1. Neuroscience

NBI-921352, a first-in-class, NaV1.6 selective, sodium channel inhibitor that prevents seizures in Scn8a gain-of-function mice, and wild-type mice and rats

  1. JP Johnson  Is a corresponding author
  2. Thilo Focken
  3. Kuldip Khakh
  4. Parisa Karimi Tari
  5. Celine Dube
  6. Samuel J Goodchild
  7. Jean-Christophe Andrez
  8. Girish Bankar
  9. David Bogucki
  10. Kristen Burford
  11. Elaine Chang
  12. Sultan Chowdhury
  13. Richard Dean
  14. Gina de Boer
  15. Shannon Decker
  16. Christoph Dehnhardt
  17. Mandy Feng
  18. Wei Gong
  19. Michael Grimwood
  20. Abid Hasan
  21. Angela Hussainkhel
  22. Qi Jia
  23. Stephanie Lee
  24. Jenny Li
  25. Sophia Lin
  26. Andrea Lindgren
  27. Verner Lofstrand
  28. Janette Mezeyova
  29. Rostam Namdari
  30. Karen Nelkenbrecher
  31. Noah Gregory Shuart
  32. Luis Sojo
  33. Shaoyi Sun
  34. Matthew Taron
  35. Matthew Waldbrook
  36. Diana Weeratunge
  37. Steven Wesolowski
  38. Aaron Williams
  39. Michael Wilson
  40. Zhiwei Xie
  41. Rhena Yoo
  42. Clint Young
  43. Alla Zenova
  44. Wei Zhang
  45. Alison J Cutts
  46. Robin P Sherrington
  47. Simon N Pimstone
  48. Raymond Winquist
  49. Charles J Cohen
  50. James R Empfield
  1. In Vitro Biology, Xenon Pharmaceuticals Inc, Canada
  2. Chemistry, Xenon Pharmaceuticals Inc, Canada
  3. In Vivo Biology, Xenon Pharmaceuticals Inc, Canada
  4. Compound Properties, Xenon Pharmaceuticals Inc, Canada
  5. Translational Drug Development, Xenon Pharmaceuticals Inc, Canada
  6. Scientific Affairs, Xenon Pharmaceuticals, Inc, Canada
  7. Executive Team, Xenon Pharmaceuticals Inc, Canada
https://doi.org/10.7554/eLife.72468
Share this article
Cite this article
  1. JP Johnson
  2. Thilo Focken
  3. Kuldip Khakh
  4. Parisa Karimi Tari
  5. Celine Dube
  6. Samuel J Goodchild
  7. Jean-Christophe Andrez
  8. Girish Bankar
  9. David Bogucki
  10. Kristen Burford
  11. Elaine Chang
  12. Sultan Chowdhury
  13. Richard Dean
  14. Gina de Boer
  15. Shannon Decker
  16. Christoph Dehnhardt
  17. Mandy Feng
  18. Wei Gong
  19. Michael Grimwood
  20. Abid Hasan
  21. Angela Hussainkhel
  22. Qi Jia
  23. Stephanie Lee
  24. Jenny Li
  25. Sophia Lin
  26. Andrea Lindgren
  27. Verner Lofstrand
  28. Janette Mezeyova
  29. Rostam Namdari
  30. Karen Nelkenbrecher
  31. Noah Gregory Shuart
  32. Luis Sojo
  33. Shaoyi Sun
  34. Matthew Taron
  35. Matthew Waldbrook
  36. Diana Weeratunge
  37. Steven Wesolowski
  38. Aaron Williams
  39. Michael Wilson
  40. Zhiwei Xie
  41. Rhena Yoo
  42. Clint Young
  43. Alla Zenova
  44. Wei Zhang
  45. Alison J Cutts
  46. Robin P Sherrington
  47. Simon N Pimstone
  48. Raymond Winquist
  49. Charles J Cohen
  50. James R Empfield
(2022)
NBI-921352, a first-in-class, NaV1.6 selective, sodium channel inhibitor that prevents seizures in Scn8a gain-of-function mice, and wild-type mice and rats
eLife 11:e72468.
https://doi.org/10.7554/eLife.72468
  2. Download BibTeX
  3. Download .RIS

Abstract

NBI-921352 (formerly XEN901) is a novel sodium channel inhibitor designed to specifically target NaV1.6 channels. Such a molecule provides a precision-medicine approach to target SCN8A-related epilepsy syndromes (SCN8A-RES), where gain-of-function (GoF) mutations lead to excess NaV1.6 sodium current, or other indications where NaV1.6 mediated hyper-excitability contributes to disease (Gardella and Møller, 2019; Johannesen et al., 2019; Veeramah et al., 2012). NBI-921352 is a potent inhibitor of NaV1.6 (IC500.051 µM), with exquisite selectivity over other sodium channel isoforms (selectivity ratios of 756 X for NaV1.1, 134 X for NaV1.2, 276 X for NaV1.7, and >583 Xfor NaV1.3, NaV1.4, and NaV1.5). NBI-921352is a state-dependent inhibitor, preferentially inhibiting inactivatedchannels. The state dependence leads to potent stabilization of inactivation, inhibiting NaV1.6 currents, including resurgent and persistent NaV1.6 currents, while sparing the closed/rested channels. The isoform-selective profile of NBI-921352 led to a robust inhibition of action-potential firing in glutamatergic excitatory pyramidal neurons, while sparing fast-spiking inhibitory interneurons, where NaV1.1 predominates. Oral administration of NBI-921352 prevented electrically induced seizures in a Scn8a GoF mouse,as well as in wild-type mouse and ratseizure models. NBI-921352 was effective in preventing seizures at lower brain and plasma concentrations than commonly prescribed sodium channel inhibitor anti-seizure medicines (ASMs) carbamazepine, phenytoin, and lacosamide. NBI-921352 waswell tolerated at higher multiples of the effective plasma and brain concentrations than those ASMs. NBI-921352 is entering phase II proof-of-concept trials for the treatment of SCN8A-developmental epileptic encephalopathy (SCN8A-DEE) and adult focal-onset seizures.

Editor's evaluation

This exciting study reports on the characterization of a novel compound that preferentially targets Nav1.6 voltage-gated sodium channels and shows substantial activity against epilepsy associated SCN8A mutations and seizure activity in a variety of animal models. This compound and approach has significant promise to be translated into a therapeutic for individuals with treatment resistant epilepsy.

https://doi.org/10.7554/eLife.72468.sa0

Introduction

NaV1.6 voltage-gated sodium channels are widely expressed in the brain and are important contributors to neural excitability (Meisler, 2019; Royeck et al., 2008).Mutations in the SCN8A gene result in malfunction of NaV1.6 sodium channels and cause a spectrum of SCN8A-related syndromes in humans, and disruptions of mouse NaV1.6 likewise disrupt normal physiology (Burgess et al., 1995; Gardella and Møller, 2019; Johannesen et al., 2019; Meisler, 2019; Veeramah et al., 2012; Wagnon et al., 2015). Variants of NaV1.6 channels can result in either gain or loss of function. Loss-of-function (LoF) variants in humans are generally associated with autism spectrum disorders with cognitive and developmental delay without epilepsy (Inglis et al., 2020; Liu et al., 2019), but, in some cases, can lead to late-onset seizures.In mice, LoF variants of NaV1.6 lead to motor impairment but increase seizure resistance (Hawkins et al., 2011; Martin et al., 2007). Gain-of-function (GoF) variants in human SCN8A generally result in early-onset SCN8A-related epilepsy syndromes (SCN8A-RES). The most severe of these epilepsy syndromes is SCN8A developmental and epileptic encephalopathy (SCN8A-DEE) (Gardella and Møller, 2019; Hammer et al., 2016; Johannesen et al., 2019). Most SCN8A-RES patients carry de novo heterozygous missense variants that lead to a gain of function of the NaV1.6 channel, although inherited and bi-allelic variants have been reported (Gardella and Møller, 2019; Wengert et al., 2019).SCN8A-DEE patients present early in life with seizure onset usually occurring in the first year of life. After seizure onset, patients begin to miss developmental milestones and display additional symptoms, including cognitive and motor delay, hypotonia, and cortical blindness. SCN8A-DEE individuals are predisposed to early death, including sudden unexplained death in epilepsy (SUDEP). While SCN8A-RESpatients often have treatment-resistant seizures, many can achieve seizure reduction or seizure freedom upon treatment with anti-seizure medicines (ASMs) that non-selectively inhibit voltage-gated sodium channels, like phenytoin (Boerma et al., 2016; Braakman et al., 2017). SCN8A-RES patients may require doses that are higher than those prescribed for most epilepsy patients and, as a result, can be more prone to drug-related adverse events (Boerma et al., 2016; Gardella and Møller, 2019). Even with high doses and multiple ASMs, many patients continue to have uncontrolled seizures as well as extensive comorbidities. The aggressive pharmacotherapy required to protect SCN8Apatients from life-threatening seizures often comes with attendant side effects that would not be tolerated in less severely impacted populations.

Existing sodium channel inhibitor ASMs are nonselective, blocking all voltage-gated sodium channel isoforms at similar plasma or brain concentrations. This lack of selectivity likely limits the benefits of sodium channel inhibitors since LoF variants of NaV1.1 are known toimpair inhibitory interneuron function and causegeneralized epilepsy with seizures plus (GEFS+) and SCN1A-DEE (Dravet Syndrome) (Catterall et al., 2010; Claes et al., 2001; Escayg et al., 2000; Gennaro et al., 2003).Thus, inhibiting NaV1.1 may counter the benefit of inhibiting the sodium channels of excitatory neurons. Inhibiting NaV1.4 and NaV1.5 currents is also undesirable since those channels are critical for facilitating contraction of skeletal and cardiac muscles, respectively (Chen et al., 1998; Ptácek et al., 1991; Rojas et al., 1991).

We hypothesized that a selective inhibitor of NaV1.6 could provide a safer and more effective treatment for patients with SCN8A-RES and might also be more broadly efficacious in more common forms of epilepsy. An extensive medicinal-chemistry effort produced NBI-921352, the firstpotent and selective inhibitor of NaV1.6 channels (Neurocrine, 2019). We explored the profile of NBI-921352 in vitro, ex vivo and in threepreclinical in vivo rodent seizure models, including electrically induced seizure assays in genetically engineered mice bearing heterozygousScn8a GoFNaV1.6 channels (N1768D), as well as in wild-type mice and rats.

Results

In vitro NaV potency and selectivity

Human NaV channel isoforms hNaV1.1, hNaV1.2, hNaV1.3, hNaV1.4, hNaV1.5, hNaV1.6, and hNaV1.7 were heterologously expressed in HEK-293 cells, and the potency and isoformselectivity of NBI-921352 (Figure 1, Table 1) was determined by automated patch-clamp techniques.NBI-921352 potently inhibitedhNaV1.6 channel currents with an inhibitory concentration 50% (IC50)of 0.051 µM (95% CI: 0.030–0.073 µM; N = 3) calculated from three biological replicates. Inhibition of other human NaV1.X isoforms required higher concentrations of NBI-921352 with IC50’s of 39 µM (95% CI: 31–47 µM; N = 3)for hNaV1.1, 6.9 µM (95% CI: 1.6–12 µM; N = 3) for hNaV1.2, > 30 µM for hNaV1.3, > 30 µM for hNaV1.4, > 30 µM for hNaV1.5, and 14 µM (95% CI: 6.4–22 µM; N = 3) for hNaV1.7. These potencies provide selectivity ratios for hNaV1.6 versus the other hNaV isoforms (IC50 hNaV1.X / IC50 hNaV1.6) of 756 (NaV1.1), 134 (NaV1.2), 276 (NaV1.7) and >583 (NaV1.3, NaV1.4, NaV1.5). Inhibition of NaV1.8 and NaV1.9 was not assessed since both channels arebelieved to be limited to peripheral sensory neurons and are not expected tohave much impact on the efficacy and tolerability of sodium channel inhibitors in epilepsy. Inhibition of NaV1.8 and NaV1.9 might reduce pain perception, but this would not be considered a significant liability for an anti-seizure medicine.

Figure 1 with 2 supplements see all
Download asset Open asset
Potency and isoform selectivity of NBI-921352 for human and mouse NaV channels.

Concentration-response curves were generated by automated patch-clamp electrophysiology using the SophionQube. Concentration-response curves were generated for human (A) or mouse (B) NaV channel isoforms heterologously expressed in HEK293 cells. The analysis included only those cells that met pre-specified acceptance criteria for seal quality, current amplitude, and series resistance. Normalized data from all cell recordings at a concentration were grouped together and plotted with GraphPad Prism 8. Details regarding the number of cells analyzed for each NaV channel and concentration can be found in the source data sheet. Error bars indicating the standard error of the mean fraction were plotted for all points, but, in some cases, they were smaller than the data point symbols and, therefore, not visible. The chemical structure of NBI-921352 is shown (C).

Figure 1—source data 1

Quantification of potency and isoform selectivity of NBI-921352.

https://cdn.elifesciences.org/articles/72468/elife-72468-fig1-data1-v2.xlsx
Table 1
Potency and isoform selectivity of NBI-921352 for human and mouse NaV channels.

Note that IC50s for the neuronal sodium channels, NaV 1.1, NaV 1.2, and NaV1.6, have been more accurately defined than those for non-neuronal sodium channels. Explicit IC50’s for NaV1.3, NaV1.4, and NaV1.5 were not determined since the inhibition at the highest concentration tested (30 µM) was <50%. IC50’s are the mean of 3 separate biological replicates of the IC50 determinations for each channel. The error is shown as the 95% confidence interval of the mean IC50. None of the other tested isoforms displayed IC50’s within the 95% confidence interval of the NaV1.6 IC50 and the confidence intervals were well separated as well.

NaV1.6NaV1.1NaV1.2NaV1.3NaV1.4NaV1.5NaV1.7
Human IC50 (µM)0.051396.9> 30> 30> 3014
95% CI0.030–0.07331–471.6–12---6.4–22
Human Selectivity hNaV1.X / hNaV1.61756134> 583> 583> 583276
Mouse IC50 (µM)0.0584111
Mouse Selectivity mNaV1.X / mNaV1.61709191

Since we intended to evaluate in vivo effects of NBI-921352 in mouse seizure models, we also assessed the potency of NBI-921352 in the mouse NaV isoforms that are most highly expressed in the brain, NaV1.6, NaV1.1, and NaV1.2. The potency and selectivity in mouse NaV channels closely paralleled that seen in the human orthologues with IC50’s of 0.058 µM (95% CI: 0.046–0.070 µM; N = 3) for mNaV1.6, 41 µM (95% CI: 30–52 µM; N = 3) for mNaV1.1, and 11 µM (95% CI: 8.2–14 µM; N = 3) for mNaV1.2. Selectivity ratios (IC50 mNaV1.X / IC50 mNaV1.6) were 709 (NaV1.1), and 191 (NaV1.2). These data indicate that NBI-921352 potently inhibits both human and mouse NaV1.6 channels, and that it does so at concentrations ≥ 134 fold lower than for any of the other channel isoforms tested.

NBI-921352 inhibited patient-identified variants of NaV1.6 channels

Patients with SCN8A-RES carry missense variants in the NaV1.6 channel. A great number of variants have been identified, with a range of biophysical defects. Since most variants are de novo, many have been identified in only one or a few patients. For this reason, we determined the effectiveness of NBI-921352 to inhibit nine patient identified variants spread across the channel (Figure 2, Table 2; Gardella and Møller, 2019; Wagnon and Meisler, 2015). The nine variants studied have all been identified in SCN8A-RES patients and are in Domains II, III, IV, and the C-terminus. Inhibition of the mutant channel constructs was evaluated by automated patch-clamp electrophysiological techniques after transient transfection of the human NaV1.6 variant construct of interest into Expi293F cells. All the variants were sensitive to inhibition by NBI-921352. Observed IC50s for inhibition were 0.051 µM (WT mean from Figure 1), 0.031 µM (95% CI: 0.027–0.037 µM)(T767I), 0.021 µM (95% CI: 0.017–0.026 µM)(R850Q), 0.032 µM (95% CI: 0.029–0.036 µM)(N984K), 0.035 µM (95% CI: 0.029–0.043 µM)(I1327V), 0.039 µM (95% CI: 0.031–0.050 µM)(N1466K), 0.34 µM (95% CI: 0.26–0.44 µM)(R1617Q), 0.055 µM (95% CI: 0.046–0.064 µM)(N1768D), 0.068 µM (95% CI: 0.054–0.085 µM)(R1872W), and 0.035 µM (95% CI: 0.029–0.041 µM)(N1877S). We found that eight of the nine variants were inhibited with a potencysimilar to that of the wild-type channel, with most being slightly more potently inhibited. Only one variant, R1617Q, required markedly higher concentrations of NBI-921352 for inhibition, with an IC50for inhibition 6.6-fold higher than that of the wild-type NaV1.6 channel. The reduced potency for R1617Qis consistent with the variant residing in the predicted binding site of NBI-921352 in the domain IV voltage sensor, see discussion.

Figure 2
Download asset Open asset
Comparison of NBI-921352 potency on human wild-type NaV1.6 and patient-identified variants of NaV1.6.

All constructs were transiently transfected into Expi293F cells and evaluated by automated patch-clamp electrophysiology using the SophionQube. The voltage-clamp methods were identical to those used for evaluation of the wild-type channels. The error bars indicate the 95% confidence interval of the fitted IC50 generated in Prism. The horizontal dotted line is at the IC50 for wild type NaV1.6 (51 nM, see Figure 1). The gray shaded band indicates the 95% confidence range for the IC50 for wild-type NaV1.6. Only two variants have fitted IC50’s outside of the 95% confidence interval for the wild-type channel IC50. R850Q was slightly more potently inhibited and R1617Q was less potently inhibited than the wild-type channel. R1617Q is near the proposed binding site for NBI-921352.

Figure 2—source data 1

Quantification inhibition of patient variants.

https://cdn.elifesciences.org/articles/72468/elife-72468-fig2-data1-v2.xlsx
Table 2
Comparison of NBI-921352 potency on human wild-type NaV1.6 and patient-identified gain-of-function variants of NaV1.6.

IC50s corresponding to Figure 2 are shown in the table and were calculated as indicated for Table 1. The 95% confidence intervals are those determined by the fit of the IC50 in Prism.

WTT767IR850QN984KI1327VN1466KR1617QN1768DR1872WN1877S
hNaV1.6 IC50 (µM)0.0510.0310.0210.0320.0350.0390.3490.0540.0670.034
95% CI0.030–0.0730.027–0.0370.017–0.0260.029–0.0350.029–0.0420.031–0.0490.28 to0.400.047–0.0650.053–0.0850.029–0.040
Fold changeWT / Variant-0.60.40.60.70.86.81.11.30.7

NBI-921352 is a state-dependent inhibitor

Many small molecule inhibitors of NaV channels bind preferentially to open and or inactivated states (Bean et al., 1983; Courtney et al., 1978; Strichartz, 1976).

NaV inhibitors that are structurally similar to NBI-921352 are known to act by binding to the VSD4 in the ‘UP’, position and hence stabilize inactivated states (Ahuja et al., 2015; McCormack et al., 2013). Charge movement in VSD4 has been linked to the voltage dependence of fast inactivation (Ahern et al., 2016). It is likely that the anionic aryl sulfonamide headgroup of NBI-921352 interacts with the fourth arginine in S4 of VSD4 and prevents return of VSD4 to the rested position and recovery from inactivation as seen for related compounds (Ahuja et al., 2015). For these reasons, we expect that binding will be encouraged by VSD4 residing in the UP, inactivated, state.

We designed our voltage-clamp protocols to encourage high occupancy of inactivated states by holding at depolarized membrane potentials (–45 mV for NaV1.1, NaV1.2, NaV1.3, NaV1.4, and NaV1.6). The isoforms that inactivate at the most negative membrane potentials (NaV1.5 and NaV1.7) accumulate excess slow inactivation at –45 mV, reducing the signal size. For these two isoforms, the membrane potential was held at –60 mV to preserve robust assay performance while fully inactivating the channels. Our investigation into the state dependence of NBI-921352 did not attempt to rigorously differentiate binding to open channels versus fast inactivated or fast inactivated versus slow inactivated states.

To query the state dependence of NBI-921352, we measured the apparent potency with two different voltage protocols that favor either the closed (rested) state or inactivatedstates (Figure 3). Holding the membrane potential at –120 mV induces most channels to reside in the resting state. Brief depolarizations to measure NaV1.6 current enabled the determination of an IC50of 36 µM (95% CI: 29–47 µM) for rested-state channels. Holding the membrane potential at –45 mV encourages channels to transition into inactivated states. Brief hyperpolarizations allow rapid recovery from inactivation for channels that are not bound to drug followed by a short 20ms test pulse to –20 mV to measure currents from unbound channels (see Materials and methods for details). Measuring the ability of NBI-921352 to inhibit reopening of activated channels leads to an apparent IC50 of 0.051 µM (Figures 1 and 3). Thus, NBI-921352 strongly prefers inactivated channels, inhibiting them at concentrations more than 750-fold less than those needed to inhibit rested or ‘peak’ sodium currents.

Figure 3 with 1 supplement see all
Download asset Open asset
NBI-921352 is a state-dependent inhibitor of NaV1.6 and preferentially targets inactivated channels.

Concentration-response curves were generated for human WT and N1788D channel isoforms heterologously expressed in HEK293 cells. The analysis included only those cells that met pre-specified acceptance criteria for seal quality, current amplitude, and series resistance. Normalized data from all cell recordings at a concentration were grouped together and plotted with GraphPad Prism 8. Details regarding the number of cells analyzed for each NaV channel isoform and concentration can be found in the source data sheet. Error bars indicating the standard error of the mean fraction were plotted for all points. The blue dotted line indicates the concentration-response curve for wild-type NaV1.6 from Figure 1. When NaV1.6 channels were equilibrated with NBI-921352 at voltages that allow equilibration with inactivated states (–45 mV), the compound provided potent inhibition, as seen in Figure 1. NBI-921352 also exhibited potent block of NaV1.6 when measuring persistent or resurgent sodium current using distinct voltage protocols (see Materials and methods and text). Forcing channels to the rested, closed state by hyperpolarizing to –120 mV resulted in very weak inhibition. Current evoked from very negative potentials is sometimes referred to as ‘peak current’. The 95% confidence intervals for the IC50’s reported in the results are those provided for the error of the fit by Prism.

Figure 3—source data 1

Quantification of state dependence of NBI-921352.

https://cdn.elifesciences.org/articles/72468/elife-72468-fig3-data1-v2.xlsx

At more hyperpolarized potentials, potency for all isoforms will tend to be somewhat less. Nonetheless potency on NaV1.6 at a more physiologic potential (–62 mV) is shown in Figure 1—figure supplement 2 where the push towards inactivated states is not so strong. NBI-921352 remains potent (IC500.053 µM) in this assay, suggesting that potency and selectivity under physiologic conditions will remain high.

NBI-921352 inhibited persistent and resurgent currents from mutant NaV1.6 channels

The state-dependent nature of inhibition is also revealed in other types of voltage-clamp protocols, including those designed to measure persistent or resurgent sodium currents. Some drugs or candidate drugs, like PRAX-330 and Riluzole, have been touted based on their preference for persistent currents, but this appears to be a feature of many of the compounds in the NaV inhibitor class since they are generally poor inhibitors of closed/rested state channels and bind preferentially to activated channels (Colombo et al., 2013; Mason et al., 2019; Wengert and Patel, 2021). Apparent differences in persistent current selectivity are largely driven by differential kinetics and concentration dependences in combination with the electrophysiological protocols chosen for the measurements.

Elevated persistent and or resurgent currents are believed to underlie or contribute to the pathology of many sodium channel related pathologies (Mason et al., 2019; Pan and Cummins, 2020; Potet et al., 2020; Tidball et al., 2020; Zaman et al., 2019). In most conditions, normal NaV1.6 channels inactivate rapidly and nearly completely after opening. Persistent currents result from channels that are not stably inactivated – a common phenotype for epilepsy-inducing variants in NaV1.6, including N1768D (Tidball et al., 2020; Wagnon et al., 2015). We found that NBI-921352 inhibited N1768D NaV1.6 persistent currents (measured as the non-inactivating current 10ms after initiation of the depolarizing test pulse) with a similar potency as for activated wild-type NaV1.6 channels with an IC50 of 0.059 µM (95% CI: 0.044–0.082 µM) (Figure 3). A more traditional approach to measuring persistent currents is to step from a very hyperpolarized voltage (for example –120 mV) to a strong depolarization for 50ms or longer. This approach is not viable for NBI-921352 since equilibration of block after from such negative, non-physiological, voltages takes several seconds.

Resurgent currents occur after repolarizing following a strong depolarization as channels redistribute between closed, open, and inactivated states (Raman and Bean, 1997). These resurgent currents are enhanced in many SCN8A-RES variants (Pan and Cummins, 2020; Raman et al., 1997). NBI-921352 also effectively inhibited resurgent currents from N1768D channels with apparent IC50 of 0.037 µM (95% CI: 0.025–0.060 µM). While the resurgent currents cannot be measured with the same voltage protocols used to measure inactivated state inhibition, these data indicate that N1768D NaV1.6 resurgent currents are susceptible to inhibition by NBI-921352 and that inhibition occurs at similar concentrations.

NBI-921352 preferentially inhibited excitatory pyramidal neurons and spared inhibitory interneurons

A primary goal of creating NaV1.6 selective inhibitors was to spare NaV1.1, the voltage-gated sodium channel that is most prevalent in inhibitory interneurons. This should allow the selective targeting of excitatory neurons, where NaV1.6 and NaV1.2 are believed to be dominant, over inhibitory interneurons. To test this hypothesis, we performed current-clamp experiments in glutamatergic pyramidal neurons from mouse layer five neocortex and from fast spiking interneurons in the same region. Application of 0.250 µM NBI-921352 decreased the maximum firing rate in all three pyramidal neurons tested (Figure 4A). The difference in the cumulative area under the curves (AUC) for control versus NBI-921352 treated conditions was evaluated by a paired two tailed t-test (N = 3 or 4 cells for each point). Current injection levels > 160 pA led to a significant reduction of the cumulative area under the input output curve (p < 0.05 in a paired two-tailed t-tests) relative to the control condition. Specific p values are shown in the data transparency Excel file. See Figure 4—figure supplement 1 for more individual neuron comparisons.

Figure 4 with 1 supplement see all
Download asset Open asset
NBI-921352 inhibits firing in pyramidal neurons but spares fast-spiking interneurons.

Current input versus action-potential output evaluations in wild-type mouse brain slices treated with vehicle or 0.25 µM NBI-921352 (A & B), or 100 µM carbamazepine (C & D) was plotted. In cortical pyramidal neurons, both NBI-921352 (A) and carbamazepine (C) reduced action-potential spiking. In fast-spiking cortical interneurons, treatment with NBI-921352 resulted in a trend toward slightly increased firing frequency (B), while carbamazepine markedly reduced firing (D). The main upper panels compare average action-potential count of 3–4 neurons in each condition±the standard error of the mean. The lower panels show recordings for individual representative neurons for each condition. No inhibitors of synaptic inputs were used for these experiments. The difference in the cumulative area under the curves (AUC) for control versus NBI-921352 treated conditions was evaluated by a paired two tailed t-test (N = 3 or 4 cells for each point). *Indicates a p < 0.05 relative to the control condition. Specific p values are shown in the data transparency Excel file. See Figure 4—figure supplement 1 for more individual neuron comparisons.

Figure 4—source data 1

Quantification of cortical neuron current clamp input output.

https://cdn.elifesciences.org/articles/72468/elife-72468-fig4-data1-v2.xlsx

In contrast, NBI-921352 had no significant effect on the fast-firing inhibitory interneurons tested (Figure 4B). Carbamazepine significantly inhibited action-potential firing in both pyramidal neurons and in fast-spiking interneurons and the degree of inhibition was similar in both types of neurons. A before and after comparison for all tested neurons is shown in Figure 4—figure supplement 1.

NBI-921352 inhibited electrically induced seizures in Scn8aN1768D/+ mice

A selective inhibitor of NaV1.6 should lend itself to the treatment of disease states caused by pathologic gain of function of NaV1.6 channels. Hence, we examined the ability of NBI-921352 to inhibit electrically induced seizures in mice with a patient-identified GoF variant in the Scn8a gene encoding NaV1.6. N1768D is a variant of NaV1.6 identified in the first reported SCN8A-DEE patient (Veeramah et al., 2012). N1768D NaV1.6 channels have impaired voltage-dependent inactivation gating that results in persistent sodium currents and enhanced resurgent currents. Because NaV1.6 channels are highly expressed in the neurons of the brain, increased sodium flux in excitatory neurons leads to seizures. Genetically modified mice bearing the same variant (Scn8aN1768D/+) were created and found to be seizure prone, producing a mouse model with a similar phenotype as that observed in SCN8A-DEE patients (Wagnon et al., 2015).Some Scn8aN1768D/+ mice develop spontaneous seizures at age p60 to p100, but seizure onset and frequency is quite variable, making spontaneous seizure studies challenging. In addition, mice rapidly clear NBI-921352, making it extremely difficult to maintain drug plasma and brain levels in an efficacious range for chronic or subchronic dosing experiments. NBI-921352 is more stable in humans with a half-life of elimination of approximately 8.5 hr (Beatch et al., 2020).

As an alternative means of assessing NBI-921352’s ability to engage NaV1.6 channels in vivo, we designed a modified version of the 6 Hz psychomotor seizure assay in Scn8aN1768D/+ mice (Barton et al., 2001; Focken et al., 2019). A mild current stimulation (12 mA) evoked robust generalized tonic-clonic seizures (GTC) with hindlimb extensionin Scn8aN1768D/+ mice, but not in wild-type littermates.

Oral administration of NBI-921352 2 hr prior to electrical stimulation prevented induction of GTC with hindlimb extension in Scn8aN1768D/+ mice in a dose-dependent manner with a 50% effective dose (ED50) of 15 mg/kg (95% CI 9.6–23 mg/kg, see Figure 5A).

Figure 5 with 2 supplements see all
Download asset Open asset
NBI-921352 inhibited electrically induced seizures in rodents.

Dose of NBI-921352 is plotted versus efficacy in Scn8aN1768D+/- mice in the modified 6 Hz psychomotor seizure assay in A. Plasma concentration of NBI-921352 is plotted versus efficacy in Scn8aN1768D+/- mice in the modified 6 Hz psychomotor seizure assay in B. Brain concentration of NBI-921352 is plotted versus efficacy in Scn8aN1768D+/- mice in the modified 6 Hz psychomotor seizure assay in C.Green open triangles represent data from animals that received a single dose 2 hr before testing in the seizure assay. Red open triangles represent data from animals that received two daily doses (once every 12 hr) for 6 days. On day 7, these animals were given a final dose (the 13th dosing) 2 hr before testing in the seizure assay. Dose of NBI-921352 is plotted versus efficacy in wild-type mice in the DC-MES assay in D. Plasma concentration of NBI-921352 is plotted versus efficacy inwild-type mice in the DC-MES assay in E. Brain concentration of NBI-921352 is plotted versus efficacy inwild-type mice in the DC-MES assay in F. Blue open circles represent data from animals that received a single dose 2 hr before testing in the seizure assay. Red open circles represent data from animals that received two daily doses (once every 12 hr) for 6 days. On day seven these animals were given a final dose (the 13th dosing) 2 hr before testing in the seizure assay. Dose of NBI-921352 is plotted versus efficacy inwild-type rats in the DC-MES assay in G. Plasma concentration of NBI-921352 is plotted versus efficacy inwild-type rats in the DC-MES assay in H. Brain concentration of NBI-921352 is plotted versus efficacy inwild-type rats in the DC-MES assay in I. Black open diamonds represent data from animals that received a single dose 2 hr before testing in the seizure assay. Red open diamonds represent data from animals that received two daily doses (once every 12 hr) for 6 days. On day 7, these animals were given a final dose (the 13th dose) 2 hr before testing in the seizure assay. Each point represents the fraction of animals exhibiting a GTC with hindlimb extension after stimulus from a dosing group of six to eight animals. Horizontal error bars show the standard error of the mean plasma (B, E, H) or brain (C, F, I) concentrations measured from the animals in that dosing group immediately after assay. Where error bars are not visible, they are smaller than the symbols. No error bars are shown for the dose levels (A, D, G), since those were dictated by the experimenter. Groups receiving the same dose level were combined for statistical analysis. Between-group differences were compared to vehicle response and were analyzed using a Kruskal-Wallis test followed by Dunn’s multiple comparisons test. Statistical significance was reached at values of p < 0.05 and is indicated by the stars at the top of A, D, and G. For specific p values and more details see Figure 5—figure supplement 1 Symbols in red represent animals tested after twice daily dosing for 6.5 days (see description above). Statistical analysis was completed as above for comparison between NBI-921352 treated and vehicle treated groups, and significance is indicated by a red star above the x-axis. In some cases, dose levels that were not statistically significant after a single dose became significant after repeat dosing when compared to vehicle controls. However, in no case were the groups treated with repeatedly with NBI-921352 statistically significant relative to their single dose comparison groups from the same experiment. For more data and p-values see Figure 5—figure supplement 2.

Figure 5—source data 1

Quantification of impact of NBI-921352 on rodent seizures.

https://cdn.elifesciences.org/articles/72468/elife-72468-fig5-data1-v2.xlsx

After seizure assessment, all animals were euthanized, and the concentration of NBI-921352 was determined in the plasma and brain tissue from each mouse. The average concentrations for each dose group were used to generate plasma concentration and brain concentration versus efficacy relationships (Figure 5B and C, respectively). The plasma 50% effective concentration (EC50) was 0.037 µM (95% CI 0.018–0.090 µM, see Figure 5B). The brain EC50 was 0.064 µM (95% CI 0.045–0.091 µM, see Figure 5C).

NBI-921352 inhibited electrically induced seizures in wild-type mice

NaV1.6 is an important mediator of neuronal excitability even in animals without GoFmutations. For this reason, we wondered whether NBI-921352 might have broader application in epilepsy beyond SCN8A-RES and in other syndromes of neural hyperexcitability. To gain insight into this possibility, we assessed NBI-921352 in a MES assay induced by direct-current electrical stimulus (DC-MES, see Materials and methods) in wild-type mice. Figure 5D and E & F show that NBI-921352 prevented GTC with hindlimb extension induction in the DC-MES assay in a dose- and concentration-dependent manner. The ED50for NBI-921352 was 23 mg/kg (95% CI 16–34 mg/kg, see Figure 5D). The efficacy of NBI-921352 was also concentration dependent with a plasma EC50 of 0.52 µM (95% CI 0.25–1.2 µM, see Figure 5E) and a brain EC50 of 0.20 µM (95% CI 0.12–0.38 µM, see Figure 5F).

NBI-921352 inhibited electrically induced seizures in wild-type rats

To further explore the preclinical efficacy of NBI-921352, we assessed NBI-921352 in a MES assay induced by direct-current electrical stimulus in wild-type Sprague Dawley rats (see Materials and methods). Figure 5G & show that NBI-921352 prevented GTC with hindlimb-extension induction in the rat DC-MES assay in a dose- and concentration-dependent manner. The ED50 for NBI-921352 was 3.7 mg/kg (95% CI 2.3–7.6 mg/kg, see Figure 5G). The efficacy of NBI-921352 was also concentration dependent with a plasma EC50 of 0.15 µM (95% CI 0.09–0.31 µM, see Figure 5H) and a brain EC50 of 0.037 µM (95% CI 0.028–0.054 µM, see Figure 5I).

Repeated-dosing efficacy in mice and rats

We found that repeated dosing tended toincrease efficacy at lower doses and exposures of NBI-921352 than after a single dose (Figure 5, red symbols). Animals were dosed every 12 hr, morning, and evening for 13 doses. Two hours after the 13th dose, on the seventh day, efficacy was tested as in acute-dosing studies. A trend toward improved efficacy was noted in all three assays (see red symbols in Figure 5), but the improvement was not statistically significant when comparing single-dose groups to repeated-dose groups at the same dose level in the same experiment. NBI-921352 did not appreciably accumulate in the plasma or tissue and therefore any trends in improved efficacy were not explained by higher drug concentrations.

NBI-921352 is effective at lower brain concentrations than three NaV inhibitor ASMs

The efficacy of NaV inhibitors in epilepsy is presumed to bedue to the drugs’ action in the CNS, and CNS side effects are also commonly reported for this class of drugs (Walia et al., 2004). Hence brain concentrations are likely to drive both efficacy and many adverse events. We found that NBI-921352 was effective in the three preclinical seizure models evaluatedat markedly lower brain concentrations than carbamazepine, phenytoin, and lacosamide (Figure 6).The brain EC50s for carbamazepine were 9.4 µM, 44 µM, and 36 µM for the Scn8aN1768D/+ 6 Hz model, WT mouse DC-MES, and WT rat DC-MES models, respectively. The brain EC50s for phenytoin were 18 µM, 13 µM, and 2.6 µM for the Scn8aN1768D/+ 6 Hz model, WT mouse DC-MES, and WT rat DC-MES models, respectively. The brain EC50s for lacosamide were 3.3 µM, 7.1 µM, and 4.3 µM for the Scn8aN1768D/+ 6 Hz model, WT mouse DC-MES, and WT rat DC-MES models, respectively. Thelower brain concentrations required for efficacy with NBI-921352 are consistent with the potent inhibition of NaV1.6 produced by NBI-921352 (Figure 1).

Figure 6 with 3 supplements see all
Download asset Open asset
NBI-921352 is more potent than three commonly prescribed NaV inhibitor ASMs.

Brain concentration versus fraction of animals exhibiting is plotted for NBI-921352 versus that for phenytoin, carbamazepine, and lacosamide in the Scn8aN1768D+/- modified 6 Hz model (A), the wild-type mouse DC-MES model (B), and the wild-type rat DC-MES model (C). Each point represents the fraction of animals exhibiting a GTC with hindlimb extension after stimulus from a dosing group of six to eight animals. Horizontal error bars show the standard error of the mean brain concentrations measured from the animals in that dosing group immediately after assay. Where error bars are not visible, they are smaller than the symbols. Statistical analysis of significance of the dose groups for the concentrations shown were performed as in Figure 5 and can be found in Figure 5—figure supplement 1 (NBI-921352), Figure 6—figure supplement 1 (Carbamazepine), Figure 6—figure supplement 2 (Phenytoin), and Figure 6—figure supplement 3 (Lacosamide).

Figure 6—source data 1

Quantification of effective brain concentrations of NBI-921352 versus common AEDs.

https://cdn.elifesciences.org/articles/72468/elife-72468-fig6-data1-v2.xlsx

NBI-921352 provided improved separation between efficacy in rats and behavioral signs

The intent of creating a highly selective NaV1.6 antagonist was to reproduce or improve on the efficacy of classic, nonselective, sodium channel inhibitor drugs while reducing or preventing the adverse events caused by polypharmacy with other sodium channel and non-sodium channel targets. If sparing NaV1.1 and other off-target interactions does, in fact, reduce adverse events, then even higher receptor occupancy of NaV1.6 might be achievable, therebyfurther improving efficacy.

To evaluate our hypothesis, we compared the window between the plasma concentrations required for efficacy (plasma EC50) relative to the minimal plasma concentration at which treated rats showed behavioral signs of adverse effects as reported by the blinded experimenter (Figure 7). We made this comparison both for NBI-921352 and for several widely used NaV inhibitor ASMs: carbamazepine, phenytoin, and lacosamide.

Figure 7
Download asset Open asset
Rat efficacy compared to acute tolerability for NBI-921352 relative to NaV inhibitor ASMs.

Plasma concentration versus efficacy data is shown for the rat DC-MES assay for NBI-921352 (A), carbamazepine (B), phenytoin (C), and lacosamide (D). The vertical dotted lines indicate the lowest plasma concentration at which a rat was observed to exhibit atypical behavioral signs indicative of an adverse reaction to drug in the assay format. Animals exhibiting such signs were excluded from efficacy evaluation. The shaded bars in B, C, and D indicate the approximate human plasma concentrations observed in clinical practice. Panel E shows the ratio of the (rat plasma EC50/ the plasma concentration where behavioral signs were noted for each compound). Statistical analysis of significance of the dose groups for the concentrations shown can be found in Figure 5—figure supplement 1 (NBI-921352), Figure 6—figure supplement 1 (Carbamazepine), Figure 6—figure supplement 2 (Phenytoin), and Figure 6—figure supplement 3 (Lacosamide).

Figure 7—source data 1

Quantification of effective and tolerated plasma concentrations of NBI-921352 versus common AEDs.

https://cdn.elifesciences.org/articles/72468/elife-72468-fig7-data1-v2.xlsx

NBI-921352 was well tolerated in these studies up to a plasma concentration of 71 µM. Dividing this concentration by the plasma EC50 of 0.15 µMin the rat DC-MES study results in a behavioral signs concentration (BSC) / plasma EC50 ratio of 473-fold. The same calculation was repeated for the established ASMs. The minimal plasma concentrations provoking behavioral signs for carbamazepine, phenytoin, and lacosamide were 110 µM, 11 µM, and 123 µM, respectively. Their plasma EC50s were 30 µM, 4.5 µM, and 24.4 µM, respectively. Figure 7E shows BSC / Plasma EC50 ratios for carbamazepine (3.7-fold), phenytoin (2.4-fold), and lacosamide (5.0-fold). This data indicatesthat increasing NaV1.6 selectivity can improve the tolerability of NaV inhibitors in rodent-seizure models.

Discussion

Sodium channel inhibitors have long been, and remain, a mainstay of pharmacotherapy for epilepsy, as well as for pain and other neurologic, cardiac, and skeletal muscle disorders. The diverse range of indications and systems affected by these drugs is a testament to their critical biological role in cellular excitability. A fundamental challenge for these currently marketed sodium channel drugs is that none of them are selective amongst the nine sodium channel isoforms. As a result, drugs targeting the sodium channels of the brain for epilepsy can inhibit both excitatory and inhibitory neurons, limiting their ability to restore balanced neuronal firing.

Reducing NaV1.1 current is proconvulsant due to the predominance of NaV1.1 in inhibitory interneurons (Catterall et al., 2010; Claes et al., 2001; Escayg et al., 2000; Mistry et al., 2014; Yu et al., 2006), while the opposite is true for NaV1.2 and NaV1.6 currents (Ben-Shalom et al., 2017; Hawkins et al., 2011; Martin et al., 2007). NaV1.2 and NaV1.6 are more highly expressed in excitatory neurons (Catterall et al., 2010; Du et al., 2020; Encinas et al., 2020).

Additionally, these nonselective agents can block the channels associated withskeletal muscle (NaV1.4), cardiac tissue (NaV1.5), and peripheral neurons (NaV1.7, NaV1.8, NaV1.9). Inhibiting these off-target channels can compromise muscular, cardiovascular,and sensory function. These risks are highlighted by the FDA’s recent drug-safety communication for the nonselective NaV inhibitor ASM lamotrigine. Lamotrigine has been linked to cardiac liabilities as a consequence of NaV1.5 inhibition (FDA, 2021). Likewise, NaV inhibitors intended as local anesthetics for trigeminal neuralgia or other pain syndromes and class I cardiac antiarrhythmic drugs are often dose limited by CNS adverse events like dizziness, sedation, and cognitive or motor impairment caused by inhibition of CNS NaV channels (Caron and Libersa, 1997).

An obvious solution to this isoform selectivity challenge is to pursue a precision-medicine approach and create selective pharmacologic agents that preferentially target the sodium channels specific to the desired target tissue or cell type. This selective approach has been pursued for multiple channel isoforms - particularly for the peripheral NaVs associated with pain (NaV1.3, NaV1.7, and NaV1.8). This approach has proven challenging because the nine isoforms of sodium channels, NaV1.1-NaV1.9, share a high degree of primary and tertiary structural similarity. Achieving selectivity with compounds that have tractable pharmaceutical properties has been particularly difficult.

Previous attempts to optimize NaV inhibitors for epilepsy have focused on either drug properties or channel-state dependence. To our knowledge, this is the first description of acentrally penetrant, isoform-selective NaV inhibitor for use in CNS indications, including epilepsy. NBI-921352 represents the first selective inhibitor of NaV1.6 that is suitable for systemic oral administration.

The NaV1.6 selective profile of NBI-921352 was designed to inhibit activity in excitatory neurons while sparing firing in the inhibitory interneurons where NaV1.1 is preferentially expressed. We found that NBI-921352 did reduce firing in cortical excitatory pyramidal cells. In contrast, fast-firing inhibitory interneuron firing was not impaired and actually showed a trend toward increasing (although not statistically significant). The reason that interneuron action-potential firing might increase is unclear but could be a consequence of network effects that arise from interactions with other neurons synapsed onto the target neurons in the experiment. These data confirm that selective NaV1.6 inhibitors can display differential inhibition of neurons in a manner that nonselective inhibitors, like carbamazepine, generally do not.

The distribution of NaV1.6 to excitatory pyramidal neurons is not absolute, and NaV1.6 expression has also been found in inhibitory interneurons. This is most well established for Purkinje neurons (Raman et al., 1997; Woodruff-Pak et al., 2006). Deletion of NaV1.6 in Purkinje neurons in mice causes deficits in both cognition and motor control (Levin et al., 2006; Woodruff-Pak et al., 2006). NaV1.6 is also highly expressed at the nodes of Ranvier of both central and peripheral neurons (Boiko et al., 2001; Caldwell et al., 2000; Herzog et al., 2003). We have found that selective inhibitors of NaV1.6 can cause ataxia and motor symptoms when plasma and brain concentrations are very high (Figure 7). It is unclear whether these effects are due to concentrations exceeding those that retain high selectivity, or whether high occupancy of NaV1.6 is directly responsible. In either case, our data would suggest that selective inhibition of NaV1.6 provides a robust window between the level of inhibition required to reduce seizure induction and the levels that impair normal behavior.

SCN8A-RES patients most often carry de novo genetic variants. While some variants are known to be recurrent, many variants are represented by a single patient (Meisler, 2019). We therefore wanted to assure that NBI-921352 inhibition was not limited to wild-type NaV1.6 channels. We tested nine distinct, patient-identified variants and found that 8 of them were inhibited by NBI-921352 at similar concentrations as wild-type channels (Figure 2). One variant, R1617Q, was found to be 6.8-fold less sensitive to inhibition than the wild-type channel. R1617Q has been identified in multiple SCN8A-DEE patients and is in the domain IV voltage-sensor domain (VSD4). NBI-921352 is an aryl sulfonamide with structural similarity to the NaV1.7 targeted aryl sulfonamides where the binding site has been identified as the NaV1.7 VSD4 (Ahuja et al., 2015; McCormack et al., 2013). It is likely that the R1617Q directly or allosterically impairs the tight association of NBI-921252 with NaV1.6 due to its proximity to the binding site. Despite the reduction in potency, NBI-921352 remains markedly more potent than existing NaV inhibitor drugs on the R1617Q variant NaV1.6 channel. This would suggest that while NBI-921352 may be an effective treatment for SCN8A-DEE patients carrying R1617Q variants, higher plasma levels of the compound could be required for efficacy in those patients.

Many SCN8A-RES-associated variants produce their GoF effects by disrupting or destabilizing the inactivation-gating machinery of NaV1.6 channels. This can lead to pathological persistent or resurgent currents that contribute to neuronal hyperexcitability (Pan and Cummins, 2020).

Most known small molecule inhibitors of NaV channels, except some marine toxins like tetrodotoxin, bind preferentially to inactivated gating states of the channels and stabilize the channels in inactivated, non-conductive conformations. This state dependence is manifested as a protocol dependence of the apparent drug potency. State dependence inhibition has been described in many ways. Use-dependent, frequency-dependent, resurgent current -selective, and persistent current-selective inhibition are all consequences of a preference for binding to inactivated channels. Stabilizing inactivated states of the channel reduces persistent and resurgent currents, and this feature has been suggested to contribute to the efficacy of many NaV targeted ASMs including phenytoin, carbamazepine, oxcarbazepine, lacosamide, cannabidiol, and lamotrigine (Wengert and Patel, 2021).NBI-921352 is also highly state dependent, with a > 750 fold preference for inactivated channels vs. rested, closed-state channels (sometimes referred to as peak current). Forcing all NaV1.6 channels into the closed state by applying voltages more hyperpolarized than physiological (–120 mV) results in very weak inhibition of the channels (Figure 3). Biasing the channels toward inactivated states by holding the membrane potential more positive in a protocol designed to monitor currents recovered from the inactivated state, resurgent current, or persistent current protocol all yielded potent inhibition. In physiologic conditions, channels are distributed among closed, open, and inactivated states, thus allowing equilibration of potent inhibition of the channel by NBI-921352.

Based on structural similarities and biophysical behavior, we believe that NBI-921352 binds to the domain IV voltage sensor domain (VSD4), much like GX-936 binds to the chimeric NaV1.7 constructs in Ahuja et al. In this structure the VSD4 is captured in the up (activated) state. The down (rested) state of the voltage sensor is not available for high-affinity interaction with the compound. We expect that the high affinity binding events for NBI-921352 occur only when the VSD4 is in the up state, promoting inactivation. We believe this is the likely mechanism by which NBI-921352 traps channels in the inactivated state and reduces NaV1.6 currents.

Increasing the selectivity of a NaV inhibitor provides the expectation of an improved safety profile by reducing adverse events caused by off-target activity. An inherent risk of this approach is the potential loss of efficacy that could come from reduced polypharmacy. We have developed a potent, highly selective NaV1.6 inhibitor in NBI-921352. Our studies with NBI-921352 indicate that a NaV1.6-specific compound can retain a robust ability to prevent seizures in rodent models at modest plasma and brain concentrations, consistent with the important role of NaV1.6 in seizure pathways. Our data also suggests that this selectivity profile does improve the tolerability of NBI-921352 relative to commonly employed nonselective sodium channel ASMs in rodents. Whether these results will translate to humans is not yet established, but Phase I clinical trials have shown that NBI-921352was well tolerated at plasma concentrations higher than were required for efficacy in the preclinical rodent studies described here. NBI-921352 is currently being developed for both SCN8A-DEE epilepsy and adult focal-onset seizures by Neurocrine, 2019. Phase II clinical trials will soon evaluate the efficacy of NBI-921352 in patients (Neurocrine, 2021). These clinical trials will provide the first evidence for whether the robust efficacy and tolerability demonstrated in rodents translates to human epilepsy patients.

Materials and methods

Key resources table
Reagent type (species) or resourceDesignationSource or referenceIdentifiersAdditional information
Cell line (H. sapiens)Expi293FThermo Fischercat# A14527SCN8A mutant transient transfections
Cell line (H. sapiens)FreeStyle 293 FThermo Fischercat# R710-07Stable transfections
Strain, strain background (M. musculus)C57BL/6 J male Scn8aN1768D/+ (stock#400690)x C3HeB/FeJ female (strain#000658).Both Male and Female heterozygous Scn8aN1768D/+ micewere testedLicensed from Miriam Meisler, Univ. of Michigan.Wagnon et al., 2015 Colony maintained at The Jackson Laboratory
Strain, strain background (M. musculus)CF-1, maleCharles RiverCode: 023 
Strain, strain background(R. norvegicus)Sprague Dawley, maleEnvigoCode: 002 
Chemical compound, drugNBI-921352US Patent #10246453 B2Compound ID #101Synthesized at Xenon Pharmaceuticals
Chemical compound, drugCarbamazepineSigma-AldrichC4024 
Chemical compound, drugPhenytoinSigma-Aldrich D4505 
Chemical compound, drugLacosamideToronto Research Chemicals L098500 
Gene (H. sapiens)SCN1AGenBankNM_006920 
Gene (H. sapiens)SCN2AGenBankNM_021007 
Gene (H. sapiens)SCN3AGenBankNM_0069220 
Gene (H. sapiens)SCN4AGenBankNM_000334 
Gene (H. sapiens)SCN5AGenBankNM_198056 
Gene (H. sapiens)SCN8AGenBankNM_014191 
Gene (H. sapiens)SCN9AGenBankNM_002977 
Gene (M. musculus)SCN1AGenBankNM_018733.2 
Gene (M. musculus)SCN2AGenBankNP_001092768.1 
Gene (M. musculus)SCN8AGenBankNM_001077499 
Gene (H. sapiens)SCN1BGenBankNM_199037 
Gene (H. sapiens)FGF13GenBankNM_033642 
Gene (H. sapiens)CNTN1GenBankNM_001843 
Recombinant DNA reagentpcDNA4/TO (vector)Thermo Fischercat #V102020Vector for SCNxA genes
Recombinant DNA reagentpcDNA6/TR(regulatory vector for tetracycline repressor protein)Thermo Fischercat#V102520Vector to generate inducible FreeStyle 293 F and Expi293F
Recombinant DNA reagentpcDNA3.1 (+)(vector)Thermo Fischercat#V79020Vector for SCN1B gene
Recombinant DNA reagentpcDNA3.1/Hygro(+)Thermo Fischercat# V87020Vector for FGF13 and CNTN1 genes

Electrophysiological determination of potency and selectivity

Cell lines

Request a detailed protocol

Electrophysiology experiments were performed with HEK293 cells either stably transfected or transiently transfected. All of the cell lines tested negative for mycoplasma. The cell lines were authenticated by STR profiling at EuroFins Genomics.Stable cell lines were transfected with an expression vector containing the full-length cDNA coding for specific human and mouse sodium channel α-subunit, grown in culture media containing 10% fetal bovine serum, and 0.5 mg/mL Geneticin (G418) at 37 °C with 5% CO2. The NaV1.x stable cell lines and accessory constructs used correspond to the following GenBank accession numbers: Human NaV1.1 (NM_006920); mouse NaV1.1 (NM_018733.2); human NaV1.2 (NM_021007); mouse NaV1.2 (NP_001092768.1); human NaV1.5 (NM_198056); human NaV1.6 (NM_014191); mouse NaV1.6 (NM_001077499); human NaV1.7 (NM_002977); human NaV1.4 (NM_000334); human NaV1.3 (NM_0069220). The human NaV β1 subunit (NM_199037) was co-expressed in all cell lines. Human and mouse NaV1.6 channels were also coexpressed with human FHF2B (NM_033642) to increase functional expression. Human NaV1.2 channels were also coexpressed with Contactin 1 (NM_001843) to increase functional expression.

For studies of mutant channels, cDNA plasmids in pcDNA4/TO Mammalian Expression Vector were transiently transfected into Expi293F cells stably expressing human FHF2b and human SCN1B subunit (polyclonal) background using ExpiFectamine 293 Transfection Kits (Gibco,Thermo Fisher Scientific CAT #: A14524). Induction was achieved using Tetracycline (Sigma Aldrich). Transfected cells were used in automated patch-clamp experiments 24 hours postinduction.

NaV channel automated Qube 384 planar patch-clamp assays

Request a detailed protocol

NBI-921352 requires several seconds to equilibrate with activated channels, and this property of the compound must be taken into consideration in the design of state-dependent assay voltage-clamp protocols.

Data was collected using the Qube 384 (Sophion) automated voltage-clamp platform using single hole plates. To measure inactivated state inhibition, the membrane potential was maintained at a voltage where inactivation is complete. For each NaV channel subtype, the Vhused to quantify compound inhibition were as follows: NaV1.6 (-45 mV), NaV1.1 (-45 mV), NaV1.2 (-45 mV), NaV1.3 (-45 mV), NaV1.5 (-60 mV), NaV1.7 (-60 mV), NaV1.4 (-45 mV). The mutant channel hNaV1.6N1768D was found to have accelerated run-down compared with wild-type hNaV1.6, so the holding potential was adjusted to –60 mV to provide sufficient signal window. The voltage was briefly repolarized to a negative voltage (–150 mV) for 20 ms for NaV1.5, NaV1.7, NaV1.3, NaV1.4or for 60 ms for NaV1.1, NaV1.2, and NaV1.6 to allow recovery from fast inactivation, followed by a test pulse to –20 or 0 mV for 10 ms to quantify the compound inhibition. The repolarization step allows compound-free channels to recover from fast inactivation, but compound-bound channels remain inhibited during the subsequent test step. For rested state ‘Peak’ current Vhwas set to –120 mV. Appropriate filters for minimum seal resistance were applied (typically >500 MΩ membrane resistance), and series resistance was compensated at 100%. The pulse protocols were run at 1 Hz for hNaV1.7, hNaV1.5, hNaV1.3, and hNaV1.4 or 0.04 Hz for NaV1.6, NaV1.1 and NaV1.2.

To construct concentration response curves, baseline currents were established after 20 min in vehicle (0.5% DMSO). Full inhibition response amplitudes were determined by adding tetrodotoxin (TTX, 300 nM) or tetracaine for NaV1.5 (10 µM) to each well at the end of the experiment. Compounds were then exposed at a single concentration for 20 min. One-sixth of every experimental plate was dedicated to vehicle-only wells that enabled correction for nonspecific drift (i.e. rundown) of the signal in each experiment. For all channel subtypes, inhibition by the compound reached steady state within 20 min of incubation. The current inhibition values (I(CPD)) were normalized to both the vehicle (Icontrol) and the full response defined by supramaximal TTX (ITTX) or tetracaine (for NaV1.5) addition responses according to Equation 1.

Inorm(CPD)=(ICPD−Icontrol)/(ITTX−Icontrol).

This normalized inhibition was then further normalized to the span of the assay to account for the run-down seen in cells exposed to vehicle alone for 20 min as follows:

Equation 2

Inorm,span=(Inorm(CPD)−Inorm(VEH))/(1−Inorm(VEH)),where:

  • Inorm, span = the current response normalized to within the span of the assay.

  • Inorm(CPD) = the normalized response in the presence of compound.

  • Inorm(VEH) = the normalized response in the absence of compound.

This normalization ensures that the data ranges were between 0 and 1, and there is no rundown in the plots. The normalized data from all cell recordings at a concentration were grouped together and plotted with GraphPad Prism 8, and IC50 values were calculated for grouped data using the following version of the Hill equation:

Equation 3

Y=RD+(1−RD)∗[CPD]/(IC50+[CPD]),where:

  • Y = the fraction of sodium current blocked in the presence of the compound.

  • [CPD] = the concentration of compound.

  • IC50 = the IC50 concentration.

RD = the ‘rundown’ of sodium current in vehicle alone, which is equal to 0 in this case, as the inhibition has already been normalized to the span.

  • The Hill slope was fixed to 1.

  • The 95% CI for the IC50 from the fitted curve to the mean data were reported unless otherwise noted.

To evaluate inhibition of hNaV1.6(N1768D) resurgent currents, synthetic NaVβ4 peptide (KKLITFILKKTREKKKECLV) was added to the intracellular recording solution at 200 μM and a dedicated protocol to elicit resurgent currents was employed (Barbosa et al., 2015). Cells were voltage clamped at (Vh = –80 mV) and subjected to a strong depolarization ( + 60 mV) for 20 ms. Following the strong depolarization, cells were partially repolarized to the voltage where resurgent current was maximal (–20 mV) for 50 ms, and resurgent current amplitude was measured. This resurgent current-specific waveform was repeated at 5 Hz for 100 s in vehicle, followed by 100 s in test compound, then 300 nM TTX. Fractional inhibition was calculated using the same normalization procedure as above.

Experiments were all performed at 27°C ± 2°C.

Automated patch-clamp recording solutions

Request a detailed protocol

The recording solutions for NaV1.1, NaV1.2, NaV1.3, NaV1.4 and NaV1.6 cell line studies contained: Intracellular solution (ICS): 5 mM NaCl, 10 mMCsCl, 120 mMCsF, 0.1 mM CaCl2, 2 mM MgCl2, 10 mM HEPES (4-(2-hydroxyethyl)–1-piperazineethanesulfonic acid buffer), 10 mM EGTA (ethylene glycol tetraacetic acid); adjusted to pH 7.2 with CsOH. Extracellular solution (ECS): 140 mM NaCl, 5 mMKCl, 2 mM CaCl2, 1 mM MgCl2, 10 mM HEPES; adjusted to pH 7.4 with NaOH. Solutions with a reversed Na+ gradient were used for NaV1.5 and NaV1.7 studies since they improved technical success. ICS: 120 mM NaF, 10 mMCsCl, 0.1 mM CaCl2, 2 mM MgCl2, 10 mM HEPES, 10 mM EGTA; adjusted to pH 7.2 with CsOH. ECS: 1 mM NaCl, 139 mMCholineCl, 5 mMKCl, 2 mM CaCl2, 1 mM MgCl2, 10 mM HEPES; adjusted to pH 7.4 with NaOH. Osmolarity in all ICS and ECS solutions was adjusted with glucose to 300 mOsm/kg and 310 mOsm/kg, respectively.

Current-clamp recording of cortical pyramidal neurons and inhibitory interneurons

Slice preparation

Request a detailed protocol

Parasagittal cortical brain slices were prepared from> P21 mice using standard procedures (adapted from Tai et al., PNAS 2014). Briefly, the mouse was deeply anaesthetized with isoflurane and decapitated. The brain was removed and placed into chilled artificial cerebrospinal fluid (aCSF) solution containing (in mM): 125 NaCl, 25 NaHCO3, 2.5 KCl, 1.25 NaH2PO4, 2 CaCl2, 2 MgCl2, 10 d-glucose, pH 7.3, osmolarity adjusted to ~306 mOsm using sucrose. All solutions were saturated with 95% O2 and 5% CO2 constantly perfused with 95% O2/5% CO2. Slices with a thickness of 400 µm were prepared using a vibratome (Ted Pella, Inc). Following sectioning, the slices were placed in a holding chamber and incubated in a water bath at 34 °C for 15 min. The brain slices were removed from the water bath and held at room temperature for 60 min prior to recording.

Brain slice electrophysiology assay

Request a detailed protocol

All experiments involving rodent subjects were performed in accordance with the guidelines of the Canadian Council on Animal Care (CCAC). Following a 60-min incubation at room temperature, a brain slice was selected and placed on the stage of an upright microscope (SliceScope Pro 2000, Scientifica). The slice was constantly perfused with room temperature aCSF, containing 0.1% DMSO as a vehicle control, and oxygenated with 95% O2/5% CO2.The slice was visualized using brightfield microscopy, and a healthy neuron was selected from neocortical layer 5. Whole-cell configuration was achieved with a pipette (bath resistance 4–6 MΩ) containing internal solution. Stimulation was applied in current-clamp mode, and consisted of a series of 1000ms square pulses, beginning at –20 pA and increasing by +20 pA increments (3000ms between pulses).

Once the recordings in vehicle were completed, and while still holding the patch on the same neuron, the bath solution was changed from 0.1% DMSO in aCSF to 0.25 µM NBI-921352 or 100 µM Carbamazepine in aCSF. The slice was incubated in circulating compound for 10 min before repeating the series of square pulse stimulations. Working stock solutions were prepared in DMSO at a concentration of 20 mM.

All data analysis was done offline using ClampFit 10.7 (Molecular Devices). Data are presented as a mean ± SEM. For each sweep, the number of evoked APs was counted, and plotted as a function of current injection (beginning with –20 pA). These generated ‘input/output’ (or ‘F/I’) curves demonstrating the relationship between stimulus and average AP frequency. Statistical significance was assessed using paired, two-way, student’s t-test applied at each current injection level with significance considered p < 0.05.

Formulation and oral dosing of NBI-921352

Vehicle preparation

Request a detailed protocol

The vehicle for oral dosing solutions was 0.5% methyl cellulose and 0.2% Tween-80 in deionized (DI) water. DI water (0.8 L) was heated up to 70–80°C. Five grams of methyl cellulose was slowly added to heated DI water. The mixture was stirred until it formed a homogeneous milky suspension. The suspension was moved to a cold room and stirred overnight to get a clear solution. Two milliliters of Tween-80 was added to the clear solution and diluted up to 1 L with DI water. The vehicle solution was stored at 2–8°C.

Drug formulation

Request a detailed protocol

NBI-921352 was weighed into vials. An appropriate amount of vehicle was added to the NBI-921352 powder then mixed on a T18 ULTRA TURRAX homogenizer (IKA, Wilmington, NC) to create a uniform suspension at the desired concentration. The vials were then wrapped in aluminum foil to protect them from light and placed on a stir plate until the time of dosing. Carbamazepine and lacosamide were formulated in the same manner.Phenytoin was formulated in 0.9% physiological saline.

Dosing

Request a detailed protocol

NBI-921352, carbamazepine, and lacosamide were administered orally using a stainless-steel gavage needle at a dose volume of 10 ml/kg. Phenytoin was formulated in physiologic saline and was administered intraperitoneally (i.p.) using a 25-gauge needle at a dose volume of 10 mL/kg. All compounds were administered 2 hr prior to electrical seizure induction for all seizure models employed in this study.

Bioanalytical assessment of plasma and brain concentrations

Sample collection:

Request a detailed protocol

Approximately 0.5 mL of blood was collected from each mouse at the end of the assay via cardiac puncture under deep anesthesia. The blood samples were collected in a syringe and transferred to tubes containing EDTA. Blood was stored at 4 °C until centrifuged within 30 min of collection. Plasma was harvested and placed on dry ice and stored in a freezer set to maintain a temperature of –70°C to –80°C until analysis. Brains were harvested immediately after blood collection and placed on dry ice prior to storage in a freezer set to maintain a temperature of –70°C to –80°C until analysis.

Plasma samples:

Request a detailed protocol

Extraction of plasma samples was carried out by protein precipitation using acetonitrile. Plasma samples (50 µL) were mixed with 50 µL of internal standard (IS) solution in water followed by addition of 10 µL of concentrated ortho-phosphoric acid and 200 µL of acetonitrile. Samples were vortexed for 30 seconds, centrifuged at 13,000 rpm for 20 min, decanted in to a 96-well plate, and further centrifuged at 4000 rpm for 20 min. The samples were analyzed by UHPLC-ESI-MS/MS as described below.

Brain samples:

Request a detailed protocol

Prior to extraction, pre-weighed whole brains were homogenized in 1:1 acetonitrile/water (v/v) (4 mL per mouse brain) using an IKA T18 ULTRA-TURRAX Homogenizer at the setting of 4 for approximately 2 min. The homogenate was centrifuged at 13,000 rpm for 20 min and 50 µL of the supernatant were treated exactly as described above for plasma samples. Fifty µL of the brain homogenate were then treated exactly as the plasma samples described above.

Standards and quality control (QC) samples:

Request a detailed protocol

K2EDTA Blank mouse plasma purchased from Valley Biomedical, California, USA was used to prepare standards and QC samples for plasma quantitation and as surrogates for brain homogenate quantitation. Calibration samples ranged from 2.34 ng/mL to 4800 ng/mL. QC samples concentration included 14 ng/mL (QC-L), 255 ng/mL (QC-M), and 3600 ng/mL (QC-H). Standards and QC samples were processed the same way as the sample extracts described above.

Analytical methods and statistics for plasma and tissue samples:

Request a detailed protocol

Samples were analyzed by UHPLC-ESI MS/MS using a TQ-5500 Sciex triple quadrupole mass spectrometer equipped with a Shimadzu Nexera UHPLC pump and auto-sampler system using an ACE C18 PFP, 2.50 × 50 mm, 1.7 µm particle size column and gradient elution consisting of solvent A (0.1% formic acid in water) and solvent B (0.1% formic acid in acetonitrile) starting at 20% B from 0 min to 0.4 min and then increased to 100% B from 0.4 min to 0.6 min. At 2.0 min, the mobile phase composition was switched back to 60% B for 1 min. The flow rate used throughout the experiment was 0.4 min/mL. The analyte, NBI-921352, and the IS were detected by electrospray in the positive ion mode using the following transitions: m/z 460/91 for NBI-921352 and m/z 503/341 m/z for the IS. The UHPLC-ESI MS/MS system was controlled by Analyst 1.6.

Sample concentrations were determined using a linear calibration function, weighted 1 /X, generated by the regression of analyte to IS peak area ratios in the standard samples to their respective concentrations. Acceptance criteria for the analytical run required that the back calculated values of the standards and the QC samples fell within ±20% of their nominal values, except for the lowest standard or lower limit of quantitation (LLOQ), for which the acceptance criterion was ±25%. At least 6 out of 12 standard points had to show back-calculated values within ±20% of their nominal concentrations for the calibration to be accepted. At least three QC samples, one at each level, had to show back-calculated values within ±20% of their nominal concentrations for the whole sample batch to be valid.

Animals

After delivery, animals were allowed sufficient time to acclimate prior to testing (~1 week). All animals were housed in plastic cages in rooms with controlled humidity, ventilation, and lighting (12 hr/12 hr light–dark cycle). All animal procedures were performed using protocols approved by Xenon Animal Care Committee and the Canadian Council on Animal Care.

Scn8aN1768D/+ mice:

Request a detailed protocol

Xenon Pharmaceuticals Inc licensed the mouse with the missense mutation p.Asn1768Asp (N1768D) in the neuronal sodium channel NaV1.6, characterized and developed by Dr. M Meisler (University Of Michigan, MI, USA). The Scn8aN1768D knock-in allele was generated by TALEN targeting of (C57BL/6JXSJL) F2 eggs at the University of Michigan Transgenic Animal Model Core. The line was propagated by backcrossing N1768D/+ heterozygotes to C57BL/6 J wild-type mice (The Jackson Laboratory, Bar Harbor, ME). Male N1768D/+ heterozygotes on a C57BL/6 J background were subsequently backcrossed to C3HeB/FeJ female mice. All the experiments were performed using animals following at least seven such backcrosses. Experiments were performed using (B6× C3 He) F7 (F7. N1768D/+) offspring aged 35–42 days.

WT mice:

Request a detailed protocol

Adult male CF-1 WT albino mice 26–35 g were obtained from Charles River, Senneville, Quebec, Canada. All the assays were carried out in mice 9–12 weeks of age.

Sprague-Dawley rats:

Request a detailed protocol

Adult male Sprague-Dawley albino rats weighing 150–200 g were obtained from Envigo, Livermore, CA, USA. All the assays were carried out in rats aged 5–6 weeks.

The modified 6 hz psychomotor seizure assay

Request a detailed protocol

Scn8aN1768D/+heterozygous mice were tested at 5 weeks of age (p35). The modified 6 Hz seizure assay in Scn8aN1768D/+heterozygous mice was adapted from the traditional 6 Hz assay psychomotor seizure assay to provide a measure of in vivo on target (NaV1.6 mediated) efficacy (Barton et al., 2001). The modified assay used a low frequency (6 Hz) but long-duration stimulation (3 s) to induce seizures. We identified 12 mA and a 0.3-ms pulse interval as a suitable current for testing in Scn8aN1768D/+ mice, since it differentiated mutant and wild-type (WT) mice. An electroshock (6 Hz, 12 mA) was delivered for 3 s (at 0.3-ms pulse interval) by corneal electrodes (Electro Convulsive Therapy Unit 57,800 from Ugo Basile). Immediately prior to the electroshock, the animals’ eyes were anesthetized with a drop of Alcaine (0.5% proparacaine hydrochloride). Upon corneal stimulation, WT mice experienced mild seizure behaviors such as facial clonus, forelimb clonus, Straub tail, rearing, and falling, but did not experience a generalized tonic-clonic seizure (GTC) with hindlimb extension. Scn8aN1768D/+ animals, however, in addition to mild seizure behaviors, experienced a GTC with hindlimb extension. The modified assay showed a clear differentiation of seizure behavior between WT and Scn8aN1768D/+ mice. Scn8aN1768D/+ mice exhibited GTC with hindlimb extension but not WT mice.

For the single-dose and repeated-dose efficacy experiments, Scn8aN1768D/+ animals were dosed PO with vehicle or NBI-921352 two hours before the administration of the electric stimulation. An animal was considered protected in the assay upon prevention of GTC with hindlimb extension and was then scored ‘0’. An animal displaying GTC with hindlimb extension was considered not protected and is then scored ‘1’. The experimenter scoring the seizure behavior was blinded to the treatment.

DC-Maximal electroshock seizure assay in rodents

Request a detailed protocol

The maximal electroshock seizure (MES) assay has been extensively used in the search for anticonvulsant substances (Löscher et al., 1991; Piredda et al., 1985; White et al., 1995). The MES assay is sensitive to nonselective NaV inhibitors. It is considered a model for generalized tonic-clonic (GTC) seizures and provides an assessment of seizure spread. Briefly, an electroshock of direct current (DC) was delivered by corneal electrodes (Electro Convulsive Therapy Unit 57,800 from Ugo Basile). The parameters of stimulation were different between mice and rats. In CF1 mice aged 9–12 weeks, a direct current of 50 mA (60 Hz) was delivered for 0.2 s (pulse width of 0.5ms), whereas in Sprague Dawley (SD) rats aged 5–6 weeks, a direct current of 150 mA (60 Hz) was delivered for 0.3 s (pulse width of 0.5ms). Immediately prior to the electroshock, the animals’ eyes were anesthetized with a drop of Alcaine (0.5% proparacaine hydrochloride). Upon corneal stimulation, naive animals experienced a generalized tonic-clonic seizure (GTC) with hindlimb extension.

For the efficacy experiments, single dose and repeated dose, animals were dosed PO with vehicle or NBI-921352 2 hr before the administration of the electric stimulation. An animal was considered protected in the assay in the absence of a GTC with hindlimb extension and is then scored ‘0’. An animal displaying GTC with hindlimb extension was considered not protected and is then scored ‘1’. The experimenter scoring the seizure behavior was blinded to the treatment.

Blinding of in vivo efficacy experiments

Request a detailed protocol

On each testing day, individual treatment groups were assigned a random label (e.g. A, B, C, etc.) by the technical staff administering the compound. To ensure blinding, the technical staff member performing drug administration differed from the person performing the test. Therefore, the experimenter conducting testing was blinded to treatment group (e.g. drug or vehicle treatment, dose, and time point).

Randomization of in vivo efficacy experiments

Request a detailed protocol

Randomization of animals into various treatment groups occurred on a per-animal (e.g. rather than a per-cage) basis. Therefore, each animal was randomly assigned to a treatment group, and all animals tested in each experiment had an equal chance of assignment to any treatment group. Prior to each study, a randomization sequence was obtained (https://www.graphpad.com/quickcalcs/).

Observation of potential adverse behavior for Figure 7

Request a detailed protocol

Animals were observed by the scientist(s) responsible for dosing for the first 30 min after dosing for and again at the time of assay (2 hr post dose) for signs of abnormal behavior. Abnormal behavioral signs were noted for all compounds – tremor, ataxia and hypoactivity – when plasma concentrations were sufficiently high. The lowest concentration at which such behavioral signs were noted in any of the animals tested is illustrated by the dotted line in Figure 7. Animals that exhibited abnormal behavior were not assessed in efficacy assays.

Data processing and analysis

Request a detailed protocol

All statistics were calculated using GraphPad Prism version eight software. All fraction seizing data plotted by dose groups is expressed and is an absolute fraction of all tested animals that seized (number of animals that seized / the number of animals tested). Racine score data is reported in Figure 5—figure supplements 1 and 2, and in Figure 6—figure supplements 1 and 2, and 3. Racine score data are plotted by dose groups and are expressed as mean ± SEM. Between-group differences were compared to vehicle response and were analyzed using a Kruskal-Wallis test followed by Dunn’s multiple comparisons test (fraction seizing data) or ordinary one-way ANOVA followed by Dunnett’s multiple comparison test (Racine score data). Statistical significance was reached at values of p < 0.05. More detailed results of the statistical analysis, with p values, is presented in the figure supplements and legends.

Data availability

All the numerical data used to generate the figures in contained in the excel data source file included in the submission.

References

    1. Courtney KR
    2. Kendig JJ
    3. Cohen EN
    (1978)
    The rates of interaction of local anesthetics with sodium channels in nerve
    The Journal of Pharmacology and Experimental Therapeutics 207:594–604.
    1. Gennaro E
    2. Veggiotti P
    3. Malacarne M
    4. Madia F
    5. Cecconi M
    6. Cardinali S
    7. Cassetti A
    8. Cecconi I
    9. Bertini E
    10. Bianchi A
    11. Gobbi G
    12. Zara F
    (2003)
    Familial severe myoclonic epilepsy of infancy: truncation of Nav1.1 and genetic heterogeneity
    Epileptic Disorders 5:21–25.
  26. Book
    1. Hammer MF
    2. Wagnon JL
    3. Mefford HC
    4. Meisler MH
    (2016)
    GeneReviews((R))
    Seattle , WA.
    1. Piredda SG
    2. Woodhead JH
    3. Swinyard EA
    (1985)
    Effect of stimulus intensity on the profile of anticonvulsant activity of phenytoin, ethosuximide and valproate
    The Journal of Pharmacology and Experimental Therapeutics 232:741–745.
    1. Raman IM
    2. Bean BP
    (1997)
    Resurgent sodium current and action potential formation in dissociated cerebellar Purkinje neurons
    The Journal of Neuroscience 17:4517–4526.

Decision letter

  1. Baron Chanda
    Reviewing Editor; Washington University in St. Louis, United States
  2. Richard W Aldrich
    Senior Editor; The University of Texas at Austin, United States
  3. Theodore Cummins
    Reviewer

Our editorial process produces two outputs: (i) public reviews designed to be posted alongside the preprint for the benefit of readers; (ii) feedback on the manuscript for the authors, including requests for revisions, shown below. We also include an acceptance summary that explains what the editors found interesting or important about the work.

Decision letter after peer review:

Thank you for submitting your article "NBI-921352, a First-in-Class, Na V 1.6 Selective, Sodium Channel Inhibitor That Prevents Seizures in Scn8a Gain-of- Function Mice, and Wild-Type Mice and Rats" for consideration by eLife. Your article has been reviewed by 2 peer reviewers, and the evaluation has been overseen by a Reviewing Editor and Richard Aldrich as the Senior Editor. The following individual involved in review of your submission has agreed to reveal their identity: Theodore Cummins (Reviewer #2).

The reviewers have discussed their reviews with one another, and the Reviewing Editor has drafted this to help you prepare a revised submission. Please note that some of the points noted below are related and should be addressed comprehensively.

Essential revisions:

1) The compound was not tested against Nav1.8 and Nav1.9 channels. It may be unlikely that NBI-921352 would inhibit either of these as they are more divergent in the likely binding region than the other isoforms test, this is not a given and is a slight weakness. While testing these two isoforms is not impossible, it may not be necessary. But the lack of data at least needs to be discussed.

2) To facilitate the binding/action of NBI-921352, a depolarized holding potential (-45 mV) was used in the voltage-clamp recording for most VGSC isoforms (except Nav1.7 and Nav1.5) and Nav1.6 variants (except N1768D). This EP protocol was used for Nav1.1, Nav1.2, and Nav1.6, the three major VGSC isoforms in the central nervous system, and therefore, the comparison among those three isoforms should be reliable, despite the depolarized holding potential (-45 mV). However, it is unlikely for neurons to experience such sustained depolarizing state (-45 mV) under physiological conditions, and therefore, the selectivity of NBI-921352 on Nav1.6 vs other VGSCs under physiological conditions could be different compared to the values reported here. It is better to hold cells at physiologically relevant membrane potentials or using action potential waveforms derived from real AP recordings of pyramidal neurons or interneurons. The authors should discuss these limitations, and possible impact on their assessment of selectivity against other VGSCs in their native cellular backgrounds.

3) While the electrophysiological data clearly show that NBI-921352 shows voltage-dependence and likely state dependence, the study lacks an assay of use-dependence using repetitive higher frequency depolarizations. Something like 10 or 20 Hz pulsing might be informative.

4) In several places it is stated that NBI-921352 preferentially inhibits activated (inactivated or open) channels. There is no direct assay looking at inhibition of open but not-inactivated channels, so this conclusion is potentially misleading. The authors could try something like a WCW non-inactivating construct, but I think just rewording those places where it says inactivated or open to maybe "inactivating and possibly also open" would be reasonable.

5) Different voltage-clamp protocols were adopted among different sodium channel isoforms (lines 406-418) to measure IC50 concentrations of the drug, which requires clarification/justification. Since NBI-921352 is "a state-dependent inhibitor" (line 30) and "requires several seconds to equilibrate with activated channels" (line 401), the difference in holding potentials (-60 mV vs -45 mV), recovery period (20 ms vs 60 ms at -150 mV), and data acquisition rates (1 Hz vs 0.04 Hz) may alter the binding/interaction of NBI-921352 with tested channels, and therefore, introduce bias in the comparison of selectivity among VGSCs. The authors should explain their rationale for choosing these protocols and address the potential confounding effects in data analysis.

6) The statement that a preference for persistent currents is, "in fact, a feature of all the compounds in the Nav inhibitor class" is not justified by either data presented in the study or by an appropriate reference. The "in fact" is particularly not justified since it really states a bias that would need to be bolstered by hard data. Not much in biology is proven as a fact. This bold statement may have some validity, but it is not really important to the major conclusions of the study.

7) It is stated that NBI-921352 is rapidly cleared in mice, but it is unclear if this is different in humans, which could limit the translational potential. It would help the reader if this was discussed.

8) Lines 248-249 states unequivocally that the efficacy and adverse events of Nav inhibitors is driven by action in the CNS. There is no citation for this and it comes across as another bias. It seems intuitive that the anti-seizure activity is dependent on CNS activity, but this is not demonstrated in the current study. Adverse events could involve sodium channels in muscle and/or the peripheral nervous system. Therefore this statement should be modified.

9) Details on what behaviors were monitored and how they were monitored for figure 7 need to be included or else the bars and discussion on this aspect of the study should be removed.

10) Nav1.6 is highly expressed in Purkinje neurons and motor neurons, and plays important roles in motor system. Did the authors observe any motor impairment in the behavior studies? It would have been informative to examine the effect of NBI-921352 on AP firing and resurgent currents in Purkinje neurons. As Nav1.6 is the major isoform found at nodes of Ranvier in myelinated fibers, including motor neurons axons, it is critical to know if NBI-921352 alters motor neuron and/or motor activity.

11) Three seizure models were used to evaluate the effect of NBI-921352 on electrically induced seizure. The data shows that NBI-921352 prevented the GTC with hindlimb extension in all three assays, demonstrating its potential to become an antiepileptic drug. In an earlier work from the same group (J Med Chem 2019; 62: 9618-9641), a modified Racine scale (0-5) was used to score the severity of different seizure activities, while only GTC with hindlimb extension (score 5 in Racine scale) is considered as seizure in this study. Is there any reason to choose this evaluation criteria instead of Racine scale used in previous work?

12) The authors tested seizure behaviors after repeated dosing of the drug, but claimed that there was no statistically significant difference between single-dose group vs repeated-dose group (lines 238-246). I was unable to find the description of statistical analysis method for behavioral studies in the manuscript.

13) There are several problems in the brain slice recording experiments. First, paired student t-test was used to compare the AP firing before and after drug treatment, which is inappropriate. Repeated measures two-way ANOVA should be used here, or use area under curve rather than number of APs for statistical analysis. Second, despite the dominant effect of Nav1.1 in inhibitory interneurons, Nav1.6 is also expressed in interneurons, and to confirm selectivity in the mouse model, it is better to examine the effect of NBI-921352 on AP firing in Nav1.6-knockout neurons. Third, the experiments lack a vehicle control data set. Fourth, the number of cells for each group is low (3-4 cells), more recordings should be performed.

14) Please add a section of statistical analysis in Materials and methods, and list the statistical analysis method used in each experiment.

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

Thank you for resubmitting your work entitled "NBI-921352, a First-in-Class, Na V 1.6 Selective, Sodium Channel Inhibitor That Prevents Seizures in Scn8a Gain-of- Function Mice, and Wild-Type Mice and Rats" for further consideration by eLife. Your revised article has been evaluated by Richard Aldrich (Senior Editor) and a Reviewing Editor.

The manuscript has been improved but there are some remaining issues that need to be addressed, as outlined below:

1. The reviewer 1 pointed out the lack of NBI-921352 data on Nav1.8 and Nav1.9, and the authors performed limited study on hNav1.8 co-expressed with β3 in CHO cells.

I am not clear why the authors selected CHO cells as the expression system, as it is known that expression of Nav1.8 is poor in CHO and HEK293 cells, even in the presence of β subunits (Nature 2002, 417: 653-656). The ND7/23 is a more commonly used system for Nav1.8 expression (for example, Toxins 2021, 13:501). I have obtained decent Nav1.8 currents (over 10 nA, with properties similar to those measured in DRG neurons) in ND7/23 cells, but barely saw Nav1.8 currents in HEK 293 cells co-transfected with hNav1.8 and β3. It is unclear how good the Nav1.8 expression was in CHO cells in authors' hands, the authors should include raw traces of Nav1.8 in Figure 1-Supplement 3 to demonstrate the reliable Nav1.8 expression in this assay. If the expression of Nav1.8 is not good, I would recommend removing the Nav1.8 data, though the authors still need to explain the lack of data on those two isoforms and discuss possible effect of NBI-921352.

Typos: there are quite some typos in figure legend of Figure 1-Supplement 3. For example, line 1045, "% Block (Tonic) = (1 ITPI,NBI921352 / ITPl,Control) X 100%", should "ITPl" be "ITP1"? And the calculation should be "(1-ITP1,NBI921352/ITP1,Control) x 100%"? Same typo is shown in line 1046.

In line 1049, "% Block (inactivation state) -", should "-" be "="?

2. I have concerns on different holding potentials being used in determining IC50 of NBI-921352 for different isoforms. The authors responded that "We chose to focus on what we refer to as "molecular selectivity," the fundamental ability for a compound to bind to the channel and stabilize the high affinity conformation. We accomplish this by choosing voltages that promote the same fraction of channels to be in the high affinity (inactivated) state." I wonder how the authors determined that "-45 mV" is the potential for NBI-921352 to bind and reach the high affinity conformation of Nav1.1, 1.2, 1.3, 1.4, and 1.6, and "-60 mV" for Nav1.7 and Nav1.5, there is no justification of this in the manuscript. In addition, when a medicine is applied to a patient, the adverse effect is determined by its affinity and effect on all targeted channels/molecules physiological or pathological conditions, therefore, the relative selectivity of NBI-921352 determined using the same protocol/condition would be more meaningful for clinical practice.

The new data that NBI-921352 showed similar potency in Nav1.6 channel at -62 mV when compared to that at -45 mV is important, this data suggests that the IC50 of NBI-921352 is not significantly changed by the holding potential (or membrane potential) between -62 mV and -45 mV. Although I don't agree with the authors' explanation, I will not ask for extra experiments. However, unless the authors provide convincing evidence that the two holding potentials (-45 mV and -60 mV) used in this study do allow different isoforms to be in their "highest affinity state (inactivated)", the explanation in Page 6 (line 124 to line 138) needs to be modified/removed.

Typo: line 137, "is this assay" should be "in this assay".

3. Regarding to the state of channels that NBI-921352 preferentially inhibits, the abstract wrote "NBI-921352 is a state-dependent inhibitor, preferentially inhibiting inactivated channels", while the figure legend 3 shows "NBI-921352 is a state-dependent inhibitor of NaV1.6 and preferentially inhibited activated (open or inactivated) channels". In main text (page 20, line 424-426), "NBI-921352 is also highly state dependent, with a >750-fold preference for activated channels vs. rested, closed-state channels (sometimes referred to as peak current)". Which state does NBI-921352 preferentially act?

4. The figure legend 5 is confusing (page 63, line 1159-1169), please rephrase the sentences.

https://doi.org/10.7554/eLife.72468.sa1

Author response

Essential revisions:

1) The compound was not tested against Nav1.8 and Nav1.9 channels. It may be unlikely that NBI-921352 would inhibit either of these as they are more divergent in the likely binding region than the other isoforms test, this is not a given and is a slight weakness. While testing these two isoforms is not impossible, it may not be necessary. But the lack of data at least needs to be discussed.

Response: We have a limited dataset on inhibition of hNaV1.8, and we have added this data to the Results section, and to the table in Supplementary file 1. We do not have ready access to an hNaV1.9 assay (internally or commercially). Since neither NaV1.8 or NaV1.9 are believed to be involved in epilepsy or central excitability we chose not to pursue additional data. We have now discussed this gap in the text. See Results line 108.

2) To facilitate the binding/action of NBI-921352, a depolarized holding potential (-45 mV) was used in the voltage-clamp recording for most VGSC isoforms (except Nav1.7 and Nav1.5) and Nav1.6 variants (except N1768D). This EP protocol was used for Nav1.1, Nav1.2, and Nav1.6, the three major VGSC isoforms in the central nervous system, and therefore, the comparison among those three isoforms should be reliable, despite the depolarized holding potential (-45 mV). However, it is unlikely for neurons to experience such sustained depolarizing state (-45 mV) under physiological conditions, and therefore, the selectivity of NBI-921352 on Nav1.6 vs other VGSCs under physiological conditions could be different compared to the values reported here. It is better to hold cells at physiologically-relevant membrane potentials or using action potential waveforms derived from real AP recordings of pyramidal neurons or interneurons. The authors should discuss these limitations, and possible impact on their assessment of selectivity against other VGSCs in their native cellular backgrounds.

There are pros and cons to any method of determining selectivity and we acknowledge that none of them are ideal for all purposes. We chose to focus on what we refer to as “molecular selectivity,” the fundamental ability for a compound to bind to the channel and stabilize the high affinity conformation. We accomplish this by choosing voltages that promote the same fraction of channels to be in the high affinity (inactivated) state. This contrasts with “functional selectivity” that may be largely driven by the distinct state-dependence of different isoforms. Our approach avoids assumptions about what the physiologically relevant voltage is since that voltage can vary depending on the tissue or cell type. For any given isoform there may be multiple physiologically relevant voltages.

Consistent with this philosophy, we bias all the channels to be in their highest affinity state (inactivated) and then use this maximal potency to compare selectivity. At more hyperpolarized, voltages, potency for all isoforms will tend to be somewhat less. We are adding more explanation of our rationale to the text, and we are adding supplemental data giving more insight into the impact of voltage on potency.

Figure 1—figure supplement 2 shows the potency of NBI-921352 after holding at a membrane potential nearer the physiologic range (-62mV). Potency at this voltage (IC50 = 53 nM) was similar to that at fully inactivated potentials evaluated in the primary potency assay described shown in Figure 1. For this reason, we anticipate that the selectivity ratios described in the manuscript will be similar to those in physiologic conditions.

A note to this effect has been added to the Results section. See line 126.

3) While the electrophysiological data clearly show that NBI-921352 shows voltage-dependence and likely state dependence, the study lacks an assay of use-dependence using repetitive higher frequency depolarizations. Something like 10 or 20 Hz pulsing might be informative.

We are in the process of preparing a manuscript on related compounds that will show a more detailed evaluation of the kinetics and state dependence of inhibition of NaV1.6. We have not yet explored this behavior as thoroughly with NBI-921352, and we feel that a more fulsome treatment of the biophysics may distract from the selectivity and pharmacology that we intend as the focus for the present work. We do know from our work in progress that block equilibrates over several seconds, and thus we do not expect a great deal of use-dependence at physiologically relevant firing rates.

4) In several places it is stated that NBI-921352 preferentially inhibits activated (inactivated or open) channels. There is no direct assay looking at inhibition of open but not-inactivated channels, so this conclusion is potentially misleading. The authors could try something like a WCW non-inactivating construct, but I think just rewording those places where it says inactivated or open to maybe "inactivating and possibly also open" would be reasonable.

Our data suggests preferential binding to inactivated states, but as the reviewer rightly points out, we cannot rule out (or in) binding to open states. Our intent with the “inactivated or open” wording was to account for this possibility. We are changing the wording to that recommended by the reviewer in hopes of clarifying this point. Our experience with constructs that impair fast inactivation (like WCW) have not been helpful for clarifying this issue because we find that these changes accelerate slow inactivation, and our compounds do not appear to clearly distinguish between slow and fast inactivated channels. Changes throughout document to replace “inactivated or open.”

5) Different voltage-clamp protocols were adopted among different sodium channel isoforms (lines 406-418) to measure IC50 concentrations of the drug, which requires clarification/justification. Since NBI-921352 is "a state-dependent inhibitor" (line 30) and "requires several seconds to equilibrate with activated channels" (line 401), the difference in holding potentials (-60 mV vs -45 mV), recovery period (20 ms vs 60 ms at -150 mV), and data acquisition rates (1 Hz vs 0.04 Hz) may alter the binding/interaction of NBI-921352 with tested channels, and therefore, introduce bias in the comparison of selectivity among VGSCs. The authors should explain their rationale for choosing these protocols and address the potential confounding effects in data analysis.

This question relates to the same underlying issues as Essential Question #2. I believe that the changes we made for #2 will also address this question.

6) The statement that a preference for persistent currents is, "in fact, a feature of all the compounds in the Nav inhibitor class" is not justified by either data presented in the study or by an appropriate reference. The "in fact" is particularly not justified since it really states a bias that would need to be bolstered by hard data. Not much in biology is proven as a fact. This bold statement may have some validity, but it is not really important to the major conclusions of the study.

We have altered our language to allow for the uncertainty pointed out by the Reviewer. See line 191.

7) It is stated that NBI-921352 is rapidly cleared in mice, but it is unclear if this is different in humans, which could limit the translational potential. It would help the reader if this was discussed.

Rodents, both rats and mice, rapidly eliminate NBI-921352, but the compound is much more stable in humans, where the half-life of elimination is approximately 8.5 hours. We have added this information to the Results section. See line 277.

8) Lines 248-249 states unequivocally that the efficacy and adverse events of Nav inhibitors is driven by action in the CNS. There is no citation for this and it comes across as another bias. It seems intuitive that the anti-seizure activity is dependent on CNS activity, but this is not demonstrated in the current study. Adverse events could involve sodium channels in muscle and/or the peripheral nervous system. Therefore this statement should be modified.

We agree and have modified the language to account for this uncertainty. See line 322.

9) Details on what behaviors were monitored and how they were monitored for figure 7 need to be included or else the bars and discussion on this aspect of the study should be removed.

Animals were observed by the scientist(s) responsible for dosing for the first 30 minutes after dosing for and again at the time of assay (2 hrs post dose) for signs of abnormal behavior. Abnormal behavioral signs were noted for all compounds – tremor, ataxia and hypoactivity – when plasma concentrations were sufficiently high. The lowest concentration at which such behavioral signs were noted in any of the animals tested is illustrated by the dotted line in figure 7. We have added more detail to the methods (see line 790) to clarify this point.

10) Nav1.6 is highly expressed in Purkinje neurons and motor neurons, and plays important roles in motor system. Did the authors observe any motor impairment in the behavior studies? It would have been informative to examine the effect of NBI-921352 on AP firing and resurgent currents in Purkinje neurons. As Nav1.6 is the major isoform found at nodes of Ranvier in myelinated fibers, including motor neurons axons, it is critical to know if NBI-921352 alters motor neuron and/or motor activity.

As noted in the response to Essential Revision #9, we did see signs of motor impairment at sufficiently high plasma concentrations of NBI-921352. While we cannot assure that these motor effects are due to inhibition of NaV1.6, we agree that it is consistent with the known role of NaV1.6 in the Purkinje neurons and in the nodes of Ranvier of motor neurons. Mice with sufficient knock down of NaV1.6 by genetic interventions are also known to exhibit motor dysfunction. Our data suggest that seizure protection occurs at lower concentrations/occupancy than those that induce ataxia. We believe that sparing other central NaV’s (NaV1.1 and NaV1.2) improves the profile of these inhibitors, but we do expect to see ontarget (NaV1.6 mediated) toxicity when receptor occupancy becomes too high. See additional discussion beginning at line 408.

11) Three seizure models were used to evaluate the effect of NBI-921352 on electrically induced seizure. The data shows that NBI-921352 prevented the GTC with hindlimb extension in all three assays, demonstrating its potential to become an antiepileptic drug. In an earlier work from the same group (J Med Chem 2019; 62: 9618-9641), a modified Racine scale (0-5) was used to score the severity of different seizure activities, while only GTC with hindlimb extension (score 5 in Racine scale) is considered as seizure in this study. Is there any reason to choose this evaluation criteria instead of Racine scale used in previous work?

We believe that both the Racine endpoint shown in Focken et al., 2019, and The GTC with hindlimb extension endpoint shown here are both valid. Our experience with this assay has been that the seizure phenotype is rather binary. That is, individual animals show very little seizure behavior prior to the GTC with hindlimb extension. As a result, we have found, based on studies with NBI-921352 and other compounds not shown here, that there is little difference in the efficacy measured by the method shown here versus the Racine evaluation used in the previous work. We chose to use the GTC with Hindlimb extension endpoint in this presentation to more easily compare and discuss the results in SCN8A mice to those presented for wild-type mice and rats in the DC-MES assays.

We have added the Racine score evaluation for these experiments in Figure 5—figure supplements 1 and 2, Figure 6—figure supplements 1, 2, and 3 to enable comparison with the SCN8A mouse fraction seizing results shown in the primary figures.

12) The authors tested seizure behaviors after repeated dosing of the drug, but claimed that there was no statistically significant difference between single-dose group vs repeated-dose group (lines 238-246). I was unable to find the description of statistical analysis method for behavioral studies in the manuscript.

We are including supplementary figures (Figure 5—figure supplements 1 and 2, Figure 6—figure supplements 1, 2, and 3) to better illustrate the statistical evaluation of the dose response data sets for all the experiments, and we have amended the text to indicate this. See Methods line 798, and the legends for Figure 5, Figure 5—figure supplements 1 and 2, Figure 6, and Figure 6—figure supplements 1, 2, and 3.

13) There are several problems in the brain slice recording experiments. First, paired student t-test was used to compare the AP firing before and after drug treatment, which is inappropriate. Repeated measures two-way ANOVA should be used here, or use area under curve rather than number of APs for statistical analysis. Second, despite the dominant effect of Nav1.1 in inhibitory interneurons, Nav1.6 is also expressed in interneurons, and to confirm selectivity in the mouse model, it is better to examine the effect of NBI-921352 on AP firing in Nav1.6-knockout neurons. Third, the experiments lack a vehicle control data set. Fourth, the number of cells for each group is low (3-4 cells), more recordings should be performed.

We have changed the statistical analysis to an area under the curve analysis as suggested by the reviewer. See text line 234, Figure 4 legend and transparency Excel data file. NaV1.6 knockout experiments would be of interest, but we do not have access to those neurons. We agree that cells would be desirable. Unfortunately, the pandemic disrupted these slice experiments. Subsequently, NBI-921352 advanced in its clinical development. This creates additional challenges and costs to adhere to regulatory reporting requirements and has prevented us from extending this dataset.

14) Please add a section of statistical analysis in Materials and methods, and list the statistical analysis method used in each experiment.

We’ve added statistical methods to the methods and to the figure legends where appropriate. See methods line 798, figure legends, and supplemental figure legends.

[Editors' note: further revisions were suggested prior to acceptance, as described below.]

The manuscript has been improved but there are some remaining issues that need to be addressed, as outlined below:

1. The reviewer 1 pointed out the lack of NBI-921352 data on Nav1.8 and Nav1.9, and the authors performed limited study on hNav1.8 co-expressed with β3 in CHO cells.

I am not clear why the authors selected CHO cells as the expression system, as it is known that expression of Nav1.8 is poor in CHO and HEK293 cells, even in the presence of β subunits (Nature 2002, 417: 653-656). The ND7/23 is a more commonly-used system for Nav1.8 expression (for example, Toxins 2021, 13:501). I have obtained decent Nav1.8 currents (over 10 nA, with properties similar to those measured in DRG neurons) in ND7/23 cells, but barely saw Nav1.8 currents in HEK 293 cells co-transfected with hNav1.8 and β3. It is unclear how good the Nav1.8 expression was in CHO cells in authors' hands, the authors should include raw traces of Nav1.8 in Figure 1-Supplement 3 to demonstrate the reliable Nav1.8 expression in this assay. If the expression of Nav1.8 is not good, I would recommend removing the Nav1.8 data, though the authors still need to explain the lack of data on those two isoforms and discuss possible effect of NBI-921352.

Typos: there are quite some typos in figure legend of Figure 1-Supplement 3. For example, line 1045, "% Block (Tonic) = (1 ITPI,NBI921352 / ITPl,Control) X 100%", should "ITPl" be "ITP1"? And the calculation should be "(1-ITP1,NBI921352/ITP1,Control) x 100%"? Same typo is shown in line 1046.

In line 1049, "% Block (inactivation state) -", should "-" be "="?

The primary reason we did not pursue experiments with NaV1.8 or NaV1.9 for this work is that those channels are not directly relevant to our primary goal of understanding sodium channel inhibition in epilepsy. Both NaV1.8 and NaV1.9 are largely restricted to peripheral neurons and are not expected to impact the efficacy of sodium channel inhibitors as antiseizure drugs. We also do not expect inhibition of Nav1.8 or Nav1.9, if it were to occur, to create safety liabilities.

For these reasons we chose not to invest in developing assays for these channels for this work. We have endeavored to make this clear in the manuscript.

The NaV1.8 data provided was performed at a commercial laboratory (Charles River) and the assay is not well matched to our own internal assays. This mismatch was the reason that we initially excluded the data from the manuscript. We will remove the NaV1.8 data since it does not impact our conclusions.

Our apologies for the typographical errors in the legend, we have removed it since the NaV1.8 data in Figure 1—figure supplement 3 will no longer be included in the manuscript.

We will address question 3 before question 2 since the response to question 3 informs our response to question 2.

3. Regarding to the state of channels that NBI-921352 preferentially inhibits, the abstract wrote "NBI-921352 is a state-dependent inhibitor, preferentially inhibiting inactivated channels", while the figure legend 3 shows "NBI-921352 is a state-dependent inhibitor of NaV1.6 and preferentially inhibited activated (open or inactivated) channels". In main text (page 20, line 424-426), "NBI-921352 is also highly state dependent, with a >750-fold preference for activated channels vs. rested, closed-state channels (sometimes referred to as peak current)". Which state does NBI-921352 preferentially act?

We believe that NBI-921352 preferentially binds to inactivated channels. This is based on the failure of the compound to potently block rested (closed) channels. Also, the presumed binding site (the UP, activated, conformation of the Domain IV voltage sensor, VSD4) only becomes available once VSD4 has reacted to the membrane field to trigger channel inactivation (see Ahuja et al., 2015, McCormack et al., 2013). We have not done detailed studies to exclude the possibility that open channels can be bound, though it seems unlikely to be a major driver of inhibition based on the limited time that the channels are in open, relative to inactivated, states. While the assays are designed with a bias toward inactivated states, we have not attempted to differentiate between fast and slow inactivated state inhibition. Our main objective for this manuscript is to define the importance of selective inhibition of NaV1.6 for treating epilepsy. We expect further clarification of the interaction of drug binding and channel gating will be a topic for future work.

We have corrected the language in the manuscript and in the legend for Figure 3 to make clear that NBI-921352 targets inactivated states but that resolution between open, fast inactivated, and slow inactivated states was not attempted.

2. I have concerns on different holding potentials being used in determining IC50 of NBI-921352 for different isoforms. The authors responded that "We chose to focus on what we refer to as "molecular selectivity," the fundamental ability for a compound to bind to the channel and stabilize the high affinity conformation. We accomplish this by choosing voltages that promote the same fraction of channels to be in the high affinity (inactivated) state." I wonder how the authors determined that "-45 mV" is the potential for NBI-921352 to bind and reach the high affinity conformation of Nav1.1, 1.2, 1.3, 1.4, and 1.6, and "-60 mV" for Nav1.7 and Nav1.5, there is no justification of this in the manuscript. In addition, when a medicine is applied to a patient, the adverse effect is determined by its affinity and effect on all targeted channels/molecules physiological or pathological conditions, therefore, the relative selectivity of NBI-921352 determined using the same protocol/condition would be more meaningful for clinical practice.

The new data that NBI-921352 showed similar potency in Nav1.6 channel at -62 mV when compared to that at -45 mV is important, this data suggests that the IC50 of NBI-921352 is not significantly changed by the holding potential (or membrane potential) between -62 mV and -45 mV. Although I don't agree with the authors' explanation, I will not ask for extra experiments. However, unless the authors provide convincing evidence that the two holding potentials (-45 mV and -60 mV) used in this study do allow different isoforms to be in their "highest affinity state (inactivated)", the explanation in Page 6 (line 124 to line 138) needs to be modified/removed.

We were fortunate that the voltage dependence of gating for the channels most relevant to our seizure studies (the CNS channels) aligned well enough to enable a common voltage protocol for all 4 CNS NaV channel isoforms. The conclusions on the advantage of selective inhibition of NaV1.6 for treating epilepsy are most dependent upon this selectivity profile and for these isoforms molecular selectivity and functional selectivity coincide.

As indicated in the response to question 3, NBI-921352 prefers inactivated channels. To create a robust assay suitable for screening compounds (about 2000) over the course of several years, we sought to create assays that were highly robust and reproducible over time. We have found that this is sometimes not possible with a single voltage protocol for every channel of interest.

To achieve robust and sensitive assays, we hold the membrane potential positive enough to ensure that most channels are inactivated. We then briefly step to a hyperpolarized potential (150 mV) that allows fast inactivated channels to recover but doesn’t allow for recovery from slow inactivation. If drug binding stabilizes the inactivated state sufficiently then drug bound channels also do not recover – leading to current inhibition. Holding at potentials that are too depolarized leads to an accumulation of slow inactivated channels and causes rundown. This reduces the size of the current signal and increases the challenge of separating gating drift from compound inhibition. These issues can compromise the fidelity and reproducibility of the inhibition assay. We empirically determined that the voltages used here result in full inactivation and are stable enough to enable robust measures over the duration of the assay.

For most isoforms (NaV1.1, NaV1.2, NaV1.3, NaV1.4, and NaV1.6) -45 mV was found to have complete inactivation and a stable recovered current amplitude. The isoforms that inactivate at the most negative potentials (NaV1.5 and NaV1.7) needed a more negative holding potential (60 mV) to preserve robust assay performance while fully inactivating the channels.

While unlikely to impact efficacy, NaV1.5 inhibition is important from a safety perspective as inhibition of NaV1.5 would introduce risk of cardiac events. Cardiac cells generally have more negative resting membrane potentials (around -90 mV) than neurons so inhibition in cardiac tissue will likely be less potent than inhibition in our assay.

We have removed the section (lines 124-138) indicated by the reviewer and instead added more text on protocols to the beginning of the section, “NBI-921352 is a state-dependent inhibitor” to improve the clarity of our methods and rationale for the readers.

Typo: line 137, "is this assay" should be "in this assay".

This typo was removed.

4. The figure legend 5 is confusing (page 63, line 1159-1169), please rephrase the sentences.

We have expanded the legend for Figure 5 to clarify.

https://doi.org/10.7554/eLife.72468.sa2

Article and author information

Author details

  1. JP Johnson

    In Vitro Biology, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Conceptualization, Data curation, Formal analysis, Methodology, Project administration, Supervision, Visualization, Writing – original draft, Writing – review and editing
    For correspondence
    jpjohnson@xenon-pharma.com
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-5762-5138

  2. Thilo Focken

    Chemistry, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Conceptualization, Data curation, Formal analysis, Supervision, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  3. Kuldip Khakh

    In Vitro Biology, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Supervision, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  4. Parisa Karimi Tari

    In Vivo Biology, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    was previously employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  5. Celine Dube

    In Vivo Biology, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  6. Samuel J Goodchild

    In Vitro Biology, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  7. Jean-Christophe Andrez

    Chemistry, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  8. Girish Bankar

    In Vivo Biology, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  9. David Bogucki

    1. Chemistry, Xenon Pharmaceuticals Inc, Burnaby, Canada
    2. Compound Properties, Xenon Pharmaceuticals Inc, Burnaby BC, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    was previously employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  10. Kristen Burford

    Chemistry, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  11. Elaine Chang

    In Vitro Biology, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  12. Sultan Chowdhury

    Chemistry, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  13. Richard Dean

    In Vitro Biology, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  14. Gina de Boer

    Compound Properties, Xenon Pharmaceuticals Inc, Burnaby BC, Canada
    Contribution
    Data curation, Formal analysis, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  15. Shannon Decker

    Chemistry, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Project administration, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  16. Christoph Dehnhardt

    Chemistry, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  17. Mandy Feng

    In Vitro Biology, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  18. Wei Gong

    Chemistry, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  19. Michael Grimwood

    Chemistry, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  20. Abid Hasan

    Chemistry, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  21. Angela Hussainkhel

    In Vitro Biology, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  22. Qi Jia

    Chemistry, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  23. Stephanie Lee

    Compound Properties, Xenon Pharmaceuticals Inc, Burnaby BC, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  24. Jenny Li

    In Vitro Biology, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  25. Sophia Lin

    In Vitro Biology, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  26. Andrea Lindgren

    Compound Properties, Xenon Pharmaceuticals Inc, Burnaby BC, Canada
    Contribution
    Data curation, Formal analysis, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  27. Verner Lofstrand

    Chemistry, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  28. Janette Mezeyova

    In Vitro Biology, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  29. Rostam Namdari

    Translational Drug Development, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  30. Karen Nelkenbrecher

    In Vivo Biology, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  31. Noah Gregory Shuart

    In Vitro Biology, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  32. Luis Sojo

    Compound Properties, Xenon Pharmaceuticals Inc, Burnaby BC, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  33. Shaoyi Sun

    Chemistry, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  34. Matthew Taron

    Chemistry, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  35. Matthew Waldbrook

    In Vivo Biology, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  36. Diana Weeratunge

    In Vitro Biology, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  37. Steven Wesolowski

    Chemistry, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Supervision, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  38. Aaron Williams

    In Vitro Biology, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Supervision, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  39. Michael Wilson

    Chemistry, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  40. Zhiwei Xie

    In Vitro Biology, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  41. Rhena Yoo

    In Vitro Biology, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  42. Clint Young

    In Vitro Biology, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  43. Alla Zenova

    Chemistry, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  44. Wei Zhang

    Chemistry, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Methodology, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  45. Alison J Cutts

    Scientific Affairs, Xenon Pharmaceuticals, Inc, Burnaby BC, Canada
    Contribution
    Data curation, Formal analysis, Project administration, Writing – original draft, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-4986-573X

  46. Robin P Sherrington

    Executive Team, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Conceptualization, Funding acquisition, Project administration, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  47. Simon N Pimstone

    Executive Team, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Funding acquisition, Project administration, Supervision, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  48. Raymond Winquist

    Executive Team, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Data curation, Formal analysis, Project administration, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  49. Charles J Cohen

    Executive Team, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Conceptualization, Funding acquisition, Project administration, Supervision, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

  50. James R Empfield

    Executive Team, Xenon Pharmaceuticals Inc, Burnaby, Canada
    Contribution
    Conceptualization, Funding acquisition, Project administration, Supervision, Writing – review and editing
    Competing interests
    is employed by Xenon Pharmaceuticals, Inc and may hold equity in Xenon Pharmaceuticals, Inc

Funding

Xenon Pharmaceuticals, Inc

  • JP Johnson
  • Thilo Focken
  • Kuldip Khakh
  • Parisa Karimi Tari
  • Celine Dube
  • Samuel J Goodchild
  • Jean-Christophe Andrez
  • Girish Bankar
  • David Bogucki
  • Kristen Burford
  • Elaine Chang
  • Sultan Chowdhury
  • Richard Dean
  • Gina de Boer
  • Shannon Decker
  • Christoph Dehnhardt
  • Mandy Feng
  • Wei Gong
  • Michael Grimwood
  • Abid Hasan
  • Angela Hussainkhel
  • Qi Jia
  • Stephanie Lee
  • Jenny Li
  • Sophia Lin
  • Andrea Lindgren
  • Verner Lofstrand
  • Janette Mezeyova
  • Rostam Namdari
  • Karen Nelkenbrecher
  • Noah Gregory Shuart
  • Luis Sojo
  • Shaoyi Sun
  • Matthew Taron
  • Matthew Waldbrook
  • Diana Weeratunge
  • Steven Wesolowski
  • Aaron Williams
  • Michael Wilson
  • Zhiwei Xie
  • Rhena Yoo
  • Clint Young
  • Alla Zenova
  • Wei Zhang
  • Alison J Cutts
  • Robin P Sherrington
  • Simon N Pimstone
  • Raymond Winquist
  • Charles J Cohen
  • James R Empfield

All of this work was funded by Xenon Pharmaceuticals, and all of the authors are, or were previously, employees of Xenon Pharmaceuticals.

Acknowledgements

We thank Dr. Fiona Scott, PhD of Neurocrine Biosciences for thoughtful discussions and commentary on this manuscript. We thank Ian Mortimer of Xenon Pharmaceuticals for strategic discussions. We thank Dr. Miriam Meisler, PhD for making Scn8aN1768D/+ mice available. We thank the SCN8A-RES patients, families and patient advocates for detailed discussions regarding the clinical presentation of SCN8A mutations and unmet medical needs in their community.

Ethics

All animal research was overseen by the Xenon Animal Care Committee and the Canadian Animal Care Council (CACC) according the recommendations of the CACC (https://ccac.ca/).

Senior Editor

  1. Richard W Aldrich, The University of Texas at Austin, United States

Reviewing Editor

  1. Baron Chanda, Washington University in St. Louis, United States

Reviewer

  1. Theodore Cummins

Version history

  1. Received: July 24, 2021
  2. Preprint posted: August 31, 2021 (view preprint)
  3. Accepted: February 23, 2022
  4. Accepted Manuscript published: March 2, 2022 (version 1)
  5. Version of Record published: March 8, 2022 (version 2)

Copyright

© 2022, Johnson et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 4,487
    Page views
  • 620
    Downloads
  • 15
    Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Open citations (links to open the citations from this article in various online reference manager services)

Cite this article (links to download the citations from this article in formats compatible with various reference manager tools)

  1. JP Johnson
  2. Thilo Focken
  3. Kuldip Khakh
  4. Parisa Karimi Tari
  5. Celine Dube
  6. Samuel J Goodchild
  7. Jean-Christophe Andrez
  8. Girish Bankar
  9. David Bogucki
  10. Kristen Burford
  11. Elaine Chang
  12. Sultan Chowdhury
  13. Richard Dean
  14. Gina de Boer
  15. Shannon Decker
  16. Christoph Dehnhardt
  17. Mandy Feng
  18. Wei Gong
  19. Michael Grimwood
  20. Abid Hasan
  21. Angela Hussainkhel
  22. Qi Jia
  23. Stephanie Lee
  24. Jenny Li
  25. Sophia Lin
  26. Andrea Lindgren
  27. Verner Lofstrand
  28. Janette Mezeyova
  29. Rostam Namdari
  30. Karen Nelkenbrecher
  31. Noah Gregory Shuart
  32. Luis Sojo
  33. Shaoyi Sun
  34. Matthew Taron
  35. Matthew Waldbrook
  36. Diana Weeratunge
  37. Steven Wesolowski
  38. Aaron Williams
  39. Michael Wilson
  40. Zhiwei Xie
  41. Rhena Yoo
  42. Clint Young
  43. Alla Zenova
  44. Wei Zhang
  45. Alison J Cutts
  46. Robin P Sherrington
  47. Simon N Pimstone
  48. Raymond Winquist
  49. Charles J Cohen
  50. James R Empfield
(2022)
NBI-921352, a first-in-class, NaV1.6 selective, sodium channel inhibitor that prevents seizures in Scn8a gain-of-function mice, and wild-type mice and rats
eLife 11:e72468.
https://doi.org/10.7554/eLife.72468

Categories and tags

Research organisms

Further reading

    1. Neuroscience

    Event-related modulation of alpha rhythm explains the auditory P300-evoked response in EEG

    Alina Studenova, Carina Forster ... Vadim Nikulin
    Research Article

    Evoked responses and oscillations represent two major electrophysiological phenomena in the human brain yet the link between them remains rather obscure. Here we show how most frequently studied EEG signals: the P300-evoked response and alpha oscillations (8–12 Hz) can be linked with the baseline-shift mechanism. This mechanism states that oscillations generate evoked responses if oscillations have a non-zero mean and their amplitude is modulated by the stimulus. Therefore, the following predictions should hold: (1) the temporal evolution of P300 and alpha amplitude is similar, (2) spatial localisations of the P300 and alpha amplitude modulation overlap, (3) oscillations are non-zero mean, (4) P300 and alpha amplitude correlate with cognitive scores in a similar fashion. To validate these predictions, we analysed the data set of elderly participants (N=2230, 60–82 years old), using (a) resting-state EEG recordings to quantify the mean of oscillations, (b) the event-related data, to extract parameters of P300 and alpha rhythm amplitude envelope. We showed that P300 is indeed linked to alpha rhythm, according to all four predictions. Our results provide an unifying view on the interdependency of evoked responses and neuronal oscillations and suggest that P300, at least partly, is generated by the modulation of alpha oscillations.

    1. Neuroscience

    Catecholaminergic neuromodulation and selective attention jointly shape perceptual decision-making

    Stijn A Nuiten, Jan Willem de Gee ... Simon van Gaal
    Research Article

    Perceptual decisions about sensory input are influenced by fluctuations in ongoing neural activity, most prominently driven by attention and neuromodulator systems. It is currently unknown if neuromodulator activity and attention differentially modulate perceptual decision-making and/or whether neuromodulatory systems in fact control attentional processes. To investigate the effects of two distinct neuromodulatory systems and spatial attention on perceptual decisions, we pharmacologically elevated cholinergic (through donepezil) and catecholaminergic (through atomoxetine) levels in humans performing a visuo-spatial attention task, while we measured electroencephalography (EEG). Both attention and catecholaminergic enhancement improved decision-making at the behavioral and algorithmic level, as reflected in increased perceptual sensitivity and the modulation of the drift rate parameter derived from drift diffusion modeling. Univariate analyses of EEG data time-locked to the attentional cue, the target stimulus, and the motor response further revealed that attention and catecholaminergic enhancement both modulated pre-stimulus cortical excitability, cue- and stimulus-evoked sensory activity, as well as parietal evidence accumulation signals. Interestingly, we observed both similar, unique, and interactive effects of attention and catecholaminergic neuromodulation on these behavioral, algorithmic, and neural markers of the decision-making process. Thereby, this study reveals an intricate relationship between attentional and catecholaminergic systems and advances our understanding about how these systems jointly shape various stages of perceptual decision-making.

    1. Neuroscience

    Two human brain systems micro-structurally associated with obesity

    Manfred G Kitzbichler, Daniel Martins ... Neil A Harrison
    Research Article Updated

    The relationship between obesity and human brain structure is incompletely understood. Using diffusion-weighted MRI from ∼30,000 UK Biobank participants, we test the hypothesis that obesity (waist-to-hip ratio, WHR) is associated with regional differences in two micro-structural MRI metrics: isotropic volume fraction (ISOVF), an index of free water, and intra-cellular volume fraction (ICVF), an index of neurite density. We observed significant associations with obesity in two coupled but distinct brain systems: a prefrontal/temporal/striatal system associated with ISOVF and a medial temporal/occipital/striatal system associated with ICVF. The ISOVF~WHR system colocated with expression of genes enriched for innate immune functions, decreased glial density, and high mu opioid (MOR) and other neurotransmitter receptor density. Conversely, the ICVF~WHR system co-located with expression of genes enriched for G-protein coupled receptors and decreased density of MOR and other receptors. To test whether these distinct brain phenotypes might differ in terms of their underlying shared genetics or relationship to maps of the inflammatory marker C-reactive Protein (CRP), we estimated the genetic correlations between WHR and ISOVF (rg = 0.026, P = 0.36) and ICVF (rg = 0.112, P < 9×10−4) as well as comparing correlations between WHR maps and equivalent CRP maps for ISOVF and ICVF (P<0.05). These correlational results are consistent with a two-way mechanistic model whereby genetically determined differences in neurite density in the medial temporal system may contribute to obesity, whereas water content in the prefrontal system could reflect a consequence of obesity mediated by innate immune system activation.