spatial selectivity of go/no-go neurons in monkey prefrontal cortex

Exp Brain Res (1994) 100:165-169 �9 Springer-Verlag 1994

M. Sakagamm �9 H. Niki

Spatial selectivity of go/no-go neurons in monkey prefrontal cortex

Received: 27 September 1993 / Accepted: 30 March 1994

Abstraet We examined single-unit activity in the inferi- or prefrontal cortex during a visual go/no-go discrimination task under maintained visual fixation. The monkeys had to base their response on either the color, shape, or position of a discriminative cue, and the relevant task condition was indicated by the color of the fixation spot. We analyzed the spatial selectivity of 128 go/no-go neurons showing a marked differential cue-period activity that depended on whether the stimulus signaled a go or no-go response. Most of these neurons (n= 106, 83%) showed asymmetry between their responses to stimuli in the contralateral and ipsilateral visual fields. Seventy-seven of these neurons had a contralateral preferential field, and 29 had an ipsilateral preferential field. These results show that in many inferi- or prefrontal neurons a degree of differentiation in their responses to go and no-go stimuli depends on the cue positions, and that the coding of behavioral meaning is carried out mainly in the contralateral hemisphere.

Key words Prefrontal cortex - Single-unit activity Go/no-go discrimination task �9 Spatial selectivity Monkey

Introduction

Previous studies have found that many prefrontal neurons showed differential cue-period activity that depended on whether the stimulus signaled one type of upcoming response or the other; for example, a go or no-go response, or a go-left or go-right response (Ko- matsu 1982; Quintana et al. 1988; Watanabe 1986). But because the retinotopic location of the stimulus was not precisely controlled in those studies, it was not clear whether prefrontal neurons code the behavioral mean-

M. Sakagami �9 H. Niki ( [~) Department of Psychology, Faculty of Letters, University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113, Japan, Fax no: 81-3-3816-4743, e-mail: niki@ tansei.cc, u-tokyo.ac.jp

ing of the stimulus equally well in both the contralateral and the ipsilateral visual fields.

We have recently analyzed the effects of changing the relevant task conditions (color, shape, and position) on the prefrontal neuronal responses to the go and no-go stimuli (Sakagami and Niki 1994). The aim of the present study was to examine the spatial selectivity of go/no-go neurons (which showed differential activity for the go and no-go stimuli) and the effects of interaction between the cue positions and the degree of differentiation in neuronal response.

Materials and methods

The experimental subjects were the same two adult male Japanese monkeys (Macacafuscata) used in our previous study (Sakagami and Niki 1994), and because the methods we used were the same as those detailed in our previous paper, they are described here only briefly. The animals were trained to perform a symmetrically rewarded go/no-go discrimination task in which they had to base their response on either the (1) color, (2) shape, or (3) position of a cue that was presented during fixation of a colored light spot (0.2 ~ in diameter). The relevant task condition was indicated by the color of the fixation spot: purple for the color condition, blue for the shape condition, and green for the position condition. Each trial was initiated by the monkey pressing a lever that turned on a fixation spot on the TV monitor. If the monkey continued depressing the lever for 1-2 s, a 160-ms cue stimulus (1.8 ~ x 1.8 ~ was presented at the left, right, upper, or lower position (each 1 ~ from the central fixation spot by center-to-center measurement). After the post-cue delay period (1-2 s) the color of the fixation spot changed. At the time of this color change, the monkey had to make an immediate release within 0.8 s (go) or to withhold the releasing response at least for 1.2 s (no-go), depending on the cur- rently relevant task condition and the cue stimulus presented pre- viously. The monkey was given fruit juice as reward for correct responses on both go and no-go trials, which were interspersed in random sequence. A rerun correction procedure was used, and we typically used two sets of eight stimuli. The first set combined two colors (red and green), two shapes (circle and cross), and two positions (left and right). The second set combined two other colors, two other shapes, and two other positions (yellow-purple, stripes-diamond, upper-lower). These two stimulus sets were usually presented in separate blocks. If color was relevant, red or yellow signaled go and green or purple signaled no-go. If shape was relevant, a circle or stripes signaled go and a cross or diamond

166 B ONSET

GO H ~ V

N G vH~

FIX. SPOT COLOR CHANGE

_ _ 1 , o" 0 , 5 S E C

C LEFT RIGHT UPPER GO NG GO NG GO NG

- +1 fJi, ,l I ..... " t I t . . . . . ++' I ] i i i i i i i i

I I I I l l l l ' I I I IH I I I I I i i i i I I p I i I . . . . . . . . , , , ,,:,, ,,,,, ,'; I I I I I i l m | I I I I i I / l l l l l I I 0 H l l l l l l I I t I I I l l p l l I I I I t T M I I I l l l l l I I I III IIII

I I I IHHIIIII I II IIIPI I I I IIHII I I II I I t II I I J i l l IN I

LOWER GO NG

-

.... I l l I I I I I I l I l l I l l l I

II I I I I 1 |

', r+~'," " " " '''~ I I I I i l l I . . . . ,~ ,,

I I I I I I I I I I I l l l N i I I I I I I I

I' '+ I I I+'I i ' -"'"' . . . . . " ....... ' '"+"ll i ..... """ + ' ..... " -+''" ......... " n , n ' " ' " " ' ",,'% " " " " , ' , " , I ' ~,,,,,~ , , , , ,, , .......

l i l l i l i I I I I I H I I IHIIH I ~1~11 I I I l l l l i I H I l i I F l a i l I Q _ , , ,H ',,II < . . . . . . . . . . . . . . . . . . . . . . . . . , ,~

II IIII ~ l n I I II

i [,.,, ..... 'i H I I l l H i l l I

I t II m l l i Z I l l I i l l I I

I l l I I H I I I I l l l l l

0 I I l l l I I I I I I I i l H I

i i i I I11 i ~ l l l I

I I l i l l l I I I I ,, I,-,,

. . . . . i;,",' I I llll

K R I I 0 1

+ F + ; 7 1 III I l l l~ I

I I I I D i l l I I|11 I I I I | l l I

I I I I I I I I I I I n I I | I I l l I I I I I I In ill

i I I I i n l l i I I i l i b I ir I I I I I I HIIIII I I

I i I I I [ I III I I I l l I I IIII III I I I l l IIIIII I

I I I I I I~ I l l l l li t I I I I I I I IF nllHI

i I i l i M I I II I I I U I M I Ill I ii irianHt t

I I l l l l I II II I i I I l l IIIIIII

Fig. 1 A Brain diagram showing the approximate recording area in the present study. (AS arcuate sulcus, CS central sulcus, PS principal sulcus, SF Sylvian fissure, ST | superior temporal sulcus.) B Record excerpts of horizontal (H) and vertical (l/) eye movements for eight go (GO) and eight no-go (NG) trials. In horizontal eye movement records, up is right and down is left. Vertical dotted lines show the 1-s pericue period (0.5 s before and 0.5 s after the cue onset) during which monkeys had to fixate rigidly. Arrows on the left indicate the time of onset of fixation spot (FIX. SPOT ONSET), and arrows on the right indicate the time of fixation spot color change (FIX. SPOTCOLOR CHANGE). Note the mainte- nance of fixation during the pericue period. C An example of differential cue-period activity of a prefrontal neuron studied under three task conditions (color, shape, and position). LEFT, RIGHT, UPPER, and LO WER indicate that the cue stimulus was

i I I I I IIII I

I INI I H | [ I I I I I I

I I I I I J I I1| l u l l

I I I I H i l l II I r~ I H i l l

IIlll DI fl I l l n I in l l l

i I iiJ i i i i I I l l

H I 1 fl i i i i i ii i i i i i

I I e l tH

80S/$ L 0.5 SEC

presented at the left, right, upper, or lower position relative to the central fixation spot. Distances from the fixation spot to each of the cue positions were 1 ~ by center-to-center measurement. The center line in each raster indicates the onset of the cue stimulus. Activity is displayed for 0.5 s before and after the cue onset. Cue duration, indicated by horizontal bars, was 160 ms. Histogram binwidth is 20 ms. Under the color condition (COLOR), go trials consisted of five red-circle trials and five red-cross trials, whereas no-go trials consisted of five green-circle trials and five green-cross trials. Under the shape condition (SHAPE), go trials consisted of five red-circle trials and five green-circle trials, whereas no-go trials consisted of five red-cross trials and five green-cross trials. Under the position condition (POSITION), both go and no-go trials consisted of five red-circle, five red-cross, five green-circle, and five green-cross trials

signaled no-go. If position was relevant, the cue presented in the left or upper position signaled go and the cue presented in the right or lower position signaled no-go.

After the task training was completed, a head-fixation device, unit recording chamber, and scleral eye coil were implanted using standard aseptic surgical techniques. Before unit recording sessions the monkey was trained to gaze at the fixation spot during the pericue period: in this training, and in subsequent recording sessions, trials were aborted if the eyes deviated from the fixation spot by more than 1 ~ during the 0.5-s precue and 0.5-s postcue period. We searched for visually responsive neurons, mainly in the inferior prefrontal cortex (Fig. 1A), while the monkey performed the task under maintained fixation (Fig. 1 B). When a visually responsive neuron was found, we usually used the first stimulus set and examined the neuron's activity under three conditions.

Neurons showing excitatory responses under both color and shape conditions were selected for analysis of spatial selectivity. Using a statistical criterion, we determined (1) whether the response was differential for the go and no-go trials, and (2) whether the response was differential for cue positions. Two-way analysis of variance (ANOVA) was performed on the neuronal data (dis- charge rate from 100 to 300 ms after the cue onset), with trial type (go vs no-go) and cue position as main factors. Post hoc tests (Tukey's honestly significant difference method) were also used for each comparison. Statistical significance was attributed when the probability of a type I error was less than 0.05.

Midway through the experiment, we analyzed 12 neurons by presenting the same stimuli in one of the four positions under both color and shape conditions. For the other 15 neurons we analyzed spatial selectivity by presenting the same set of cue stimuli randomly in one of the 20 positions under the color condition.

For the position condition, we could not always determine to what extent the observed differential activity was due to the difference in trial type or to the difference in cue position. Consequent- ly, our analysis in the present study was restricted to the color and shape conditions.

At the end of the experiment each monkey was killed and its brain was removed for histological analysis. The recording, data analysis, and histological procedures are detailed in our previous paper (Sakagami and Niki 1994).

Results

Figure 1 A shows the recording area in this study. Most of the neurons we examined were within the caudal inferior convexity, the area lateral to the principal sulcus (ventral portion of area 46 and area 12 of Walker) (Walker 1940). Some of the most caudal neurons were within area 45 of Walker (inferior arcuate area).

The database was the 249 go/no-go differential neurons described in our previous paper (Sakagami and Niki 1994). In the present paper our analysis of spatial selectivity centered on the 128 excitatory go/no-go neurons (100 go and 28 no-go neurons) that responded dif- ferentially to the go and no-go stimuli of the first set presented on the left or right under both color and shape conditions. Of these 128 neurons, 106 (83%) showed spatial selectivity between the left and right positions (73 for both color and shape conditions, 12 for the color condition only, and 21 for the shape condition only). Under the color condition, out of 85 spatially selective neurons, 62 (73 %) showed significant interaction (trial type x position). For the shape condition, 52 out of 94 (55%) spatially selective neurons exhibited significant interaction. In all neurons showing the significant

167

interaction effect (except for one neuron for the color condition), the degree of differentiation in neuronal response for go and no-go stimuli on the more responsive side (preferential field) was greater than that for the stimuli on the less responsive side. Seventy-seven (60%) neurons had their more responsive side in the contralateral visual field (56 for color and shape conditions, 8 for color condition only, and 13 for shape condition only), and 29 (23%) in the ipsilateral visual field (17 for color and shape conditions, 4 for the color condition only, and 8 for the shape condition only). Thus, contralateral bias was evident.

When we compared the side preference of our go/no- go neurons between color and shape conditions, 73 neurons had consistent side preference under both conditions (contralateral n=56, ipsilateral n= 17). The re- maining 33 neurons showed side preference under one condition and no such preference under the other condition.

For 12 neurons the same go and no-go stimuli were presented at four positions (left, right, upper, and lower). Two-way ANOVA with position (left, right, upper, lower) and trial type (go vs no-go) as main factors showed a significant main effect of trial type for all 12 neurons, a significant main effect of position for 10 of them, and significant interaction for 11 of them for the color condition. The results for the shape condition were as follows: effect of trial type, 12 of 12; position, 11 of 12; interaction, 5 of 12. Figure 1 C, illustrating the activity of a neuron whose spatial selectivity was examined for four positions using the same stimuli, shows that this neuron (KRll01, recorded from the right hemisphere) had a preference for the contralateral side. The magnitude of differential responses for go and no-go trials was larger for the contralateral field (left) than for the ipsilateral field (right). This neuron did not show much difference in reactivity to the go stimuli presented in the upper and lower positions under both color and shape conditions. In the position condition the stimuli on the left and upper positions (go stimuli in this condition) evoked intense activity.

For 15 neurons whose spatial selectivity under the color condition was analyzed by presenting the cue stimuli randomly in one of the 20 positions, two-way ANOVA for each of the neurons revealed a significant main effect of trial type for all the neurons, a significant main effect of position for all the neurons, and significant interaction for 13 of them. Figure 2 shows exam- ples of spatially selective activity by four neurons examined in 20 positions. Figure 2A, B is an example of activity of a go (KL1103) and a no-go neuron (KL5901) with narrow tuning. Figure 2 C is an example of spatially selective activity of the neuron (KL6601) with relatively broad tuning. Figure 2D shows activity of another broadly tuned neuron (KL2803), which was responsive to the stimuli presented at almost all positions examined. Two-way ANOVAs performed on the data shown in Fig. 2A, B, D revealed significant effects of position (P < 0.0001), trial type (P < 0.0001), and significant inter-

168

A uL2 u35 uR2

D1

Uo., oR, �89 oR2

C uLl uR,

Fig. 2A-D Spatially selective activity of four prefrontal neurons for 20 different cue positions. A Narrowly tuned go neuron (KLl103, recorded from the left hemisphere); B narrowly tuned no-go neuron (KL5901, recorded from the left hemisphere); C relatively broadly tuned go neuron (KL6601, recorded from the left hemisphere); D broadly tuned go neuron (KL2803, recorded from the left hemisphere). Note that spatial selectivity was also observed in the no-go neuron shown in B. In the paired his- tograms for each of the 20 positions, the upper histogram repre- sents the activity for the go stimuli and the lower histogram repre-

B I UL2

[~I ULI

L3 L2 L1 R1 R2 R3

D1

DL1 D 2 ~ D R I ~

~DL2

D ~ U L 2 U3

~ UL1 U2

U1 L3 L2 L1

D1

~ D L 1 D2

~ D L 2

UR2

o3E 0RI

DR2

sents the activity for the no-go stimuli. Each histogram displays the mean number of spikes per second over four trials. The center line indicates the cue onset. Activity is displayed for 0.5 s before and after the cue onset. Cue duration was 160 ms. Binwidth is 20 ms. The ordinate scales (maximum) are 150 spikes/s in A, 63 spikes/s in B, 225 spikes/s in C, and 250 spikes/s in D. Distances from the fixation spot are 1 ~ for U1, D1, L1, R1 ; 4 ~ for U2, D2; 5 ~ for L2, R2; 7 ~ for U3, D3; 8 ~ for UL1, UR1, DL1, DR1; and 10 ~ for L3, R3; 12 ~ for UL2, UR2, DL2, DR2. (U up, D down, L left, R right)

ac t ion be tween pos i t ion and trial type (P < 0.0001). For the n e u r o n whose activi ty is shown in Fig. 2 C, the sig- nif icant effects are o f pos i t ion ( P < 0 . 0 0 0 1 ) and of trial type (P < 0.0001), bu t no t of in te rac t ion (P > 0.1).

Discussion

Our da t a d e m o n s t r a t e tha t the great ma jo r i ty of go /no - go neu rons in the inferior pref ronta l cor tex exhibit spatial selectivity, and tha t for m o r e than half of the neurons the degree of differentiat ion depends on the cue posi t ion. The ma jo r i ty of g o / n o - g o neu rons were m o r e

responsive to stimuli on the cont ra la te ra l side of the visual field. This f inding is consis tent with the results of an ab la t ion s tudy in which o c u l o m o t o r work ing m e m o - ries were found to be lateral ized pr imar i ly to the con- t ra la teral prefronta l cor tex (Funahash i et al. 1993). It should be no ted tha t we found no neurons whose side preference (contra la tera l versus ipsilateral) was reversed between co lor and shape condi t ions .

Because the pos i t ion of the s t imulus in the present s tudy is one of the task condi t ions on which animals h a d to base their response, this migh t have had some influ- ence on the observed spatial selectivity. It seems to be unlikely, however , tha t the observed field a s y m m e t r y

was due to the association between specific position and specific response, since there was contralateral bias in both hemispheres of the 2 monkeys.

Previous studies have indicated that the receptive fields of prefrontal neurons are usually large and pri- marily represented the visual field contralateral to the recording hemisphere (Boch and Goldberg 1989; Funa- hashi et al. 1990; Mikami et al. 1982; Suzuki and Azu- ma 1983). An enhancement of sensory response has also been observed when the stimulus became a saccade target (Boch and Goldberg 1989; Funahashi et al. 1990; Goldberg and Bushnell 1981). The "preferential" field of our go/no-go neurons, however, differed from either the receptive field or saccade-target field mentioned above in the sense that the activity of our neurons was related to the associated behavioral meaning of the stimulus.

The prefrontal cortex has been thought to receive information about the color and shape of a visual stimulus mainly from the inferior temporal cortex (areas TE and TEO; Ungerleider et al. 1989). The preferential fields of our go/no-go neurons in the inferior prefrontal cortex seem to be more similar to the receptive fields of TEO neurons than to those of TE neurons, since the latter typically included portions of ipsilateral as well as the contralateral visual field, whereas the former did not (Boussaoud et al. 1991; Gross et al. 1972). On the other hand, considering well-documented anatomical connections between the dorsolateral prefrontal cortex and the inferior parietal cortex (Cavada and Goldman-Rakic 1989) and the intrinsic connections between the subar- eas within the prefrontal cortex (Barbas and Pandya 1989), it is possible that visuospatial information from the inferior parietal cortex contributes to the formation of the preferential field of inferior prefrontal neurons.

Acknowledgements The authors thank T. Miyakawa and S. Yoshida for animal care, K. Tsutsui for his assistance in data analysis, and H. Takenaka for computer programming. Thanks are also due to Dr. P.M.B. Cahusac of the University of Stifling for his help in checking the English. This work was supported by Grants-in-Aid for Scientific Research on Priority Areas (02255103, 03251103, 04246102) to H. Niki during 1990-1992 and Grants-in-Aid for Encouragement of Young Scientists (03710025, 04710042) to M. Sakagami during 1991-1992 from the Ministry of Education, Science and Culture of Japan.

169

References Barbas H, Pandya DN (1989) Architecture and intrinsic connec-

tions of the prefrontal cortex in the rhesus monkey. J Comp Neurol 286:353-375

Boch RA, Goldberg ME (1989) Participation of prefrontal neurons in the preparation of visually guided eye movements in the rhesus monkey. J Neurophysiol 61:1064 1084

Boussaoud D, Desimone R, Ungerleider LG (1991) Visual topog- raphy of area TEO in the macaque. J Comp Neurol 306:554 575

Cavada C, Goldman-Rakic PS (1989) Posterior parietal cortex in rhesus monkey. II. Evidence for segregated corticocortical net- works linking sensory and limbic areas with the frontal lobe. J Comp Neurol 287:422-445

Funahashi S, Bruce CJ, Goldman-Rakic PS (1990) Visuospatial coding in primate prefrontal neurons revealed by oculomotor paradigms. J Neurophysiol 63 : 814-831

Funahashi S, Bruce CJ, Goldman-Rakic PS (1993) Dorsolateral prefrontal lesions and oculomotor delayed-response perfor- mance: evidence for mnemonic "scotomas". J Neurosci 13:1479-1497

Goldberg ME, Bushnell MC (1981) Behavioral enhancement of visual responses in monkey cerebral cortex. II. Modulation in frontal eye fields specifically related to saccades. J Neurophys- iol 46:773-787

Gross CG, Rocha-Miranda CE, Bender DB (1972) Visual properties of neurons in inferotemporal cortex of the macaque. J Neurophysiol 35:96-111

Komatsu H (1982) Prefrontal unit activity during a color discrimination task with GO and NO-GO responses in the monkey. Brain Res 244: 269-277

Mikami A, Ito S, Kubota K (1982) Visual response properties of dorsolateral prefrontal neurons during visual fixation task. J Neurophysiol 47: 593-605

Quintana J, Yajeya J, Fuster JM (1988) Prefrontal representation of stimulus attributes during delay tasks. I. Unit activity in cross-temporal integration of sensory and sensory-motor information. Brain Res 474:211-221

Sakagami M, Niki H (1994) Encoding of behavioral significance of visual stimuli by primate prefrontal neurons: relation to relevant task conditions. Exp Brain Res 97:423M36

Suzuki H, Azuma M (1983) Topographic studies on visual neurons in the dorsolateral prefrontal cortex of the monkey. Exp Brain Res 53:47-58

Ungerleider LG, Gaffan D, Pelak VS (1989) Projections from in- fero-temporal cortex to prefrontal cortex via the uncinate fas- cicle in rhesus monkeys. Exp Brain Res 76:473-484

Walker AE (1940) A cytoarchitectural study of the prefrontal area of the macaque monkey. J Comp Neurol 73 : 59-86

Watanabe M (1986) Prefrontal unit activity during delayed condi- tional go/no-go discrimination in the monkey. I. Relation to the stimulus. Brain Res 382:1-14

spatial selectivity of go/no-go neurons in monkey prefrontal cortex

Documents