Single Neurons in the Insular Cortex of a Macaque Monkey Respond to Skin Brushing: Preliminary Data of the Possible Representation of Pleasant Touch

Abstract

Pleasant touch may serve as a foundation for affiliative behavior, providing a mechanism for the formation and maintenance of social bonds among conspecifics. In humans, this touch is usually referred to as the caress. Dynamic caressing performed on the hairy skin with a velocity of 1-10 cm/s is perceived as being pleasant and determines positive cardio-physiological effects. Furthermore, imaging human studies show that affiliative touch activates the posterior insular cortex (pIC). Recently, it was demonstrated that pleasant touch in monkeys (i.e., sweeping in a grooming-like manner) is performed with velocities similar to those characteristics of human caress (9.31 cm/s), and causes similarly positive autonomic effects, if performed with velocity of 5 cm/s and 10 cm/s, but not lower or higher. Due to similarities between the human caress and non-human primate sweeping, we investigated for the first time whether single neurons of the perisylvian regions (secondary somatosensory cortex [SII] and pIC) of a rhesus monkey can process sweeping touch differently depending on the stimulus speed. We applied stimulation with two speeds: one that optimally induces positive cardio-physiological effects in the monkey who receives it, and includes the real speed of sweep (5-15 cm/s, sweep fast), and a non-optimal speed (1-5 cm/s, sweep slow). The results show that single neurons of insular cortex differently encode the stimulus speed. In particular, even the majority of recorded somatosensory neurons (82.96%) did not discriminate the two speeds, a small set of neurons (16.59%) were modulated just during the sweep fast. These findings represent the first evidence that single neurons of the non-human primates insular cortex can code affiliative touch, highlighting the similarity between human and non-human primates' social touch systems. This study constitutes an important starting point to carry out deeper investigation on neuronal processing of pleasant sweeping in the central nervous system.

Keywords: Macaca mulatta; grooming; insular cortex; perisylvian region; pleasant touch; single neurons.

Figure 1

(A) Somatosensory task: Sweep Slow and Sweep Fast. The two sweeping were randomly applied, but for both the task was the following: (1) Green_On—Green_Off: the green light-emitting diode (LED) turned on for 0.5–2 s. We requested monkey to fixate the LED for all duration. (2) Red_On: the green LED switched to red and the experimenter moved the hand toward the monkey’s body part to be stimulated. We requested monkey to maintain the fixation. (3) Trg2_On—Trg2_Off: the experimenter touched the monkey (Trg2_On). The red LED switched off and the green LED was turned on until the sensory stimulation was finished (Trg2_Off). We requested monkey to maintain the fixation until the end of the sweeping. (4) Once the sweep had finished, the green LED switched off and after 0.8–1 s the monkey received drops of juice as a reward. After 1.5–2 s, the next trial began. From the end of stimulation to the start of the next trial we did not request the monkey to carry out any fixation task, but just to maintain the hands in their position. (B) The two epochs employed for statistical analyses. (1) Baseline: from 500 to 900 ms after Green_On (total of 400 ms); and (2) Stimulation: from Trg2_On to Trg2_Off. (For details see “Recording of Behavioral Events and Definition of Epochs of Interest” Section).

Figure 2

(A) Example of speed selective (SS) neuron (Fast Selective). The neuron was only modulated during the Sweep Fast. The neuron was recorded in the right hemisphere and the receptive field (RF) was located on the hairy side of the bilateral hand. (B) Example of SS neuron (Fast Selective). The neuron was modulated during both sweeping velocities, but the activity during the Sweep Fast was higher than that during the Sweep Slow. The neuron was recorded in the right hemisphere and the RF was located on the hairy side of the bilateral hand. (C) Slow Selective neuron, activated only during the Sweep Slow. The neuron was recorded in the left hemisphere and the RF was located on the bilateral hairy side of the mouth. (D) Example of speed unselective (SU) neuron. The neuron was modulated during both sweeping speeds, independently of the velocity. The neuron was recorded in the right hemisphere and the RF was located on the bilateral hairy side of the hand. The stimulation was performed on the contralateral side, from the distal (fingers) to the proximal part (near wrist). In (A–D) the histograms are aligned at the starting point of the stimulation, indicated by the light blue triangles. The end of the stimulation is indicated by the green triangles. The time between the red dots and yellow circles indicates the baseline, while the time between the light blue and green triangles represents the stimulation epoch. The histograms are normalized for the maximal spike/s value between the two conditions.

Figure 3

Distribution of the R-square values of the SS and SU neurons showing significant correlation between the speed of the stimulus and the firing rate. The lines indicate the median value for the SS (in green) and SU neurons (in blue). In order to investigate the potential similarity between the speed/firing rate correlation of SS neurons and the correlation showed by the SU neurons, we compared the distribution of R-square obtained in the two groups. The Mann-Whitney U Test indicates that the R-square of the SS neurons (median = 0.33) is higher (p < 0.001) than the R-square of the SU neurons (median = 0.18), thus speed variation account for a greater part of the variability of SS than SU neurons discharge.

Figure 4

Time course of the activity of neuronal populations aligned (gray line) on the tactile stimulation onset in the both conditions, Sweep Fast and Sweep Slow. Green lines indicate the time course of activity during Sweep Fast, while blue lines represent the time course of activity during Sweep Slow. On the Y axis is represented the normalized activity of the neuronal populations, while on the X axis is represented the time (sec). (A) The population of the SS neurons is activated stronger during the Sweep Fast compared with the Sweep Slow, in particular from 80 ms after alignment to 280 ms. (B) The population of the SU neurons showing correlation between the speed of stimulation and firing rate. From 40 ms after the alignment to 220 ms, the population is activated stronger during the Sweep Fast than Sweep Slow, while from 460 to 680 ms the activity is higher during the Sweep Slow than during the Sweep Fast. (C) The population of the SU neurons that did not show correlation between the speed of sweeping and the firing rate. The population does not discriminate between the speed of stimulation and the two tasks do not differ from each other. In (A–C) the activity is in alignment when the experimenter touched the monkey’s body. The median times of the end of sweep fast and sweep slow are indicated with the green and blue markers, respectively, above each population plot. Shaded areas around each marker represent the 25th and 75th percentile times. In (A,B) the light green shaded regions indicate the period in which the paired samples t-tests evidence a significant separation of the two curves (p < 0.05).

Figure 5

A 2D reconstruction of the lateral fissure of the left (A) and right (B) hemisphere, aligned to the middle of the insula; the continuous lines mark the lips of the sulcus, the border of the insula with the upper and lower bank of the sulcus, and the fundus. The blue area indicates the recorded region related to the SU neurons, while the green area indicates the recorded area of the SS neurons. The arrow marks the rostral-most level of the central sulcus (C).

Figure 6

(A) Distribution of the SS and SU neurons in relation of the depths. (B) Drawing of the section in which the dashed box indicates the location of the photomicrograph in (C). The arrow indicates the depth of track showed in (C). (C) High power photomicrograph of Nissl section from the right hemisphere in which it is possible to identify tracks of the recording electrode through the insular cortex and SII region. Abbreviations: L, lateral fissure; SII, secondary somatosensory cortex; ST, temporal sulcus; C, central sulcus.

Figure 7

Distribution of the task related neurons in relation to the localization of RFs.