Phenotypic Mutation 'Possum' (pdf version)
Mutation Type missense
Coordinate119,638,705 bp (GRCm38)
Base Change T ⇒ C (forward strand)
Gene Scn10a
Gene Name sodium channel, voltage-gated, type X, alpha
Synonym(s) Nav1.8
Chromosomal Location 119,608,456-119,719,032 bp (-)
MGI Phenotype Homozygotes for a targeted null mutation exhibit impaired perception of pain.
Accession Number

NCBI RefSeq: NM_009134; MGI: 108029

Mapped Yes 
Amino Acid Change Threonine changed to Alanine
Institutional SourceBeutler Lab
Ref Sequences
T790A in NCBI: NP_033160.2 (fasta)
Gene Model not available
SMART Domains

Pfam:Ion_trans 155 395 2.9e-71 PFAM
coiled coil region 398 434 N/A INTRINSIC
low complexity region 557 572 N/A INTRINSIC
Pfam:Ion_trans 698 888 1.4e-38 PFAM
Pfam:Na_trans_assoc 903 1162 2.2e-65 PFAM
Pfam:Ion_trans 1188 1417 4.6e-52 PFAM
PDB:1BYY|A 1419 1471 N/A PDB
Pfam:Ion_trans 1510 1721 2.2e-46 PFAM
Pfam:PKD_channel 1559 1728 4e-9 PFAM
IQ 1850 1872 7.57e0 SMART
Predicted Effect probably damaging

PolyPhen 2 Score 1.000 (Sensitivity: 0.00; Specificity: 1.00)
(Using NCBI: NP_033160.2)
Phenotypic Category behavior/neurological, nervous system
Penetrance 100% 
Alleles Listed at MGI

All alleles(5) : Targeted(5)

Lab Alleles
AlleleSourceChrCoordTypePredicted EffectPPH Score
IGL00088:Scn10a APN 9 119672226 missense probably damaging 1.00
IGL01339:Scn10a APN 9 119622766 unclassified probably damaging 1.00
IGL01467:Scn10a APN 9 119658412 missense probably benign 0.33
IGL01472:Scn10a APN 9 119617763 missense probably damaging 1.00
IGL01481:Scn10a APN 9 119609194 missense possibly damaging 0.71
IGL01539:Scn10a APN 9 119638698 missense possibly damaging 0.95
IGL01580:Scn10a APN 9 119627159 missense probably damaging 1.00
IGL01676:Scn10a APN 9 119672165 nonsense probably null
IGL01681:Scn10a APN 9 119694077 missense probably damaging 1.00
IGL01748:Scn10a APN 9 119627084 missense probably damaging 1.00
IGL01866:Scn10a APN 9 119635502 nonsense probably null
IGL01998:Scn10a APN 9 119609676 missense probably damaging 1.00
IGL02015:Scn10a APN 9 119664951 unclassified probably benign 0.08
IGL02098:Scn10a APN 9 119691478 missense possibly damaging 0.90
IGL02113:Scn10a APN 9 119609890 missense probably damaging 1.00
IGL02245:Scn10a APN 9 119672152 missense probably damaging 1.00
IGL02262:Scn10a APN 9 119658433 missense probably benign 0.07
IGL02317:Scn10a APN 9 119638555 missense probably benign 0.00
IGL02428:Scn10a APN 9 119691562 unclassified probably damaging 1.00
IGL02439:Scn10a APN 9 119618848 missense probably benign 0.40
IGL02583:Scn10a APN 9 119691440 unclassified 0.00
IGL02597:Scn10a APN 9 119610123 missense probably damaging 0.99
IGL02680:Scn10a APN 9 119666059 missense probably damaging 1.00
IGL02733:Scn10a APN 9 119616705 missense probably damaging 1.00
IGL02851:Scn10a APN 9 119671608 missense probably damaging 1.00
IGL02992:Scn10a APN 9 119609560 missense possibly damaging 0.90
IGL03040:Scn10a APN 9 119622985 missense probably damaging 1.00
IGL03049:Scn10a APN 9 119665990 missense probably damaging 1.00
IGL03407:Scn10a APN 9 119648171 missense probably damaging 0.99
R0025:Scn10a UTSW 9 119670484 missense probably damaging 1.00
R0030:Scn10a UTSW 9 119669990 missense probably benign 0.00
R0328:Scn10a UTSW 9 119694102 missense possibly damaging 0.92
R0494:Scn10a UTSW 9 119624100 missense probably damaging 0.99
R0511:Scn10a UTSW 9 119613700 missense probably damaging 0.99
R0548:Scn10a UTSW 9 119665928 missense probably benign 0.00
R0584:Scn10a UTSW 9 119670531 missense probably damaging 1.00
R0595:Scn10a UTSW 9 119666063 missense probably benign 0.01
R0894:Scn10a UTSW 9 119630147 missense probably damaging 1.00
R1022:Scn10a UTSW 9 119609274 missense probably damaging 1.00
R1024:Scn10a UTSW 9 119609274 missense probably damaging 1.00
R1263:Scn10a UTSW 9 119617733 missense probably damaging 1.00
R1456:Scn10a UTSW 9 119691478 missense probably benign 0.01
R1466:Scn10a UTSW 9 119666490 missense probably damaging 1.00
R1466:Scn10a UTSW 9 119666490 missense probably damaging 1.00
R1573:Scn10a UTSW 9 119613626 missense probably benign
R1704:Scn10a UTSW 9 119609394 missense probably damaging 1.00
R1933:Scn10a UTSW 9 119609998 missense probably benign 0.43
R1945:Scn10a UTSW 9 119691454 missense possibly damaging 0.89
R2013:Scn10a UTSW 9 119613736 missense probably benign 0.21
R2155:Scn10a UTSW 9 119609448 missense probably benign 0.00
R2196:Scn10a UTSW 9 119609004 missense probably benign
R2231:Scn10a UTSW 9 119633850 missense probably benign 0.00
R2353:Scn10a UTSW 9 119638687 missense possibly damaging 0.79
R2392:Scn10a UTSW 9 119627202 missense possibly damaging 0.86
R2895:Scn10a UTSW 9 119661401 missense probably benign 0.00
R2926:Scn10a UTSW 9 119638701 missense possibly damaging 0.74
R3783:Scn10a UTSW 9 119691562 missense probably damaging 1.00
R3821:Scn10a UTSW 9 119638633 missense probably benign
R4003:Scn10a UTSW 9 119608968 missense probably benign 0.00
R4208:Scn10a UTSW 9 119616776 missense probably damaging 0.99
R4231:Scn10a UTSW 9 119631544 missense possibly damaging 0.86
R4626:Scn10a UTSW 9 119631505 missense probably benign 0.35
R4702:Scn10a UTSW 9 119633791 missense probably benign 0.20
R4713:Scn10a UTSW 9 119609651 missense probably damaging 1.00
R4729:Scn10a UTSW 9 119671526 missense probably damaging 1.00
R4782:Scn10a UTSW 9 119622910 missense possibly damaging 0.70
R4822:Scn10a UTSW 9 119638672 missense probably damaging 1.00
R4856:Scn10a UTSW 9 119694309 missense probably benign 0.12
R4856:Scn10a UTSW 9 119694310 missense possibly damaging 0.49
R4932:Scn10a UTSW 9 119687874 splice site probably null
R5015:Scn10a UTSW 9 119622921 missense possibly damaging 0.93
R5193:Scn10a UTSW 9 119609655 missense probably damaging 1.00
R5211:Scn10a UTSW 9 119661232 missense probably benign 0.01
R5320:Scn10a UTSW 9 119648109 missense probably damaging 1.00
R5400:Scn10a UTSW 9 119609034 missense probably benign 0.04
R5448:Scn10a UTSW 9 119687947 missense probably benign 0.24
R5457:Scn10a UTSW 9 119694127 missense probably damaging 1.00
R5554:Scn10a UTSW 9 119694130 missense probably benign 0.01
R5680:Scn10a UTSW 9 119624136 missense probably damaging 1.00
R5762:Scn10a UTSW 9 119635441 critical splice donor site probably null
R5935:Scn10a UTSW 9 119627171 missense probably damaging 0.98
R5956:Scn10a UTSW 9 119631560 missense probably damaging 1.00
R6041:Scn10a UTSW 9 119609469 missense probably damaging 1.00
R6047:Scn10a UTSW 9 119622831 missense probably benign 0.20
R6132:Scn10a UTSW 9 119613695 missense possibly damaging 0.91
R6156:Scn10a UTSW 9 119635583 missense probably benign
X0058:Scn10a UTSW 9 119609364 nonsense probably null
Mode of Inheritance Autosomal Dominant
Local Stock Embryos, Sperm, gDNA
MMRRC Submission 030627-UCD
Last Updated 2016-05-13 3:09 PM by Stephen Lyon
Record Created unknown
Record Posted 2012-03-02
Phenotypic Description
Figure 1. Heightened sensitivity to cold pain in Possum mice. (A) Number of hindpaw lifting or jumping responses of mice placed on a -1°C metal plate for 2 min.  N≥6 per genotype including male and female littermates. (B) Time interval between contact with a 52°C metal plate and first response (jumping or hindpaw lifting). (C) Mechanical thresholds eliciting paw withdrawal determined with the von Frey test before and after injection of complete Freund’s adjuvant (CFA).  B and C, N≥5 per genotype of males and females. Psm, Possum.  *P≤0.05, one-way ANOVA.  Error bars represent s.e.m.  Obtained from (1).

Figure 2. Normal conditioned fear responses in Possum mice. On day 1, mice were habituated to the apparatus and then subjected to fear conditioning by exposure to the  context and conditioned stimulus (30-s period of light and a tone of 3000 Hz, 80 dB) in association with foot shock during the last 2 seconds (0.50 mA scrambled current). On day 2, contextual conditioning (determined by freezing behavior) was measured in a 5 minute test in the chamber where the mice were trained. Twenty four hours later, on day 3, cued conditioning was measured by habituating mice to a novel context for 3 minutes, after which they were exposed to the conditioned stimulus (light+tone) for 3 minutes. (A) Average duration of freezing of homozygous Possum (N=10 males, 10 females) and wild type mice (N=10 males, 10 females) during conditioning to associate environmental context or cue with foot shock.  The conditioning protocol is shown below the X axis.  CS, cue presentation 30 s. Asterisk, foot shock 2 s. (B) The percentage of time frozen during each interval of the conditioning trial shown in (A). (C) Sum total of the average duration of each instance of freezing in test sessions before fear conditioning or during context exposure after fear conditioning. (D) Following fear conditioning and in a novel environment, sum total of the average duration of each instance of freezing in test sessions during habituation (pre-cue) or cue presentation.  Psm, Possum.  Error bars represent s.e.m.  P values (one-way ANOVA) are indicated.  ns, not significant. Obtained from (1).

Figure 3. Shift in EEG from θ rhythm to δ rhythm upon initiation of tonic immobility in Possum mice. Average parietal cortical EEG power recorded from homozygous Possum (N=8) or wild-type mice (N=8) during the indicated stages.  Stages were consecutive periods of baseline recording, immobility episode 1 induced by scruffing, recovery period, immobility episode 2 induced by scruffing, and a second recovery period.  The baseline stage and recovery periods were a minimum of 5 min and immobility stages were delineated according to visual inspection of the mouse.   During immobility stages wild-type mice did not display tonic immobility and continued to move normally.  *P≤0.05, **P≤0.01, ***P≤0.001, one-way ANOVA.  Error bars represent s.e.m. Obtained from (1).

Figure 4. Alertness of Possum mice during immobility. (A) Grand average ERPs to standard tones recorded from Possum mice (N=8) at frontal and parietal electrodes before (baseline) or during tonic immobility induced by scruffing.  Time 0 corresponds to the time the auditory stimulus was given. (B) Mean amplitude of P1, N1a, N1b, P2, N2, and P3 peaks of ERPs to standard, rare, or noise tones recorded from the parietal or frontal cortices of Possum mice (N=8) during a baseline period (before auditory stimulation) or during immobility induced by scruffing.  *P≤0.05, **P≤0.01, ***P≤0.001, one-way ANOVA.  Error bars represent s.e.m. Obtained from (1).

Figure 5. Sinus bradycardia and irregular RR intervals during immobility of Possum mice. (A) Mean heart rate before and after scruffing of homozygous Possum mice left untreated or injected i.p. with atropine. Each point represents an individual mouse. (B) Representative ECG traces from a homozygous Possum mouse. Upper panel demonstrates normal sinus rhythm with P-waves (arrows) preceding the QRS complexes. The middle panel shows an ECG pattern consistent with sinus bradycardia and RR interval variability during immobility induced by scruffing. All homozygous Possum mice tested (N=12) had a bradycardic response upon scruffing. Lower panel shows a normal ECG pattern in a scruffed Possum mouse pre-treated with atropine. Tick marks, 0.1 s. (C) Mean systolic blood pressure before and after scruffing in Possum homozygotes. Each point represents an individual mouse. Possum mice exhibited tonic immobility after scruffing in all experiments shown. P values were determined by unpaired Student’s t test. Obtained from (1).

Figure 6. Nav1.8Possum current is enhanced compared to wild-type. (A-H) Families of whole cell currents in representative wild-type (A) and Scn10aPsm/Psm (B) small sensory neurons in culture. Currents were elicited by step depolarizations from -91.5 mV in 10 mV increments for 300 ms from a holding potential of -111.5 mV every 10 s (voltage protocol inset below). Peak currents were measured and plotted versus voltage (insets in A, B). (C) Wild-type currents elicited by a step to -11.5 mV are inactivated by 30 ms prepulse voltages in a voltage dependent manner. Prepulse voltage steps are followed by a step to -131.5 mV for 0.4 ms to close channels in the open state (protocol shown in inset). (D) Average peak inward current density for wild-type (N=28) and Possum homozygous DRG neurons (N=22).  ***P<10-4. (E) Population activation data (G/Gmax) (calculated taking into account the Vm-ENa driving force) for N=18 wild-type and N=10 Scn10aPsm/Psm cells is fit by a Boltzmann function to obtain V0.5,act and slope k (Table S1). (F) Population inactivation data (I/Imax) are fit by a Boltzmann function to obtain V0.5,inact and slope k (Table S1). (G) Percent of current remaining at the end of 150-300 ms test pulses to various indicated potentials. (H) Decay of currents elicited at the indicated membrane voltages were fit with 2 exponentials and the percentage of the slow component is shown as a function of voltage. Numbers of neurons are indicated. Error bars indicate s.e.m.  P values obtained using Student’s t-test.  Psm, Possum. Obtained from (1).

Table 1.Obtained from (1).

The Possum phenotype was identified in the G3 population of ENU-induced germline mutants.  Heterozygous and homozygous Possum mice arrested all movement, exhibited transient apnea (5-10 seconds), and assumed a rigid posture following pinching of the skin at the back of the neck (“scruffing”) (see video) (1).  Scruffed Possum mice were unable to right themselves when placed on their side or back and also displayed ‘waxy flexibility’ in which the tail could retain a manually set position.  The immobility response lasts for one to five minutes, after which they regain normal behavior and exhibit no discernible residual effects.  Heterozygotes displayed a slightly shorter response (~1 minute) than homozygotes.  Recovery from immobility could be promoted by auditory (e.g., clapping) or mechanical stimulation (e.g., gentle prodding), and after as little as 30 seconds, scruffing resulted in a repetition of the behavior.  The behavior recapitulates precisely the features of experimentally induced tonic immobility (2;3).  Habituation did not occur when mice were scruffed on a daily basis.  Possum homozygotes displayed normal levels of activity in tests of locomotion.


After identification of the Possum mutation within the voltage-gated sodium channel α subunit Nav1.8, pain responses of Possum mice to various stimuli were examined since Nav1.8 is expressed in dorsal root ganglia (DRG) (4-7) where it functions to propagate nociceptive signaling in response to noxious cold, heat, and mechanical stimuli (8-10).  Homozygous Possum mice displayed significantly increased sensitivity in the cold-plate test (Figure 1A). However, homozygous Possum and wild type mice exhibited similar latencies to paw withdrawal in the hot-plate test, demonstrating normal pain responses to heat stimuli (Figure 1B). Paw withdrawal responses elicited by von Frey filaments applied to the plantar surface did not differ between Possum and wild type mice, either before or after injection of complete Freund’s adjuvant into the paw (Figure 1C). None of the stimuli tested induced immobility in Possum mice.  In addition, treatment of neonatal Possum mice with capsaicin to ablate TRPV1-expressing nociceptive C fibers did not abrogate scruffing-induced immobility, indicating that the phenotype is not a response to pain sensed by capsaicin-susceptible sensory neurons. Neither sudden immersion in cold water nor pinching of the tail or feet induced immobility in Possum mice.


In tests of conditioned fear elicited by foot shock paired with visual and auditory cues, Possum homozygotes displayed similar or reduced fear responses relative to wild-type mice (Figure 2).  Foot shock alone did not induce the immobility phenotype.


Electroencephalographic (EEG) activity was measured in homozygous Possum mice before and after scruffing, and compared to that of wild type mice given the same treatment (Figure 3). During the baseline period before scruffing, frontal and parietal EEG patterns were similar in Possum and wild-type mice, with no epileptiform or other abnormal activity. Wild-type mice maintained an identical EEG pattern after scruffing. In contrast, scruffing of Possum homozygotes produced a shift in EEG activity from frequencies predominantly in the 4 to 10 Hz range (θ rhythm) during baseline recording to δ frequencies in the 1 to 4 Hz range concomitant with immobility. Coincident with recovery from immobility, EEG activity abruptly returned to and maintained baseline patterns indefinitely. Following a 5-min recovery period, scruffing again resulted in immobility and development of altered EEG activity similar to that observed during the first episode. These data implicate Nav1.8 in modulating the activity of CNS neurons. Measurements of evoked response potentials (ERP) to auditory stimuli in Possum mice suggested that although immobile, they are alert rather than unconscious (Figure 4).


Nav1.8 transcripts have been detected in the heart (11), and three studies reported association between common sequence variants of human SCN10A and prolongation of PR interval, P wave, and QRS complex on electrocardiogram (ECG) (11-13).  ECG recordings from wild type and homozygous Possum mice before scruffing revealed similar heart rates (Figure 5A). In contrast, the heart rates of Possum mice were reduced by approximately 50% during the immobile period after scruffing relative to baseline heart rates (Figure 5A). The ECGs recorded for wild type and homozygous Possum mice showed no significant differences in PR interval, P wave duration, or QRS duration prior to scruffing. However, scruffing of Possum mice resulted in irregular RR intervals following clearly defined P waves (Figure 5B), consistent with sinus bradycardia with increased heart rate variability coincident with the Possum response. No significant ECG changes were observed following scruffing of wild type mice. The changes in heart rate and rhythm resolved as the mice became ambulatory. Since bradycardia can cause hypotension and loss of consciousness, the blood pressure of Possum mice was measured, but no fluctuation in blood pressure following scruffing was detected, indicating that immobility is not associated with or caused by changes in blood pressure (Figure 5C).


Since increased vagus nerve discharge can decrease heart rate and cause heart rate variability, we tested whether the bradycardia and marked RR variability in Possum mice resulted from increased vagal tone. Possum homozygotes were treated prior to scruffing with atropine, an anti-cholinergic drug that inhibits the action of the vagus nerve on the heart. Atropine prevented the reduction in heart rate (Figure 5A) and normalized cardiac rhythm during immobility (Figure 5B), but failed to prevent the Possum immobility response itself. No ECG changes were observed with scruffing or atropine administration in wild-type mice. Thus, whereas excessive vagal output to the heart is responsible for bradycardia and RR variability in Possum mice, it is not responsible for the tonic immobility induced by scruffing. Neither epinephrine nor propranolol, administered at large pharmacologic doses, abolished Possum behavior or induced it in wild-type mice.

Electrophysiological properties of the Nav1.8Possum channel
To understand how the Possum mutation affects the function of Nav1.8, whole cell voltage and current clamp methods were used to characterize Nav1.8 currents and excitability of acutely cultured homozygous Possum and wild-type small DRG sensory neurons (1). The Nav1.8 current was isolated from those of other channels, including the other TTX-resistant channel Nav1.9, and the voltage dependence and channel kinetics were determined for the parameters described in Table 1.

TTX-resistant Nav1.8-like currents expressed in homozygous Possum DRG neurons were 4-fold larger than those of wild-type DRG (Figure 6B & D; Table 1) (1). A 3.6-fold increase in ramp-induced inward current was also observed (-180±45 pA/pF (N=6); P=0.034). During long duration steps to subpeak voltages inactivation of Nav1.8-like currents in Possum neurons was impaired compared to that in wild-type neurons (Figure 6G) due to a 1.6-fold slowing of both fast and slow components of fast inactivation (Table 1) and a 2-fold increase in the proportion of the slow component (Figure 6H). There was a significant slowing of activation as well (time to peak, Table 1). No differences between Possum and wild-type DRG neurons were observed for V0.5,act, V0.5,inact, k values, deactivation rates, or recovery from fast or slow inactivation (Figure 6E & F; Table 1) (1). Therefore, the Possum mutation increases the current mediated by Nav1.8 channels by increasing the apparent current density and proportion of slow:fast components and slowing the kinetics of fast inactivation. Since these changes likely heighten the excitability of Nav1.8-expressing neurons, we performed current clamp recordings in the presence of TTX to determine whether Possum neurons are hyperexcitable. Although no detectable differences in resting potential, action potential duration, and membrane resistance were measured at -60 mV, the amount of depolarizing current required to elicit an action potential (rheobase) was profoundly decreased in Possum compared to wild-type neurons as was the voltage threshold, and the number of action potentials elicited during 150 and 900 ms injection of current at twice the rheobase. Thus, Possum small DRG neurons with long duration action potentials are hyperexcitable compared to those of littermate controls (1).


Nature of Mutation
The Possum mutation was mapped to Chromosome 9, and corresponds to an A to G transition at position 2403 of the Scn10a transcript, in exon 15 of 27 total exons.
785  -A--L--G--N--L--T--F--I--L--A--I-
The mutated nucleotide is indicated in red lettering, and results in a threonine to alanine change at residue 790 of the Nav1.8 protein.
Protein Prediction
Figure 7. The SCN10A protein. A, Domain structure of SCN10A. B, Transmembrane organization of sodium channel subunits. Cylinders represent probable α-helical transmembrane segments (numbered 1-6 for each domain) with green cylinders indicating the pore-lining segments and yellow cylinders indicating the S4 voltage sensors (positively charged residues are represented by the + signs). The extracellular immunoglobulin-like domains of the β1 and β2 subunits are shown and a site of interaction between α and β1 subunit is indicated by the wavy, gray lines. Consensus phosphorylation sites are indicated by the orange circles. Purple circles represent the outer (EEDD) and inner (DEKA) rings of amino acid residues forming the ion selectivity filter and the TTX binding site. White circles indicate residues critical for fast inactivation. Binding sites for α- and β-scorpion toxins and anesthetic drugs are also shown. The residue affected by the possum mutation results in a threonine to alanine change at residue 790 of the SCN10A protein.
The sodium channels form part of the ion channel superfamily that includes voltage-gated potassium and calcium channels.  The 1957 amino acid protein Scn10a (also called PN3, SNS, and hereafter Nav1.8) is a voltage-gated sodium channel α subunit.  The α subunit, of approximately 220-260 kD, is the pore-forming entity of sodium channels, and may be joined by one or more auxiliary β subunits that modify the kinetics and voltage dependence of channel gating, and regulate channel localization and interaction with cell adhesion molecules, extracellular matrix and cytoskeletal components (14).  Nine proteins (referring to α subunits) comprise the mammalian sodium channel family, and all are greater than 50% identical in the transmembrane and extracellular domains, although the intracellular domains are not highly conserved [reviewed in (15)].  They are named by principal permeating ion (Na), principal physiological regulator (voltage, indicated as the abbreviated subscript, v), gene subfamily number (currently only 1), and specific channel isoform number (1-9, assigned based on the order in which they were identified).  Splice variants are identified by lowercase letters following the channel isoform number.  (See record for TremorD for a description of a Nav1.6 mutation.)

Nav1.8, like the other sodium channel α subunits, contains four homologous domains (I-IV) each consisting of six transmembrane α helices (S1-S6) (Figure 7) (5;6;16).  A membrane-reentrant extracellular loop called the “pore loop” is located between the S5 and S6 helices in each of the four domains [see (15;17) and references therein for detailed information on sodium channel structure].  The pore loops line the outer, narrow entry to the pore, and each loop contains a pair of amino acids that form outer and inner rings serving as the ion selectivity filter in the outer pore of the channel.  The outer ring of acidic residues is EEDD, while the inner ring is DEKA.  A single amino acid residue in the pore loop of domain I determines the affinity of different sodium channels for the pore blocker tetrodotoxin.  For example, rat cardiac sodium channels bind tetrodotoxin with 200-fold lower affinity than brain or skeletal sodium channels because of the substitution of an aromatic tyrosine or phenylalanine residue with cysteine (18;19).  In Nav1.8, this position (amino acid 356) is occupied by a serine residue (2;4;5), which renders the channel tetrodotoxin-insensitive (IC50 ≈ 60-100 μM) (20).  The S4 helices constitute the voltage sensors of the channel, and contain positively charged amino acid residues (arginine and/or lysine) at every third position.  Depolarization of the membrane allows the S4 segments to move outward and initiate a conformational change that opens the pore.  The short, intracellular loop between domain III, S6 and domain IV, S1 functions as the inactivation gate, which folds into the channel and blocks the pore from the inside during sustained depolarization of the membrane.  A hydrophobic triad of isoleucine, phenylalanine and methionine (IFM) within this loop is critical for fast inactivation (see Background), and glycine and proline residues flanking the IFM motif serve as hinges to allow closure of the inactivation gate.  The S5 and S6 helices line the inner, wider exit from the pore, and local anesthetics (e.g. procaine, etidocaine, lidocaine) and related antiepileptic (e.g. phenytoin) and antiarrhythmic drugs (e.g. lidocaine, quinidine) all bind from inside the cell to overlapping sites within this pore to block ionic current through the channel.  Amino acids in the S6 segments from at least three of the four domains contribute to drug binding, with residues from the S6 helix of domain IV (F1764 and Y1771) being the most important (21;22).  Together, these residues are thought to comprise the only known receptor site for local anesthetics.


Figure 8. Crystal structure of a voltage-gated Na+ channel (closed-pore conformation) from Arcobacter butzleri (NavAb). From PDB: 3RW0. One subunit is shown (1-6, transmembrane segments S1-S6; P and P2, pore (P)-helices). The voltage sensing domain (VSD) is not shown. See text for more details.

The crystal structure of a voltage-gated Na+ channel (closed-pore conformation) from Arcobacter butzleri (NavAb) has been solved at 2.7Å resolution [Figure 8; PDB:3RVY & PDB: 3RW0 (23)]. NavAb is a member of the NaChBac bacterial channel family that is considered an ancestor of vertebrate Nav channels. Similar to other voltage-gated ion channels, NavAb has four subunits/domains that create an ion-conducting pore surrounded by voltage sensors (23).  Payandeh et al. (2011) propose that the S1N helix and S2-S3 loop shield the intracellular surface of the voltage sensing domain and that the voltage-sensor domains and S4-S5 linkers dilate the central pore by pivoting together around a hinge at the base of the pore (23). The pore module of NavAb has an outer funnel-like vestibule, a selectivity filter, a central cavity, and an intracellular activation gate (23). The pore (P)-helices stabilize cations in the central cavity through helical-dipole interactions and the second pore-helix (P2-helix) forms an extracellular funnel in NavAb (23).


The Possum mutation is a threonine to alanine substitution at residue 790 of Nav1.8, located in the S5 helix of domain II, at the cytoplasmic interface of the channel.



Northern blot and RT-PCR analysis indicates that Nav1.8 is highly expressed in small-diameter sensory neurons of dorsal root and trigeminal ganglia, in both newborn and adult mice (5;6).  Among dorsal root ganglion sensory neurons, Nav1.8 is predominantly expressed in nociceptive neurons with C fibers (89% positive for Nav1.8 immunoreactivity) or Aδ fibers (93% positive), and less in neurons with Aα/β fibers (21% of total neurons of this type positive) (7).  Neurons with C fibers unresponsive to nociceptive stimuli also express Nav1.8 (7).  C fibers (unmyelinated) and A fibers (myelinated) convey “slow” (0.5-2 m/s) and “fast” (3-75 m/s) pain signals from the skin, respectively.  C fibers respond to thermal, mechanical, and chemical stimuli, as well as to changes in homeostasis of cardiovascular, respiratory, energy, and fluid systems (24).  A fibers are divided into Aα/β and Aδ fibers, which respond to stretch stimuli or pressure, cold, and homeostatic stimuli, respectively.


Nav1.8 is not detected in other peripheral or central neurons, nor in glia or non-neuronal tissues.  However, Nav1.8 mRNA and protein are expressed within cerebellar Purkinje cells in mice with experimental allergic encephalomyelitis (EAE), a model for multiple sclerosis, and in post-mortem tissue of humans with multiple sclerosis (25).  The degree of upregulation of Nav1.8 expression positively correlates with EAE duration and worsening, and expression of Nav1.8 has been shown to perturb the firing patterns of cerebellar Purkinje cells (26;27).  Abnormal Nav1.8 expression in the CNS has been hypothesized to contribute to clinical abnormalities such as ataxia in multiple sclerosis and EAE (25).  Analysis of EST counts suggests that Nav1.8 is also expressed at low levels in the inner ear (3 Scn10a ESTs present out of a total pool of 38356 ESTs; data from UniGene’s EST ProfileViewer; UniGene Accession # 333688).


A recent report from Verkerk and colleagues examined NaV1.8 expression in the heart (28).  They observed staining within intracardiac ganglia located in the pulmonary vein region; staining within the heart was less evident with most expression within the myocardium in between cardiomyocytes (28).   Co-localization with neural proteins (i.e., Tubb3) indicates that NaV1.8 is localized to neural tissues within the myocardium (28).  Examination of NaV1.8 localization in isolated intracardiac neurons found that NaV1.8 confirmed that NaV1.8 is expressed in neurons of the heart (28).  Expression levels of NaV1.8 were variable within the intracardiac neurons.  Furthermore, NaV1.8 was expressed in only a small subset of neurons of sympathetic origin; the majority of neurons are cholinergic in origin (28).

Figure 9. α-subunit of voltage-gated sodium channels. A, Structure of the Nav channel. Domains are numbered I through IV. The four domains form an ion selective pore in the membrane. The inactivation gate is between domains III and IV.  In a resting state, the interior of the cell is negatively charged, while the exterior is positively charged. B, Deactivated, activated, and inactivated states of channel in cross-section. Positively charged residues in the voltage sensor S4 helices are indicated by the + signs. See text for details.

Figure 10.  Conformational changes in the sodium channel upon activation.  The movie is adapted from Yarov-Yarovoy,V. et al. (2012). PNAS, 109:E93-102 based on studies with the bacterial NaV channel, NaChBac.  Mammalian NaV and CaV channels consist of four homologous domains (I through IV), each containing six transmembrane segments (S1 through S6) and a membrane-reentrant pore loop between the S5 and S6 segments.  The colors in Figure 10 are the same as in Figure 7, domain I and the voltage-sensing domain (VSD) of domain III were omitted for clarity. (Step 1) The hydrophobic side chains of I33 (in S1), F67 (in S2), and I96 (in S3) create the hydrophoibc constriction site (HCS) within the VSD core. (Step 2)  S4 moves outward, rotates, and tilts as it passes through the HCS. (Step 3) The movements of S4 moves the S3-S4 loop closer to the S1-S2 loop to generate a cleft present only in the activated state on the extracellular side of the VSD.  (Step 4) The movement of S4 in "Step 2" results in a sideways gating movement of the S4-S5 linker on the intracellular side of the VSD.  (Step 5) The secondary structure of the VSD changes and allows for release of energy stores to drive pore opening.

The voltage-gated sodium channels (VGSCs or Nav) mediate the rapid influx of sodium that underlies the depolarizing rising phase of action potentials in neurons and excitable cells [Figure 9 & 10; reviewed in (17)].  In a seminal paper, Hodgkin and Huxley demonstrated the three key features that now characterize the sodium channel: voltage-dependent activation, rapid inactivation, and selective ion conductance (29).  VGSCs typically open (“activate”) within a millisecond in response to membrane depolarizations (Figure 10), often leading to a regenerative all or none depolarization typical of action potentials in neurons.  Once activated, VGSCs do not remain open for long, and current through VGSCs can be terminated within approximately one millisecond by a process known as “fast inactivation” mediated by the intracellular inactivation gate folding into the channel and blocking the pore (see Protein Prediction).  Recovery from inactivation occurs after the cell membrane potential is repolarized to negative membrane potentials (mediated by other ion channels), allowing sodium channels to return to the resting or closed state in which they are again ready to open in response to membrane depolarization.


Although all sodium channels generally function as described, distinct sodium channels differ in their precise functioning, and the sodium currents measured in various neuronal cell types are not homogeneous.  In particular, inactivation kinetics can vary widely between distinct channel isoforms.  Based on sensitivity to blockage by tetrodotoxin (TTX), a toxin found in the liver of the puffer fish, sodium channels are divided into two groups: TTX-sensitive channels that are typically fast inactivating (decay time from peak current <5 ms at room temperature), and TTX-resistant channels that are typically slowly inactivating (decay time from peak current 5-10 ms at room temperature).  Central nervous system (CNS) neurons are known to express relatively homogeneous fast inactivating TTX-sensitive currents, while peripheral dorsal root ganglia neurons express currents containing both fast inactivating TTX-sensitive and slowly inactivating TTX-resistant components (30;31).  It has been proposed that the slower TTX-resistant currents might serve to prolong the duration of the action potential to modulate neurotransmitter release.


The dorsal root ganglion (DRG) neurons of the peripheral nervous system project axons to the CNS, and serve to relay sensory information detected by dendrites innervating organs such as the skin, muscles, and internal organs to the CNS.  Thus, DRG neurons sense and transmit information about pain (as well as other) sensations from the periphery, and are a large focus of pain research.  Increased or decreased pain sensations can result from the altered electrical excitability of DRG neurons, and VGSCs, which determine the excitability of neurons, have been implicated in this process.  Nav1.8 is a slowly inactivating TTX-resistant channel expressed in neurons of the dorsal root ganglia.


Targeted deletion of Nav1.8 in mice has provided insights into the specific role of Nav1.8 in pain states.  Nav1.8-null mice display reduced thermo- and mechanoreceptive pain responses, and reduced responses to inflammatory pain stimuli in several tests (8;32).  Slightly but significantly increased latencies are observed in the paw-withdrawal and tail-flick responses to a radiant heat stimulus, although latency in the hot plate test is similar between mutants and wild type animals.  Pain thresholds in the tail pressure test are increased in Nav1.8-null mice.  Mutant mice show reduced thermal hyperalgesia induced by inflammatory stimuli such as intraplantar injection of carrageenan or systemic injection of nerve growth factor (8;32).  In these tests, injection of the inflammatory stimulus decreases paw-flick latency to heat applied to the paw; Nav1.8-null mice display a slightly longer paw-flick latency compared to wild type animals.  Electrophysiological analyses demonstrate that small DRG neurons from Nav1.8-null mice are less likely to fire action potentials, consistent with the analgesia observed in behavioral studies (33).  However, Nav1.8-deficient small DRG neurons have also been reported to exhibit lowered thresholds of electrical activation and increased current densities of TTX-sensitive channels (8).  These may be compensatory changes for the lack of Nav1.8 (8).


Nav1.8 is required for sensing cold pain and pain in the cold (10).  Electrophysiological recordings from cold-sensitive neurons innervating the skin demonstrate that TTX-sensitive VGSCs become inactive at low temperatures, while the inactivation properties of Nav1.8 are cold-resistant.  Responsiveness to mechanical stimulation at low temperatures is impaired in neurons lacking Nav1.8.  Nav1.8-null mice display greatly reduced foot-lifting behavior relative to wild type mice in the cold-plate test, in which animals are placed on a plate held at 0° C.  Thus, Nav1.8 function is critical and may be specialized for perception of cold pain and pain in the cold.


In addition to a role in primary pain states, Nav1.8 has been implicated in neuropathic pain, although this function remains controversial.  Neuropathic pain is chronic pain such as shooting or burning pain, or tingling and numbness, and is frequently induced by tissue and nerve damage.  Neuropathic pain is hypothesized to result from the spontaneous firing of pain-sensing neurons, in part due to the aberrant expression and function of VGSCs.  Spontaneous activity in damaged sensory neurons is reportedly reduced in Nav1.8-null mice relative to wild type (34), and knockdown of Nav1.8 has been shown to reverse neuropathic pain induced by spinal nerve injury in rats (35).  However, other studies have found no role for Nav1.8 in neuropathic pain (32;36).  In addition, Nav1.8 mRNA and protein expression in axotomized small DRG neurons are dramatically reduced in models of neuropathic pain (37;38).  It has been suggested that compensatory changes in expression of other channels may support the ectopic firing of injured neurons, or that altered Nav1.8 activity in uninjured neurons contributes to their hyperactivity, both possibilities that may lead to development of neuropathic pain.


Nav1.8 does appear to modulate neuropathic pain caused by mutations in the Nav1.7 channel. Nav1.7 is expressed in small DRG neurons (39), and exhibits TTX sensitivity, as well as rapid activation and inactivation properties (40).  These channels likely contribute the predominant TTX-sensitive current found in small DRG neurons (41).  Strikingly, nonsense mutations in Nav1.7 result in a complete lack of pain sensation in humans that appear otherwise healthy, except for deficits in olfaction in some patients (42;43) (OMIM *603415, #243000).  Similarly, selective deletion of Nav1.7 in small DRG neurons in mice (using a Cre-lox system in which Cre is expressed from the Nav1.8 promoter) reduces mechanical and thermal hypersensitivity associated with acute or inflammatory pain (44).  Mice completely lacking Nav1.7 die just after birth because of a failure to feed, possibly due to anosmia or hyposmia similar to that documented in human patients (44).  Conversely, point mutations that alter the gating properties of Nav1.7 can cause increased pain sensations and neuropathy, as in inherited erythromelalgia (also called erythermalgia, OMIM #133020) for which the primary symptom is severe chronic burning pain sensations in the hands and feet.  The disease is caused by Nav1.7 mutations that increase neuronal hyperexcitability, for example by causing hyperpolarizing shifts in the voltage-dependence of channel activation, thereby lowering the threshold of the neurons for firing action potentials (45).  Nav1.8 channels have been shown to contribute to the hyperexcitability of DRG neurons with mutant Nav1.7 channels.  The erythromelalgia mutation L858H produces a depolarizing shift in resting membrane potential and decreases action potential threshold in DRG neurons (46).  This hyperexcitability is supported by the presence of Nav1.8, which, due to its depolarized voltage-dependence of activation and inactivation, permits the neurons to generate action potentials and sustain repetitive firing despite a depolarized resting membrane potential.  Superior cervical ganglion neurons do not express Nav1.8, and the same Nav1.7 mutation depolarizes the resting membrane potential but renders these cells hypoexcitable.

Putative Mechanism

The Possum mutation, a threonine to alanine substitution in transmembrane helix S5 in domain II, alters the physiology of Nav1.8 channels by increasing current density and slowing the kinetics of inactivation. Precedent exists for the hyperactivation of sodium channels by dominant missense mutations located adjacent to the cytoplasmic leaflet of the plasma membrane lipid bilayer.  Hyperkalemic periodic paralysis syndrome (HYPP, OMIM #170500) is caused by mutations in SCN4A, encoding the skeletal muscle sodium channel Nav1.4 (47). The disease is an autosomal dominant disorder characterized by temporary limb weakness that occurs spontaneously or is triggered by rest after exercise, consumption of potassium-rich foods, stress or fatigue (48).  A progressive, fixed, interattack weakness may also develop in some patients. The two most common mutations causing HYPP are T704M and M1592V, accounting for 75% of all cases (48).  The threonine to methionine substitution at position 704 of human Nav1.4, hyperactivates the channel by shifting the voltage-dependence of activation in the hyperpolarizing direction and dramatically impairing slow inactivation (49). The gain-of-function mutation is dominant because only a small amount of noninactivating sodium current is sufficient to cause repetitive firing and steady depolarization of muscle cells leading to periodic paralysis.  Notably, T704 of Nav1.4 is orthologous to T790 of Nav1.8 on sequence alignment.


Similar dominant mutations in human Nav1.7 cause erythromelalgia by increasing neuronal hyperexcitability.  When their sequences are aligned, the L858H mutation of Nav1.7 would lie adjacent to the threonine mutated in Possum (it would correspond to L789 of mouse Nav1.8).  Nav1.7 L858 is the last amino acid of the linker between helices S4 and S5 in domain II.  The L858H mutation shifts the voltage-dependence of activation in the hyperpolarizing direction and also slows the deactivation time of the channel (50).  Another mutation, A863P (within domain II, helix 5, four amino acids from the Possum mutation in the corresponding Nav1.8 sequence), causes the same types of defects in channel function, again resulting in hyperexcitability of neurons (51).


The threonine mutated in Possum is conserved in all nine of the known VGSCs in both human and mouse, as are at least 7 of the 10 surrounding amino acids.  It is likely that alanine, which is nonpolar and hydrophobic, fails to form the electrostatic interactions normally mediated by the polar threonine residue, thereby changing channel conformation and properties. Another factor that may be involved in the effect of the Possum mutation is the change in size of the amino acid side chain. Side chain size of S241 in the domain I S4-S5 linker of Nav1.7 was found to be a critical determinant of channel properties (52).


The normal responses of Possum mice to noxious heat and punctate mechanical stimuli argue against the idea that excessive pain exists and that it triggers the immobility phenotype in scruffed Possum mice. Among DRG, Nav1.8 is predominantly expressed in nociceptive neurons with C fibers and Aδ fibers, and less in neurons with Aα/β fibers (7).  Nociceptive C fibers ablated by capsaicin are not required for the induction of immobility in Possum mice.  Thus, sensory fibers resistant to neonatal capsaicin ablation, which include Aδ fibers and Mrgprd+ mechanosensitive C fibers (53), may instead play an integral role in the immobility response to scruffing, and may be necessary and sufficient for sensing and propagating the signals induced by such physical manipulation. However, it cannot be ruled out that developmental compensatory mechanisms impart susceptibility to the immobility phenotype, or that the phenotype is mediated by events other than acute activation of mutant Nav1.8 channels in nociceptors during scruffing.


It was reported previously that nonsynonymous coding sequence variants of human SCN10A are associated with increased PR interval, P wave duration, and QRS duration in ECGs, and that Nav1.8-deficient mice have decreased PR intervals (11-13).  Scruffed Possum mice experienced vagus nerve-stimulated sinus bradycardia and highly irregular RR intervals on ECG concomitant with immobility. These data do not permit us to discern whether the effects of Nav1.8Possum on cardiac conduction are intrinsic to cardiomyocytes or secondary to neurotransmission via DRG neurons.  Expression and localization analysis by Verkerk et al. (2012) indicate that the expression of Nav1.8 is predominant in intracardiac neurons; expression in cardiomyoctes was not detected (28). Furthermore, Verkerk et al. (2012) show that Nav1.8 within the intracardiac neurons regulates neuronal action potential firing frequency (28).  The pathways leading the effects on cardiac conduction remain unclear.  Verkerk et al. (2012) speculate that the neurons that are controlled by Nav1.8 regulate the release of neurotransmitters and/or neuropeptides (28).  The neurotransmitters/neuropeptides can subsequently act by a direct effect on a myocyte or indirectly through other neurons. Observation of normal blood pressures during immobility and the inability to block immobility with atropine support the conclusion that the cardiac abnormalities are independent of the behavioral Possum phenotype. The ability to restore normal cardiac rhythm and prevent bradycardia using atropine demonstrates the existence of a novel reflex arc activated by a discrete stimulus (scruffing), and having as one output a strong vagal response.


Neither excessive pain nor altered cardiac conduction was responsible for the immobility phenotype, yet the mechanistic basis of the behavior remains unknown.  We hypothesize that scruffing-induced immobility in Possum mice is equivalent to the phenomenon of manipulation-induced tonic immobility or pinch-induced catalepsy reported to result from repeated pinching of the scruff of mice (54;55). Although no quantitative standard exists to define tonic immobility, experimentally induced tonic immobility is well accepted to be induced by some form of physical restraint, and to be characterized by a transient state of enhanced muscular rigidity, ‘waxy flexibility’, and altered EEG patterns indicative of enhanced alertness (56-58), that terminate abruptly after varying duration (2;3), exactly as observed in Possum mutants. Existing data support a genetic basis for tonic immobility (59), but genes contributing to this behavior, whether experimentally or naturally induced, have not been identified. In addition, no class of sensory neurons has previously been implicated in this response. Sodium conduction in DRG neurons, occurring via Nav1.8, may be one factor responsible for this behavior.


Although the neural substrates of manipulation-induced tonic immobility remain poorly defined, evidence suggests that they include the brainstem (60), particularly the parabrachial region (PB) (61), and the ventrolateral periaqueductal gray (PAG) (62;63) that, respectively, relay nociceptive information to the cortex and amygdala, and modulate sensory afferent output from the spinal cord. Dopaminergic, cholinergic, GABAergic, and opioid signaling in these and other areas are thought to mediate manipulation-induced tonic immobility (63;64). It may be that scruffing, an arguably intense stimulus, super-activates these nuclei through stimulation of DRG neurons rendered hyperexcitable by Nav1.8Possum and leads to the dramatic immobility observed in Possum mice. The sensitivity of the scruff to stimulation leading to immobility may reflect the evolutionarily ancient underpinnings of this behavior, which is thought to represent an antipredatory action that reduces both the appeal and detection of prey by predators (2;65). Possum mice represent a novel animal model for studying this state, which has been variable and difficult to elicit in mice.


Altered EEG patterns, observed only during immobility, also raise the possibility of direct function(s) for Nav1.8 in CNS neurons, which may contribute mechanistically to the behavior. Although brain expression of Nav1.8 has not been reported (4-7), small amounts or highly localized expression below the level of detection by standard assays cannot be ruled out.

Primers Primers cannot be located by automatic search.
Possum genotyping is performed by amplifying the region containing the mutation using PCR, followed by sequencing of the amplified region to detect the single nucleotide transition.
PCR program
1) 95°C             2:00
2) 95°C             0:30
3) 56°C             0:30
4) 72°C             1:00
5) repeat steps (2-4) 29X
6) 72°C             7:00
7) 4°C               ∞
Primers for sequencing
The following sequence of 602 nucleotides (from Genbank genomic region NC_000075 for linear genomic sequence of Scn10a) is amplified:
55315                                                            tgaaac
55321 agatggcaga caagcctcac taatgtgagg ccctggatct gccagttggt ggagtacatt
55381 tcaatttagc gaccactgga aggtgatcag agctctgagt gagatgtaaa aagaaaagtc
55441 tccacaaaca gccccacccc aaactcatga gatgtcttgg tgcttgtccc acaagcttgg
55501 cttccagtga agacttcagt tatgagatgg tgagaagaga cctgagacag tgtctctttc
55561 cacagcttcg ggtcttcaag ctggccaagt cctggcccac cctgaacatg ctcatcaaga
55621 tcatcgggaa ctctgtgggg gccctgggca acctgacctt catcctggcc atcatcgtct
55681 ttatcttcgc cctggtggga aagcagctcc tctcagagaa ctatgggtgc cgcagggatg
55741 gcatctccgt gtggaatggt gagaggctgc gctggcacat gtgtgacttc ttccattcct
55801 tcctcgtcgt cttccggatc ctctgcgggg agtggatcga gaacatgtgg gtctgcatgg
55861 aggtcagcca ggactacatc tgcctcaccc tcttcttgac agtgatggtg ctaggc
Primer binding sites are underlined; sequencing primer binding sites are highlighted in gray; the mutated A is indicated in red.
  13. Pfeufer, A., van Noord, C., Marciante, K. D., Arking, D. E., Larson, M. G., Smith, A. V., Tarasov, K. V., Muller, M., Sotoodehnia, N., Sinner, M. F., Verwoert, G. C., Li, M., Kao, W. H., Kottgen, A., Coresh, J., Bis, J. C., Psaty, B. M., Rice, K., Rotter, J. I., Rivadeneira, F., Hofman, A., Kors, J. A., Stricker, B. H., Uitterlinden, A. G., van Duijn, C. M., Beckmann, B. M., Sauter, W., Gieger, C., Lubitz, S. A., Newton-Cheh, C., Wang, T. J., Magnani, J. W., Schnabel, R. B., Chung, M. K., Barnard, J., Smith, J. D., Van Wagoner, D. R., Vasan, R. S., Aspelund, T., Eiriksdottir, G., Harris, T. B., Launer, L. J., Najjar, S. S., Lakatta, E., Schlessinger, D., Uda, M., Abecasis, G. R., Muller-Myhsok, B., Ehret, G. B., Boerwinkle, E., Chakravarti, A., Soliman, E. Z., Lunetta, K. L., Perz, S., Wichmann, H. E., Meitinger, T., Levy, D., Gudnason, V., Ellinor, P. T., Sanna, S., Kaab, S., Witteman, J. C., Alonso, A., Benjamin, E. J., and Heckbert, S. R. (2010) Genome-Wide Association Study of PR Interval. Nat. Genet.. 42, 153-159.
Science Writers Eva Marie Y. Moresco
Illustrators Diantha La Vine
AuthorsAmanda L. Blasius, Adrienne E. Dubin, Matt J. Petrus, Byung-Kwan Lim, Anna Narezkina, Jose R. Criado, Derek N. Wills, Yu Xia, Cindy L. Ehlers, Kirk U. Knowlton, Ardem Patapoutian, Bruce Beutler