegenerative disorder is amyotrophic lateral sclerosis (ALS) or amyotrophic lateral sclerosis (ALS) combined with frontotemporal dementia (FTD) and/or frontotemporal lobar degeneration (FTLD).

BRIEF DESCRIPTION OF THE DRAWINGS

The patent or application file contains at least one drawing executed in color. Copies of this patent or patent application publication with color drawings will be provided by the Office upon request and payment of the necessary fee.

FIGS. 1A-1G demonstrate identification of specific cellular pathways perturbed in sporadic ALS and C90RF72 -ALS. FIG. 1A provides a diagram of RNA-seq datasets obtained from the frontal cortex and cerebellum by Prudencio et al. FIG. IB shows a comparison of the significant (FDR <0.05) differentially expressed transcripts in C90RF72-ALS (C9-ALS) and sporadic ALS (sALS). Note, there were no common transcripts between C90RF72-ALS and sporadic ALS in either brain region. FIG. 1C shows a comparison of the differentially expressed transcripts by brain region. FIG. ID provides a correlation of the fold change (log2) of changed transcripts in C90RF72 -ALS that were common to both brain regions (Spearman's R ² ). A gene ontology (GO) analysis revealed cellular processes affected in C90RF72-ALS (FIG. IE) and sALS (FIG. IF). FIG. 1G provides protein-protein interaction analysis of proteins encoded by the transcripts changed in C90RF72 -ALS revealed a protein chaperone network.

FIGS 2A-2B demonstrate activation of HSFl in C90RF72-ALS, FTD, and combined ALS/FTD patients. FIG. 2A provides quantitative real-time PCR (qRT- PCR) for HSFl target genes in the frontal cortex of sporadic and C90RF72- associated disease (n=56 C90RF72-ALSfFJO, n=46 sporadic ALS/FTD, n=9 controls) (one-way ANOVA with Bonferonni post-hoc test for multiple comparisons, * p<0.05, ** p<0.01, *** p<0.001). No significant changes were detected between the sporadic cases and controls. FIG. 2B shows qRT-PCR for HSFl in the frontal cortex and cerebellum of these same cases.

FIGS. 3A-3D demonstrate DPRs induce expression of C90RF72 signature transcripts in human neurons. FIG. 3 A provides a diagram of generation of human neurons from stem cells. FIG. 3B provides a viability dose response curve of human stem cells and stem-cell derived neurons exposed to various DPRs (n=6). FIGS. 3C- 3D provide qRT-PCR of C9<9i? 72-chaperome transcripts (FIG. 3C) and HSFl (FIG. 3D) in human stem cell-derived motor neurons following treatment with DPRs (poly- GA, poly-GP, poly-GR) or a scrambled poly-GAPR (5 uM for 24 h) normalized to control (DMSO-treated) neurons (mean±SD, n=3, one-way ANOVA with Dunnett's post-hoc test for upregulated genes in DPR-treated neurons compared to control, * p<0.05, ** p<0.01, *** p<0.001, **** p<0.0001).

FIGS. 4A-4G demonstrate detection of C9ftKF72-associated transcriptional changes in gain-of-function Drosophila models. FIG. 4 A shows control UAS- (G4C2)8 and expanded UAS-(G4C2)49 transgenes were expressed in the adult fly nervous system using the drug-inducible Gal4 driver, elavGS, for 16d. Quantitative PCR (qPCR) analysis of endogenous dHSFl and HSFl -regulated genes revealed significant upregulation with (G4C2)49 expression compared to (G4C2)8 controls. Differences in expression are likely underestimated as the analyses include neuronal and non-neuronal tissue while (G4C2)n was expressed only in neurons. FIG. 4B provides qPCR analysis of a dHSFl overexpression mutant fly line shows endogenous HSF is upregulated approximately 2-fold in mutant flies compared to control. FIG. 4C shows Western immunoblot analysis of expression of a control UAS-LacZ transgene confirmed that the HSF OE mutant did not affect the Gal4 UAS expression system. FIG. 4D shows (G4C2)49 was expressed in the optic system of control animals or HSF OE animals using Gmr-Gal4. (G4C2)49 causes toxicity seen by pigment loss, reduced eye size, and disruptions in the normal ommatidial organization. In HSF OE animals, toxicity of (G4C2)49 is enhanced - animals have increased pigment loss, increased disruption of ommatidial organization, and further reduced eye size. Expression of control (G4C2)8 in the fly optic system (Gmr-Gal4) of control and HSF OE animals does not affect the external eye. FIG. 4E provides quantification of the external eye degenerative phenotype caused by (G4C2)49 expression shows enhancement in HSF OE animals versus control animals to be consistent and statistically significant (n=6). Animals received a score between 0 (WT eye) and 8 (lethality caused by extreme degeneration in the optic system). (G4C2)49 expression causes an average score of 4 in controls. FIG. 4F provides Gmr-GAL4 driven expression of (GR)36 shows toxicity in control scenarios like (G4C2)49. HSF OE in these animals also causes enhanced toxicity (increased pigment loss, increased disruption of ommatidial organization, and reduced eye size). FIG. 4G provides quantification of the external eye degenerative phenotype caused by (GR)36 expression shows enhancement in HSF OE animals versus control animals to be consistent and statistically significant (n=7). Animals received a score between 0 (WT eye) and 11 (lethality caused by extreme degeneration in the optic system) while (GR)36 causes an average score of 5 in controls. (All plots: mean +/- SD, unpaired, student's t-test, * /? < 0.05, ** p < 0.0\, *** /? < 0.001, **** /? < 0.0001.)

FIGS. 5A-5F demonstrate bioinformatic method comparison for gene expression analysis of C90RF72-associated ALS and sporadic ALS (sALS) in the frontal cortex and cerebellum. FIGS. 5A-5D provide gene density plots comparing the expression levels of differentially expressed transcripts as determined by the prior double cut-off method (|log2 fold change] > 2 and p-value <0.05) and the FDR method (FDR <0.05), which was used in FIG. 1 and subsequent analysis. Note, no significant changes were detected in the sALS cerebellum using with FDR method. FIG. 5E provides Venn diagrams demonstrating considerable differences in the designated disease-associated transcripts between these bioinformatics methods in both brain regions. FIG. 5F provides table of the number of differentially expressed transcripts as determined by Prudencio et al., the double cut-off method and the FDR method used here.

FIGS. 6A-6B demonstrate gene networks in C90RF72-ALS and sporadic ALS. FIG. 6A shows protein-protein interaction network derived from differentially expressed transcripts in C90RF72-ALS cerebellum. Those transcripts that are differentially expressed in both the frontal cortex and the cerebellum in C90RF72- ALS are highlighted by dashed yellow circles, and predominantly consist of heat shock proteins and protein chaperones. FIG. 6B shows protein-protein interaction network derived from differentially expressed transcripts in the sporadic ALS cortex.

FIGS. 7A-7B demonstrate activation of HSF1 in C90RF72-ALS, FTLD, and combined ALS/FTLD patients. FIG. 7A provides quantitative real-time PCR (qRT- PCR) for HSF1 target genes in the cerebellum of sporadic and C90RF72-associated disease (n=56 C90RF72-ALS/FTLD, n=42 sporadic ALS FTLD, n=7 controls) (oneway ANOVA with Bonferonni post-hoc test for multiple comparisons * p<0.05, ** p<0.01, *** p<0.001. Note, no significant changes were detected between the sporadic cases and controls. FIG. 7B shows correlation of HSF1 levels and HSF1 target gene levels in the frontal cortex and cerebellum in C90RF72-ALS/FTLD. Spearman's R ² values are plotted for each target gene, error bars denote 95% confidence interval, p-value <0.0001 in all cases.

FIG. 8 demonstrates poly-GR expression results in the upregulation of heat shock response genes and dHSFl in the adult fly nervous system. UAS-(GR)36 was expressed in the adult fly nervous system using the drug-inducible Gal4 driver, elavGS, for 16d. qPCR analysis of endogenous HSF1 -regulated genes and dHSFl revealed significant upregulation of the Drosophila orthologs of many of the genes identified in patient studies, suggesting that poly-GR is contributing to the altered transcriptome in C90RF72-ALS FTLD patients. Control animals did not express a transgene. Differences in expression are likely underestimated as the analyses include neuronal and non-neuronal tissue while (GR)36 was expressed only in neurons. (n=6, mean +/- SD, unpaired, student's t-tests, p-value * < 0.05, ** < 0.01).

FIG. 9 provides external eye quantification scale for (GR)36 animals. For quantification of the enhancement effects of increased expression of HSF in the eye, (GR)36 animals received a score between 0 (normal eye) and 11 (extreme toxicity causing lethality). Across multiple studies control Gmr-GAL4 > (GR)36 animals receive a score between 5-6.

FIGS. 10A-10I demonstrate gene expression analysis of C90RF72-associated ALS (C9-ALS) and sporadic ALS (sALS) patients. FIG. 10A provides a diagram of RNA-seq datasets grouped by disease state and bioinformatics analysis and a Table of the number of brain samples used for RNA-seq obtained from the frontal cortex and cerebellum by Prudencio et al. FIGS. 10B and IOC provide gene density plots comparing the expression levels of significantly changed transcripts as determined by the double cut-off method (FIG. 10B) and FDR method (FIG. IOC) compared to the distribution of all transcripts. FIG. 10D provides Venn diagrams comparing the number of differentially expressed transcripts as determined by double cut-off method and the FDR method based on disease state. Note no significantly changed transcript were detected in the cerebellum of C90RF72-ALS by the FDR method. FIG. 10E provides Venn diagrams comparing the differentially expressed transcripts of C90RF72-ALS (C9-ALS) and sporadic ALS (sALS). Note there was no common transcripts between C90RF72-ALS and sporadic ALS in either brain region based on the FDR method. FIGS. 10F and 10G show gene ontology (GO) term biological process clusters based on FDR method differentially expressed transcript from the cortex of C90RF72-ALS and sporadic ALS. FIG. 10H shows protein-protein interaction network of differentially expressed transcripts from C9-ALS cortex based on the FDR method. FIG. 101 provides correlation of the log2 fold change (FC) gene expression of transcripts that were significantly changed (FDR <0.05) in both the cortex and the cerebellum in C90RF72-ALS.

FIGS. 11A-11D demonstrate activation of HSFl in C90RF72-ALS/FTO. FIGS. 11 A and 1 IB provide quantitative RT-PCR of HSFl levels in the frontal cortex (FIG. 11 A) and cerebellum (FIG. 1 IB) from samples of patients with ALS and/or FTD harboring mutations in C90RF72 vs. ALS and/or FTD patients without C90RF72 mutations and controls. FIG. 11C provides representative

immunohistochemistry for HSFl in cerebellum samples from a control case, sporadic ALS case, a C90RF72-ALS/FTD case, and a control stain of the same C90RF72- ALS/FTD case performed with the secondary antibody but without the HSFl primary antibody. Scale bar is 20 uM. FIG. 1 ID shows qPCR of HSFl targets genes in the cerebellum of C90RF72 (C9) or sporadic (non-C9) ALS and/or FTD cases and controls cases.

FIGS. 12A-12E demonstrate that the connectivity map connects the

C90RF72-ALS gene signature to compounds that inhibit the proteasome. FIG 12A provides a diagram of connectivity map analysis. FIG. 12B provides volcano plot of compounds associated with the C90RF2-ALS cortex signature genes. FIG. 12C provides volcano plot of compounds associated with the C90RF72-ALS cerebellum signature genes. Compounds found in both sets are highlighted in bold. FIG. 12D shows connectivity map analysis yields hit compounds that perturb proteostasis through various mechanisms, as well as other compounds with their known target of action provided. C90RF72-ALS connectivity scores are in parentheses. FIG. 12E shows heat map of the connectivity scores of compounds that affect pathways implicated in ALS. Significantly connected compounds (p<0.001) are highlighted with an asterisk.

FIGS. 13A-13D demonstrate inhibition of the proteasome recapitulates the C90RF72-ALS gene signature in neurons in vitro. FIG. 13A provides schematic for the generation of purified neurons from stem cells and subsequent drug treatment with a proteasome inhibitor, MG132. FIG. 13B shows gene ontology (GO) term biological process clusters based on transcripts induces by MG132 treatment in motor neurons, which is similar to the C90RF72-ALS GO term clusters. FIG. 13C provides transcripts levels of all coding C90RF72-ALS signature genes. Genes that significantly changed in expression after proteasome inhibition at both doses are bolded (adjusted p-value <0.0001). FIG. 13D shows correlation of log2 fold change in chaperome-related genes in C90RF72-ALS cortex, sporadic ALS (sALS) cortex, and C90RF72-ALS cerebellum compared to changes induced after proteasome inhibition in motor neurons in vitro. Linear regression and R2 values are displayed.

FIGS. 14A-14F demonstrate activation of the HSF1 pathway by C90RF72 hexanucleotide repeat expansion-associated dipeptide repeat proteins. FIG. 14A shows cell viability in stem cells following synthetic dipeptide repeat protein (DPR) treatment. Note GAPR is a scrambled control peptide. Cell viability in neurons following DPR treatment. FIG. 14B shows mislocalization of TDP-43 after DPR treatment. Correlation between DAPI (nuclear) and TDP-43 immunofluorescence in motor neurons following DPR treatment. FIG. 14C provides quantitative RT-PCR of heat-shock related C9-ALS signature genes in neurons following DPR treatment. Normalization is performed to the geometric mean of GAPDH and actin. FIG. 14D shows correlation between log2 fold changes of tested C9-ALS signature genes in poly- PR treated neurons and log2 fold changes in C9-ALS frontal cortex. FIG. 14E provides a model of neuron proteostasis in healthy neurons. FIG. 14F shows putative mechanisms of disruption of proteostasis in C90RF72-ALS/FTD neurons

DETAILED DESCRIPTION OF THE INVENTION

The present invention relates to the characterization of a specific C90RF72 gene signature, specifically the C90RF72 gene signature with respect to ALS patients. Disclosed herein are methods for inducing the C90RF72 gene signature in neurons, as well as methods of using the C90RF72 gene signature to screen for neurodegenerative disorders, diagnose neurodegenerative disorders, and identify agents for treating neurodegenerative disorders.

In certain embodiments, the inventions disclosed herein relate to a C90RF72 gene signature. The C90RF72 gene signature as described herein is a C90RF72 gene signature that may be identified in patients having amyotrophic lateral sclerosis (ALS), frontotemporal dementia (FTD), and/or frontotemporal lobar degeneration (FTLD). FTD may be characterized by FTLD of the brain. The C90RF72 gene transcript exhibits increased activation of HSF1, as well as increased expression of one or more HSF1 target genes. The one or more HSF1 targets genes include BAG2, CHORDC1, CRYAB, DEDD2, DNAJB1, DNAJB4, FKBP4, HSPA1A, HSPA1B, HSPB1, SERPINH1, and/or STIP1.

In some embodiments, the inventions disclosed herein relate to methods of inducing the C90RF72 gene signature, e.g., the C90RF72-ALS gene signature. In some aspects a C90RF72 gene signature is induced in a neuron. The C90RF72 gene signature may be induced by contacting the neuron with an agent. The neuron may be a human neuron or a non -human neuron. In some aspects the neuron is a human neuron, e.g., a stem cell-derived neuron. In some aspects the agent induces an HSF1 response, e.g., activation of HSF1. In some aspects the agent is a dipeptide repeat protein (DPR), and in some embodiments is a synthetic DPR. The DPR may be selected from the group consisting of poly-glycine-arginine (poly-GR), poly-glycine-alanine (poly-GA), poly- glycine-proline (poly-GP), and combinations thereof. In some embodiments a C90RF72 gene signature is induced in a neuron by contacting the neuron with poly- GR DPR. In some embodiments a C90RF72 gene signature is induced in a neuron by contacting the neuron with poly-GA DPR. In some embodiments a C90RF72 gene signature is induced in a neuron by contacting the neuron with poly-GP DPR. In some aspects treatment of a neuron with one or more DPRs results in increased expression of one or more C90i? 77-chaperome transcripts (e.g., GAPDH, ACTIN, SERPINH1, STIP1, BAG3, CHORDC 1, HSPA1B, and HSPA6) and/or increased expression of HSF 1.

In some aspects the agent is selected from the group consisting of a proteasome inhibitor, an HSP90 inhibitor, a translation inhibitor, and combinations thereof. In some embodiments the agent is a proteasome inhibitor (e.g., MG-132, MG-262, or thiostreptin). In some embodiments the agent is a translation inhibitor (e.g., puromycin). In some embodiments the agent is an HSP90 inhibitor (e.g., geldanamycin, monorden/radcicol, alvespimycin/17-DMAG, trichostatin A, tanespimycin/17-AAG, or parthenolide). In some aspects the agent is selected from the group consisting of 15-delta prostaglandin J2, securinine, astemizole, thioridazine, and nordihydroguaiaretic acid. In some aspects the agent is any combination of agents disclosed herein. In some aspects treatment of a neuron with one or more proteasome inhibitors results in increased expression of one or more C90RF71- chaperome transcripts (e.g., DNAJB4, CHORDC1, SERPINHl, BAG3, FKBP4, STIP1, HSPA6, HSPB 1, HSPA1B, DNAJB 1, and HSPA1A).

In some embodiments, the inventions disclosed herein relate to methods of screening subjects or patients for a C90RF72 gene signature. In some aspects methods include detecting a C90RF72 gene signature in a subject by obtaining a sample from the subject and screening the sample to detect if the gene signature is present in the sample. In some aspects the C90RF72 gene signature is a C90RF72- ALS gene signature. In some aspects the C90RF72 gene signature exhibits increased expression of one or more genes selected from the group consisting of BAG2, CHORDC1, CRYAB, DEDD2, DNAJB1, DNAJB4, FKBP4, HSPA1A, HSPA1B, HSPB1, SERPINH1, and STIP1. In some aspects the C90RF72 gene signature exhibits increased activity of HSFl . Detection of a C90RF72 gene signature in a patient may be indicative of the patient having ALS, FTD, and/or FTLD.

In some aspects a method of screening for ALS in a subject comprises screening a sample from the subject to detect a C90RF72 -ALS gene signature, wherein the presence of the gene signature is indicative of the patient having ALS. The patient may further exhibit symptoms of FTD and/or FTLD.

In some embodiments, the inventions disclosed herein relate to methods of screening one or more test agents to identify candidate agents which modulate HSF l activation in a cell. "Modulate" as used herein means to decrease (e.g., inhibit, reduce) or increase (e.g., stimulate, activate) a level, response, property, activity, pathway, or process. A "modulator" is an agent capable of modulating a level, response, property, activity, pathway, or process. In some aspects a modulator is an inhibitor or blocker of HSFl activation. The term "agent" as used herein means any compound or substance such as, but not limited to, a small molecule, nucleic acid, polypeptide, peptide, drug, ion, etc. An "agent" can be any chemical, entity or moiety, including without limitation synthetic and naturally-occurring proteinaceous and non- proteinaceous entities. In some embodiments, an agent is nucleic acid, nucleic acid analogues, proteins, antibodies, peptides, aptamers, oligomer of nucleic acids, amino acids, or carbohydrates including without limitation proteins, oligonucleotides, ribozymes, DNAzymes, glycoproteins, siRNAs, lipoproteins, aptamers, and modifications and combinations thereof.

In some embodiments a method of screening comprises (a) contacting a neuron (e.g., a stem cell-derived neuron) exhibiting a C90RF72-ALS gene signature with a test agent; (b) measuring level or activity of HSFl or one or more HSF l target genes in the contacted neuron; and (c) identifying the test agent as a candidate agent that modulates HSF l activation if the level or activity of HSF l or the one or more HSF l target genes in the neuron is decreased as compared with the level or activity of HSF l or the one or more HSFl target genes in the neuron that has not been contacted with the test agent. An HSFl target gene may be selected from the group consisting of BAG2, CHORDC1, CRYAB, DEDD2, DNAJB 1, DNAJB4, FKBP4, HSPA1A, HSPA1B, HSPB1, SERPINH1, DEDD2, and STIP1.

In some embodiments a method of identifying a candidate agent that inhibits HSFl pathway activation comprises: (a) contacting a neuron (e.g., a stem cell-derived neuron) with a proteasome inhibitor; (b) contacting the proteasome inhibitor contacted neuron with a test agent; (c) measuring activation of the HSFl pathway in the presence of the test agent; and (d) identifying a candidate agent that inhibits HSFl pathway activation, wherein the test agent is the candidate agent for inhibiting HSFl pathway activation if the test agent decreases activation of the HSFl pathway in the proteasome inhibitor contacted neuron.

In some embodiments a method of identifying a candidate agent that inhibits HSFl pathway activation comprises (a) contacting a neuron (e.g., a stem cell-derived neuron) with a dipeptide repeat protein (e.g., poly-GR DPR); (b) contacting the DPR contacted neuron with a test agent; (c) measuring activation of the HSFl pathway in the presence of the test agent; and (d) identifying a candidate agent that inhibits HSFl pathway activation, wherein the test agent is the candidate agent for inhibiting HSFl pathway activation if the test agent decreases activation of the HSFl pathway in the DPR contacted neuron.

The present invention also provides a method for disrupting HSFl activation. In some aspects HSFl activation is disrupted by genomic modification (e.g., using CRISPR/Cas or TALEN systems). CRISPR/Cas systems can employ a variety of Cas proteins (Haft et al. PLoS Comput Biol. 2005; l(6)e60). In some embodiments, the CRISPR/Cas system is a CRISPR type I system. In some embodiments, the

CRISPR/Cas system is a CRISPR type II system. In some embodiments, the CRISPR/Cas system is a CRISPR type V system.

In some embodiments, the inventions disclosed herein relate to methods of treating neurodegenerative disorders. In some aspects methods of treating a neurodegenerative disorder comprise blocking or inhibiting activation of the HSFl pathway. In some embodiments activation of the HSFl pathway is disrupted using genome editing (e.g., CRISPR/Cas). In some embodiments a method of treating a neurodegenerative disorder comprises administering an effective amount of an agent to a subject, wherein the agent inhibits or blocks HSF 1 pathway activation (e.g., in a neuron). In some aspects the agent inhibits DPR activation of the HSF1 pathway.

In some aspects blocking HSF 1 pathway activation or inactivating the HSF 1 pathway results in inhibiting (e.g., decreasing expression of) one or more HSF1 target genes. The HSF 1 target genes may be selected from the group consisting of BAG2, CHORDC1, CRYAB, DEDD2, DNAJB1, DNAJB4, FKBP4, HSPA1A, HSPA1B, HSPB1, SERPINH1, and STIP1. In some aspects the agent inhibits activation of the HSF 1 pathway by poly-GA DPR, poly-GP DPR, and/or poly-GR DPR.

In some embodiments the neurodegenerative disorder is a disorder that exhibits increased HSF1 pathway activation. In some aspects the neurodegenerative disorder is a disorder that exhibits a C90RF72gene signature (e.g., is a C90RF72- associated neurodegenerative disorder). In some embodiments the neurodegenerative disorder is selected from the group consisting of amyotrophic lateral sclerosis (ALS), frontotemporal dementia (FTD), frontotemporal lobar degeneration (FTLD), and combinations thereof. In some aspects the neurodegenerative disorder is ALS. In some aspects the neurodegenerative disorder is ALS in combination with FTD and/or FTLD.

For administration to a subject, the agents disclosed herein can be provided in pharmaceutically acceptable compositions. These pharmaceutically acceptable compositions comprise a therapeutically-effective amount of one or more of the agents, formulated together with one or more pharmaceutically acceptable carriers (additives) and/or diluents. As described in detail below, the pharmaceutical compositions of the present invention can be specially formulated for administration in solid or liquid form, including those adapted for the following: (1) oral administration, for example, drenches (aqueous or non-aqueous solutions or suspensions), gavages, lozenges, dragees, capsules, pills, tablets (e.g., those targeted for buccal, sublingual, and systemic absorption), boluses, powders, granules, pastes for application to the tongue; (2) parenteral administration, for example, by subcutaneous, intramuscular, intravenous or epidural injection as, for example, a sterile solution or suspension, or sustained-release formulation; (3) topical application, for example, as a cream, ointment, or a controlled-release patch or spray applied to the skin; (4) intravaginally or intrarectally, for example, as a pessary, cream or foam; (5) sublingually; (6) ocularly; (7) transdermally; (8) transmucosally; or (9) nasally. Additionally, agents can be implanted into a patient or injected using a drug delivery system. (See, for example, Urquhart, et al., Ann. Rev. Pharmacol. Toxicol. 24: 199-236 (1984); Lewis, ed. "Controlled Release of Pesticides and

Pharmaceuticals" (Plenum Press, New York, 1981); U.S. Pat. No. 3,773,919; and U.S. Pat. No. 35 3,270,960, content of all of which is herein incorporated by reference.) As used herein, the term "pharmaceutically acceptable" refers to those agents, materials, compositions, and/or dosage forms which are, within the scope of sound medical judgment, suitable for use in contact with the tissues of human beings and animals without excessive toxicity, irritation, allergic response, or other problem or complication, commensurate with a reasonable benefit/risk ratio.

As used herein, the term "pharmaceutically-acceptable carrier" means a pharmaceutically-acceptable material, composition or vehicle, such as a liquid or solid filler, diluent, excipient, manufacturing aid (e.g., lubricant, talc magnesium, calcium or zinc stearate, or steric acid), or solvent encapsulating material, involved in carrying or transporting the subject agent from one organ, or portion of the body, to another organ, or portion of the body (e.g., to one or more muscle cells or satellite cells). Each carrier must be "acceptable" in the sense of being compatible with the other ingredients of the formulation and not injurious to the subject. Some examples of materials which can serve as pharmaceutically-acceptable carriers include: (1) sugars, such as lactose, glucose and sucrose; (2) starches, such as corn starch and potato starch; (3) cellulose, and its derivatives, such as sodium carboxymethyl cellulose, methylcellulose, ethyl cellulose, microcrystalline cellulose and cellulose acetate; (4) powdered tragacanth; (5) malt; (6) gelatin; (7) lubricating agents, such as magnesium stearate, sodium lauryl sulfate and talc; (8) excipients, such as cocoa butter and suppository waxes; (9) oils, such as peanut oil, cottonseed oil, safflower oil, sesame oil, olive oil, corn oil and soybean oil; (10) glycols, such as propylene glycol; (11) polyols, such as glycerin, sorbitol, mannitol and polyethylene glycol (PEG); (12) esters, such as ethyl oleate and ethyl laurate; (13) agar; (14) buffering agents, such as magnesium hydroxide and aluminum hydroxide; (15) alginic acid; (16) pyrogen-free water; (17) isotonic saline; (18) Ringer's solution; (19) ethyl alcohol; (20) pH buffered solutions; (21) polyesters, polycarbonates and/or polyanhydrides; (22) bulking agents, such as polypeptides and amino acids (23) serum component, such as serum albumin, HDL and LDL; (22) C2-C12 alcohols, such as ethanol; and (23) other non-toxic compatible substances employed in pharmaceutical formulations. Wetting agents, coloring agents, release agents, coating agents, sweetening agents, flavoring agents, perfuming agents, preservative and antioxidants can also be present in the formulation. The terms such as "excipient", "carrier", "pharmaceutically acceptable carrier" or the like are used interchangeably herein.

The phrase "therapeutically-effective amount" as used herein means that amount of an agent, material, or composition comprising an agent described herein which is effective for producing some desired therapeutic effect in at least a sub- population of cells in an animal at a reasonable benefit/risk ratio applicable to any medical treatment. For example, an amount of an agent administered to a subject that is sufficient to produce a statistically significant, measurable decrease in the activation of HSFl .

The determination of a therapeutically effective amount of the agents and compositions disclosed herein is well within the capability of those skilled in the art. Generally, a therapeutically effective amount can vary with the subject' s history, age, condition, sex, and the administration of other pharmaceutically active agents.

As used herein, the term "administer" refers to the placement of an agent or composition into a subject (e.g., a subject in need) by a method or route which results in at least partial localization of the agent or composition at a desired site such that desired effect is produced. Routes of administration suitable for the methods of the invention include both local and systemic routes of administration. Generally, local administration results in more of the administered agents being delivered to a specific location as compared to the entire body of the subject, whereas, systemic

administration results in delivery of the agents to essentially the entire body of the subject.

The compositions and agents disclosed herein can be administered by any appropriate route known in the art including, but not limited to, oral or parenteral routes, including intravenous, intramuscular, subcutaneous, transdermal, airway (aerosol), pulmonary, nasal, rectal, and topical (including buccal and sublingual) administration. Exemplary modes of administration include, but are not limited to, injection, infusion, instillation, inhalation, or ingestion. "Injection" includes, without limitation, intravenous, intramuscular, intraarterial, intrathecal, intraventricular, intracapsular, intraorbital, intracardiac, intradermal, intraperitoneal, transtracheal, subcutaneous, subcuticular, intraarticular, sub capsular, subarachnoid, intraspinal, intracerebro spinal, and intrasternal injection and infusion. In preferred embodiments of the aspects described herein, the compositions are administered by intravenous infusion or injection.

As used herein, a "subject" means a human or animal (e.g., a mammal).

Usually the animal is a vertebrate such as a primate, rodent, domestic animal or game animal. Primates include chimpanzees, cynomologous monkeys, spider monkeys, and macaques, e.g., Rhesus. Rodents include mice, rats, woodchucks, ferrets, rabbits and hamsters. Domestic and game animals include cows, horses, pigs, deer, bison, buffalo, feline species, e.g., domestic cat, canine species, e.g., dog, fox, wolf, avian species, e.g., chicken, emu, ostrich, and fish, e.g., trout, catfish and salmon. Patient or subject includes any subset of the foregoing, e.g., all of the above, but excluding one or more groups or species such as humans, primates or rodents. In certain

embodiments of the aspects described herein, the subject is a mammal, e.g., a primate, e.g., a human. The terms, "patient" and "subject" are used interchangeably herein. A subject can be male or female.

# # #

One skilled in the art readily appreciates that the present invention is well adapted to carry out the objects and obtain the ends and advantages mentioned, as well as those inherent therein. The details of the description and the examples herein are representative of certain embodiments, are exemplary, and are not intended as limitations on the scope of the invention. Modifications therein and other uses will occur to those skilled in the art. These modifications are encompassed within the spirit of the invention. It will be readily apparent to a person skilled in the art that varying substitutions and modifications may be made to the invention disclosed herein without departing from the scope and spirit of the invention.

The articles "a" and "an" as used herein in the specification and in the claims, unless clearly indicated to the contrary, should be understood to include the plural referents. Claims or descriptions that include "or" between one or more members of a group are considered satisfied if one, more than one, or all of the group members are present in, employed in, or otherwise relevant to a given product or process unless indicated to the contrary or otherwise evident from the context. The invention includes embodiments in which exactly one member of the group is present in, employed in, or otherwise relevant to a given product or process. The invention also includes embodiments in which more than one, or all of the group members are present in, employed in, or otherwise relevant to a given product or process.

Furthermore, it is to be understood that the invention provides all variations, combinations, and permutations in which one or more limitations, elements, clauses, descriptive terms, etc., from one or more of the listed claims is introduced into another claim dependent on the same base claim (or, as relevant, any other claim) unless otherwise indicated or unless it would be evident to one of ordinary skill in the art that a contradiction or inconsistency would arise. It is contemplated that all embodiments described herein are applicable to all different aspects of the invention where appropriate. It is also contemplated that any of the embodiments or aspects can be freely combined with one or more other such embodiments or aspects whenever appropriate. Where elements are presented as lists, e.g., in Markush group or similar format, it is to be understood that each subgroup of the elements is also disclosed, and any element(s) can be removed from the group. It should be understood that, in general, where the invention, or aspects of the invention, is/are referred to as comprising particular elements, features, etc., certain embodiments of the invention or aspects of the invention consist, or consist essentially of, such elements, features, etc. For purposes of simplicity those embodiments have not in every case been specifically set forth in so many words herein. It should also be understood that any embodiment or aspect of the invention can be explicitly excluded from the claims, regardless of whether the specific exclusion is recited in the specification. For example, any one or more active agents, additives, ingredients, optional agents, types of organism, disorders, subjects, or combinations thereof, can be excluded.

Where ranges are given herein, the invention includes embodiments in which the endpoints are included, embodiments in which both endpoints are excluded, and embodiments in which one endpoint is included and the other is excluded. It should be assumed that both endpoints are included unless indicated otherwise. Furthermore, it is to be understood that unless otherwise indicated or otherwise evident from the context and understanding of one of ordinary skill in the art, values that are expressed as ranges can assume any specific value or subrange within the stated ranges in different embodiments of the invention, to the tenth of the unit of the lower limit of the range, unless the context clearly dictates otherwise. It is also understood that where a series of numerical values is stated herein, the invention includes

embodiments that relate analogously to any intervening value or range defined by any two values in the series, and that the lowest value may be taken as a minimum and the greatest value may be taken as a maximum. Numerical values, as used herein, include values expressed as percentages. For any embodiment of the invention in which a numerical value is prefaced by "about" or "approximately", the invention includes an embodiment in which the exact value is recited. For any embodiment of the invention in which a numerical value is not prefaced by "about" or "approximately", the invention includes an embodiment in which the value is prefaced by "about" or "approximately" .

"Approximately" or "about" generally includes numbers that fall within a range of 1% or in some embodiments within a range of 5% of a number or in some embodiments within a range of 10% of a number in either direction (greater than or less than the number) unless otherwise stated or otherwise evident from the context (except where such number would impermissibly exceed 100% of a possible value). It should be understood that, unless clearly indicated to the contrary, in any methods claimed herein that include more than one act, the order of the acts of the method is not necessarily limited to the order in which the acts of the method are recited, but the invention includes embodiments in which the order is so limited. It should also be understood that unless otherwise indicated or evident from the context, any product or composition described herein may be considered "isolated" .

As used herein the term "comprising" or "comprises" is used in reference to compositions, methods, and respective component(s) thereof, that are essential to the invention, yet open to the inclusion of unspecified elements, whether essential or not.

As used herein the term "consisting essentially of refers to those elements required for a given embodiment. The term permits the presence of additional elements that do not materially affect the basic and novel or functional

characteristic(s) of that embodiment of the invention.

The term "consisting of refers to compositions, methods, and respective components thereof as described herein, which are exclusive of any element not recited in that description of the embodiment.

It is to be understood that the inventions disclosed herein are not limited in their application to the details set forth in the description or as exemplified. The invention encompasses other embodiments and is capable of being practiced or carried out in various ways. Also, it is to be understood that the phraseology and terminology employed herein is for the purpose of description and should not be regarded as limiting.

While certain compositions and methods of the present invention have been described with specificity in accordance with certain embodiments, the following examples serve only to illustrate the methods and compositions of the invention and are not intended to limit the same.

EXAMPLES:

Example 1:

A hexanucleotide (GGGGCC) repeat expansion in C90RF72 is the most common genetic contributor to amyotrophic lateral sclerosis (ALS) and

frontotemporal lobar degeneration (FTLD). Reduced expression of the C90RF72 gene product has been proposed as a potential contributor to disease pathogenesis. Additionally, repetitive RNAs and dipeptide repeat proteins (DPRs), such as poly-GR, can be produced by this hexanucleotide expansion that disrupt a number of cellular processes, potentially contributing to neural degeneration. To better discern which of these mechanisms leads to disease-associated changes in patient brains, gene expression data generated from the cortex and cerebellum was analyzed. It was found that transcripts encoding heat shock proteins (HSPs) regulated by the HSF1 transcription factor were significantly induced in C90RF72- ALS/FTLD patients relative to both sporadic ALS/FTLD cases and controls. Treatment of human neurons with chemically synthesized DPRs was sufficient to activate a similar transcriptional response. Expression of GGGGCC repeats and also poly-GR in the brains of Drosophila lead to the upregulation of HSF 1 and the same highly-conserved HSPs. Additionally, HSF1 was a modifier of poly-GR toxicity in Drosophila. The results suggest that the expression of DPRs are associated with upregulation of HSF 1 and activation of a heat shock response in C90RF72 -ALS FTLD.

Identification of a C90R 72-associated transcriptional signature in patient brain samples

There remains much to be learned concerning the mechanisms by which the repeat expansion in C90RF72 contributes to ALS and FTLD. Recently, RNA- sequencing datasets were generated from the frontal cortex as well as the cerebellum of sporadic ALS cases, C90RF72-ALS cases and controls [17]. In Prudencio et al., a "double-cutoff method" was used for identifying genes whose expression was significantly changed in each class of ALS patient relative to controls (methods). Although such methods are useful for identifying changes in gene expression, they tend to be more sensitive to large fold-changes in less abundant transcripts, while modest fold-changes in abundant transcripts may go undetected (FIG. 5) [18, 19], It was reasoned that further analyses of these data might provide new insights into the disease mechanisms acting in C90RF72 and sporadic patients, respectively. Utilizing a false discovery rate (FDR) threshold of 5%, it was sought to identify changes in abundantly-expressed transcripts and 56 transcripts were found that were

differentially expressed between C90RF72 -ALS cortex and controls at this confidence interval (FIG. 5, methods). Comparison of sporadic ALS patient and control cortex with these same metrics identified 65 differentially expressed transcripts, most (61) of which were downregulated. Consistent with the previous report that sporadic and C90RF72 ALS patients display distinct transcriptional signatures relative to controls, no overlap was found in the identity of transcripts that were identified as differentially expressed in the cortex of sporadic ALS and

C90RF72 ALS patient classes (FIG. IB) [17]. However, the majority of the genes found as likely to be differentially expressed in sporadic and C90RF72-ALS patients had not been previously identified [17], validating the importance of reanalyzing the sequencing data using the methods employed herein (FIG. 5, see methods). In C90RF72-ALS, the cortex contains distinct p62-positive DPR neuronal inclusions, and the cerebellum is characterized by abundant DPR inclusions [11, 17, 20, 21]. Hence, it was reasoned that identifying transcripts with expression changes shared in both the frontal cortex and the cerebellum might lead to genes and pathways that were reproducibly induced by the C90RF72 repeat expansion. Strikingly, 27 of the 56 transcripts differentially expressed in the C90RF72-ALS cortex were also significantly changed in the C90RF72-ALS cerebellum (p = 2.93 * 10 ^_40 ; FIG 1C) Comparison of the fold expression changes in these 27 transcripts between these two regions in C90RF72 -ALS brains revealed a strong positive correlation (R ² = 0.88). Notably, increased abundance for 26 of these 27 transcripts was identified in both brain regions relative to controls (FIG. ID). The one exception was the C90RF72 transcript itself which showed reduced abundance (57% cortex, FDR=0.0169; 42% cerebellum, FDR=2.75* 10 ^'5 ), in agreement with previous studies of patient brains [3, 5, 22], In contrast to a prior analysis, no significant transcriptional changes were detected between the cerebellum of sporadic ALS cases and controls (FIG. IB), consistent with this region being histologically unremarkable in sporadic cases [23].

To determine if the transcripts significantly changed in C90RF72 and sporadic ALS cortex pointed to particular pathways that might be responding to disease processes, gene ontology (GO) analysis was employed. Transcripts identified in C90RF72-ALS were significantly associated with response to topologically incorrect proteins (p=2.13* 10 ^-6 ) and protein folding (4.57* 10 ^"7 ) (FIG. IE). In contrast, differentially expressed transcripts detected in sporadic ALS were associated with functions in the mitochondrial respiratory chain complex assembly (8.53 * 10 ^"21 ) and related terms (FIG. IF), and included 9 members of the NADH dehydrogenase (complex I) enzyme and 6 components of cytochrome oxidase C (complex IV) (FIG. 6B). These findings suggest that the transcriptional responses in the C90RF72 and sporadic ALS cortex might be reflective of changes in protein and mitochondrial homeostasis, respectively. It was also examined whether any of the C90RF72- associated transcripts encoded proteins that interact in particular complexes. Using the In Web protein-protein interaction network [24], analysis of the 56 differentially expressed transcripts from C90RF72 cortex identified an interaction network involving heat shock proteins (HSPs) and protein chaperones, with HSPAIB (HSP70) and HSPB l (HSP27) at its hubs (FIG. 1G). Examination of protein interactions from the 221 transcripts differentially expressed in the C90RF72 cerebellum revealed a similar and expanded network of more than 80 interactors that was centered on the same core HSPs (FIG. 6A).

Activation of the HSF1 pathway in C90R 72-ALS/FTLD

A well-established regulator of HSP and protein chaperone expression is the transcription factor heat shock factor 1 (HSF 1) [25]. Hence, it was considered if the transcriptional response observed in the C90RF72 brain might be at least in part mediated by activation of HSF 1. To explore this possibility, established HSF1 target genes previously identified by ChlP-seq and genome-wide methods were examined [26-28]. Consistent with the notion that a portion of the response in C90RF72 patient brain was mediated by HSF1, 13 of the 27 transcripts identified as significantly changed in both the cerebellum and frontal cortex were among 812 genes bound by HSF 1 after heat shock treatment across three human cell lines (p=1.22* 10 ^'12 ) [28], including several HSPs upregulated in the initial small C90RF72-ALS cohort [17].

As a next step towards investigating whether activation of the HSF 1 might be responsible for the induction of these genes in C90RF72 patients, quantitative RT- PCR was used to measure the transcript abundance of HSF1 and 11 of these conserved HSP-associated transcripts in a much larger patient cohort that also included patients diagnosed with FTLD and both ALS and FTLD (n=56 C90RF72- ALS/FTLD, n=46 sporadic ALS/FTLD, n=8 controls). In the frontal cortex, expression of each of these 11 HSF1 target genes was significantly increased in the Ci>QKF72-ALS/FTLD cohort relative to controls (p<0.05 or lower for each gene) and to sporadic cases (p<0.01 or lower) (FIG. 2A). Next, the expression analyses of these 11 HSF 1 targets was extended to the cerebellum, and again it was found that the abundance of each transcript gene was significantly elevated in C90RF72- ALS/FTLD relative to both controls and sporadic ALS cases (FIG. 7A). For example, a significant, two orders of magnitude induction of the HSP70 transcript HSPA IB in C90RF72- ALS/FTLD cases was found relative to controls. To investigate if the larger number of genes initially detected only in the C90RF72 cerebellum might also be reflective of a heat shock response, another HSF 1 target gene CRYAB was examined and found to be significantly upregulated in both C90RF72 brain regions in this larger patient cohort (FIG. 2A, FIG. 7A).

Evaluation of HSF1 expression in these same samples demonstrated that it was significantly more abundant in both the cortex and cerebellum relative to sporadic ALS cases (P< 0.05) (FIG. 2B). A strong and consistent correlation was found between the levels of HSF1 and each of these C90RF72 -chaperome transcripts in both brain regions (p<0.0001 for each gene, FIG. 7B). For instance, the relationship between the transcript levels of HSF1 and HSPB1 yielded an R ² value of 0.73 (95% CI 0.63 - 0.81) in the cortex and 0.65 (95% CI 0.52 - 0.75) in the cerebellum. Taken together, these findings indicate that HSF 1 is activated in C90RF72-ALS and FTLD patient brains and suggests that it is regulating the expression of the HSPs found to be induced there.

DPRs are sufficient to induce a C90R 72-associated transcriptional changes

The C90RF72 GGGGCC repeat expansion is translated from both sense and anti-sense transcripts through non-ATG translation to generate 5 distinct dipeptide repeat proteins (DPRs), e.g. poly-glycine-arginine (poly-GR) [5, 11, 29]. It was considered if DPRs alone were sufficient to induce the upregulation of C90RF72 signature transcripts. Therefore, the effects of synthetic DPRs were tested in human stem cell-derived neurons [13]. GP ₁₀ , GAi ₀ and a scrambled GAPR ₅ control were not acutely toxic to the parental human stem cell line or stem-cell derived neurons. In contrast, poly-GRio resulted in a dose-dependent decrease in the viability of stem cell- derived neurons, but not the stem cell from which they were produced (FIG. 3B). In human neurons, it was found that both poly-GA and poly-GR led to the significant upregulation oiHSPAlB (p<0.01), as well as additional C90RF72 signature transcripts (FIG. 3C). Given that poly-GA is not associated with decreased viability in these conditions, this suggests that the observed transcriptional changes are not simply a consequence of general neuronal toxicity. There was a strong correlation (R ² =0.58) between the degree of induction of these transcripts in human neurons by poly-GR and the changes present specifically in C90RF72 brains. Upon measuring HSF1, although there was a trend for increased levels with poly-GA and poly-GP, the greatest increase was again observed with poly-GR (FIG. 3D). These findings support the notion that gain-of-function effects from DPRs are sufficient to induce HSF l target genes that are upregulated in C90i? 72-associated disease.

Detection of C90R 72-associated transcriptional changes in gain-of-function Drosophila models

To test for correlations in DPR production and altered HSF l target gene expression in vivo, a Drosophila gain-of-function transgenic model engineered to express 49 pure GGGGCC repeats driven by a drug-inducible neuronal-specific ElavGS-GAL4 driver was evaluated [14, 30]. Fly models expressing toxic GGGGCC repeats produce DPRs and RNA foci [14, 31, 32]. Significant increased expression of the Drosophila orthologs of conserved C9(9i? 72-associated HSPs and protein chaperones was found in flies expressing 49 repeats in neurons compared to controls (FIG. 4A). Upregulation of HSFl -associated genes was observed in the absence of significant animal death, arguing that the effect is specific to expression of

{GGGGCC)n9. Note that these expression changes are likely to be an underestimation of the actual changes caused by the repeats in vivo since (GGGGCC)^ was only expressed in neurons while gene expression changes were assayed using whole heads, including non-neuronal tissue. Additionally, an increase in HSF l expression similar to that observed in C90RF72 patient brains was detected (FIG. 4A). This demonstrates that at least part of the transcriptional response to the C90RF72 repeat expansion is conserved in Drosophila, and is consistent with gain-of-function effects of the C90RF72 repeat expansion driving the expression of HSF l target genes.

Activation of the HSF l pathway has been proposed to be protective in several neurodegenerative diseases associated with protein aggregation as a means to combat the cellular effects of toxic proteins [33]. Given that an HSFl heat shock response was observed in C90RF72 patients and model systems, it was considered whether HSF l may be a potential modifier of C90RF72 gain-of-function toxicity. To investigate this idea, a fly line harboring an additional allele of the Drosophila HSFl ortholog (dHSFl) was selected [34]. Increased dHSFl expression in this line was confirmed and noted that it was comparable to the relative increase in dHSFl expression observed in response to the GGGGCC repeat expansion (FIG. 4B). The presence of additional dHSFl did not affect the expression of a control LacZ transgene (FIG. 4C). It was next asked if this increase in dHSFl would have an effect on GGGGCC-mediated toxicity and the Gmr-Ga driver was used to specifically express the repeats in the fly optic system to assess the effect on the eye. Consistent with prior observations, GGGGCC ₄₉ expression in the eye during development led to generation of animals with eye degeneration and disruption of the highly regular ommatidial structure, reduced eye size, and loss of pigment (FIG. 4D) [14, 30].

dHSFl upregulation by itself did not affect eye structure in the presence of a control GGGGCC _% (FIG. 4D). Surprisingly, it was found that GGGGCC ₄₉ -induced toxicity in the external eye was enhanced in the presence of dHSFl overexpression (FIGS. 4D and 4E).

Among the repeat expansion encoded DPRs, arginine-rich DPRs are particularly toxic in model systems, including Drosophila [14]. Given that the expression of GGGGCC ₄₉ is associated with the production of both DPRs and potentially toxic RNA, the transcriptional effects of poly-GR in vivo were assayed. There was significant upregulation of dHSFl and many HSFl -regulated transcripts in Drosophila expressing a poly-GRioo transgene in neurons compared to non-transgenic controls (FIG. 8). Additionally, the effects of modulating dHSFl levels in the optic system of poly-GR Drosophila was tested again using Gmr-GAL4 to drive transgene expression. Exacerbation of poly-GR36 external eye toxicity was observed in the presence of dHSFl upregulation (FIGS. 4H and 4J). These results argue that the changes in toxicity caused by added dHSFl in the GGGGCCng model is in part due to the effects of GR-dipeptide. Taken together, these findings suggest that augmentation of HSFl activity may enhance DPR-mediated toxicity in Drosophila.

Discussion

In this study, novel differentially expressed transcripts in C90RF72-ALS were identified based on analysis of two brain regions compared to controls. Every C9(9i? 72-associated transcript was not significantly altered in sporadic ALS, suggesting that the observed changes in this set of transcripts are not just a sign of neuronal loss, but rather are reflective of Ci>QKF72-specific pathogenesis.

Furthermore, a C90RF72 transcriptional signature was validated in a large

ALS/FTLD patient cohort and gain-of-function models. The findings specifically link activation of the HSFl pathway to C90RF72- ALS/FTLD. The HSFl pathway is highly conserved from budding yeast to mammals and is an important mediator of the compensatory response to disruptions in proteostasis, such as heat shock [27], Impairment of HSFl activity and loss of protein chaperone function have been reported to occur with ageing and in the setting of age- related neurodegeneration [33, 35, 36]. For instance, in models of poly-glutamine repeat-associated Huntington disease decreased expression of HSFl target genes is observed and may contribute to protein aggregation [37]. Likewise, decreased expression of a particular set of protein chaperones, including HSP90, occurs in Alzheimer disease and Parkinson disease [35]. In C90RF72- ALS/FTLD, robust increased expression of a family of protein chaperones and co-chaperones was found, consistent with activation of a heat shock response in this particular disease. This study may provide the first evidence of increased, rather than impaired, activity of HSF l based on human brain samples for a specific neurodegenerative disease. In addition, HSFl is generally not thought to be regulated at the transcriptional level in the context of neurodegeneration. Upregulation of HSFl itself in C90RF72- ALS/FTLD was found, as well as a strong correlation between levels of HSFl and its target genes.

Prior studies in model systems have suggested that HSFl is a protective factor that helps neurons cope with stress associated with misfolded proteins and protein aggregates [38]. Unexpectedly, it was observed that having additional HSFl in the developing eye in two Drosophila models of C9OR 72- ALS/FTLD was not beneficial. Pharmacological activation of HSFl has been proposed as a therapeutic strategy to enhance protein chaperone function and neuronal survival in

neurodegenerative disease [39]. For instance, arimoclomol, which may act to enhance HSF l -pathway activation, has been shown to delay disease progression in an SOD1 overexpression mouse model [40]. A phase II clinical trial for arimoclomol was recently conducted for a subtype of familial ALS associated with mutations in SOD1 and was found to be well -tolerated [41]. These findings suggest that additional preclinical studies may be warranted if this strategy is applied to other forms of ALS, especially the most common type of ALS, C90RF72 -ALS, and associated dementias. Additionally, the transcriptional differences present among distinct cohorts of ALS/FTLD patients re-emphasizes the potential importance of patient stratification by genotype for future clinical trials.

Several studies have aimed to identify the specific transcriptional changes and pathways affected by the C90RF72 repeat expansion in patient-based cellular models, including iPSC-derived neurons, with little concordance among them [42-45]. Gene expression changes have also been explored in a few animal models for this disease. In a loss-of-function mouse model lacking both copies of C90RF72, transcriptional analysis of the spinal cord from C90RF72-/- animals revealed significant changes in several pathways related to inflammation [7]. On the other hand, gain-of-function bacterial artificial chromosome (BAC) transgenic mouse models harboring the human C90RF72 repeat expansion have been generated with varying phenotypes and transcriptional changes. In one model containing exons 1 -6 of human C90RF72 with approximately 500 hexanucleotide repeats, no significant changes in the transcriptome of the frontal cortex at 6 months of age were reported [46]. In another BAC mouse model with 100-1000 repeats, immunomodulatory and extracellular matrix pathways were identified as being altered in the frontal cortex also at 6 months of age [47]. Although both of these BAC mouse models exhibit DPR inclusions in the nervous system that increase with age, evidence of neurodegeneration was not observed. One possibility is that DPRs did not reach sufficient levels in these models at the examined time points to induce neurodegeneration or the transcriptional changes described herein. Indeed, robust expression of DPRs using an adeno-associated viral vector with 66 repeats was sufficient to induce DPR aggregates, TDP -43 -positive inclusions, neuronal loss, and behavioral deficits in mice [12].

Using two gain-of-function Drosophila models, upregulation of many

Drosophila orthologs of the same genes that were upregulated in C90RF72 patient brains were found. This is consistent with the notion that more potent expression of DPRs in models is essential to recapitulate C90RF72 transcriptional changes and disease phenotypes. The methods and findings described herein, starting with unbiased transcriptional analysis of patient samples, may be useful for the

characterization and assessment of existing and new models employed to study C90RF72 disease. Based on these findings, the following model is proposed. The presence of the C90RF72 repeat expansions results in the production of various toxic DPRs. In early life, neurons can degrade DPRs or perhaps sequester them into protective p62-positive inclusions. With aging, there is a decreased capacity of neurons to maintain proteostasis, and environmental insults may be associated with additional proteotoxic stress. This leads to the gradual accumulation of DPRs and the activation of a heat shock response to increase protein chaperones, perhaps in an attempt to refold inherently unstructured DPRs. However, increased HSF1 activity may actually contribute to DPR-dependent toxicity. One possibility is that the resulting increased levels of protein chaperones may promote the solubility or the stability of toxic DPRs. This model could partially explain the variable disease penetrance and expressivity by which the C90RF72 repeat expansion acts to cause ALS and/or FTLD. It could be that natural human variation in the HSF 1 response influences when and where the repeat expansion results in neurodegeneration.

Methods

Bioinformatics

The processed gene expression count matrix of the brain-derived RNA-seq datasets from Prudencio et al. were obtained via GEO (GSE67196). The data was analyzed using the R library "edgeR" as described by Prudencio et. al., with modifications as follows [17, 48]. Statistical inference was performed with two methods which is referred to as "double cut-off and "FDR". For the "double cut-off method, as described by Prudencio et al., differentially expressed genes called by this approach had to pass two filters: one cut-off of absolute log2fold change > 2 and a second cut-off of unadjusted p-value < 0.05. For the "FDR" method, the false discovery rate was controlled using the Benjamini-Hochberg method [49] and all genes below a threshold FDR of 0.05 were considered to be significantly

differentially expressed. Additionally, a generalized linear model (glmFit() in edgeR) was used to model the effect of gender rather than the exactTest() function, which resulted in slight differences in the number of differentially expressed genes found using the double cut-off method when compared to the original published analysis. Note that a pseudocount of 0.01 was used for plotting log2(CPM). Protein-protein interaction networks were generated using GeNets hosted at the Broad Institute (apps.broadinstitute.org/genets) based on the InWeb network[24]. Associated gene ontology (GO) terms were obtained based on the Ensembl database. GO term clustering was performed with evigo (reduce and visualize gene ontology, revigo.irb.hr/) to assist in the selection of representative biological processes terms.

Brain samples

Protocols were approved by the Mayo Clinic IRB and Ethics Committee on Human Experimentation. Informed consent for post-mortem tissue was obtained from all individuals or the appropriate next-of-kin. The diagnosis of ALS and/or FTLD was based on neurological and pathological examination and C90RF72 repeat expansion status was determined using repeat-primed PCR and the cohort was described in Prudencio et al. [50]. See Table 1 for patient characteristics. For transcript measurements by quantitative RT-PCR on human brains, total RNA was extracted and 500 ng of RNA with RNA integrity values (REN) higher than 7, measured by an Agilent Bioanalyzer, and was used for reverse transcription to synthesize cDNA as previously described [17]. Using a SYBR green assay (Life Technologies) samples were run in triplicate on an ABI Prism 7900HT Real-Time PCR System (Applied Biosystems). Relative mRNA expression of examined genes was normalized to GAPDH and RPLPO values, the endogenous transcript controls. Primer sequences are provided in Table 2. Statistical differences were calculated by one-way ANOVA followed by Dunn' s multiple comparison tests. Associations between HSF1 and heat shock related transcripts were evaluated using a Spearman's test of correlation.

Neuron production and cell culture experiments

Human embryonic stem cells were cultured in mTESR (Stemcell

technologies) on matrigel (Corning). Neurons were generated from HuES-3-Hb9:GFP based on the neuron differentiation protocol described [51, 52]. Human embryonic stem cells were cultured in mTeSR (Stemcell technologies) on matrigel(Corning)- coated plates. For motor neuron differentiation, the media was changed to 1 : 1 Neurobasal :DMEM F 12 (Life Technologies) supplemented with N2 (StemCell Technologies), B27 (Life technologies), Glutamax (Life Technologies), non-essential amino acids (Life technologies). For the first week, this neural media was

supplemented with retinoic acid (Sigma Aldrich, Ι μΜ), smoothened agonist (SAG, DNSK, ΙμΜ), BMP inhibitor (LDN-193189, DNSK, 100 nM) and TGF-beta inhibitor (SB431542, DNSK, lOuM). Then, for the next week, this neural media was supplemented with retinoic acid, smoothened agonist, GSK3-beta inhibitor (SU-5402, DNSK, 4μΜ), and gamma-secretase inhibitor (DAPT, DNSK, 5uM). After two weeks, cells were dissociated with accutase (Innovative Cell Technologies) to single cells. GFP-positive motor neuron was purified with FACS and then plated on poly-D- lysine(Sigma Aldrich)/laminin(Life Technologies)-coated plates and allowed to mature two weeks in media supplemented neurotrophic factors (BDNF, CNTF and GDNF). Upon completion of the differentiation protocol, cells were sorted via flow- cytometry for GFP-positive cells to yield GFP-positive neurons plated on

PDL/laminin-coated plates (Sigma, Life technologies). Neurons were maintained in Neurobasal medium supplemented with N2, B27, Glutamax, non-essential amino acids, and neurotrophic factors (BDNF, GDNF, CNTF), and allowed to mature for two weeks before experiments with dipeptide repeat proteins (DPRs). Recombinant biotin-tagged DPRs, (each 20 amino acids in length (poly-GA, poly-GP, or poly-GR with 10 repeats or scrambled control poly-GAPR with 5 repeats) were synthesized by Anaspec with >95% purity and dissolved in DMSO (Sigma). Following DPR treatment, RNA was extracted after 24 h via an RNeasy Minikit (Qiagen), and cDNA prepared with iScript (Bio-Rad). qRT-PCR reactions were performed with iTaq SYBR green (Bio-Rad) on a C I 000 touch thermal cycler with CFX real-time system (Bio-rad). Relative expression was normalized to GAPDH. Primers were designed from the MGH PrimerBank and synthesized by IDT. Primer sequences are provided in Table 2. Viability was measured with CellTiter-Glo (Promega) on a Cytation3 reader (Biotek). All cell lines tested negative for mycoplasma using the MycoAlert detection kit (Lonza LT07-518).

Drosophila lines

Animals were raised and maintained at 18°C on standard cornmeal-molasses food. The UAS-(G4C2)n transgenic models[30, 53], UAS-(GR)36 model[14], and the HSF overexpression (OE) mutant, HSF[+t8][34], are previously defined. UAS- (GR)36, control, and mutant HSF[+t8] were obtained from Bloomington Drosophila Stock Center. qPCR in the adult fly nervous system

UAS-(G4C2)n or UAS-(GR)36 transgenes were driven by elavGS, a drug- inducible Gal4 driver that expresses only in neurons. Crosses were setup and maintained at 24°C. Female progeny with the desired genotype were collected and matured to l -3d before being transferred to vials containing 40ug/ml of RU486. Animals were aged on RU486-infused food 16d while being flipped onto fresh drug- infused food every 2-days. Total RNA was collected from heads of frozen animals using Trizol, converted to cDNA using random primers, and analyzed by qPCR using SYBR Green. All primers were previously developed with the exception of dHSFl, dHSP70, dBAG3, dStipl, dFkbp4 (Fkbp59), and dChordcl [54, 55]. Data was normalized to the housekeeping gene, RP49 [56]. Primer sequences are provided in Table 2. Full genotypes for (G4C2)n are as follows: wl 118/yw;; UAS-(G4C2)n, elavGS/+. (GR)36 animals, wl l 18/yw; UAS-(GR)36/+; elavGS/+, were compared to controls, wl 1 18/yw;;. For analysis of HSF mutant expression, briefly, male HSF OE mutant flies were crossed to wl 118 virgin females and maintained at 24°C. Male progeny were collected and aged to 5d before analysis. Full genotype: wl 118;;

HSF[+t8]/+. Control wl 1 18 males were maintained and aged in parallel.

External eye analysis

Scoring of the external eye phenotype for (G4C2)49 was done using a 0-8 scale previously defined where 0 = WT eye and 8 = lethality (extreme toxicity) [30]. (G4C2)49 expression causes an average degenerative score of 4-5 across multiple studies. Scoring of the external eye phenotype for (GR)36 was done using a 0-1 1 scale where 0 = WT eye and 11 = lethality (extreme toxicity)(FIG. 9). (GR)36 expression causes an average degenerative score of 5-6 across multiple studies.

For optimal eye phenotypes, crosses for (G4C2)n were setup and maintained at 24°C and (GR)36 at 21°C. Male progeny with the desired genotype were collected daily and matured to l-2d before imaging on a Leica Apol6 microscope. Severity of the external eye phenotype was determined post-imaging while looking for changes in red pigmentation, ommatidial organization, and eye size. Full genotypes for (G4C2)n are as follows: "Contror ^« wl 1 18;; UAS~(G4C2)n, Gmr~GaS4/+ and "HSF QE" ^ W1118;; UAS~(G4C2)a Gmr-Gal4/FISF[+t8]. Full genotypes for (GR)36 are as follows: "Control" - wl 118; UAS~(GR)36/ ; Gmr~Gal4/ - and "HSF OE" - wl l!S; UAS-(GR)36/- ; Gmr-GaI4/HSF[+t8].

Drosopbila beta-gaiactosidase western blots

Western blots are as previously described [30].

Table i: Patieitt coho t for brain samples used I qRT-FCR analysis

-3ϊ-

Tabie 2 : qf€R Primers

Gene Forward (5 -3') (SEQ ID NOS: 1.-15) Reverse (5'-3') (SEQ ID NOS: 16-30

HAGS GCACCACTACGTGGAACGA GGTGGCCTTCCCTAGCAG

CHORDCI CCTAAGCCAGTAGAAGCAATAAA TGATGACAGTTTAAGTTTATCAAGTG AA C

CRYAB CAGCTGGTTTGACACTGGAC GCTTCACATCCAGGTTGACA

DEDD2 CAGTGCGACGAGAGCAAC TAGGGTTCTGGAGACACTGG

DNAJH! GGCCTACGACGTGCTCAG GTGTAGCTGAAAGAGGTACCATTG

DNAJB4 AGGTCGCAGAAGCTTATGAAGT CCTCCTGCTCCTCCTTTCA

FKBP4 CGGGAGAAGAAGCTCTATGC GCTCTCCTGAGGAAGCCTCT

GAPDH GTTCGACAGTCAGCCGCATC GGAATTTGCCATGGGTGGA

HSFl CAAGCTGTGGACCCTCGT TCGAACACGTGGAAGCTGT

HSFAIA CGGCAAGGTGGAGATCAT GGTGTTCTGCGGGTTCAG

HSPA1B AAGGGTGTTTCGTTCCCTTT TAGTGTTTTCGCCAAGCAAA

HSFB1 TCCCTGGATGTCAACCACTT GATGTAGCCATGCTCGTCCT

RPIPO TCTACAACCCTGAAGTGCTTGAT CAATCTGCAGACA.GACACTGG

SERFINH1 GGTGGAGGTGACCCATtiA CTTG TC A ATGGCCTC AGTC A

sripi CGAAAGATGCCAAATTATACAGC CTGGATACATTCCTCACAGTCCT

For neuron experiments

Gene Forward (5 ^* -3 ^! ) (SEQ ID NOS; 31 -39) Reverse (5'-3') SEQ ID NOS: 40-48)

ACTS AGATCAAGATCATTGCTCCTCCT CGGACTCGTCATACTCCTGC

BAGS TGGGAGATCAAGATCGACCC GGGCCATTGGCAGAGGATG

CHORDCI CCTTGCTGTGCTACAACCG CGGAACACCTGGGTGGTATG

GAPDH TTGTCAAGCTCATTfCTJTGGTATG TCCTCTTGTGCTCTTGCTGG

HSFl CAAGCTGTGGACCCTCGT TCGAACACGTGGAAGCTGT

SPAJB TTTGAGGGCATCGACTTCTACA CCAGGACCAGGTCGTGAATC

HSPAfi CAAGGTGCGCGTATGCTAC GCTCATTGATGATCCGCAACAC

SERPINHl TCAGTGAGCTTCGCTGA.TGAC CATGGCGTTGACTAGCAGGG

STIPJ CCTTACAGTGCTACTCCGAAGC ATAGGCAGCAGAACGGTTGC

For Drosophiia experiments

Gene Forward (5'-3') (SEQ ID NOS; 49-59) Reverse (5'-3') (SEQ ID NOS; 60-70) tffiSPl ATCGCTTGATTTGCTGGACC GAAGCTGGCCATGTTGTTGT dflSFJO GGCACCTGCTGGCAAATT CGGATAGTGTCGTTGCACTTGT dBAG 3 GTACGCCTACCTGGACGAGA CAACTGCACCTGTGTCAACC (Stv)

(''Sap! TGGCCTAAGGGTTACTCACG GCATTCGTGGGATCGTACTT dPkbp4 GCTCCAAA.CTACGCTTACGG TGGTTGGG A CTC A A ATCCTC (Fkbp59)

dChord.1 TATTTCCCGAGCA.GGACAAC TTGCATTGCTCACATTCACA

HSP6S GAAGGCACTCAAGGACGCTAAAATG CTGAACCTTGGGAATACGAGTG dH. } GGACAAGGATGCCAAGAAGAAGAAG CAGTCGTTGGTCAGGGATTTGTA (HSP83) G

iiHSP4() GAGATCATCAAGCCCACCACAAC CGGGAAACTTAATGTCGAAGGAG

AC

JHSF27 GGCCACCACAATCAAATGTCAC CTCCTCGTGCTTCCCCTCTACC

RP49 TGTCCTTCCAGCTTCAAGATGACCATC CTTGGGCTTGCGCCATTTGTG References:

1. Taylor, J.P., R.H. Brown, Jr., and D.W. Cleveland, Decoding ALS: from genes to mechanism. Nature, 2016. 539(7628): p. 197-206

2. Robberecht, W. and T. Philips, The changing scene of amyotrophic lateral sclerosis. Nat Rev Neurosci, 2013. 14(4): p. 248-64.

3. DeJesus-Hernandez, M., et al., Expanded GGGGCC hexanucleotide repeat in noncoding region of C90RF72 causes chromosome 9p-linked FED and ALS. Neuron, 201 1. 72(2): p. 245-56.

4. Renton, A.E., et al., A hexanucleotide repeat expansion in C90RF72 is the cause of chromosome 9p21 -linked ALS-FTD. Neuron, 2011. 72(2): p. 257-68.

5. Todd, T.W. and L. Petrucelli, Insights into the pathogenic mechanisms of Chromosome 9 open reading frame 72 (C9orf72) repeat expansions. J Neurochem, 2016. 138 Suppl l : p. 145-62.

6. Neumann, M., et al., Ubiquitinated TDP-43 in frontotemporal lobar degeneration and amyotrophic lateral sclerosis. Science, 2006. 314(5796): p. 130-3.

7. O'Rourke, J.G., et al., C9orf72 is required for proper macrophage and microglial function in mice. Science, 2016. 351(6279): p. 1324-9.

8. Jiang, J., et al., Gain of Toxicity from ALS/FTD-Linked Repeat Expansions in C90RF72 Is Alleviated by Antisense Oligonucleotides Targeting GGGGCC- Containing RNAs. Neuron, 2016. 90(3): p. 535-50.

9. Burberry, A., et al., Loss-of-function mutations in the C90RF72 mouse ortholog cause fatal autoimmune disease. Sci Transl Med, 2016. 8(347): p. 347ra93. 10. Haeusler, A.R., C.J. Donnelly, and J.D. Rothstein, The expanding biology of the C9orf72 nucleotide repeat expansion in neurodegenerative disease. Nat Rev Neurosci, 2016. 17(6): p. 383-95.

11. Mori, K., et al., The C9orf72 GGGGCC repeat is translated into aggregating dipeptide-repe at proteins in FTLD/ALS. Science, 2013. 339(6125): p. 1335-8.

12. Chew, J., et al., Neurodegeneration. C90RF72 repeat expansions in mice cause TDP-43 pathology, neuronal loss, and behavioral deficits. Science, 2015. 348(6239): p. 1151-4.

13. Kwon, L, et al., Poly-dipeptides encoded by the C9orj72 repeats bind nucleoli, impede RNA biogenesis, and kill cells. Science, 2014. 345(6201): p. 1 139-45.

14. Mizielinska, S , et al., C9orf72 repeat expansions cause neurodegeneration in Drosophila through arginine-rich proteins. Science, 2014. 345(6201): p. 1 192-4.

15. Zhang, Y.J., et al., C90RF72 poly(GA) aggregates sequester and impair HR23 and nucleocytoplasmic transport proteins. Nat Neurosci, 2016. 19(5): p. 668- 77.

16. Zhang, Y.J., et al., Aggregation-prone c9FTD/ALS poly(GA) RAN-translated proteins cause neurotoxicity by inducing ER stress. Acta Neuropathol, 2014. 128(4): p. 505-24.

17. Prudencio, M., et al., Distinct brain transcriptome profiles in C9orf72- associated and sporadic ALS. Nat Neurosci, 2015. 18(8): p. 1 175-82.

18. Anders, S. and W. Huber, Differential expression analysis for sequence count data. Genome Biol, 2010. 11(10): p. R106. 19. Love, M.I., W. Huber, and S. Anders, Moderated estimation of fold change and dispersion for RNA-seq data with DESeq2. Genome Biol, 2014. 15(12): p. 550.

20. Agosta, F., et al., Structural and functional brain signatures of C9orf72 in motor neuron disease. Neurobiol Aging, 2017.

21. Mackenzie, I R , et al , Quantitative analysis and clinico-pathological correlations of different dipeptide repeat protein pathologies in C90RF72 mutation carriers. Acta Neuropathol, 2015. 130(6): p. 845-61.

22. Waite, A.J., et al., Reduced C9orp2 protein levels in frontal cortex of amyotrophic lateral sclerosis and frontotemporal degeneration brain with the C90RF72 hexanucleotide repeat expansion. Neurobiol Aging, 2014. 35(7): p. 1779 e5-1779 el3.

23. Al-Sarraj, S., et ., ρ62 positive, TDP-43 negative, neuronal cytoplasmic and intranuclear inclusions in the cerebellum and hippocampus define the pathology of C9orf72-linked FTLD andMND/ALS. Acta Neuropathol, 2011. 122(6): p. 691-702.

24. Li, T., et al., A scored human protein-protein interaction network to catalyze genomic interpretation. Nat Methods, 2017. 14(1): p. 61-64.

25. Zou, J., et al., Repression of heat shock transcription factor HSF1 activation by HSP90 (HSP90 complex) that forms a stress-sensitive complex with HSF1. Cell, 1998. 94(4): p. 471-80.

26. Mahat, D.B., et al . , Mammalian Heat Shock Response and Mechanisms Underlying Its Genome-wide Transcriptional Regulation. Mol Cell, 2016. 62(1): p. 63-78. 27. Solis, E.J., et al., Defining the Essential Function of Yeast Hsfl Reveals a Compact Transcriptional Program for Maintaining Eukaryotic Proteostasis. Mol Cell, 2016. 63(1): p. 60-71.

28. Mendillo, M.L., et al., HSF1 drives a transcriptional program distinct from heat shock to support highly malignant human cancers. Cell, 2012. 150(3): p. 549-62.

29. Gendron, T.F., et al., Antisense transcripts of the expanded C90RF72 hexanucleotide repeat form nuclear RNA foci and undergo repeat-associated non- ATG translation in c9FTD/ALS. Acta Neuropathol, 2013. 126(6): p. 829-44.

30. Kramer, N.J., et al., Spt4 selectively regulates the expression of C9orfi2 sense and antisense mutant transcripts. Science, 2016. 353(6300): p. 708-12.

31. Freibaum, B.D., et al., GGGGCC repeat expansion in C9orf72 compromises nucleocytoplasmic transport. Nature, 2015. 525(7567): p. 129-33.

32. Zhang, K , et al., The C9orpi repeat expansion disrupts nucleocytoplasmic transport. Nature, 2015. 525(7567): p. 56-61.

33. Morimoto, R.I., The heat shock response: systems biology of proteotoxic stress in aging and disease. Cold Spring Harb Symp Quant Biol, 201 1. 76: p. 91-9.

34. Jedlicka, P., M.A. Mortin, and C. Wu, Multiple functions of Drosophila heat shock transcription factor in vivo. EMBO J, 1997. 16(9): p. 2452-62.

35. Brehme, M., et al., A chaperome subnetwork safeguards proteostasis in aging and neurodegenerative disease. Cell Rep, 2014. 9(3): p. 1 135-50.

36. Hartl, F.U., A. Bracher, and M. Hayer-Hartl, Molecular chaperones in protein folding and proteostasis. Nature, 2011. 475(7356): p. 324-32. 37. Chafekar, S.M. and M L. Duennwald, Impaired heat shock response in cells expressing full-length poly glutamine -expanded huntingtin. PLoS One, 2012. 7(5): p. e37929.

38. Gomez-Pastor, R., E.T. Burchfiel, and DJ. Thiele, Regulation of heat shock transcription factors and their roles in physiology and disease. Nat Rev Mol Cell Biol, 2018. 19(1): p. 4-19.

39. Neef, D.W., A.M. Jaeger, and D.J. Thiele, Heat shock transcription factor 1 as a therapeutic target in neurodegenerative diseases. Nat Rev Drug Discov, 201 1. 10(12): p. 930-44.

40. Kieran, D., et al., Treatment with arimoclomol, a coinducer of heat shock proteins, delays disease progression in ALS mice. Nat Med, 2004. 10(4): p. 402-5.

41. Benatar, M., et al., Randomized, double-blind, placebo-controlled trial of arimoclomol in rapidly progressive SOD 1 ALS. Neurology, 2018. 90(7): p. e565- e574.

42. Lagier-Tourenne, C, et al., Targeted degradation of sense and antisense C9orf72 RNA foci as therapy for ALS and frontotemporal degeneration. Proc Natl Acad Sci U S A, 2013. 110(47): p. E4530-9.

43. Donnelly, C.J., et al., RNA toxicity from the ALS/FTD C90RF72 expansion is mitigated by antisense intervention. Neuron, 2013. 80(2): p. 415-28.

44. Sareen, D., et al., Targeting RNA foci in iPSC-derived motor neurons from ALS patients with a C90RF72 repeat expansion. Sci Transl Med, 2013. 5(208): p. 208ral49. 45. Cooper-Knock, J., et al., C90RF72 GGGGCC Expanded Repeats Produce Splicing Dysregulation which Correlates with Disease Severity in Amyotrophic Lateral Sclerosis. PLoS One, 2015. 10(5): p. e0127376.

46. Peters, O.M., et al., Human C90RF72 Hexanucleotide Expansion Reproduces RNA Foci andDipeptide Repeat Proteins but Not Neurodegeneration in BAC Transgenic Mice. Neuron, 2015. 88(5): p. 902-9.

47. O'Rourke, J.G., et al., C9orf72 BAC Transgenic Mice Display Typical Pathologic Features ofALS/FTD. Neuron, 2015. 88(5): p. 892-901.

48. Robinson, M.D., D.J. McCarthy, and G.K. Smyth, edgeR: a Bioconductor package for differential expression analysis of digital gene expression data.

Bioinformatics, 2010. 26(1): p. 139-40.

49. Reiner, A., D. Yekutieli, and Y. Benjamini, Identifying differentially expressed genes using false discovery rate controlling procedures. Bioinformatics, 2003. 19(3): p. 368-75.

50. Prudencio, M., et al., Repetitive element transcripts are elevated in the brain ofC9orp2 ALS/FTLD patients. Hum Mol Genet, 2017. 26(17): p. 3421-3431.

51. Boulting, G.L., et al., A functionally characterized test set of human induced pluripotent stem cells. Nat Biotechnol, 2011. 29(3): p. 279-86.

52. Hill, S.J., et al., Two familial ALS proteins function in prevention/repair of transcription-associated DNA damage. Proc Natl Acad Sci U S A, 2016. 113(48): p. E7701-E7709.

53. Burguete, A.S., et al., GGGGCC microsatellite RNA is neuritically localized, induces branching defects, and perturbs transport granule function. Elife, 2015. 4: p. e08881. 54. Colinet, H , S.F. Lee, and A. Hoffmann, Temporal expression of heat shock genes during cold stress and recovery from chill coma in adult Drosophila melanogaster. FEBS J, 2010. 277(1): p. 174-85.

55. Berson, A., et al., TDP-43 Promotes Neurod generation by Impairing Chromatin Remodeling. Curr Biol, 2017. 27(23): p 3579-3590 e6

56. Gabler, M., et al., Trans-splicing of the mod(mdg4) complex locus is conserved between the distantly related species Drosophila melanogaster andD. virilis. Genetics, 2005. 169(2): p. 723-36.

Example 2:

This Example both re-presents certain data from Example 1 and provides additional data.

Transcriptome analysis of patient brain samples

Recently, Prudencio and colleagues (16) reported the generation of RNA sequencing data from the frontal cortex and cerebellum of control individuals, as well as sporadic (sALS) and C90RF72 ALS cases (FIG. 10A). Here, the analysis of these datasets is extended by comparing two approaches for identifying differentially expressed transcripts (FIG. 10A). In Prudencio et al., a "double-cutoff method" was used for identifying genes whose expression was significantly changed in each class of ALS patients relative to controls. While such methods are useful for identifying changes in gene expression, they tend to be more sensitive to large fold-changes in less abundant transcripts while modest fold-changes in abundant transcripts may go undetected (17, 18). Comparing the transcript abundance of genes identified by the double cut-off method in the cortex and cerebellum of sALS and C90RF72-ALS cases with the abundance of all unique transcripts detected confirmed that the majority of identified transcripts were indeed present at lower levels (Log2CPM < 0) (FIG. 10B). The initial study was therefore complemented with a false discovery rate (FDR) ranking method to help identify abundant transcripts displaying significant, but more modest fold-changes between ALS cases and controls (expressed at Log2CPM > 0). RNA sequencing data from cortex and cerebellum samples analyzed at an FDR < 0.05 (irrespective of fold change) showed that this method effectively detected changes in the levels of more abundant transcripts within both sporadic and C90RF72 patients (FIG. IOC). As a result of this distinction in mean abundance of transcripts detected by the two methods, there was sparse overlap between the transcripts identified in each case as differentially expressed in sALS and C90RF72 ALS (FIG. 10D). Notably, the double-cutoff method had not detected C90RF72 as differentially expressed between C90RF72 mutant cases, sporadic cases and controls. In contrast, utilizing FDR ranking identified the reduced C90RF72 transcript abundance previously observed in repeat expansion carriers relative to controls (57% cortex, FDR 0.0169; 42% cerebellum, FDR 2.75 E-05)(3, 19, 20).

It was intriguing to see that there was no overlap between the FDR-identified transcripts in sALS and C90RF72-ALS in the cortex or cerebellum (FIG. 10E), hinting at molecularly distinct transcriptional responses in each patient class. Gene ontology (GO) pathway analysis of transcripts identified by FDR in C90RF72-ALS cortex demonstrated obvious enrichment for a response to unfolded proteins and related terms, such as chaperone-mediated protein folding and cellular response to heat (FIG. 10F).

Strikingly, 27 of the 56 differentially expressed transcripts in the C90RF72- ALS cortex were also significantly changed in the cerebellum (p = 2.93 * 10-40).

Comparison of the fold change in expression of these shared genes in the cortex and cerebellum revealed a strong positive correlation (R ² = 0.88) with increased levels of 26 out of 27 transcripts and decreased transcription of C90RF72 in both regions (FIG. 101).

On the other hand, analysis of sALS cortex revealed enrichment for transcripts encoding proteins with functions in the electron transport chain and energy metabolism (FIG. 10G). In the cerebellum, no changes were found in sporadic ALS cases relative to controls at this FDR (FIG. 10E). Consistent with this later observation, the cerebellum of these sALS cases was histologically unremarkable (27).

It was noted that the transcript most increased in abundance in both the cerebellum and cortex of C90RF72 cases, was the heat-shock protein (HSP) HSPA6 (FIG. 101). Including HSPA6, a significant number (1 1) of the transcripts identified by FDR in both C90RF72 -ALS cortex and cerebellum encoded for HSPs, protein chaperones, and other components of the "chaperome", which consists of over 300 proteins that are involved in maintaining protein homeostasis or "proteostasis" (p=l .53 * 10-14) (22, 23). Interaction analysis of the proteins encoded by transcripts significantly changed in the C90RF72-ALS cortex revealed a network with HSPs at the hubs (FIG. 101). When performing this analysis using transcripts significantly changed in the C90RF72-ALS cerebellum, an expanded, but highly related network of HSP regulated protein products was observed. Given that the cortex and cerebellum of C90RF72 patients are both histologically characterized by marked p62- positive neuronal inclusions (24), it was surmised that the gene signature detected in these two brain regions reflected a specific biological process.

Activation of the HSFl pathway in C90R 72-ALS/FTD

One well known regulator of chaperone protein expression is the transcription factor HSFl (heat shock factor-1) (25, 26). Under normal conditions, HSFl is complexed with HSPs in an inactive state (27). Upon heat shock, HSFl is released and accumulates in the nucleus to induce the transcription of HSPs and other chaperones (28). It was noticed that the 11 chaperome-associated transcripts in the C90RF72-ALS gene signature were established target genes of HSF l, suggesting that activation of HSF l could be driving, at least a subset of, the transcriptional changes observed ((26), (29), (30)). It was next examined if pathological alterations in HSFl could be detected in human brains with C90RF72-ALS. It was found that HSFl mRNA levels were upregulated specifically in C9(9i? 72-associated

neurodegenerative disease (ALS, FTD, and combined ALS-FTD) compared to either controls or non-C90RF72 patients (p<0.05) (FIG. 11A and 11B). Additionally, histological analysis revealed increased nuclear HSF l protein expression in the cerebellar neurons of C90RF72 -ALS/FTD cases, most notably in Purkinje neurons, but not in controls (FIG. 11C).

Next, it was considered whether the abundance of transcripts from the HSF 1 target genes initially identified in C90RF72 -ALS could be validated in a larger cohort of cases (n=56 C90RF72-ALSfFTO, n=46 sporadic ALS/FTD, n=9 controls). In the frontal cortex, RNA levels of each of the 12 HSF1 targets tested were significantly increased in C90RF72-ALS/FTO compared to sporadic cases (p<0.01) (FIG. 11D) and all were significantly increased relative to controls (p<0.05 or lower for each gene). For example, BAGS, which encodes an HSP70 co-chaperone protein, was markedly upregulated in C90RF72 -ALS FTD compared to controls (p<0.01) and also compared to sporadic disease (p<0.001) (FIG. 1 1D). Additionally, a significant correlation between HSF1 expression levels and the expression levels of each of these target genes in this cohort of ALS, FTD, and ALS FTD patients was found (p<0.0001 for each gene). For instance, the relationship between transcript levels of HSF1 and HSPA1A, encoding an HSP70 protein, yielded an R2 value of 0.63 (95% CI 0.51 - 0.74). These analyses were extended to the cerebellum, and again all 12 HSF1 targets were significantly increased in C90RF72- ALS/FTD compared to sporadic disease (p<0.05 or lower for each gene). These findings strongly suggest that activation of the HSF 1 pathway in C90RF72 -ALS/FTD patients is driving the gene signature originally observed through analysis of RNA sequencing data.

Compounds connected to ALS disease states

It was then considered what specific molecular targets might be perturbed to induce the particular HSF 1 response that was observed in C90RF72 -ALS/FTD patients or the signature of gene expression found in sporadic ALS. To this end the "Connectivity Map" was looked to, which is a public database of cell line expression profiles following exposure to chemical compounds and coupled to a pattern- matching algorithm tool based on the Kolmogorov-Smirnov statistic (3 J). The comparison of a gene signature of interest to reference datasets of cells treated with diverse chemical agents with known mechanisms of action yields an enrichment score or "connectivity score", as an indication of the degree of similarity (FIG. 12A). In this unbiased manner, the transcriptional profiles of C90RF72-A S and sporadic ALS brain samples were compared with those induced by 1261 distinct chemical compounds in 5 cell lines. Starting with either the C90RF72 -ALS cortex or cerebellum yielded a similar profile of most significant (p<0.001) compounds with a high connectivity score (n=10 cortex; n=l l cerebellum), none of which were observed when using the sporadic ALS gene signature (FIGS. 12B and 12C). Interestingly, hit compounds broadly fell into three categories: proteasome inhibitors (e.g. G262), HSP90-inhibitors (e g geldanamycin), and translation inhibitors (e g. puromycin) (FIG. 12D). MG-262 is a cell-permeable 26S proteasome inhibitor, and had near maximum connectivity scores with the C90RF72-ALS signature from both brain regions (0.957 cortex, 0.979 cerebellum). Geldanamycin, which is an antibiotic that inhibits the ATPase activity of HSP90 chaperone proteins, and its derivatives (e.g. tanespimycin/17-AAG) were also hits. Puromycin is an aminonucleoside antibiotic that resembles tRNA and inhibits translation through premature chain termination, resulting in the production of truncated puromycyl-containing peptides. Some compounds identified through connectivity analysis also had known off-target effects on these molecular targets. For instance, 15d-PGJ2 is an anti-inflammatory agent, but has been reported to induce aggregation of ubiquitinated proteins and inhibit the proteasome (32).

Disease mechanisms ranging from retroviral activation to oxidative stress have been proposed to underlie ALS pathogenesis (J, 5). The connectivity map database allowed for the question to be asked whether the gene expression differences observed in the two patient classes of ALS supported any of these hypotheses (FIG.12D). Azacytidine, which inhibits DNA methylation and promotes the expression of retroviral elements, had a low connectivity score. This was also the case for compounds known to activate ER stress pathways, such as thapsigargin, or to inhibit the autophagy-lysosome pathway as exemplified by chloroquine. Disruption of nucleocytoplasmic transport has been proposed as mechanism of disease in

C90RF72-ALS, in particular (33, 34). However, the response to ivermectin, which inhibits nuclear import, was not significantly associated with the C90RF72 -ALS patient gene signature. Finally, dysregulation of RNA metabolism and

ribonucleoproteins have been implicated in ALS and FTD. Of note, agents that globally alter transcription, such as thioguanosine, were also not connected to either C90RF72- ALS or sALS gene expression changes (FIG. 12D).

Proteasome inhibition recapitulates the C90RF72-ALS gene signature

It was then considered if inhibition of the proteasome would be sufficient to recapitulate the C90RF72-ALS signature in neurons. To this end, purified stem-cell derived human neurons were generated, a commonly used proteasome inhibitor (MG- 132) was applied, and samples for RNA-sequencing were processed (FIG. 13A). A significant upregulation of transcripts encoding heat shock proteins and chaperones was found, with gene ontology analysis showing enrichment primarily for the unfolded protein response and related terms- noticeably similar to the C90RF72 -ALS analysis (FIG. 13B). Protein-protein interaction network analysis of the significantly upregulated transcripts revealed high connectivity to HSPs.

Notably, all of the chaperome-associated C90RF72-ALS signature transcripts were significantly upregulated after proteasome inhibition (adjusted p<0.0001 at both doses) (FIG. 13C). Moreover, HSPA6, the top gene (by fold change), was also the top differentially expressed gene in this dataset (log2 fold change >10). Upregulation of these transcripts after proteasome inhibition by RT-PCR was then validated. The gene expression changes in proteasome-inhibited neurons showed a strong correlation with C90RF72-ALS brains, but not sALS (R2=0.67 between MG-132 treated human neurons vs. C90RF72-ALS, and R2=0.01 between MG-132 treated neurons vs. sALS, FIG. 13). Moreover, querying the signature from proteasome-inhibited neurons in the connectivity map yielded a connectivity score of 0.999 for MG-262.

Additionally, many compounds identified associated with the C90RF72- ALS gene signature were also associated with the gene expression profiles of proteasome- inhibited neurons. Overall, this analysis suggests that the disease-state present in C90RF72-ALS brain mimics the neuronal response in vitro to proteasome inhibition.

DPRs are sufficient to induce a heat shock-associated transcriptional response

The C90RF72 GGGGCC repeat expansion is translated in the sense and anti- sense direction through non-ATG translation to generate 5 distinct dipeptide repeat proteins (DPRs). The sense DPRs (poly-GA, -GP, -GR) are however thought to be more abundant in patient brains (5, 10, 35). Given that DPRs may be in an intrinsically unfolded state, it was considered whether they interacted with protein chaperones in general, and the specific components of the chaperome network that were found upregulated in C90RF72-ALS. Recently, independent mass spectrometry efforts were conducted to identify DPR interacting proteins (36, 37). Though, the hypothesis that DPRs might interact with chaperones was left unexplored. Here, within these recent data sets, significant overlap was found between the "chaperome" and DPR interacting proteins; for instance, a dataset of poly-PR interacting proteins included 77 members of the chaperome (p=2.87 e-25). Additionally, at least 14 of the DPR-interacting protein chaperones are encoded by genes that were found to be upregulated in C90RF72-ALS, such as DNAJA1 and FKBP4. These connections suggest that DPRs may perturb chaperome function, and hence lead to compensatory upregulation of genes encoding chaperome components.

It was next asked whether both histological and transcriptional signatures from C90RF72 patients could be recreated in human neurons using synthetic DPRs. It was first noticed that neither DPRs, nor a scrambled GAPR-repeat control, were toxic to human embryonic stem cells (FIG. 14A). However, there was a dose-dependent decrease in neuronal viability when stem cell-derived neurons were treated with poly- GR (FIG. 14A) for more than 24 hours. This is consistent with prior observations that poly-GR can be cytotoxic, and demonstrates that this toxicity is cell-context dependent (12). Mislocalization of the DNA/RNA-binding protein TDP-43 from the nucleus to the cytoplasm is one of the pathological hallmarks of neurodegeneration in ALS. It was observed that poly-GR alone also induced a modest change in TDP-43 localization in human neurons (FIG. 14B).

It was also found that all DPRs tested, even those that were apparently nontoxic, lead to the significant induction of HSPA6 when compared to controls (p<0.01) (FIG. 14C), as well as upregulation of 5 out of 6 additional HSF1 target genes that were tested, with the most substantial response generated by poly-GR (FIG. 14C). Although SERPINHl was increased in abundance in C90RF72-ALS brains, its levels were not significantly changed after DPR treatment, suggesting that its upregulation in disease may be a secondary effect of chronic HSF 1 activation or other pathways. Nonetheless, there was a strong correlation (R2=0.49) between the induction of these transcripts in human neurons by DPRs and the changes present in C90RF72-ALS patients (FIG. 14D). Finally, GFP -tagged DPRs were expressed in a non-neuronal human cell line, 293 cells, and it was found that each of the five DPRs lead to the robust induction of HSPA6 and many other disease-associated HSF1 target genes compared to GFP alone. Together, these results suggest that several individual DPRs are sufficient to lead to the transcriptional changes specific to the brains of patients with C90RF72-ALS/FTD

Concluding remarks

Here, it has been found and validated that activation of the HSF1 transcription factor and increased downstream expression of select target chaperones are features of C90RF72-ALS/FTO. Engagement of this pathway in C90RF72 patients was distinguishing from sporadic patients, which showed significantly less expression of HSF 1 and its targets. It should be noted that the sporadic ALS cases analyzed did not carry known predisposing genetic variants and are likely to represent a less uniform patient population. This heterogeneity, however, did not impede the ability to detect significant alterations in the abundance of transcripts associated with mitochondrial respiration, which were not observed in C90RF72 patients. Overall, these

observations further the hypothesis that distinct pathways may be activated in discrete ALS populations as part of their disease {16).

The HSF 1 pathway, which was found to be induced in C90RF72 patients, is highly conserved from budding yeast to mammals and is an important mediator of the compensatory response to disruptions in proteostasis such as heat shock (30).

Consistent with the notion that the C90RF72 repeat expansion was leading to a response in the patient brain similar to heat shock, the changes in transcript abundance that were detected were highly correlated to those observed in cultured human cells, including human stem cell-derived neurons, treated with chemical inhibitors of the proteasome and HSP90 family of chaperones. It has not escaped notice that mutations in several components of the HSF1 pathway result in neuromuscular disease. For instance, variants in DNAJB2 and HSPB8 cause forms of Charcot-Marie-Tooth disease with motor neuropathy, and mutations in DNAJB6 manifest with limb-girdle muscular dystrophy and/or FTD (38-40). Beyond providing further insights into the molecular heterogeneity of ALS, the gene expression changes found in the C90RF72 brain provided a validated signature that could be used as a probe for determining which of the proposed pathogenic mechanisms for the repeat expansion was most likely acting in patients. While this study cannot rule out a role for C90RF72 haploinsufficiency or RNA- toxicity, these findings that even transient exposure of human neurons to DPRs is sufficient to induce a response highly correlated to the patient brain lends significant credence to the notion that they are a key driver of the disease process.

Based on these findings, the following model for neural degeneration in C90RF72-ALSfFTO is proposed (FIG. 14). The presence of the C90RF72 repeat expansions results in the production of various toxic DPRs. In early life, neurons can degrade DPRs or perhaps sequester them into protective p62-positive inclusions. With aging, there is a decreased capacity of neurons to maintain proteostasis, and environmental insults may be associated with additional proteotoxic stress, leading to the further accumulation of DPRs. Neurons initially compensate through upregulation of an HSFl response and increased expression of the chaperome networks.

Ultimately, the continued production of DPRs exhaust the capacity of the proteasome and chaperome to maintain proteostasis resulting in neurodegeneration and neurological sequelae. This framework is attractive as it provides for several features that could ultimately contribute to the variable disease penetrance and age of onset of the C90RF72 expansion. For instance, it is well established that the extent which the HSF l pathway can be activated is a governor of lifespan in C. elegans (41, 42). It could likewise be that natural human variation in the HSF l response influences when, where or indeed if, the repeat expansion results in neurodegenerative disease.

Finally, given the similarity between the transcriptional response in the brains of C90RF72 patients and a classical heat shock response, these findings raise the possibility that therapeutic cooling, as used in neonatal encephalopathy (43), or other approaches promoting restoration of normal protein homeostasis might be of therapeutic benefit. References:

1. J. P. Taylor, R. H. Brown, Jr., D. W. Cleveland, Decoding ALS: from genes to mechanism. Nature 539, 197-206 (2016).

2. W. Robberecht, T. Philips, The changing scene of amyotrophic lateral sclerosis. Nat Rev Neurosci 14, 248-264 (2013).

3. M. DeJesus-Hernandez et al., Expanded GGGGCC hexanucleotide repeat in noncoding region of C90RF72 causes chromosome 9p-linked FTD and ALS. Neuron 72, 245-256 (2011).

4. A. E. Renton et al, A hexanucleotide repeat expansion in C90RF72 is the cause of chromosome 9p21-linked ALS-FTD. Neuron 72, 257-268 (2011).

5. T. W. Todd, L. Petrucelli, Insights into the pathogenic mechanisms of Chromosome 9 open reading frame 72 (C9orf72) repeat expansions. J Neurochem 138 Suppl 1, 145-162 (2016).

6. J. G. O'Rourke et al, C9orf72 is required for proper macrophage and microglial function in mice. Science 351, 1324-1329 (2016).

7. J. Jiang et al. , Gain of Toxicity from ALS/FTD-Linked Repeat Expansions in C90RF72 Is Alleviated by Antisense Oligonucleotides Targeting GGGGCC- Containing RNAs. Neuron 90, 535-550 (2016).

8. A. Burberry et al., Loss-of-function mutations in the C90RF72 mouse ortholog cause fatal autoimmune disease. Sci Transl Med 8, 347ra393 (2016). 9. A. R. Haeusler, C. J. Donnelly, J. D. Rothstein, The expanding biology of the C9orf72 nucleotide repeat expansion in neurodegenerative disease. Nat Rev Neurosci 17, 383-395 (2016).

10. K. Mori et al, The C9orf72 GGGGCC repeat is translated into aggregating dipeptide-repeat proteins in FTLD/ALS. Science 339, 1335-1338 (2013).

11. J. Chew et al, Neurodegeneration. C90RF72 repeat expansions in mice cause TDP-43 pathology, neuronal loss, and behavioral deficits. Science 348, 1151-1154 (2015).

12. 1. Kwon et al, Poly-dipeptides encoded by the C9orf72 repeats bind nucleoli, impede RNA biogenesis, and kill cells. Science 345, 1139-1145 (2014).

13. S. Mizielinska et al, C9orf72 repeat expansions cause neurodegeneration in Drosophila through arginine-rich proteins. Science 345, 1192-1194 (2014).

14. Y. J. Zhang et al, C90RF72 poly(GA) aggregates sequester and impair HR23 and nucleocytoplasmic transport proteins. Nat Neurosci 19, 668-677 (2016).

15. Y. J. Zhang et al, Aggregation-prone c9FTD/ALS poly(GA) RAN-translated proteins cause neurotoxicity by inducing ER stress. Acta Neuropathol 128, 505-524 (2014).

16. M. Prudencio et al, Distinct brain transcriptome profiles in C9orf72-associated and sporadic ALS. Nat Neurosci 18, 1175-1182 (2015). 17. S. Anders, W. Huber, Differential expression analysis for sequence count data. Genome Biol 11, R106 (2010).

18. M. I. Love, W. Huber, S. Anders, Moderated estimation of fold change and dispersion for RNAseq data with DESeq2. Genome Biol 15, 550 (2014).

19. V. V. Belzil et al, Reduced C9orf72 gene expression in c9FTD/ALS is caused by histone tnmethylation, an epigenetic event detectable in blood. Acta Neuropathol 126, 895-905 (2013).

20. A. J. Waite et al , Reduced C9orf72 protein levels in frontal cortex of amyotrophic lateral sclerosis and frontotemporal degeneration brain with the C90RF72

hexanucleotide repeat expansion. Neurobiol Aging 35, 1779 el775-1779 el713 (2014).

21. S. Al-Sarraj et al, p62 positive, TDP-43 negative, neuronal cytoplasmic and intranuclear inclusions in the cerebellum and hippocampus define the pathology of C9orf72-linked FTLD and MND/ALS. Acta Neuropathol 122, 691-702 (201 1).

22. M. Brehme et al, A chaperome subnetwork safeguards proteostasis in aging and neurodegenerative disease. Cell Rep 9, 1 135-1150 (2014).

23. F. U. Haiti, A. Bracher, M. Hayer-Hartl, Molecular chaperones in protein folding and proteostasis. Nature 475, 324-332 (2011).

24. 1. R. Mackenzie et al., Quantitative analysis and clinico-pathological correlations of different dipeptide repeat protein pathologies in C90RF72 mutation carriers. Acta Neuropathol 130, 845-861 (2015). 25. R. I. Morimoto, The heat shock response: systems biology of proteotoxic stress in aging and disease. Cold Spring Harb Symp Quant Biol 76, 91-99 (2011).

26. M. L. Mendillo et al , HSF1 drives a transcriptional program distinct from heat shock to support highly malignant human cancers. Cell 150, 549-562 (2012).

27. J. Zou, Y. Guo, T. Guettouche, D. F. Smith, R. Voellmy, Repression of heat shock transcription factor HSF 1 activation by HSP90 (HSP90 complex) that forms a stress- sensitive complex with HSF1. Cell 94, 471-480 (1998).

28. K. D. Sarge, S. P. Murphy, R. I. Morimoto, Activation of heat shock gene transcription by heat shock factor 1 involves oligomerization, acquisition of DNA- binding activity, and nuclear localization and can occur in the absence of stress. Mol Cell Biol 13, 1392-1407 (1993).

29. D. B. Mahat, H. H. Salamanca, F. M. Duarte, C. G. Danko, J. T. Lis, Mammalian Heat Shock Response and Mechanisms Underlying Its Genome-wide Transcriptional Regulation. Mol Cell 62, 63-78 (2016).

30. E. J. Solis et al, Defining the Essential Function of Yeast Hsfl Reveals a Compact Transcriptional Program for Maintaining Eukaryotic Proteostasis. Mol Cell 63, 60-71 (2016).

31. J. Lamb et al., The Connectivity Map: using gene-expression signatures to connect small molecules, genes, and disease. Science 313, 1929-1935 (2006). 32. H. Liu et al, Prostaglandin D2 toxicity in primary neurons is mediated through its bioactive cyclopentenone metabolites. Neurotoxicology 39, 35-44 (2013).

33. K. Zhang et al, The C9orf72 repeat expansion disrupts nucleocytoplasmic transport. Nature 525, 56-61 (2015).

34. B. D. Freibaum et al, GGGGCC repeat expansion in C9orf72 compromises nucleocytoplasmic transport. Nature 525, 129-133 (2015).

35. T. F. Gendron et al, Antisense transcripts of the expanded C90RF72

hexanucleotide repeat form nuclear RNA foci and undergo repeat-associated non- ATG translation in c9FTD/ALS. Acta Neuropathol 126, 829-844 (2013).

36. K. H. Lee et al, C9orf72 Dipeptide Repeats Impair the Assembly, Dynamics, and Function of Membrane-Less Organelles. Cell 167, 774-788 e717 (2016).

37. Y. Lin et al, Toxic PR Poly-Dipeptides Encoded by the C9orf72 Repeat

Expansion Target LC Domain Polymers. Cell 167, 789-802 e712 (2016).

38. B. S. Tang et al, Small heat-shock protein 22 mutated in autosomal dominant Charcot-Marie-Tooth disease type 2L. Hum Genet 116, 222-224 (2005).

39. S. C. Blumen et al, A rare recessive distal hereditary motor neuropathy with HSJ1 chaperone mutation. Ann Neurol 71, 509-519 (2012).

40. 1. Yabe et al, Pathology of frontotemporal dementia with limb girdle muscular dystrophy caused by a DNAJB6 mutation. Clin Neurol Neurosurg 127, 10-12 (2014). 41. J. F. Morley, R. I. Morimoto, Regulation of longevity in Caenorhabditis elegans by heat shock factor and molecular chaperones. Mol Biol Cell 15, 657-664 (2004).

42. A. L. Hsu, C. T. Murphy, C. Kenyon, Regulation of aging and age-related disease by DAF-16 and heat-shock factor. Science 300, 1142-1145 (2003).

43. H. Wang et al., Brain temperature and its fundamental properties: a review for clinical neuro scientists. Front Neurosci 8, 307 (2014).