Rapid and cumulative adult plasticity in the mouse visual cortex

Abstract

Experience-dependent neural plasticity enables the brain to adapt to diverse and dynamic environments by reshaping circuits. In the adult visual system, this plasticity can be elicited by repeated sensory stimuli; however, its temporal dynamics and underlying mechanisms remain unclear. Here, we investigated the regulation of visual response potentiation induced by repeated light flashes in the primary visual cortex of awake adult mice. Our findings revealed two distinct temporal phases of potentiation: a rapid phase occurring within seconds and a cumulative phase developing over hours to days. Notably, the identification of this rapid phase phenomenon adds to and refines the prevailing view that visual plasticity in the adult cortex is predominantly slow. Additionally, exposure to visual stimuli enhanced spontaneous slow-wave activity in the visual cortex during non-REM sleep. This plasticity was significantly impaired in Grin2a (NR2A) knockout mice, a model of schizophrenia, which mirrors visual plasticity deficits observed in human patients. The dual temporal characteristics of flash-evoked visual plasticity likely reflect multifaceted aspects of adult brain functionality, encompassing processes related to memory, learning, and neurological disorders. This model of visual plasticity in defined neural circuits provides a simplified yet robust and extensible framework for exploring the neural mechanisms underlying adaptive and maladaptive behavioral changes.

Keywords: NMDA receptors; NREM sleep; experience-dependent adult plasticity; flash-evoked potentials; mouse visual cortex; stimulus-selective response plasticity (SRP).

Figure 1

Flashing light visual stimulation of awake mice (A) Experimental setup for visual stimulation in a head-fixed awake mouse. Optical fibers connected to LEDs were positioned in front of each eye. (B) Local field potentials (LFPs) and multiunit activity (MUA) were recorded from the binocular zone of the primary visual cortex (V1). Flashing light stimuli (10 msec duration) were alternately delivered to the right or left eye at 1 s intervals. (C) Representative traces showing somatosensory EEG, cervical EMG, and V1 LFP (right hemisphere) during visual stimulation (ticks indicate the stimulation events). Stimulation of either eye evoked visual responses in both ipsilateral and contralateral visual cortices. (D) Representative visually evoked potential (VEP) in V1 (averaged LFP responses to 300 stimulations per eye). N1, N2, and N3 represent the first, second, and third negative peaks, respectively, while P1 and P2 represent the first and second positive peaks, respectively. (E) Timeline of the visual experiments. Visual stimulation sessions consisted of alternating presentations to the left and right eyes, with 100 presentations per eye for a total of 200 presentations (200 s). Each day included 3 sessions, conducted over at least 4 days.

Figure 2

Potentiation of response amplitudes induced by repetitive visual stimulation (A) Concurrent recordings of somatosensory EEG and visual cortex LFP over 6 consecutive days. Each day, 200 visual stimulations (10 msec light flashes) were provided to each eye (400 stimulations in total). (B) Time course of color-scaled individual visual responses in V1, showing responses to 200 flashes per day over 6 days (total 1,200 flashes). Yellow and blue indicate the positive and negative VEP peaks, respectively. (C) Representative VEPs evoked by flashing light in V1 on day 1 and day 6. (D) Comparison of VEP peak amplitudes on days 1 and 5 across 8 mice, showing responses in the right and left binocular cortices to contralateral and ipsilateral inputs. Mann–Whitney test, N1:15 recording sites, 30 recordings of contralateral or ipsilateral eye stimulation, p = 0.4001; P1:15 recording sites, 30 recordings, **p = 0.0080; N2:15 recording sites, 30 recordings, **p = 0.0089; P2:14 recording sites, 27 and 28 recordings, *p = 0.0450; N3:14 recording sites, 27 and 26 recordings, ***p = 0.0007; P3:6 recording sites, 12 and 7 recording sites, p = 0.1956. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001. (E) Potentiation of VEP peak amplitudes following contralateral or ipsilateral eye visual stimulation. The ratios of day 5 to day 1 VEP peaks (1.0: no change) are plotted. Visual responses of the right and left cortices were averaged for each mouse. N = 8 mice, 7 and 8 recording sites. One-sample t-test, theoretical mean = 1. Contralateral responses, N1: p = 0.2259, t = 1.328, degree of freedom (df) = 7; P1: **p = 0.0013, t = 5.168, df = 7; N2: *p = 0.0035, t = 4.305, df = 7; P2: *p = 0.0162, t = 3.310, df = 6; N3: **p = 0.0049, t = 4.340, df = 6. Ipsilateral responses, N1: p = 0.2590, t = 1.228, df = 7; P1: p = 0.4229, t = 0.8510, df = 7; N2: p = 0.1014, t = 1.885, df = 7; P2: p = 0.0536, t = 2.396, df = 6; N3: *p = 0.0484, t = 2.386, df = 7. (F) Pooled VEP data (contralateral and ipsilateral responses are averaged for each mouse). In each animal, the VEP peak values were calculated as follows: the ipsilateral responses in the right-hemispheric visual cortex to right-eye stimulation and in the left-hemispheric visual cortex to left-eye stimulation were averaged for each mouse, as were the contralateral responses under both conditions. The individual averages for ipsilateral and contralateral responses were then averaged across all 8 animals. One-sample t-test, N1:p = 0.1769, t = 1.501, df = 7; P1: p = 0.2272, t = 1.324, df = 7; N2: *p = 0.0104, t = 3.472, df = 7; P2: **p = 0.0026, t = 4.567, df = 7; N3: **p = 0.0024, t = 4.626, df = 7. (G) Time course of VEP N2 amplitude (contralateral VEP in the left V1, blue line) in another set of animals (6 mice averaged) with daily visual stimulation (3 sets of 100 light pulses interleaved for 400 s) for 4–6 consecutive days. The average VEP amplitudes for each 100 pulses session are indicated by filled circles. (H) Average N2 amplitudes (every 100 stimulations of the contralateral input). N = 6 mice. Statistical comparisons were performed between the first and subsequent days (average of 300 stimulations). Paired t-test, day 1 vs. day 2, *p = 0.0197, t = 3.380, df = 5; day 1 vs. day 3, *p = 0.0161, t = 3.565, df = 5; day 1 vs. day 4: ***p = 0.0005, t = 7.924, df = 5.

Figure 3

Rapid growth of visual responses during stimulation (A) Visual responses during the first 200 stimulations (100 per eye) on day 1, with responses for the first 25 and last 38 stimulations. (B) Rapid increase in the VEP N2 peak amplitude within seconds of stimulation onset. N = 14 mice, 1 s stimulus intervals. (C) A significant increase in the V1 visual response (area under waveform) from the initial (0–10 pulses) to late (190–200 pulses) stimulation on day 1. Paired t-test, **p = 0.0054, 8 mice, t = 3.971, df = 7. At the beginning of the stimulation, individual visual responses were weak; thus, the area of the waveforms was quantified instead of the N2 amplitude.

Figure 4

Flash-evoked visual plasticity (FVP) in visual processing regions (A) LFP recording sites in the somatosensory cortex, anterior cingulate cortex (ACC), V1 (medial and lateral binocular regions) and V2LM. (B) (Left) VEP changes from the initial session (day 1, 300 pulses, gray line) to the late session (day 6, 300 pulses, blue line) in the ACC (4 mice). (Right) VEP changes from the initial session (day 1, 300 pulses, gray line) to the late session (day 6, 300 pulses, blue line) in V1BL (6 mice). The LFPs of the ACC and V1BL were simultaneously recorded. (C) Day 6 to day 1 ratio of VEP N2 peaks (ipsilateral and contralateral responses) across the regions. Wilcoxon signed-rank test, N = 6 mice. ACC: 4 recording sites, p = 0.8750; V1M, 19 recording sites; ***p = 0.0001; V1L: 21 recording sites, ***p = 0.0001; V2LM: 12 recording sites, ***p = 0.0005.

Figure 5

Eye input specificity for FVP (A) Sequential visual stimulation of one eye (left panel, 1st period, 300 pulses/day for 4–5 days) and then to the other eye (right panel, 2nd period, 300 pulses/day for 4–5 days) in the same mice. (B) Representative mouse data are shown. Stimulation of the right eye (blue) activates both the contralateral left visual cortex and the ipsilateral right visual cortex. After stimulating the right eye for only 4 days, the left eye was stimulated exclusively for another 4 days. Horizontal bars represent one session of visual stimulation (100 trials). (C) Both visual stimulation sequences induce potentiation at the same recording sites. VEP peak N2 amplitudes of 3 mice and visual responses of the right and left cortices to contralateral and ipsilateral eye inputs were included (12 samples). Paired t-test, 3 mice, 12 recording sites, 1st period: **p = 0.007, t = 4.625, df = 11; 2nd period: ****p < 0.0001, t = 10.06, df = 11.

Figure 6

Neuronal spiking activity associated with VEP (A) Example of multiunit activity (MUA) from V1 (putative pyramidal neurons, 58,120 spikes overlaid). (B) Perievent histogram of MUA peaks M1 and M2, corresponding to VEP peaks N1 and N2 (average of 300 pulses). (C) Paired t-test, day 1 vs. day 4, MUA M2:12 recording sites (ipsilateral and contralateral responses), 17 MUAs, t = 3.470, df = 16, **p = 0.0032; VEP N2:24 recording sites (ipsilateral and contralateral responses), t = 7.685, df = 23. ****p < 0.0001.

Figure 7

Decoding brain responses using a support vector machine (SVM) classifier (A) Estimation of eye input (right or left) from single brain responses using SVM. (B) VEP parameters (positive peak, negative peak, and area) used for the SVM classification. The time windows corresponding to each peak of VEPs were defined as follows: N1: 0–100 msec, N2: 100–300 msec, N3: 300–500 msec, P1: 0–200 msec, P2: 200–400 msec, and Area (shaded waveform area): 0–500 msec after stimulus onset (t = 0). (C) Classification accuracy using data from single or combinations of electrodes listed in the panel below. R: right hemisphere; L: left hemisphere. N = 6 mice. 50% is chance level. (D) Comparison of initial (300 pulses, day 1) and late sessions (300 pulses, day 4). N = 6 mice, Paired t-test, N1: *p = 0.0107, t = 3.965, df = 5; N2: **p = 0.0092, t = 4.118, df = 5; N3: p = 0.9825, t = 0.02304, df = 5; P1: *p = 0.0146, t = 3.661, df = 5; P2: P2 = 0.0760, t = 2.231, df = 5; Area: **p = 0.0091, t = 4.130, df = 5. Note that the simultaneously recorded visual responses of V1 (medial and lateral binocular regions, both hemispheres) and V2LM were used to train the SVM. (E) Accuracy using LFP waveform areas under the curve with different time windows ranging from 0 to 500 msec. Accuracies were calculated using information from all electrodes (6 mice).

Figure 8

NMDA receptor-dependent FVP (A) VEPs in NR2A knockout mice on days 1 and 6 (contralateral eye stimulation). (B) Absence of changes in VEP peaks (day 1 vs. day 6) in NR2A knockout mice after visual stimulation (N = 6 mice, 12 recording sites, 24 recordings of contralateral or ipsilateral eye stimulation). Mann–Whitney test, N1: p = 0.5775; P1: p = 0.5844; N2: p = 0.4964. (C) VEPs of wild-type mice injected with MK801 intraperitoneally (0.5 mg/kg body weight) (day 1 vs. day 6, contralateral eye stimulation) 30 min before the start of visual stimulation. (D) Absence of FVP (day 1 vs. day 6) in MK801 treated wild type mice after visual stimulation. VEP peak N2 amplitudes (N = 6 mice, 12 recording sites, 24 recordings of contralateral or ipsilateral eye stimulation). Mann–Whitney test, N1: p = 0.1305; P1: p = 0.4552, N2: p = 0.1060. (E) Impaired rapid potentiation in NR2A KO mice. Visual response (area under waveform) during the initial stimulation (0–10 pulses) vs. late stimulation (190–200 pulses) on day 1. Paired t-test, p = 0.924, 6 mice, t = 0.1002, df = 5.

Figure 9

FVP in freely behaving mice (A) Visual stimulation (10 msec duration, 100 pulses at 1 Hz, 3 times of 100 pulses interleaved for 400 s per day) was applied to an awake behaving mouse from the ceiling of a recording chamber. (B) Representative LFP visual responses of a behaving mouse for consecutive 4 days. (C) VEP (day 1 vs. day 4) of a behaving mouse after visual stimulation. (D) Time course of the VEP N2 amplitude of a behaving mouse across visual stimulations (300 pulses/day for 4 days). (E) A significant increase in VEP N2 amplitudes (day 1 vs. day 4) in behaving mice after visual stimulation. N = 6 mice, paired t-test, *p = 0.0215, t = 3.299, df = 5.

Figure 10

Increased slow-wave activity in V1 during non-REM sleep after FVP (A) (Left) Spontaneous 3–6 Hz spike-and-wave-like activity in V1 during waking. Spontaneous activity in V1 is sporadic and localized. (Right) Visually evoked activity in the same animal. (B) Increased frequency of spontaneous activity (day1 vs. day 4) after visual stimulation. N = 6 head-fixed mice, paired t-test, **p = 0.0072, t = 4.374, df = 5. (C) Time course of slow-wave activity (delta power, 0.5–4 Hz, 60 s bin) in a behaving mouse before, during, and after visual stimulation sessions. EEG and LFP delta activity did not include data when visual stimulation was provided and data during waking. Chronic recording was performed under light and dark (12:12) cycles over 8 days. (c_1) Somatosensory EEG delta activity. (c_2) V1 LFP delta activity concurrently recorded with EEG. (c_3) Normalized V1 LFP delta activity (divided by the EEG delta activity). Dotted horizontal line for visual inspection. (c_4) Visual stimulation. Each vertical bar indicates 300 pulse stimulations (10 msec duration, 1 Hz) that were interleaved for 3 h. (D) A significant increase in delta activity during sleep after visual stimulation (baseline before visual stimulations vs. days 4–10 after visual stimulation). N = 6 behaving mice, paired t-test, *p = 0.0374, t = 2.813, df = 5. (E) (Top) Theta activity (6–10 Hz) during REM sleep in the same animals. N = 6 behaving mice, paired t-test, p = 0.5898, t = 0.5756, df = 5. (Bottom) Gamma activity (40–60 Hz) during waking in the same animals, 6 behaving mice, paired t-test, p = 0.2327, t = 1.358, df = 5.

Figure 11

Two-process model of FVP schematic illustration of rapid and cumulative phases of FVP.