Brain Research Bulletin,Vol. 27, pp. 751-751. 0 Pergamon Press pk. 1991. Printed in the U.S.A.
0361-9230/91
$3.00 + .OO
BRIEF COMMUNICATION
Sensory Response Enhancement and Suppression of Monkey Primary Somatosens~ry Cortical Neurons R. J. NELSON, ~epurtment
B. LI AND V. D. DOUGLAS
of Anatomy and Neurohiol~gy, College of medicine, ~niver~i~ 875 Monroe Avenue, Memphis, TN 38163
Received
of Tennessee
at Memphis
16 July 1990
NELSON, R. J., B. LI AND V. D. DOUGLAS. Sensory response enhancement and suppression of monkey primary somatosensory co&al neurons. BRAIN RES BULL 2’7(5)7.51-757, 1991.-Vibratory stimulus-related responses were recorded from monkey primary somatosensory cortical (SI) neurons while animals performed two tasks. In the movement task, vibratory stimuli served as the go-cue for wridt flexion or extension. In the no-movement task, movements normally made in response to vibratory stimuli were extinguished. Area 3a, 3b, and 1 neurons with deep receptive fields (RFs) exhibited greater stimulus-related activity during the movement task than during the no-movement task. Area 3b neurons with cutaneous RFs were similarly enhanced during the movement task, whereas area 1 neurons with cutaneous RFs were less responsive to vibratory stimuli during the move~nt task. These results suggest that motor-set and/or selective attention may modulate the responsiveness of SI neurons to peripheral stimuli and that changes in sensory responsiveness in SI neurons differ as a function of their cortical location and RF type, SI cortex
Rhesus monkeys
Sensory responses
Motor set
Selective attention
SI neurons to vibratory stimuli were ~investigated. In one behavioral task, the vibratory stimuli served as the go-cue for wrist movement. In the other, the same stimuli were cues for wrist position maintenance. The studied neurons were grouped by receptive field type (RF, cutaneous or deep) and by cortical location (areas 3a, 3b and 1) to determine if there were any dflerential effects upon sensory responsiveness that were related to these factors.
AXONS in the primary somatosenso~ cortex (SI) of primates receive information from the sensory periphery dealing with a multitude of stimulus-related parameters. They are capable, under the proper conditions, of encoding the spatial location and temporal characteristics of stimuli and often do so with great fidelity (13). Yet, during goal-oriented behavior, only a small portion of the information that is transmitted through the ascending somatosenso~ system may be consciously perceived at any given time (5). The responses of SI neurons may be different depending upon whether an animal makes a movement in response to a stimulus (3) (“motor set”) or whether a stimulus is relevant or irrelevant for the task at hand (5) (“selective attention”). However, observations from studies designed to demonstrate effects of these two factors have been somewhat con~dicto~. The activity of SI neurons associated with somesthetic stimuli may be unchanged (10) or change dramatically (15) after movements initially made in response to the stimuli are extinguished. SI neurons may be more responsive to a relevant stimulus as compared to when the same stimulus is made irrelevant (5) or SI neurons may be unaffected by changes in attentive behavior (18). These observations suggest that motor set and selective attention may influence the responsiveness of SI neurons to peripheral stimuli only under certain behavioral conditions. Because of these conflicting observations, the responses of
METHOD
Behavioral
Training
Three adult rhesus monkeys (Macacu muluttu) were taught to make wrist flexion and extension movements or maintain their wrist position following delivery of a vibratory stimulus to the palm. The monkeys were cared for in accordance with the NIH Guide for Care and Use of Laboratory Animals, revised 1985. Each monkey sat in an acrylic monkey chair with his right forearm on an armrest and his right palm on a plate attached, at the end nearest the wrist, to the axle of a brushless IX torque motor. The load of 0.07 Nm was applied to the plate, which assisted extension and opposed flexion movements. The monkeys exerted a flexion force when maintaining wrist position. Animals viewed a visual display, located 35 cm in front of them at eye level. This display was coupled to the wrist position signal com-
7.51
752
NELSON,
ing from the torque motor and consisted of 31 light-emitting diodes (LEDs). A central, red LED indicated a centered wrist position. Each smaller. yellow LED above or below the central lamp indicated a deviation from the adjacent lamp of 1” of wrist position. The monkeys performed two tasks. During the “movement task,” they made wrist flexion or extension in response to vibratory stimuli. In the “no-movement task,” they withheld movement following the presentation of vibratory stimuli. The display was turned off at the start of the first trial during the no-movement task to indicate which task the animal should perform. Trials for each task began in the same manner, with the monkey centering the plate. If the monkey maintained this centered position for 0.5, 1.O, or 1.5 s (pseudorandomized), a 27-, 57- or 127-Hz low-amplitude stimulus (sine wave CO.057” angular deflection or < 100 wrn peak-to-peak measured 10 cm from the axle of the torque motor) set the plate into vibration. Movement of more than 0.5” from center during the hold period cancelled the trial. At this point, the two tasks differed. In the movement task. the vibratory stimulus, which served as the go-cue, remained on until the monkey either flexed or extended at least 5” from the held position. If he made the requested movement, the monkey received a fruit juice reward. A movement direction instruction was given at the start of each trial by the presence or lack of illumination of red LED located in the upper left comer of the visual display. If this LED was illuminated, extension was the appropriate movement. Otherwise, the monkey was instructed to flex. This instruction lamp remained on throughout the duration of each extension trial. In the no-movement task, the vibratory stimulus remained on for 1 s. If the monkey held the centered position within 0.5” for this l-s period, he was rewarded. Otherwise, the trial was cancelled. In both tasks, a new trial was begun when the monkey once again held a steady position. Electrophysiological
Recording and Hisrology
Once animals reached a steady daily performance level, stainless steel recording chambers were surgically implanted and extraceliular recordings were made of SI neurons using platinumiridium microelectrodes. Transdural penetrations were made daily into the somatosensory cortex, and the activity of single units was amplified, discriminated and stored in a PDP-11/23+ microcomputer by conventional means (2, 11, 16, 1’7). On the last recording day, electrolytic lesions were made in cortical locations of interest by passing 10 p,A of current for 10 s. Then the animals were deeply anesthetized with sodium pentobarbital and transcardially perfused with 10% buffered formolsaline. Histological sections of the cortex were prepared. Electrode tracks were reconstructed based upon the depth of each recording site from the point where the cortical activity was first encountered and the location of the marking lesions (16,17). Only those neurons from recording sites that could be clearly identified as to their cortical area location were included in this analysis.
Ll AND DOUGLAS
of an increase or decrease in firing rate that was 240% of the background discharge rate for more than 30 consecutive ms (17). Upper and lower t~sholds of at least 40% from the mean activity during the vibratory response were then set. The vibratory response was defined as ending when the activity crossed one of these thresholds for 30 consecutive ms. The duration of the vibratory response was determined by this method. The neuronal response during no-movement trials was then measured during the same period relative to stimulus onset. Measurement of stimulus-related responses during the same interval for recordings from the two task conditions seemed reasonable because a previous report has shown the relative independence between response latency and enhancement index in another sensory system (1). The magnitude of the stimulus-related change in activity during movement and no-movement trials was calculated for each neuron by subtracting the neuron’s background activity from its vibratory response. This change was considered to be a measure of the neuron’s vibratory responsiveness. RESULTS
Selection Crireria Several selection criteria were used to determine which neurons should be included in this study. A total of 55 in area 3a. 65 in area 3b and 136 in area 1 were initially examined because each had short latency (O.O5). These groups had enhancement indices distributed over larger ranges. Each of the four remaining groups had mean enhancement indices that were significantly greater than 1.0 (Table 1). Using a paired f-test, the absolute value of the magnitudes of the stimulus-related activity during movement and no-movement trials were compared, grouping the samples by the same variables listed above. Area 1 samples from neurons with cutaneous RFs showed significantly smaller vibratory activities during the movement task when compared with the no-movement task [mean difference: - 17.49; f(25)=5.49, pO. 1). The enhance-
NELSON, LI AND DOUGLAS
5
1
n
Area 3a Deep RFs
%3 a. s @2 1 0
8 7
a q Area 3b Deep RFs m Area 3b Cut. RFs
6
q Area 1 Deep RFs m Area 1 Cut. RFs
FIG. 2. The distribution of enhancement indices for all recorded samples, grouped by cortical location and receptive field type. ment indices for each of the subgroups for areas 3a and 3b neurons with deep RFs were not significantly different from those of the larger groups (unpaired z-test; p>>O.l). However, due to the small size of the subgroups, neither were these indices different from unity. The general lack of differences between the treatment of all of the data as independent samples and that of choosing a single value for each recorded neuron suggest that combining all samples from similar neurons together probably did not mask any significant trends and suggest that the conclusion made with regard to samples from a given cortical region may be extended to neurons in that region.
755
SI RESPONSE ENHANCEMENT AND SUPPRESSION TABLE 1 THEDISTRIBUTION OF ALL
SAMPLES GROUPED BY CORTICAL LOCATION AND RFs TYPE
h&X
Flex
Location Ext
2(l) 2(l)
4(l) 13(3)
O(0) O(0)
6(l) g(2)
4(21 4(l)
170) 7(3)
Ventral
Location Dorsal
30) 12(3) O(0)
18(4) 2(l) 4(l)
g(4) 6(2) 4(l)
20) 6(l) O(0)
Deep RPS Area 3a Digits: wrist: Area 3b Digits: wrist: Area 1 Digits: wrist: cut. RFS Area 3b Digits: Hand: Wrist: Area 1 Digits: Hand: Wrist:
SD
Mean
SD
Mean
SD
Mean
Total 21(6)
2.01$
1.18
6.30
14.25
30.27
42.7
14t3>
1.42”
0.59
5.19
23.40
18.9*
32.0
32(13)
1.75$
0.94
5.54t
10.11
27.6f
37.7
Total 39(10)
1.72j
1.08
7.76$
14.65
24.7$
35.6
26(9)
0.62$
0.29
- 17.49$
16.23
- 38.4)
29.6
Samples (Neurons)
RIJ Type
Percentage Change
Activity Difference
E~~cement
Numbers in parentheses indicate the number of neurons from which samples were recorded. Means and standard deviations of the three comparisons of the &mums-related activity during movement as compared with no-movement trials. Comparisons were made using all samples from a given data group. Enhancement indices and percentage changes were compared using a one-factor ANOVA; the differences were compared using a paired t-test. The enhancement indices were tested for significance against the hypothesis that they were equal to 1.O; the differences and percentage changes were tested against the hypothesis that they were equal to 0.0. SD = standard deviation. *p