Reduction of Kcnt1 is therapeutic in mouse models of SCN1A and SCN8A epilepsy

Developmental and epileptic encephalopathies (DEEs) are severe seizure disorders with inadequate treatment options. Gain- or loss-of-function mutations of neuronal ion channel genes, including potassium channels and voltage-gated sodium channels, are common causes of DEE. We previously demonstrated that reduced expression of the sodium channel gene Scn8a is therapeutic in mouse models of sodium and potassium channel mutations. In the current study, we tested whether reducing expression of the potassium channel gene Kcnt1 would be therapeutic in mice with mutation of the sodium channel genes Scn1a or Scn8a. A Kcnt1 antisense oligonucleotide (ASO) prolonged survival of both Scn1a and Scn8a mutant mice, suggesting a modulatory effect for KCNT1 on the balance between excitation and inhibition. The cation channel blocker quinidine was not effective in prolonging survival of the Scn8a mutant. Our results implicate KCNT1 as a therapeutic target for treatment of SCN1A and SCN8A epilepsy.


Introduction
Developmental and epileptic encephalopathies (DEEs) are among the most severe epileptic disorders.The typical disease course begins with onset of seizures during the first year of life, followed by developmental delay, movement disorders, intellectual disability, sleep disturbances, and feeding difficulties (Scheffer and Nabbout, 2019;Meisler et al., 2021;Johannesen et al., 2022).Seizures are often resistant to treatment with current antiepileptic drugs (Scheffer and Nabbout, 2019;Meisler et al., 2021;Johannesen et al., 2022).
Many DEEs result from mutations in sodium and potassium channel genes (Lindy et al., 2018;Symonds et al., 2019).Based on their roles in the neuronal action potential, excessive sodium current or insufficient potassium current would be predicted to cause hyperexcitability and epilepsy.In agreement with expectation, many missense mutations in the voltage-gated sodium channel gene SCN8A result in excessive sodium current ("gain-of-function", or GOF, mutations) and SCN8A-DEE (Veeramah et al., 2012;Meisler et al., 2021;Johannesen et al., 2022).Experimental expression of an SCN8A GOF mutation in excitatory neurons is sufficient to cause seizures and premature death, while expression limited to inhibitory neurons does not (Bunton-Stasyshyn et al., 2019).
Reducing expression of Scn8a prolongs survival of epileptic mice with mutations in the potassium channel genes Kcna1 and Kcnq2 (Hill et al., 2022).Here, we asked whether modulating expression of a potassium channel can improve the phenotype of sodium channel mutants.Administration of the Kcnt1 ASO (Burbano et al., 2022) on postnatal day 2 doubled the lifespan of Scn8a mutant mice and extended survival of Scn1a haploinsufficient mice.Our results suggest a new therapeutic intervention for DEEs caused by mutations of SCN1A and SCN8A.

Mice
The Scn8a cond allele, abbreviated W, contains two tandem copies of exon 26, the final coding exon of Scn8a (Bunton-Stasyshyn et al., 2019).The upstream copy, designated 26a, is a floxed exon that encodes the wildtype channel.Deletion of exon 26a by Cre results in expression of exon 26b encoding the variant p.R1872W.This variant has been identified in multiple individuals with SCN8A epilepsy (Bunton-Stasyshyn et al., 2019;Johannesen et al., 2022).Scn8a cond/cond male mice were crossed with EIIa-Cre/+ female mice (JAX 003724) to generate Scn8a cond/+ ,EIIa-Cre double heterozygous mice expressing the R1872W variant (designated W/+ mice).Both the Scn8a cond allele and the EIIa-Cre transgene were maintained on a C57Bl/6 J genetic background.
Scn1a +/− mice with deletion of exon 1 were maintained on the protective 129S6/SvEvTac strain background and activated in (C57BL/6 J X 129S6/SvEvTac) F1 mice (Miller et al., 2014).Both male and female mice were used for all experiments.Experiments were approved by the Committee on the Use and Care of Animals at the University of Michigan.

ASOs
ASOs were synthesized by Ionis Pharmaceuticals as described (Swayze et al., 2007).Both the non-targeting control and Kcnt1 ASOs are 20-bp gap-mers with 5 2'-O-methoxyethyl modifications on the first and last 5 bases and phosphorothioate modifications on all 20 bases.The Kcnt1 ASO (5' GCT TCA TGC CAC TTT CCA GA 3′) is complementary to the 3' UTR of mouse Kcnt1 and was previously described (Burbano et al., 2022).The non-targeting control ASO (5' CCT ATA GGA CTA TTC AGG AA 3′) is well-tolerated and is not complementary to any transcript encoded by the mouse genome (Swayze et al., 2007).Animals treated with control ASO received a 30 μg dose.

Intracerebroventricular (ICV) injections
At postnatal day 2 (P2), mice were cryo-anesthetized for 3 min.ASO was diluted in PBS (2 μL injection volume) and manually injected into the left ventricle as described (Lenk et al., 2020).Animals were allowed to recover for 10 min at 37°C before being returned to the home cage.

Quinidine administration
Quinidine (Sigma Aldrich, St. Louis, MO) was diluted in phosphate-buffered saline (50 or 100 mg/kg) and administered by daily intraperitoneal injection beginning at P10, the youngest age at which daily intraperitoneal injections were feasible.

Quinidine does not extend survival in the SCN8A DEE mouse
Quinidine is a nonspecific cation channel blocker used to treat cardiac arrhythmia.In vitro, quinidine blocks KCNT1 channel activity, suggesting that it could be a precision therapy for patients with gain-of-function KCNT1 mutations (Mori et al., 1998;Milligan et al., 2014).In vivo, quinidine has mixed efficacy in KCNT1 epilepsy patients (Mikati et al., 2015;Numis et al., 2018;Fitzgerald et al., 2019;Cole et al., 2021).
To determine whether inhibition of KCNT1 channels by quinidine would be therapeutic in Scn8a mutant mice, we administered 50 or 100 mg/kg quinidine by daily intraperitoneal injection beginning at P10 (Figure 3).Untreated mice exhibited median survival of 15 days (n = 47).Treatment with 50 or 100 mg/kg quinidine did not extend the lifespan of the SCN8A-DEE mice (median survival = 14 days; n = 7 & 9, respectively; Figure 3).

ASO-mediated reduction of Kcnt1 extends the lifespan of a mouse model of SCN1A haploinsufficiency
We also tested the effect of the Kcnt1 ASO in Scn1a +/− mice, a model of Dravet Syndrome.Consistent with previous studies (Miller et al., 2014;Favero et al., 2018), approximately 1/3 of untreated Scn1a +/− mice died between 3 and 4 weeks of age, and during the remaining 6-month monitoring period, there were several sporadic deaths (Figure 4).We administered 45 μg Kcnt1 ASO to Scn1a +/− mice at P2. None of the treated mice died in the first 4 weeks, indicating that reduced Kcnt1 expression during this critical period is sufficient to prevent death (p = 0.0513, Mantel-Cox log-rank test, Figure 4).There were four deaths during the 6-month monitoring period, all after 9 weeks of age (Figure 4).Since quinidine was not effective in the Scn8a mutant mice, we did not treat the Scn1a mutant mice.

Discussion
Developmental and epileptic encephalopathies are frequently caused by pathological variants of ion channel genes.Here, we showed that reduction of Kcnt1 expression is protective in mouse models of Scn1a and Scn8a epilepsy.Our findings suggest that patients with mutations of SCN1A and SCN8A could benefit from treatment with a KCNT1 ASO or KCNT1-specific channel blocker.
A previous study demonstrated that the Kcnt1 ASO improved the survival, seizure, and behavioral phenotypes of Kcnt1 GOF mice (Burbano et al., 2022).Interestingly, the Kcnt1 ASO was therapeutic at lower doses in Kcnt1 mutant mice than in the Scn8a mutant studied here.In neonatal Kcnt1 mutant mice, 3.4 μg extended the median survival by more than 100 days (Burbano et al., 2022).In contrast, doses of 15-45 μg added only 20 days to survival of the Scn8a mutant mice.These observations suggest that the effect in the Scn8a mutant may be indirect.For example, gain-of-function of Kcnt1 reduces excitability of parvalbumin interneurons (Gertler et al., 2022).Reduced expression of Kcnt1 may enhance excitability of parvalbumin interneurons and thereby reduce seizure susceptibility.Consistent with this hypothesis, homozygous knockout of Kcnt1 reduces the lethality of electrically-induced seizures by more than half (Quraishi et al., 2020).Further investigation may identify other types of epilepsy that respond to reduction of KCNT1.
We previously demonstrated that reducing Scn8a expression is therapeutic in Scn1a +/− mice (Lenk et al., 2020) and in mice with epilepsy caused by loss of the potassium channel genes Kcna1 and Kcnq2 (Hill et al., 2022).P2 administration of the Scn8a ASO completely rescued the Scn1a +/− mice (Lenk et al., 2020).In contrast, 4/14 of the Scn1a +/− mice treated with the Kcnt1 ASO died between two and six months of age.The deaths after 2 months may result from turnover of the Kcnt1 ASO; alternatively, reduced Kcnt1 may be effective only in the interval between 3-4 weeks.The long-term effectiveness of the Scn8a ASO in Scn1a mutant mice is interesting, since the effect on Scn8a expression persists for only 6 weeks (Lenk et al., 2020).Viral overexpression of the Kcna1 channel is protective against seizures induced by tetanus neurotoxin or pentylenetetrazole (Snowball et al., 2019;Qiu et al., 2022).Taken together, these observations suggest that modulation of ion channel expression to compensate for epileptogenic mutations is a promising therapeutic strategy.Hill et al. 10.3389/fnins.2023.1282201Frontiers in Neuroscience frontiersin.org Among ion channel genes that could be targeted to treat channelopathies, KCNT1 is an attractive choice because reduced expression is well tolerated.Heterozygous loss-of-function mutations of KCNT1 are present in the general population and not associated with disease (Lek et al., 2016;Karczewski et al., 2020).Kcnt1 −/− mice are healthy and fertile, with minor abnormalities such as impaired reversal learning and slightly elevated pain sensitivity (Bausch et al., 2015;Lu et al., 2015;Martinez-Espinosa et al., 2015;Quraishi et al., 2020).In contrast, heterozygous loss of Scn8a is not present in the healthy population (probability of lossof-function intolerance, pLI = 1) (Lek et al., 2016;Karczewski et al., 2020) and homozygous loss is lethal in the mouse (Burgess et al., 1995).
Quinidine has been proposed as a therapy for patients with KCNT1 epilepsy because of the effectiveness of high doses for correction of GOF mutations in vitro (Milligan et al., 2014;Numis et al., 2018).The effects of quinidine are not specific to KCNT1 (Roden, 2014).Clinical application of quinidine in KCNT1 epilepsy has mixed success.Some individuals achieved seizure freedom (Mikati et al., 2015;Fitzgerald et al., 2019), but most patients report no benefit or worsening seizures (Mikati et al., 2015;Numis et al., 2018;Cole et al., 2020).Quinidine concentration sufficiently high to block KCNT1 may be difficult to achieve in vivo without deleterious effects on other ion channels (Liu et al., 2022).We found that quinidine was not protective in Scn8a mutant mice.More specific KCNT1 channel blockers (Cole et al., 2020;Griffin et al., 2021) may be more effective for treatment of KCNT1, SCN8A, and SCN1A epilepsy.

FIGURE 3
FIGURE 3Quinidine does not prolong survival of Scn8a mutant mice.Survival of Scn8a cond/+ ,EIIa-Cre (W/+) mice daily treated with 50 or 100 mg/kg quinidine compared to untreated mice.Grey shading indicates treatment period.