Fig. 2
From: Cortical recruitment determines learning dynamics and strategy
Spectro-temporal differences impact on cortical recruitment. a Example field of view illustrating GCAMP6s labeling of L2/3 auditory cortex neurons. b Result of automated cell segmentation run on the data acquired in the example shown in a. c Example single trial responses to sounds (different colors) for two neurons (top and bottom). Gray bars = sound duration. d Population responses (n = 27 sessions, 15,511 neurons in 6 mice) to Sound B (red) and A (blue). Both normalized fluorescence (light colors) and deconvolved (dark colors) calcium signals are shown. e Mean deconvolved signal and fraction of significantly responding neurons to sounds A, B, and C. Mean calcium responses to sound A (0.05 ± 0.03% ΔF/F.s-1, mean±SEM) were significantly smaller than to B (0.10 ± 0.02% ΔF/F.s-1) and C (0.12 ± 0.02% ΔF/F.s-1; sign test, p = 0.0008 and p = 0.026, n = 27 sessions, 15,511 neurons in 6 mice). Sound B and C also activated a larger fraction of neurons (24 and 25%; two-sided Signed test across 20 sound repetitions, p < 0.05) than A (18%; χ2 test, p = 10–41 and 6 × 10–31, n = 15,511). f Population response reliability (diagonal) and similarity (off-diagonal) matrix for sounds A, B, and C. The pair-wise discriminability value, computed with a linear classifier is indicated in red. g Mean deconvolved calcium signals for 6757 auditory cortex neurons in 12 awake mice during 29 calcium imaging sessions for 2-s-long white noise sounds modulated in intensity between 60 dB and 85 dB upwards and downwards (mean between 0 and 0.5 s after sound onset for up: 0.80 ± 0.08% vs down: 0.62 ± 0.11%, Signed test p = 3.75 × 10–4, n = 29 sites, in 12 mice; mean between 0 and 2.5 s: 0.166 ± 0.58% vs down: 0.157 ± 0.57%, Signed test, p = 0.034). The inset show the significantly different (χ2 test, p = 10–68) fraction of responsive cells (75 and 61%; two-sided Signed test, p < 0.05). h Mean deconvolved calcium signals for 59,590 auditory cortex neurons in seven awake mice during 60 calcium imaging sessions for white noise sounds modulated in intensity at 1Hz and at 20Hz (0.357 + −0.032% and 0.15 + − 0.034% ΔF/F.s−1, signed test, p = 0.0009). The inset show the significantly different (χ2 test, p = 10–189) fraction of responsive cells (37 and 29%; two-sided Signed test, p < 0.05). Error bars represent standard errors (SEM)
