Multisensory integration in neurons of the medial pulvinar of macaque monkey

Abstract

The pulvinar is a heterogeneous thalamic nucleus, which is well developed in primates. One of its subdivisions, the medial pulvinar, is connected to many cortical areas, including the visual, auditory, and somatosensory cortices, as well as with multisensory areas and premotor areas. However, except for the visual modality, little is known about its sensory functions. A hypothesis is that, as a region of convergence of information from different sensory modalities, the medial pulvinar plays a role in multisensory integration. To test this hypothesis, 2 macaque monkeys were trained to a fixation task and the responses of single-units to visual, auditory, and auditory-visual stimuli were examined. Analysis revealed auditory, visual, and multisensory neurons in the medial pulvinar. It also revealed multisensory integration in this structure, mainly suppressive (the audiovisual response is less than the strongest unisensory response) and subadditive (the audiovisual response is less than the sum of the auditory and the visual responses). These findings suggest that the medial pulvinar is involved in multisensory integration.

Keywords: audiovisual; medial pulvinar; multisensory; single-units.

Fig. 1

Experimental design. A) The monkey sat in a primate chair in front of a computer screen. After fixation of the central point (for a random duration comprised between 500 and 1200 ms), an auditory, visual, or audiovisual stimulus was presented for 250 ms. The monkey had to maintain fixation during stimulus presentation to obtain a reward. B) Three static visual stimuli were presented (always the same picture for each): A picture of random dots, a macaque face and a rattlesnake. C) T3 MRI slices of the brain of monkey C in a frontal section (top, left) and in a sagittal section (top, right) showing electrode position. Drawings below MRI pictures show the cerebral regions delimited by the boxes in the MRI slices, labeled according to macaque brain atlases (Paxinos et al. 2000; Saleem and Logothetis 2007). PuM: Medial pulvinar, PuL: Lateral pulvinar, PuI: Inferior pulvinar, PuA: Anterior pulvinar, BrSC: Superior colliculus brachium, LV: Lateral ventricle, cc: Corpus callosum.

Fig. 2

Global responses and individual neuron examples. a and b) Global PSTHs representing the mean firing rate (sp/s) of neuronal populations according to the response types: increase (A) or decrease (B) in firing rate. Significant responses were averaged over all stimuli and all modalities. Since a given neuron could yield between 1 and 15 significant responses (9 individual stimuli, 3 modalities with all stimuli pooled, 3 categories with all modalities pooled), the total number of responses (159 + 170) is larger than the number of neurons with significant responses (213). Black line corresponds to the mean and gray shading to ±95% CI. c–e) Examples of auditory neuron (C), visual neuron (D), and classic bimodal audiovisual neuron (E). Top: PSTHs were calculated across all trials (all stimulus categories pooled) for a given modality (auditory in light gray, visual in gray, and audiovisual in black), with a bin width of 10 ms and smoothed with a 45-ms moving average. Bottom: Boxplots summarizing distributions of mean firing rates per trial for each stimulus presentation in response to auditory (A), visual (V), and audiovisual (AV) stimuli. Paired spontaneous activity measured during 500 ms before stimulus onset has been subtracted from each response. Responses were considered as significant when the distribution of firing rates during stimulus presentation was different from the distribution of preceding spontaneous firing rates, as tested with a Wilcoxon signed rank test. Sharps below boxplots indicate significance level: #: P < 0.05, ##: P < 0.01, and ###: P < 0.001. Modulation of response amplitude by the modalities was evaluated by the Kruskal–Wallis test, followed by Mann–Whitney tests between modality (P-values adjusted by Bonferroni correction). Asterisks above boxplots indicate significance level: ^*: P < 0.05, ^**: P < 0.01, and ^***: P < 0.001.

Fig. 3

Proportions of single-units classified by sensory modalities and multimodal integration types. A) Pie chart representing the proportions of single-units classified as unimodal visual, unimodal auditory, classic audiovisual, complex audiovisual, or nonspecific to a modality. B) Pie chart representing the proportions of complex audiovisual neurons, classified according to the modalities to which they responded.

Fig. 4

Example of a complex audiovisual neuron. First column: Responses to noise stimuli; second column: Responses to macaque stimuli; third column: Responses to snake stimuli. Boxplots summarizing distributions of mean firing rate per trial (spontaneous activity subtracted) for each stimulus are shown below each corresponding PSTHs (auditory in light gray, visual in gray and audiovisual in black). This complex neuron was assigned to the visual modality for macaque stimulation and to the multisensory modality for noise and snake stimuli. Significance of responses was tested using Wilcoxon test (#: P < 0.05, ##: P < 0.01, and ###: P < 0.001). There was no significant modulation (enhancement or suppression) with the bimodal stimulation (Mann–Whitney test, adjusted P-values > 0.05).

Fig. 5

Neuronal selectivity and sparseness. The SPIs of the 213 responding cells are plotted as a function of the SEIs (bottom left). Indexes distributions are shown on the top (SEI) and on the right (SPI). A SPI close to 1 is indicative of a neuron responding similarly to all stimuli. The SEI indicates whether the neuron responded to only one stimulus (index close to 1) or whether it responded equally well to the preferred and least preferred stimulus (index close to 0).

Fig. 6

Global responses of neuronal populations according to the response types. Global PSTHs representing the average firing rate (sp/s) of neuronal populations according to the responses types, for each stimulus and modality: Excitatory (A) or inhibitory (B) responses. Each row of PSTHs (top to bottom) corresponds to auditory, visual, audiovisual, and pooled modality conditions. Each column of PSTHs (left to right) corresponds to noise, macaque, snake, and pooled stimuli conditions. For each PSTH, we only used the neurons showing a significant response to the corresponding stimulus (or pooled stimuli). Responses were averaged over all neurons showing a statistically significant excitatory (A) or inhibitory (B) response for each stimulus, without (white background) or after (gray background) pooling across stimuli or modalities. The mean spontaneous activity has been subtracted before averaging. The PSTH with the dark gray background represents population average for all enhanced or all decreased responses (all the responses used for the PSTHs with white backgrounds) (same as Fig. 2A, B). Black line corresponds to the mean and gray shading to ±95% CI.

Fig. 7

Multisensory integration: additivity and AMIs. The AMIs of 96 multisensory cells are plotted as a function of ADIs (bottom left). Indexes distributions are shown on the top (ADI) and on the right (AMI). The brackets are merely indicative of the meaning of the index values and are not intended to indicate statistical significance.

Fig. 8

Latencies of medial pulvinar neurons. A) Cumulative distribution of half-rise latencies for each modality: Visual (V, 30 latencies) in gray, auditory (A, 35 latencies) in light gray, and audiovisual (AV, 46 latencies) in black. In order to have a single value per neuron and modality, only the shortest latency within each modality was taken into account. B) Distribution of half-rise latencies for visual responses produced by the presentation of noise (n = 13), macaque (n = 18), and snake (n = 15) pictures. Dots represent individual data, horizontal central bars correspond to the medians, and error bars indicate the 95% confidence interval of the medians. Latencies were significantly longer for responses to macaque stimuli. Asterisks indicate significance level obtained with Posthoc Tukey’s test: ^*: P < 0.05; ^**: P < 0.01.

Fig. 9

Sensory modalities are intermingled in the medial pulvinar. Topographic representation of neuronal responses to each modality. Representation of response modalities of the 213 cells with significant responses as a function of the anteroposterior (y axis, from 2 to 6 mm) and mediolateral (x axis, from 5 to 9 mm) coordinates of electrode penetration sites. Response modality is gray-scale-coded: Light gray corresponds to auditory neurons, dark gray to visual neurons and black to classic and complex audiovisual neurons pooled together. The radius of the circles is proportional to the number of neurons that responded to each modality.