Figure 1.
Moderate cooling evokes large current responses in hippocampal neurons.
A, Time course of whole-cell current at −60 mV in a hippocampal neuron subjected to a cooling ramp from 35°C. Insets show individual current events (marked with an arrow) at baseline temperature and during the initial cooling period on an expanded time scale. The arrowhead marks the occurrence and temperature threshold of the first cold-evoked event. B, Response of the same neuron to heating from 35°C to 39°C and subsequent cooling. Note the absence of response during heating or during cooling from a higher baseline value. C, Number of events during cooling ramp for the neuron shown in A quantified in 2-second bins. D, Mean event frequency, E, mean amplitude and F, mean area during basal and cooling conditions (n = 15). Statistical significance in panels D–F was assessed with Student’s paired t-test (*p<0.05; **p<0.01. G–H, Cumulative probability histogram of event amplitude (G) and event area (H) at 35°C vs. during cooling of the recording shown in A. Note the tendency of cooling to shift both event amplitude and area towards larger values. I–J, Time course of cell-attached action currents recorded in the same hippocampal neuron, during identical cooling and heating protocol. Note the very similar characteristics of threshold and pattern as recorded in the whole-cell configuration. All recordings in this figure were performed in LCS (see Methods).
Figure 2.
Differential effect of synaptic blockers on cooling-evoked responses.
A–D, Time courses of whole-cell current at a holding potential of −60 mV in four hippocampal neurons in the absence and presence of the synaptic blockers A, CNQX; B, AP-V; C, Bicuculline; D, Baclofen. E–G, Summary histogram of the number of events during cooling in the absence and presence of E, CNQX (n = 5); F, AP-V (n = 4); and G, Baclofen (n = 4). Statistical significance in panels D–F was assessed with repeated-measures 1-way ANOVA in combination with Dunnett’s post-test with respect to the first cooling stimulus in control conditions, and indicated with: *p<0.05; **p<0.01. All recordings, except B, were obtained in LCS. Time scale shown in A, applies to all traces (A–D).
Figure 3.
Hippocampal neurons fire action potentials in response to cooling.
A–B, Time course of membrane potential change of two hippocampal neurons recorded in whole-cell current-clamp mode at −60 mV showing the cooling-evoked firing of action potentials. Note how the response of the neuron in A is completely abolished in the presence of a cocktail of synaptic blockers (20 µM CNQX+50 µM AP-V+5 µM bicuculline), whereas the neuron in B continues firing. The vertical lines in some of the voltage traces correspond with a pulse protocol for determination of membrane resistance. C–D, Effect of temperature on electrophysiological properties in current-clamp mode in the presence of a cocktail of synaptic blockers C, Membrane resistance was obtained from the membrane potential change in response to a 25 pA pulse of 250-millisecond-duration. D, Action potentials were evoked at rheobase using depolarizing current pulses of 450-ms-duration at a membrane potential of −60 mV. The neuron shown in C–D was silent in the presence of synaptic inhibitors. All recordings in [LCS].
Table 1.
Effect of temperature on the electrophysiological properties of hippocampal neurons recorded in current-clamp configuration in the presence of synaptic blockers (20 µM CNQX+50 µM AP-V+5 µM bicuculline), which in these abolished cooling-evoked action potential firing.
Figure 4.
Glutamate spillover does not generate the cooling-evoked responses.
A, Time course of whole-cell current [HCS] at −60 mV in a hippocampal neuron subjected to cooling ramps in the absence and presence of 150 µM DL-TBOA. Note, that TBOA does not affect synaptic activity at baseline temperature. B, Summary histogram of mean frequency, amplitude and area of synaptic events at baseline temperature 36–37°C. C, Summary histogram of mean frequency, amplitude and area of cooling-evoked events. In panels B–C, parameters in the presence of TBOA and during washout are normalized to data prior to TBOA application. Statistical significance was assessed with 1-way-ANOVA in combination with Dunnett’s post test with respect to the data prior to TBOA application, and is indicated with *p<0.05.
Figure 5.
Action potential firing is required for cooling-evoked responses in the neuronal network.
A–B, Time courses of whole-cell current [LCS] at a holding potential of −60 mV in a hippocampal neuron in the A, absence and B, presence of 1 µM TTX before and during cooling. Below, events detected in the same recordings are quantified in 2-second bins. C–D, Mean frequency, mean amplitude, and mean area of the synaptic currents of hippocampal neurons during cooling in C, control solution and D, TTX. Parameters during cooling are represented as percent of the values at baseline temperature. Note that in the absence of TTX, cooling increases all the synaptic event parameters, while in the presence of TTX, event frequency is reduced and the remaining parameters are unchanged. Statistical significance in panels C–D was assessed between each parameter during cooling and at baseline temperature with Student’s paired t-test: *p<0.05; **p<0.01, n = 6.
Figure 6.
Temperature sensitive TRP channels are not involved in generating the cooling-evoked responses.
A, Time course of whole-cell current at −60 mV in a hippocampal neuron during repetitive cooling ramps in the absence and presence of thermo-TRP agonists menthol (100 µM), AITC (20 µM) and capsaicin (1 µM). The time scale bar indicates the zero current level. B−D, Summary histograms showing the average effect of the agonists on the B, threshold, C, average frequency, and D, relative amplitude of the temperature-induced responses during the descending part of a cooling ramp (n = 5). E, Time course of whole-cell current at −60 mV in a hippocampal neuron during cooling ramps in the absence and presence of the thermo-TRP antagonist BCTC (10 µM); F, Summary histogram showing the effect of BCTC on the mean frequency, amplitude and area of the cooling-evoked responses, n = 3. G−H, 20 µM HC-030031, n = 4; I−J, 10 µM ruthenium red, n = 2. In panels F, H, J, parameters in the presence of antagonist and during washout are normalized to data prior to antagonist application. In the different panels, statistical significance was assessed with repeated-measures 1-way-ANOVA in combination with Tukey’s post test, and is indicated with *p<0.05 where applicable. All records shown were obtained in HCS except the neuron in panel A.
Figure 7.
Immunoblot detection of TREK/TRAAK family two-pore domain potassium channels in hippocampal tissue and hippocampal cultures.
Western blots of TREK-1, TREK-2 and TRAAK from mouse hippocampus tissue (Hippo), HEK293 cells transfected with TREK-1 (HEK+TREK-1), TREK-2 (HEK+TREK-2), TRAAK (HEK+TRAAK) and hippocampal cultured cells (Cell Hippo). A, Upper panel, hippocampus (20 µg of protein) was probed with an anti-TREK-1 antibody with (+) or without (−) pre-absortion with the microsomal fraction of transfected HEK293 cells expressing TREK-2 and TRAAK, as described in Materials and Methods. HEK293 cells transfected with TREK-1 (10 µg of protein) and hippocampal cultured cells (20 µg of protein) were probed with the pre-absorbed anti-TREK-1. Lower panel, the pre-absorbed anti-TREK-1 was incubated with the corresponding antigenic peptide. B, Upper panel, hippocampus (20 µg of protein) was probed with (+) or without (−) pre-absortion with microsomal fraction of transfected HEK293 cells expressing TREK-1 and TRAAK, HEK293 cells transfected with TREK-2 (10 µg of protein) and hippocampal cultured cells (20 µg protein) were probed with the pre-absorbed anti-TREK-2. Lower panel, the pre-absorbed anti-TREK-2 was incubated with the corresponding antigenic peptide. C. Upper panel, hippocampus (20 µg of protein) was probed with an anti-TRAAK antibody with (+) or without(−) pre-absortion with microsomal fraction of transfected HEK293 cells expressing TREK-1 and TREK-2. HEK293 cells transfected with TRAAK (5 µg of protein) and hippocampal cultured cells (20 µg of protein) were probed with the pre-absorbed anti-TRAAK. Lower panel, the pre-absorbed anti-TRAAK was incubated with the antigenic peptide. D The levels of the major bands for each antibody in the hippocampus tissue and hippocampal cultured cells were calculated from the pixel intensity values (minus background) normalized to the pixel intensity values of HEK293- transfected cells and presented as optical density (OD). In this figure same batch of HEK293 transfected cells was used for hippocampus and cultured hippocampal cells, same batch of hippocampus tissue and hippocampal cultured cells were used to probe each antibody. Data correspond to one representative experiment of several experiments using different batches of samples with essentially the same results.
Figure 8.
Sensitivity of cooling-evoked responses to arachidonic acid, riluzole, chloroform and nifedipine supports the involvement of TREK channels.
Figure 9.
Chloroform blocks the effects of cold temperature on spontaneous ictal activity and evoked field EPSP amplitude in hippocampal slices.
A, Sample experiment illustrating the increase in CA3 spontaneous ictal activity caused by moderate, progressive cooling, in a disinhibited hippocampal slice. The trace has been filtered at 50 Hz (low-pass). B, Averaged values (n = 6 slices) of ictal activity, quantified as number of ictal spikes following each individual SPW-R event, at different temperatures. Values have been normalized to the ictal activity measured at 36°C (*p<0.05, one-way ANOVA). C, Sample recordings of SPW-R events followed by ictal activity in a representative experiment before, during and after mild cooling, in control conditions and in 20 mM CHCl3. D, Summary of ictal activity associated to SPW-R events (n = 11 slices from 7 mice for controls; n = 7 slices from 5 mice for CHCl3). Cooling from 37 to 34°C increased ictal activity reversibly and CHCl3 abolished such increase. E, Representative traces of evoked fEPSPs recorded in CA1 area at 30°C (blue), 37°C (black) and 40°C (red) in control, in 20 mM CHCl3 and during washout. Each trace is the average from 5 consecutive fEPSPs. F, Average of 5 consecutive traces, illustrating the effect of temperature on fEPSP amplitude. CHCl3 (20 mM) applied at 37°C diminished basal fEPSP amplitude and blocked further increases in fEPSP size caused by lowering temperature. (n = 9 slices from 6 mice for control, n = 6 slices from 5 mice for CHCl3 and washout). The effects of CHCl3 were fully reversible. For clarity, error bars have been removed for the mean values obtained during wash. A two-way ANOVA showed significant differences (p<0.001) between control and CHCl3 recordings when including data from all temperature values in the analysis. Posthoc analysis showed significant differences (*p<0.05) between amplitudes at 30 and 37°C in control and wash, with no differences in CHCl3.