Download Neuronal activity in dorsomedial frontal cortex and prefrontal cortex

Exp Brain Res (2001) 139:116–119
DOI 10.1007/s002210100760
RESEARCH NOTE
Nan-Hui Chen · Ilsun M. White · Steven P. Wise
Neuronal activity in dorsomedial frontal cortex and prefrontal cortex
reflecting irrelevant stimulus dimensions
Received: 27 February 2001 / Accepted: 22 March 2001 / Published online: 11 May 2001
© Springer-Verlag 2001
Abstract Previous studies of the dorsomedial frontal
cortex (DMF) and the prefrontal cortex (PF) have shown
that, when monkeys respond to nonspatial features of a
discriminative stimulus (e.g., color) and the stimulus
appears at a place unrelated to the movement target,
neurons nevertheless encode stimulus location. This
observation could support the idea that these neurons
always encode stimulus location, regardless of its
relevance to an instrumentally conditioned behavior. Past
studies, however, leave open the possibility that activity
observed during one operant task might reflect the
contingencies of a different task, performed at different
times. To test these alternatives, we examined the activity
of DMF and PF neurons in two rhesus monkeys conditioned to perform an operant eye-movement task in
which only the color and shape of visual stimuli served
as salient discriminative features. Each of eight stimuli was
associated with a response to a different eye-movement
target. The location of these stimuli varied from trial to
trial but was of no behavioral relevance, and the
monkeys did not perform any operant task in which
stimulus location controlled behavior. A substantial
minority of neurons in both DMF and PF nevertheless
encoded stimulus location, which indicates that this
property does not depend on its relevance in an instrumentally conditioned behavior.
N.-H. Chen
KunMing Institute of Zoology, 32 JiaoChang DongLu, KunMing,
YunNan 650223, P.R. China
I.M. White
Behavioral Neurobiology, Institute of Toxicology,
Swiss Federal Institute of Technology Zurich,
8603 Schwerzenbach, Switzerland
S.P. Wise (✉)
Section on Neurophysiology, Laboratory of Systems Neuroscience,
National Institute of Mental Health, National Institute of Health,
49 Convent Drive, MSC 4401, Building 49/Room B1EE17,
Bethesda, MD 20892-4401, USA
e-mail: [email protected]
Tel.: +1-301-4025481, Fax: +1-301-4025441
Keywords Visually guided movement ·
Supplementary eye field · Frontal cortex ·
Medial eye field · Monkey
Introduction
A persistent issue in behavioral neurophysiology is the
extent to which neuronal activity reflects operant conditioning. For example, Mann et al. (1988) have concluded
that the properties of the dorsomedial frontal cortex
(DMF), also known as the supplementary eye field,
changed dramatically to reflect the task that a monkey
had been conditioned to perform. Notwithstanding more
recent results (Tehovnik and Slocum 2000), the possibility
that instrumental conditioning determines the response
properties of cortical neurons remains an open issue, one
with special importance to frontal cortex physiology.
Stimulus location has been consistently shown to affect
activity in both DMF (Olson et al. 2000; White and Wise
1999) and prefrontal cortex (PF) neurons (Asaad et al.
2000; Hoshi et al. 1998; White and Wise 1999), even in
operant tasks involving responding exclusively to nonspatial stimulus dimensions such as color and shape.
This finding suggests that these frontal networks process
spatial information regardless of its relevance to an operant
behavior. An alternative hypothesis, tested here, is that
stimulus location affects neuronal activity because that
feature is relevant in other operant tasks that the subjects
perform at different times.
Materials and methods
Two male rhesus monkeys (Macaca mulatta), 6–8 kg, sat in a
primate chair facing a video screen. The study was approved by
the NIMH ACUC and conformed with the Guide for the care and
use of laboratory animals (NIH publication, 1996). The head of
each monkey was fixed and eye movements monitored at 200
samples/s with an infrared oculomotor (Bois Instruments).
The first monkey pressed a bar to initiate a trial, which began
with the presentation of a fixation spot (0.4°, 5° window) at screen
center. The monkey needed to fixate that spot for 0.75–3.25 s,
117
Fig. 1 A Stimuli used, in monochromatic form, placed at the target
location that each stimulus signaled. Each stimulus could appear
at any of nine places: one of the eight targets or the fixation point
(FP). B Example stimulus and response (arrow). C Spatial tuning
for saccade direction, mean ± SEM discharge rate, in a PF neuron.
Saccade direction is indicated vectorially, below. Note the preference
for saccades with upward and leftward components. D Activity of
a different PF neuron, showing preference for leftward stimulus
locations. (X central location)
otherwise the trial was aborted. If fixation was maintained, a
stimulus appeared at either the central fixation point or one of
eight potential saccade targets 10° from center (Fig. 1B) for
0.12–0.20 s. There were eight stimuli, each measuring ~2.5°
(Fig. 1A) and composed of two elements of various shapes, hues,
orientations, and sizes. Each stimulus, selected and located pseudorandomly, signaled a correct eye movement or, equivalently, a
response target. Stimulus location was irrelevant. Next, the eight
potential response targets appeared simultaneously. The monkey
had to make a saccade to the correct target within 2.75 s and maintain gaze (2° window) for 0.6–2.7 s, after which the target spot
dimmed. Then the monkey was required to release the bar quickly
to receive reinforcement, a 0.1 to 0.3-ml drop of water. The task
for the second monkey differed somewhat. There was no bar to
press; each trial began automatically. The discriminative stimuli
were divided into two sets of four, those with targets separated by
90°. One of these two sets was associated with targets in the cardinal
directions from center (0°, 90°, 180°, and 270°), the other with the
intermediate target locations (45°, 135°, 225°, and 315°). On each
trial, only the targets associated with that set appeared, and reinforcement was given for the fixation contingencies alone.
Cranial implantation was performed aseptically and with the
monkeys under isofluorane anesthesia (1%–3%). The monkeys
received banamine (0.5 mg/kg i.m.) postoperatively. Glass-coated,
platinum-iridium electrodes (1–2 MΩ measured at 1 kHz) recorded
frontal cortex activity. Single-unit potentials were filtered with a
Table 1 Numbers of neurons
showing significant effects of
stimulus location and saccade
direction (one-way ANOVA,
P<0.05). A neuron can contribute to both stimulus location
and saccade direction tabulations. (N number of task related
units, DMF dorsomedial frontal
cortex, PF prefrontal cortex)
Subject area
bandpass of 600 Hz to 6 kHz, amplified and discriminated using a
multi-spike detector (Alpha-Omega Engineering).
The location of DMF was confirmed with intracortical microstimulation. Trains of 11 cathodal constant-current pulses were
delivered at 350 pulses/s, using a stimulus isolation unit (PSIU-6;
Grass Instruments). Each pulse was 200–350 µs in duration. A ball
electrode was used for less precise localization.
Near the end of physiological data collection, electrolytic
lesions (15 µA for 10 s, anodal current) were made along five
tracks in the first monkey, six in the second. After the monkey was
perfused with 10% formol-saline, marking pins were inserted.
Each brain was sectioned on a freezing microtome at 40 µm thickness, mounted on glass slides, and stained with thionin. We plotted
recording sites by reference to the electrolytic lesions and pin
locations.
For each neuron’s activity, spatial tuning was evaluated by
separate one-way ANOVAs (P<0.05) for the effects of saccade
direction and stimulus location. Reaction time was tested similarly.
A significant task relationship was examined separately for neurons
showing significant spatial tuning. Modulations above referenceperiod activity levels were tested by a one-way ANOVA, comparing
reference-period activity (100–500 ms before stimulus onset) and
the poststimulus response (50–500 ms after stimulus onset).
Results
Task performance was 95% and 91% correct, for the first
and second monkey, respectively. The first monkey
responded with a mean saccadic reaction time of
372±45 ms (SD), the second at 642±303 ms. Reaction
time varied by saccade direction and stimulus location,
but the differences were small.
Figure 1C, D shows the spatial tuning properties of
two PF neurons. The properties of DMF neurons were
similar. One of the illustrated cells showed significant
spatial tuning for saccade direction, regardless of the
location of the discriminative stimulus (Fig. 1C), but no
spatial tuning for stimulus location (not shown). The
other PF neuron was spatially tuned to stimulus location
(Fig. 1D), notwithstanding the fact that the location of
each discriminative stimulus was behaviorally irrelevant.
The spatial tuning for saccade direction was complex,
but also statistically significant, in this neuron (not
shown). Table 1 presents the numbers of neurons with
significant spatial tuning for saccade direction and for
stimulus location in both DMF and PF. Although more
neurons in both areas showed selectivity for saccade
direction, a significant minority was affected by stimulus
location. Overall, the spatial tuning functions resembled
those reported previously in both DMF (Schall 1991)
and PF (Funahashi et al. 1991).
Monkey 1 (N=112)
Monkey 2 (N=87)
Total (N=199)
Stimulus
location
Stimulus
location
Saccade
direction
Stimulus
location
Saccade
direction
Saccade
direction
DMF
PF
8
6
27
23
4
5
29
7
12
11
56
30
Total
n
%
14
13
50
45
9
10
36
41
23
12
86
43
118
Fig. 2A, B Electrode penetration sites. A Significant stimuluslocation effects (squares) or both saccade-direction and stimuluslocation effects (crosses). Symbol size indicates the number of
neurons of the indicated class at each coordinate. B Significant
saccade-direction effects (squares) or both effects (crosses). The
filled circles indicate the location of marking pins in the first
monkey. All plots are two-monkey composites referenced to the
PA landmark. (Arc arcuate sulcus, PA posterior limit of the arcuate
sulcus, Prin principal sulcus)
Figure 2 shows the recording sites in a composite map
based on bilateral recordings from DMF and unilateral
recordings from PF. Figure 2B shows that neurons with
statistically significant effects of saccade direction predominated in both DMF and PF (see also Table 1).
Figure 2A shows the location of the smaller number of
cells, in both areas, showing stimulus-location effects. A
proportion of neurons showed significant effects of both
stimulus location and saccade direction (Fig. 2A, B), but
most saccade-direction cells lacked stimulus-location
effects. Cells with activity reflecting stimulus location
were scattered throughout the sampled area, with no
evidence of segregation.
Discussion
As reviewed in detail (Tehovnik et al. 2000), DMF has
been viewed as an oculomotor area (Schall 1991),
although possible skeletomotor functions have received
increased attention recently (Mushiake et al. 1996; Schlag
et al. 1998). DMF represents space in both objectcentered (Olson and Gettner 1999; Olson and Tremblay
2000) and craniocentric coordinates (Lee and Tehovnik
1995; Schlag and Schlag-Rey 1987; Tehovnik and Lee
1993). However, DMF cells also respond to nonspatial
stimuli (Chen and Wise 1995; Olson et al. 2000). PF has
properties generally similar to DMF in that its cells,
including those in the posterior region of PF sampled
here, respond to both nonspatial and spatial aspects of
visual stimuli (Rainer et al. 1998; Rao et al. 1997) and
are affected by both nonspatial and spatial rules (Asaad
et al. 2000; Hoshi et al. 1998; White and Wise 1999).
The present study examined whether activity in DMF
and PF was affected by stimulus location, even when
that stimulus dimension was behaviorally irrelevant.
Previous studies of both PF (Rainer et al. 1998; Rao et
al. 1997; White and Wise 1999) and DMF (Olson et al.
2000; White and Wise 1999) have shown that stimulus
location influences neuronal activity in nonspatially
guided operant tasks. However, in each of those studies,
the monkeys alternated between tasks in which cue location was the relevant stimulus dimension and other tasks
in which it was not. Under such circumstances, one
might argue that cells in DMF and PF were affected by
stimulus location in nonspatially guided tasks because
spatial factors controlled responding in other tasks. The
present experiment overcame that problem because
stimulus location was never a differential discriminative
stimulus for responding. We found that stimulus location
was nevertheless encoded in a minority of DMF and PF
neurons.
Of course, it is impossible to rule out the possibility
that this finding resulted from the monkeys’ experience
in their home cage, in which the location of objects was
highly relevant to behavior. But DMF, at least, appears to
be specialized for operantly conditioned behavior. Its
cells are modulated in relation to eye movements made
for juice reinforcement, but not when monkeys make
saccades without primary reinforcement (Bon and
Lucchetti 1992; Lee and Tehovnik 1995). DMF neurons
also respond to juice delivery only in the context of
instrumental behavior, not when juice is delivered
randomly (Mann et al. 1988). Accordingly, we think it
unlikely that unconditioned cage behavior underlies the
encoding of stimulus location.
Although this discussion emphasizes stimulus-location
effects, such signals were relatively rare in both DMF
and PF. The activity of most tuned cells reflected
saccade direction and its correlated variables instead
(Table 1). We note the existence of covariates because
the present study did not entail any attempt to distinguish
whether neuronal selectivity reflected features of the
discriminative stimuli, the fixation targets or the saccadic
eye movements. We assume, however, that the spatial
factors predominate, in accordance with the spatial tuning
of cells in both DMF (Schall 1991; Schlag and
Schlag-Rey 1985, 1987; Schlag et al. 1998; Tehovnik et
al. 2000) and the parts of PF sampled here (Funahashi et
al. 1991), as well as with the presumed role of these
areas in the selection of goals and the guidance of movement. In comparison with previous reports, the present
results suggest that the irrelevancy of stimulus location
led to a paucity of neurons encoding that dimension of
the stimulus. This idea complements a finding from the
frontal eye field, where cells show increased selectivity
for nonspatial aspects of a visual stimulus after monkeys
have been trained to respond to those features (Bichot et
al. 1996).
We conclude that neurons in DMF and PF reflect
features of the sensory environment that are encountered
during operant conditioning, but that are irrelevant to the
119
conditioned behavior and to the function of DMF and PF
in the guidance of that behavior. Alternatively, it remains
possible that this minority of cells might use stimulus
location as a signal that the task does not call for analyzing
this variable. The former possibility might shed some
light on recent findings of unexpected signals in the
primary motor cortex (M1). Signals have been found in
M1 that reflect variables such as relative vibrotactile
frequency (Mountcastle et al. 1992; Salinas and Romo
1998) and the ordinal rank of visual cues (Carpenter et
al. 1999). In those studies, monkeys had been operantly
conditioned for extensive periods. Perhaps M1 neurons,
like the DMF and PF cells described here, also reflect
information encountered during operant conditioning,
even if that information is irrelevant to its function.
Acknowledgements Author contributions: experimental design
and animal training (S.W., I.W., N.-H.C.), data collection (N.-H.C.),
data analysis (N.-H.C.). We thank Mr. Robert Gelhard for preparing
the histological material.
References
Asaad WF, Rainer G, Miller EK (2000) Task-specific neural
activity in the primate prefrontal cortex. J Neurophysiol 84:
451–459
Bichot NP, Schall JD, Thompson KG (1996) Visual feature
selectivity in frontal eye fields induced by experience in
mature macaques. Nature 381:697–699
Bon L, Lucchetti C (1992) The dorsomedial frontal cortex of the
macaca monkey: fixation and saccade-related activity. Exp
Brain Res 89:571–580
Carpenter AF, Georgopoulos AP, Pellizzer G (1999) Motor cortical
encoding of serial order in a context-recall task. Science 283:
1752–1757
Chen LL, Wise SP (1995) Neuronal activity in the supplementary
eye field during acquisition of conditional oculomotor associations. J Neurophysiol 73:1101–1121
Funahashi S, Bruce CJ, Goldman-Rakic PS (1991) Neuronal activity
related to saccadic eye movements in the monkey’s dorsolateral
prefrontal cortex. J Neurophysiol 65:1464–1483
Hoshi E, Shima K, Tanji J (1998) Task-dependent selectivity of
movement-related neuronal activity in the primate prefrontal
cortex. J Neurophysiol 80:3392–3397
Lee K, Tehovnik EJ (1995) Topographic distribution of fixationrelated units in the dorsomedial frontal cortex of the rhesus
monkey. Eur J Neurosci 7:1005–1011
Mann SE, Thau R, Schiller PH (1988) Conditional task-related
responses in monkey dorsomedial frontal cortex. Exp Brain
Res 69:460–468
Mountcastle VB, Atluri PP, Ramo R (1992) Selective outputdiscriminative signals in the motor cortex of waking monkeys.
Cereb Cortex 2:277–294
Mushiake H, Fujii N, Tanji J (1996) Visually guided saccade
versus eye-hand reach: contrasting neuronal activity in the
cortical supplementary and frontal eye fields. J Neurophysiol
75:2187–2191
Olson CR, Gettner SN (1999) Macaque SEF neurons encode
object-centered directions of eye movements regardless of the
visual attributes of instructional cues. J.Neurophysiol. 81:
2340–2346
Olson CR, Tremblay L (2000) Macaque supplementary eye field
neurons encode object-centered locations relative to both
continuous and discontinuous objects. J.Neurophysiol. 83:
2392–2411
Olson CR, Gettner SN, Ventura V, Carta R, Kass RE (2000)
Neuronal activity in macaque supplementary eye field during
planning of saccades in response to pattern and spatial cues. J
Neurophysiol 84:1369–1384
Rainer G, Asaad WF, Miller EK (1998) Memory fields of neurons
in the primate prefrontal cortex. Proc Natl Acad Sci USA 95:
15008–15013
Rao SC, Rainer G, Miller EK (1997) Integration of what and
where in the primate prefrontal cortex. Science 276:821–824
Salinas E, Romo R (1998) Conversion of sensory signals into
motor commands in primary motor cortex. J.Neurosci 18:
499–511
Schall JD (1991) Neuronal activity related to visually guided
saccadic eye movements in the supplementary motor area of
rhesus monkeys. J Neurophysiol 66:530–558
Schlag J, Schlag-Rey M (1985) Unit activity related to spontaneous
saccades in frontal dorsomedial cortex of monkey. Exp Brain
Res 58:208–211
Schlag J, Schlag-Rey M (1987) Evidence for a supplementary eye
field. J Neurophysiol 57:179–200
Schlag J, Dassonville P, Schlag-Rey M (1998) Interaction of the
two frontal eye fields before saccade onset. J Neurophysiol 79,
64–72
Tehovnik EJ, Lee K (1993) The dorsomedial frontal cortex of the
rhesus monkey: topographic representation of saccades evoked
by electrical stimulation. Exp Brain Res 96:430–442
Tehovnik EJ, Slocum WM (2000) Effects of training on saccadic
eye movements elicited electrically from the frontal cortex of
monkeys. Brain Res 877:101–106
Tehovnik EJ, Sommer MA, Chou IH, Slocum WM, Schiller PH
(2000) Eye fields in the frontal lobes of primates. Brain Res
Brain Res Rev 32:413–448
White IM, Wise SP (1999) Rule-dependent neuronal activity in
the prefrontal cortex. Exp Brain Res 126:315–335