Figure 1.
Two-alternative spatial discrimination task based on auditory and/or visual cues.
(A) Timing of task events. Nose poking into the central hole initiated a trial. After a variable foreperiod, a cue stimulus was delivered from the left or right, randomly chosen from visual, auditory and audiovisual stimuli. Rats responded to the stimulus by withdrawing from the central hole and selected the direction of the cue stimulus by poking their heads into the hole ipsilateral to the stimulus. (B) Mean reaction time for each type of stimulus across sessions and rats (*p<0.001 in ANOVA and post hoc Tukey test). Error bars, standard deviation.
Figure 2.
Direction preference preceding locomotion.
(A) Example of tetrode isolation for single unit. (Top) Average waveforms of spikes recorded on the four tetrode channels correspond to a cluster in black in the scatter plots below. The scatter plots indicate the peaks of waveforms from channel 3 plotted against principal component 1 (PC1) from channel 3 recorded from a tetrode. (Right) Corresponding autocorrelation functions with a window of ±30 ms. The bin size is 0.1 ms. (B) Rasters and spike density functions (SDFs) aligned to the onset of the target for isolated single unit (black cluster) during spatial discrimination task. The data were derived from only correct trials in responses to contralateral stimuli divided into visual, auditory and audiovisual stimulus conditions. Trials in the raster plot are sorted according to reaction time. (C) The data of the same neuron as above were aligned to movement onset and divided into ipsiversive and contraversive movement trials regardless of the modality of the stimulus. (D) Direction preference index of above unit was calculated using receiver operating characteristics (ROC) analysis for each time point (p<0.05, permutation test). (E) Direction preference curves for all cells (225 cells). Each row corresponds to one cell. Cells were sorted by the time of center of mass of significant positive preferences. Red and blue colors indicate preference indices with significant positive and negative values (p<0.05, permutation test), respectively, and gray points correspond to insignificant values. (F) Distribution of direction preference index in premovement period across population (225 cells). (G) Preference calculated for correct trials plotted against preference calculated for erroneous trials in premovement period. (H) Direction preferences calculated for each stimulus modality in each neuron and were averaged for each stimulus modality condition. The blue period shows a significant modulation of the preference under the multisensory condition (AV>unimodal, ANOVA with post hoc tukey test, p<0.05).
Figure 3.
Correlation between reaction speed and firing rates.
(A) Raster and spike density function (SDF) for example cell showing negative correlation between behavioral reaction time and firing rate in premovement period. (B) Scatter plots of reaction time and spike density in premovement period of cell. (C) Distribution of correlation coefficients between reaction time and firing rate in the premovement period for contraversive (left) and ipsiversive movements (right).
Figure 4.
Typical examples and their population average of 1–4 types of neuron for contraversive (A–B) and ipsiversive movements (C–D). Trials were sorted according to reaction time and divided into faster half and slower half reaction time trials by the median reaction time and SDFs were averaged for each reaction time group (A, C). The neurons shown in the type 1 and type 2 are the same as those shown in Fig. 2A–C and Fig. 3A, respectively. Population average of mean normalized SDFs of neurons that satisfied criteria (see Methods) for contraversive (B) and ipsiversive movements (D). * p<0.05, ** p<0.001 in paired t-test. (E) Scatter plots indicating the premovement activity index plotted against correlation coefficient between reaction time and firing rate in the premovement period. The premovement activity index is a normalized premovement activity so that 0 equals the spike density during the prestimulus baseline period (see Methods). The four colors of dots are used to code neuron types 1–4 that fall in one of four quadrants in the scatter plots with defined criteria (see Methods). (F) Percentage of each type of neuron for ipsiversive and contraversive movements (n = 171).
Figure 5.
Enhancement and depression in premovement activity by multisensory stimuli.
(A) Examples of neurons showing enhanced or depressed premovement activity for contraversive movement under audiovisual (AV) stimulus condition compared with those under visual (Vis) and auditory (Aud) stimulus conditions. MSI (multisensory modulation index, see Methods) in the premovement period for each neuron is shown. (B) Graphs show the distributions of MSIs in the premovement period for the ipsiversive and contraversive movements. (C) Mean MSI divided into cell type groups and directions of the movements, showing significant deviation from zero in some cell types (* p<0.05, ** p<0.01, *** p<0.001 in t-test).
Figure 6.
(A) Examples of a sensory-motor neuron showing visually-evoked responses. Raster plots and spike density functions (SDFs) were aligned to stimulus presentation. Trials in the raster plot were sorted according to reaction time. The data were derived from only correct trials in response to the stimuli presented to contralateral stimuli. (B) Comparisons of latencies and firing rates of sensory-evoked responses between unisensory and multisensory stimulus conditions in the multisensory cells. The dotted lines show latencies and mean firing rates of evoked responses for each cell (n = 29, 21 for auditory- and visually-evoked responses, respectively). The histograms and error bars show the means and standard error across cells, respectively. *p<0.05, **p<0.01 in paired t-test. (C) Distribution of correlation coefficients between reaction time and firing rate in the sensory-evoked response periods of visual or auditory responsible cells (n = 32, 17 cells for auditory and visual responses, respectively).
Figure 7.
Inverse effects of muscimol on ipsiversive and contraversive movements.
Distribution of reaction times for ipsiversive movement and contraversive movement after unilateral muscimol injections. Control is the mean reaction times of the pooled data from left and right movements after bilateral saline injections of each animal. The dotted lines denote mean reaction time for each injection condition of each animal. The circles and error bars denote the means and standard error across animals, respectively. *p<0.01, **p<0.001 in paired t-test comparisons using the Bonferroni correction.
Figure 8.
Intact multisensory facilitation after muscimol injections.
Mean reaction times for audiovisual stimuli were plotted against mean reaction times for corresponding faster unisensory stimuli for ipsilateral (blue) and contralateral choices. Filled circles indicate individual animals in which reaction times for audiovisual stimuli are shorter than those for unisensory stimuli (p<0.05, t-test). Error bars, ± standard error.