Body Temperature rises to fever levels during regular activity and upon exposure to higher ambient temperatures.

A. Left: Setup for recording mouse body temperature (Tb) at room temperature using an implanted transponder for non-invasive measurement. Right: Average Tb readings over 6 hours at 5-minute intervals for 10 mice. Tb typically hovers around 36.5°C during the day. B. Tb may briefly elevate into fever range (shaded red bars) during regular daily activity. Median with upper and lower limits indicated. C. Tb readings over 6 hours, at 5-minute intervals for animal 3 in Figure 1B. Shaded red dots indicate time points when Tb enters the fever range. D. Left: Setup for Tb recordings during infrared light exposure. Right: Tb elevates into the fever range (>38°C) for extended durations, lasting minutes to hours. Average Tb readings over 6 hours at 5-minute intervals for 10 mice.

Spiking in P12-14 cortical excitatory pyramidal neurons remains stable as temperature enters the fever range.

A. Setup for recording L4-evoked postsynaptic potential and spiking in a cortical excitatory pyramidal neuron (PN) at just-subthreshold Vm at 30°C, 36°C and 39°C in mouse primary somatosensory (S1) cortex. B. Example traces of L4-evoked spikes in L2/3 PN during consecutive recordings at 30°C (black), 36°C (gray), and 39°C (red). C. Depolarization required to reach spike threshold (ST) in L2/3 PNs during temperature elevations from 30°C to 39°C. D. Same as (C) for input resistance (Rin). E. Same as (C) for the percentage of spiking L2/3 PNs. F. Same as in (C) for the spiking distribution. G. Same as in (C) for the number of spikes. H. Rheobase in L2/3 PNs during temperature elevations from 30°C to 39°C. 8 PNs recorded from 3 animals. I. Quantification of evoked spiking for F-I curves in L2/3 PNs during temperature elevations from 30°C to 39°C. Each recorded cell was exposed to the full temperature range (30 to 39°C). 8 PNs recorded from 3 animals. In C, D and G, each data point represents an individual cell. In C-F, 37 PNs recorded from 14 animals, with each cell exposed to temperature elevations from 30°C to 39°C. Data in C, D, and G are shown as mean ± SEM. Statistical significance was assessed as follows: C, D, G :one-way repeated-measures ANOVA with Tukey’s post-hoc test (α = 0.05); E – two-tailed binomial test; F – Kolmogorov–Smirnov (K–S) test, H: two-way repeated-measures ANOVA with Tukey’s post-hoc test(α = 0.05).

Spiking in P7-8 cortical excitatory pyramidal neurons decreases as temperature enters the fever range.

A. Setup for recording L4-evoked postsynaptic potential and spiking in an excitatory pyramidal neuron (PN) at just-subthreshold Vm at 30°C, 36°C and 39°C in mouse primary somatosensory (S1) cortex. B. Example traces of L4-evoked spikes in L2/3 PN during consecutive recordings at 30°C (black), 36°C (gray), and 39°C (red). C. Depolarization required to reach spike threshold (ST) in L2/3 PNs during temperature elevations from 30°C to 39°C. D. Same as (C) for input resistance (Rin). E. Same as (C) for the percentage of spiking L2/3 PNs. F. Same as in (C) for the spiking distribution. G. Same as in (C) for the number of spikes. In C, D and G, each data point represents an individual cell. In C-G, 19 PNs recorded from 5 animals, with each cell exposed to temperature elevations from 30°C to 39°C. Data in C, D, and G are shown as mean ± SEM. Statistical significance was assessed as follows: C, D, G :one-way repeated-measures ANOVA with Tukey’s post-hoc test (α = 0.05); E – two-tailed binomial test; F – Kolmogorov–Smirnov test.

Increased depolarization at higher spike thresholds helps maintain stable spiking activity in cortical pyramidal neurons during temperature elevations.

A. Illustration demonstrating how temperature-induced changes in spike threshold make spiking more challenging for a neuron, requiring larger levels of depolarization to sustain spiking. B. Example traces of L4-evoked postsynaptic potentials (PSPs) in L2/3 PN during consecutive recordings at 30°C (black), 36°C (gray), and 39°C (red). C. Correlation of PSP peak versus spike threshold (ST). r = Pearson correlation coefficient with Deming linear regression. D. Correlation of input resistance versus ST. r = Pearson correlation coefficient with simple linear regression. E. Same as D but for input resistance versus PSP. F. Illustration showing how temperature-induced loss in inhibition (blue IPSP) could lead to larger levels of depolarization. IPSPs typically bring the membrane potential away from the spike threshold. G. The late PSP peak in L2/3 PNs during temperature elevations from 30°C to 39°C. Statistical significance was evaluated on log-transformed data using one-way repeated measures ANOVA with Tukey’s test, with significance at α=0.05. H. Same as D but for PSP versus Late PSP Peak. r = Spearman correlation coefficient with simple linear regression.

Cortical excitatory neurons that remain active at febrile temperatures exhibit increased spiking rates.

A. Setup for in vivo recording of spiking in mouse S1 cortex during temperature increases from 36°C to 39°C and cooling back to 36°C. B. Color maps (with yellow and red indicating spiking activity) illustrate multiunit activity obtained in the left and right cortices during baseline temperature (left), warming to fever Tb (middle), and cooling back to 36°C (recovery upon heat removal). C. Example traces of spiking activity recorded from three cortical positions (marked by black, gray, and white asterisks on the right side of panels in B. D. Mean spike waveforms (black), single spikes (gray) and autocorrelograms (bottom) of single units isolated, with unit 3 representing a putative interneuron. E. Normalized firing rates of putative excitatory PNs and interneurons obtained from recordings in A. F. The normalized firing rates of five representative units during the baseline, warming (red shaded area), and recovery periods. F presents the same data, with color coding consistent with the units in D. G. Left: Setup for ex vivo recording of L4-evoked postsynaptic potentials and spiking in L2/3 cortical PN at just-subthreshold Vm at 30°C, 36°C and 39°C in mouse S1 cortex. Right: Percent distribution for recorded PNs that never spiked, stopped spiking, stayed spiking, or started spiking during temperature increases from 36°C to 39°C. H. L2/3 PN spiking activity in neurons that stayed spiking (left), started spiking (middle), and stopped spiking (right) upon temperature increases from 36°C to 39°C. I. Activity distribution for ex vivo recorded neurons that never spiked, and those that stopped spiking, stayed spiking, or started spiking upon temperature increases from 36°C to 39°C. P-value = 0.01 indicates a significant increase in the fraction of neurons that stayed spiking during temperature increases from 36°C to 39°C (solid red bar). In E and H, each data point represents an individual animal and cell, respectively. Data were collected from 5 animals in B–F, 11 animals in G–H, and 9–14 animals in I. Statistical significance was assessed using one-way repeated-measures ANOVA with Tukey’s post-hoc test (E), paired two-tailed t-test (H), and binomial test (I), with significance set at α = 0.05.

Excitatory pyramidal neurons that remain spiking with temperature elevations into fever range exhibit unique intrinsic properties.

Neurons that spiked at all temperatures (30°C, 36°C, and 39°C) are STAY PNs, while those that stopped spiking at 36°C or 39°C are STOP neurons A. Depolarization required to reach spike threshold (ST) in STOP and STAY L2/3 PNs during temperature elevations from 30°C to 39°C. B. Same as (A) for L4-evoked postsynaptic potentials (PSPs) C. Correlation of PSP peak versus ST in STAY L2/3 PNs at 30°C, 36°C, and 39°C. r = Pearson correlation coefficient with Deming linear regression. D. Same as (C) for STOP cells. E. Same as (A) for the L4-evoked late PSP peak. F. Same as (A) for spike height. G. Same as (A) for spike afterhyperpolarization (AHP). H. Same as (A) for input resistance. Each data point in A–H represents an individual cell. Data were collected from 14 animals. Mean ± SEM is shown in A–B and E–H. Statistical significance was assessed using one- or two-way repeated-measures ANOVA with Tukey’s or Sidak post-hoc test (A–B, E–H; α = 0.05). In C–D, correlations were evaluated using Pearson’s r with Deming linear regression.

TRPV3 protein expression by immunohistochemistry at postnatal days 7, 14, and 21.

A. TRPV3 immunostaining (green) in mouse primary somatosensory (S1) cortex, striatum, thalamus, and hippocampus, with DAPI counterstain (blue). Top row left: Postnatal day (P)7; bottom row left: P14; bottom row right: P21. Top row right: Sections incubated with TRPV3 antibody plus TRPV3 blocking peptide in S1 cortex, striatum, thalamus, and hippocampus (green) with DAPI counterstain (blue). Remaining TRPV3 staining is shown in green with DAPI in blue. B. TRPV3 and TRPV4 immunostaining in S1 cortex, striatum, thalamus, and hippocampus at P14. Top row left: TRPV3 (green) and TRPV4 (brick red). Top row right: TRPV3 (green) and TRPV4 (brick red) with DAPI counterstain (blue). Bottom row left: TRPV4 immunostaining in S1BF (inset from top row left) at 10×. Bottom row right: TRPV3 immunostaining in S1BF (inset from top row left) at 10×.

Temperature elevations in fever range increase TRPV3 currents in cortical pyramidal neurons.

A. Setup for recording whole-cell TRPV3 currents at 30°C (black), 36 °C (grey) and 39°C (red). in cortical excitatory pyramidal neurons (PNs) with bath application of camphor (5mM), a TPRV3 agonist. B. Current density-voltage (I-V) relationship of TRPV3 currents at 30°C (black), 36 °C (grey) and 39°C (red) in WT mice: 11 cells from 4 mice. C. Scatter dot plots of the current density-voltage measurements. D. Current density-voltage (I-V) relationship of TRPV3 currents at 30°C (black) in the presence of camphor (5mM), a TPRV3 agonist, or camphor (5mM) + TRPV3 blocker (Forsythoside B, 50 µM) (blue). E. Same as (D) but for 36°C. F. Same as (D) but for 39°C. G. Current density-voltage (I-V) plot showing the net TRPV3 current (opener – (opener+ blocker) condition). In B-F, statistical significance was assessed using a two-way repeated-measures ANOVA with Tukey’s or Sidak post-hoc test (α = 0.05)

Inhibiting TRPV3, but not TRPV4 channels, significantly reduced the population of STAY pyramidal neurons and spiking levels at fever temperature.

A. Setup for recording L4-evoked postsynaptic potential and spiking in an excitatory cortical pyramidal neuron (PN) with an intracellular blocker of TRPV3 channels (Forsythoside B, 50 µM) (left) or TRPV4 channels (RN1734, 10 µM) (right) at just-subthreshold Vm at 30°C, 36°C, and 39°C in mouse S1 cortex. B. Percentages of cell types obtained from experiment in A. C. Evoked spikes in L2/3 cortical PNs during temperature elevations to 30°C, 36°C, and 39°C under three conditions: no blockers, TRPV3 blocker (Forsythoside B, 50 µM), or TRPV4 blocker (RN1734, 10 µM). D. Correlation between PSP peak and spike threshold (ST). r = Pearson correlation coefficient with Deming linear regression. E. Same as (C) for the L4-evoked late PSP peak. F. Same as (C) for input resistance (Rin). Each data point in C–F represents an individual cell. Data were collected from 26 cells in 7 animals for the TRPV3 blocker, 24 cells in 6 animals for the TRPV4 blocker, and 37 cells in 14 animals for the no-block condition. Mean ± SEM is shown in C, E, and F. Statistical significance was assessed using one- or two-way repeated-measures ANOVA with Tukey’s or Sidak post-hoc test (α = 0.05). In D, correlations were evaluated using Pearson’s r with Deming linear regression.

TRPV3 knockout mice exhibit reduced spiking activity at febrile temperatures and delayed seizure onset.

A. Setup for recording L4-evoked postsynaptic potentials and spiking in excitatory cortical pyramidal neurons (PNs) at just-subthreshold Vm at 30 °C, 36 °C, and 39 °C in mouse S1 cortex of wildtype (Trpv3+/+) and Trpv3 knockout (Trpv3-/-) mice. B. Depolarization required to reach spike threshold (ST) in Trpv3+/+ and Trpv3-/- mice at 30°C, 36°C, and 39°C. C. Same as (B) but for input resistance (Rin). D. Same as (B) but for number of spikes. E. Same as (B) but for postsynaptic potential (PSP). F. Setup for recording mouse body temperature (Tb) at room temperature and during fever-range and higher, using an implanted transponder for non-invasive measurement, with exposure to infrared light. G. Time to loss of postural control (LPC), defined as collapse and failure to maintain upright posture, in wildtype (Trpv3+/+), heterozygous (Trpv3+/-) and Trpv3 knockout (Trpv3-/-) mice. H. Same as in (G) but showing Tb at seizure onset. I. Same as in (G) but for the time from LPC to seizure onset. Each data point in B–E represents an individual cell (3 animals per genotype). Statistical significance was assessed using two-way repeated-measures ANOVA with Tukey’s or Sidak post-hoc test (α = 0.05). Each data point in G–I represents an individual animal. Statistical significance was assessed using one-way repeated-measures ANOVA with Tukey’s or Sidak post-hoc test (α = 0.05).

Summary model of how STAY neurons achieve firing stability through unique ion channel compositions, including TRPV3.

A. In cortical L2/3 pyramidal neurons (PNs) with synaptically evoked spiking, gradual increases in brain slice temperature from 30 °C to 36 °C to 39 °C result in four possible outcomes: neurons remain inactive, continue spiking (STAY), stop spiking, or initiate spiking. To spike, PNs must reach spike threshold (ST), defined as the minimal Vm that elicits an action potential, which requires sufficient depolarization via the postsynaptic potential (PSP). STAY neurons consistently reach ST and continue spiking. Neurons that stop spiking fall below the level of depolarization required to reach ST, while neurons that initiate spiking achieve sufficient depolarization to newly reach ST. B. STAY neurons may contain unique ion channels, such as TRPV3. TRPV3 channels are highly permeable to Ca²⁺ ions, and Ca²⁺ influx contributes to PN depolarization. The presence of TRPV3 facilitates greater ion entry, enabling depolarization sufficient to reach ST and sustain spiking, whereas its absence reduces depolarization to levels insufficient for spiking.