| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Center for the Neural Basis of Cognition, Mellon Institute; and Department of Neuroscience, University of Pittsburgh, Pittsburgh, Pennsylvania
Submitted 19 January 2005; accepted in final form 27 March 2005
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
; Kalivas and Nakamura 1999; Mogenson et al. 1980; Ono et al. 2000). The limbic system is crucial for the initial stage of evaluating rewards (Roesch and Olson 2004; Tremblay and Schultz 1999, 2000). The transition from evaluating a reward to initiating action in pursuit of it is thought to depend on structures interposed functionally between the limbic system and motor system. Areas that may play a role in this transition include the dorsolateral prefrontal cortex (PFC), the frontal eye field (FEF), the supplementary eye field (SEF), the premotor cortex (PM), and the supplementary motor area (SMA). Neuronal activity in all of these areas is influenced both by the nature of the action the monkey is planning and by the value of the reward to which it will lead (Coe et al. 2002; Hikosaka and Watanabe 2000; Hikosaka et al. 1989; Kobayashi et al. 2002; Leon and Shadlen 1999; Matsumoto et al. 2003; Roesch and Olson 2003; Wallis and Miller 2003; Watanabe 1990, 1992, 1996; Watanabe et al. 2002). Thus it makes sense to think of them as a potential watershed between limbic evaluative functions and motor output.
The PFC, FEF, SEF, PM, and SMA are thought to serve a variety of functions with cognitive, oculomotor, and skeletomotor components. It would be surprising, for that reason, if reward-related activity took the same form or had the same significance in all of them. To test for possible differences among these areas with respect to the nature of reward-related activity, we recently carried out a comparative study in which we recorded from neurons in all of them while monkeys performed a memory-guided saccade task in which a cue presented early in each trial indicated whether the reward delivered on successful completion of the trial would be large or small (Roesch and Olson 2003). We found that the tendency for neurons to fire more strongly when a large reward was expected substantially increased as the recording site moved posteriorly (from PFC to FEF to PM in the lateral frontal lobe and from SEF to SMA in the medial frontal lobe).
The pattern of interareal differences observed in the previous study was of interest because it cast light on the possible significance of reward-related activity. This might either 1) represent the value of the anticipated reward in the service of an economic decision process or 2) reflect motivational modulation of the state of motor preparation, motor output, arousal, or attention. To distinguish definitively between these possibilities is not possible in experiments that manipulate only the value of the predicted reward because, under this manipulation, perceived value and motivational state are correlated (Maunsell 2004; Roesch and Olson 2004). Nevertheless, in light of the fact that reward-related activity was strongest by far in PM and SMA, which is to say in motor areas, the second interpretation, based on motivational modulation of the monkey's preparatory state, carries greatest weight.
We were concerned that these results might be specific to manipulations of value based on reward size. To address this issue, we devised an alternate approach in which the size of the reward was the same on every trial but the monkey was informed by an early cue whether the delay before delivery of the reward would be long or short. It is well known that inserting a longer anticipated delay before an anticipated reward reduces its perceived value, a phenomenon known as time-discounting (Lowenstein and Elster 1992). We report here that varying the delay before delivery of a constant reward had very much the same effect on neuronal activity as varying the size of a reward delivered at a constant delay. In particular, the tendency for neurons to fire more strongly after a cue that predicted a short delay was much more robust in areas behind the arcuate sulcus (FEF/PM, PM, and SMAr) than in areas in front of it (PFC, FEF, and SEF).
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Four adult male rhesus monkeys were used (Macaca mulatta; laboratory designations N, P, A, and F). Experimental procedures were approved by the Carnegie Mellon University Animal Care and Use Committee and were in compliance with the guidelines set forth in the United States Public Health Service Guide for the Care and Use of Laboratory Animals.
Preparatory surgery
At the outset of the training period, each monkey underwent sterile surgery under general anesthesia maintained with isofluorane inhalation. The top of the skull was exposed, bone screws were inserted around the perimeter of the exposed area, a continuous cap of rapidly hardening acrylic was laid down so as to cover the skull and embed the heads of the screws, a head-restraint bar was embedded in the cap, and scleral search coils were implanted on the eyes, with the leads directed subcutaneously to plugs on the acrylic cap (Robinson 1963). After initial training, recording chambers were implanted into the acrylic. At each selected site, a 2-cm-diameter disk of acrylic and skull was removed. A cylindrical recording chamber was cemented into the hole with its base just above the exposed dural membrane.
Chambers were placed either at a medial location (over SEF and SMAr) or at a lateral location (over PFC, FEF, FEF/PM, and PM). Recording was carried out from a medial chamber in monkeys A, N, and F; a left lateral chamber in monkey P; and a right lateral chamber in monkeys F and N. The medial chambers placed over SEF and SMAr were centered on the midline of the brain approximately 21 mm anterior to the Horsley–Clarke interaural plane. The lateral chambers placed over PFC, FEF, and PM were centered approximately at anterior 23 mm and lateral 23 mm.
Memory-guided saccade task
The aim of this task was to allow initial characterization of the spatial selectivity of each neuron. The monkeys performed memory-guided saccades to six targets forming a hexagonal array at an eccentricity of 10° (Fig. 1A). Each trial began with the monkey's fixating a central spot. At 500 ms after attainment of fixation, the six targets appeared. After an additional 300 ms a cue was presented for 250 ms in superimposition on one of the targets. After a random delay in the range of 500 to 1,000 ms, the fixation spot was extinguished, whereupon the monkey had to make a saccade directly to the previously cued target. Trials involving the six targets were interleaved in random order subject to the constraint that each block of six successful trials had to contain one trial involving each target. Testing continued until it was possible to identify the target eliciting maximal activity. Subsequent testing in other tasks involved this target and the one diametrically opposite with respect to the fixation point (Fig. 1A: 1 and 1', 2 and 2', or 3 and 3').
|
The monkeys performed a memory-guided saccade task in which a cue presented early in the trial predicted a short (500 ms) or a long (2,500 ms) delay period. Essential features of the task are summarized in Fig. 1. Each trial began with onset of a central fixation spot (Fig. 1B1). At a time 50 ms after attainment of fixation, the spot was transformed to a cue the shape and color of which signified the length of the upcoming delay period (Fig. 1C1). After 400 ms two targets appeared (Fig. 1D1) at diametrically opposed locations. A directional cue identical to the fixation cue except in size was then presented for 250 ms in superimposition on one of the targets (Fig. 1E1). After a 500 ms (or 2,500 ms) delay period (Fig. 1F1), the fixation spot was extinguished (Fig. 1G1), whereupon the monkey was required to make a saccade directly to the previously cued target (Fig. 1H1) and to maintain fixation on it for 300–450 ms after saccade completion until delivery of a juice reward (Fig. 1I1). The intertrial interval after a correct trial was set to 1,000 ms whereas, after an error, either a fixation break or a wrong choice, it was set to 5,000 ms. There were four conditions representing all possible combinations of delay length (short or long) and direction (preferred or antipreferred). The conditions were interleaved in random order subject to the constraint that one trial conforming to each condition had to be completed successfully before initiation of the next block of trials. Because of this constraint, no long-term advantage attached to breaking fixation on an undesirable (long-delay) trial. The rejected condition would simply be presented repeatedly until a trial was successfully completed. To prevent confounding activity related to delay length with selectivity for the visual properties of the cues, the cue convention was reversed after each block of 40 successful trials. The collection of data from a given neuron commonly continued until 80 trials had been completed successfully.
Stimuli in the variable-delay task
The fixation spot was a 0.38° white square presented at the center of the screen. Targets were 0.38° white squares presented 10° from central fixation. The central delay cues, which spanned 0.96°, were a red square and a blue circle. The directional cue shared all of the properties of the foveal delay cue with the exception that it spanned 1.32°. The background of the display had a luminance of 1.5 cd/m2 and CIE x and y chromaticity coefficients of 0.26 and 0.26. White stimuli had a luminance of 126.5 cd/m2 and CIE x and y chromaticity coefficients of 0.28 and 0.32. Red stimuli had a luminance of 112.5 cd/m2 and CIE x and y chromaticity coefficients of 0.27 and 0.61. Blue stimuli had a luminance of 110.2 cd/m2 and CIE x and y chromaticity coefficients of 0.15 and 0.17.
Variable-reward task
Many of the neurons studied in the context of the variable-delay task were also studied in the context of the variable-reward task. In the variable-reward task, the delay was fixed at 1,500 ms, whereas the cue at the beginning of the trial predicted a big (0.3 ml) or small (0.1 ml) juice reward. Essential features of the task are summarized in Fig. 1. For further details of task design, see Roesch and Olson (2003). Data collected in the context of the variable-reward task were considered in a previous paper concerned with that task (Roesch and Olson 2003). Here, we consider data from the variable-reward task solely in connection with the question whether neurons sensitive to delay length were also sensitive to reward size.
Order of tasks
Neuronal activity was first monitored in the context of the memory-guided saccade task with reward size and delay length fixed and with targets at six locations spaced at 60° intervals around fixation (Fig. 1A). Any neuron appearing to exhibit task-related activity in this task was selected for study in the variable-delay task and the variable-reward task. In these tasks, the possible target locations were confined to the neuron's preferred direction (as determined in the memory-guided saccade task) and the opposite direction. The order in which the two tasks were run alternated across sessions. Some neurons were studied in the context of only one of the tasks because recording instability or satiation of the monkey prevented running both.
Single-neuron recording
At the beginning of each day's session, a varnish-coated tungsten microelectrode with an initial impedance of several megohms at 1 kHz (Frederick Haer, Bowdoinham, ME) was advanced vertically through the dura into the immediately underlying cortex. The dura was debrided at intervals commonly spanning a few months to ensure penetrability by the electrode. The electrode could be placed reproducibly at points forming a square grid with 1-mm spacing (Crist et al. 1988). The action potentials of a single neuron were isolated from the multineuronal trace by means of an online spike-sorting system using a template-matching algorithm (Signal Processing Systems, Prospect, Australia). The spike-sorting system, on detection of an action potential, generated a pulse the time of which was stored with 1-ms resolution.
Electromyographic measurements
Adhesive surface electrodes were placed on the shaved skin overlying the right splenius capitus and masseter muscles. The voltage threshold was set as low as possible subject to the constraint that the voltage did not cross threshold at rest. Muscle activity was stored as time-marked records of threshold crossings. From these, we constructed histograms representing the mean instantaneous threshold-crossing rate as a function of time during the trial under each condition.
Experimental control and data collection
All aspects of the behavioral experiment, including presentation of stimuli, monitoring of eye movements, monitoring of neuronal activity, and delivery of reward, were under the control of a Pentium-based computer running Cortex software provided by R. Desimone, Laboratory of Neuropsychology, National Institute of Mental Health. Eye position was monitored by means of a scleral search coil system (Riverbend Instruments, Birmingham, AL). The X and Y coordinates of eye position were stored with 4-ms resolution. Stimuli generated by an active matrix LCD projector were rear-projected on a frontoparallel screen 25 cm from the monkey's eyes.
Analysis of the dependency of behavior on delay length
We used paired t-tests to compare, across sessions, the session means of the following measures obtained on short-delay versus long-delay trials: reaction time, error rate, and fixation-break rate. Reaction time was defined as the delay from offset of the fixation spot to the moment when the eye left the central fixation window. Error rate was defined as the number of trials on which a saccade was directed to the wrong target expressed as a percentage of all trials on which a saccade was directed to either target. Fixation-break rate was defined as the percentage of all trials on which the eye left the central fixation window at any time before offset of the fixation spot.
Analysis of the dependency of firing rate on task factors
We used two-factor ANOVAs to analyze the dependency of the firing rate of each neuron on delay length and response direction. We independently analyzed data from seven trial epochs: 1) from delay cue onset to directional cue onset (700 ms), 2) from onset to offset of the directional cue (250 ms), 3) 250 ms beginning with directional cue offset, 4) 250 ms before fixation spot offset, 5) 200 ms before saccade initiation, 6) from saccade onset to 100 ms after saccade completion, 7) 100 ms before to 100 ms after initiation of reward delivery. In all tests, the criterion for statistical significance was taken as P 
0.05.
Assessing contribution of reaction time to activity related to delay length
To determine whether neuronal activity continued to depend on delay length when the effects of behavioral reaction time were factored out, we performed a multivariate regression analysis, fitting three models
1) Y = a0 + a1RT
2)Y = a0 + a2DELAY
3)Y = a0 + a1RT + a2DELAY
where Y is the firing rate measured from onset of the delay cue to offset of the fixation spot and RT is the behavioral reaction time. The variable DELAY was set to 1 or 0 for trials with short or long delays, respectively. To determine whether adding the variable DELAY produced a significant improvement in performance, we compared model 3 to model 1. To determine whether adding the variable RT produced a significant improvement, we compared model 3 to model 2. Significance was assessed with an F-test using
 |
0.05. Localization of recording sites
To characterize the location of the recording sites relative to gross anatomical landmarks, we projected the sites onto structural magnetic resonance (MR) images (Fig. 2). The images were collected by use of a Bruker BioSpin 4.7 T magnet in which the anesthetized monkey was supported by an MR-compatible stereotaxic device. Fiducial marks made visible by means of a contrast agent included the centers of the ear bars and selected locations inside the recording chamber. Frontoparallel 2-mm-thick slices spanning the entire brain were collected. In addition, 2-mm-thick slices were collected parallel to the cortical surface underlying each lateral chamber. To determine the location of recording sites relative to functional divisions of cortex, we mapped out regions under each chamber from which motoric responses (eye, face, and limb movements) could be elicited at low threshold (40 µA) by electrical microstimulation (1.65-ms biphasic pulses delivered through the recording microelectrode at a frequency of 300 Hz in 200-ms-long trains).
|
, 2000). Supplementary eye field (SEF): a region, located rostral to the genu of the arcuate sulcus and extending 2–5 mm from the hemispheric midline, in which microstimulation elicited saccades. Rostral supplementary motor area (SMAr): a region immediately caudal to the SEF in which microstimulation elicited movements of the face and limbs. |
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
In the following sections, we will describe the impact of delay length on behavior and neuronal activity in PFC, FEF, FEF/PM, PM, SEF, and SMAr. At each stage of analysis, we will consider (first) effects related to anticipated delay and (second) effects related to elapsed delay. We will examine the impact of anticipated delay by comparing short-delay to long-delay trials during epochs aligned on the delay-predicting cue and preceding the moment at which, in short-delay trials, the signal to respond was given. We will examine the impact of elapsed delay by comparing short-delay to long-delay trials during epochs near in time to the response and aligned on response initiation. The analyses concerned with anticipated and elapsed delay involve comparing roughly identical periods of neuronal activity, as observed on short-delay trials, to nonoverlapping early and late periods of activity, as observed on long-delay trials. Thus the two analyses are not entirely independent. However, as will be seen, they reveal qualitatively different effects.
Behavior: anticipated delay
To analyze the impact of anticipated delay on behavior we assessed how fixation breaks were distributed across time during the early part of the trial under short- versus long-delay conditions. The results, shown in Fig. 3C, indicate 1) that fixation breaks were more frequent under long- than under short-delay conditions and 2) that the tendency to break fixation declined over the course of the trial under both conditions. To determine whether the impact of anticipated delay length was significant, we compared the fixation-break frequencies (number of trials prematurely terminated by cessation of fixation expressed as a percentage of all trials) observed under short- and long-delay conditions in each monkey during the first 1,000 ms beginning at the presentation of the delay cue (Fig. 3D). This analysis epoch spans a period in which equivalent events occurred at equivalent times on short- and long-delay trials, with the only difference between them ascribed to anticipation. The tendency for fixation breaks to occur more often in anticipation of a long delay was present and significant in every monkey (two-tailed paired t-test, P < 0.05) and was highly significant in data collapsed across monkeys (P < 0.001). We conclude that the monkeys were less motivated to perform the task when anticipating a long delay than when anticipating a short delay.
|
We will refer to effects occurring at the end of the trial (and revealed by comparing measures from short- and long-delay trials taken around the time of the behavioral response) as related to "elapsed delay." This is an expositional tool; it does not imply that the effects were a direct consequence of elapsed delay. There is no sure way of establishing whether effects present at the time of the response depended on the monkey's experiencing the antecedent delay or, alternatively, were a result of the monkey's having been put by the delay cue into a state that persisted until the end of the trial. With this qualification, we note that behavioral measures taken at the end of the trial were sensitive to the duration of the antecedent delay. This was evident in two behavioral measures computed for every neuronal data collection session. The error rate (percentage of trials when the incorrect target was selected relative to all trials when one target or the other was selected) was lower on short-delay (0.6%) than on long-delay (2.0%) trials (Fig. 3A). This trend was present in data from every monkey, achieving significance (two-tailed paired t-test, P < 0.05) in three out of four of them (Fig. 3D), and was highly significant in data collapsed across all monkeys (P < 0.0001). The average behavioral reaction time was faster on short-delay (233 ms) than on long-delay (248 ms) trials (Fig. 3B). This trend was present and achieved significance (two-tailed paired t-test, P < 0.05) in three out of four monkeys (Fig. 3D), and was highly significant in data collapsed across all monkeys (P < 0.0001). These results indicate that monkeys were in a state of heightened preparation (reflected by a simultaneous improvement of accuracy and speed) after a short delay as compared to a long delay.
Neuronal data analysis: anticipated delay
To determine whether neuronal activity was influenced by the length of the anticipated delay, we compared neuronal activity occurring before the imperative command (offset of the fixation spot) on short-delay trials to neuronal activity occurring during the identical period (at the end of which the fixation spot remained on) on long-delay trials. Neurons in many frontal areas exhibited activity related to the length of the expected delay. This activity commonly took the form of a main effect (with the net firing rate higher or lower on short-delay trials) and less frequently took the form of an interaction effect (with the strength of the directional signal stronger or weaker on short-delay trials). For the neurons illustrated in Fig. 4, A and B, the net firing rate during the "anticipated-delay" comparison period (time-locked to delay-cue onset and highlighted in yellow) was higher on short-delay trials (top row for each neuron) than on long-delay trials (bottom row for each neuron).
|
Neuronal data analysis: elapsed delay
On examination of histograms representing the activity of single neurons, it was evident that the firing rate at the end of the delay period, around the time of the saccade (epoch highlighted in green in Fig. 4), could differ according to whether the antecedent delay had been long or short. The neuron of Fig. 4A clearly fired more strongly at the end of a short than at the end of a long delay. In contrast, on trials requiring a response in the preferred (leftward) direction, the neuron of Fig. 4B fired more strongly at the end of a long delay. To analyze the nature and rate of incidence of effects dependent on elapsed delay, we will proceed through three steps of analysis. 1) Population histograms. These depict activity during a 1,500 ms epoch beginning 500 ms before saccade initiation (righthand column in Figs. 5– 9). The beginning of this epoch coincides with a point in time about 250 ms after offset of the directional cue on short-delay trials and 2,250 ms after offset of the directional cue on long-delay trials. 2) Individual neurons by epoch. We will consider whether firing rate depended on delay length or its interaction with response direction during late epochs of the trial, including epoch IV (250 ms before fixation spot offset), epoch V (200 ms before saccade initiation), epoch VI (from saccade onset to 100 ms after saccade completion), and epoch VII (from 100 ms before to 100 ms after initiation of reward delivery). 3) Individual neurons across a long premovement epoch. To obtain one robust statistical measure for each neuron, to facilitate comparison across areas, we will determine whether the firing rate of each neuron depended significantly on delay length or its interaction with response direction during a long epoch beginning 250 ms before the imperative cue (offset of the fixation spot) and ending with saccade initiation.
|
|
POPULATION. We collected data from 204 neurons in the PFC of two monkeys (Table 1). As a basis for qualitative assessment of the effect of anticipated delay on the activity of these neurons, we constructed population curves representing firing rate as a function of time under the four trial conditions (Fig. 5A). Population histograms shown to the left (Fig. 5A1) represent activity subject to influence by the duration of the anticipated delay. In these displays, aligned on cue onset, thick and thin lines represent population activity on trials requiring responses in the preferred and antipreferred directions, respectively. Neuronal activity was strongly affected by response direction as indicated by the consistent elevation of thick above thin lines after appearance of the directional cue. Effects of anticipated delay length would be manifest as differences in firing rate between trials in which the response direction (indicated by line thickness) was the same but expected delay length (indicated by color) was different. It is evident from the close coincidence of red and blue curves that the impact of anticipated delay on firing rate was weak. To characterize the time course of activity modulated by anticipated delay length, we independently computed indices reflecting 1) the impact of expected delay length on net firing rate independent of direction and 2) the impact of expected delay length on the strength of the directional signal. The impact on net firing rate was measured with an index representing the average amount by which the firing rate increased under the short-delay condition. It was computed as (SP + SA – LP – LA)/2, where SP is the firing rate under the short-delay, preferred-direction condition; LA is the firing rate under the long-delay, antipreferred-direction condition; and so on. The impact of anticipated delay length on firing rate was variable over time and negligible in strength (Fig. 5B1). The effect of delay on the directional signal was represented by an index that corresponded to the average amount by which short-delay caused the difference in firing rate between preferred-direction and antipreferred-direction trials to increase. It was computed as (SP – SA – LP + LA)/2. This index hovered around zero before the directional signal and then became slightly positive (Fig. 5D1).
|
Main effect of delay.
Counts of neurons exhibiting a significant main effect of delay length on firing rate are shown in Fig. 5C1, where blue (or red) symbols represent the percentage of cases in which firing was increased (or decreased) for short compared with long delay trials. During each epoch (I–III), the full count of neurons exhibiting a main effect of delay (whether this took the form of significantly stronger or significantly weaker firing under the short-delay as compared with the long-delay condition) was significantly in excess of the frequency (0.05) expected by chance from type 1 errors (
2 test, P < 0.05). This observation can be reconciled with the observation that there was little effect of delay on the average activity of the neuronal population (as indicated in the population histogram) by noting that neurons increasing and decreasing their firing rate on short-delay trials were equally common, so that the effects canceled at the population level. During no epoch was there a significant difference in number between neurons firing more strongly and those firing more weakly before a short delay (2 test, P > 0.05).
Interaction between delay and direction.
Counts of neurons exhibiting a significant interaction effect are shown in Fig. 5E1, where blue (or red) symbols represent the percentage of cases in which the directional signal was stronger (or weaker) on short-delay trials. Counts during epoch I necessarily represent type 1 errors because it was only after this epoch that the directional cue appeared. During epoch III, the proportion of neurons exhibiting an interaction effect was significantly in excess of the frequency expected by chance (2 test, P < 0.05). During no epoch was there a significant difference in number between neurons exhibiting stronger direction selectivity and those exhibiting weaker direction selectivity before a short delay (2 test, P > 0.05).
INDIVIDUAL NEURONS ACROSS A LONG ANTICIPATORY EPOCH. To generate for each neuron a single statistical measure of the impact of predicted delay length on the neuronal firing rate, we carried out an ANOVA using, as the dependent variable, the mean firing rate across the entire period from onset of the delay cue to 250 ms after offset of the directional cue, taking as factors both expected delay length and instructed response direction. The results are presented in Table 1 and Fig. 11.
|
2 test, P < 0.05). The difference in frequency between neurons firing more strongly and those firing more weakly before a short delay was not significant (2 test, P > 0.05).
Interaction between delay and direction.
Interaction effects were no more common than expected by chance (2 test, P > 0.05). The difference in frequency between cases in which direction selectivity was stronger before a short delay and those in which it was weaker was not significant (2 test, P > 0.05). These results did not differ significantly between monkeys.
SUMMARY. Among PFC neurons, the length of the anticipated delay exerted a subtle influence on firing rate. A few neurons fired more strongly and a few more weakly in anticipation of a short delay.
Prefrontal cortex (PFC): elapsed delay
POPULATION. In curves representing the population firing rate as a function of time, it is evident that activity was enhanced after a long delay as compared to a short delay (Fig. 5A2). This was true both when the response was in the neuron's preferred direction (the thick red curve lies above thick blue curve) and when it was in the opposite direction (the thin red curve lies above the thin blue curve). The enhancement was present before saccade initiation but was most marked after the saccade (downward-directed red regions in the difference histogram of Fig. 5B). In contrast to the impact of elapsed delay on net firing rate, there was almost no effect on the strength of the directional signal (Fig. 5D2).
INDIVIDUAL NEURONS BY EPOCH. The results, presented in Fig. 5, C2 and E2, can be summarized in the following terms.
Main effect of delay.
During each epoch (IV–VII), the proportion of neurons exhibiting a main effect was significantly in excess of the frequency expected by chance (2 test, P < 0.05). During epochs IV and VI, the preponderance of neurons that fired more strongly after a long delay was significant (2 test, P < 0.05).
Interaction between delay and direction.
During epochs IV, VI, and VII, the proportion of neurons exhibiting an interaction effect was significantly in excess of the frequency expected by chance (2 test, P < 0.05). During no epoch was there a significant difference in number between neurons exhibiting stronger direction selectivity and those exhibiting weaker direction selectivity after a long delay (2 test, P > 0.05).
INDIVIDUAL NEURONS ACROSS A LONG PREMOVEMENT EPOCH. The results, presented in Table 3 and Fig. 17, can be summarized in the following terms.
|
|
2 test, P < 0.05). The preponderance of neurons firing more strongly after a long delay was significant (2 test, P < 0.05).
Interaction between delay and direction.
Interaction effects were significantly more common than expected by chance (2 test, P < 0.05). The preponderance of neurons in which the directional signal was stronger after a long delay was significant (2 test, P < 0.05). These results did not differ significantly between monkeys.
SUMMARY. Among PFC neurons, the length of the elapsed delay exerted a substantial influence on firing rate. A majority of delay-sensitive neurons fired more strongly after a long delay.
Frontal eye field (FEF): anticipated delay
POPULATION. We collected data from 124 neurons in the FEF of three monkeys (Table 1). Curves representing the activity of this population as a function of time during the trial indicate that the level of anticipated delay exerted only minor effects on neuronal activity early in the trial (Fig. 6A1).
|
Main effect of delay.
During each epoch (I–III), the proportion of neurons exhibiting a main effect was significantly in excess of the frequency expected by chance (2 test, P < 0.05). During no epoch was there a significant difference in number between neurons firing more strongly and those firing more weakly before a short delay (2 test, P > 0.05).
Interaction between delay and direction.
During no epoch was the proportion of neurons exhibiting an interaction effect significantly in excess of the frequency expected by chance (2 test, P > 0.05). During no epoch was there a significant difference in number between neurons exhibiting stronger direction selectivity and those exhibiting weaker direction selectivity before a short delay (2 test, P > 0.05).
INDIVIDUAL NEURONS ACROSS A LONG ANTICIPATORY EPOCH. The results, presented in Table 1 and Fig. 11, can be summarized in the following terms.
Main effect of delay.
There was a significant main effect of expected delay in 19% of FEF neurons. This percentage was significantly in excess of that expected by chance (2 test, P < 0.05). The difference in frequency between neurons firing more strongly and those firing more weakly before a short delay was not significant (2 test, P > 0.05).
Interaction between delay and direction.
Interaction effects were no more common than expected by chance (2 test, P > 0.05). The difference in frequency between cases in which direction selectivity was stronger before a short delay and those in which it was weaker was not significant (2 test, P > 0.05). These results did not differ significantly between monkeys.
SUMMARY. Among FEF neurons, the length of the anticipated delay exerted a moderate influence on firing rate. Some neurons fired more strongly and some more weakly in anticipation of a short delay.
Frontal eye field (FEF): elapsed delay
POPULATION. Population activity was enhanced after a long delay as compared to a short delay (Fig. 6A2). The time course of the enhancement is summarized in Fig. 6B2. There was no consistent effect of elapsed delay on the strength of the directional signal (Fig. 6D2).
Individual neurons by epoch. The results, presented in Fig. 6, C2 and E2, can be summarized in the following terms.
Main effect of delay.
During each epoch (IV–VII), the proportion of neurons exhibiting a main effect was significantly in excess of the frequency expected by chance (2 test, P < 0.05). During epochs IV and VI, the preponderance of neurons that fired more strongly after a long delay was significant (2 test, P < 0.05).
Interaction between delay and direction.
During epochs IV–VI, the proportion of neurons exhibiting an interaction effect was significantly in excess of the frequency expected by chance (2 test, P < 0.05). During no epoch was there a significant difference in number between neurons exhibiting stronger direction selectivity and those exhibiting weaker direction selectivity after a long delay (2 test, P > 0.05).
INDIVIDUAL NEURONS ACROSS A LONG PREMOVEMENT EPOCH. The results, presented in Table 3 and Fig. 17, can be summarized in the following terms.
Main effect of delay.
There was a significant main effect of expected delay in 43% of FEF neurons. This percentage was significantly in excess of that expected by chance (2 test, P < 0.05). The preponderance of neurons firing more strongly after a long delay was significant (2 test, P < 0.05).
Interaction between delay and direction.
Interaction effects were significantly more common than expected by chance (2 test, P < 0.05). The difference in number between neurons exhibiting stronger and weaker direction selectivity after a long delay was not significant (2 test, P > 0.05). These results did not differ significantly between monkeys.
SUMMARY. Among FEF neurons, the length of the elapsed delay exerted a substantial influence on firing rate. A majority of delay-sensitive neurons fired more strongly after a long delay.
FEF/PM: anticipated delay
POPULATION. We collected data from 34 neurons in FEF/PM of two monkeys (Table 1). Curves representing the activity of this population as a function of time during the trial indicate that the length of anticipated delay exerted a strong effect on neuronal activity (Fig. 7A1). The net firing rate was elevated shortly after the presentation of cues predicting short delays and the elevation persisted throughout the delay period (Fig. 7B1).
|
Main effect of delay.
During each epoch (I–III), the proportion of neurons exhibiting a main effect was significantly in excess of the frequency expected by chance (2 test, P < 0.05). During each epoch, the preponderance of neurons firing more strongly before a short delay was significant (2 test, P < 0.05).
Interaction between delay and direction.
During no epoch was the proportion of neurons exhibiting an interaction effect significantly in excess of the frequency expected by chance (2 test, P > 0.05). During no epoch was there a significant difference in number between neurons exhibiting stronger direction selectivity and those exhibiting weaker direction selectivity before a short delay (2 test, P > 0.05).
INDIVIDUAL NEURONS ACROSS A LONG ANTICIPATORY EPOCH. The results, presented in Table 1 and Fig. 11, can be summarized in the following terms.
Main effect of delay.
There was a significant main effect of expected delay in 24% of FEF/PM neurons. This percentage was significantly in excess of that expected by chance (2 test, P < 0.05). The preponderance of neurons firing more strongly before a short delay was significant (2 test, P < 0.05).
Interaction between delay and direction.
Interaction effects were no more common than expected by chance (2 test, P > 0.05). The difference in frequency between cases in which direction selectivity was stronger before a short delay and those in which it was weaker was not significant (2 test, P > 0.05). These results did not differ significantly between monkeys.
SUMMARY. Among FEF/PM neurons, the length of the anticipated delay exerted a substantial influence on firing rate. Most delay-sensitive neurons fired more strongly in anticipation of a short delay.
FEF/PM: elapsed delay
POPULATION. Population activity was markedly enhanced toward the end of a short delay versus toward the end of a long delay (Fig. 7A2). The sign of the effect reversed after saccade completion, as indicated by the transition from an upward-directed blue region to a downward-directed red region in Fig. 7B2. There was an apparent slight tendency for the directional signal to be stronger after a short delay (Fig. 7D2).
Individual neurons by epoch. The results, presented in Fig. 7, C2 and E2, can be summarized in the following terms.
Main effect of delay.
During epochs IV–VI, the proportion of neurons exhibiting a main effect was significantly in excess of the frequency expected by chance (2 test, P < 0.05) and the preponderance of neurons that fired more strongly after a short delay was significant (2 test, P < 0.05). During epoch VII, neurons that fired more strongly after a long delay were significantly preponderant.
Interaction between delay and direction.
During epoch IV, the proportion of neurons exhibiting an interaction effect was significantly in excess of the frequency expected by chance (2 test, P < 0.05). During no epoch was there a significant difference in number between neurons exhibiting stronger direction selectivity and those exhibiting weaker direction selectivity after a long delay (2 test, P > 0.05).
INDIVIDUAL NEURONS ACROSS A LONG PREMOVEMENT EPOCH. The results, presented in Table 3 and Fig. 17, can be summarized in the following terms.
Main effect of delay.
There was a significant main effect of elapsed delay in 38% of FEF/PM neurons. This percentage was significantly in excess of that expected by chance (2 test, P < 0.05). The preponderance of neurons firing more strongly after a short delay was significant (2 test, P < 0.05).
Interaction between delay and direction.
Interaction effects were no more common than expected by chance (2 test, P > 0.05). There was no difference in number between neurons exhibiting stronger and weaker direction selectivity after a long delay. These results did not differ significantly between monkeys.
SUMMARY. Among FEF/PM neurons, the length of the elapsed delay exerted a substantial influence on firing rate. Before and during the saccade, a majority of delay-sensitive neurons fired more strongly after a short delay. During a period beginning after the saccade and extending through reward delivery, this pattern was reversed.
Premotor cortex (PM): anticipated delay
POPULATION. We collected data from 76 neurons in PM of two monkeys (Table 1). Curves representing the activity of this population as a function of time during the trial indicate that the length of the anticipated delay exerted a strong effect on neuronal activity (Fig. 8A1). The net firing rate was sharply elevated throughout the period after presentation of the short-delay cue (Fig. 8B1). The strength of the directional signal was also moderately elevated (Fig. 8D1).
|
Main effect of delay.
During each epoch (I–III), the proportion of neurons exhibiting a main effect was significantly in excess of the frequency expected by chance (2 test, P < 0.05). During each epoch, the preponderance of neurons firing more strongly before a short delay was significant (2 test, P < 0.05).
Interaction between delay and direction.
During no epoch was the proportion of neurons exhibiting an interaction effect significantly in excess of the frequency expected by chance (2 test, P > 0.05). During no epoch was there a significant difference in number between neurons exhibiting stronger direction selectivity and those exhibiting weaker direction selectivity before a short delay (2 test, P > 0.05).
INDIVIDUAL NEURONS ACROSS A LONG ANTICIPATORY EPOCH. The results, presented in Table 1 and Fig. 11, can be summarized in the following terms.
Main effect of delay.
There was a significant main effect of expected delay in 28% of PM neurons. This percentage was significantly in excess of that expected by chance (2 test, P < 0.05). The preponderance of neurons firing more strongly before a short delay was significant (2 test, P < 0.05).
Interaction between delay and direction.
Interaction effects were no more common than expected by chance (2 test, P > 0.05). The difference in frequency between cases in which direction selectivity was stronger before a short delay and those in which it was weaker was not significant (2 test, P > 0.05). These results did not differ significantly between monkeys.
SUMMARY. Among PM neurons, the length of the anticipated delay exerted a dramatic influence on firing rate. Most delay-sensitive neurons fired more strongly in anticipation of a short delay.
Premotor cortex (PM): elapsed delay
POPULATION. Population activity was considerably enhanced toward the end of a short delay versus toward the end of a long delay (Fig. 8A2). There was an apparent slight tendency for the directional signal to be stronger after a short delay (Fig. 8D2).
INDIVIDUAL NEURONS BY EPOCH. The results, presented in Fig. 8, C2 and E2, can be summarized in the following terms.
Main effect of delay.
During each epoch (IV–VII), the proportion of neurons exhibiting a main effect was significantly in excess of the frequency expected by chance (2 test, P < 0.05). During epochs IV and V, the preponderance of neurons that fired more strongly after a short delay was significant (2 test, P < 0.05).
Interaction between delay and direction.
During epoch V, the proportion of neurons exhibiting an interaction effect was significantly in excess of the frequency expected by chance (2 test, P < 0.05). During no epoch was there a significant difference in number between neurons exhibiting stronger direction selectivity and those exhibiting weaker direction selectivity after a long delay (2 test, P > 0.05).
INDIVIDUAL NEURONS ACROSS A LONG PREMOVEMENT EPOCH. The results, presented in Table 3 and Fig. 17, can be summarized in the following terms.
Main effect of delay.
There was a significant main effect of elapsed delay in 42% of PM neurons. This percentage was significantly in excess of that expected by chance (2 test, P < 0.05). The preponderance of neurons firing more strongly after a short delay was significant (2 test, P < 0.05).
Interaction between delay and direction.
Interaction effects were no more common than expected by chance (2 test, P > 0.05). There was no difference in number between neurons exhibiting stronger and weaker direction selectivity after a long delay. These results did not differ significantly between monkeys.
SUMMARY. Among PM neurons, the length of the elapsed delay exerted a marked influence on firing rate. During the period leading up to the saccade, a majority of delay-sensitive neurons fired more strongly after a short delay.
Supplementary eye field (SEF): anticipated delay
POPULATION. We collected data from 147 neurons in the SEF of two monkeys (Table 1). Curves representing the activity of this population as a function of time during the trial indicate that the length of anticipated delay had only a minor impact on neuronal activity (Fig. 9, A1 and B1).
Individual neurons by epoch. The results, presented in Fig. 9, C1 and E1, can be summarized in the following terms.
Main effect of delay.
During epoch II, the proportion of neurons exhibiting a main effect was significantly in excess of the frequency expected by chance (2 test, P < 0.05). During no epoch was there a significant difference in number between neurons firing more strongly and those firing more weakly before a short delay (2 test, P > 0.05).
Interaction between delay and direction.
During no epoch was the proportion of neurons exhibiting an interaction effect significantly in excess of the frequency expected by chance (2 test, P > 0.05). During no epoch was there a significant difference in number between neurons exhibiting stronger direction selectivity and those exhibiting weaker direction selectivity before a short delay (2 test, P > 0.05).
INDIVIDUAL NEURONS ACROSS A LONG ANTICIPATORY EPOCH. The results, presented in Table 1 and Fig. 11, can be summarized in the following terms.
Main effect of delay.
There was a significant main effect of expected delay in 12% of SEF neurons. This percentage was significantly in excess of that expected by chance (2 test, P < 0.05). The difference in frequency between neurons firing more strongly and those firing more weakly before a short delay was not significant (2 test, P > 0.05).
Interaction between delay and direction.
Interaction effects were no more common than expected by chance (2 test, P > 0.05). The difference in frequency between cases in which direction selectivity was stronger before a short delay and those in which it was weaker was not significant (2 test, P > 0.05). These results did not differ significantly between monkeys.
SUMMARY. Among SEF neurons, the length of the anticipated delay exerted a very weak influence on firing rate. A few neurons fired more strongly and a few more weakly in anticipation of a short delay.
Supplementary eye field (SEF): elapsed delay
POPULATION. Population activity was enhanced after a long delay as compared to a short delay (Fig. 9A2). The enhancement was present during the late phase of the delay period and carried over into the saccadic period (Fig. 9B2). There was little or no apparent effect on the strength of the directional signal (Fig. 9D2).
INDIVIDUAL NEURONS BY EPOCH. The results, presented in Fig. 9, C2 and E2, can be summarized in the following terms.
Main effect of delay.
During each epoch (IV–VII), the proportion of neurons exhibiting a main effect was significantly in excess of the frequency expected by chance (2 test, P < 0.05). During epochs IV–VI, the preponderance of neurons that fired more strongly after a long delay was significant (2 test, P < 0.05).
Interaction between delay and direction.
During epoch IV, the proportion of neurons exhibiting an interaction effect was significantly in excess of the frequency expected by chance (2 test, P < 0.05). During no epoch was there a significant difference in number between neurons exhibiting stronger direction selectivity and those exhibiting weaker direction selectivity after a long delay (2 test, P > 0.05).
INDIVIDUAL NEURONS ACROSS A LONG PREMOVEMENT EPOCH. The results, presented in Table 3 and Fig. 17, can be summarized in the following terms.
Main effect of delay.
There was a significant main effect of expected delay in 20% of SEF neurons. This percentage was significantly in excess of that expected by chance (2 test, P < 0.05). The preponderance of neurons firing more strongly after a long delay was significant (2 test, P < 0.05).
Interaction between delay and direction.
Interaction effects were significantly more common than expected by chance (2 test, P < 0.05). The difference in number between neurons exhibiting stronger and weaker direction selectivity after a long delay was not significant (2 test, P > 0.05). These results did not differ significantly between monkeys.
SUMMARY. Among SEF neurons, the length of the elapsed delay exerted a dramatic influence on firing rate. A majority of delay-sensitive neurons fired more strongly after a long delay.
Rostral supplementary motor area (SMAr): anticipated delay
POPULATION. We collected data from 84 neurons in the SMAr of one monkey (Table 1). Curves representing the activity of this population as a function of time during the trial indicate that the length of the anticipated delay exerted a substantial effect on neuronal activity (Fig. 10A1). There was a considerable increase of firing rate beginning shortly after the delay cue on short-delay trials (Fig. 10B1).
|
Main effect of delay.
During each epoch (I–III), the proportion of neurons exhibiting a main effect was significantly in excess of the frequency expected by chance (2 test, P < 0.05). During each epoch, the preponderance of neurons firing more strongly before a short delay was significant (2 test, P < 0.05).
Interaction between delay and direction.
During epoch III, the proportion of neurons exhibiting an interaction effect was significantly in excess of the frequency expected by chance (2 test, P < 0.05). During no epoch was there a significant difference in number between neurons exhibiting stronger direction selectivity and those exhibiting weaker direction selectivity before a short delay (2 test, P > 0.05).
INDIVIDUAL NEURONS ACROSS A LONG ANTICIPATORY EPOCH. The results, presented in Table 1 and Fig. 11, can be summarized in the following terms.
Main effect of delay.
There was a significant main effect of expected delay in 45% of SMAr neurons. This percentage was significantly in excess of that expected by chance (2 test, P < 0.05). The preponderance of neurons firing more strongly before a short delay was significant (2 test, P < 0.05).
Interaction between delay and direction.
Interaction effects were no more common than expected by chance (2 test, P > 0.05). The difference in frequency between cases in which direction selectivity was stronger before a short delay and those in which it was weaker was not significant (2 test, P > 0.05). These results did not differ significantly between monkeys.
SUMMARY. Among SMAr neurons, the length of the anticipated delay exerted a dramatic influence on firing rate. Most delay-sensitive neurons fired more strongly in anticipation of a short delay.
Rostral supplementary motor area (SMAr): elapsed delay
POPULATION. Population activity was markedly enhanced toward the end of a short delay versus toward the end of a long delay (Fig. 10A2). However, as the time for the saccade approached and passed, there was a complex series of reversals in the pattern of dependency of firing rate on antecedent delay length (Fig. 10B2). There was no consistent relation between delay length and the strength of the directional signal (Fig. 10D2).
INDIVIDUAL NEURONS BY EPOCH. The results, presented in Fig. 10, C2 and E2, can be summarized in the following terms.
Main effect of delay.
During epochs IV–VI, the proportion of neurons exhibiting a main effect was significantly in excess of the frequency expected by chance (2 test, P < 0.05). During epochs IV and VI, the preponderance of neurons that fired more strongly after a short (or long) delay was significant (2 test, P < 0.05).
Interaction between delay and direction.
During no epoch was the proportion of neurons exhibiting an interaction effect significantly in excess of the frequency expected by chance (2 test, P > 0.05). During no epoch was there a significant difference in number between neurons exhibiting stronger direction selectivity and those exhibiting weaker direction selectivity after a long delay (2 test, P > 0.05).
INDIVIDUAL NEURONS ACROSS A LONG PREMOVEMENT EPOCH. The results, presented in Table 3 and Fig. 17, can be summarized in the following terms.
Main effect of delay.
There was a significant main effect of expected delay in 32% of SMAr neurons. This percentage was significantly in excess of that expected by chance (2 test, P < 0.05). The preponderance of neurons firing more strongly after a short delay was not significant (2 test, P > 0.05).
Interaction between delay and direction.
Interaction effects were no more common than expected by chance (2 test, P > 0.05). The difference in number between neurons exhibiting stronger and weaker direction selectivity after a long delay was not significant (2 test, P > 0.05). These results did not differ significantly between monkeys.
SUMMARY. Among SMAr neurons, the length of the elapsed delay exerted a mixed influence on firing rate. During the period leading up to the saccade, a majority of delay-sensitive neurons fired more strongly after a short delay. This pattern reversed, however, in the perisaccadic period.
Selectivity for anticipated delay: comparison among areas
To determine whether the impact of anticipated delay length varied systematically across areas, we carried out area-to-area comparisons based on counts of neurons in which firing depended significantly on expected delay length, direction, and a delay x direction interaction (Table 1; Fig. 11). Several systematic interareal differences were clearly evident on comparison.
ENHANCEMENT IN ANTICIPATION OF SHORT DELAY. Neurons firing significantly more strongly under the short-delay condition became steadily more frequent with progress in a posteriorward direction across the cortical surface (Fig. 11A, black bars). This was true in both the lateral frontal lobe (PFC < FEF < FEF/PM < PM) and the medial frontal lobe (SEF < SMAr). Most pairwise interareal differences in the frequency of neurons exhibiting stronger firing on short-delay trials were highly significant (Table 2).
|
2 test) was significantly higher in the FEF than in the PFC (P < 0.01) and PM (P < 0.05) and was significantly higher in the SEF than in the PFC (P < 0.001), PM (P < 0.0001), and SMAr (P < 0.01). IMPACT OF ANTICIPATED DELAY ON DIRECTIONAL SIGNALS. Neurons exhibiting a statistically significant dependency of firing rate on the interaction between delay length and response direction were rare in all areas. The frequency with which these neurons were observed in the total sample (0.048) was in fact no greater than the expected rate of type 1 errors (0.05).
Selectivity for anticipated delay: impact of reversing the cue-delay associations
At the end of every 40 successful trials in the variable-delay task, the cue previously associated with short delay became associated with long delay and vice versa. Consequently, in each 80-trial data collection session, there was one block conforming to each cue convention. This manipulation allowed us to consider the influence of anticipated delay length on neuronal activity independently of any selectivity neurons may have possessed for the visual attributes of the stimuli. However, it may have resulted in an attenuation of activity related to the anticipated delay. This would be true if it took monkeys many trials to adjust their expectations after each switch. We addressed this concern by asking how long it took monkeys to adjust to the new cue-delay contingencies.
We first assessed the impact of the switch in cue significance on the behavioral reaction time. To achieve power in this analysis, we combined reaction time measures across all data collection sessions in all monkeys, considering only those trials that the monkey completed successfully. The results are shown in Fig. 12A. This plots, as a function of trial number, the reaction time to the preswitch short-delay cue minus the reaction time to the preswitch long-delay cue. The values were negative before the switch because the monkeys responded more swiftly on trials in which there was a short delay. The values became positive after the switch for the same reason. The index was not initially positive after the switch but became reliably positive by the sixth postswitch trial. The fact that the shift did not occur instantly indicates, surprisingly, that it depended on the monkey's registering the information conveyed by the cue and did not depend on just the experienced duration of the delay. We conclude that it took the monkeys around five trials to register in their performance an awareness of the switched cue-delay contingencies.
|
It is conceivable that neuronal activity after the switch might have adjusted at different rates in different areas. To assess this possibility, we analyzed data separately from each area with a substantial predelay firing rate enhancement on short-delay trials. To compensate for the smaller sample sizes, we carried out a coarser analysis, analyzing firing rates on blocks of four consecutive correct trials, with blocks demarcated so that the time of the switch fell at a between-block boundary. The results are shown in Fig. 12, C–E. These plots indicate that neuronal activity adjusted quickly to the new rule regardless of area. We conclude that the rule-switching design did not lead to a major attenuation of neuronal activity dependent on expected delay length.
Selectivity for anticipated delay: location of selective neurons
To analyze the fine distribution of neurons sensitive to anticipated delay, we projected recording sites onto the cortical surface. For three lateral chambers (in monkeys P, N, and F), we collected MR images parallel to the cortical surface. Using a set of slices that contained both the cortex and fiducial markers at known locations relative to the recording grid, we determined the locations of recording sites relative to the arcuate (AS) and principal (PS) sulci. The results are shown in Fig. 13, A–C. In this figure, the size of each symbol indicates the proportion of neurons at the corresponding site in which there was a significant enhancement of firing rate on short-delay trials during the predelay epoch (delay-cue onset to directional-cue offset). The general tendency for enhancement under short-delay conditions to occur at posterior sites and, in particular, at sites behind the arcuate sulcus, is clear. In the vicinity of the FEF and FEF/PM, neurons represented by a single symbol occupied depths ranging from 0 to 4 mm (mean = 2.60 mm; median = 2.00 mm). However, there was no consistent trend with respect to depth. Neurons showing a significant effect of delay length were observed throughout the range of recording depths (mean = 2.34 mm; median = 2.25 mm). For three midline chambers (in monkeys A, N, and F), we collected frontoparallel MR images showing the cortex and fiducial markers at known locations relative to the recording grid. These images were used to determine the locations of recording sites relative to the interhemispheric midline and the frontal plane containing the genu of the arcuate sulcus (AS genu). The results are shown in Fig. 13, D–F. The general tendency for enhancement under conditions of anticipated short delay to occur at relatively posterior sites—in monkey F, as contrasted to relatively anterior sites, in monkeys A and N—is clear.
|
Like anticipation of a big reward, anticipation of a short delay increased the monkey's level of task engagement as demonstrated by a reduced fixation-break rate (Fig. 2C). Were the two effects related? To answer this question, we recorded from a large number of neurons in the context of both the variable-reward and the variable-delay task: 186, 113, 34, 66, 136, and 84 neurons in PFC, FEF, FEF/PM, PM, SEF, and SMAr, respectively. To determine whether neurons that fired more strongly (or more weakly) under big-reward conditions also fired more strongly (or more weakly) under short-delay conditions we computed a reward index and a delay index on firing rates during the predelay epoch for each neuron averaged across direction. The reward index was (BR – SR)/(BR + SR) where BR and SR were the firing rates on big and small reward trials during the predelay epoch (reward-cue onset to directional-cue offset). The delay index was (SD – LD)/(SD + LD), where SD and LD were the firing rates on short- and long-delay trials during the predelay epoch (delay-cue onset to directional-cue offset). On plotting the reward index against the delay index for all neurons in each area, we found that there was a significant positive correlation in each area (Fig. 14). The correlation coefficient, like the incidence of big-reward and short-delay enhancement, increased with posteriorward movement across the cortical surface both in lateral frontal cortex (PFC < FEF < FEF/PM < PM) and in medial frontal cortex (SEF < SMAr). We conclude that neurons that fired more strongly when a big reward was anticipated also tended to fire more strongly when a short delay was anticipated.
|
|
|
To determine whether the impact of elapsed delay length varied systematically across areas, we carried out area-to-area comparisons based on counts of neurons in which firing depended significantly on delay length, direction, and their interaction (Table 3; Fig. 17). Several systematic interareal differences were clearly evident on comparison.
SIGN OF DELAY SENSITIVITY. Among delay-sensitive neurons, those firing more strongly after a long delay (gray bars in Fig. 17A) outnumbered those firing more strongly after a short delay (black bars in Fig. 17A) in three relatively anterior areas (PFC, FEF, and SEF), whereas the opposite pattern was present in three relatively posterior areas (FEF/PM, PM, and SMAr). Statistical analysis confirmed this picture (Table 4). The ratio of long-delay–preferring to short-delay–preferring neurons was significantly different in every case where an area from the anterior triad was compared with an area from the posterior triad, whereas it fell short of significance in every case where two areas from the same triad were compared, with the sole exception of SMAr and FEF/PM.
|
2 test). As an incidental point, we may note that the direction selectivity in both FEF and SEF appeared to reverse after completion of the saccade (Figs. 6A2 and 9A2). The apparent reversal may have reflected the preparation and initiation of a return saccade—in the neuron's preferred direction after an initial saccade in its nonpreferred direction and vice versa.
IMPACT OF ELAPSED DELAY ON DIRECTIONAL SIGNALS.
In data pooled across all areas, neurons exhibiting a statistically significant dependency of firing rate on the interaction between delay length and response direction were observed with a frequency (0.11) significantly in excess of the frequency (0.05) expected by chance (2 test, P < 0.0001). Nevertheless, the counts in individual areas were so low as to render all interareal differences insignificant.
Selectivity for elapsed delay: time course of long-delay enhancement in rostral areas
The tendency for neurons in PFC, FEF, and SEF to exhibit enhanced activation at the end of a long delay (Fig. 17A, gray bars; Table 3) could have arisen from either of two mechanisms. 1) The activity of these neurons could have depended solely on elapsed time. In other words, their firing rate could have increased gradually over the course of the delay, attaining a higher level after a long delay because the increase was more prolonged. 2) Their activity could have depended on some process set in motion at the time of presentation of the delay cue. In other words, they could have exhibited enhanced firing even early in a long-delay trial.
To distinguish between these possibilities, we constructed population histograms representing firing rate as a function of time during the trial for all neurons in PFC, FEF, and SEF that fired significantly more strongly at the end of a long delay than at the end of a short delay (Fig. 18, A–C). As a basis for comparison, we also constructed population histograms for the remaining neurons, which did not show significant long-delay enhancement (Fig. 18, G–I). In each display, blue and red lines represent population activity on preferred-direction short-delay and long-delay trials respectively, with alignment on directional-cue onset. For short-delay conditions, activity starting 250 ms before the imperative cue (offset of the fixation spot) is replotted (in light blue) with alignment on the imperative cue, to allow for comparison of premovement activity between short-delay and long-delay conditions.
|
Selectivity for elapsed delay: relation to saccadic reaction time
Behavioral reaction times were significantly shorter on short-delay than on long-delay trials in all monkeys. This raises the question: Was delay-related neuronal activity directly correlated with the length of the delay or, alternatively, was it directly correlated with the behavioral reaction time and correlated with delay only through this dependency? To resolve this issue, for each neuron, we used a multiple least-squares regression approach to optimize the parameters of three models representing firing rate as a linear function of reaction time and delay length: 1) a reduced model incorporating reaction time only, 2) a reduced model incorporating delay length only, and 3) a full model incorporating both. This analysis was based on a period (immediately before the behavioral response) extending from 250 ms before to 100 ms after the imperative cue (offset of the fixation spot). It was carried out independently for trials involving each neuron's preferred and antipreferred directions because the relation between reaction time and neuronal activity might vary according to whether the saccade was toward or away from the neuron's response field. We compared each of the reduced models to the full model using a nested F-test (see METHODS). Then we calculated the percentage of neurons in each area that showed a significant improvement of fit only when the variable of delay length was added to the model (Fig. 19, black bars), only when reaction time was added (Fig. 19, white bars), or when either was added (Fig. 19, hatched bars).
|
2 test, P < 0.05) and 2) that the difference between groups in the relative frequency of long-delay–preferring and short-delay–preferring neurons was significant (2 test, P < 0.05). This was true both for trials requiring a response in the neuron's preferred direction and for trials requiring a response in the antipreferred direction.
RELATION OF FIRING RATE TO REACTION TIME.
In all areas, neurons firing significantly more strongly on trials with short reaction times were preponderant (Fig. 19, downward-pointing white and lightly hatched bars). The preponderance achieved significance in all cases except the case of neurons in the posterior group of areas (FEF/PM, PM, SMAr) before saccades in the antipreferred direction (2 test, P < 0.05).
RELATION OF DELAY-RELATED TO REACTION-TIME-RELATED EFFECTS. Delay-related effects frequently occurred in the absence of reaction-time–related effects (black portions vs. hatched portions of dark bars in Fig. 19). Moreover, in anterior areas (PFC, FEF, and SEF), the preponderant form of delay-related activity (stronger firing after a long delay) was opposite to the preponderant form of reaction-time–related activity (stronger firing before a fast reaction). Thus we can rule out the interpretation that delay-related firing was simply a secondary consequence of the relation between neuronal firing rate and reaction time. Nevertheless, in the posterior group of areas (FEF/PM, PM, and SMAr), the tendency for a neuron to exhibit one kind of effect might have been correlated with its tendency to exhibit the other. To investigate this possibility, we computed, for each neuron in the posterior group, 1) the mean firing rate on short-delay trials minus the mean firing rate on long-delay trials and 2) the slope of a line fitted to points representing firing rate versus reaction time across all trials requiring a response in a given direction. We then analyzed the correlation across neurons between the two measures. On preferred-direction trials, the correlation was weak (r2 = 0.196) but significant (P < 0.0001). On antipreferred-direction trials, it was very weak (r2 = 0.010) but still significant (P < 0.0001). Thus in posterior areas, there was a weak but significant tendency for neurons that fired more strongly after a short delay (with reaction time factored out) to fire more strongly before a faster behavioral reaction (with the length of the antecedent delay factored out).
EMG measures
Activity related to the length of the delay might have been present during the delay period in PM, FEF/PM, and SMAr because monkeys were engaging in skeletomotor behavior or were preparing to make skeletomotor responses at the trial's end. The monkeys in this study did not engage in consistent overt licking during the delay period, such as has been observed in tasks involving multisecond delays (Hassani et al. 2001; Watanabe 2001). Nevertheless, they may have contracted the muscles of the jaw and neck during the delay period or may have maintained a state of preparation for contraction of these muscles. If they did so at all, then they might have done so to different degrees when anticipating a long or a short delay. Thus firing dependent on the length of the anticipated delay might have been related primarily to contracting or preparing to contract the neck and jaw muscles and only secondarily to delay length. To assess this possibility, we recorded EMG activity in the three monkeys (P, N, and F) still available for study after the collection and analysis of single-neuron data, focusing on the splenius capitus, responsible for ipsilateral version and torsion of the head, and the masseter, responsible for jaw closure and a commonly used indicator of oral activity (Apicella et al. 1991; Huang et al. 1989; Murray et al. 1991). We collected data from the splenius capitus during eight, twelve, and ten 400-trial sessions and from the masseter during eight, eight, and nine 400-trial sessions in monkeys N, P, and F, respectively. Combining the data for each muscle in each monkey across all trials in all sessions, we analyzed the dependency of the EMG on experimental condition by carrying out an ANOVA with delay length and direction as factors for each of the epochs (I–VII in Fig. 20) on which neuronal data analysis had been based.
|
EMG IN MASSETER. Effects of predicted length of delay on masseter EMG for the most part paralleled those on splenius capitus EMG (Fig. 20, D–F). In monkey P (Fig. 20D), EMG activity was enhanced under short-delay conditions, with the enhancement attaining significance in epochs I and III (P < 0.05). EMG activity was stronger under short-delay conditions in multiple epochs in monkey N (I and IV–VII) and in monkey F (I–VII). Thus expectation of a short delay induced stronger contraction of the jaw muscles in all three monkeys.
EMG IN RELATION TO NEURONAL ACTIVITY. The fact that motor output before the onset of the delay reflected the delay's anticipated length underscores the complexity of the processes set in motion by the delay cue. These were not restricted to central processes concerned with attention, arousal, and motor readiness but included overt motor output. Thus delay-dependent neuronal activity could have been related either to covert or to overt processes.
DELAY-RELATED EMG IN COMPARISON TO REWARD-RELATED EMG.
There is a striking similarity between EMG activity dependent on predicted delay, as described here, and EMG activity dependent on predicted reward, as described in a previous study (Roesch and Olson 2003; Fig. 16). Under each manipulation, two monkeys (P and N) showed relatively weak anticipatory EMG effects, whereas the third monkey (F) showed intense anticipatory activation of both the splenius capitus and the masseter muscles from the instant of the predictive cue onward. This parallel argues for the view that predicting a large reward and predicting a short delay called into play equivalent state changes in each monkey.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
We have recorded single-neuron activity in a large set of frontal cortical areas in monkeys performing an ocular delayed-response task in which a cue presented early in the trial predicted whether a short or long delay would elapse before the signal to respond. The areas studied were dorsolateral prefrontal cortex in and around the principal sulcus (PFC), the frontal eye field (FEF), a rostral zone of premotor cortex in which electrical stimulation elicited both upper-body and eye movements (FEF/PM), premotor cortex proper (PM), the supplementary eye field (SEF), and a rostral region of the supplementary motor area (SMAr). During the early period of the trial, when events and their timing were identical but the anticipated delay was either short or long, neuronal activity in comparatively posterior areas (FEF/PM, PM, and SMAr) was considerably higher when a short delay was anticipated. During the interval preceding and accompanying saccade initiation, when the immediately preceding elapsed delay had been either short or long, neuronal activity in comparatively anterior areas (PFC, FEF, and SEF) was substantially higher when a long delay had occurred.
Neuronal activity dependent on anticipated delay: does it reflect motivation?
Delay-related neuronal activity might have arisen from motivational modulation of neuronal activity related to motor preparation, attention, or arousal. Monkeys perform with greater speed and accuracy when a more valued reward is expected (Hassani et al. 2001; Hollerman et al. 1998; Kawagoe et al. 1998, 2004; Kobayashi et al. 2002; Lauwereyns et al. 2002a,Lauwereyns et al. 2002b; Leon and Shadlen 1999; Musallam et al. 2004; Roesch and Olson 2003, 2004; Takikawa et al. 2002; Tremblay et al. 1998; Watanabe 1990; Watanabe et al. 2001) just as rats run faster for a higher concentration of sucrose (Stellar 1982). This indicates that neural processes underlying response preparation and execution are enhanced when motivation is high (Stellar and Stellar 1985). In the present experiment, speed and accuracy measures were not appropriate indicators of motivational level because they were influenced by elapsed delay as distinct from anticipated delay. However, the fact that fixation breaks were less frequent on short- than on long-delay trials (Fig. 3C) indicates that the monkeys were more motivated.
If one grants the proposition that the prominent delay-related activity of neurons in FEF/PM, PM, and SMAr could have arisen from motivational modulation of preparatory processes, then the question arises: What was the nature of the modulated processes? There are at least four plausible scenarios that involve motivational modulation of motor signals in delay tasks such as the one used here. 1) Response preparation. Neurons representing the plan for the instructed response might fire more strongly when the monkey is more motivated. This scenario can account for the relatively infrequent effect whereby directional signals were enhanced on short-delay trials (Fig. 11C, black bars). However, it cannot explain the much more frequent cases in which enhancement occurred independently of response direction (Fig. 11A, black bars), nor can it account for the fact that areas in which directionally selective neurons were most prominent (Fig. 11B, FEF and SEF) do not coincide areas in which delay effects were most prominent (Fig. 11A, FEF/PM, PM, and SMAr). 2) Arousal. Neuronal activity sensitive to arousal or responsible for generalized motor readiness might be enhanced when the monkey is more motivated. It is not clear why a state of behavioral arousal should be accompanied by an increase in population activity in premotor cortex or how such an increase could give rise to generalized motor readiness. However, with such issues resolved, this scenario could explain the fact that neuronal activity tends to increase on short-delay trials. 3) Synergistic muscle activity. Neuronal activity governing overt behaviors that automatically accompany response planning, such as increases of axial tonus, might be enhanced when the monkey is more motivated. This scenario is compatible with our finding that delay-period neck EMG activity was enhanced on short-delay trials in one monkey. The absence of delay-induced modulations of neck EMG in the other two monkeys seems to argue against it. However, the other monkeys may have engaged in overt behaviors involving other muscles in which we did not monitor EMG. 4) Ingestive preparation. Neurons involved in preparing ingestive movements might be more active when the monkey is more motivated. None of these scenarios can be absolutely ruled out. Each may contribute to delay-related activity. To assess the contribution of each will require further experiments.
Neuronal activity dependent on anticipated delay: does it represent value?
Neurons that fired more strongly when the monkey anticipated receiving reward at a shorter delay might have represented the time-discounted value of the reward. It is reasonable to assume that monkeys placed greater value on the reward if they expected it after a shorter delay. It has been observed across a wide range of species, extending from pigeons to humans, that animals prefer a smaller, immediate reward over a larger, delayed one (Cardinal et al. 2001; Evenden and Ryan 1996; Herrnstein 1961; Ho et al. 1999; Lowenstein 1992; Mobini et al. 2002; Montague and Berns 2002; Thaler 1981). The view that monkeys placed greater value on a reward anticipated at short delay is supported both by the fact that they were more motivated (as discussed in the preceding section) and by the fact that neurons in orbitofrontal cortex fired more strongly in response to the cue predicting a reward at short delay (unpublished observations from ongoing experiments). Neuronal activity in orbitofrontal cortex is related to the magnitude of reward value and not to the intensity of movement preparation (Roesch and Olson 2004).
These considerations support the view that the monkeys placed greater value on the reward when the anticipated delay was short but they do not establish that delay-related neuronal activity represented value in an economically meaningful sense—as a factor in the decision how hard to work. There is a fundamental difference between 1) neuronal activity representing reward value in the service of an economic decision process and 2) neuronal activity reflecting motivational modulation of preparatory processes. However, in experimental paradigms involving manipulation of the value of the anticipated reward, value and motivation are correlated, so that it is impossible to determine whether neuronal activity is related to one or the other (Maunsell 2004; Roesch and Olson 2004). Under these circumstances, neuronal firing even in parietal area LIP is correlated with value (Dorris and Glimcher 2004; Glimcher 2002; Platt and Glimcher 1999; Sugrue et al. 2004). However, it would be wrong to conclude from the presence of a correlation that this activity represents value.
A simple example will suffice to make this point. In some monkeys, EMG activation of the neck muscles is greater when they are working for a more valued reward (Roesch and Olson 2004). It does not follow that neck muscle activation represents value in an economically meaningful sense (in other words, that the monkey decides to work harder because the neck muscles are more tense). Causality works in the other direction: the neck muscles are more tense because the monkey has decided to work harder. The same cautionary note applies to neuroeconomic interpretations of reward-related activity anywhere in the brain. Neuronal activity may be correlated with "experienced value" (Sugrue et al. 2004), "action value" (Barraclough et al. 2004), or "subjective desirability of action" (Dorris and Glimcher 2004) and yet not represent value in any sense more meaningful than the sense in which the neck muscles represent it.
It might be argued that neurons active in conjunction with planning a particular behavior (for example, making a saccade to the left) represent value in at least a narrow sense—that they represent the value to the monkey of executing the action with which their firing is associated (Barraclough et al. 2004; Dorris and Glimcher 2004; Glimcher 2002, 2003; Platt and Glimcher 1999; Sugrue et al. 2004). This argument is questionable for two reasons. First, if we consider the case of value-dependent neck-muscle activation, then it is clear that activity correlated with value need not represent value in an economically meaningful sense. Second, it seems unlikely that decision processes are based on the values associated with actions (as encoded by neurons representing those actions) rather than on the values associated with goals (as encoded by neurons representing those goals). Animals decide between possible goals on the basis of their value (an affective process). They select actions on the basis of their efficacy in contributing to the attainment of the chosen goal (a strategic process). In Lashley's classic study, rats trained to run to the baited arm of a maze, if prevented by a cerebellar lesion from running, would roll (Lashley and McCarthy 1926). They chose the goal on the basis of its value and used whatever means they could to get to it. For these reasons, we reject the conclusion that neuronal activity in any of the areas that we have studied represents the time-discounted value even in the limited sense of representing the value attached to the action with which their firing is associated.
Neuronal activity dependent on anticipated delay: does it reflect temporal preparation?
The tendency of neurons in FEF/PM, PM, and SMAr to fire more strongly after a cue predicting a short delay than after a cue predicting a long delay might be related to the timing of the monkey's preparation to respond—early under expectation of a short delay but late, and therefore outside the immediate postcue period—under expectation of a long delay. In tasks where a warning cue precedes an imperative cue at a predictable interval, preparation is maximal (as indicated by the fact that reaction times are minimal) at the predicted time (Nobre 2001). Neuronal signals apparently related to temporal preparation have been observed in several cortical areas. In the supplementary eye field of monkeys performing a smooth pursuit task, neuronal activity peaks around the time of an expected change in direction of the target (Heinen and Liu 1997). In area V4 of monkeys performing a detection task, attentional modulation of neuronal visual responses peaks around the time when the target event is most likely to occur (Ghose and Maunsell 2002). In the lateral intraparietal area of monkeys performing an interval-estimation task, neuronal activity shifts from representing the saccade by which a short interval will be reported to representing the saccade by which a long interval will be reported at the point in time demarcating a short from a long interval (Leon and Shadlen 2003). Might it be true, by analogy to the findings described above, that stronger firing immediately after the cue on short-delay trials reflected early preparation to respond? We believe that this cannot be the entire explanation of the effect. If the neuronal firing rate were related simply to timed preparation, then one would expect that the firing should be as strong on long-delay as on short-delay trials immediately before the behavioral response. Contrary to this prediction, firing was consistently weaker on long-delay trials even during the period immediately preceding the response. This was true in all three areas showing robust short-delay enhancement: FEF/PM (Fig. 7A2), PM (Fig. 8A2), and SMAr (Fig. 10A2).
Comparison to results obtained by manipulating reward size
In a previous paper (Roesch and Olson 2003), we reported that the incidence of reward-related neuronal activity increased steadily along the anterior–posterior axis in both the lateral frontal lobe (PFC < FEF < FEF/PM < PM) and the medial frontal lobe (SEF < SMAr). This finding was surprising because the comparatively posterior areas in which reward-related activity was strongest (FEF/PM, PM, and SMAr) were farthest removed connectionally from limbic cortex and thus might have been supposed to be least likely to process information related to reward value. As a means for resolving this paradox, we suggested that the robust reward-related activity of neurons in FEF/PM, PM, and SMAr reflected the monkey's motivationally modulated level of motor readiness, arousal, or attention and did not represent the value of the reward in an economically meaningful sense. The results of the previous study (concerned with anticipated reward size) and the results of the present study (concerned with anticipated delay to reward) are mutually illuminating. 1) The results of the current study replicate, under a new set of experimental conditions, the results of the old. The tendency for neurons to fire strongly in anticipation of the more valuable outcome (a short delay) was far more robust in posterior areas, FEF/PM, PM, and SMAr, than in anterior areas, PF, FEF, and SEF (cf. Fig. 11A in this report and Fig. 11A in Roesch and Olson 2003). 2) The results of the previous study constrain the interpretation of results from the current one. Inasmuch as areas showing a strong dependency on reward size also showed a strong dependency on delay length, and inasmuch as individual neurons sensitive to one factor also tend to be sensitive to the other (Fig. 14), we conclude i) that one mechanism underlies both forms of sensitivity and ii) that short-delay enhancement therefore cannot be attributed exclusively to timed readiness.
Neuronal activity dependent on elapsed delay
Anterior areas: enhanced activity after long delays.
In around a fifth of neurons in PFC, FEF, and SEF, the firing rate was significantly enhanced, immediately before the behavioral response, at the end of a long delay as compared with a short delay (Fig. 17A). These neurons significantly outnumbered neurons showing the opposite effect in all three areas. To understand better the nature of long-delay enhancement, we analyzed firing rate as a function of time for all neurons that exhibited the effect in each area. In all three areas, we found that neurons with long-delay enhancement exhibited a steady buildup of firing rate during the delay period as if their activity were indeed dependent on elapsed time (Fig. 18, A–C). However, we also found that their firing rate was somewhat enhanced, even very early in the trial, when the monkey merely anticipated a long rather than a short delay (Fig. 18, A–C). In the SEF, the enhancement appeared at a visual-response latency after onset of the delay cue (Fig. 18C). In PFC and FEF, it appeared only in response to onset of the target array, as if dependent on gating of neuronal visual responsiveness by the monkey's anticipatory state. One possible explanation for these effects is that PFC, FEF, and SEF are recruited into a higher rate of activity under more challenging circumstances. This would fit with other findings on the SEF indicating that net neuronal activity is higher when monkeys 1) must select targets according to a pattern-location–association rule versus a spatial rule (Olson et al. 2000), 2) are being guided by a novel versus familiar set of pattern-location associations (Chen and Wise 1995a,Chen and Wise 1995b), 3) are preparing to look away from versus toward a cue (Olson and Gettner 2002), or 4) are preparing to follow an antisaccade versus a prosaccade rule (Amador et al. 2004; Schlag-Rey et al. 1997). The nature of the contribution to task performance arising from difficulty-related enhancement is not known. One important insight arising from the present study is that the enhancement of firing rate occurring on long-delay trials is not ubiquitous but rather is confined to comparatively anterior frontal areas.
|
GRANTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
FOOTNOTES |
|---|
Address for reprint requests and other correspondence: M. R. Roesch, University of Maryland School of Medicine, Department of Anatomy and Neurobiology, HSF-2, Room S251, 20 Penn Street, Baltimore, MD 21201 (E-mail: mroes001{at}umaryland.edu)
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Apicella P, Ljungberg T, Scarnati E, and Schultz W. Responses to reward in monkey dorsal and ventral striatum. Exp Brain Res 85: 491–500, 1991.[Web of Science][Medline]
Barraclough DJ, Conroy ML, and Lee D. Prefrontal cortex and decision making in a mixed-strategy game. Nat Neurosci 7: 404–410, 2004.[CrossRef][Web of Science][Medline]
Cardinal RN, Pennicott DR, Sugathapala CL, Robbins TW, and Everitt BJ. Impulsive choice induced in rats by lesions of the nucleus accumbens core. Science 292: 2499–2501, 2001.
Chen LL and Wise SP. Neuronal activity in the supplementary eye field during acquisition of conditional oculomotor associations. J Neurophysiol 73: 1101–1121, 1995a.
Chen LL and Wise SP. Supplementary eye field contrasted with the frontal eye field during acquisition of conditional oculomotor associations. J Neurophysiol 73: 1122–1134, 1995b.
Coe B, Tomihara K, Matsuzawa M, and Hikosaka O. Visual and anticipatory bias in three cortical eye fields of the monkey during an adaptive decision-making task. J Neurosci 22: 5081–5090, 2002.
Crist CF, Yamasaki DSG, Komatsu H, and Wurtz RH. A grid system and a microsyringe for single cell recording. J Neurosci Meth 26: 117–122, 1988.[CrossRef][Web of Science][Medline]
Dorris MC and Glimcher PW. Activity in posterior parietal cortex is correlated with the relative subjective desirability of action. Neuron 44: 365–378, 2004.[CrossRef][Web of Science][Medline]
Evenden JL and Ryan CN. The pharmacology of impulsive behaviour in rats: the effects of drugs on response choice with varying delays of reinforcement. Psychopharmacology (Berl) 128: 161–170, 1996.[CrossRef][Medline]
Fujii N, Mushiake H, and Tanji J. An oculomotor representation area within the ventral premotor cortex. Proc Natl Acad Sci USA 95: 12034–12037, 1998.
Fujii N, Mushiake H, and Tanji J. Rostrocaudal distinction of the dorsal premotor area based on oculomotor involvement. J Neurophysiol 83: 1764–1769, 2000.
Ghose GM and Maunsell JHR. Attentional modulation in visual cortex depends on task timing. Nature 419: 616–620, 2002.[CrossRef][Medline]
Glimcher P. Decisions, decisions, decisions: choosing a biological science of choice. Neuron 36: 323–332, 2002.[CrossRef][Web of Science][Medline]
Glimcher P. Decisions, Uncertainty and the Brain: The Science of Neuroeconomics. Cambridge, MA: MIT Press, 2003.
Hassani OK, Cromwell HC, and Schultz W. Influence of expectation of different rewards on behavior-related neuronal activity in the striatum. J Neurophysiol 85: 2477–2489, 2001.
Heinen SJ and Liu M. Single-neuron activity in the dorsomedial frontal cortex during smooth-pursuit eye movements to predictable target motion. Vis Neurosci 14: 853–865, 1997.[Web of Science][Medline]
Herrnstein R. Relative and absolute strength of response as a function of frequency of reinforcement. J Exp Anal Behav 4: 267–272, 1961.[CrossRef][Web of Science][Medline]
Hikosaka K and Watanabe M. Delay activity of orbital and lateral prefrontal neurons of the monkey varying with different rewards. Cereb Cortex 10: 263–271, 2000.
Hikosaka O. Role of basal ganglia in saccades. Rev Neurol 145: 580–586, 1989.[Medline]
Ho MY, Mobini S, Chiang TJ, Bradshaw CM, and Szabadi E. Theory and method in the quantitative analysis of "impulsive choice" behaviour: implications for psychopharmacology. Psychopharmacology (Berl) 146: 362–372, 1999.[CrossRef][Medline]
Hollerman JR, Tremblay L, and Schultz W. Influence of reward expectation on behavior-related neuronal activity in primate striatum. J Neurophysiol 80: 947–963, 1998.
Huang CS, Hiraba H, Murray GM, and Sessle BJ. Topographical distribution and functional properties of cortically induced rhythmical jaw movements in the monkey (Macaca fascicularis). J Neurophysiol 61: 635–650, 1989.
Kalivas PW and Nakamura M. Neural systems for behavioral activation and reward. Curr Opin Neurobiol 9: 223–227, 1999.[CrossRef][Web of Science][Medline]
Kawagoe R, Takikawa Y, and Hikosaka O. Expectation of reward modulates cognitive signals in the basal ganglia. Nat Neurosci 1: 411–416, 1998.[CrossRef][Web of Science][Medline]
Kawagoe R, Takikawa Y, and Hikosaka O. Reward-predicting activity of dopamine and caudate neurons—a possible mechanism of motivational control of saccadic eye movement. J Neurophysiol 91: 1013–1024, 2004.
Kobayashi S, Lauwereyns J, Koizumi M, Sakagami M, and Hikosaka O. Influence of reward expectation on visuospatial processing in macaque lateral prefrontal cortex. J Neurophysiol 87: 1488–1498, 2002.
Lashley KS and McCarthy DA. The survival of the maze habit after cerebellar injuries. J Comp Psychol 6: 423–433, 1926.[CrossRef]
Lauwereyns J, Takikawa Y, Kawagoe R, Kobayashi S, Koizumi M, Coe B, Sakagami M, and Hikosaka O. Feature-based anticipation of cues that predict reward in monkey caudate nucleus. Neuron 33: 463–473, 2002a.[CrossRef][Web of Science][Medline]
Lauwereyns J, Watanabe K, Coe B, and Hikosaka O. A neural correlate of response bias in monkey caudate nucleus. Nature 418: 413–417, 2002b.[CrossRef][Medline]
Leon MI and Shadlen MN. Effect of expected reward magnitude on the response of neurons in the dorsolateral prefrontal cortex of the macaque. Neuron 24: 415–425, 1999.[CrossRef][Web of Science][Medline]
Leon MI and Shadlen MN. Representation of time by neurons in the posterior parietal cortex of the macaque. Neuron 38: 317–327, 2003.[CrossRef][Web of Science][Medline]
Lowenstein G and Elster J. Choice Over Time: New York: Russell Sage Foundation, 1992.
Matsumoto K, Suzuki W, and Tanaka K. Neuronal correlates of goal-based motor selection in the prefrontal cortex. Science 301: 229–232, 2003.
Maunsell JH. Neuronal representations of cognitive state: reward or attention? Trends Cogn Sci 8: 261–265, 2004.[CrossRef][Web of Science][Medline]
Mobini S, Body S, Ho MY, Bradshaw CM, Szabadi E, Deakin JF, and Anderson IM. Effects of lesions of the orbitofrontal cortex on sensitivity to delayed and probabilistic reinforcement. Psychopharmacology (Berl) 160: 290–298, 2002.[CrossRef][Medline]
Mogenson GJ, Jones DL, and Yim CY. From motivation to action: functional interface between the limbic system and the motor system. Prog Neurobiol 14: 69–97, 1980.[CrossRef][Web of Science][Medline]
Montague PR and Berns GS. Neural economics and the biological substrates of valuation. Neuron 36: 265–284, 2002.[CrossRef][Web of Science][Medline]
Murray GM, Lin LD, Moustafa EM, and Sessle BJ. Effects of reversible inactivation by cooling of the primate face motor cortex on the performance of a trained tongue-protrusion task and a trained biting task. J Neurophysiol 65: 511–530, 1991.
Mussallam S, Corneil BD, Greger B, Scherberger H, and Anderson RA. Cognitive control signals for neural prosthetics. Science 9: 162–163, 2004.
Nobre AC. Orienting attention to instants in time. Neuropsychologia 39: 1317–1328, 2001.[CrossRef][Web of Science][Medline]
Olson CR. Object-based vision and attention in primates. Curr Opin Neurobiol 11: 171–179, 2001.[CrossRef][Web of Science][Medline]
Olson CR and Gettner SN. Neuronal activity related to rule and conflict in macaque supplementary eye field. Physiol Behav 77: 663–670, 2002.[CrossRef][Medline]
Olson CR, Gettner SN, Ventura V, Carta R, and Kass RE. Neuronal activity in macaque supplementary eye field during planning of saccades in response to pattern and spatial cues. J Neurophysiol 84: 1369–1384, 2000.
Ono T, Nishijo H, and Nishino H. Functional role of the limbic system and basal ganglia in motivated behaviors. J Neurol 247 Suppl. 5: V23–V32, 2000a.
Platt ML and Glimcher PW. Neural correlates of decision variables in parietal cortex. Nature 400: 233–238, 1999.[CrossRef][Medline]
Robinson DA. A method of measuring eye movements using a scleral search coil in a magnetic field. IEEE Trans Biomed Eng 10: 137–145, 1963.[Medline]
Roesch MR and Olson CR. Impact of expected reward on neuronal activity in prefrontal cortex, frontal and supplementary eye fields and premotor cortex. J Neurophysiol 90: 1766–1789, 2003.
Roesch MR and Olson CR. Neuronal activity related to reward value and motivation in primate frontal cortex. Science 304: 307–310, 2004.
Schlag-Rey M, Amador N, Sanchez H, and Schlag J. Antisaccade performance predicted by neuronal activity in the supplementary eye field. Nature 390: 398–401, 1997.[CrossRef][Medline]
Stellar E. The Physiological Mechanisms of Motivation and Reward, edited by Pfaff D.W. New York, Springer-Verlag, 1982. p 377–407.
Stellar JR and Stellar E. The Neurobiology of Motivation and Reward. New York: Springer-Verlag, 1985.
Sugrue LP, Corrado GS, and Newsome WT. Matching behavior and the representation of value in the parietal cortex. Science 304: 1782–1787, 2004.
Takikawa Y, Kawagoe R, and Hikosaka O. Reward-dependent spatial selectivity of anticipatory activity in monkey caudate neurons. J Neurophysiol 87: 508–515, 2002.
Takikawa Y, Kawagoe R, Itoh H, Nakahara H, and Hikosaka O. Modulation of saccadic eye movements by predicted reward outcome. Exp Brain Res 142: 284–291, 2002c.[CrossRef][Web of Science][Medline]
Thaler R. Some empirical evidence on dynamic inconsistency. Econ Lett 8: 201–207, 1981.
Tremblay L, Hollerman JR, and Schultz W. Modifications of reward expectation-related neuronal activity during learning in primate striatum. J Neurophysiol 80: 964–977, 1998.
Tremblay L and Schultz W. Relative reward preference in primate orbitofrontal cortex. Nature 398: 704–708, 1999.[CrossRef][Medline]
Tremblay L and Schultz W. Reward-related neuronal activity during go–no go task performance in primate orbitofrontal cortex. J Neurophysiol 83: 1864–1876, 2000.
Wallis JD and Miller EK. From rule to response: neuronal processes in the premotor and prefrontal cortex. J Neurophysiol 90: 1790–1806, 2003a.
Wallis JD and Miller EK. Neuronal activity in primate dorsolateral and orbital prefrontal cortex during performance of a reward preference task. Eur J Neurosci 18: 2069–2081, 2003d.[CrossRef][Web of Science][Medline]
Watanabe M. Prefrontal unit activity during associative learning in the monkey. Exp Brain Res 80: 296–309, 1990.[Web of Science][Medline]
Watanabe M. Frontal units of the monkey coding the associative significance of visual and auditory stimuli. Exp Brain Res 89: 233–247, 1992.[CrossRef][Web of Science][Medline]
Watanabe M. Reward expectancy in primate prefrontal neurons. Nature 382: 629–632, 1996.[CrossRef][Medline]
Watanabe M, Cromwell HC, Tremblay L, Hollerman JR, Hikosaka K, and Schultz W. Behavioral reactions reflecting differential reward expectations in monkeys. Exp Brain Res 140: 511–518, 2001.[CrossRef][Web of Science][Medline]
Watanabe M, Hikosaka K, Sakagami M, and Shirakawa S. Coding and monitoring of motivational context in the primate prefrontal cortex. J Neurosci 22: 2391–2400, 2002.
This article has been cited by other articles:
|
|
 |
|
  N. D. Walsh and M. L. Phillips Interacting Outcome Retrieval, Anticipation, and Feedback Processes in the Human Brain Cereb Cortex, May 8, 2009; (2009) bhp098v1. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. Gregorios-Pippas, P. N. Tobler, and W. Schultz Short-Term Temporal Discounting of Reward Value in Human Ventral Striatum J Neurophysiol, March 1, 2009; 101(3): 1507 - 1523. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. K. Berdyyeva and C. R. Olson Monkey Supplementary Eye Field Neurons Signal the Ordinal Position of Both Actions and Objects J. Neurosci., January 21, 2009; 29(3): 591 - 599. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. Minamimoto, G. La Camera, and B. J. Richmond Measuring and Modeling the Interaction Among Reward Size, Delay to Reward, and Satiation Level on Motivation in Monkeys J Neurophysiol, January 1, 2009; 101(1): 437 - 447. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Rezvani and B. D. Corneil Recruitment of a Head-Turning Synergy by Low-Frequency Activity in the Primate Superior Colliculus J Neurophysiol, July 1, 2008; 100(1): 397 - 411. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J.-W. Sohn and D. Lee Order-Dependent Modulation of Directional Signals in the Supplementary and Presupplementary Motor Areas J. Neurosci., December 12, 2007; 27(50): 13655 - 13666. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Heinzle, K. Hepp, and K. A. C. Martin A Microcircuit Model of the Frontal Eye Fields J. Neurosci., August 29, 2007; 27(35): 9341 - 9353. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Genovesio, S. Tsujimoto, and S. P. Wise Neuronal Activity Related to Elapsed Time in Prefrontal Cortex J Neurophysiol, May 1, 2006; 95(5): 3281 - 3285. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. R. Roesch and C. R. Olson Neuronal Activity in Primate Orbitofrontal Cortex Reflects the Value of Time J Neurophysiol, October 1, 2005; 94(4): 2457 - 2471. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
