Differential roles of NaV1.2 and NaV1.6 in regulating neuronal excitability at febrile temperature and distinct contributions to febrile seizures

Dysregulation of voltage-gated sodium channels (VGSCs) is associated with multiple clinical disorders, including febrile seizures (FS). The contribution of different sodium channel subtypes to environmentally triggered seizures is not well understood. Here we demonstrate that somatic and axonal sodium channels primarily mediated through NaV1.2 and NaV1.6 subtypes, respectively, behave differentially at FT, and might play distinct roles in FS generation. In contrast to sodium channels on the main axonal trunk, somatic ones are more resistant to inactivation and display significantly augmented currents, faster gating rates and kinetics of recovery from inactivation at FT, features that promote neuronal excitabilities. Pharmacological inhibition of NaV1.2 by Phrixotoxin-3 (PTx3) suppressed FT-induced neuronal hyperexcitability in brain slice, while up-regulation of NaV1.2 as in NaV1.6 knockout mice showed an opposite effect. Consistently, NaV1.6 knockout mice were more susceptible to FS, exhibiting much lower temperature threshold and shorter onset latency than wildtype mice. Neuron modeling further suggests that NaV1.2 is the major subtype mediating FT-induced neuronal hyperexcitability, and predicts potential outcomes of alterations in sodium channel subtype composition. Together, these data reveal a role of native NaV1.2 on neuronal excitability at FT and its important contribution to FS pathogenesis.


Ethics statement
All procedures involving animals followed the protocols approved by the Animal Research Advisory Committee at the Shanghai Institutes of Biological Sciences, and accorded with the guidelines for the care and use of laboratory animals approved by School of Brain and Cognitive Sciences, Beijing Normal University. All possible efforts were made to minimize the number and suffering of animals used in this study.

Animals
To accord with a general analogy of rodent model to the first year and toddler years of human life 1 , age P13-P17 of both Sprague-Dawley (SD) rat and C3Fe.Cg-Scn8a med/+ J strain mice were used in this study.
Initial electrophysiology studies on different sodium channel subtypes were carried out on brain slices from SD rat ( Fig.1-4). C3Fe strain mice were used for behavior studies and electrophysiology ( Fig. 5-7).
We made somatic nucleated patch recordings for currents primarily contributed from NaV1.2 channels and isolated axonal bleb recordings for currents mainly mediated by NaV1.6 2,3 . For whole-cell recordings or nucleated patch recordings, low resistance pipettes (4-6 M ) were made from borosilicate glass tubing using Sutter P97 (Sutter Instrument, USA). For axonal recordings, high resistance pipettes (7-9 M ) were used. A positive air pressure was applied before lowering the pipette into the bath solution and maintained during approaching the neurons of interest. When a dent on the surface of a neuron or an axonal bleb was formed, the air pressure was released to form Giga seal instantly. Pulses of negative air pressure was applied to form whole cell or axonal bleb recording configuration. To achieve nucleated patch recording 3 configuration, negative air pressure was applied and maintained after forming whole cell recording configuration to attract cell nucleus to the pipette tip. Patch pipette was gently retracted out from the brain tissue. Giga seal was maintained during the whole process of making a nucleated patch. The access resistance was less than 20 M for whole cell recordings, and around 25 M for axonal bleb recordings and nucleated patch recordings.
To obtain purely Na + currents, we added 20 TEA, 3  The protocols for voltage-dependent activation/inactivation, kinetics of recovery from inactivation and onset inactivation of sodium channels are referred to suppl. Fig. 2, 3. The normalized G-V curves were averaged and fitted by a Boltzmann equation, where G is the conductance, Gmax is the maximal conductance, V50 is the voltage at which sodium channels are half activated or inactivated, and K is the slope factor. The time constants were determined by single-exponential fits.
Data were low-pass filtered at 10 kHz and sampled at 100 kHz using MultiClamp 700B and Digidata 1440A with pClamp 10.2 (Molecular Devices, CA, USA) or Micro1401 with Spike2 software (CED, Cambridge, UK). Leak currents were online subtracted by a P/4 protocol. Only those recordings with resting leak currents no greater than -80 pA were used for data analysis. All presented membrane potentials (Vm) in voltage clamp recordings were not corrected for liquid junction potentials (2.5 mV).
Temperature of brain slices were monitored by a closely placed thermistor and controlleVmd by a dual automatic temperature controller (TC-344B, Warner Instr. Inc.).

Immunostaining
A detailed procedure for immunostaining was described previously 2,4,5 . Briefly, prefrontal cortical brain tissues were cut into 15µm thick sections on cryostat. Slices were incubated in a blocking solution (5% normal goat serum, 0.3% Triton X-100 in PBS) at room temperature for 2 h, followed by overnight incubation in primary antibodies at 4 °C. The primary antibodies used in this study are mouse anti-NaV1.1

Neuron modeling
To evaluate the contribution of sodium channel subtypes to neuronal excitability at FT, we constructed a single compartment model and a realistic model using Neuron 7.2 6 . Single compartment model (10 μm diameter, 190.98 μm length, 1 compartment, 6000 μm 2 surface area) was implanted with NaV1.2 (gbar_na12 = 20 pS/μm 2 ), NaV1.6 (gbar_na16 = 50 pS/μm 2 ), KV (gbar_KV = 30 pS/μm 2 ) channels and a passive conductance (g_pas = 1 pS/μm 2 ). The membrane capacitance is 0.75 μF/cm 2 . For evaluating the 5 components of sodium conductance underlying backpropagating APs, AP waveforms experimentally recorded at different temperatures were assigned to an expanded SEClamp6 to measure the underlying sodium channel subtypes' conductance. Trapezoidal numerical integration method was used to calculate the area of sodium conductance. To evaluate contributions of different sodium channel subtypes to neuronal excitability at different temperatures, we designed channel subtype replacement experiments.
The initial density of each sodium channel subtype was set at 200 pS/μm 2 . The total sodium channel density was kept constant. To simulate knockout of one channel subtype with a compensatory upregulation of the other channel subtype, if the density of one channel subtype was set to 0 pS/μm 2 , the other one was correspondingly increased to 400 pS/μm 2 . To evaluate the effect of channel gating time constants or gating rate on neuronal excitability, we set temperature at 36.5C, then alter the state variable time constants (τm, τh) or channel open/close rate constants.
HCN channels have an exponential distribution from dendrites to the soma, and are absent in the axon 8 .
Ion channel mechanisms of NaV1.2 and NaV1.6 were revised from previous models 7 . Sodium currents (INa) were described by: where gNa is the local sodium conductance density, m and h are the activation and inactivation state dependent variables, V is the local Vm, ENa is the reversal potential for Na + ion (ENa = 70 mV, measured: Ionic mechanisms for voltage-gated non-inactivating potassium channels (KV), muscarinic activated slow non-inactivating potassium channels (KM), calcium-dependent potassium channels (KCa), voltagedependence HCN channels and high-voltage activated calcium channels were adopted from published models 7,8 . The integration time step in our simulation is 0.025 ms. Based on our data, the Q10s for current 7 amplitudes of NaV1.2 and NaV1.6 were set as 2.04 and 1.50, respectively, if temperature is higher than PT; and 1.36 and 1.16 when temperature lower than PT.

Behavior analysis of febrile seizures
Mice were placed in a 2L flask in a WP-25A electrothermal incubator (Taisite Instrument Co., LTD, Tianjin, China) prewarmed at 42.5 ± 1.0 °C for 30 min before returning to home cage. Both the core (rectal) and ambient temperatures were monitored in real-time. Seizure responses were videoed and scored based on a modified Racine scales which include 5 stages. Seizure scores were judged independently by two persons and discussed with third person to reach a consensus on scoring. We adopted the following formula 9 for evaluating seizure severity.
(Eq. 17) FS temperature threshold and latency refer to the point when the mice developed stage 4 phenotypes that can be unambiguously defined.

Statistic Analysis
Data were processed in GraphPad Prism5.0 (GraphPad Software Inc. San Diego, CA) and Matlab 2011b (MathWorks, Natick, MA) software. Data were presented as mean ± s.e.m., and compared by one-way ANOVA with post-hoc Bonferroni test or student t-test as indicated. The n value represents the number of neurons from at least three mice. The p values less than 0.05 were considered statistically significant.