Downregulation of the silent potassium channel Kv8.1 increases motor neuron vulnerability in amyotrophic lateral sclerosis

Abstract While voltage-gated potassium channels have critical roles in controlling neuronal excitability, they also have non-ion–conducting functions. Kv8.1, encoded by the KCNV1 gene, is a ‘silent’ ion channel subunit whose biological role is complex since Kv8.1 subunits do not form functional homotetramers but assemble with Kv2 to modify its ion channel properties. We profiled changes in ion channel expression in amyotrophic lateral sclerosis patient–derived motor neurons carrying a superoxide dismutase 1(A4V) mutation to identify what drives their hyperexcitability. A major change identified was a substantial reduction of KCNV1/Kv8.1 expression, which was also observed in patient-derived neurons with C9orf72 expansion. We then studied the effect of reducing KCNV1/Kv8.1 expression in healthy motor neurons and found it did not change neuronal firing but increased vulnerability to cell death. A transcriptomic analysis revealed dysregulated metabolism and lipid/protein transport pathways in KCNV1/Kv8.1-deficient motor neurons. The increased neuronal vulnerability produced by the loss of KCNV1/Kv8.1 was rescued by knocking down Kv2.2, suggesting a potential Kv2.2-dependent downstream mechanism in cell death. Our study reveals, therefore, unsuspected and distinct roles of Kv8.1 and Kv2.2 in amyotrophic lateral sclerosis–related neurodegeneration.


Introduction
Amyotrophic lateral sclerosis is a devastating neurodegenerative disease that results in a rapid and progressive loss of motor neurons (MNs) in the motor cortex and spinal cord. 1,2dvances in induced pluripotent stem cell (iPSC) technology not only enable in vitro modelling of amyotrophic lateral sclerosis but also facilitate the discovery of novel molecular targets in diseased neurons. 3][6][7] Previously, our lab found that a positive feed-forward cycle of endoplasmic reticulum stress and abnormal excitability drives neuronal death in patient-derived MNs harbouring the familial amyotrophic lateral sclerosis A4V mutation in superoxide dismutase 1 (SOD1 A4V/+ ) and other familial amyotrophic lateral sclerosis MNs. 4,5,8ecause of the association between abnormal excitability and accelerated disease progression in amyotrophic lateral sclerosis, 9 we set out to find its molecular basis in MNs with the SOD1(A4V) mutation.However, while we identified a marked downregulation of the silent ion channel subunit KCNV1/Kv8.1 in patient MNs, we found that a loss of KCNV1 did not affect excitability but did increase vulnerability to cell death.We also found that KCNV1 interacts with Kv2.2 that may contribute to amyotrophic lateral sclerosis pathogenesis.

HB9::GFP transfection using zinc finger nuclease
For transfection of a HB9::GFP reporter into each of the four iPSC lines, a 1 kb HB9 promoter fragment (gift from Hynek Wichterle) controlling the expression of myristoylated GFP was inserted into a donor plasmid specific for the AAVS1 locus (Sigma).Subsequently, 2.5 million iPSCs were treated with accutase and electroporated using the Neon Transfection System (100 μL tip; 1600 V voltage, 20 ms width, 1 pulse; Life Technologies) with 2 μg of AAVS1 ZFN plasmid and 6 μg of the modified AAVS1 donor plasmid.After nucleofection, cells were plated on Matrigel with mTeSR1 in the presence of 10 μM ROCK inhibitor.After 48-h puromycin selection, surviving clonal colonies were individually passaged and genomic DNA was extracted.Polymerase chain reaction (PCR) was used to confirm proper targeting of the cassette.Expression of the reporter was verified using expression of the GFP and the MN marker Isl1.

MN differentiation
MN differentiation for patch-seq and patch-quantitative PCR (qPCR) was carried out as described previously 10 with modifications 11 in a 24-day protocol based on initial neuralization with SB431542 (10 μM, Sigma Aldrich) and dorsomorphin (1 μM, Stemgent) and MN patterning with retinoic acid (1 μM, Sigma) and a small smoothened agonist 1.3 (1 μM, Calbiochem).Differentiated MNs were dissociated using accutase, filtered with a 70 µm filter and isolated by activated flow cell sorting using HB9::GFP expression.MNs were plated in the presence of P0 mouse glial cells and were maintained in Neurobasal media; supplemented with N2 and B27 (Invitrogen) and 10 ng/mL each of brainderived neurotrophic factor, glial cell line-derived neurotrophic factor and ciliary neurotrophic factor (R&D) and ascorbic acid (0.4 μg/mL, Sigma); and fed every 2-3 days.
MN differentiation for cell death assays and reverse transcription (RT)-qPCR was carried out with a modified 14-day protocol described in Klim et al. 12 Briefly, iPSC were dissociated using accutase and plated with 1.5 million cells per 10 cm dishes coated with Geltrex.One day after plating, the medium was changed to differentiation medium (half Neurobasal and half DMEM/F12 supplemented with B27 and N2 supplements, GlutaMax and non-essential amino-acids) from StemFlex™ Medium (Gibco).SB431542 of 10 µM, 100 nM LDN-193189, 1 µM retinoic acid and 1 µM smoothened agonist were treated on Days 0-5, and 5 µM DAPT, 4 µM SU-5402, 1 µM retinoic acid and 1 µM smoothened agonist were treated on Days 6-14.The differentiated cells were dissociated using accutase and then filtered with a 70 µm filter.The single-cell suspension was incubated with PE Mouse Anti-Human CD56 (BD Biosciences #555516) and then Anti-R-Phycoerythrin Magnetic Particles (BD Biosciences #557899) in magnetic cell separation buffer (phosphate-buffered saline with 2 mM ethylenediaminetetraacetic acid and 0.5% bovine serum albumin) on ice.The sorted MNs were plated in the presence of P0 mouse glial cells and were maintained in the same medium and condition as described above.

Electrophysiology
Current clamp of iPSC-derived MNs: Purified and differentiated neurons on P0 mouse glial cells were identified microscopically, and whole-cell current clamp recordings were conducted using a MultiClamp 700B amplifier.Data were digitized with a Digidata 1440A A/D interface and recorded using pCLAMP10 (Molecular Devices).Borosilicate glass pipettes were pulled on a P-97 puller (Sutter Instruments).The external solution consisted of 140 mM NaCl, 5 mM KCl, 2 mM CaCl 2 , 2 mM MgCl 2 , 10 mM HEPE (4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid) and 10 mM D-glucose, pH 7.4, adjusted with NaOH.The internal solution consisted of 135 mM K-gluconate, 10 mM KCl, 1 mM MgCl 2 , 5 mM EGTA (ethylene glycol bis(2-aminoethyl)tetraacetic acid) and 10 mM HEPE (4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid).Fast and slow capacitance transients and whole-cell capacitance were compensated using the automatic capacitance compensation in voltage clamp mode on the MultiClamp 700B.After currents were injected to bring the membrane potential to −60 mV in current clamp mode, a depolarizing current ramp (0-700 pA in 1 s) or series of steps (500 ms steps from 0 to 100 pA) were applied to measure the firing frequency and the action potential characteristics.Spikes were counted using a criterion of a peak voltage > −10 mV and amplitude > 20 mV.Frequency of firing was calculated from all the spikes during stimulation and instant frequency from the first two spikes.Action potential characteristics including trough voltage were analysed from the first action potential waveform.

Singl-cell collection
For patch-qPCR, single neurons were picked immediately after recording, transferred into 8-well strip tubes containing 5 μL lysis buffer (CellsDirect™ One-Step RT-qPCR Kit, #11753100) and then placed on dry ice.Tubes were kept at −80°C until the next pre-amplification step.For patchseq, the single neurons were collected in 5 μL of 1× TCL buffer (Qiagen, #1031576) in each well of 96-well PCR plates (Eppendorf, #951020401) on dry ice and then kept at −80°C until complementary DNA (cDNA) construction.

Single-cell RNA sequencing
Single-cell RNA sequencing (RNA-seq) libraries were generated using the Smart-seq2 protocol 13 with minor modifications. 14All libraries were prepared by the Broad Technology Labs and sequenced at the Broad Genomics Platform.Briefly, total RNA from single cells was purified using RNA-SPRI beads.Poly(A)+ mRNA was converted to cDNA and amplified.cDNA was subjected to transposonbased fragmentation using dual-indexing to barcode each transcript fragment with a combination of barcodes specific to a single cell.Barcoded cDNA fragments were then pooled for sequencing.Sequencing was carried out as paired-end 2 × 25 bp with additional eight cycles for each index.To obtain expression values for each cell, the data were demultiplexed and aligned to the human genome (hg19) using Tophat version 2.0.10 with default settings. 15Transcripts were quantified by the BTL computational pipeline using Cuffquant version 2.2.1, 16 and raw counts of reads mapped per gene were extracted from aligned bam using FeatureCounts (subread-1.5.0-p1) with default settings.For visualization purposes, expression levels were converted to log-space by taking the log2(FPKM + 1) or by normalizing read counts to transcripts per million using log2(TPM + 1).To identify genes that are differentially expressed between distinct electrophysiological classes or genotypes, we first filtered cells with poor electrophysiological recordings or low sequencing quality (library size greater or less than three median absolute deviations from the median).Single-cell differential expression analysis was performed using SCDE (single-cell differential expression) 17 with default settings and batch effects corrected for when calculating expression differences.

Single-cell RT-qPCR: primer design
DELTAgene assays (Fluidigm) were designed for 279 genes (listed in the Supplementary Table 3), including housekeeping genes (GAPDH and ACTB), MN markers (MNX1, CHAT, ISL1 and SLC18A3), a glial cell marker (mGfap), voltage-gated ion channels and ligand-gated ion channels.The assays are designed to cross an intron.A nested primer strategy (outer and inner pairs of forward and reverse primers) was utilized.Outer primer pairs were used for the preamplification, and inner primer pairs were used for RT-qPCR in the Biomark.For outer primers, the oligos were synthesized by IDT and dissolved at a concentration of 200 μM in H 2 O, and 267 gene primer pairs were mixed in a 15 mL falcon tube at a final concentration of 200 nM each primer.For inner primers, the oligos were synthesized by IDT and dissolved at a concentration of 50 μM in H 2 O.To make 10 μL of 20 μM primer pairs per each gene in V-shaped 96-well plate (Eppendorf, #951020401), 4 μL of forward and reverse primers were added to 2 μL1 H 2 O.

Single-cell RT-qPCR: cDNA synthesis and pre-amplification
We used a protocol adapted from Fluidigm (Application Note MRKT00075e) that combines reverse transcription and preamplification (called reverse transcription-specific target amplification) using CellsDirect™ One-Step RT-qPCR Kit.A mixture consisting of 2.8 μL of multiplex outer primers plus 0.2 μL of the enzyme mix (SuperScript® III Reverse Transcriptase and Platinum® Taq DNA Polymerase) was added to each single cell in lysis buffer in the 8-well strips.Following centrifugation, the 8-well strips were moved to a thermal cycler and subjected to the following protocol: 50°C, 60 min for reverse transcription; 95°C, 2 min for Taq enzyme activation; and 20 cycles of 95°C, 15 s and 60°C, 4 min.The reactions were kept at −80°C until the next tests.
To determine optimal dilution for following singlecell qPCR, a few randomly picked reactions were diluted in H 2 O at 25-, 50-, 75-and 100-fold and tested with GAPDH, a housekeeping gene, by qPCR.A dilution of 25-fold was selected for subsequent experiments.All the reactions were diluted and tested with GAPDH, a positive control, and mGfap, a negative control, to determine whether we had properly picked neurons without glial cells.

Single-cell RT-qPCR: data acquisition and analysis
Diluted reactions were analysed by qPCR using 96.96Dynamic Array™ Integrated Fluidic Circuits (IFCs) and the Biomark™ HD system from Fluidigm.The IFCs were processed, and the instruments were operated in the BCH IDDRC Molecular Genetics Core Facility according to the manufacturer's procedures for analysing DELTAgene assays.Three IFCs were used to analyse the 96 samples for the total of 267 assays, and all IFCs included housekeeping genes, GAPDH and ACTB, and a blank.In single-cell qPCR analysis, a Ct of 30 was used as the background value for all real-time signals.Expression levels were calculated by 30 Ct value, assigning a value of 0 when Ct > 30.Hierarchical clustering and heat map were generated with R software.

RNA extraction and RT-qPCR
RNA was extracted from MN culture using RNeasy Micro Kit (Qiagen), and then, cDNA was generated using SuperScript Vilo Synthesis Kit (Invitrogen) following manufacturer's manuals.qPCR was performed on Applied Biosystems 7500 machine (Life Technologies) using Fast SYBR Green Master Mix (Roche).

RNA-seq analysis
RNA-seq libraries were prepared using the TruSeq Stranded RNA Kit with Ribo-Zero Gold to enrich messenger RNA (mRNA).The libraries were then sequenced on a NovaSeq6000 platform, generating paired-end reads of 100 base pairs in length.To ensure the accuracy of downstream analysis, the reads were classified into five categories using Xenome (v1.0.0), a tool designed to remove reads derived from unwanted sources such as contaminating DNA or RNA from other species.Reads classified as human were then aligned to the human genome (Hg38) using STAR aligner (v2.7.5c).Average input read counts were 31.00 ± 3.61 M(SD) and average percentage of uniquely aligned reads were 88.04 ± 0.03(SD)%.Total counts of read fragments aligned to known gene regions within the human ensemble gene model annotation (GRCh38) are used as the basis for quantification of gene expression.Fragment counts were derived using HTSeq program (ver 0.12.4).Quality control measures were performed to assess the quality of the data, including base quality, mismatch rate and mapping rate to the whole genome.Additionally, repeats, chromosomes, key transcriptomic regions (exons, introns, UTRs and genes), insert sizes, AT/GC dropout, transcript coverage and GC bias were assessed to identify potential issues in the library preparation or sequencing.To identify genes that were differentially expressed between conditions, lowly expressed genes were removed and genes with counts per million (CPM) > 0.15 in at least three samples were selected for downstream analysis.Differential expression analysis was conducted using the Bioconductor package EdgeR (ver 3.14.0),which uses a negative binomial model to account for the variability in the data.Expressed genes were sorted by directional pVal and used as input for Gene Set Enrichment Analysis.MsigDB (ver7.0)was used as reference gene set.Raw and processed data were deposited within the Gene Expression Omnibus repository (www.ncbi.nlm.nih.gov/geo).

Statistical analysis
Students' t-test was performed using GraphPad.

Results
To explore the cellular pathways involved in the hyperexcitability component of amyotrophic lateral sclerosis pathogenesis, we employed a 'patch-seq' strategy that enables both neuronal excitability and gene expression to be measured simultaneously at a single cell level. 8Compared to prior studies where a heterogeneous population of differentiated neurons were used and neuronal maturity not characterized, 4,9 we focused on a homogenous population of mature MNs that fire action potentials robustly.We introduced a GFP reporter under control of the Hb9 promoter into SOD1 A4V/+ iPSC lines and their isogenic controls 4 using an AAVS1-sHb9-GFP plasmid.The MNs were FACS purified after differentiation.Immunostaining for ISL1 (a MN marker) and MAP2 (a neuronal marker) revealed that most of the neurons had a MN identity (91.5 ± 1.4% of 39b SOD1 A4V/+ neurons were ISL1+ MAP2 + and 92.7 ± 1.7% of 39b-cor SOD1+/+; Supplementary Fig. 1A).To accelerate neuronal maturation, we cultured the purified MNs with glia for 3 weeks and then performed whole-cell current clamp recordings followed by single MN isolation for single-cell RNA-seq (Smart-seq2; Fig. 1A).Two independent batches comprising a total of 181 MNs were analysed, and ∼8000 genes per neuron were detected (Supplementary Table 1).
There were no significant differences in housekeeping, MN or neuronal marker gene expression between the SOD1 mutant and control MNs (Supplementary Fig. 1B).A significant difference in the expression of 63 genes was however detected by the single-cell patch-seq analysis of 39b SOD1 A4V/+ and 39b-cor SOD1 +/+ MNs (Fig. 1A), including a marked downregulation of KCNV1 (Supplementary Table 2).We independently confirmed this finding with a single-cell patch-RT-qPCR analysis focused specifically on the differential expression of 279 ion channel genes (Fig. 1B; Supplementary Table 1).There was no difference in voltage-gated Na + or Ca 2+ channel expression (Supplementary Fig. 2), and the expression signatures of most K + channel subtypes were also generally very similar between diseased and healthy MNs.However, KCNV1/Kv8.1 was significantly downregulated in 39b SOD1 A4V/+ MNs relative to their isogenic controls (39b-cor SOD1 +/+ MNs) both in the single-cell patch-RT-qPCR analysis and the patch-seq study (Fig. 1B).
To determine whether KCNV1 downregulation is consistent across different cell lines, an independent patient-derived iPSC line with the same mutation, RB9d SOD1 A4V/+ , and its isogenic control RB9d-cor SOD1 +/+ were also investigated.RB9d SOD1 A4V/+ MNs showed a similar downregulation of KCNV1/Kv8.1 compared to the control RB9d-cor SOD1 +/+ neurons (Fig. 1C).To test if this observation also holds for other amyotrophic lateral sclerosis mutations, we examined KCNV1 expression in MNs differentiated from iPSCs harbouring C9orf72 hexanucleotide repeat expansions.We also found KCNV1 downregulation in C9orf72 repeat expansion iPSC MNs compared to their isogenic control MNs (provided by Dr. Coppola; Fig. 1D), as well as in another independent C9orf72 mutant line compared to healthy controls (Fig. 1E).The downregulation of KCNV1 in diverse amyotrophic lateral sclerosis MN models and lines suggests that this could contribute to disease progression.KCNV1/Kv8.1 is a 'silent' potassium channel subunit that, while it has a structure typical of voltage-activated potassium channels, cannot form functional homomeric channels.Instead, when it forms heteromeric channels with Kv2, it downregulates Kv2 channel density and modulates channel gating. 18,19To study the role of KCNV1/Kv8.1 in healthy MNs, we produced a knockout of KCNV1 by CRISPR gene editing in the control 39b-cor SOD1 +/+ iPSCs (Supplementary Fig. 3A).Introduction of INDEL mutations resulted in early stop codon formation at the N-terminal that significantly reduced KCNV1 expression (Supplementary Fig. 3B).The KCNV1 knockout MNs did not exhibit altered spontaneous action potential firing, as assessed by multi-electrode array recordings, when compared to KCNV1 intact controls, indicating that loss of KCNV1 does not directly affect MN excitability (Fig. 2A).
Amyotrophic lateral sclerosis is characterized by a progressive loss of MNs, and we wondered whether KCNV1 may impact neuronal survival.To test this, we induced cell death using MG132, a proteasome inhibitor that promotes mutant SOD1 protein aggregation. 5KCNV1 knockout MNs exhibited a higher rate of MG132-induced cell death when compared to non-CRISPR-edited controls (Fig. 2B).To independently confirm this result, we also knocked down KCNV1 expression in control 39b-cor SOD1 +/+ MNs to mimic the KCNV1 downregulation in amyotrophic lateral sclerosis, using two different small hairpin RNAs (shRNAs) targeting KCNV1/Kv8.1,and found that the knockdown of KCNV1 in these healthy control MNs increased MG132-induced cell death significantly, while not affecting basal survival (Fig. 2C-E; Supplementary Fig. 4).A downregulation of KCNV1 in MNs, increases, therefore, their vulnerability to cell death without changing their excitability.
How does a reduction of KCNV1/Kv8.1 contribute to amyotrophic lateral sclerosis-related MN cell death?To explore this, we studied transcriptomic alterations caused by a reduction in KCNV1 expression.Many genes were significantly differently expressed in KCNV1 shRNA knockdown MNs compared to their controls (Fig. 3A; Supplementary Table 3), including genes involved in lipid metabolism, protein translation and membrane transport pathways (Supplementary Fig. 5A).Gene sets related to the endoplasmic reticulum membrane, metabolism and catabolism were among the top sets downregulated, while gene sets associated with neuron projection and intracellular transport were among the top sets that were upregulated (Fig. 3B).We validated the differential expression of several of the genes (Fig. 3C), including NEK1 and OPTN, two amyotrophic lateral sclerosis-associated genes that participate in proteostasis regulation 20,21 ; RPS3A, a ribosomal and chaperone protein that counteracts α-synuclein aggregation 22 ; STMN2, a microtubule-associated protein involved in TDP43 pathology 12,23 ; and VAMP3, a vesicle-associated membrane protein that directs transport of proteolipid proteins. 24ollectively, the transcriptomic analysis identified several overlapping molecular pathways between amyotrophic lateral sclerosis and KCNV1 knockdown MNs.Expression of KCNV1 was also reduced by tunicamycin treatment that interferes with protein glycosylation and results in misfolding, suggesting a possible involvement of KCNV1 in intracellular protein transport (Supplementary Fig. 5B).KCNV1/Kv8.1 forms heteromers with Kv2 channels, and we therefore examined if Kv2 was also involved in the MN cell death induced by a reduction of KCNV1 expression.1][32] Recent studies show that endoplasmic reticulum-plasma membranelocated Kv2 ion channels regulate endoplasmic reticulum Ca 2+ uptake and release in neurons [33][34][35] as well as lipid metabolism 36 that could impact multiple biological processes, including synaptic transmission, receptor signalling, membrane trafficking and cytoskeletal dynamics. 37e found that the expression of KCNB2, which encodes the Kv2.2 ion channel, was significantly decreased in KCNV1 knockdown MNs (Fig. 3D).Potentially, this change could be either protective or pathogenic.To explore this, we knocked down KCNB2/Kv2.2expression in 39b SOD1 A4V/+ ALS MNs and then induced cell death using MG132.We found that the survival of SOD(A4V) amyotrophic lateral sclerosis MNs was significantly increased by KNCB2 shRNA knockdown when compared to SOD1(A4V) MNs with no KNCB2 knockdown (Fig. 3E).Suppression of Shab, a Drosophila gene orthologous to several human genes including KCNB2 (https://ncbi.nlm.nih.gov/gene/38352), was recently reported to rescue the eye degeneration found in a C9orf72 expansion Drosophila model. 38We also observed a trend of increased survival in C9orf72 patient MNs when the expression of KCNB2 was suppressed (Supplementary Fig. 6).It is therefore possible that a reduction in KCNV1/Kv8.1 increases MN vulnerability through alterations in Kv2.2 ion channels and that the transcriptional downregulation of Kv2.2 in KCNV1/Kv8.1 knockdown MNs is, therefore, a protective compensatory response.

Discussion
Potassium ion channels are the most diverse and widely distributed ion channels in mammals and play critical roles in controlling membrane excitability.Recently, other functions of potassium channels have been reported, suggesting they modulate cellular Ca 2+ metabolism, 33,34 lipid metabolism, 36 apoptosis 39 and neuron-microglia interactions. 40Our data now reveal that an experimental reduction in the expression of the silent Kv8.1 subunit increases MN vulnerability and that vulnerable amyotrophic lateral sclerosis MNs have a reduction in Kv8.1 expression.We suggest this may occur through a loss of interaction with those Kv2.2 channels that could lead to the dysregulation of multiple pathways,  including lipid metabolism and membrane transport.While the expression of Kv8.1 and Kv2.2 have not been reported to be altered in SOD(A4V) transgenic mice, KCNV1/Kv8.1 is reduced in the motor cortex of sporadic amyotrophic lateral sclerosis patients. 2he precise underlying molecular machinery of Kv8.1-Kv2.2interactions in amyotrophic lateral sclerosis MNs need now to be defined, since this may open opportunities for novel therapeutic interventions.We cannot completely rule out a potential involvement of the ion-conducting functions of Kv2 in the MN disease state.Kv2 channels are important in regulating neuronal firing, 41 and BACE2, a protease that generates amyloid β, cleaves the C-terminal of Kv2, resulting in reduced K currents and increased neuronal apoptosis. 42Kv2 inhibition also protects against the degeneration of pancreatic β cells. 43Nevertheless, we did not observe any differential effects of the Kv2 blocker GXTX 44 on the KCNV1 knockout MNs (Supplementary Fig. 7A) or a significant rescue of etoposide-induced cell death on blocking Kv2 (Supplementary Fig. 7B).It is possible that Kv2 plays a multifaceted and complex role in neurodegeneration, and additional work will be required to reveal if its involvement in amyotrophic lateral sclerosis is via its ion-conducting action or other functions.
While we have identified a novel role for a silent potassium ion channel subunit in driving amyotrophic lateral sclerosisrelated neurodegeneration, this does not explain the altered membrane excitability previously identified in SOD1(A4V) MNs.Besides KCNV1, another silent potassium channel subunit, KCNK1, was also downregulated in our study (Fig. 1B).KCNK1 belongs to the two-pore domain potassium channel family that has diverse functions including osteoclastogenesis. 45Exactly what drives the altered excitability present in amyotrophic lateral sclerosis MNs and whether ion channel subunits like KCNK1 play a role now need to be addressed, together with discovering how Kv8.1 protects MNs.

Figure 2 Figure 3
Figure 2 Loss of KCNV1/Kv8.1 increases MG132-induced neuronal cell death.(A) Spontaneous activity of KCNV1 CRISPR knockout(D6) and control (G2) MNs recorded on a multiple electrode array.Average spike rates of active electrodes measured (n = 23-30 different 4-week co-culture replicates and average spike rates calculated from eight electrodes for each co-culture).(B) After culturing MNs with glia for 4 weeks, MG132 (10 µM) was added into the culture to induce cell death.Forty-eight-hour post-treatment, the cell culture was fixed and stained with MN markers: hN (human nucleus marker), MAP2 (mature neuron marker), and Isl1 (MN nucleus marker).The number of hN + Isl1 + cells or hN + MAP2 + cells were quantified and normalized to dimethylsulfoxide-treated control for each cell line (n = 12 different co-culture replicates from two independent differentiations, with n = 6 for each batch; average count calculated from >10 image fields for each co-culture replicate).A greater cell death occurred in KCNV1 knockout than control MNs.(C) Outline of 2D differentiation, NCAM magnetic cell separation, glia co-culture, shRNA treatment, followed by RT-qPCR analysis and cell death assays.(D) KCNV1 expression reduction in non-mutant 39b-cor SOD1 +/+ MNs treated with a lentivirus encoding KCNV1 shRNA (n = 4 different co-culture replicates from one batch of differentiation) compared to those treated with a control GFP targeting shRNA.(E) 39b-cor SOD1 +/+ MNs treated with KCNV1 shRNA show less survival after MG132 treatment than control GFP shRNA-treated MNs (n = 6 different co-culture replicates; average count calculated from >10 image fields for each co-culture replicate).Statistical significance by Student's t-test (**P ≤ 0.01, ***P ≤ 0.001, **** P ≤ 0.0001).