letters to nature
.........................................................................................................................

Methods

The treatment of the monkeys was in accordance with the guidelines set by the
US Department of Health and Human Services (NIH) for the care and use of
laboratory animals.
Behavioural tasks. The behavioural task and physiological techniques used
here have been described1. Briefly, in each trial, a fraction of the dots moved
coherently in one direction while the remaining dots appeared at random
locations within the stimulus aperture, creating a masking motion noise. The
strength of the coherent motion (per cent coherence) ranged from 0% (fully
random noise) to 51.2%. Each trial began when the monkey fixated a small
point of light on a video display (Fig. 1a). The monkey then viewed the random
dot display for a brief interval while gazing steadily at the fixation point.
Following a brief delay period, the monkey indicated the perceived direction of
motion by making a saccadic eye movement to one of two visual targets
corresponding to the two possible directions of motion. Extinction of the
fixation point and appearance of the two visual targets signalled the end of the
delay period. Eye movements were measured throughout all experiments by
the scleral search coil technique18. The monkey received fluid rewards for
correct choices.
Microstimulation techniques. For each microstimulation experiment, the
multi-unit receptive field was mapped and the preferred direction identified
using a bar of light or a patch of moving dots. To maximize the likelihood that
the monkey would use directional information supplied by neurons at the
stimulation site, the dot patterns were presented within the receptive field, and
the axis of the directional discrimination was aligned with the preferred-null
axis of the receptive field. Microstimulation (200 Hz, 10 mA, biphasic pulses)
was applied during the stimulus presentation interval or during the delay
interval for monkey K, and also during a fixation interval before the onset of the
visual stimulus (‘previsual’) for the other three monkeys (Fig. 1b). For monkeys
R, S and T, each microstimulation interval lasted 800 ms; 200 ms separated
stimulation intervals. For monkey K, microstimulation duration was 1 s with
500 ms separation between the two microstimulation intervals. In an extra ten
experiments, monkey R was tested with same time course as monkey K; the
results remained essentially the same as in Fig. 3c. The latter time course was
also used in the red–green experiment. In the variable delay experiment (Fig. 4),
microstimulation was applied in the last 800 ms of the delay period; the
duration of the delay period was varied randomly among three values—900 ms
(short), 1,800 ms (medium) and 2,700 ms (long).
Data analysis. Data were analysed using a standard logistic regression model19.
Microstimulation effects were measured as the horizontal shift of the psychometric function in stimulated trials relative to control trials. The amplitude of
this shift was measured in units of the abscissa (coherence, in percentage of
dots), where leftward shifts were considered to be positive. Throughout this
study, microstimulation did not significantly affect the slopes of the psychometric functions.
Received 20 October 1997; accepted 17 April 1998.
1. Salzman, C. D., Murasugi, C. M., Britten, K. H. & Newsome, W. T. Microstimulation in visual area
MT: effects on direction discrimination performance. J. Neurosci. 12, 2331–2355 (1992).
2. Murasugi, C. M., Salzman, C. D. & Newsome, W. T. Microstimulation in visual area MT: effects of
varying pulse amplitude and frequency. J. Neurosci. 13, 1719–1729 (1993).
3. Salzman, C. D. & Newsome, W. T. Neural mechanisms for forming a perceptual decision. Science 264,
231–237 (1994).
4. Dubner, R. & Zeki, S. M. Response properties and receptive fields of cells in an anatomically defined
region of the superior temporal sulcus. Brain Res. 35, 528–532 (1971).
5. Baker, J. F., Petersen, S. E., Newsome, W. T. & Allman, J. M. Visual response properties of neurons in
four extrastriate visual areas of the owl monkey (Aotus trivirgatus). J. Neurophysiol. 45, 397–416
(1981).
6. Van Essen, D. C., Maunsell, J. H. R. & Bixby, J. L. The middle temporal visual area in the macaque:
myeloarchitecture, connections, functional properties and topographic representation. J. Comp.
Neurol. 199, 293–326 (1981).
7. Maunsell, J. H. R. & Van Essen, D. C. Functional properties of neurons in the middle temporal visual
area (MT) of the macaque monkey: I. Selectivity for stimulus direction, speed and orientation. J.
Neurophysiol. 49, 1127–1147 (1983).
8. Albright, T. D. Direction and orientation selectivity of neurons in visual area MT of the macaque. J.
Neurophysiol. 52, 1106–1130 (1984).
9. Albright, T. D., Desimone, R. & Gross, C. G. Columnar organization of directionally selective cells in
visual area MT of macaques. J. Neurophysiol. 51, 16–31 (1984).
10. Treue, S. & Maunsell, J. H. R. Attentional modulation of visual motion processing in cortical areas MT
and MST. Nature 382, 539–541 (1996).
11. Posner, M. & Petersen, S. The attention system of the human brain. Annu. Rev. Neurosci. 13, 25–42
(1990).
12. Desimone, R. & Duncan, J. Neural correlates of selective visual attention. Annu. Rev. Neurosci. 18,
193–222 (1995).

NATURE | VOL 394 | 2 JULY 1998

13. Moran, J. & Desimone, R. Selective attention gates visual processing in the extrastriate cortex. Science
229, 782–784 (1985).
14. Motter, B. C. Neural correlates of feature selective memory and pop-out in extrastriate area V4. J.
Neurosci. 14, 2190–2199 (1994).
15. Mountcastle, V. B., Motter, B. C., Steinmetz, M. A. & Duffy, J. in Dynamic Aspects of Neurocortical
Function (eds Edelman, G. M., Gall, W. E. & Cowan, W. M.) 159–193 (Wiley, New York, 1984).
16. Corbetta, M., Miezin, F., Dobmeyer, S., Shulman, G. & Petersen, S. Selective and divided attention
during visual discrimiantions of shape, color, and speed: functional anatomy by positron emission
tomography. J. Neurosci. 11, 2383–2402 (1991).
17. Corbetta, M., Miezin, F., Shulman, G. & Petersen, S. A PET study of visuospatial attention. J. Neurosci.
13, 1202–1226 (1993).
18. Robinson, D. A. A method of measuring eye movement using a scleral search coil in a magnetic field.
IEEE Trans. Biomed. Eng. 10, 137 (1963).
19. Cox, D. R. Analysis of Binary Data (Methuen, London, 1970).

8

Acknowledgements. We thank J. Groh, G. DeAngelis, G. Horwitz, D. Chen and J. Nichols for criticism of
the manuscript, and J. Stein for technical assistance. We were supported by the National Eye Institute.
W.T.N. is an investigator of the Howard Hughes Medical Institute.
Correspondence should be addressed to E.S. (e-mail: eyal@monkeybiz.stanford.edu).

Seeing only the right half
of the forest but
cutting down all the trees?
Fabrizio Doricchi*† & Chiara Incoccia*
* Centro Ricerche di Neuropsicologia, IRCCS S. Lucia, Via Ardeatina 306,
00179 Roma, Italy
† Istituto di Psicologia ‘‘Luigi Meschieri’’, Universita’ degli Studi di Urbino,
Via Saffi 15, 61029, Urbino, Italy
.........................................................................................................................

Unilateral neglect following damage to the right hemisphere of
the brain can be characterized by failure of the global attentional
mechanisms of the right hemisphere to direct the local detail
processors of the left hemisphere towards the contralesional left
hemispace. This is suggested by patients who recognize the global
form of the left side of shapes (the forest) but fail to cancel out its
local details (the trees)1. Here we report the opposite behavioural
dissociation in a patient (Q.M.) with damage to the right hemisphere
of the brain. Q.M. detected local details (such as the tail of a dog)
on the left or right side of visual shapes, regardless of whether
these details belonged to predefined target shapes (a dog in this
case) or to distractor shapes differing on the opposite side (a dog
with a swan’s neck and head, for example). Psychological testing
showed an abnormal tendency of this patient to respond to local
features, but perfect accuracy in interpreting global features when
the local features could not interfere in global processing. The
results indicate that the left hemisphere can integrate multiple
local features simultaneously but loses global awareness as soon as
local features individually compete for response selection. However, awareness of the whole is not necessary for the sequential
processing of the parts.
Q.M. is a 57-year-old right-handed woman who suffered righthemisphere ischaemia (Fig. 1). Two months after the stroke, the
patient had severe contralesional neglect and hemianopia. We
examined this patient extensively in light of the following observations. When presented with an array of 3s and Bs (Fig. 2a, test a1)
and required to cancel out the 3s, Q.M. cancelled out both 3s and
Bs on the right side of the array, disregarding the presence of
the vertical bar on the left side of the Bs. Tilting the items and
varying the thickness of the vertical bar (Fig. 2a, tests a2, a3) did not
modify the patient’s performance. Unexpectedly, the same behaviour was observed when the items were mirror-reversed about the
vertical axis so that the difference between the two stimuli was on
the right, not left, side. These findings were replicated using 6s and
Ss (Fig. 2a, test b). When asked about her strategy, the patient
answered that she looked for items with ‘two little humps’ when
cancelling out Bs, and for items with an upper ‘hook’ when
cancelling out 6s.

Nature © Macmillan Publishers Ltd 1998

75

letters to nature

8
Figure 1 NMR scans of patient Q.M. Scans were taken at a, 5 days, and b,15 days

neglect were the following: line cancellation, 6 out of 21; letter cancellation,14 out

post-stroke. The right hemisphere lesion involves the occipital, temporal and

of 104; Wundt–Jastrow area illusion, 0 out of 20; sentence reading; 0 out of 6.

parietal lobes (Brodmann’s areas 17, 18, 19, 22, 37, 39 and 40) and the thalamic–

Higher scores indicate better performance.

capsular area. The patient’s scores on a battery of tests12 for assessment of

Figure 2 Cancellation tasks. a, Items of cancellation tasks a1–a3, b, and c1–c5.
b, Example of Q.M.’s performance on task c3 (rightward-heading items). The
ratios of frequencies of total cancelled targets to total cancelled distractors in the
canonically oriented (can), mirror-reversed (mrev), leftward-heading (lhead) and
rightward-heading (rhead) conditions were as follows: task a1 45:45 (can), 45:45
(mrev); task a2 39:39 (can), 42:42 (mrev); task a3 36:36 (can), 42:42 (mrev); task b
28:30 (can), 48:48 (mrev); task c1 16:16 (lhead),16:16 (rhead); task c2 16:16 (lhead),
12:12 (rhead); task c3 16:16 (lhead), 18:18 (rhead); task c4 20:20 (lhead), 16:20
(rhead); task c5 16:10 (lhead),16:15 (rhead).

Initially, the patient’s behaviour might have been attributed to
neglect based on object-centred coordinates2: in this case whatever
the orientation of targets and distractors, which were items
endowed with a canonical left–right orientation, the patient
would have paid attention only their canonical right side, neglecting
the left one. This hypothesis was ruled out by the finding that she
treated items having no canonical left–right orientation in the same
way that she treated 3s and 6s. When required to cancel out the line
drawings of dogs with elongated tails (intermixed with short-tailed
items), Q.M. omitted items on the left side of the array but cancelled
out both long-tailed dogs and distractors such as a dog with a
donkey’s head and neck, a dog with a deer’s head, or a chimaera
consisting of half-dog and a half-swan (Fig. 2a, b, tests c1–c3). In
76

searching for critical detail (that is, the elongated tail), the patient
disregarded the opposite half of the animal shape. This happened
independently of the leftward or rightward orientation of the items
and the type of array (rows by columns, tests c1–c3; scattered items,
tests c4, c5). Q.M.’s performance could not be explained by the
inability to recognize living items, as she correctly named (15 out of
15) line drawings of animals3. Object-centred neglect was not found
in Q.M. when copying a multi-item scene4, in a shape-comparison
task5 or when reading mirror-reversed words6.
These initial observations indicate the presence of both neglect
for the contralesional space and exacerbated narrowing of attention
towards local visual details. A pathological bias towards local
processing was demonstrated by tachistoscopic testing with hierarchical Navon letters7 (Fig. 3a) presented in the right hemifield.
When Q.M. had to pay attention both to local and global features,
she showed a severe deficit in detecting global ones (divided
attention task, Fig. 3c). The patient suffered from exaggerated
local-on-global interference and slowing of global processing even
when asked to actively disregard local features (global-directed
attention task, Fig. 3b). However, Q.M. recognized with perfect
accuracy the global shape of stimuli in which neutral local small
rectangles could neither help nor interfere with the recognition of
the global letter (Fig. 3a, b). This demonstrated the retained
simultaneous integration of multiple local elements in the attended
ipsilesional space when these elements could not compete for
response selection. Lack of interference from local rectangles following damage of the right hemisphere could also have been related
to the possible involvement of this hemisphere in the processing of
local features based on non-living objects8.
We evaluated the processing of global and local visual features in
the neglected space by presenting Q.M. with symmetrical and
asymmetrical hierarchical shapes (Fig. 4a). Q.M. cancelled out all
the local elements of all items, scanning each element by eye and
head movements with the spared visual hemifield, but she never
acknowledged the different global form of the left side of asymmetrical shapes. Q.M. was also asked to mark the centre of
symmetrical and asymmetrical hierarchical shapes having a longer
or shorter leftward extension (Fig. 4b). The centre of symmetrical
shapes was marked with a negligible rightward shift from the true
centre. This could have suggested effective analysis of global shape
on both sides of space. However, the centre of the asymmetrical
shapes was marked extremely close to the true centre of the
symmetrical ones. This showed that the influence of the global
contralesional extension of shapes was dramatically small. The
centre of symmetrical and asymmetrical lines lacking any vertical
component favouring the enlargement of the attentional focus
(Fig. 4c) was marked with a severe rightward shift, irrespective of
the leftward extension of the line.

Nature © Macmillan Publishers Ltd 1998

NATURE | VOL 394 | 2 JULY 1998

letters to nature
Q.M.’s case shows that left unilateral neglect together with
upregulated processing of local features does not necessarily
impair the direction of attention towards the contralesional local
components of hierarchical stimuli. Q.M.’s lesion spared the frontal
areas in the damaged hemisphere. Therefore, the sequential detection of local elements might have been based on the access of the
Congruent

a

Uncongruent

8

Local neutral

b

67

Q.M.

44
72
71

N+C

72

Global task

71
C1

70
71

C2

posterior local processors of the left hemisphere to anterior premotor and motor control mechanisms in both hemispheres. On the
other hand, Q.M.’s right hemisphere lesion destroyed the entire
network of parietal, temporal and occipital areas activated by global
processing in humans9,10. This finding and the associated maintenance of integration of multiple local features in the ipsilesional
space support the theory11 that, in patients with left unilateral
neglect, global processing can still be mediated by the intact left
hemisphere. However, the cancellation and tachistoscopic tasks
indicated that Q.M. was prone to the loss of conscious global
attention as soon as local features competed for response selection.
This result specifies, for the first time to our knowledge, a sharp
functional limit of the left hemisphere in global processing, and
provides a new view of the assignment of local and global processing
to merely relative left and right hemispheric specialization9,10.
Findings from positron emission tomography9 show that shifting
attention between local and global features depends on the influence of parietotemporal areas on prestriate visual cortex. Our
observations raise a related problem, namely, how to define
mechanisms allowing the maintenance of conscious awareness at
one attentional level, local or global, while processing the opposite

63
71

C3

65
72
72

C4
0

100

200

300

400

500

600

700

800

900

1000

800

900

1,000

69

Q.M.

72
71
70

N+C

Local task

72
71

C1
70
69

C2
67
69

C3

71
72

C4
0

100

200

300

400

500

600

700

ms

c

17 (48 missed)

Global

1
Q.M.

Local
0

200

68 (1 missed)
400

600

800

1,000
ms

1,200

1,400

1,600

1,800

2,000

Figure 3 Bias towards local processing in patient Q.M. a, Examples of
hierarchical Navon letters. b, Mean manual reaction times (in milliseconds) and
frequencies of correct responses (in 72 trials; shown as numbers next to bars) of
Q.M., of a control neglect patient (N þ C) and of age-matched normal controls
(C1–C4). In the direct attention task, Q.M. was slower (F-test P , 0:001) and less
accurate (x2 test, P , 0:01) in the global than in the local task. In the global task:
first, with incongruent stimuli Q.M.’s error frequency (in responding to local rather
than global features, that is, local-on-global interference) was higher than that of
the control subject (C2) with the worst performance (x2 test, P , 0:001); second,

Figure 4 Processing of local and global visual features in the neglected space.

with congruent stimuli Q.M.’s performance was at ceiling (x2 test with continuity

a, Examples of cancellation by Q.M. of local elements of symmetrical and

correction, P . 0:5); third, the perfect performance of Q.M. in the global task with

asymmetrical hierarchical shapes. b, Mean centre marking of hierarchical

local neutral features indicates that accuracy with congruent stimuli was probably

shapes. Open dot indicates true centre; filled dot indicates subjective centre.

not due to the processing of local, rather than global, features; and fourth, Q.M.’s

Taking the respective true centres as reference points, subjective centres were

reaction times were slower than the slowest control subject (N þ C) (F-test,

marked as follows: symmetrical shapes 2.8 mm rightward (t-test, P , 0:01),

P , 0:001). In the local task: first, Q.M.’s performance did not differ from that of

asymmetrical longer shapes 41.3 mm rightward (t-test, P , 0:001), asymmetrical

any control subject (x2 test, not significant in each comparison); and second,

shorter shapes 22.8 mm leftward (t-test, P , 0:001), asymmetrical shorter shapes

Q.M.’s reaction times were not statistically different from those of N þ C and C1,

22.8 mm leftward (t-test, P , 0:001). Taking the true centres of symmetrical

and were slower than those of C2, C3 and C4 (F-test, P , 0:001 in each case).

shapes as reference points, subjective centres were marked as follows: asym-

N þ C (who correctly discriminated targets in the cancellation tasks with 3s, 6s

metrical longer shapes 3.7 mm leftward, symmetrical shapes 2.8 mm rightward,

and animals) had a parietal–frontal lesion which spared the posterior areas that

asymmetrical shorter shapes 7.2 mm rightward (F omnibus test, P , 0:001).

were damaged in Q.M. c, Reaction times and frequencies of hits (and missed

c, Mean rightward bisection error of asymmetrical longer, symmetrical, and

responses) in the divided attention tasks for Q.M. The detection of global targets

asymmetrical shorter lines were, respectively, 39 mm, 40 mm and 37 mm (omni-

was worse (x2 test, P , 0:0001) and slower (F-test, P , 0:001) than the detection of

bus F test, not significant). The vertical dashed lines in a–c emphasize the relative

local ones.

position of items with respect to the patient’s trunk midsagittal plane.

NATURE | VOL 394 | 2 JULY 1998

Nature © Macmillan Publishers Ltd 1998

77

letters to nature
level. Q.M. exhibited a sort of transitory global simultanagnosia
when attention was reflexively captured by local details which were
relevant, but not sufficient, for task solution. Q.M.’s failure consciously to perceive the left half of the forest does not preclude her
from cutting down all the trees one by one, but does not prevent
her from failing to realize that she is cutting down the wrong
M
woods.

Glutamate mediates an
inhibitory postsynaptic
potential in dopamine neurons

Methods

C. D. Fiorillo & J. T. Williams

.........................................................................................................................

On cancellation tasks a1–a3 and b (Fig. 2), the items (2 3 2 cm) were arranged
in an array of 9 rows 3 10 columns. In tasks a2 and a3, items were tiled 458 and
908 clockwise and anticlockwise. In tasks c1–c3, items (4 3 4 cm) were
arranged in a 8 3 6 array. In tasks c4 and c5 the items were scattered. Each
task was performed twice for each target/distractor and item orientation
condition. A total of 60 targets and 60 distractors was presented in each
orientation condition of tasks a and b; 24 targets and 24 distractors were
presented in each condition of tasks c.
In the tachistoscopic Navon tests, stimuli (Fig. 3a) were large letters (global
features, 4 3 6 degrees of visual angle) composed of smaller ones (local
features, 0:8 3 1:2 degrees). In the divided attention task (Fig. 3c), stimuli
were obtained by combining, between local and global levels, letters H, S, A and
E; the same letter never appeared simultaneously at both levels. The patient had
to detect H and S at whatever level they appeared. In the global- and localdirected attention tasks (Fig. 3b), congruent items were letters H and S
composed of smaller letters H and S and incongruent ones were letters H
and S composed of smaller letters S and H, respectively. In the global task,
subjects discriminated between the large letter configurations (H or S); in the
local task, the small letters were discriminated between. In the global task with
local neutral features, stimuli were large letters H or S composed of small
rectangles. Items were presented for 150 ms, with their inner edge at 1.58 to the
right of central fixation.
The hierarchical stimuli of the recognition–cancellation task (Fig. 4a) were
geometrical shapes (180 3 180 mm) composed of small letters (5 3 5 mm).
Symmetrical stimuli were a circle (n ¼ 4), a square (n ¼ 4) and a diamond
(n ¼ 4). Asymmetrical stimuli (n ¼ 12) were obtained by combining the half
shapes of the symmetrical ones. In half of the trials, a target letter was alternated
with a non-target letter.
In the centre-marking task (Fig. 4b), the items were symmetrical Navon
squares (180 3 180 mm, n ¼ 12) and asymmetrical Navon rectangles with
longer (270 mm, n ¼ 12) or shorter (120 mm, n ¼ 12) horizontal leftward
extensions. The items of the bisection task (Fig. 4c) were symmetrical (180 mm,
n ¼ 12) and asymmetrical (270 mm, n ¼ 12; 120 mm, n ¼ 12) lines.
Received 16 December 1997; accepted 14 April 1998.
1. Marshall, J. C. & Halligan, P. W. Seeing the forest but only half the trees? Nature 379, 521–523 (1995).
2. Walker, R. Spatial and object based neglect. Neurocase 1, 371–383 (1995).
3. Snodgrass, J. G. & Vanderwart, M. A standardized set of 260 pictures: norms for name agreement,
image agreement, familiarity and visual complexity. J. Exp. Psychol. Hum. Learn. Mem. 6, 174–215
(1980).
4. Gainotti, G., Messerli, P. & Tissot, R. Qualitative analysis of unilateral spatial neglect in relation to
laterality of cerebral lesion. J. Neurol. Neurosurg. Psychiatry 35, 545–550 (1972).
5. Driver, J. & Halligan, P. W. Can visual neglect operate in object centred coordinates? An affirmative
single case study. Cogn. Neuropsychol. 8, 475–496 (1991).
6. Hillis, A. E. & Caramazza, A. A framework for interpreting distinct patterns of hemispatial neglect.
Neurocase 1, 189–207 (1995).
7. Navon, D. Forest before trees: the precedence of global features in visual perception. Cogn. Psychol. 9,
353–383 (1977).
8. Fink, G. R. et al. Hemispheric specialization for global and local processing: the effect of stimulus
category. Proc. R. Soc. Lond. B 264, 487–494 (1997).
9. Fink, G. R. et al. Where in the brain does visual attention select the forest and the trees? Nature 382,
626–628 (1996).
10. Robertson, L. C., Lamb, M. R. & Knight, R. T. Effects of lesions of temporal-parietal junction on
perceptual and attentional processing in humans. J. Neurosci. 8, 3757–3769 (1988).
11. Halligan, P. W. & Marshall, J. C. Focal and global attention modulate the expression of visuo-spatial
neglect: a case study. Neuropsychologia 32, 13–21 (1994).
12. Pizzamiglio, L. et al. Cognitive rehabilitation of the hemineglect disorder in chronic patients with
unilateral brain damage. J. Clin. Exp. Neuropsychol. 14, 901–923 (1992).
Acknowledgements. We thank Q.M. for her collaboration, L. Pizzamiglio, G. Vallar and G. Antonucci for
suggestions, C. Montagna, E. Catani and B. Fane for text revision, and G. Paganelli and his team for help in
the radiological examination of Q.M. This work was supported by a grant from the Centro Residenziale
IRCCS S. Lucia.
Correspondence and requests for materials should be addressed to F.D. at the IRCCS S. Lucia (e-mail:
fdoricchi@uniroma1.it).

78

Vollum Institute, Oregon Health Sciences University, 3181 South West Sam
Jackson Park Road, Portland, Oregon 97201, USA
.........................................................................................................................

Rapid information transfer within the brain depends on chemical
signalling between neurons that is mediated primarily by glutamate and GABA (g-aminobutyric acid), acting at ionotropic
receptors to cause excitatory or inhibitory postsynaptic potentials
(EPSPs or IPSPs), respectively. In addition, synaptically released
glutamate acts on metabotropic receptors to excite neurons on a
slower timescale through second-messenger cascades, including
phosphoinositide hydrolysis1. We now report a unique IPSP
mediated by the activation of metabotropic glutamate receptors.
In ventral midbrain dopamine neurons, activation of metabotropic glutamate receptors (mGluR1) mobilized calcium from
caffeine/ryanodine-sensitive stores and increased an apaminsensitive potassium conductance. The underlying potassium conductance and dependence on calcium stores set this IPSP apart
from the slow IPSPs described so far2–4. The mGluR-induced
hyperpolarization was dependent on brief exposure to agonist,
because prolonged application of exogenous agonist desensitized
the hyperpolarization and caused the more commonly reported
depolarization1,5,6. The rapid rise and brief duration of synaptically released glutamate in the extracellular space can therefore
mediate a rapid excitation through activation of ionotropic
receptors, followed by inhibition through the mGluR1 receptor.
Thus the idea that glutamate is solely an excitatory neurotransmitter must be replaced with a more complex view of its dual
function in synaptic transmission.
Intracellular recordings were made in dopamine neurons of the
ventral tegmental area (VTA) and substantia nigra. In the presence
of ionotropic receptor antagonists, a train of stimuli (10 at 66 Hz)
elicited a hyperpolarization that had two overlapping components
of similar amplitude. The early component was mediated by GABAB
receptors (Fig. 1a), as previously reported7. The late component was
insensitive to GABAB antagonists (Fig. 1a, b) and reversibly blocked
by tetrodotoxin (TTX, 300 nM; 91 6 2%, n ¼ 4), suggesting that it
might also be a slow IPSP.
The kinetics of this late hyperpolarization suggested that it was
mediated by a mechanism distinct from GABAB-mediated IPSP.
The mGluR1-subtype is present on dopamine cells8,9, is coupled to
phosphoinositide (PtdIns) hydrolysis1 and releases calcium from
intracellular stores in other cells1. Calcium could activate a calciumsensitive potassium conductance (rSK3) present on dopamine
cells that is sensitive to the bee venom toxin, apamin10,11. Apamin
irreversibly blocked the late hyperpolarization (by 86 6 3%, n ¼ 7,
Fig. 1a), but did not affect the GABAB IPSP. Thus the late
hyperpolarization was dependent on a rise in intracellular Ca2+
concentration and activation of a K+ current.
The mGluR antagonists (S)-MCPG (0.5–1.0 mM), (S)-4-CPG
(500 mM) and (RS)-AIDA (100–500 mM) reduced the amplitude of
the late hyperpolarization by 75 6 5% (n ¼ 6, Fig. 1b, c), 73 6 7%
(n ¼ 4), and 44 6 8% (n ¼ 6), respectively. This pharmacological
profile suggests that the late hyperpolarization is an IPSP mediated
by mGluR112–14 and the mGluR1d splice variant has been specifically
demonstrated in dopamine cells15. The inhibition caused was
selective for the IPSP because (S)-MCPG and (S)-4-CPG did not
affect the amplitude of the apamin-sensitive component of the
after-hyperpolarization (AHP) following the action potential ((S)-

Nature © Macmillan Publishers Ltd 1998

NATURE | VOL 394 | 2 JULY 1998

8