Expanded intronic G₄C₂ repeats in the C9ORF72 gene cause amyotrophic lateral sclerosis (ALS) and frontotemporal dementia (FTD). These intronic repeats are translated through a non-AUG-dependent mechanism into five different dipeptide repeat proteins (DPRs), including poly-glycine-arginine (GR), which is aggregation-prone and neurotoxic. Here, we report that Kapβ2 and GR interact, co-aggregating, in cultured neurons in vitro and CNS tissue in vivo. Importantly, this interaction significantly decreased the risk of death of cultured GR-expressing neurons. Downregulation of Kapβ2 is detrimental to their survival, whereas increased Kapβ2 levels mitigated GR-mediated neurotoxicity. As expected, GR-expressing neurons displayed TDP-43 nuclear loss. Raising Kapβ2 levels did not restore TDP-43 into the nucleus, nor did alter the dynamic properties of GR aggregates. Overall, our findings support the design of therapeutic strategies aimed at up-regulating Kapβ2 expression levels as a potential new avenue for contrasting neurodegeneration in C9orf72-ALS/FTD.

Human tissues: Human post-mortem CNS tissue from healthy controls and patients carrying the C9ORF72 intronic nucleotide repeat expansion was obtained from the Target ALS and the Jefferson Weinberg ALS center biobanks. Information and demographics of patients and controls are presented in Table 4. Human post-mortem tissues (fresh frozen) were stored at -80 C.

Cell cultures: HEK293 cells were cultured in DMEM medium (Cytiva Cat.# SH30243.LS ) supplemented with 10% FBS (Cytiva Cat.# SH30071.01HI), penicillin, and streptomycin (Thermo Fisher Scientific Cat.# SV30010). Cells were passaged every 3-4 days using 0.05% trypsin (Corning Cat.# 25-051-C).

Primary cortical neurons: After meninges removal, brains were dissected from embryonic day 16 (E16) rat embryos. Cortices and midbrain regions were cut into small pieces and incubated on a shaker at 80 rcf for 45 min at 37°C in 0.2% trypsin in HBSS without Ca²⁺ and Mg²⁺ (Cytiva Cat.# SH30588.01). FBS (Cytiva Cat.# SH30071.01HI) was added, and the cell suspension was centrifuged at 800 rcf for 10 min. at 4°C. Cells were washed with Ca²⁺ and Mg²⁺ free HBSS and centrifuged at 800 rcf for 10 min. at 4°C. The cell suspension was then passed through a 70 mm strainer (Foxx Life Sciences Cat.# 410-0002-OEM) to remove undigested connective tissue and large cell clumps. Dispersed cells were then counted and plated in poly-D lysine-coated plates. The following plating densities were used: 150,000 cells/well in 24 well plates, 300,000 cells/well in 12 well plates, and 5,000,000 cells/dish in 10 cm dishes. Arsenite 0.5mM was applied for 30 min. immediately before processing the neurons for analysis. DSP (dithiobis(succinimidyl propionate)) was added to the neurons before harvesting at a concentration of 0.1mM to the neurons in culture before processing them for western blot analysis.

Animals: B6.C-Tg(CMV-cre)1Cgn/J::B6.C-GR50 (GR₅₀) or B6.C-Tg(CMV-cre)1Cgn/J::B6.C-GFP (GFP) mice were generated as described. At 12 months, male mice were sacrificed, and cortices and spinal cords were collected and frozen in liquid N₂ or embedded in OCT. Frozen samples were stored at -80°C till the time of processing. All animal procedures were compliant with the ARRIVE guidelines and approved by IACUC. We have complied with all relevant ethical regulations for animal use.

Plasmids and siRNA: The following plasmids were used: pCDNA3_hSyn_Td_Tomato; pCDNA3_hU6_Flag_mCherry; pCDNA3_hU6_Flag_GR₅₀_mCherry; pCDNA3_hU6_Flag_GR₁₀₀_mCherry; pCDNA3_CMV_eGFP; pCDNA3_CMV_eGFP_Kapβ2. eGFP and eGFP_Kapβ2 were co-transfected with Td_Tomato at a DNA concentration ration of 1:2 (200ng/400ng) or 1:4 (200ng/800ng). eGFP and eGFP_Kapβ2 were co-transfected with mCherry, mCherry_GR₅₀ and mCherry_GR₁₀₀ at 1:1 ratio (400ng/400ng). Human TDP-43 and TDP-43-GFP were subcloned into pE-SUMO (LifeSensors, Malvern, PA) as described (McGurk et al., 2018). All plasmid inserts were sequenced and confirmed to be correct. To silence Kapβ2, 50 mM of ON-TARGETplus siRNA-TNPO1-Smart Pool (Horizon Discovery, L-086861-02-0005) was used. 50mM of AllStars Negative Control scramble siRNA (Qiagen Cat. #1027280) was used as control.

Transfection: Lipofectamine 2000 (Fisher Scientific Cat. #11-668-019) was used to transfect HEK cells and cortical neurons (1 mL Lipo2000/500ng total D.N.A.). HiPerfect Transfection Reagent (Qiagen Cat. #301705) was used to transfect siRNA (8mL HiPerfect/50nM siRNA). siRNA was transfected 24 hours before plasmids transfection. For cortical neurons, Lipo/DNA or RNA mix was incubated with the neurons for 1h, and then cell media were changed with fresh ones. For HEK293 cells, the mixture was incubated with the cells overnight.

Lentivirus constructs production: HEK293 cells were transfected with the following plasmids: pLenti_hSyn_Flag_GR₅₀_eGFP, psPAX2 (Addgene Plasmid #12260), pMD2.G (Addgene Plasmid #12259). Plasmids were diluted in DMEM, and PEI MAX was added as a transfection reagent. The mix was incubated with HEK293 cells confluent at 70% for 5h; the media was changed with a fresh one. Cells supernatant was collected 48h after transfection and centrifuged for 10 min at 2,000 rcf at 4°C. Lentivirus Concentrator (OriGene Technologies Cat. #TR30025) was added to the supernatant and incubated overnight at 4°C. To pellet the virus, the media was centrifuged at 500 rcf for 45 min at 4°C. Viral particles were then resuspended in Ca²⁺ and Mg²⁺ free PBS (Cytiva Cat. #SH30028. L.S.), aliquoted, and immediately frozen in liquid N₂.

Purification of Recombinant Proteins: His₆-SUMO1 N-terminally tagged TDP-43-WT and TDP-43-GFP were purified as described (McGurk et al., 2018) and overexpressed in BL21(DE3) RIL E.coli cells. These cells were then sonicated on ice in 50 mM HEPES (pH 7.5), 2% TritonX-100, 300 mM NaCl, 30 mM imidazole, 5% glycerol, 2 mM β-mercaptoethanol, and protease inhibitors (cOmplete, EDTA-free, Roche). The recombinant TDP-43 proteins were purified over Ni-NTA agarose beads (Qiagen) and eluted using 50 mM HEPES (pH 7.5), 150 mM NaCl, 300 mM imidazole, 5% glycerol, and 5 mM DTT. The buffer was exchanged with the same buffer without imidazole. The eluate containing the recombinant TDP-43 proteins was then frozen in liquid N₂ and stored as aliquots at -80°C until use. Kapβ2 was purified as described (Guo et al., 2018). In brief, E. coli BL21-CodonPlus(DE3)-RIL cells (Agilent) were transformed with GST-Tev-Kapβ2 plasmid, and expression was induced overnight at 25°C with 1 mM IPTG. Cells were pelleted and resuspended in Tris buffer (50 mM Tris pH 7.5, 100 mM NaCl, 1 mM EDTA, 20% (v/v) glycerol, 2 mM DTT, supplemented with protease inhibitors), then lysed by sonication. Cell lysate was then loaded onto glutathione Sepharose^TM 4 Fast Flow resin (GE Healthcare) and washed with Tris buffer, followed by ATP buffer (50 mM Tris pH 7.5, 100 mM NaCl, 1 mM EGTA, 0.5 mM MgCl₂, 5 mM ATP, 20% glycerol, 2 mM DTT, supplemented with protease inhibitors), then washed and eluted with Buffer A (20 mM imidazole, 75 mM NaCl, 1 mM EDTA, 20% (v/v) glycerol, 2 mM DTT). Finally, the protein was cleaved with Tev protease and purified on a HiTrap Q HP column (GE Healthcare) using a salt gradient. Purified protein was concentrated, flash-frozen, and stored at -80°C. Before the aggregation assay, the protein was thawed and centrifuged at 16,100 x g for 10 min to remove any preformed aggregates. Protein concentration was determined by Bradford assay (Bio-Rad, Hercules, CA). Equimolar concentrations of Kapβ2, TDP-43, and GR₂₀ (5mM) were used. 200nM HIS₆-SUMO-TDP-43-GFP and 100nM TAMRA-GR₂₀ were added to visualize TDP-43 and GR₂₀.

Peptides: Chemically synthesized GR₂₀ and Tetramethylrhodamine (TAMRA)-tagged GR₂₀ were obtained as a lyophilized powder from Peptide2.0 and dissolved in PBS.

Sample preparation: 100 mg of fresh frozen lumbar spinal cord samples were homogenized in 1% SDS using a 15ml Dounce Homogenizer. The homogenate was centrifuged at 3000 rcf for 20 minutes at 4°C to remove tissue debris. Protein estimation was performed on the clear homogenate using a BCA assay (Pierce BCA kit #23225). 30 mg of protein samples are loaded for Western blot analysis.

For mouse tissues: 50 mg of CNS tissues were chopped into small pieces, incubated in RIPA for 2 h, and homogenized by pipetting every 20/25 minutes. Samples were then sonicated using Bioruptor Pico-Diagenode.

For cell cultures: Cells were harvested in PBS and centrifuged at 500 rcf for 3 minutes. PBS was removed by suction, and samples were resuspended in 80 mL of RIPA with protease inhibitor (Sigma-Aldrich Cat# P2714-1BTL). Cells were incubated for 15 minutes at 4°C on an orbital shaker and then sonicated. After sonication, samples were centrifuged at 16,000 rcf for 5 minutes at 4°C. The supernatant was kept, and BCA was used to measure the protein content of each sample.

Western blot: Cell cultures and animal tissues: 15mg of protein extracts were loaded on pre-cast Stain Free gels (Bio-Rad Cat.#4568034) and run for 1hr at 100V. Gels were then transferred on PVDF membrane (Millipore Cat.#IPVH00010) overnight at 4°C at 30V and developed with primary and secondary antibodies at the indicated concentration. Before transfer, gels were cross-linked by UV light exposure for 5 minutes, and total protein content was then revealed on the membrane by UV light exposure. Membranes were developed with SuperSignal™ West Femto Maximum Sensitivity Substrate (Thermo Fisher Scientific Cat#34094) and imaged using ChemiDOC XRS+ System (Bio-Rad).

Human samples: The gels were UV-activated for 5 minutes before transferring onto the 0.22 mm nitrocellulose membrane. The semi-dry transfer was done using a Trans-Blot Turbo Transfer System (Bio-Rad) at 25mV for 10 minutes. After the transfer, a stain-free blot image was captured and used as total protein content. The blot was processed as previously described. The quantification was done using Image Lab Software (Bio-Rad); the intensity of the Kapβ2-positive band is normalized to its total proteins. Quantification was performed using Image Lab software. Adjusted volume was measured for each band and divided by the adjusted volume calculated, considering the entire lane of total proteins. The abundance of the protein of interest was normalized to the total amount of proteins in each lane, therefore removing variations associated with comparing abundance to a single protein arbitrarily chosen as housekeeping protein (e.g. GAPDH). Total protein normalization is more compatible with detecting and quantifying proteins of lower abundance.

Immunoprecipitation: Flag: Immunoprecipitation was carried out using anti-Flag agarose beads (mouse IgG agarose beads were used as control). Rat primary cortical neurons and tissues were extracted in IP buffer containing 50 mM Tris-HCl (pH 7.5), 150 mM NaCl, 1mM EDTA, 0.5% NP-40, 100 mM EDTA (pH 8.0) supplemented with protease inhibitor (Sigma-Aldrich Cat.# P2714-1BTL). After incubating at 4⁰C for 30 min., the samples were sonicated and then centrifuged for 10 min. at 10,000 rcf. Bradford assay was used to determine protein concentration. The agarose beads were added to 600 mg of cell lysates and incubated at 4°C for 24h under constant shaking. Cell lysates were centrifuged for 1 min. at 1,000 rcf to collect the supernatant. The pellets were washed five times, but only the first wash was collected. After the last centrifugation, the samples were eluted with SDS-PAGE loading dye at 95°C for 5 min. The entire eluate was run for Western blot. 2% of total lysate was used as input.

GFP and Kapβ2: Following manufacturer instructions, immunoprecipitation was carried out using magnetic beads (Thermo Fisher Scientific Cat.# 88804). Briefly, 25 mL of magnetic beads were conjugated with 10 mg of the desired antibody overnight and then crosslinked for 30 minutes using DSS (disuccinimidyl suberate). The obtained beads-antibody complex was incubated with cell or tissue lysates overnight. The next day, beads were washed, and the bound protein was eluted. The entire eluate was run for Western blot. 2% of total lysate was used as input. For Kapβ2 IP, a double volume of beads was used.

Quantification: The signal specifically corresponding to Kapβ2 and GFP bands was measured with ImageLab and the ratio was calculated between the respective adjusted volumes.

Immunofluorescence: Primary cortical neurons were fixed for 20 minutes at 37°C with 4% PFA and incubated in PermBlock solution (2% BSA, 0.3% Triton-x, 5% donkey serum in PBS). Primary antibodies were diluted in 0.1% BSA cells and were incubated overnight at 4°C on an orbital shaker. The following day, cells were washed two times with PBS at room temperature for 10 minutes and incubated with the appropriate secondary antibody for 1h at room temperature. Cells were washed twice in PBS and incubated with Hoechst solution (Thermo Fisher Scientific Cat.# 62249) for 10 minutes at room temperature to stain the nuclei. After a final rinse in PBS, coverslips were mounted on glass slides using AquaMount (Lerner Laboratories Cat.# 13800). Tissues were cryo-sectioned using Cryostar NX50. 20 mm thick sections were placed on glass slides and incubated for 10 minutes in 4% PFA. Slides were then washed in TBS-T for 10 minutes and incubated for 1h at room temperature in PermBlock solution (2% BSA, 0.3% Triton-x, 5% donkey serum in PBS) and then incubated with primary antibody diluted in 1% BSA, 0.3% Triton-x and 1% donkey serum overnight at 4°C in a humidified chamber. The following day, slides were washed with TBS-T, incubated with the appropriate fluorescent secondary antibody for 1 hour at room temperature, rinsed twice in TBS-T, incubated for 10 minutes with Hoechst (Thermo Fisher Scientific Cat.# 62249), and mounted on coverslips using AquaMount (Lerner Laboratories Cat.# 13800) for imaging analysis.

Antibodies and dilutions: Primary antibodies: anti-Kapβ2 (RRID:AB_2206884, Santa Cruz Biotechnology, Inc Cat.# sc-101539; WB dilution 1:1000); anti-Kapβ2 (RRID:AB_262123, Sigma Cat.# T0825; IF dilution 1:500); anti-Map2 (RRID:AB_2138178 Novus Cat.#NB300-213; IF dilution 1:2000); anti-GFP (RRID:AB_11042881, Proteintech Cat.#50430-2-AP, WB dilution 1:1000); anti-TDP-43 (RRID:AB_615042, Proteintech Cat.#10782-2-AP; IF dilution 1:500); anti-GAPDH (RRID:AB_2107436, Proteintech Cat.#60004-1-Ig; WB dilution 1:1000); Neun (D4G40) XP (RRID:AB_2651140, Cell Signaling Technology Cat.#24307; IF dilution 1:500): FUS (RRID:AB_2247082, Proteintech Cat.#11570-1-AP; IF:1:500)

Secondary HRP-conjugated antibodies: HRP-Conjugated anti-rabbit (RRID:AB_772191, Cytiva Cat.# NA9340-1ML; WB dilution 1:10,000); HRP-Conjugated anti-mouse (RRID:AB_772193, Cytiva Cat.# NA9310-1ML; WB dilution 1:10,000); HRP-Conjugated anti-rat (RRID:AB_2936877, Sigma Cat.# AP136P; WB dilution 1:3,000);

Secondary fluorescent antibodies: AlexaFluor-546 anti-mouse (RRID:AB_2534012, Thermo Fisher Scientific Cat.# A10036; IF dilution 1:1,000); AlexaFluor-647 anti-rabbit (RRID:AB_2536183, Thermo Fisher Scientific Cat.# A31573; IF dilution 1:1,000); AlexaFluor-647 anti-chicken (RRID:AB_2762845, Thermo Fisher Scientific Cat.# A32933; IF dilution 1:1,000).

Extraction of total RNA and quantitative Real-Time PCR: Total RNA was extracted using Trizol Reagent (Ambion Cat.# 15596026) following the manufacturer’s instructions. Briefly, cells were lysed in Trizol Reagent and homogenized using a syringe for insulin. RNA was then extracted in isopropanol. Samples were centrifuged for 15 min. at 12,000 rcf. The RNA-containing pellet was washed with 75% ethanol and centrifuged for 5 min. at 7,500 rcf. Dried pellets were resuspended in DEPC water, and concentration was assessed with Nanodrop. 500ng of total RNA was first treated with DNAse I (Thermo Fisher Scientific Cat.# EN0521) and retrotranscribed using SuperScript™ IV Reverse Transcriptase (Thermo Fisher Scientific Cat.# 18090010). The cDNA was diluted 1:2 and used for qRT-PCR, which was performed using the following TaqMan assay: GAPDH-VIC (Thermo Fisher Scientific Cat# 4352338E) and Kapβ2-FAM (Thermo Fisher Scientific Cat.# 4331182 Assay# Rn01489969_m1). 50ng of cDNA was used per reaction. Samples were analyzed using the Design and Analysis tool by Thermo Fisher Scientific. Results are shown as fold change using the 2^DDCt method.

Confocal Imaging Analysis: Confocal images were taken on a Nikon A1R confocal microscope with a 60X objective. Z-stacks were taken for each image. 0.2 mm Z-stacks were taken to perform colocalization studies and puncta counting using a built-in function in Imaris. One mm stacks were taken for TDP-43 nuclear intensity studies. Maximum intensity projection was then generated through a built-in function of the A1R analysis software. The nuclei of each co-transfected cell were selected as the region of interest, and TDP-43 intensity was assessed. The mask was made on the mCherry channel to analyze aggregate content, and the TDP-43 fluorescence was subsequently evaluated. For Kapb2 nuclear and cytoplasmic quantification maximum intensity projections were generated, and ROI were drawn around nucleus and cytoplasm and intensity referred to the Kapβ2 chnnel was measured. The ratio between nuclear and cytoplasmic intensities in the same cell was then calculated. FRAP experiment was performed in green (GFP or GFP-Kap2) and red (mCherry, GR50-mCherry, GR100-mCherry) fluorescent cells. Aggregates were selected as regions of interest and photobleached for 2 sec. with blue light with laser power at 25% power. Following photobleaching, images were taken every 2 seconds for 3 minutes and analyzed with Nikon A1R software. Normalization and curve fitting were performed using the online tool EasyFRAP (normalization was performed using double normalization, and curve fitting was performed using double equation). The fluorescence recovery intensity curves were represented over time using Prism GraphPad 9.0. The mobile fraction in FRAP experiments represents the proportion of molecules or particles that can freely move within the sample after photobleaching. The mobile fraction was calculated using the following formula: Mobile Fraction=Mobile Fluorescence Recovery\Initial Fluorescence Loss where the mobile fluorescence recovery is the increase in fluorescence signal after photobleaching, indicating the recovery of mobile molecules, and the initial fluorescence Loss is the initial decrease in fluorescence signal due to photobleaching.

Time-lapse microscopy: Cells were imaged every two days after transfection. 20´ objective was used, and 25 fields/wells were acquired. Only cells positives for both Kapβ2 and mCherry were counted. Cells were scored dead when they disappeared from the field of view or clear signs of neurite fragmentation appeared. Kaplan Meier curves were then constructed in Prism GraphPad 9.0.

Visualization of TDP-43 Condensates: His₆-SUMO1-TDP-43 (5 mM) was incubated with an equimolar concentration of (GR)₂₀ for 1 hour at room temperature in PBS. 200 nM His₆-SUMO1-TDP-43-GFP and 100 nM TAMRA-(GR)₂₀ were added to visualize TDP-43 and (GR)₂₀. The reaction mixture was incubated at room temperature for 1h, and samples were spotted onto a coverslip and imaged by DIC and fluorescent microscopy using a Leica DMi8 Inverted microscope. To inhibit condensation, 5mM Kapβ2 was added at the beginning of the assay.

Proteomics analysis: Publicly available data sets from Liu, F., Morderer, D., Wren, M.C. et al. were downloaded from massive.ucsd.edu (MassIVE MSV000088581). Analysis and statistical comparison of protein abundance were performed using basic packages (ggplot, ggplot2) of R (version 4.0.5) and Rstudio (version 2022.07.2).

Transcriptomics analysis: In this study, we performed a transcriptomic analysis using publicly available datasets to investigate Kapβ2 expression patterns in iPS-derived cortical neurons and ALS patients' cerebral tissue. For the AnswerALS data: The raw sequencing data were processed using quality control metrics and aligned to a reference genome using Kallisto. Differential gene expression analysis was performed using DESeq2. Kapβ2 expression levels across samples were analyzed with ggplo2, and statistical analysis was carried out using GGPUBR. For the human tissue data, raw counts were downloaded from https://github.com/NathanSkene/ALS_Human_EWCE. Kapβ2 expression levels across samples were analyzed with ggplo2; statistical analysis was done using GGPUBR.

Statistics and Reproducibility: All experiments were performed at least 3 times (n³3). At least 3 mice/group were used. For live-cell imaging, >150 neurons/condition were counted for each n. For colocalization, puncta counting, and TDP-43 intensity, ³10 neurons/condition were counted for each n. For the FRAP experiment, ³5 cells/conditions were analyzed for each n. When two groups were compared, the Kolmogorov-Smirnov test assessed the normality of each population. If the two populations' results were normally distributed, then the unpaired t-student test was performed using the built-in function in Prism GraphPad 9.0. If one or both populations were not normally distributed, then the Mann-Whitney test was used to compare ranks, or the Kolmogorov-Smirnov test was used using Prism GraphPad 9.0. The log-rank and Cox proportional hazards tests compared the Kaplan Meier curves and the statistical significance between curves was assessed with Cox (GraphPad software). For the FRAP experiment, statistical differences between the curves were evaluated by curve analysis (one-way ANOVA test) in Prism GraphPad 9.0.