DOPAMINERGIC CONTROL OF WORKING MEMORY
AND ITS RELEVANCE TO SCHIZOPHRENIA: A CIRCUIT
aDepartment of Electrical and Electronics Engineering, Sophia Univer-
sity, 7-1 Kioicho, Chiyoda-ku, Tokyo 102-8554, Japan
bDepartment of Psychiatry, Columbia University Medical Center, New
York State Psychiatric Institute, 1051 Riverside Drive, Unit 31, New
York, NY 10032, USA
Abstract—This article argues how dopamine controls work-
ing memory and how the dysregulation of the dopaminergic
system is related to schizophrenia. In the dorsolateral pre-
frontal cortex, which is the principal part of the working
memory system, recurrent excitation is subtly balanced with
intracortical inhibition. A potent controller of the dorsolateral
prefrontal cortical circuit is the mesocortical dopaminergic
system. To understand the characteristics of the dopaminer-
gic control of working memory, the stability of the circuit
dynamics under the influence of dopamine has been studied.
Recent computational studies suggest that the hyperdopa-
minergic state is usually stable but the hypodopaminergic
state tends to be unstable. The stability also depends on the
efficacy of the glutamatergic transmission in the corticomes-
encephalic projections to dopamine neurons. When this cor-
tical feedback is hypoglutamatergic, the circuit of the dorso-
lateral prefrontal cortex tends to be unstable, such that a
slight increase in dopamine releasability causes a cata-
strophic jump of the dorsolateral prefrontal cortex activity
from a low to a high level. This may account for the seemingly
paradoxical overactivation of the dorsolateral prefrontal cor-
tex observed in schizophrenic patients. Given that dopamine
transmission is abnormal in the brains of patients with
schizophrenia and working memory deficit is a core dysfunc-
tion in schizophrenia, the concept of circuit stability would be
useful not only for understanding the mechanisms of work-
ing memory processing but for developing therapeutic strat-
egies to enhance cognitive functions in schizophrenia.
© 2005 IBRO. Published by Elsevier Ltd. All rights reserved.
Key words: cognitive, computational, dopamine, glutamate,
mesocortical, prefrontal cortex.
Working memory is not a unitary system but consists of
subsystems (Baddeley, 1986, 2003). Numerous functional
imaging studies have suggested that performing working
memory tasks activates several different areas of the cor-
tex, including the dorsolateral prefrontal cortex (DLPFC),
the anterior cingulate area, and the posterior parietal cor-
tex (PPC) (for example: Fletcher and Henson, 2001;
Jonides et al., 1998; Leung et al., 2002; Sakai et al., 2002).
This network may include the attentional control system
(Awh and Jonides, 2001; Awh et al., 2000; Chafee and
Goldman-Rakic, 1998, 2000; de Fockert et al., 2001;
Courtney et al., 1998; Curtis and D’Esposito, 2003; Kast-
ner and Ungerleider, 2000; Stedron et al., 2005; Tanaka,
2003). Some of the areas that are activated by the perfor-
mance of working memory tasks are, however, different for
different tasks, suggesting that they are primarily related to
task-specific processing or modalities the tasks employ. In
contrast, the DLPFC is commonly activated by all kinds of
working memory tasks, and is therefore considered to be
involved in a working memory process that is common
across tasks, such as maintenance of working memory.
However, further specification of working memory pro-
cesses is not straightforward. So far, neuroimaging studies
have suggested that the DLPFC plays a central role in the
maintenance and manipulation of working memory (Curtis
and D’Esposito, 2003; D’Esposito et al., 1999, 2000;
Leung et al., 2002; Owen et al., 1996; Rowe et al., 2000;
Smith and Jonides, 1999). Executive functions would re-
quire cooperation of subprocesses (Funahashi, 2001; Per-
ner and Lang, 1999) and would be mediated by a large
cortical–subcortical network (Elliott, 2003; Royall et al.,
2002). Still, it would be the case that the DLPFC plays a
central role also in those functions (for reviews: Carpenter
et al., 2000; Funahashi, 2001; Fuster, 2000; Robbins,
1996; Smith and Jonides, 1999; Tanji and Hoshi, 2001).
Since the pioneering works (Fuster and Alexander,
1971; Kubota and Niki, 1971), neuroanatomical, neuro-
physiological and neurocomputational studies have been
elucidating the machinery of the DLPFC for working mem-
ory (for reviews: Castner et al., 2004; Constantinidis and
Wang, 2004; Durstewitz et al., 2000b; Funahashi, 2001;
Fuster, 2000, 2001; Goldman-Rakic, 1987, 1995, 1996;
Goldman-Rakic et al., 2000; Tanaka, 2001; Ungerleider,
1998; Wang, 2001). With anatomical basis (Goldman-Ra-
kic, 1987; Gonzalez-Burgos et al., 2000; Melchitzky and
Lewis, 2003; Melchitzky et al., 1998, 2001), it has been
suggested that recurrent excitation or mutual excitation
between pyramidal neurons having similar preferences of
memory fields plays a primary role in the maintenance of
working memory. However, computational model networks
with recurrent excitation easily go runaway, showing epi-
leptiform activity of all the neurons in the network without
coding any meaningful information. A number of computa-
tional studies have suggested that stable activity of neu-
rons in a cortical circuit is attained on a subtle balance
Correspondence to: S. Tanaka, Department of Electrical and Electron-
ics Engineering, Sophia University, 7-1 Kioicho, Chiyoda-ku, Tokyo
102-8554, Japan. Tel: ?81-3-3238-3331; fax: ?81-3-3238-3321.
E-mail address: email@example.com (S. Tanaka).
Abbreviations: DA, dopamine; DLPFC, dorsolateral prefrontal cortex;
FS, fast-spiking; HVA, high-voltage-activated; NFS, non-fast-spiking;
PFC, prefrontal cortex.
Neuroscience 139 (2006) 153–171
0306-4522/06$30.00?0.00 © 2005 IBRO. Published by Elsevier Ltd. All rights reserved.
between the recurrent excitation and intracortical inhibition
(Amit and Brunel, 1997; Amit and Mongillo, 2003; Brunel,
2003; Brunel and Wang, 2001; Compte et al., 2000; Durst-
Nirenberg, 2004; Tanaka 1999, 2001; Wang, 1999, 2001).
The intracortical inhibition would have different roles in work-
ing memory, being consistent with the existence of a wide
variety of subtypes of inhibitory interneurons in the cortex
(Cauli et al., 1997; Gupta et al., 2000; Kawaguchi, 1993,
1995; Kawaguchi and Kubota, 1993, 1996, 1997; Krimer
and Goldman-Rakic, 2001; Miles, 2000; Somogyi et al.,
1998; Tamas et al., 1998). In the visuospatial working
memory, roles of two types of the intracortical inhibition,
the iso- and cross-directional inhibition, have been studied
(Rao et al., 1999, 2000; Tanaka, 1999, 2000, 2001,
2002a). Their studies suggested that a primary role of the
isodirectional inhibition is to restrict or tune memory fields,
while the cross-directional inhibition contributes to the sup-
pression of background activity (Rao et al., 1999, 2000;
Tanaka, 1999, 2000, 2001). This notion becomes clearer
when neurons represent spatial working memory with mul-
tiple targets (Tanaka, 2002a). In this case, the cross-direc-
tional inhibition would play an important role in controlling
the competition between targets by mediating mutual inhi-
bition. Thus established scarcely stable representation of
working memory is subject to modulation by neuromodu-
Catecholaminergic modulation of working memory ac-
tivity has been studied neurophysiologically using the de-
layed-response task paradigm (Arnsten, 1998; Sawaguchi,
1991; Sawaguchi and Goldman-Rakic 1991, 1994; Sawagu-
chi et al., 1990a,b, 1994; Williams and Goldman-Rakic,
1995). Local injection of selective dopamine (DA) D1 recep-
tor antagonists into the DLPFC in monkeys impaired spa-
tial working memory but had no effect on performance in
the control task requiring visually guided saccades, indi-
cating that sensory and motor functions were unaltered
(Sawaguchi and Goldman-Rakic, 1991, 1994). In contrast,
stimulation of D1 receptors in aged monkeys enhanced
spatial working memory (Castner and Goldman-Rakic,
2004). This suggests that D1 receptors play a selective
role in spatial working memory of the primate. In accor-
dance with this, computational studies have illustrated
how dopaminergic modulation by D1 receptor activation
changes the circuit dynamics for working memory of their
models (Brunel and Wang, 2001; Dreher and Burnod,
2002; Dreher et al., 2002; Durstewitz et al., 1999, 2000a;
Tanaka, 2002a,b, 2005; Yamashita and Tanaka, 2002,
2003, 2005). Here, circuit dynamics are dynamic circuit
properties that emerge synergistically from interacting neu-
ronal dynamics. In contrast, the dopaminergic modulation
that was incorporated into their models is at an individual
neuron level. These studies, therefore, filled a gap be-
tween the circuit dynamics and the neuronal dynamics.
The first part of this article argues that D1 receptor activa-
tion changes the circuit dynamics so that it can alter the
representation of working memory. We propose that, by
controlling D1 receptor activation, the system can differ-
ently represent spatial working memory with multiple target
locations. The closed-loop, prefronto–mesoprefrontal sys-
tem would control DA release in the DLPFC. In this sys-
tem, while the well-known inverted-U shape characteristic
of dopaminergic modulation comes from a circuit property
of the DLPFC, the actual operating point on the inverted-U
curve for working memory processing is determined by the
activity of the cortico-mesocortical system. Both the DA
releasability in the DLPFC and the efficacy of the corti-
comesencephalic glutamatergic transmission would play
critical roles in controlling the activity of the cortico-meso-
cortical system. We argue, secondly, how changes in
these parameters affect the circuit dynamics for working
Working memory research benefits from schizophrenia
research. In fact, the impairment of working memory and
other cognitive functions is one of the major symptoms of
schizophrenia (Fleming et al., 1995, 1997; Goldman-Ra-
kic, 1994; Gooding and Tallent, 2004; Park and Holzman,
1992, 1993), and the study of working memory in the
context of schizophrenia could provide insights into its
mechanisms. In the history of the research and treatment
of schizophrenia, tremendous effort has been devoted to
understanding the relationship between the symptoms of
schizophrenia and neurotransmitters/neuromodulators, such
as DA, serotonin, and glutamate. This is reasonable because
typical antipsychotics, such as haloperidol, are DA D2 recep-
tor antagonists, and many atypical antipsychotics are highly
selective for DA and serotonin receptors. The brains of
schizophrenic patients are in a hypodopaminergic state in
the cortex, especially in the prefrontal cortex (PFC), and a
hyperdopaminergic state in the striatum (Abi-Dargham,
2004; Abi-Dargham and Moore, 2003; Abi-Dargham et al.,
1998, 2000, 2002; Breier et al., 1997; Davis et al., 1991;
Frankle et al., 2003; Goldman-Rakic et al., 2004; Kahn and
Davis, 2000; Kegeles et al., 2000; Koh et al., 2003; Laru-
elle, 2003; Laruelle and Abi-Dargham, 2003; Laruelle et
al., 1999, 2003; Manoach, 2003; Seeman and Kapur,
2000; Suhara et al., 2002). N-methyl-D-aspartate (NMDA)
receptor antagonists, such as phencyclidine and ketamine,
induce schizophrenia-like symptoms in healthy subjects
(Balla et al., 2001, 2003; Bunney et al., 2000; Halberstadt,
1995; Jentsch and Roth, 1999; Jentsch et al., 1997, 1999,
2000; Krystal et al., 2003; Le Pen et al., 2003; Noda et al.,
2000; Olney and Farber, 1995a,b; Tsai and Coyle, 2002).
These facts suggest that abnormal transmission of DA,
serotonin, and glutamate is critically related to the symp-
toms of schizophrenia.
The brains of patients with schizophrenia exhibit incon-
sistent lower or higher activation of the DLPFC when they
perform working memory tasks (Andreasen et al., 1992;
Callicott et al., 2000, 2003a,b; Carter et al., 1998; Mano-
ach, 2003; Meyer-Lindenberg et al., 2001, 2002; Paulman
et al., 1990; Weinberger and Berman, 1998), and this
activity is associated with the impairment of cognitive func-
tions including working memory, which are commonly seen
in schizophrenia (Elvevag and Goldberg, 2000; Goldberg
and Green, 2000, 2002; Goldberg et al., 2003; Gooding and
Tallent, 2004; Kuperberg and Heckers, 2000; Lewis, 2004;
Weinberger and Gallhofer, 1997; Winterer and Weinberger,
S. Tanaka / Neuroscience 139 (2006) 153–171 154
2004). However, the mechanisms of such inconsistent ac-
tivation of the DLPFC in schizophrenia are largely un-
known. A theoretical analysis with the basic prefronto–
mesoprefrontal circuit model recently attempted to account
for this kind of activation of the DLPFC in schizophrenia
(Tanaka et al., 2005). It suggests that, when the efficacy of
the glutamatergic transmission in the corticomesence-
phalic projections to midbrain DA neurons is low, the circuit
of the DLPFC tends to be unstable, such that a slight
increase in DA releasability causes a catastrophic jump of
the DLPFC activity from a low to a high level. The third and
last part of this article will argue how this may be related to
the inconsistent activation of the DLPFC in the brains of
the patients with schizophrenia.
OF WORKING MEMORY
Dopaminergic modulatory systems
There are four dopaminergic pathways existing in the
brain, the mesocortical system, the mesolimbic system,
the nigrostriatal system, and the tuberoinfundibular sys-
tem, each of which has activity with a unique set of phys-
iological and psychological effects (Stahl, 2000). In the
following, we focus on the issues of working memory rep-
resentation and its modulation by DA. The system we
consider is, therefore, the DLPFC circuit for working mem-
ory representation and the mesocortical dopaminergic sys-
tem that is supposed to modulate working memory activity
in the DLPFC.
Dopaminergic modulation in the circuit
Since the pioneering work of Brozoski et al. (1979), the
critical involvement of DA in cognitive functions has been
suggested by many investigations. It is, however, only
recently that the circuit mechanisms by which DA controls
or modulates cognitive functions have been coming to light
through neuroanatomical and neurophysiological findings
(Gao and Goldman-Rakic, 2003; Gao et al., 2001, 2003;
Gonzalez-Islas and Hablitz, 2001, 2003; Le Moal, 2000;
Seamans and Yang, 2004; Seamans et al., 2001b; Yang et
al., 1999). These studies have disclosed several aspects
of complicated actions of DA in the circuit. Some of them
have suggested that dopaminergic modulation in the cor-
tical circuit is not homogeneous but rather circuit specific.
For example, Goldman-Rakic and coworkers (Gao and
Goldman-Rakic, 2003; Gao et al., 2001, 2003) have found
that DA depressed inhibitory transmission from fast-spik-
ing (FS) interneurons to pyramidal neurons but enhanced
inhibitory transmission from non-fast-spiking (NFS) inter-
neurons to pyramidal neurons. Because FS neurons form
synaptic contacts on perisomatic domains while NFS neu-
rons synapse on peridendritic domains, this implicates
circuit-dependent dopaminergic modulation of intracortical
inhibition (Gao et al., 2001, 2003). This circuit perspective
would be particularly important for the understanding of
dopaminergic mechanisms of working memory and other
Modulation of working memory activity via D1
Electrophysiological recordings from the brains of behav-
ing monkeys showed alterations of working memory per-
formance by the local application of D1 receptor antago-
nists in the DLPFC (Sawaguchi, 2001; Sawaguchi and
Goldman-Rakic, 1991, 1994). By combining iontophoretic
analysis of DA receptors with single-cell recording in mon-
key during working memory performance, Williams and
Goldman-Rakic (1995) suggested, for the first time, that
D1 receptors selectively modulate the memory fields of
DLPFC neurons. Goldman-Rakic and coworkers (Gold-
man-Rakic et al., 2000; Muly et al., 1998) proposed a
hypothesis that accounts for the cellular mechanisms of the
modulation of memory fields by D1 receptors. In this hypoth-
esis, D1 receptor activation causes a change in the balance
between the excitation and the inhibition in the DLPFC circuit,
which changes the firing rates of the DLPFC neurons during
the delay period. Their experimental result suggests that
stimulation of D1 receptors in the DLPFC modulates
working memory with an inverted-U shape profile (Desi-
mone, 1995; Goldman-Rakic et al., 2000; Williams and
Goldman-Rakic, 1995). Currently, this profile of dopami-
nergic modulation is widely known, but it actually has
dual interpretations. One is physiological, in which the
average firing rate during the delay period is modulated by
DA with this profile. The other is behavioral, in which
working memory performance also shows the inverted-U
shape modulation by DA (Arnsten, 1997; Cai and Arnsten,
1997; Honey and Bullmore, 2004; Mattay et al., 2000;
Zahrt et al., 1997). Both seem to have grounds, but they
are not identical. At least at present, there is no evidence
that the relationship between the activity level in the delay
period and working memory performance is monotonic.
The inverted-U shape modulation of the activity of
DLPFC neurons has been reproduced in computational
studies (Brunel and Wang, 2001; Tanaka, 2002b; Ya-
mashita and Tanaka, 2002, 2003, 2005). Brunel and Wang
(2001) used a leaky integrate-and-fire neuron model with
NMDA, AMPA, and GABAAchannels as well as the leak
channel. In their model, DA was assumed to change the
values of NMDA, AMPA, and GABAA conductances.
Among these parameters, the NMDA conductance is the
most critical for the generation of the inverted-U shape
characteristic. Both the pyramidal neurons and the inter-
neurons in their model have NMDA conductances whose
dependences on D1 receptor activation are given by sigmoi-
dal curves with the same height. The essential point of the
generation of the inverted-U shape characteristic is that the
sigmoidal curve for the interneurons is rightward shifted, so
that the NMDA conductance of the interneurons increases
even in the high D1 receptor activation region. Conse-
quently, the GABAergic inhibition becomes relatively stron-
ger when D1 receptors are highly activated. Tanaka and
Yamashita (Tanaka, 2002b; Yamashita and Tanaka, 2003)
also showed computationally the inverted-U shape char-
acteristic of dopaminergic modulation. They used a leaky
integrate-and-fire neuron model with NMDA, AMPA,
S. Tanaka / Neuroscience 139 (2006) 153–171155
GABAA, persistent sodium, calcium-dependent potassium,
and the leak channels. Their models assumed that D1
receptor stimulation increases NMDA conductance and
persistent sodium currents (by leftward shifting the current-
voltage curve) and decreases AMPA and calcium-depen-
dent potassium conductances, in accordance with experi-
mental results (Chen et al., 2004; Gao et al., 2001; Gore-
lova and Yang, 2000: Gorelova et al., 2002; Pedarzani and
Storm, 1995; Seamans et al., 2001a,b; Zhou and Hablitz,
1999) and the previous computational model of dopami-
nergic modulation (Durstewitz et al., 2000a). By taking into
account the hypothesis that D1 receptor activation effec-
tively increases glutamatergic input acting on NMDA re-
ceptors of pyramidal neurons when the extracellular DA
concentration is low but further activation of D1 receptors
increases that of interneurons (Muly et al., 1998), Tanaka
and Yamashita (Tanaka, 2002b; Yamashita and Tanaka,
2003) assumed a greater increase of the NMDA conduc-
tance of the interneurons in the hyperdopaminergic region
compared with that of the pyramidal neurons. The studies
of both groups thus suggest that the decrease of the
activity of DLPFC neurons in the hyperdopaminergic re-
gion of the inverted-U shape profile is due to stronger
GABAergic inhibition over the circuit while the increase of
the activity in the hypodopaminergic region is due to an
increase in glutamatergic transmission via NMDA recep-
tors. The maximum activation of DLPFC neurons is in
D1 receptor stimulation has another important charac-
teristic: it increases the robustness of working memory
representation (Durstewitz and Seamans, 2002; Durst-
ewitz et al., 1999, 2000a). The robustness was examined
by adding a distractor to the circuit in the delay period to
observe if the target pattern activity was disrupted. Robust-
ness against distractors increases with D1 receptor activa-
tion. The neuron model used by Durstewitz et al. (2000a) is a
Hodgkin-Huxley type with sodium, persistent sodium, high-
voltage-activated (HVA) calcium, delayed rectifier potassium,
slowly inactivating potassium, fast BK calcium- and voltage-
dependent C-type potassium, NMDA, AMPA, and GABAA
channels on the pyramidal neurons and sodium and delayed
rectifier potassium, NMDA, AMPA, and GABAAchannels on
the GABAergic interneurons. This model assumed that D1
receptor activation modulates persistent sodium, slowly
inactivating potassium, HVA calcium, NMDA, AMPA, and
GABAAcurrents. The simulation showed that the robust-
ness depended more strongly on NMDA conductance than
AMPA and GABAAconductances.
Operational control hypothesis
The role of DA in working memory might not be restricted
to merely increasing the robustness of the working mem-
ory representation of a single target location. Computer
simulation of spatial working memory with multiple targets
further suggests that DA can have an active role in working
memory processing (Tanaka 2002a,b). These simulations
used a leaky integrate-and-fire neuron model for DLPFC
neurons; AMPA, NMDA, calcium-dependent potassium
channel, persistent sodium channel, and leak conductances
were assumed to depend on D1 receptor activation. The task
in the simulation was based on the oculomotor delayed-
response task (Funahashi et al., 1989) but multiple target
locations were cued simultaneously or sequentially. When
simultaneously cued, increasing the NMDA conductance
tends to allow acceptance of increasing numbers of targets
to be represented as working memory (Tanaka, 2002a).
With two sequentially cued targets of spatial working mem-
ory, the simulation further showed that changes in D1 re-
ceptor activation in the DLPFC could change the storage
strategy of the DLPFC (Tanaka, 2002b). When the D1 re-
ceptor activation is low, the circuit replaces the first target
by the second target (Fig. 1A). With an intermediate level
of D1 receptor activation, the circuit accepts both targets.
That is, it stores two targets after receiving the second cue
(Fig. 1B). When D1 receptor activation is high, the circuit
rejects the second target, maintaining only the first target
as working memory even after receiving the cue for the
second target (Fig. 1C). Note that the NMDA/AMPA con-
ductance ratio is increasing from Fig. 1A to Fig. 1C. In this
situation, NMDA receptors play a critical role, again. As
noted in the last subsection, D1 receptor stimulation in-
creases the robustness of working memory representation
primarily by increasing the NMDA conductance. The ro-
bustness of the representation of the first target thus in-
creases with D1 receptor activation in this simulation.
Therefore, the representation of the first target with lowest
robustness is easily replaced by the second target (Fig.
1A). In contrast, the representation of the first target with
highest robustness persists even after receiving the sec-
ond cue (Fig. 1C). The response to the second cue is only
transient because the cross-directional inhibition from the
robust first target representation is so strong that it sup-
presses the activity to represent the second target. In the
intermediate case (Fig. 1B), the representation of the first
target is not replaced but coexists with the second target.
In this case, the robustness of the representation of the
first target is strong enough to maintain the representation
but still weak so that the cross-directional inhibition does
not eliminate the second target. The circuit thus responds
differently to the same second cue by changing D1 recep-
tor activation. Most cognitive operations are obviously
more complex than these. The simulated operations
shown in Fig. 1 are, as they were, fundamental cognitive
operations. Still, this model suggests an interesting feature
of dopaminergic modulation of cognitive functions of the
brain: DA can control fundamental cognitive operations by
changing the D1 receptor activation (the “operational con-
The gating hypothesis of DA proposes that a phasic
dopaminergic signal controls the gating of afferent infor-
mation into the DLPFC to allow it, when necessary, to
establish a new representation in the DLPFC (Braver and
Cohen, 2000; Cohen et al., 2002; Montague et al., 2004;
O’Reilly et al., 1999). The operational control hypothesis
proposes a similar but different function of DA. It would,
therefore, be interesting to see the differences between the
gating hypothesis of DA and the operational control hy-
pothesis. First of all, the former is a procedure for updating
S. Tanaka / Neuroscience 139 (2006) 153–171156
the representation. On the other hand, the essential fea-
ture of the latter is that DA can alter the circuit dynamics
qualitatively so that it represents working memory differ-
ently depending on how much the D1 receptors are acti-
vated. This is a kind of selective processes for working
memory loading. Second, the gating hypothesis requires
an external input to gate (Moody et al., 1998), while the
operational control hypothesis does not. The gating func-
tion may be mediated by a thalamocortical circuit or, more
widely, a prefronto–striato–pallido–thalamo–prefrontal net-
work (Kalivas et al., 2001; Tanibuchi and Goldman-Rakic,
2003, 2005; Watanabe and Funahashi, 2004a,b), as sug-
gested by connectionist model studies (Braver and Cohen,
2000; Braver et al., 1999; Cohen et al., 2002; Frank et al.,
2001). The phasic activity of DA neurons in the ventral
tegmental area and substantia nigra (Hollerman et al.,
1998; Schultz, 1998; Waelti et al., 2001) may be related to
the gating function (Braver and Cohen, 2000; Cohen et al.,
2002; Montague et al., 2004; O’Reilly et al., 1999). In
contrast, the DA function in the operational control hypoth-
esis is based on the tonic action of DA on DLPFC neurons.
Midbrain DA neurons exhibit both phasic and tonic modes
of activity, and these modes could contribute to different
aspects of working memory. Therefore, the operational
control hypothesis proposes a new function of DA rather
than an alternative to the gating function. That DA has a
role in working memory that depends on the tonic DA
action is supported by recent microdialysis monitoring of
the DA release in the medial PFC of rats performing a
delayed-response task, suggesting that the accuracy of
performance depends on the magnitude of the DA release
(Phillips et al., 2004). Interestingly, the DA release for
accurate performance was not dependent on the presence
of reward. Moreover, the new function would not be merely
to increase robustness of working memory representation
because, if this were the case, it would be incompatible
with a decrease in DA efflux during the delay period.
Rather, it is more likely to be related to an operation of
working memory during the test phase as well as the
training phase. Further argument on this issue requires
additional experimental and theoretical studies.
Contribution of D2 receptors
Some functional imaging studies have suggested the in-
volvement of D2 receptors in working memory (Kimberg et
al., 2001; Luciana et al., 1992; Luciana and Collins, 1997;
Mehta et al., 2001, 2004), but other studies have not
replicated these results (Kimberg et al., 1997, 2001; Muller
et al., 1998). From these imaging studies, however, it is still
uncertain which aspect of working memory processing
depends on D2 receptors. Recently, how D1 and D2 ago-
nism differentially influences executive functions has been
assessed at the behavioral level (Roesch-Ely et al., 2005).
They used pergolide (a mixed D1/D2 agonist) and bro-
mocriptine (a selective D2 agonist) for healthy humans
performing Stroop tasks and delayed response tasks. The
results were that only bromocriptine showed decreased
interference in the Stroop tasks and that neither agonist
had any effect on the delayed response tasks. They ar-
gued that D2 agonism facilitated the switching from read-
ing to color naming. This interpretation is consistent with
Fig. 1. Three different operations of spatial working memory; i.e. replacement (A), addition (B), and rejection (C). The circuit sequentially received two
transient inputs cueing different target locations. The first cue was given during 300–500 ms and the second cue was given during 1300–1500 ms.
The representation of the first target in the first delay period is the same across A, B, and C. However, the second target is represented differently
in the second delay period in A, B, and C. The NMDA/AMPA ratios are 0.032 (A), 0.047 (B), and 0.054 (C).
S. Tanaka / Neuroscience 139 (2006) 153–171157
the suggestion made by Seamans and Yang (2004) that
the predominance of D2 modulation provides a net reduc-
tion in inhibition. McDowell et al. (1998) suggested, in
functional imaging, that bromocriptine improved executive
functions but had no effect on the maintenance of working
memory without significant executive demands. Taken to-
gether, it seems that the role of D2 receptors is more
relevant to executive functions than to the maintenance of
In the cortex, DA D2 receptors are relatively scarce
(Lidow and Goldman-Rakic 1994; Lidow et al., 1989), yet
they have a role in working memory processing. Recently,
by the iontophoretic application of a D2 agonist and an
antagonist, Wang et al. (2004a) suggested that D2 recep-
tors contribute only to response-related activity and have
little effect on delay-period activity of spatial working
memory in the DLPFC of monkeys. Their interpretation
of this result is that the response-related activity is corol-
lary discharge, informing the DLPFC that a motor com-
mand has been completed, which is modulated by D2
receptors. This idea has led us to construct a model to
reproduce their result (Ebi et al., 2005). The model of the
cortical circuit has two populations (Population D and Pop-
ulation S); the excitatory and inhibitory neurons in Popu-
lation D have D1 receptors, while those in Population S
have D2 receptors (Fig. 2). The neurons in Population D
receive cue inputs and maintain the information as working
memory. The neurons in Population S receive response-
related signal as an efference copy (Wang et al., 2004a).
There would be overlapping of both the neuronal distribu-
tion across the populations and the distributions of D1 and
D2 receptors over the populations in the cortex. This di-
chotomy in the model is just for clarifying the roles of D1
and D2 receptors. Note, however, that this is consistent
with their conclusion that “neurons with saccade-related
activity either have a high density of postsynaptic D2
receptors or receive an external input from neurons that
do, whereas cells expressing memory-related persistent
activation are preferentially modulated by D1 receptors”
(Wang et al., 2004a). One of the interesting features in this
study is that, in spite of generally inhibitory actions of D2
receptors on neuronal activity and GABAergic actions, the
synergistic result of D2 receptor activation is the facilitation
of the transient response-period activity (Fig. 2). This is a
kind of disinhibition due to D2 receptor stimulation. The
neurons of Population S do not exhibit the cue-related or
delay-period activity, as shown in the figure. In contrast,
the neurons of Population D exhibit delay-period activity.
However, D2 receptor stimulation does not affect the de-
lay-period activity, in accordance with their result (Wang et
al., 2004a). Moreover, though the neurophysiological data
in Fig. 2 show that D2 receptor activation also seems to
affect the cue-period activity to some extent, the activities
of the neurons they examined that exhibited cue-related
activity were not affected by the D2 receptor agonist or
antagonist (Wang et al., 2004a). These results suggest
that D1 and D2 receptors have different roles in working
memory processing in the DLPFC.
Further dissociation of the roles of D1 and D2 recep-
tors in working memory along with anatomical studies
would provide knowledge about the circuit-specific mech-
anisms of working memory processing. This would enable
receptor type- and circuit-specific models of working mem-
ory. In psychiatry, intensive efforts to develop cognitive
enhancers are currently being made (for example: Barch,
2004; Farah et al., 2004). Such models may be useful to
develop means for process-specific enhancement of cog-
STABILITY OF THE
Open-loop and closed-loop controls of DA release in
The operational control hypothesis, argued in the last sec-
tion, requires the regulation of the extracellular DA level in
the DLPFC. Midbrain DA neurons can release DA phasi-
cally and tonically under various regulating mechanisms
(Dreher and Burnod, 2002; Dreher et al., 2002; Fiorillo et
al., 2003; Grace, 1991, 1995, 2000; Schultz, 1998, 2001,
2002). Phasic activity of DA neurons, which is supposed to
be triggered by an input from outside the mesocortical
system, would be able to increase the extracellular DA
concentration in the DLPFC rapidly. The released DA is
removed by reuptake. In the mesocortical system, this
mode of DA control, the open-loop control, may play a role
in loading or updating of working memory. In the mainte-
nance of working memory, however, the reciprocal con-
nectivity between the DLPFC and the midbrain DA nuclei
(Sesack and Carr, 2002; Tzschentke, 2001; Williams and
Goldman-Rakic, 1998) would enable another mode of DA
control, the closed-loop control. In this mode of control, the
corticomesencephalic projections to the DA neurons would
modulate tonic activity of the DA neurons. Recent compu-
tational studies suggest that this closed-loop system forms
a “regulator” of DA release in the DLPFC (Tanaka, 2005;
Yamashita and Tanaka, 2003). This is in accord with the
multiple regulatory system view in the mesolimbic system:
“burst firing induces massive synaptic DA release, which is
rapidly removed by reuptake before escaping the synaptic
cleft, whereas increased population activity modulates
tonic extrasynaptic DA levels that are less influenced by
reuptake” (Floresco et al., 2003). It is, therefore, likely that
the brain uses different control modes for different pro-
cesses of working memory.
Stability analysis of the prefronto–mesoprefrontal
The dynamics of the prefronto–mesoprefrontal system
have been analyzed further. The basic architecture of the
model is shown in Fig. 3. The model consists of the DLPFC
and the midbrain DA nuclei. It has been anatomically and
physiologically confirmed in rodents that mesocortical DA
neurons send axons to the cortex and receive feedback
projections from the cortex (Sesack and Carr, 2002;
Sesack et al., 2003; Tzschentke, 2001). In primates, the
DA afferents originate from a widespread continuum of
S. Tanaka / Neuroscience 139 (2006) 153–171158
neurons in the substantia nigra, ventral tegmental area
(VTA), and the retrorubral area (Williams and Goldman-
Rakic, 1998). The present computational model deals with
the population of the mesocortical DA neurons as a unit.
The activity of this unit determines the DA release in the
DLPFC, which then modulates the activity of the DLPFC
neurons by D1 receptor activation. The activity of the
DLPFC feeds back to the DA unit. Beside this cortical
feedback, this unit receives phasic and tonic inputs, which
generate the phasic and tonic activities of the DA unit,
respectively. The phasic input triggers the DA release in
the DLPFC. In the simulation, it is usually strong enough to
increase the DA release in the DLPFC rapidly. In contrast,
the tonic input to the DA unit is utilized just for biasing the
activity of the DA unit and is rather weak or often vanishing.
In the cortex, the cue input triggers the sustained activity of
Fig. 2. Effects of D2 receptor activation on the saccade-related activity of DLPFC neurons. (A) The model has two populations of neurons, each of
which is related to delay-period activity (PD and ND) or saccade-related activity (PS and NS). P denotes pyramidal neurons and N denotes GABAergic
interneurons. (B) Time courses of the activity of PS for various levels of D2 receptor stimulation. The colored lines in the figure are the time courses
of the neuronal activity recorded from monkey DLPFC for low (green), intermediate (blue), and high (red) D2 receptor stimulation (Wang et al., 2004a).
The three-dimensional surface plot shows the result of the computer simulation of the model.
S. Tanaka / Neuroscience 139 (2006) 153–171159
the DLPFC neurons, which mimics working memory activ-
ity (Fig. 3).
The essential features of the dynamics of the prefronto–
mesoprefrontal system can be described mathematically
with the following set of state equations:
where the state variables, xp, xn, xd, and yd, denote the
activity of the pyramidal neurons in the DLPFC, the activity
of the GABAergic interneurons in the DLPFC, the activity
of the DA neurons in the midbrain, and the DA release in
the DLPFC, respectively, z?zmaxf ?yd?t?? is the level of D1
receptor activation and its effects on the signal transduc-
tion pathways (Duman and Nestler, 2000; Greengard,
2001) (we simply call this “D1 receptor activation” in the
following), zmaxis the sensitivity of the D1 receptors, and
The pyramidal neurons in the DLPFC receive a transient
cue input, Icue?t?, during 100 ms in the beginning of the
simulation, which triggers circuit dynamics. The DA neu-
rons receive phasic ?Iphasic?t?? and tonic ?Itonic?t?? inputs from
outside the circuit (Fig. 3). The phasic input is synchro-
nized with the cue input to the DLPFC. In this model, the
D1 receptor activation is assumed to change W˜
and ? ˜nas W˜
and ? ˜n??n?0.24z?0.26?. The values used in the simulation
are: ?p?20 ms, ?n?6.8 ms, Wpp?1.11, Wpn?3.84, Wnp
?0.27, Wnn?0.0 (for simplicity in this simulation), Wpd
From the above set of equations, one obtains the
equilibrium state of the system as a fixed point in the
state space. Fig. 4 shows the fixed points and the tra-
jectories showing the state transition to the fixed points.
The fixed points of this system in the f(xp)?z space (i.e.
the space of the firing rate of the pyramidal neurons of
the DLPFC and the D1 receptor activation) are obtained
by the intersections of the xp nullcline and the yd
nullcline. The fixed points indicated by small circles in
the figure are stable, while those by small triangles are
unstable. The origin is always a stable fixed point, and
another stable fixed point emerges in Fig. 4A, B, C, and
F. The gray curves show trajectories of the state transi-
tion. The figure shows the convergence of the trajecto-
ries to a stable fixed point. The locations of the fixed
points depend on two important circuit parameters,
which are the synaptic weights Wdyand Wpd. The syn-
aptic weight Wdydenotes the DA releasability from the
axon terminals of the mesocortical projections in the
DLPFC. The amount of DA released in the DLPFC
changes with this parameter even when the activity of
the DA neuron is unchanged. The synaptic weight Wpd
denotes the efficacy of glutamatergic transmission in the
Fig. 3. The architecture of the model of the prefronto–mesoprefrontal system. The DLPFC is composed of pyramidal neurons and GABAergic
interneurons. They are reciprocally connected and also have self-innervations. A transient cue input to the DLPFC triggers the dynamics of the circuit,
which mimics the loading of working memory. The pyramidal neurons in the DLPFC send glutamatergic axons to the midbrain DA neurons. These DA
neurons receive external phasic and tonic inputs as well. The afferent fibers of the DA neurons release DA in the DLPFC.
S. Tanaka / Neuroscience 139 (2006) 153–171160
corticomesencephalic projections. Though there could
be several mechanisms to change this efficacy, we re-
gard this as an independent parameter without specify-
ing the mechanism in this model. When the efficacy of
glutamatergic transmission is low and the DA releasabil-
ity is not high (Fig. 4D and E), the origin is the only stable
fixed point because there is no intersection of the xp
nullcline and the ydnullcline. When the system has two
stable fixed points (Fig. 4F), an unstable fixed point
exists between the two. From this positional relationship
of these fixed points, it can be stated that the hyperdo-
paminergic state is generally stable while the hypodo-
paminergic state tends to be unstable. These results
suggest a model of the activity of the DLPFC under
normal and abnormal conditions. In the remaining part of
this article, we will discuss how it is related to schizo-
RELEVANCE TO SCHIZOPHRENIA
Unstable activation of the PFC
Schizophrenia is a complex syndrome and the etiology is
not likely to be attributable to a single cause. Nevertheless,
there could be critical factors that account for some as-
pects of the symptoms. From a circuit dynamics perspec-
tive, the failure of the regulation of the prefronto–mesopre-
frontal dopaminergic system would be one of them. This is
consistent with the hypothesis that the PFC is hypodopam-
inergic in schizophrenia (Davis et al., 1991; Kahn and
Davis, 2000), which is supported by receptor imaging stud-
ies (Abi-Dargham et al., 2002; Guo et al., 2003; Okubo et
al., 1997). In functional imaging studies, on the other hand,
the DLPFC in schizophrenia has originally been reported
to be hypoactive when performing a working memory task
(Andreasen et al., 1992; Carter et al., 1998; Meyer-Linden-
berg et al., 2002; Paulman et al., 1990; Ramsey et al.,
2002; Weinberger and Berman, 1998). This task-related
hypofrontality has been repeatedly observed in many stud-
ies (Berman, 2002). Later, however, functional imaging
studies reported exaggerated activation of the DLPFC for
working memory performance compared with normal sub-
jects (Callicott et al., 2000, 2003a,b; Manoach, 2003;
Manoach et al., 1999, 2000; Weinberger et al., 2001). The
relationship between the activation of the DLPFC and the
performance of working memory tasks, therefore, seems
to be complicated (see also: Honey et al., 2002; Perlstein
et al., 2001, 2003).
Fig. 4. Portraits on the f(xp)?z plane of the model prefronto–mesoprefrontal system with high (A–C) and low (D–F) efficacy of corticomesencephalic
glutamatergic transmission. Here, f(xp) denotes the activity of the pyramidal neurons and z is the D1 receptor activation level. Multiple fixed points are
obtained as the intersections between xpand ydnullclines. Among these intersections, circles and triangles represent stable and unstable fixed points,
respectively. The unstable fixed point is the hyperbolic fixed point or the saddle point of the system. The dynamics of this model near the fixed points
are determined by linearizing the model equations around their fixed points and estimating the associated eigenvalues. An increase in the DA
releasability influences the location of the fixed points except for the origin. The DA releasability and the glutamatergic transmission efficacy are
indicated by the % values relative to 0.30 and 0.023, respectively.
S. Tanaka / Neuroscience 139 (2006) 153–171 161
In the following, we apply the model from the previous
section to see how the activity of the DLPFC changes with
the DA releasability and the efficacy of the glutamatergic
transmission in the corticomesencephalic projections. Fig.
5 shows the dependences of the DLPFC activity and the
DA release or the D1 receptor activation (both are identical
in this simulation because the sensitivity of the D1 recep-
tors, zmaxin the model, is fixed) on the DA releasability in
high and low corticomesencephalic glutamatergic trans-
mission cases. There is a pair of a stable branch (the blue
curve in the figure) and an unstable branch (the red
dashed curve in the figure) as well as another stable
branch on the horizontal axis. In the hyperglutamatergic
condition (Fig. 5A and B), the stable fixed point (the light
blue dot) shows a relatively low firing rate of the DLPFC
neurons with a high D1 receptor activation. Further in-
crease in the DA releasability decreases the firing rate of
the pyramidal cells in the DLPFC but the D1 receptor
activation itself does not change very much. In the hypo-
glutamatergic condition (Fig. 5C and D), there is no stable
fixed point other than zero for both the activity of the
pyramidal cell and the D1 receptor activation when the DA
release is less than 180%. That is, low or moderate DA
releasability does not lead to activation of the DLPFC. A
slight increase in the DA releasability causes a cata-
strophic jump of the activity to a high level, and only the
states with high DA releasability provoke activation of the
DLPFC (Fig. 5C). Yet, the D1 receptor activation level
remains fairly low (Fig. 5D). When the DA releasability is at
300% in the hypoglutamatergic case, the DLPFC activity is
much higher than that in the hyperglutamatergic case,
while the D1 receptor activation level is lower than that in
the hyperglutamatergic case. The distinguishing feature in
this case is that the unstable region (indicated by a red
shadow in Fig. 5C) is enlarged. When the DLPFC is acti-
vated moderately, therefore, the activity goes to either a
high level or the zero level (the two light blue dots in the
figure). That is, the unstable dynamics of the circuit
makes the activation of the DLPFC highly susceptible to
The transition of the DLPFC state from a hypoactive
state to a hyperactive state is abrupt or catastrophic under
the hypoglutamatergic condition. This catastrophic transi-
tion occurs by increasing the releasability of DA in the
DLPFC. An increase in the DA releasability may be asso-
ciated with subjective effort to improve performance at
working memory tasks. Callicott et al. (2000) showed that
a subgroup of patients with schizophrenia who performed
an n-back task relatively normally had increased DLPFC
activation compared with normal controls, while patients
who performed poorly failed to activate the DLPFC. The
above computational results show both the exaggerated
activation and the inactivation of the DLPFC, which have
been observed in patients with schizophrenia (Callicott et
al., 2000, 2003a,b; Manoach, 2003; Manoach et al., 1999,
2000). Moreover, the above result suggesting that this
Fig. 5. Bifurcation diagram of the activity of the pyramidal cells in the DLPFC (A and C) and the D1 receptor activation (B and D). The efficacy of
corticomesencephalic glutamatergic transmission is fixed at 200% (A and B) or 30% (C and D) relative to 0.023 (100%). The blue solid line indicates
the stable branch of the bifurcation. The red dashed line indicates the unstable branch of the bifurcation. The vertical green dashed line shows a
catastrophic jump of the state. The horizontal axis is the DA releasability relative to 0.30 (100%).
S. Tanaka / Neuroscience 139 (2006) 153–171162
occurs when the corticomesencephalic glutamatergic
transmission is weak is compatible with the hypoglutama-
tergic hypothesis of schizophrenia (Aghajanian and
Marek, 2000; Carlsson et al., 2001, 2004; Goff and
Coyle, 2001; Halberstadt, 1995; Jentsch and Roth, 1999;
Kim et al., 1980; Krystal et al., 2003; Laruelle et al., 2003;
Meador-Woodruff and Healy, 2000; Meador-Woodruff and
Kleinman, 2002; Moghaddam, 2003; Olney and Farber,
1995; Tsai and Coyle, 2002). Taken together, this study
proposes that the activation of the DLPFC for working
memory performance under normal and abnormal condi-
tions (as in schizophrenia) can be accounted for, at least in
part, by the dynamics of the prefronto–mesoprefrontal sys-
tem, which is produced by the synergistic actions of DA,
glutamate, and GABA. This synergism, being restricted by
the circuitry of the system, changes with the circuit param-
eters. In this study, those parameters are the DA releas-
ability and the efficacy of the glutamatergic transmission in
the corticomesencephalic projections to the midbrain DA
neurons. Both have been key in the research of schizo-
The response to DA releasers, such as amphetamine and
cocaine, would also provide important information on the
intrinsic dynamics of the prefrontal-mesoprefrontal dopa-
minergic control system. There is a tremendous amount of
research on these psychostimulants both in humans and
animals (for reviews: Barch, 2004; Elliott and Beveridge,
2005; Fone and Nutt, 2005; Goldman-Rakic et al., 2004;
Pierce and Kalivas, 1997; Seeman and Madras, 1998; Sol-
anto, 1998; Steketee, 2003; White and Kalivas, 1998). A
computational trial to simulate the response to a DA releaser
was made recently using the prefronto–mesoprefrontal
closed-circuit model described in the last section (Tanaka,
2005). In this simulation, the effect of a DA releaser is
taken into account by shifting the activity-release curve of
DA. This study addressed how the neuronal activities of
both the DLPFC and the midbrain DA unit, the DA release
in the DLPFC, and the DA-glutamate interaction changed
when the sensitivity of the D1 receptors of DLPFC neurons
A DA releaser was applied into the DLPFC in the
simulation, which increased the DA release at the termi-
nals in the DLPFC without a direct change in the activity of
the DA neurons in the midbrain. The activity of the DA
neurons is modulated indirectly via the feedback input from
the DLPFC. Apparently, the DA release is not determined
only by the DA system. The change in the DA concentra-
tion modulates the glutamatergic as well as GABAergic
transmission in the DLPFC circuit. Fig. 6 depicts some
results of the simulation (Tanaka, 2005). With hyposensi-
tive or normal D1 receptors, the DA releaser increases the
DA release in the DLPFC (Fig. 6C and E). This does not
always occur. With hypersensitive D1 receptors, the DA
releaser does not increase the DA level in the DLPFC (Fig.
6G and I). In this case, the administration of the DA re-
leaser makes the DLPFC hyperdopaminergic, which sup-
presses the activity of the DLPFC neurons (Fig. 6H). Then,
the glutamatergic transmission decreases (Fig. 6J). This
cancels out the increase of DA release by the DA releaser,
maintaining the net DA release in the DLPFC unchanged
The DA-glutamate interaction is one of the key features
to an understanding of the circuit mechanisms for cognitive
functions. Given that the DA–glutamate interaction would
occur at many sites of the system (Del Arco and Mora,
2005; Morari et al., 1998; Sesack et al., 2003; Tseng and
O’Donnell, 2003, 2004; West et al., 2003), however, more
comprehensive studies would be necessary. This issue
would also be critical in the research of schizophrenia to
consolidate the DA hypothesis and the glutamate hy-
This article has proposed that the mesocortical dopami-
nergic system can control working memory. The funda-
mental assumption is closed-loop or feedback control of
dopaminergic modulation based on the closed-loop cir-
cuitry of the prefronto–mesoprefrontal system. In this con-
trol scheme, working memory activity of the DLPFC, which
is subject to dopaminergic modulation, can control the
activity of midbrain DA neurons. The computer simulation
with this model suggests that this system works as a
regulator of DA release. We argued that this control
scheme would be particularly suitable for the regulation of
neuronal activity during delay periods in working memory
tasks without explicit input. The closed-loop circuitry has
both anatomical and physiological grounds in rodents
(Sesack and Carr, 2002; Sesack et al., 2003; Tzschentke,
2001). The importance of the closed-loop circuitry, in this
context, lies in the intrinsic nature of the system as a
regulator of DA release in the DLPFC (Tanaka, 2005;
Yamashita and Tanaka, 2003). In this model, therefore,
dysregulation of the control of dopaminergic modulation is
largely responsible for working memory impairment.
Critical parameters that govern the dynamics of the
prefronto–mesoprefrontal system are the DA releasability
in the DLPFC and the efficacy of the corticomesencephalic
glutamatergic transmission from the DLPFC to the mid-
brain DA neurons. A system dynamics analysis of the
prefronto–mesoprefrontal system revealed that a hypoglu-
tamatergic condition enlarges an unstable region so that a
slight increase in the DA releasability in the DLPFC causes
a catastrophic jump from a hypofrontal state to a hyper-
frontal state. The activation level of this hyperfrontal state
is higher than the activation level with a normal glutama-
tergic condition. We argued that this might account for the
increased activation of the DLPFC during working memory
performance in some patients with schizophrenia. At
present, however, it is uncertain how this kind of activation
of the DLPFC is related to the impairment of working
memory. Patients with schizophrenia show an overall def-
icit of working memory and executive functions, which are
often greater than those of patients with frontal lobe lesions
(Pantelis et al., 1997). Robbins and coworkers (Pantelis et
al., 1997; Robbins, 1990) suggested that cognitive deficits
S. Tanaka / Neuroscience 139 (2006) 153–171163
Fig. 6. Dynamics of the model prefronto–mesoprefrontal system with hyposensitive and hypersensitive D1 receptors in the PFC. (A) The architecture
of the model. (B) The sensitivity curves of the effect of D1 receptor activation. (a) The hyposensitive case (zmax?1.0) (C–F). (b) The hypersensitive
case (zmax?2.0) (G–J). (C, G) Time courses of the DA neuron activity (blue), DA release (green), and D1 receptor effect (red). The DA releaser was
applied at t?1000 ms. (D, H) Population average firing rate of the pyramidal cells in the superficial layer of the PFC (dark blue) and glutamate release
from the pyramidal cells in the deep layer of the PFC (magenta). (E, I) Population average firing rate of the pyramidal cells in the superficial layer of
the PFC versus the DA release in the PFC. (F, J) DA release in the PFC versus the glutamate release from the pyramidal cells in the deep layer of
S. Tanaka / Neuroscience 139 (2006) 153–171164
in patients with schizophrenia would be caused more gen-
erally by the dysfunctioning of the frontostriatal system.
Moreover, abnormalities of the DLPFC in schizophrenia
may reflect a compromised neural strategy for handling
cognitive information by the DLPFC (Callicott et al., 2003).
These issues are to be studied further.
The basic idea underlying the arguments in this article
is compatible with the disconnection hypothesis proposed
for the mechanistic description of schizophrenic brains by
Friston (1998, 2005). This perspective is no doubt impor-
tant, given that cognitive information processing is pro-
duced by the integration of mutually connected regions.
However, this hypothesis does not specify how dynamics
change as a consequence of the alteration in connectivity.
On the other hand, many computational studies with bio-
physical modeling have addressed the circuit dynamics for
the maintenance of working memory (Amit and Brunel,
1997; Amit and Mongillo, 2003; Brunel, 2003; Compte et
al., 2000; Durstewitz and Seamans, 2002; Durstewitz et
al., 1999, 2000a,b; Fellous and Sejnowski, 2003; Tanaka,
1999, 2000, 2001, 2002a,b, 2005; Wang, 1999, 2001;
Wang et al., 2004b; Yamashita and Tanaka, 2002, 2003,
2005). So far, however, they have not addressed the dy-
namics of a larger network for either the maintenance or
other processes of working memory with a few exceptions
of preliminary studies (Miyashita et al., 2003; Tabuchi and
Tanaka, 2003; Tanaka, 2003). The circuit presented in this
article is a part of a large network that mediates working
memory processes and other cognitive functions. Yet, this
article has shown how the connectivity (i.e. the DA releas-
ability and the efficacy of the corticomesencephalic gluta-
matergic transmission in this case) alters the circuit dy-
namics. This approach should be extended to assess the
whole working memory system. In such a system, not only
DA but other neuromodulators including norepinephrine
and serotonin would also play significant roles. To eluci-
date the mechanisms of neuromodulatory effect in such a
huge and complex system, a strategic approach that inte-
grates the circuit dynamics perspective and the connectiv-
ity perspective would be necessary.
In summary, based on the simulation of circuit dynamics
being modulated by DA, we have first proposed the oper-
ational control hypothesis of working memory processing.
Second, the analysis of the prefronto–mesoprefrontal
closed-loop system, a regulator for the DA release in the
DLPFC, has illustrated how the operating point on the
inverted-U shape curve of dopaminergic modulation is de-
termined dynamically. It is interesting to note that there
exists a pair of stable and unstable fixed points on the
inverted-U curve and that the unstable fixed point is always
to the left of the stable fixed point. The distance between
these fixed points influences the dynamics. When they are
located closely to each other, the dynamics tend to be
vulnerable. This occurs when the cortical feedback is hypo-
glutamatergic. This concept has been applied to account for
inconsistent lower or higher activation of the DLPFC in the
brains of patients with schizophrenia. In the circuit dynamics
perspective, such kind of activation of the DLPFC in schizo-
phrenia can be attributable, at least partly but essentially, to
the dysregulation of the prefronto–mesoprefrontal dopami-
nergic control system. This should be tested by collabora-
tive works of functional imaging studies and computational
Acknowledgments—The author acknowledges Drs. Anissa Abi-
Dargham and Mark Slifstein at Columbia University for valuable
discussions and comments on this manuscript. He also acknowl-
edges Dr. Min Wang at Yale University for providing data used in
Fig. 2. Koki Yamashita and Hiroaki Ebi in my laboratory at Sophia
University are thanked for their technical assistance in preparing
the figures in the manuscript. This study was supported in part by
the Grant-in-Aid for Scientific Research (#15500218) from the
Ministry of Education, Culture, Sports, Science and Technology,
Japan. The author was on sabbatical leave from Sophia University
Abi-Dargham A (2004) Do we still believe in the dopamine hypothesis?
New data bring new evidence. Int J Neuropsychopharmacol 7
Abi-Dargham A, Gil R, Krystal J, Baldwin RM, Seibyl JP, Bowers M,
van Dyck CH, Charney DS, Innis RB, Laruelle M (1998) Increased
striatal dopamine transmission in schizophrenia: confirmation in a
second cohort. Am J Psychiatry 155:761–767.
Abi-Dargham A, Rodenhiser J, Printz D, Zea-Ponce Y, Gil R, Kegeles
LS, Weiss R, Cooper TB, Mann JJ, Heertum RLV, Gorman JM,
Laruelle M (2000) Increased baseline occupancy of D2 receptors
by dopamine in schizophrenia. Proc Natl Acad Sci USA 97:
Abi-Dargham A, Mawlawi O, Lombardo I, Gil R, Martinez D, Huang Y,
Hwang D-R, Keilp J, Kochan L, Heertum RV, Gorman JM, Laruelle
M (2002) Prefrontal dopamine D1 receptors and working memory
in schizophrenia. J Neurosci 22:3708–3719.
Abi-Dargham A, Moore H (2003) Prefrontal DA transmission at D1
receptors and the pathology of schizophrenia. Neuroscientist 9:
Aghajanian GK, Marek GJ (2000) Serotonin model of schizophrenia:
emerging role of glutamate mechanisms. Brain Res Rev 31:
Amit DJ, Brunel N (1997) Model of global spontaneous activity and
local structured activity during delay periods in the cerebral cortex.
Cereb Cortex 7:237–252.
Amit DJ, Mongillo G (2003) Selective delay activity in the cortex:
phenomena and interpretation. Cereb Cortex 13:1139–1150.
Andreasen NC, Rezai K, Alliger R, Swayze VW 2nd, Flaum M, Kirchner
P, Cohen G, O’Leary DS (1992) Hypofrontality in neuroleptic-naive
patients and in patients with chronic schizophrenia. Assessment with
xenon 133 single-photon emission computed tomography and the
Tower of London. Arch Gen Psychiatry 49(12):943–958.
Arnsten AFT (1997) Catecholamine regulation of the prefrontal cortex.
J Psychopharmacol 11:151–162.
Arnsten AFT (1998) Catecholamine modulation of prefrontal cortical
cognitive function. Trends Cogn Sci 2:436–447.
Awh E, Anllo-Vento L, Hillyard SA (2000) The role of spatial selective
attention in working memory for locations: evidence from event-
related potentials. J Cogn Neurosci 12:840–847.
Awh E, Jonides J (2001) Overlapping mechanisms of attention and
spatial working memory. Trends Cogn Sci 5:119–126.
Baddeley AD (1986) Working memory. Oxford, UK: Oxford University
Baddeley A (2003) Working memory: looking back and looking for-
ward. Nat Rev Neurosci 4(10):829–839.
S. Tanaka / Neuroscience 139 (2006) 153–171165
Balla A, Koneru R, Smiley J, Sershen H, Javitt DC (2001) Continuous
phencyclidine treatment induces schizophrenia-like hyperreactivity
of striatal dopamine release. Neuropsychopharmacology 25:
Balla A, Sershen H, Serra M, Koneru R, Javitt DC (2003) Subchronic
continuous phencyclidine administration potentiates amphet-
amine-induced frontal cortex dopamine release. Neuropsychop-
Barch DM (2004) Pharmacological manipulation of human working
memory. Psychopharmacology 174:126–135.
Berman KF (2002) Functional neuroimaging in schizophrenia. In:
Neuropsychopharmacology: The fifth generation of progress
(Davis KL, Charney D, Coyle JT, Nemeroff C, eds), pp 745–756.
Philadelphia: Lippincott Williams and Wilkins.
Braver TS, Barch DM, Cohen JD (1999) Cognition and control in
schizophrenia: a computational model of dopamine and prefrontal
function. Biol Psychiatry 46(3):312–328.
Braver TS, Cohen JD (2000) On the control of control: The role of
dopamine in regulating prefrontal function and working memory. In:
Attention and performance XVIII: Control of cognitive processes
(Monsell S, Driver J, eds), pp 713–737. Cambridge, MA: MIT
Breier A, Su T-P, Saunders R, Carson RE, Kolachana BS, Bartolomeis
Ad, Weinberger DR, Weisenfeld N, Malhotra AK, Eckelman WC,
Pickar D (1997) Schizophrenia is associated with elevated amphet-
amine-induced synaptic dopamine concentrations: Evidence from
a novel positron emission tomography method. Proc Natl Acad Sci
Brozoski TJ, Brown RM, Rosvold HE, Goldman PS (1979) Cognitive
deficit caused by regional depletion of dopamine in prefrontal
cortex of rhesus monkey. Science 205:929–932.
Brunel N (2003) Dynamics and plasticity of stimulus-selective persis-
tent activity in cortical network models. Cereb Cortex 13:1151–
Brunel N, Wang X-J (2001) Effects of neuromodulation in a cortical
network model of object working memory dominated by recurrent
inhibition. J Comput Neurosci 11:63–85.
Bunney BG, Bunney WE, Carlsson A (2000) Schizophrenia and
glutamate: an update. Psychopharmacology: The fourth genera-
tion of progress. http://www.acnp.org/content-32.html.
Cai JX, Arnsten AF (1997) Dose-dependent effects of the dopamine
D1 receptor agonists A77636 or SKF81297 on spatial working
memory in aged monkeys. J Pharmacol Exp Ther 283:183–189.
Callicott JH, Bertolino A, Mattay VS, Langheim FJP, Duyn J, Coppola
R, Goldberg TE, Weinberger DR (2000) Physiological dysfunction
of the dorsolateral prefrontal cortex in schizophrenia revisited.
Cereb Cortex 10:1078–1092.
Callicott JH, Egan MF, Mattay VS, Bertolino A, Bone AD, Verchinksi B,
Weinberger DR (2003a) Abnormal fMRI response of the dorsolat-
eral prefrontal cortex in cognitively intact siblings of patients with
schizophrenia. Am J Psychiatry 160:709–719.
Callicott JH, Mattay VS, Verchinski BA, Marenco S, Egan MF, Wein-
berger DR (2003b) Complexity of prefrontal cortical dysfunction in
schizophrenia: more than up or down. Am J Psychiatry 160:
Carlsson A, Waters N, Holm-Waters S, Tedroff J, Nilsson M, Carlsson
ML (2001) Interactions between monoamines, glutamate, and
GABA in schizophrenia: New evidence. Annu Rev Pharmacol Toxi-
Carlsson ML, Carlsson A, Nilsson M (2004) Schizophrenia: from do-
pamine to glutamate and back. Curr Med Chem 11:267–277.
Carpenter PA, Just MA, Reichle ED (2000) Working memory and
executive function: evidence from neuroimaging. Curr Opin Neu-
Carter CS, Perlstein P, Ganguli R, Brar J, Mintun M, Cohen JD (1998)
Functional hypofrontality and working memory dysfunction in
schizophrenia. Am J Psychiatry 155:1285–1287.
Castner SA, Goldman-Rakic PS (2004) Enhancement of working
memory in aged monkeys by a sensitizing regimen of dopamine D1
receptor stimulation. J Neurosci 24(6):1446–1450.
Castner SA, Goldman-Rakic PS, Williams GV (2004) Animal models of
working memory: insights for targeting cognitive dysfunction in
schizophrenia. Psychopharmacology (Berl) 174:111–125.
Cauli B, Audinat E, Lambolez B, Angulo MC, Ropert N, Tsuzuki K,
Hestrin S, Rossier J (1997) Molecular and physiological diversity of
cortical nonpyramidal cells. J Neurosci 17(10):3894–3906.
Chafee MV, Goldman-Rakic PS (1998) Matching patterns of activity in
primate prefrontal area 8a and parietal area 7ip during a spatial
working memory task. J Neurophysiol 79:2919–2940.
Chafee MV, Goldman-Rakic PS (2000) Inactivation of parietal and
prefrontal cortex reveals interdependence of neural activity during
memory-guided saccades. J Neurophysiol 83:1550–1566.
Chen G, Greengard P, Yan Z (2004) Potentiation of NMDA receptor
currents by dopamine D1 receptors in prefrontal cortex. Proc Natl
Acad Sci USA 101:2596–2600.
Cohen JD, Braver TS, Brown JW (2002) Computational perspectives
on dopamine function in prefrontal cortex. Curr Opin Neurobiol
Compte A, Brunel N, Goldman-Rakic PS, Wang X-J (2000) Synaptic
mechanisms and network dynamics underlying spatial working
memory in a cortical network model. Cereb Cortex 10:910–923.
Constantinidis C, Wang XJ (2004) A neural circuit basis for spatial
working memory. Neuroscientist 10(6):553–565.
Courtney SM, Petit L, Maisog JM, Ungerleider LG, Haxby JV (1998)
An area specialized for spatial working memory in human frontal
cortex. Science 279:1347–1351.
Curtis CE, D’Esposito M (2003) Persistent activity in the prefrontal
cortex during working memory. Trends Cogn Sci 7:415–423.
Davis KL, Kahn RS, Ko G, Davidson M (1991) Dopamine in schizophrenia:
a review and reconceptualization. Am J Psychiatry 148:1474–
de Fockert JW, Rees G, Frith CD, Lavie N (2001) The role of working
memory in visual selective attention. Science 291:1803–1806.
Del Arco A, Mora F (2005) Glutamate-dopamine in vivo interaction in
the prefrontal cortex modulates the release of dopamine and ace-
tylcholine in the nucleus accumbens of the awake rat. J Neural
Desimone R (1995) Is dopamine a missing link? Nature 376:549–550.
D’Esposito M, Postle BR, Ballard D, Lease J (1999) Maintenance
versus manipulation of information held in working memory: an
event-related fMRI study. Brain Cogn 41:66–86.
D’Esposito M, Postle BR, Rypma B (2000) Prefrontal cortical contri-
butions to working memory: evidence from event-related fMRI
studies. Exp Brain Res 133(1):3–11.
Dreher J-C, Burnod Y (2002) An integrative theory of the phasic and
tonic modes of dopamine modulation in the prefrontal cortex. Neu-
ral Netw 15:583–602.
Dreher J-C, Guigon E, Burnod Y (2002) A model of prefrontal cortex
dopaminergic modulation during the delayed alternation task.
J Cogn Neurosci 14:853–865.
Duman RS, Nestler EJ (2000) Signal transduction pathways for cate-
cholamine receptors. Psychopharmacology: The fourth generation
of progress. http://www.acnp.org/content-32html.
Durstewitz D, Kelc M, Gunturkun O (1999) A neurocomputational
theory of the dopaminergic modulation of working memory func-
tions. J Neurosci 19:2807–2822.
Durstewitz D, Seamans JK, Sejnowski TJ (2000a) Dopamine-medi-
ated stabilization of delay-period activity in a network model of
prefrontal cortex. J Neurophysiol 83:1733–1750.
Durstewitz D, Seamans JK, Sejnowski TJ (2000b) Neurocomputa-
tional models of working memory. Nat Neurosci 3:1184–1191.
Durstewitz D, Seamans JK (2002) The computational role of dopamine
D1 receptors in working memory. Neural Netw 15:561–572.
Ebi H, Yamashita K, Tanaka S (2005) Simulation of D1 and D2
receptor effects on working memory. Abstracts of the computa-
S. Tanaka / Neuroscience 139 (2006) 153–171166
tional neuroscience meeting, CNS 2005 (July 17–21, 2005
Madison, WI, USA), p. 71.
Elliott R (2003) Executive functions and their disorders. Br Med Bull
Elliott JM, Beveridge TJ (2005) Psychostimulants and monoamine
transporters: upsetting the balance. Curr Opin Pharmacol 5:94–
Elvevag B, Goldberg TE (2000) Cognitive impairment in schizophrenia
is the core of the disorder. Crit Rev Neurobiol 14(1):1–21.
Farah MJ, Illes J, Cook-Deegan R, Gardner H, Kandel E, King P, Parens
E, Sahakian B, Wolpe PR (2004) Neurocognitive enhancement: what
can we do and what should we do? Nat Rev Neurosci 5(5):
Fellous J-M, Sejnowski TJ (2003) Regulation of persistent activity by
background inhibition in an in vitro model of a cortical microcircuit.
Cereb Cortex 13:1232–1241.
Fiorillo CD, Tobler PN, Schultz W (2003) Discrete coding of reward
probability and uncertainty by dopamine neurons. Science 299:
Fleming K, Goldberg TE, Gold JM, Weinberger DR (1995) Verbal
working memory dysfunction in schizophrenia: use of a Brown-
Peterson paradigm. Psychiatry Res 56(2):155–161.
Fleming K, Goldberg TE, Binks S, Randolph C, Gold JM, Weinberger
DR (1997) Visuospatial working memory in patients with schizo-
phrenia. Biol Psychiatry 41(1):43–49.
Fletcher PC, Henson RNA (2001) Frontal lobes and human memory:
insights from functional neuroimaging. Brain 124:849–881.
Floresco SB, West AR, Ash B, Moore H, Grace AA (2003) Afferent
modulation of dopamine neuron firing differentially regulates tonic
and phasic dopamine transmission. Nat Neurosci 6(9):968–973.
Fone KC, Nutt DJ (2005) Stimulants: use and abuse in the treatment
of attention deficit hyperactivity disorder. Curr Opin Pharmacol
Frank MJ, Loughry B, O’Reilly RC (2001) Interactions between frontal
cortex and basal ganglia in working memory: a computational
model. Cogn Affect Behav Neurosci 1(2):137–160.
Frankle WG, Lerma J, Laruelle M (2003) The synaptic hypothesis of
schizophrenia. Neuron 39:205–216.
Friston KJ (1998) The disconnection hypothesis. Schizophr Res
Friston K (2005) Disconnection and cognitive dysmetria in schizophre-
nia. Am J Psychiatry 162(3):429–432.
Funahashi S, Bruce CJ, Goldman-Rakic PS (1989) Mnemonic coding
of visual space in the monkey’s dorsolateral prefrontal cortex.
J Neurophysiol 61:331–349.
Funahashi S (2001) Neuronal mechanisms of executive control by the
prefrontal cortex. Neurosci Res 39:147–165.
Fuster JM (2000) Executive frontal functions. Exp Brain Res 133:
Fuster JM (2001) The prefrontal cortex: an update: time is of the
essence. Neuron 30:319–333.
Fuster JM, Alexander GE (1971) Neuron activity related to short-term
memory. Science 173:652–654.
Gao W-J, Krimer LS, Goldman-Rakic PS (2001) Presynaptic regula-
tion of recurrent excitation by D1 receptors in prefrontal circuits.
Proc Natl Acad Sci USA 98:295–300.
Gao WJ, Goldman-Rakic PS (2003) Selective modulation of excitatory
and inhibitory microcircuits by dopamine. Proc Natl Acad Sci USA
Gao WJ, Wang Y, Goldman-Rakic PS (2003) Dopamine modulation of
perisomatic and peridendritic inhibition in prefrontal cortex. J Neu-
Goff DC, Coyle JT (2001) The emergent role of glutamate in the
pathophysiology and treatment of schizophrenia. Am J Psychiatry
Goldberg TE, Green MF (2000) Neurocognitive functioning in patients
with schizophrenia: an overview. Psychopharmacology: The fourth
generation of progress. http://www.acnp.org/content-32.html.
Goldberg TE, Green MF (2002) Neurocognitive functioning in patients
with schizophrenia: an neuropsychopharmacology: The fifth gen-
eration of progress. (Davis KL, Charney D, Coyle JT, Nemeroff C,
eds), pp. 657–669. Philadelphia: Lippincott Williams & Wilkins.
Goldberg TE, David A, Gold JM (2003) Neurocognitive deficits in
schizophrenia. In: Schizophrenia, 2e (Hirsch SR, Weinberger DR,
eds), pp 168–184. Malden: Blackwell.
Goldman-Rakic PS (1987) Circuitry of primate prefrontal cortex and
regulation of behavior by representational memory. In: Handbook
of physiology, sec 1: The nervous system, Vol. V: Higher functions
of the brain, Part 1. pp 373–417.
Goldman-Rakic PS (1994) Working memory dysfunction in schizo-
phrenia. J Neuropsychiatr Clin Neurosci 6:348–357.
Goldman-Rakic PS (1995) Cellular basis of working memory. Neuron
Goldman-Rakic PS (1996) Regional and cellular fractionation of work-
ing memory. Proc Natl Acad Sci USA 93:13473–13480.
Goldman-Rakic PS, Muly EC 3rd, Williams GV (2000) D1 receptors in
prefrontal cells and circuits. Brain Res Rev 31:295–301.
Goldman-Rakic PS, Castner SA, Svensson TH, Siever LJ, Williams
GV (2004) Targeting the dopamine D1 receptor in schizophrenia:
insights for cognitive dysfunction. Psychopharmacology (Berl) 174:
Gonzalez-Burgos G, Barrionuevo G, Lewis DA (2000) Horizontal syn-
aptic connections in monkey prefrontal cortex: an in vitro electro-
physiological study. Cereb Cortex 10:82–92.
Gonzalez-Islas C, Hablitz JJ (2001) Dopamine inhibition of evoked
IPSCs in rat prefrontal cortex. J Neurophysiol 86:2911–2918.
Gonzalez-Islas C, Hablitz JJ (2003) Dopamine enhances EPSCs in
layer II-III pyramidal neurons in rat prefrontal cortex. J Neurosci
Gooding DC, Tallent KA (2004) Nonverbal working memory deficits in
schizophrenia patients: evidence of a supramodal executive pro-
cessing deficit. Schizophr Res 68(2–3):189–201.
Gorelova NA, Yang CR (2000) Dopamine D1/D5 receptor activation
modulates a persistent sodium current in rat prefrontal cortical
neurons in vitro. J Neurophysiol 84:75–87.
Gorelova N, Seamans JK, Yang CR (2002) Mechanisms of dopamine
activation of fast-spiking interneurons that exert inhibition in rat
prefrontal cortex. J Neurophysiol 88:3150–3166.
Grace AA (1991) Phasic versus tonic dopamine release and the
modulation of dopamine system responsivity: a hypothesis for the
etiology of schizophrenia. Neuroscience 41:1–24.
Grace AA (1995) The tonic/phasic model of dopamine system
regulation: Its relevance for understanding how stimulant abuse
can alter basal ganglia function. Drug Alcohol Depend 37:111–
Grace AA (2000) The tonic/phasic model of dopamine system regula-
tion and its implications for understanding alcohol and psycho-
stimulant craving. Addiction 95 (Suppl 2):S119–S128.
Greengard P (2001) The neurobiology of slow synaptic transmission.
Guo N, Hwang DR, Lo ES, Huang YY, Laruelle M, Abi-Dargham A
(2003) Dopamine depletion and in vivo binding of PET D1 receptor
radioligands: implications for imaging studies in schizophrenia.
Gupta A, Wang Y, Markram H (2000) Organizing principles for a
diversity of GABAergic interneurons and synapses in the neocor-
tex. Science 287(5451):273–278.
Halberstadt AL (1995) The phencyclidine-glutamate model of schizo-
phrenia. Clin Neuropharmacol 18(3):237–249.
Hollerman JR, Tremblay L, Schultz W (1998) Influence of reward
expectation on behavior-related neuronal activity in primate stria-
tum. J Neurophysiol 80:947–963.
Honey GD, Bullmore ET, Sharma T (2002) Decoupling of cognitive
performance and cerebral functional response during working
memory in schizophrenia. Schizophr Res 53(1–2):45–56.
S. Tanaka / Neuroscience 139 (2006) 153–171167
Honey G, Bullmore E (2004) Human pharmacological MRI. Trends
Pharmacol Sci 25:366–374.
Jentsch JD, Tran A, Le D, Youngren KD, Roth RH (1997) Subchronic
phencyclidine administration reduces mesoprefrontal dopamine
utilization and impairs prefrontal cortical-dependent cognition in
the rat. Neuropsychopharmacology 17:92–99.
Jentsch JD, Roth RH (1999) The neuropsychopharmacology of
phencyclidine: from NMDA receptor hypofunction to the dopamine
hypothesis of schizophrenia. Neuropsychopharmacology 20:201–
Jentsch JD, Taylor JR, Elsworth JD, Redmond DEJ, Roth RH (1999)
Altered frontal cortical dopaminergic transmission in monkeys after
subchronic phencyclidine exposure: involvement in frontostriatal
cognitive deficits. Neuroscience 90(3):823–832.
Jentsch JD, Taylor JR, Roth RH (2000) Phencyclidine model of frontal
cortical dysfunction in nonhuman primates. Neuroscientist 6:
Jonides J, Schumacher EH, Smith EE, Koeppe RA, Awh E, Reuter-
Lorenz PA, Marshuetz C, Willis CR (1998) The role of parietal
cortex in verbal working memory. J Neurosci 18:5026–5034.
Kahn RS, Davis KL (2000) New developments in dopamine and
schizophrenia. Psychopharmacology: The fourth generation of
Kalivas PW, Jackson D, Romanidies A, Wyndham L, Duffy P (2001)
Involvement of pallidothalamic circuitry in working memory. Neu-
Kastner S, Ungerleider LG (2000) Mechanisms of visual attention in
the human cortex. Annu Rev Neurosci 23:315–341.
Kawaguchi Y (1993) Groupings of nonpyramidal and pyramidal cells
with specific physiological and morphological characteristics in rat
frontal cortex. J Neurophysiol 69:416–431.
Kawaguchi Y (1995) Physiological subgroups of nonpyramidal cells
with specific morphological characteristics in layer II/III of rat frontal
cortex. J Neurosci 15:2638–2655.
Kawaguchi Y, Kubota Y (1993) Correlation of physiological subgroups
of nonpyramidal cells with parvalbumin- and calbindinD28k-immu-
noreactive neurons in layer V of rat frontal cortex. J Neurophysiol
Kawaguchi Y, Kubota Y (1996) Physiological and morphological iden-
tification of somatostatin- or vasoactive intestinal polypeptide-con-
taining cells among GABAergic cell subtypes in rat frontal cortex.
J Neurosci 16:2701–2715.
Kawaguchi Y, Kubota Y (1997) GABAergic cell subtypes and their
synaptic connections in rat frontal cortex. Cereb Cortex 7:476–
Kegeles LS, Abi-Dargham A, Zea-Ponce Y, Rodenhiser-Hill J, Mann J,
Van Heertum RL, Cooper TB, Carlsson A, Laruelle M (2000)
Modulation of amphetamine-induced striatal dopamine release by
ketamine in humans: implications for schizophrenia. Biol Psychia-
Kim JS, Kornhuber HH, Schmid-Burgk W, Holzmuller B (1980) Low
cerebrospinal fluid glutamate in schizophrenic patients and a new
hypothesis on schizophrenia. Neurosci Lett 20(3):379–382.
Kimberg DY, D’Esposito M, Farah MJ (1997) Effects of bromocriptine
on human subjects depend on working memory capacity. Neuro-
Kimberg DY, Aguirre GK, Lease J, D’Esposito M (2001) Cortical
effects of bromocriptine, a D2 dopamine receptor agonist, in hu-
man subjects, revealed by fMRI. Hum Brain Mapp 12:246–257.
Koh PO, Bergson C, Undie A, Goldman-Rakic PS, Lidow MS (2003)
Up-regulation of the D1 dopamine receptor-interacting protein,
calcyon, in patients with schizophrenia. Arch Gen Psychiatry 60:
Krimer LS, Goldman-Rakic PS (2001) Prefrontal microcircuits: mem-
brane properties and excitatory input of local, medium, and wide
arbor interneurons. J Neurosci 21:3788–3796.
Krystal JH, D’Souza DC, Mathalon D, et al. (2003) NMDA receptor
antagonist effects, cortical glutamatergic function, and schizophrenia:
Kubota K, Niki H (1971) Prefrontal cortical unit activity and delayed
alternation performance in monkeys. J Neurophysiol 34:337–347.
Kuperberg G, Heckers S (2000) Schizophrenia and cognitive function.
Curr Opin Neurobiol 10(2):205–210.
Laruelle M (2003) Dopamine transmission in the schizophrenic brain.
In: Schizophrenia, 2/e (Hirsch SR, Weinberger DR, eds), pp 365–
387. Malden: Blackwell.
Laruelle M, Abi-Dargham A, Gil R, Kegeles L, Innis R (1999) Increased
dopamine transmission in schizophrenia: relationship to illness
phases. Biol Psychiatry 46:56–72.
Laruelle M, Abi-Dargham A (2003) Dopamine imbalances in the
schizophrenic brain: imaging studies. In: Dopamine receptors and
transporters, 2/e (Sidhu A, Laruelle M, Vernier P, eds), pp 593–
642. New York: Marcel Dekker.
Laruelle M, Kegeles LS, Abi-Dargham A (2003) Glutamate, dopamine,
and schizophrenia: from pathophysiology to treatment. Ann N Y
Acad Sci 1003:138–158.
Latham PE, Nirenberg S (2004) Computing and stability in cortical
networks. Neural Comput 16:1385–1412.
Le Moal M(2000) Mesocorticolimbic dopaminergic neurons: functional
and regulatory roles. Psychopharmacology: The fourth generation
of progress. http://www.acnp.org/content-32.html.
Le Pen G, Grottick AJ, Higgins GA, Moreau J-L (2003) Phencyclidine
exacerbates attentional deficits in a neurodevelopmental rat model
of schizophrenia. Neuropsychopharmacology 28(10):1799–1809.
Leung H-C, Gore JC, Goldman-Rakic PS (2002) Sustained mnemonic
response in the human middle frontal gyrus during on-line storage
of spatial memoranda. J Cogn Neurosci 14:659–671.
Lewis R (2004) Should cognitive deficit be a diagnostic criterion for
schizophrenia? J Psychiatry Neurosci 29(2):102–113.
Lidow MS, Goldman-Rakic PS, Rakic P, Innis RB (1989) Dopamine D2
receptors in the cerebral cortex: distribution and pharmacological
characterization with [3H]raclopride. Proc Natl Acad Sci USA 86
Lidow MS, Goldman-Rakic PS (1994) A common action of clozapine,
haloperidol, and remoxipride on D1 and D2-dopaminergic recep-
tors in the primate cerebral cortex. Proc Natl Acad Sci USA 91:
Luciana M, Depue RA, Arbisi P, Leon A (1992) Facilitation of working
memory in humans by a D2 dopamine receptor agonist. J Cogn
Luciana M, Collins PF (1997) Dopaminergic modulation of working
memory for spatial but not object cues in normal humans. J Cogn
Manoach DS (2003) Prefrontal cortex dysfunction during working
memory performance in schizophrenia: reconciling discrepant find-
ings. Schizophr Res 60(2–3):285–298.
Manoach DS, Press DZ, Thangaraj V, Searl MM, Goff DC, Halpern E,
Saper CB, Warach S (1999) Schizophrenic subjects activate dor-
solateral prefrontal cortex during a working memory task, as mea-
sured by fMRI. Biol Psychiatry 45:1128–1137.
Manoach DS, Gollub RL, Benson ES, Searl MM, Goff DC, Halpern E,
Saper CB, Rauch SL (2000) Schizophrenic subjects show aberrant
fMRI activation of dorsolateral prefrontal cortex and basal ganglia
during working memory performance. Biol Psychiatry 48(2):99–
Mattay VS, Callicott JH, Bertolino A, Heaton I, Frank JA, Coppola R,
Berman KF, Goldberg TE, Weinberger DR (2000) Effects of dex-
troamphetamine on cognitive performance and cortical activation.
McDowell S, Whyte J, D’Esposito M (1998) Differential effect of a
dopaminergic agonist on prefrontal function in traumatic brain in-
jury patients. Brain 121:1155–1164.
Meador-Woodruff JH, Healy DJ (2000) Glutamate receptor expression
in schizophrenic brain. Brain Res Rev 31:288–294.
a paradigmshiftin medication development.
S. Tanaka / Neuroscience 139 (2006) 153–171168
Meador-Woodruff JH, Kleinman JE (2002) Neurochemistry of schizo-
phrenia: glutamatergic abnormalities. In: Neuropsychopharmacology:
The fifth generation of progress (Davis KL, Charney D, Coyle JT,
Nemeroff C, eds), pp. 717–728. Philadelphia: Lippincott Williams &
Mehta MA, Swainson R, Ogilvie AD, et al (2001) Improved short-term
spatial memory but impaired reversal learning following the dopa-
mine D2 agonist bromocriptine in human volunteers. Psychophar-
macology (Berl) 159:10–20.
Mehta MA, Manes FF, Magnolfi G, Sahakian BJ, Robbins TW (2004)
Impaired set-shifting and dissociable effects on tests of spatial
working memory following the dopamine D2 receptor antagonist
sulpiride in human volunteers. Psychopharmacology (Berl) 176
Melchitzky DS, Sesack SR, Pucak ML, Lewis SA (1998) Synaptic
targets of pyramidal neurons providing intrinsic horizontal connec-
tions in monkey prefrontal cortex. J Comp Neurol 390:211–224.
Melchitzky DS, Gonzalez-Burgos G, Barrionuevo G, Lewis DA (2001)
Synaptic targets of the intrinsic axon collaterals of supragranular
pyramidal neurons in monkey prefrontal cortex. J Comp Neurol
Melchitzky DS, Lewis DA (2003) Pyramidal neuron local axon termi-
nals in monkey prefrontal cortex: differential targeting of sub-
classes of GABA neurons. Cereb Cortex 13:452–460.
Meyer-Lindenberg A, Poline J-B, Kohn PD, Holt JL, Egan MF, Wein-
berger DR, Berman KF (2001) Evidence for abnormal cortical
functional connectivity during working memory in schizophrenia.
Am J Psychiatry 158:1809–1817.
Meyer-Lindenberg A, Miletich RS, Kohn PD, Esposito G, Carson RE,
Quarantelli M, Weinberger DR, Berman KF (2002) Reduced pre-
frontal activity predicts exaggerated striatal dopaminergic function
in schizophrenia. Nat Neurosci 5(3):267–271.
Miles R (2000) Diversity in inhibition. Science 287(5451):244–246.
Miyashita S, Tabuchi Y, Tanaka S (2003) Cortico-thalamocortical op-
erations of multi-target spatial working memory. Neurocomputing
Moghaddam B (2003) Bringing order to the glutamate chaos in schizo-
phrenia. Neuron 40:881–884.
Montague PR, Hyman SE, Cohen JD (2004) Computational roles for
dopamine in behavioural control. Nature 431(7010):760–767.
Moody SL, Wise SP, di Pellegrino G, Zipser D (1998) A model that
accounts for activity in primate frontal cortex during a delayed
matching-to-sample task. J Neurosci 18:399–410.
Morari M, Marti M, Sbrenna S, Fuxe K, Bianchi C, Beani L (1998)
Reciprocal dopamine-glutamate modulation of release in the basal
ganglia. Neurochem Int 33(5):383–397.
Muller U, von Cramon DY, Pollmann S (1998) D1- versus D2-receptor
modulation of visuospatial working memory in humans. J Neurosci
Muly EC, Szigeti K, Goldman-Rakic PS (1998) D1 receptor in inter-
neurons of macaque prefrontal cortex: distribution and subcellular
localization. J Neurosci 15:10553–10565.
Noda Y, Kamei H, Mamiya T, Furukawa H, Nabeshima T (2000)
Repeated phencyclidine treatment induces negative symptom-like
behavior in forced swimming test in mice: imbalance of prefrontal
serotonergic and dopaminergic functions. Neuropsychopharma-
Okubo Y, Suhara T, Suzuki K, Kobayashi K, Inoue O, Terasaki O,
Someya Y, Sassa T, Sudo Y, Matsushima E, Iyo M, Tateno Y, Toru
M (1997) Decreased prefrontal dopamine D1 receptors in schizo-
phrenia revealed by PET. Nature 385:634–636.
Olney JW, Farber NB (1995a) Glutamate receptor dysfunction and
schizophrenia. Arch Gen Psychiatry 52:998–1007.
Olney JW, Farber NB (1995b) NMDA antagonists as neurotherapeutic
drugs, psychotogens, neurotoxins, and research tools for studying
schizophrenia. Neuropsychopharmacology 13:335–345.
O’Reilly RC, Braver TS, Cohen JD (1999) A biologically-based neural
network model of working memory. In: Models of working memory
(Shah P, Miyake A, eds). pp. 375–411. Cambridge: Cambridge
Owen AM, Evans AC, Petrides M (1996) Evidence for a two-stage
model of spatial working memory processing within the lateral
frontal cortex: a positron emission tomography study. Cereb Cor-
Pantelis C, Barnes TR, Nelson HE, Tanner S, Weatherley L, Owen
AM, Robbins T (1997) Frontal-striatal cognitive deficits in patients
with chronic schizophrenia. Brain 120:1823–1843.
Park S, Holzman PS (1992) Schizophrenics show spatial working
memory deficits. Arch Gen Psychiatry 49(12):975–982.
Park S, Holzman PS (1993) Association of working memory deficit and
eye tracking dysfunction in schizophrenia. Schizophr Res 11(1):
Paulman RG, Devous MD Sr, Gregory RR, Herman JH, Jennings L,
Bonte FJ, Nasrallah HA, Raese JD (1990) Hypofrontality and cog-
nitive impairment in schizophrenia: dynamic single-photon tomog-
raphy and neuropsychological assessment of schizophrenic brain
function. Biol Psychiatry 27(4):377–399.
Pedarzani P, Storm JF (1995) Dopamine modulates the slow Ca2?-
activated K? current IAHP via cyclic AMP-dependent protein ki-
nase in hippocampal neurons. J Neurophysiol 74:2749–2753.
Perlstein WM, Carter CS, Noll DC, Cohen JD (2001) Relation of
prefrontal cortex dysfunction to working memory and symptoms in
schizophrenia. Am J Psychiat 158:1105–1113.
Perlstein WM, Dixit NK, Carter CS, Noll DC, Cohen JD (2003) Pre-
frontal cortex dysfunction mediates deficits in working memory and
prepotent responding in schizophrenia. Biol Psychiatry 53:25–38.
Perner J, Lang B (1999) Development of theory of mind and executive
control. Trends Cogn Sci 3:337–344.
Phillips AG, Ahn S, Floresco SB (2004) Magnitude of dopamine re-
lease in medial prefrontal cortex predicts accuracy of memory on a
delayed response task. J Neurosci 24:547–553.
Pierce RC, Kalivas PW (1997) A circuitry model of the expression of
behavioral sensitization to amphetamine-like psychostimulants.
Brain Res Rev 25(2):192–216.
Ramsey NF, Koning HA, Welles P, Cahn W, van der Linden JA, Kahn
RS (2002) Excessive recruitment of neural systems subserving
logical reasoning in schizophrenia. Brain 125(Pt 8):1793–1807.
Rao SG, Williams GV, Goldman-Rakic PS (1999) Isodirectional tuning
of adjacent interneurons and pyramidal cells during working
memory: Evidence for microcolumnar organization in PFC. J Neu-
Rao SG, Williams GV, Goldman-Rakic PS (2000) Destruction and
creation of spatial tuning by disinhibition: GABA(A) blockade of
prefrontal cortical neurons engaged by working memory. J Neuro-
Robbins TW (1990) The case of frontostriatal dysfunction in schizo-
phrenia. Schizophr Bull 16(3):391–402.
Robbins TW (1996) Dissociating executive functions of the prefrontal
cortex. Philos Trans R Soc Lond B Biol Sci 351(1346):1463–1470.
Roesch-Ely D, Scheffel H, Weiland S, Schwaninger M, Hundemer HP,
Kolter T, Weisbrod M (2005) Differential dopaminergic modulation
of executive control in healthy subjects. Psychopharmacology
Rowe JB, Toni I, Josephs O, Frackowiak RSJ, Passingham RE (2000)
The prefrontal cortex: response selection or maintenance within
working memory? Science 288:1656–1660.
Royall DR, Lauterbach EC, Cummings JL, Reeve A, Rummans TA,
Kaufer DI, LaFrance Jr WC, Coffey CE (2002) Executive control
function: a review of its promise and challenges for clinical re-
search. A report from the Committee on Research of the American
Neuropsychiatric Association. J Neuropsychiatry Clin Neurosci
Sakai K, Rowe JB, Passingham RE (2002) Active maintenance in
prefrontal area 46 creates distractor-resistant memory. Nat Neu-
S. Tanaka / Neuroscience 139 (2006) 153–171169
Sawaguchi T (2001) The effects of dopamine and its antagonists on
directional delay-period activity of prefrontal neurons in monkeys
during an oculomotor delayed-response task. Neurosci Res 41
Sawaguchi T, Matsumura M, Kubota K (1990a) Catecholaminergic
effects on neuronal activity related to a delayed response task in
monkey prefrontal cortex. J Neurophysiol 63:1385–1400.
Sawaguchi T, Matsumura M, Kubota K (1990b) Effects of dopamine
antagonists on neuronal activity related to a delayed response task
in monkey prefrontal cortex. J Neurophysiol 63:1401–1412.
Sawaguchi T, Goldman-Rakic PS (1991) D1 dopamine receptors in
prefrontal cortex: involvement in working memory. Science 251:
Sawaguchi T, Goldman-Rakic PS (1994) The role of D1-dopamine
receptor in working memory: local injections of dopamine antago-
nists into the prefrontal cortex of rhesus monkeys performing an
oculomotor delayed-response task. J Neurophysiol 71:515–528.
Schultz W (1998) Predictive reward signal of dopamine neurons.
J Neurophysiol 80:1–27.
Schultz W (2001) Reward signaling by dopamine neurons. Neurosci-
Schultz W (2002) Getting formal with dopamine and reward. Neuron
Seamans JK, Durstewitz D, Christie B, Stevens CF, Sejnowski TJ
(2001a) Dopamine D1/D5 receptor modulation of excitatory syn-
aptic inputs to layer V prefrontal cortical neurons. Proc Natl Acad
Sci USA 98:301–306.
Seamans JK, Gorelova N, Durstewitz D, Yang CR (2001b) Bidirec-
tional dopamine modulation of GABAergic inhibition in prefrontal
cortical pyramidal neurons. J Neurosci 21:3628–3638.
Seamans JK, Yang CR (2004) The principal features and mechanisms
of dopamine modulation in the prefrontal cortex. Prog Neurobiol
Seeman P, Madras BK (1998) Anti-hyperactivity medication: methyl-
phenidate and amphetamine. Mol Psychiatry (Nature) 3:386–396.
Seeman P, Kapur S (2000) Schizophrenia: more dopamine, more D2
receptors. Proc Natl Acad Sci USA 97(14):7673–7675.
Sesack SR, Carr DB (2002) Selective prefrontal cortex inputs to
dopamine cells: implications for schizophrenia. Physiol Behav
Sesack SR, Carr DB, Omelchenko N, Pinto A (2003) Anatomical
substrates for glutamate-dopamine interactions: evidence for
specificity of connections and extrasynaptic actions. Ann N Y Acad
Smith EE, Jonides J (1999) Storage and executive processes in the
frontal lobes. Science 283:1657–1661.
Solanto MV (1998) Neuropsychopharmacological mechanisms of
stimulant drug action in attention-deficit hyperactivity disorder: a
review and integration. Behav Brain Res 94:127–152.
Somogyi P, Tamas G, Lujan R, Buhl EH (1998) Salient features of
synaptic organisation in the cerebral cortex. Brain Res Rev 26:
Stahl SM (2000) Essential psychopharmacology: neuroscientific basis
and practical applications, 2nd Edition. Cambridge, UK: Cam-
bridge University Press.
Stedron JM, Sahni SD, Munakata Y (2005) Common mechanisms for
working memory and attention: the case of perseveration with
visible solutions. J Cogn Neurosci 17(4):623–631.
Steketee JD (2003) Neurotransmitter systems of the medial prefrontal
cortex: potential role in sensitization to psychostimulants. Brain
Res Rev 41:203–228.
Suhara T, Okubo Y, Yasuno F, Sudo Y, Inoue M, Ichimiya T, Na-
kashima Y, Nakayama K, Tanada S, Suzuki K, Halldin C, Farde L
(2002) Decreased dopamine D2 receptor binding in the anterior
cingulate cortex in schizophrenia. Arch Gen Psychiatry 59:25–30.
Tabuchi Y, Tanaka S (2003) Intracortical vs. halamocortical process-
ing of spatial working memory. Neurocomputing 52–54:403–
Tamas G, Somogyi P, Buhl EH (1998) Differentially interconnected
networks of GABAergic interneurons in the visual cortex of the cat.
J Neurosci 18:4255–4270.
Tanaka S (1999) Architecture and dynamics of the primate prefrontal
cortical circuit for spatial working memory. Neural Netw 12:
Tanaka S (2000) Post-cue activity of prefrontal cortical neurons con-
trolled by local inhibition. Neurocomputing 32–33:563–572.
Tanaka S (2001) Computational approaches to the architecture and
operations of the prefrontal cortical circuit for working memory.
Prog Neuropsychopharmacol Biol Psychiatry 25:259–281.
Tanaka S (2002a) Multi-directional representation of spatial working
memory in a model prefrontal cortical circuit. Neurocomputing
Tanaka S (2002b) Dopamine controls fundamental cognitive opera-
tions of multi-target spatial working memory. Neural Netw 15:
Tanaka S (2003) A prefronto-parietal network model with feedforward
and feedback connections. Neurocomputing 52–54:943–948.
Tanaka S (2005) State-dependent alteration of dopamine and gluta-
mate transmission in the prefrontal cortex by psychostimulants.
Tanaka S, Slifstein M, Abi-Dargham A (2005) Stability theory of pre-
frontal cortical activation in schizophrenia. Soc Neurosci Abstr 35:
Tanibuchi I, Goldman-Rakic PS (2003) Dissociation of spatial-, object-,
and sound-coding neurons in the mediodorsal nucleus of the pri-
mate thalamus. J Neurophysiol 89(2):1067–1077.
Tanibuchi I, Goldman-Rakic PS (2005) Comparison of oculomotor
neuronal activity in paralaminar and mediodorsal thalamus in the
rhesus monkey. J Neurophysiol 93(1):614–619.
Tanji J, Hoshi E (2001) Behavioral planning in the prefrontal cortex.
Curr Opin Neurobiol 11:164–170.
Tsai G, Coyle JT (2002) Glutamatergic mechanisms in schizophrenia.
Annu Rev Pharmacol Toxicol 42:165–179.
Tseng KY, O’Donnell P (2003) Dopamine-glutamate interactions in the
control of cell excitability in medial prefrontal cortical pyramidal
neurons from adult rats. Ann N Y Acad Sci 1003:476–478.
Tseng KY, O’Donnell P (2004) Dopamine-glutamate interactions con-
trolling prefrontal cortical pyramidal cell excitability involve multiple
signaling mechanisms. J Neurosci 24:5131–5139.
Tzschentke TM (2001) Pharmacology and behavioral pharmacology of
the mesocortical dopamine system. Prog Neurobiol 63:241–320.
Ungerleider LG (1998) A neural system for human visual working
memory. Proc Natl Acad Sci USA 95:883–890.
Waelti P, Dickinson A, Schultz W (2001) Dopamine responses comply
with basic assumptions of formal learning theory. Nature 412:
Wang M, Vijayraghavan S, Goldman-Rakic PS (2004a) Selective D2
receptor actions on the functional circuitry of working memory.
Wang X-J (1999) Synaptic basis of cortical persistent activity: the
importance of NMDA receptors to working memory. J Neurosci
Wang X-J (2001) Synaptic reverberation underlying mnemonic persis-
tent activity. Trends Neurosci 24:455–463.
Wang X-J, Tegner J, Constantinidis C, Goldman-Rakic PS (2004b)
Division of labor among distinct subtypes of inhibitory neurons in a
cortical microcircuit of working memory. Proc Natl Acad Sci USA
Watanabe Y, Funahashi S (2004a) Neuronal activity throughout the
primate mediodorsal nucleus of the thalamus during oculomotor
delayed-responses. I. Cue-, delay-, and response-period activity.
J Neurophysiol 92:1738–1755.
Watanabe Y, Funahashi S (2004b) Neuronal activity throughout the
primate mediodorsal nucleus of the thalamus during oculomotor
delayed-responses. II. Activity encoding visual versus motor sig-
nal. J Neurophysiol 92:1756–1769.
S. Tanaka / Neuroscience 139 (2006) 153–171170
Weinberger DR, Gallhofer B (1997) Cognitive function in schizophre- Download full-text
nia. Int Clin Psychopharmacol 12 (Suppl 4):S29–S36.
Weinberger DR, Berman KF (1998) Speculation on the meaning of
cerebral metabolic hypofrontality in schizophrenia. Schizophr Bull
Weinberger DR, Egan MF, Bertolino A, Callicott JH, Mattay VS, Lipska
BK, Berman KF, Goldberg TE (2001) Prefrontal neurons and the
genetics of schizophrenia. Biol Psychiatry 50(11):825–844.
West AR, Floresco SB, Charara A, Rosenkranz JA, Grace AA (2003)
Electrophysiological interactions between striatal glutamatergic
and dopaminergic systems. Ann N Y Acad Sci 1003:53–74.
White FJ, Kalivas PW (1998) Neuroadaptations involved in amphetamine
and cocaine addiction. Drug Alcohol Depend 51(1–2):141–153.
Williams GV, Goldman-Rakic PS (1995) Modulation of memory fields by
dopamine D1 receptors in prefrontal cortex. Nature 376:572–575.
Williams SM, Goldman-Rakic PS (1998) Widespread origin of the
primate mesofrontal dopamine system. Cereb Cortex 8:321–345.
Winterer G, Weinberger DR (2004) Genes, dopamine and cortical
signal-to-noise ratio in schizophrenia. Trends Neurosci 27(11):
Yamashita K, Tanaka S (2002) Circuit simulation of memory field
modulation by dopamine D1 receptor activation. Neurocomputing
Yamashita K, Tanaka S (2003) Circuit properties of the cortico-meso-
cortical system. Neurocomputing 52–54:969–975.
Yamashita K, Tanaka S (2005) Parametric study of dopaminergic
neuromodulatory effects in a reduced model of the prefrontal cor-
tex. Neurocomputing 65–66:579–588.
Yang CR, Seamans JK, Gorelova N (1999) Developing a neuronal
model for the pathophysiology of schizophrenia based on the
nature of electrophysiological actions of dopamine in the prefrontal
cortex. Neuropsychopharmacology 21:161–194.
Zahrt J, Taylor JR, Mathew RG, Arnsten AFT (1997) Supranormal
stimulation of D1 dopamine receptors in the rodent prefrontal cor-
tex impairs spatial working memory performance. J Neurosci 17:
Zhou F-M, Hablitz JJ (1999) Dopamine modulation of membrane and
synaptic properties of interneurons in rat cerebral cortex. J Neuro-
(Accepted 24 August 2005)
(Available online 1 December 2005)
S. Tanaka / Neuroscience 139 (2006) 153–171171