Article Figures & Data
Figures
- Figure 1.
Auditory-evoked Ca2+ responses to CS+ are enhanced in tuft dendrites. a, Schematic of the experimental design and of the auditory fear conditioning protocol. b, Freezing scores to CS+ and CS– 24 h after conditioning (n = 20 mice; paired t test). c, Schematic of the experimental design. Example field of view (FOV) of tuft dendrites from L2/3 pyramidal neurons in the auditory cortex expressing GCaMP6f. Scale bar: 10 μm; 5 μm (inset). d, Top, Heatmaps of Ca2+ transients in response to CS+ (left) and CS– (right) in an example tuft dendrite. White bars indicate timing of the tones (5 × 500 ms). Bottom, Trial-averaged Ca2+ responses to CS+ (cyan) and CS– (magenta). Bold line, Average response; light shading, SEM. e, Overlay of the trial-averaged Ca2+ responses to CS+ (cyan) and CS– (magenta) from the same tuft dendrite shown in d. Gray bars show the tones in the auditory stimuli. f, Peak amplitude of the trial-averaged Ca2+ responses to CS+ (0.74 ± 0.05 ΔF/F) and CS– (0.58 ± 0.05 ΔF/F). Mann–Whitney test. Inset, Peak amplitude of the trial-averaged Ca2+ responses in naive mice were not different (see also Extended Data Fig. 1-1). g, Integral of the trial-averaged Ca2+ responses to CS+ (1.48 ± 0.05 ΔF/F*s), CS– (1.12 ± 0.05 ΔF/F*s), and a reference stimulus (10 kHz; 1.33 ± 0.08 ΔF/F*s; Extended Data Fig. 1-2). Mann–Whitney test. All data represented as mean ± SEM. ns: p > 0.05, * p < 0.05, ** p < 0.01, **** p < 0.0001. In a subset of experiments, tuft dendrites were traced to their soma of origin to measure the correlation of activity across dendritic branches of the same neuron (Extended Data Fig. 1-3).
- Figure 2.
CS+ and CS– evoke similar responses in basal dendrites. a, Top, Schematic of the experimental design. Example FOV of basal dendrites from L2/3 pyramidal neurons expressing GCaMP6f (located 100–250 μm below pia). Scale bar: 10 μm. Inset, Zoom image of basal dendrite. Scale bar: 5 μm. b, Heatmaps of Ca2+ transients in response to CS+ (left) and CS– (right) in an example basal dendrite. White bars indicate timing of the tones (5 × 500 ms). Bottom, Trial-averaged Ca2+ responses to CS+ (cyan) and CS– (magenta). Bold line, Average response; light shading, SEM. c, Overlay of the trial-averaged Ca2+ responses to CS+ (cyan) and CS– (magenta) from the same basal dendrite shown in b. Gray bars show the tones in the auditory stimuli. Inset, Peak amplitude of the trial-averaged Ca2+ responses to CS+ (0.78 ± 0.04 ΔF/F) and CS– (0.90 ± 0.05 ΔF/F). Mann–Whitney test. d, Integral of the trial-averaged ΔF/F responses to CS+ (CS+: 1.51 ± 0.07 ΔF/F*s), CS– (1.60 ± 0.05 ΔF/F*s), and a reference stimulus (10 kHz; 1.56 ± 0.05 ΔF/F*s; Extended Data Fig. 2-1). Mann–Whitney test. All data represented as mean ± SEM. ns: p > 0.05.
- Figure 3.
Action potential output is increased during CS+. a, Schematic of the experimental design. In vivo patch clamp recordings were performed in the auditory cortex of urethane anaesthetized mice following auditory fear conditioning. b, Raster plot of somatic action potentials (top), overlay of voltage responses (middle), and average subthreshold response (bottom) in an example L2/3 pyramidal neuron during CS+ (cyan) and CS– (magenta). c, Integral of the subthreshold voltage response to each tone during CS+ (cyan) and CS– (magenta). Wilcoxon matched-pairs signed-rank test. d, Average somatic firing rate during each tone in the CS+ (cyan) and CS– (magenta). Wilcoxon matched-pairs signed-rank test. e, Average somatic firing rate during each tone in the CS+ (cyan) and CS– (magenta) after local cortical application of the NMDA channel agonist, APV. Wilcoxon matched-pairs signed-rank test. Bars represent mean ± SEM. ns: p > 0.05, * p < 0.05.
- Figure 4.
Control of the somatic output by tuft dendrites. a, Schematic of the simulation design. A reconstructed L2/3 pyramidal neuron was used for a multicompartment model. Background synapses (green dots) were distributed uniformly across the neuron, while auditory-evoked synapses were distributed uniformly across the basal (blue dots) or tuft dendrites (red dots; see also Extended Data Fig. 4-1). b, To simulate auditory-evoked responses in this L2/3 neuron, we put either 235 tuft and 200 basal (red traces), or 235 basal and 200 tuft (blue traces) auditory-evoked synapses during a 5 × 500 ms stimulation window. Example membrane potentials during these two simulation conditions are shown for two tuft locations (labels t1 and t2 in panel a), the soma, and for one basal location (label b in panel a). c, To see how changes in tuft and basal inputs change the output of the neuron, we simulated an increasing number of auditory-evoked tuft inputs with a constant 200 basal auditory-evoked inputs (red) and an increasing number of auditory-evoked basal inputs with a constant 200 tuft auditory evoked-inputs (blue). Error bars are SEM. Dotted lines denote the experimentally determined mean firing rates for the CS+ (cyan) and CS– (magenta) conditions. d, Schematic of the reduced model using two sigmoids to represent tuft inputs controlling the maximum (green) and threshold (orange) of the neuron input output function. e, Compared with CS–, the neuron input output function has a decreased threshold (orange) and increased maximum (green) in response CS+ (illustrated by arrows). f, The reduced model is able to recapitulate the simulation results from the multicompartmental model in c).
Extended Data
Extended Data Figure 4-1
Details of the L2/3 pyramidal neuron computational model. Please see materials section for further explanation. a, The reconstruction of the pyramidal neuron used in the model, with tuft and basal dendrites colored in red and orange, respectively. b, Examples of the location of background (left), basal (middle), and tuft (right), synapses used in the simulations. Each synapse is a combination of AMPA and NMDA conductances, and it is shown as a red dot. c, Table of conductances values used in the simulation. For the Ih channel in the tuft, conductance was a function of the distance from the soma, –0.8696 + 2.087 exp(x/323), where x was distance from the soma in microns. d, The somatic voltage response to 2-s-long current steps into the soma, from –1.0- to 0.7-nA steps of 0.1 nA. e, example action potential evoked in response to a current injection at the soma of 3 nA for 2 ms. f, The spike frequency output as a function of DC current steps into the soma. g, The steady state somatic voltage as a function of DC current into the soma. h, A total of 30 examples runs of only background synaptic inputs. Download Figure 4-1, EPS file.
Extended Data Figure 1-3
Fear learning did not alter the correlation of Ca2+ transients across tuft dendrites of the same neuron. a, 3D reconstruction (bottom) of a L2/3 pyramidal neurons (N1) from an example FOV (top). ROIs from the same neuron are highlighted in green. b, Ca2+ transients recorded in tuft dendrites from the reconstructed neuron in a (Roi#10 and Roi#6; N1). Orange dots, Evoked local Ca2+ transients. c, The correlation of Ca2+ responses in tuft dendrites from the same neuron is similar during CS+ (0.69 ± 0.07) and CS– (0.74 ± 0.06; p = 0.51). Mann–Whitney test. All values represent mean ± SEM. Download Figure 1-3, EPS file.
Extended Data Figure 1-1
The auditory stimuli used during fear conditioning evoked similar activity in the tuft dendrites of naive mice. a, Two-photon calcium imaging was performed in tuft dendrites in response to the same stimuli used during fear condition-ing (5 × 500 ms, 5 or 15 kHz tones) in naive mice (n = 3) that did not experience the fear conditioning protocol. b, Heatmaps of Ca2+ responses in an example tuft dendrite to both 5 and 15 kHz (top) and trial-averaged Ca2+ activity (bottom). c, The peak amplitude of the trial-averaged Ca2+ response to the auditory stimuli was not significantly different (5 kHz, 0.73 ± 0.08 vs 15 kHz, 0.70 ± 0.08; p = 0.52; 3 mice; Mann–Whitney test). Download Figure 1-1, EPS file.
Extended Data Figure 1-2
Activity in tuft dendrites evoked by a reference stimulus (10 kHz). a, Schematic of the experiment: dendritic recordings were performed from tuft dendrites of L2/3 pyramidal neurons in the auditory cortex. b, Top, Heatmap of Ca2+ responses in an example tuft dendrite to a reference auditory stimulus (5 × 500 ms, 10 kHz). Bottom, Trial-averaged Ca2+ response from the same tuft dendrite. Download Figure 1-2, EPS file.
Extended Data Figure 2-1
Activity in basal dendrites evoked by a reference stimulus (10 kHz). a, Schematic of the experiment: dendritic recordings were performed from basal dendrites of L2/3 pyramidal neurons in the auditory cortex. b, Top, Heatmap of Ca2+ responses in an example basal dendrite to a reference auditory stimulus (5 × 500 ms, 10 kHz). Bottom, Average Ca2+ response from the same basal dendrite. c, Histogram distribution of the peak amplitude values of all the Ca2+ transients for tuft and basal dendrites evoked by the reference stimulus (10 kHz; all tones; p = 0.34 Mann–Whitney test). Download Figure 2-1, EPS file.
In this issue
