Discharge properties of area 5 neurones during arm movements triggered by sensory stimuli in the monkey

Brain Research, 309 (1984) 63-77 Elsevier

63

BRE 10284

Discharge Properties of Area 5 Neurones During Arm Movements Triggered by Sensory Stimuli in the Monkey C. E. CHAPMAN, G. SPIDALIERI* and Y. LAMARRE Centre de Recherche en Sciences Neurologiques, D~partementde Physiologie, Facult~de M~decine, Universit~de Montrdal, C.P. 6128, SuccursaleA, Montreal, Quebec, H3C 3J7 (Canada) (Accepted February 7th, 1984) Key words: area 5 - - arm movement - - single units - - monkey

Unitary discharge was recorded from 157 cells in area 5 of 2 monkeys trained to perform rapid movements of the contralateral arm. Ninety-six cells were task-related. The earliest movement-related modulation in discharge for the large majority of cells (92%) followed the onset of electromyographic (EMG) activity. The discharge pattern of almost all units for which discharge was recorded during movements in opposite directions varied with direction, most often in a nonreciprocal manner. Discharge was correlated with peak velocity in 23% of the excited cells (n = 52). Almost the entire population of cells correlated with velocity were located in the upper part of the anterior bank of the intraparietal sulcus, suggesting that there may be at least two different functional subregions within the arm representation of area 5. Forty percent of the movement-related units had a short latency response to a small, brief perturbation of the elbow which served as one of the movement cues. These sensory responses were labile, not being present in every trial for a large number of cells. Thirty-six percent of the perturbation-sensitive cells were classified as reaction time (RT)-dependent on the basis of a correlation between RT and either the magnitude or the frequency of occurrence of the response. The response was clearly dependent on the subsequent motor response being absent when movement was extinguished. This dependence of the sensory response on the subsequent movement is a property which might represent a neural substrate for somatic sensory attention. The results also support the idea that the RT-dependent cells may be involved in the initiationof the shortest RT movements in response to the somaesthetic cue.

INTRODUCTION A n important role for area 5 in somatic sensation has been established o n the basis of the results of le-

parietal cortex in m a n lead to motor disturbances: misreaching, even with visual guidance, and a decrease in the spontaneous use of the affected side 11. In the monkey, similar disturbances in reaching are

sion studies in m a n and m o n k e y and the sensory properties of single n e u r o n e s in the m o n k e y (see Hyv/irinentl for a recent review). The anatomical connections of area 5, however, suggest that it may also play a role in the control of m o v e m e n t . Whereas the input to area 5 is derived largely from areas 1 and 2 of somatosensory cortex (SI) 14,16,17,36,52its output is

seen during reversible lesions of area 5, produced by local cooling48. Further evidence that area 5 is involved in the control of m o v e m e n t has been provided by unit recordings in behaving monkeys which have shown movement-related discharge of area 5 n e u r o n es, not all of which can be explained in terms of peripheral feedbackl,19,30,34, 45.

directed not only to sensory 17 but also towards motor centres, including the basal ganglia, p o n t i n e nuclei, dorsomedial p r e m o t o r cortex (area 6), supplementary motor area (SMA) and m o t o r cortex (MI)14,15,20,38,49,55,56. Indeed, lesions of the posterior

A n u m b e r of hypotheses have b e e n advanced to explain the possible role of area 5 in the control of movement. Mountcastle et al. 34 proposed that posterior parietal cortex (areas 5 and 7) has a ' c o m m a n d function' for m o v e m e n t initiation. Their putative

* Permanent address: Institute of Human Physiology, Via Fossato di Mortora 64B, 44 100 Ferrara, Italy. Correspondence: Y. Lamarre, Centre de Recherche en Sciences Neurologiques, D6partement de Physiologie, Facult6 de M6decine, Universit6 de Montr6al, C.P. 6128, Succursale A, Montr6al, Qu6bec H3C 3J7, Canada. 0006-8993/84/$03.00 © 1984 Elsevier Science Publishers B.V.

64 'command' neurones discharged with active movement, yet had no identifiable peripheral receptive field. As pointed out by these authors, however, the findings could also be interpreted as representing corollary discharge relayed from the precentral motor areas. Indeed, the results of more recent studies have shown that cells with no identifiable receptive field are not preferentially activated well in advance of movement onset, suggesting that they may not play a role in movement initiation1,19. A more likely role for area 5 is in the sensory guidance of ongoing movement. Thus, the misreaching subsequent to posterior parietal cortex lesions could be attributed to deficits in the integration of sensory and motor signals. Further evidence for such a role is provided by the findings that performance on active, but not passive, tactile discrimination tasks is reduced following posterior parietal cortex lesions in the monkey26.32,48. Finally, there is evidence that the somatic sensitivity of area 5 neurones varies with the level of arousal or attention 34, a property which suggests a role in selective sensory attention. While such a property may be particularly relevant to the proposed role for area 5 in somatic sensation, it might also be important in controlling exploratory movements. In the present study, we investigated the role of area 5 in the control of movement by recording the discharge of single neurones in monkeys trained to perform rapid elbow movements in response to 3, randomly presented, conditioning cues of 3 different modalities (visual, auditory and somaesthetic). Preliminary accounts of some of the results have been published3.4. 24. MATERIALS AND METHODS Two monkeys (1 Macaca mulatta and 1 Macaca fascicularis) were trained to perform either rapid flexions (Daly) or rapid flexions and extensions (Placide) about the elbow in response to 3 different, randomly presented, conditioning cues: a tone, a light and a small (1-2°), brief (50 ms) perturbation of the forearm. The task and the methods for data acquisition and analysis have been fully described elsewhere 21. Briefly, movements had to reach a predetermined amplitude (usually 15-20 °) within 500 ms of the cue presentation in order that the monkey re-

ceive a drop of water. There were no reference points for starting or stopping the movement. For the monkey that made flexion and extension movements (Placide), movement direction was indicated with an elastic band, attached to the manipulandum, which offered a slight resistance to movement in the desired direction. This monkey had been used for another experiment and had a small lesion in the caudal pole of the ipsilateral dentate nucleus. The results obtained in the 2 monkeys were, however, similar. Following training, the animals were anaesthetized with pentobarbital and, under aseptic conditions, a recording chamber (16 mm diameter) was implanted over the left hemisphere (centre of chamber at A2 L10 in Daly and A0 L l l in Placide), contralateral to the trained arm 22. Teflon-coated, multi-stranded, stainless steel wires were chronically implanted into selected arm and shoulder muscles for E M G recordings. Several days after surgery, recordings were begun. Extracellular recordings from single neurones were made with glass-coated tungsten microelectrodes. The action potentials of single units were discriminated and isolation was assured by monitoring the waveform of the spike using an analogue delay device (Bak Electronics, AD5). The depth of each unit with respect to the cortical surface was noted. Unitary discharge was recorded during the performance of 30-45 movements in 1 direction and, for Placide, a further 30-45 trials were then taken in the opposite direction. For units with a sensory response to the perturbation, additional recordings were taken when possible. These included testing the cell's response to perturbations in the opposite direction and to perturbations after the motor response had been extinguished by withholding the reward. Finally, most cells in one monkey (Piacide) were routinely tested for the presence of a somatic receptive field. A PDP-9 computer was used to control the task, for on-line data acquisition and for analysis. For each trial, data were collected for 2 to 2.5 s. These included neural spike intervals, pulse replicas of the EMG activity 7 and angular displacement of the elbow. Rasters of unitary activity, aligned with movement onset, were displayed and continuously updated during the recording session. Selected data were also recorded on magnetic tape. The value and timing of a number of parameters were calculated for each trial, including maximal velocity, acceleration,

65 deceleration, m o v e m e n t amplitude, and initial position of the arm. For the purposes of analysis, data were displayed in rasters and peri-event time histograms which could be aligned on any of the behavioural events which were measured. Towards the end of the recording period, electrolytic lesions were made in selected electrode tracks for later histological identification. After the final recording session, the m o n k e y was deeply anaesthetized and perfused with buffered formol-saline. The dura mater was removed from the depth of the recording chamber. Known stereotaxic points on the cortical surface were marked with India ink and the brain was photographed. Electrode tracks were reconstructed from 10 p m parasaggital sections stained with cresyl violet. Since the cytoarchitectonic border between areas 2 and 5 is difficult to determine 18,36,4° a number of anatomical and physiological criteria were used to classify the cells. Neurones were considered to be in area 5 on the basis of cortical depth (3 mm or more, i.e. within the anterior bank of the intra-parietal sulcus) and/or peripheral receptive field characteristics (passive driving from the ipsilateral side of the body) 6,34,44. For penetrations into the intraparietal sulcus, area 7 was considered to start at the depth at which sensory responses to the visual cue were encountered 47. We may therefore have excluded some area 5 cells from the analysis since there is evidence that at least a small percentage of these respond to visual stimuli 34,43. RESULTS A total of 157 units were recorded in area 5. The mean spontaneous discharge rate of all units during the 500 ms control period preceding the presentation of the stimulus was 11 imp/s (S.D. of the mean + 9.0 imp/s). Sixty-one percent (96) of the units were clearly related to movement and 40% of these (38/96) had a short latency sensory response to the perturbation of the elbow which served as one of the movement cues. No cells in area 5 showed a response linked in time to the auditory cue. As described in the Materials and Methods, cells with visual responses were considered to be in area 7. The results reported here are based upon a detailed analysis performed on 90 of the modulated cells.

Times of onset of first changes in unitary discharge Fig. 1A shows the frequency distribution of the earliest times of change of discharge in relation to movement. Data from the 2 monkeys were pooled. The mean latency for all cells was 25 + 56.6 ms before the onset of movement. Seventy-six percent of all modulated cells showed an excitation as the earliest change with respect to movement; 24% showed an initial decrease in discharge. Two-thirds of the neurones were modulated before the onset of movement, but in only 8% did this precede the earliest detected change in E M G (60-80 ms before movement onset) for muscles acting about the elbow; shoulder muscles were active after the onset of movement. For neurones which responded to the small torque pulse (shaded region in Fig. 1A), movement-related discharge began earlier (x = ----43 + 52.7 ms) than for neurones with no response (x = - - 1 3 + 49.9 ms). The difference was significant at the 5% level (Student's t-test). The mean latency for the response to the perturbation (Fig. 1B) was 40 + 15.7 ms.

Receptivefield properties In one monkey (Placide), the receptive field properties of 62 cells were studied. Seventy-nine percent (49/62) were influenced by peripheral sensory stimuli. Bilateral receptive fields were found for 35% (17) of the responsive cells; in 65% (32) the receptive A

B

20-

~

15-

~

10 °

N = 90 :~= -25 ms 20-

r

~

N = 33 ~= 40 ms

15. 117-

5-

0"

-160

-80

0

80

MOVEMENT ONSET

160 ms

0

•

i

40

•

i

80

•

i

120 ms

SOMESTHETIC CUE

Fig. 1. A: frequency distribution of the times of the first change in area 5 unit activity measured with respect to the onset of contralateral elbow movement in 2 monkeys• Only one value is plotted for each cell; for cells which had different values for each direction of movement, the earliest value was used to construct the histogram. The dotted region represents the latency values for cells which had a response to the somaesthetic cue (n = 38). B: frequency distribution of the response latencies to the somaesthetic cue measured with respect to the stimulus• Five of the cells shown in A were excluded since only a response to the offset of the perturbation was recorded.

66 field was restricted to the contralateral limb or trunk. A large majority of responsive cells (41/49) were movement-related. Twenty-one percent (13/62) of the cells tested had no identifiable receptive field• The proportion of movement-related cells within this group was smaller (62%) than for the cells responsive to somatic stimuli (84%). This proportion was not, however, different from the entire sample with respect to the percentage modulated during movement (61%).

sitivity or timing (~ = 24 ms before the onset of movement)• The majority (36/38) of the neurones whose discharge was recorded during movement in both directions were direction sensitive. These cells were classified into 2 groups depending on their pattern of discharge: reciprocal (n = 6) or non-reciprocal (n = 30).

PLA 349 STIMULUS

MOVEMENT ONSET

Movement related responses: patterns of discharge Fig. 2. is an example of a neurone which had no identifiable receptive field but whose discharge was modulated during movement• Unitary activity is displayed in the form of dot-display rasters and histograms aligned on the presentation of the stimulus (left) and the onset of movement (right). The individual trials in the rasters have been ranked in order of increasing RT. The irregular line running through the rasters (left) indicates the time of onset of movement. Both flexion (above) and extension (below) trials are shown. All 3 types of trials, i.e. those triggered by somaesthetic, visual and auditory cues, are pooled since the discharge was, for this example, the same regardless of the modality of the cue. In other words, for identical movements, the unitary response was the same. The discharge decreased about 60 ms before the onset of movement and remained de,pressed throughout the entire movement. As a result of the low level of spontaneous discharge in this cell, it was not immediately evident if the onset of the response was better related to the performance of the movement or to the presentation of the sensory cue. A more detailed analysis of the timing of the response (onset latency measured for each trial and plotted against RT), however, clearly demonstrated that the onset was linked to the motor response (see Fig. 6A and accompanying text for a more detailed description)• This cell was atypical because the pattern of discharge was the same for both flexion and extension. Only 2 of 38 cells which were tested with movements in both directions showed such an absence of direction sensitivity, i.e. the direction, magnitude and timing of the modulation were the same for both directions of movement. In general, neurones with no somatic afferent input were not different from the entire sample with respect to direction-sen-

:'

:

•

.

•

.

• •:

.

•

,

-..-. :..-.

i..,[.,., I

•

.,..,Jim

I

!

I

I

0

0.5

10

15

.-.

"+ ,

.• -.) "•j

I

20 s

flit .

•

;

'

)

.....-

•

...-

• ." ": " i: '

,

.

.', .'

•,

.

.- : -. -

..

.

•"

. ..: :.

:'." ." ~

:

•

•

•

Z.,

.....

i

"

-

:''.

,

.

"

. --

:

':

;'.:

:.-.-. ..

•

•

•

.

:

l

,,III+, I

Fig. 2. Symmetrical, nondirectional responses of a neurone during flexion (F) and extension (X) m o v e m e n t s of the contralateral elbow. The dot-display rasters, and corresponding histograms, for F and X are constructed from equal n u m b e r s of trials triggered by the somaesthetic, visual and auditory signals. Trials are aligned on the stimulus (left) and m o v e m e n t onset (right) and have been rearranged, in this and all other figures, in order of increasing reaction time. The irregular line running through the rasters on the left indicates the onset of elbow displacement• Unless indicated otherwise, the m o v e m e n t trace represents an average of all trials shown in the corresponding raster in this and all other figures. The binwidth of all histograms is 10 ms.

67 Cells with a reciprocal pattern showed an increase in discharge for movements in one direction and a decrease for movements in the opposite direction. Two different types of non-reciprocal patterns were observed with approximately equal frequency. Either the modulation had the same sign for movements in opposite directions (n = 17), the magnitude and/or the timing varying with direction, or the modulation was present with only one direction of m o v e m e n t (n = 13). The majority of the latter (8/13) were responsive to cutaneous, rather then deep, stimuli. Fig. 3 is an example of a reciprocal cell: discharge increased during flexion and decreased during extension. As in all task-related area 5 cells, the pattern of discharge associated with m o v e m e n t was the same, regardless of the modality of the conditioning stimulus. There was however, in this, and in all of the reciprocal cells, a short latency sensory response to the perturbation (flexion-directed) which was consistent with the behaviour of the unit during active movement. Thus, the unit shown in Fig. 3 (PLA343) was excited by both passive and active flexion. All but one of the reciprocal cells were located in the upper part of the anterior bank of the intraparietal sulcus, i.e. in that part of area 5 which is immediately adjacent to area 2. The discharge pattern of these cells, as well as the associated response to the perturbation, suggest that these cells are activated by joint afferent input and could, in fact, be located in area 2 where units are commonly related to joint movement41, 54. But, 3 of the reciprocal units (including PLA343) had bilateral receptive fields and so could definitely be ascribed to area 5. The remaining 2 units showed submodality convergence, a property which could be associated with area 2 as well as area 512,34. Fig. 3 is also an example of a cell with an early modulation of movement-related discharge (120 ms before the onset of flexion). As noted earlier, only 8 out of 96 modulated cells discharged in advance of the earliest detected E M G onset. These cells were not, however, obviously different from the entire sample and all of those tested had a peripheral receptive field. Figs. 4 and 5 show an example of a neurone with a non-reciprocal pattern of discharge which was particularly evident when the trials were aligned with the onset of movement (Fig. 4). There was an excitation,

PLA 343 F

X

30

0

0.5

1.0 1.5 2.0s

LIGHT

L SOUND ~:.~i

...-:---~

Fig. 3. Reciprocal responses of a neurone during flexion and extension in response to somaesthetic (somes.), visual (light) and auditory (sound) cues. The rasters and histograms are aligned on the stimulus. During flexion (left), cell discharge increased before the onset of movement and remained elevated until the end of flexion. During extension, discharge decreased at the onset of movement. There was a short latency response to the small perturbation (flexion-directed) evident in both the flexion trials (top left), initiated from an initially extended position of the elbow, and in the extension trials (rop right), initiated from an initially, flexed position. beginning at or slightly after the onset of movement, during both flexion and extension trials. The modulation was much more pronounced for flexion move-

68

PLA 3 0 0 F

X •- . 7:/" .} ,:i..

SOMES

ent in every trial. During both the flexion and the extension trials, a response was more evident in the trials with shorter RTs.

Movement-related responses." timing with respect to the motor response For 45 cells in the 2 animals, the onset of the move-

]30

PLA 300 X 0

05

1.0

1.5

!::X:/:: ::

SOM ES. :"::"I-'~i".';!-"!i':. ' ' ' ? ' / . [ ~ -"(.{-/? : :

2.0s

o

g

LIGHT ]30o

0 LIGHT SOUND ~ ' - ° ' . . . [

" ""

" "~

" ....

0.5

1,0

--

1.5 2.0s

•., -'::.::

W-"

Fig. 4. Nonreciprocal, directional response of a neurone during flexion and extension movements. Cell discharge was increased for both directions of movement but the increase was much greater for flexion (left) than extension (right). The rasters and histograms are aligned on the onset of movement. There was a response to the perturbation (top) which is not clearly seen when the trials are not aligned on the stimulus (see Fig. 5).

ments. A short latency (20-60 ms) response to the somaesthetic cue was present and is most clearly seen when the trials are aligned with the stimulus (Fig. 5). The somaesthetic response did not vary with the initial position of the arm: the magnitude of the response was approximately the same when the elbow was initially extended (flexion trials) and flexed (extension trials). The response was not, however, pres-

SOUND

I

::. . ..'..- .....:

ja, Fig. 5. Same cell as in Fig. 4, this display rasters and histograms aligned on the stimulus) showing more clearly the sensory nature of the response to the flexion-directed perturbation (top). The sensory response was very labile, being present in trials with short, but not long, RTs.

69 0.66 (D). There was no difference in the results obtained from the 2 animals. In one monkey (Daly), recordings had also been taken from neurones in area 4 and the results have been reported elsewhere 21. A comparison of the results obtained in areas 4 and 5 showed that the slopes of the R S - R T plots were similar (0.92 + 0.12 in area 4 and 0.89 + 0.30 in area 5) while the values for r were significantly different (P < 0.01, M a n n - W h i t n e y Utest), being lower in area 5 (0.66 + 0.19) than in area 4 (0.86 + 0.07). Discharge in area 5 was, thus, not as tightly coupled to the motor response as in area 4.

ment-related discharge was measured from individual trials triggered by the teleceptive cues (somaesthetic trials were excluded as it was not always possible to dissociate the sensory response, if present, from the subsequent movement-related discharge). Examples of the types of relations obtained are shown in Fig. 6A and B for the cells illustrated in, respectively, Figs. 2 and 4. Measures of the onset latency of the individual responses (RS, measured with respect to the stimulus) are plotted against R T and the corresponding linear regression lines are shown. For both cells, RS was strongly correlated with R T (P < 0.001). The results of this analysis are summarized in Fig. 6C and D. The mean slope for the linear regression lines (C) was close to 1.0, i.e. cell discharge latency increased with increasing RTs, while the mean value for the correlation coeficients (r) was

A

Movement-related responses: correlations with parameters of movement For cells in which discharge increased during movement (n = 52), the strength of the correlation

B

PLA 349 400

300

g

PLA 300 550 SLOPE = 0.91 r = 0.76

SLOPE = 0.80 r = 0.65 N = 28

450

o 200

350

100

250

e/ /

N = 30

.-./

eJ

re-

7. O Q Q

o

o

0 •° I

!

I

150

I

I

250

I

350

I

I

150

I

I

I

150

450

I

250

I

I

350

I

I

1

450

I

550

RT (ms)

RT (ms)

C

D 10

X = 0.88 +_ 0.25 N=45

o

X = 0.66 + 0.19 N =4

[ 7

u.I

=" :"

o--

l-l,', 0

I-I-q,

r]

0.2

0.4

0.6

,

0.8

1.0

1.2

, 1.4

o-

rl

0 , , , , , 0 0.2 0.4

, , , , 0.6 0.8

[7 I

1.0

SLOPE CORRELATIONCOEFFICIENT(r) Fig. 6. Timing of movement-related discharge. A and B: plots of RS as a function of RT for the units illustrated in, respectively, Figs. 2 and 4. The time of onset of unit activity (RS) was measured from individual trials triggered by the visual and auditory cues. Both flexion (filled circles) and extension (open circles) trials were combined for PLA349 (A) while only flexion trials are shown for PLA300 (B). The best fit linear regression lines are also shown. Both relations were significant (P < 0.01). C and D: distribution of slopes (C) and correlation coefficients (D) for the RS-RT plots.

70

PLA 286

PLA 343 200co

r = 067 N = 35

• "

"

75-

r~

r = 075

175-

N=54

>Q) Z ktJ

oIdJ n" It-

•

150-

50-

© Z

E 125"

7

•

•

25-

LO O o°

o•

100o ~

,

100

,

,

,

,

,

,

0 - - 7 / / ,

•

,

,

,

150

200 250 300 350 400 450 100 1~0 200 250 300 3~0 MAXIMUM VELOCITY(° s ~) MAXIMUM VELOCITY(° s ') Fig. 7. Plots of cell firing frequency and peak velocity for the neurone shown in Fig. 3 (PLA343) and for another cell (£LA286) which had a non-reciprocal discharge pattern during movement. Both flexion and extension trials are included for PLA286. Only flexion trials are shown for PLA343. For both plots, the linear regression line is illustrated. Correlation coefficients were significant (P < 0.001). between the firing frequency of the cell and maximal velocity and R T was tested (level of significance, P < 0.01). E x a m p l e s of the types of relations obtained are shown in Fig. 7 for the cell illustrated in Fig. 3 (left) and for one other cell which showed a nonreciprocal pattern of discharge (right). Discharge during flexion (PLA343) and both flexion and extension (PLA286) was significantly correlated with maximal velocity ( P < 0.001). The results of this analysis are summarized in Table I. While discharge was correlated with p e a k velocity in 23% of the cells (12/52), no relations were found with RT. Most of the units whose discharge was correlated with velocity showed a sensory response to the p e r t u r b a t i o n (10/12). It was also found that the spontaneous discharge rate of the cell, m e a s u r e d during the initial 500 ms of each trial, was correlated with the initial position of the arm in 26% of cells tested with m o v e m e n t s in both directions (n = 38). Of these, a slight m a j o r i t y (6/10) had a response to the perturbation. The results of this analysis suggest that, for most of the cells which were correlated with m o v e m e n t p a r a m e t e r s , the relationships might, therefore, be explained by peripheral feedback. Table I also shows that cells r e c o r d e d in the u p p e r part of the anterior bank of the intraparietal sulcus, i.e. anterior area 5, m a d e up almost the entire pop-

ulation of cells correlated with p e a k velocity (11/12). This observation suggests that there may be functional differences within the area 5 representation of the arm. F u r t h e r support for this suggestion comes from the observation that all but one of the reciprocal type units were located in the m o r e anterior part of area 5. RT- and movement-dependence o f the somaesthetic response The sensory nature of the response to the perturbation was evident from rasters and histograms aligned on the presentation of the cue (compare Figs. 4 and 5), i.e, the onset of the sensory response was time-locked to the presentation of the stimulus and not to the movement. This was verified from plots of TABLE I Correlation of neuronal discharge with maximal velocity

Anterior area 5 (n = 20) Posterior area 5 (n = 32)

Response to perturbation

No response to perturbation

10/16 (50%)*

1/4 (5%)

0/16

1/16 (3%)

* Percentage of the total number of neurones within the region tested for correlations. Only neurones showing a movementrelated increase in discharge are included in this analysis. Level of significance, P < 0.01.

71 the sensory response onset (RS) against RT. In 17 cells for which RS could be unequivocally measured in individual trials, no significant correlation was found between RS and RT, as would be expected for a stimulus-related response. The sensory response latency was constant regardless of changes in RT. Conversely, there was a strong correlation (P < 0.01), for all 17 cells, between RT and the time between the sensory response onset and the motor response onset. The mean slope was close to 1 (0.98 _+ 0.22, n = 17). Although the response was entirely sensory in nature, it was frequently quite labile. Fig. 8A shows a neurone with a response to the extension-directed perturbation (latency 60 ms) when the elbow was initially flexed (extension trials, right). The response was weaker, and frequently absent, with the elbow initially extended (flexion trials, left). The response varied with the RT: with short RTs, the response was present in every trial; with long RTs, the response was weaker or absent, the latter especially during the flexion trials. Examination of the accompanying individual movement traces shows that there was little or no variation in the starting position of the arm, i.e. the response modulation was independent of possible changes in position for any one direction of movement. Similar results were obtained in other cells (see, for example, Fig. 5) while a second group of neurones did not show any obvious modulation (see Fig. 3). In a few cases, the lability of the response to the perturbation varied with the recording conditions. For example, while the perturbation response of PLA300 was RT-dependent in the series of trials shown in Fig. 5, the RT-dependence was absent with further testing (only the somaesthetic cue used for the series of trials shown in Fig. 10). In order to describe the RT-dependence of the sensory response in a more quantitative manner, the correlation between the amplitude of the sensory response and RT was tested, the discharge being measured during a fixed time window which encompassed the entire burst of activity as seen on the corresponding cumulative histogram. Examples of the plots and linear regression lines obtained are shown in Fig. 8B and C for the two modulated cells shown in Figs. 8A and 5, respectively. For both cells, a negative correlation between the magnitude of the sensory response and RT was obtained, indicating decreasing

response magnitude with increasing RTs. The correlation was, however, weaker for PLA300 (P < 0.05 as compared to P < 0.01 for PLA288) because there were several trials with no response. Twenty-nine units were analyzed in this manner, excluding cells in which the response consisted of a decrease in discharge (n = 4) or only an 'off' response (n = 5). Of all cells analyzed 15 (52%) showed a negative slope while only 2 (7%) showed a positive slope (slope of 0 for the remaining 12 cells). Only 3 cells showed a strong correlation using this method (P < 0.02) and for all the slope was negative. In no case could the observed relationship be attributed to concomittant changes in position. The small number of significant correlations obtained with this method of analysis may be explained, A

PLA 288 F

X

SOMES.

0

B

05

I0

15

20 s

PLA 288

-'~

C

~

-e°°1

PLA 300

N = 19

•

• ,=o76

~

)

~10o

* 40

so

20

~

-. o

,

o

150

250 RT (ms)

350

,

,.

150

"1

lee

250

,:

,ca

i

350

RT(msl

Fig. 8. RT-dependence of the sensory response to the somaesthetic cue. A: a sensory response to the perturbation (extension-directed) was evident during both the flexion and extension trials, being weaker and sometimes absent with longer RTs. That the modulation of the response could not be explained by variations in the initial position can be seen by inspection of the individual movement traces. B and C: sensory response magnitude is plotted as a function of RT for the cells shown, respectively, in A and in Fig. 5. Flexion (filled circles) and extension (open circles) trials are pooled together. For both cells, the best fit linear regression line is illustrated.

72 in part, by the fact that the sensory response often behaved in an all-or-none manner. Also this method could not be applied to neurones in which the response consisted of a decrease in discharge. Consequently a second method of analysis was applied to the results. Rasters were constructed with the trials rearranged in order of increasing RT and divided into equal numbers of short and long RT trials. The difference between the frequency of response occurrence for the two sub-groups was determined. For the purposes of this analysis, consideration was only given to those trials in which the somaesthetic stimulus had been randomly intermixed with the two teleceptive stimuli. For PLA300, for example, only the series of trials shown in Fig. 5 was used. Pooling the values obtained from the flexion and extension trials in this cell, a response was present during 85% of the shorter as compared to 38% of the longer RT trials, a difference of 47%. This analysis was carried out on 33 neurones, including 4 cells which showed a decreased discharge in response to the perturbation. The results are shown in Fig. 9A. A large number of cells responded with the same frequency in short and long RT trials (difference of 0)'. There was, however, a group of cells (including PLA300), indicated by the dotted region, in which the observed difference was greater than would have been expected by chance (P < 0.01). The distribution by chance (Fig. 9B) was determined by randomly assigning all trials into one of two arbitrary groups and then calculating the percentage difference in the manner described above. All of the randomized measures were within the corresponding 99% confidence limits, as were the RTdependent measures of 24 cells. The 9 cells shown in the dotted region of Fig. 9A were, however, outside of these limits and hence were considered to have shown a significant RT-dependence. For all but one of the latter cells, the difference was a positive value, i.e. the response frequency decreased with longer RTs. In no case could the observed difference be attributed to any systematic change in the initial position of the arm. In summary, 36% (12/33) of the perturbation-sensitive cells tested were classified as RTdependent, including the 3 cells with strong correlations between response magnitude and RT. The somaesthetic response was also found to be dependent on the subsequent performance of the movement. Fig. 10 shows a series of trials, cued by

A 15-

10

J uJ

o

O-

i

R

I

l-'~'7"q I I

I

I

1

I

I

I

I

I 410 i 6~

20

0

1

B

or"

20Z

15

10

0-

t

4T0 I

-

DIFFERENCE

I

I

-20

BETWEEN

~ 0

i

1 80

P A I R S OF O B S E R V A T I O N S ( % )

Fig. 9. Distribution of the differences in the frequency of occurrence of a sensory response for short, as compared to long, RT trials (A). A second distribution (B), corresponding to that which would have been observed by chance, was determined using the same method and data. Trials were assigned to 2 groups, with equal numbers of trials, either by virtue of RT (A) or randomly (B). The frequency of occurrence of the sensory response within each sub-group was determined and the difference between the pairs of observations was calculated. A positive value indicates that the sensory response occurred more frequently in short RT trials (A). Cells in which the RT-dependent values fell outside of the 99% confidence intervals (~ + 2.58 S.D.) which were determined from the data in B were considered to be significantly different (dotted region in A).

the perturbation, which were taken before (left) and after (right) movement was extinguished by withholding the reward. In this series of trials only the somaesthetic cue was used and the sensory response was present in every trial, i.e. the RT-dependence seen in the trials shown in Fig. 5 was absent although both the perturbation and the initial position were identical. Similar changes were seen in a few other cells but were not a consistent finding. When movement was extinguished, the sensory response to the perturbation was virtually abolished (given with the elbow in the same initial position). A further 9 cells were tested in this manner and in all cases the sensory response was reduced or absent when the motor response was extinguished. A comparison of the RT-dependent cells with

73

PLA 300 NO MOVEMENT SOMES.

~

I1,,,I,, 0

0.5

- .....

..

"'~" .._

. I 1.0

I 1.5

! 20 s

I

I

I

I

I

Fig• 10. The somaesthetic response was dependent on the subsequent performance of the movement• A weak but consistent response to the perturbation was present in each flexion trial (left). When movement was extinguished by withholding the reward (right), there was no response to the same perturbation, given at the same initial arm position. those which were not RT-dependent did not reveal any large differences between the two groups• The proportion of cells in which the discharge was correlated with maximal velocity was the same (36% and 39%, respectively). Similarly, equal numbers of units were driven by deep or cutaneous stimulation within the two groups. Finally, RT-dependent and -independent neurones were found in both the anterior and more posterior parts of area 5. i

DISCUSSION In agreement with previous studies, the discharge pattern of area 5 neurones varied with the direction of movement19,30. Similarly, the spontaneous discharge rate of one-quarter of the cells was correlated with the initial position at the elbow• Finally, the onset latency values reported here are comparable to those reported by Bioulac and Lamarret and Kalaska et a1.19, the majority of cells being modulated near the onset of movement•

Origin of movement-related discharge Movement-related activity in area 5 may represent peripheral feedback from the moving limb, relayed through S116, and/or central feedback (corollary discharge or efference copy) from the precentral motor Pareas14,16,20. Alternatively the movement-related discharge might represent a motor command generated autonomously by area 5 neurones 34. Evidence that at least part of the movement-re-

lated discharge is central in origin has been obtained using single unit recording techniques in deafferented monkeys1,45. It is not clear, however, if the discharge which remains after deafferentation represents a corollary discharge from precentral cortex, as might be suggested by a comparison of the timing values for area 4 and 5 in the same monkeys or if it represents a motor command generated in area 545. While the present results cannot resolve this question, they do suggest a fairly close tie between area 5 and the periphery• The earliest movement-related modulation in discharge for almost all cells followed the earliest onset of E M G activity in the operant limb. Furthermore, a large majority of the modulated units (84%) had a peripheral receptive field which included the contralateral elbow and/or upper arm. Finally, almost all cells in which discharge was correlated with maximal velocity of the moving limb responded to passive displacement of the elbow, suggesting that peripheral feedback was driving these cells• Single unit recordings in normal monkeys have shown that area 5 activity generally follows that of area 41,19, suggesting that the discharge could represent a corollary discharge from motor cortex• Whereas the discharge of area 4 cells is time-locked to the subsequent motor response, as evidenced by the strong correlation between RS and R T 21, the discharge of area 5 neurones was found to be less tightly linked to movement in this study• Furthermore, a comparison of the patterns of discharge for movements in opposite directions, both in this task25 and that of Kalaska et al. 19, shows different relations to direction for populations of areas 5 and 4 neurones. For example, 33% of area 4 neurones show a nonreciprocal discharge pattern in this task as compared to 79% of area 5 neurones. Conversely, reciprocal discharge patterns are seen more frequently in area 4 (67%) than area 5 (16%)• Thus, area 5 discharge does not; in any simple manner, reflect that found in area 4, ~uggesting that if motor cortex is indeed a major source of input for area 5 cells then there must be a considerable degree of divergence and/or convergence. It should be pointed out, moreover, that dorsomedial area 6 and/or SMA are also possible sources for this central input, although area 6, at least, appears to be a poor candidate as the proportion of direction-sensitive neurones is rather low 53. A more

74 likely possibility, as suggested by MacKay et al.30 is that area 5 cells receive convergent input from peripheral and central sources during movement. Role of area 5 in the control of movement Certain categories of cells in area 5 have discharge characteristics which would be consistent with a role in the generation of movement. In the present study, 16% of the modulated cells had no obvious receptive field. This proportion is consistent with previous studies 1A9,34,45 and these cells might correspond to the so-called 'command' neurones described by Mountcastle and coworkers 34. In agreement with previous studies using conditioned limb movementsk19, these neurones were not ovbiously different from the entire sample of area 5 cells, their discharge being modulated near the onset of movement. While the latency values do not support a role for these cells in the initiation of movement, they might play a role in the control of ongoing movement. Seal at al. 45, on the other hand, using a task similar to that used here, found that neurones with no peripheral receptive field discharged well before (170 ms) the onset of movement. The latter cells might represent a very different population of cells from those described here since their recordings appear to have been further lateral ( > 5 mm lateral to the postcentral sulcus) than in this study (2-4 mm lateral). This suggestion is supported by their finding that the discharge of a number of these early cells was timelocked to the auditory cue 46. No responses to the auditory cue were found in this study. Another explanation for the different results could be that some of their cells were located in the posterior bank of the intraparietal sulcus. Neurones deep within the intraparietal sulcus, presumably in area 7, have been shown to discharge in response to both movement and auditory signals 23,24. Another category of cells which could conceivably play a role in movement initiation are those which discharge in advance of the earliest detected peripheral motor events. Such cells made up only 8% of the area 5 cells in the present study. A similar proportion has been reported by Bioulac and Lamarre I. Kalaska et a1.19, on the other hand, using a two-dimensional pointing task, reported that 35% of area 5 neurones were modulated prior to the earliest E M G activity. The difference might be related to the different tasks

used. In both this study and that of Kalaska et al., the 'early' cells were not obviously different from the entire sample with respect to either directionality or the presence of a peripheral receptive field. These cells could be involved in movement initiation although at least in this task, few cells appear to play such a role. The present results suggest that there may be at least two functional subregions within the arm representation of area 5. The discharge of cells located deep within the intraparietal sulcus ('posterior' area 5) was infrequently and weakly correlated with peak velocity whilst the discharge of cells located more anteriorly was frequently and strongly correlated with peak velocity. This finding is somewhat paradoxical in view of the fact that the two regions were indistinguishable with respect to receptive field properties, responses to the somaesthetic cue and timing. The suggestion that such functional differences exist within area 5 is, however, supported by the results of Mountcastle et al. 34 who found that the response properties of area 5 neurones were more complex in the posterior, as compared to the anterior, part 2s. An anatomical basis for these different subregions is suggested by the pattern of neural connectivity within area 5: 'anterior' area 5 (area PE according to the nomenclature used by Pandya and Seltzer37) receives input from somatosensory cortex (area 2) and projects posteriorly into the anterior bank of the intraparietal sulcus ('posterior' a r e a 5) 37. That there is indeed a close link between 'anterior' area 5 and the periphery is supported by the observation that the corticospinal projection from area 5, which terminates in the dorsal horn5 and so could modulate transmission of somatosensory inputs, originates only from this region, and not from the sulcus 35. Relation of the conditioned sensory response to R T and movement In 40% of the movement-related cells there was a short latency response to perturbation of the elbow. A similar proportion of responsive neurones has also been found in area 421 in the same experimental conditions. A major finding of this study was the lability of the somaesthetic responses in area 5. In one-third of the cells, responsiveness decreased with increasing RTs. That area 5 neurones are less securely linked to the periphery than neurones in, for example, SI has also been reported by Mountcastle et al. 34. The most

75 likely explanation for the present results is that the response lability reflects a dependence on attention to the stimulus as has been found for responses to visual stimuli in area 7 2,29,33,42. While the animal's attention to the stimulus was not specifically controlled in this task, the animal evidently perceived each stimulus since a RT movement was performed in each 'movement' trial. We therefore tentatively conclude that large changes in the monkey's arousal, which can reportedly influence the somatic sensitivity of area 5 neurones34, were not an important factor in explaining the present results. A more likely explanation is that the RT-dependence can be attributed to small variations in the level of attention to the stimulus: when working 'optimally', as evidenced by shorter RTs, the monkey was more attentive to the somaesthetic cue and so a response was present. Similarly, when the animal's attention to the somaesthetic cue was presumably increased, as in the case of a series of trials cued only by the perturbation, then the response was more consistent (see Figs. 5 and 10). When the monkey was presumably less attentive to the stimulus (longer RTs, somaesthetic cue delivered randomly with light and sound) the sensory response was frequently absent and/or reduced in magnitude. In the extreme case, the absence of the sensory response after movement was extinguished can be explained by the animal's failure to pay any attention to the now behaviourally irrelevant stimulus. Another explanation for the latter observation could be that the modulation reflects not only attention but also intention to move. These somatic responses are not, however, always predictive of movement as is characteristic of areas effectively downstream to area 5, including the frontal eyefields, premotor cortex and SMA9.50,53. The mechanisms involved in mediating the lability of the response are unlikely to be purely peripheral. The perturbation was evidently suprathreshold for activation of the appropriate mechanoreceptors as a RT movement was performed in each movement trial. Furthermore, while the sensitivity of some types of mechanoreceptors likely to be activated by the perturbation is under central control (notably muscle spindles and joint receptorslO.31.51), labile responses were found for neurones apparently activated solely by cutaneous input. These observations suggest that a central gating mechanism is involved in mediating

the modulation of the response. The level at which such modulation occurs is not known, although it does not seem to be present at the level of S113,24,39 which is the major source of peripheral afferent input to area 5. Sensory responses in area 4, which also projects to area 5, are, however, reported to vary with the behavioural state, both attention 39 and the intended direction of movement 8 influencing the somatic sensitivity of cells in MI. But, using this same paradigm, somaesthetic responses in MI are not obviously related to RT and persist after extinction of movement e4. These results suggest that the modulation may occur primarily at the level of area 5, perhaps taking origin in the limbic system 27, although a possible contribution from area 4 cannot be dismissed. The dependence of the sensory response on the subsequent performance of the conditioned movement, itself cued by the perturbation, suggests that these RT-dependent area 5 neurones may play a role in the initiation of the shortest RT movements triggered by somaesthetic cues. The results are also consistent with the idea that such neurones may represent a neural substrate for somatic sensory attention. Indeed, such a role would be consistent with the results of lesion studies in the monkey which have shown that area 5 is critically important for active, but not passive, tactile discrimination of form 26,32. ACKNOWLEDGEMENTS The authors wish to thank Drs. T. Drew, J. F. Kalaska and J. P. Lund for helpful criticism of the manuscript. The authors would also like to thank M.-T. Parent for training the animals and assisting in the experiments, R. Bouchoux for constructing the manipulandum, H. Duguay for writing the computer programmes, G. Filosi and D. Cyr for art and photographic work, S, Bergeron for electronic support and H. Dussault and H. Auzat for typing the manuscript. The research was supported by a Group grant from the Medical Research Council of Canada. C.E.C. is a Medical Research Council Postdoctoral Fellow. Y.L. is a member of the Medical Research Council Group in Neurological Sciences. G.S. was a CRNN A T O Postdoctoral Fellow (1978-79) and a H. H. Jasper Postdoctoral Fellow (1979).

76 REFERENCES 1 Bioulac, B. and Lamarre, Y., Activity of postcentral cortical neurons of the monkey during conditioned movements of a deafferented limb, Brain Research, 172 (1979) 427-437. 2 Bushnell, M. C., Goldberg, M. E. and Robinson, D. L., Behavioral enhancement of visual responses in monkey cerebral cortex. I. Modulation in posterior parietal cortex related to selective visual attention, J. Neurophysiol., 46 (1981) 755-772. 3 Chapman, C. E., Busby, L., Spidalieri, G. and Lamarre, Y., Modulation of somatic afferent input to area 5 prior to conditioned arm movements in the monkey, Proc. int. Union Physiol. Sci., 15 (1983) 293. 4 Chapman, C. E., Busby, L., Spidalieri, G. and Lamarre, Y., Sensorimotor properties of neurons in area 5 during conditioned arm movements, Abstr. Neurosci. Soc., 9 (1983) 490. 5 Coulter, J. D. and Jones, E. G., Differential distribution of corticospinal projections from individual cytoarchitectonic fields in the monkey, Brain Research, 129 (1977) 335-340. 6 Duffy, F. H. and Burchfiel, J. L., Somatosensory system: organizational hierarchy from single units in monkey area 5, Science, 172 (1971) 273-275. 7 Evarts, E. V., Precentral and postcentral cortical activity in association with visually triggered movement, J. Neurophysiol., 37 (1974) 373-381. 8 Evarts, E. V. and Tanji, J., Reflex and intended responses in motor cortex pyramidal tract neurons of monkey, J. Neurophysiol., 39 (1976) 1069-1080. 9 Goldberg, M. E. and Bushnell, M. C., Behavioral enhancement of visual responses in monkey cerebral cortex. II. Modulation in frontal eye fields specifically related to saccades, J. Neurophysiol., 46 (1981)773-787. 10 Grigg, P., Mechanical factors influencing response of joint afferent neurons from cat knee, J. Neurophysiol., 38 (1975) 1473-1484. 11 Hyv~irinen, J., The Parietal Cortex of Monkey and Man, Springer-Verlag, New York, 1982, pp. 202. 12 Hyv~irinen, J. and Poranen, A., Receptive field integration and submodality convergence in the hand area of the postcentral gyrus of the alert monkey, J. Physiol. (Lond.), 283 (1978) 539-556. 13 Hyv~irinen, J., Poranen, A. and Jokinen, Y., Influence of attentive behavior on neuronal responses to vibration in primary somatosensory cortex of the monkey, J. Neurophysiol., 43 (1980) 870-882. 14 Jones, E. G., Coulter, J. D. and Hendry, S. H. C., Intracortical connectivity of architectonic fields in the somatic sensory, motor and parietal cortex of monkeys, J. comp. Neurol., 181 (1978) 291-348. 15 Jones, E. G., Coulter, J. D., Burton, H. and Porter, R., Cells of origin and terminal distribution of corticostriatal fibers arising in the sensorimotor cortex of monkeys, J. comp. Neurol., 173 (1977) 53-80. 16 Jones, E. G. and Powell, T. P. S., Connexions of the somatic sensory cortex of the rhesus monkey. I. Ipsilateral cortical connexions, Brain, 92 (1969) 477-502. 17 Jones, E. G. and Powell, T. P. S., An anatomical study of converging sensory pathways within the cerebral cortex of the monkey, Brain, 93 (1970) 793-820. 18 Jones, E. G., Wise, S. P. and Coulter, J. D., Differential thalamic relationships of sensory-motor and parietal corti-

cal fields in monkeys, J. comp. Neurol., 183 (1979) 833-882. 19 Kalaska, J., Caminiti, R. and Georgopoulos, A. P., Cortical mechanisms related to the direction of two-dimensional arm movements: relations in parietal area 5 and comparison with motor cortex, Exp. Brain Res., 51 (1983) 247-260. 20 Kfinzle, H., Cortico-cortical efferents of primary motor and somatosensory regions of the cerebral cortex in Macaca fascicularis, Neurosci., 3 (1978)25-39. 21 Lamarre, Y., Busby, L. and Spidalieri, G., Fast ballistic arm movements triggered by visual, auditory and somesthetic stimuli in the monkey. I. Activity of precentral cortical neurons, J. Neurophysiol., 50 (1983) 1343-1358. 22 Lamarre, Y., Joffroy, A. J., Filion, M. and Bouchoux, R., A stereotaxic method for repeated sessions of central unit recording in the paralysed or moving animal, Rev. Canad. Biol., 29 (1970) 371-376. 23 Lamarre, Y., Spidalieri, G., Busby, L. and Lund, J. P., Programming of initiation and execution of ballistic arm movements in the monkey. In H. H. Kornhuber and L. Deecke (Eds.), Motivation, Motor and Sensory Processes of the Brain, Progress in Brain Res., Vol. 54, Elsevier/ North-Holland Biomedical Press, Amsterdam, 1980, pp. 157-169. 24 Lamarre, Y., Spidalieri, G. and Chapman, C. E., A comparison of neuronal discharge recorded in the sensori-motor cortex, parietal cortex and dentate nucleus of the monkey during arm movements triggered by light, sound or somesthetic stimuli. In J. Massion, J. Paillard, W. Schultz and M. Wiesendanger (Eds.), Neural Coding of Motor Performance, Exp. Brain Res., Suppl. 7, Springer-Verlag, Berlin, 1983, pp. 140-156. 25 Lamarre, Y., Spidalieri, G. and Lund, J. P., Patterns of muscular and motor cortical activity during a simple arm movement in the monkey, Canad. J. physiol. Pharmacol., 59 (1981) 748-756. 26 LaMotte, R. H. and Mountcastle, V. B., Neural processing of temporally-ordered somesthetic input: remaining capacity in monkeys following lesions of the parietal lobe. In G. Gordon (Ed.), Active Touch: the Mechanism of Recognition of Objects by Manipulation, Pergamon Press, Oxford, 1978, pp. 73-77. 27 Leichnetz, G. R. and Astruc, J., The efferent projections of the medial prefrontal cortex in the squirrel monkey (Saimiri sciureus), Brain Research, 109 (1976) 455-472. 28 Lynch, J. C., The functional organization of posterior parietal association cortex, Behav. Brain Sci., 3 (1980) 485-534. 29 Lynch, J. C., Mountcastle, V. B., Talbot, W. H. and Yin, T. C. T., Parietal lobe mechanisms for directed visual attention, J. NeurophysioL, 40 (1977) 363-389. 30 MacKay, W. A., Kwan, M. C., Murphy, J. T. and Wong, Y. C., Responses to active and passive wrist rotation in area 5 of awake monkeys, Neurosci. Lett., 10 (1978) 235-239. 31 Millar, J., Joint afferent fibres responding to muscle stretch, vibration and contraction, Brain Research, 63 (1973) 380-383. 32 Moffett, A., Ettlinger, G., Morton, H. B. and Piercy, M. F., Tactile discrimination performance in the monkey: the effect of ablation of various subdivisions of posterior parietal cortex, Cortex, 3 (1967) 59-96. 33 Mountcastle, V. B., Andersen, R. A. and Motter, B. C., The influence of attentive fixation upon the excitability of

77 the light-sensitive neurons of the posterior parietal cortex, J. Neurosci., 1 (1981) 1218-1235. 34 Mountcastle, V. B., Lynch, J. C., Georgopoulos, A., Sakata, H. and Acuna, C., Posterior parietal association cortex of the monkey: command functions for operations within extrapersonal space, J. Neurophysiol., 38 (1975) 871-908. 35 Murray, E. A. and Coulter, J. D., Organisation of corticospinal neurons in the monkey, J. comp. Neurol., 195 (1981) 339-365. 36 Pandya, D. N. and Kuypers, H. G. J. M., Cortico-cortical connections in rhesus monkey, Brain Research, 13 (1969) 13-36. 37 Pandya, D. N. and Seltzer, B., Intrinsic connections and architectonics of posterior parietal cortex in the rhesus monkey, 1. comp. Neurol., 204 (1982) 196-210. 38 Petras, J. M., Connections of the parietal lobe, J. psychiatr. Res., 8 (1971) 189-201. 39 Poranen, A. and Hyv~rinen, J., Effects of attention on multiunit responses to vibration in the somatosensory regions of the monkey's brain, EEG Clin. Neurophysiol., 53 (1982) 525-537. 40 Powell, T. P. S. and Mountcastle, V. B., The cytoarchitecture of the postcentral gyrus of the monkey Macaca mulatta, Bull. Johns Hopk. Hosp., 105 (1959) 108-131. 41 Powell, T. P. S. and Mountcastle, V. B., Some aspects of the functional organization of the cortex of the postcentral gyrus of the monkey: a correlation of findings obtained in a single unit analysis with cytoarchitecture, Bull. Johns Hopk. Hosp., 105 (1959) 133-162. 42 Robinson, D. L., Goldberg, M. E. and Stanton, G. B., Parietal association cortex in the primate: sensory mechanisms and behavioral modulations, J. Neurophysiol., 41 (1979) 910-932. 43 Sakata, H., Somatic sensory responses of neurons in the parietal association area (area 5) of monkeys. In H. H. Kornhuber (Ed.), Somatosensory System, Georg Thieme, Stuttgart, 1975, pp. 250-261. 44 Sakata, H., Takaoka, Y., Kawarasaki, A. and Shibatani, H., Somatosensory properties of neurons in the superior parietal cortex (area 5) of the rhesus monkey, Brain Research, 64 (1973) 85-102. 45 Seal, J., Gross, C. and Bioulac, B,, Activity of neurons in area 5 during a simple arm movement in monkeys before

and after deafferentation of the trained limb, Brain Research, 250 (1982) 229-243. 46 Seal, J., Gross, C. and Bioulac, B., Different neuronal populations within area 5 of the monkey. In J. Massion, J. Paillard, W. Schultz, M. Wiesendanger (Eds.), Neural Coding of Motor Performance, Exp. Brain Res. Suppl. 7, SpringerVerlag, Berlin, 1983, pp. 157-163. 47 Seltzer, B. and Pandya, D. N., Converging visual and somatic sensory cortical input to the intraparietal sulcus of the rhesus monkey, Brain Research, 192 (1980) 339-351. 48 Stein, J., The effect of parietal lobe cooling on manipulative behaviour in the conscious monkey. In G. Gordon (Ed.), Active Touch, Pergamon Press, Oxford, 1978, pp. 79-9O. 49 Strick, P. L. and Kim, C. C., Input to primate motor cortex from posterior parietal cortex (area 5). 1. Demonstration by retrograde transport, Brain Research, 157 (1978) 325-330. 50 Tanji, J. and Kurata, K., Comparison of movement-related activity in two cortical motor areas of primates, J. Neurophysiol., 48 (1982) 633-653. 51 Vallbo, A. B., Hagbarth, K.-E., Torebj6rk, H. E. and Wallin, B. G., Somatosensory, proprioceptive, and sympathetic activity in human peripheral nerves, Physiol. Rev., 59 (1979) 919-957: 52 Vogt, B. A. and Pandya, D. N., Cortico-cortical connections of somatic sensory cortex (areas 3, 1 and 2) in the rhesus monkey, J. comp. Neurol., 177 (1978) 179-192. 53 Weinrich, M. and Wise, S. P., The premotor cortex of the monkey, J. Neurosci., 2 (1982) 1329-1345. 54 Whitsel, B. L., Dreyer, D. A. and Roppolo, J. R., Determinants of body representation in postcentral gyrus of monkeys, J. Neurophysiol., 34 (1971) 1018-1034. 55 Wiesendanger, R., Wiesendanger, M. and RuEgg, D. G., An anatomical investigation of the cortico pontine projection in the primate (Macaca fascicularis and Saimiri sciureus). II. The projection from frontal and parietal association areas, Neuroscience, 4 (1979) 747-765. 56 Zarzecki, P., Strick, P. L. and Asanuma, H., Input to primate motor cortex from posterior parietal cortex (area 5). II. Identification by antidromic stimulation, Brain Research, 157 (1978) 331-335.