Dopamine Receptors And Schizophrenia Contribution Of Molecular Genetics And Clinical Neuropsychology

Prévia do material em texto

R
E
V
IE
W
 A
R
TIC
LE
International Journal of Neuropsychopharmacology (1999), 2, 197–227. Copyright # 1999 CINP
Dopamine receptors and schizophrenia :
contribution of molecular genetics and clinical
neuropsychology
Ge! rard Emilien1, Jean-Marie Maloteaux1,2, Muriel Geurts1 and Michael J. Owen3
Departments of " Pharmacology and #Neurology, UniversiteU Catholique de Louvain, Cliniques Universitaires Saint Luc, B-1200
Brussels, Belgium
$Departments of Psychological Medicine and Medical Genetics, University of Wales College of Medicine, Cardiff, CF4 4XN, UK
Abstract
Family, twin and adoption studies suggest that genetic factors play an important role in the aetiology of
schizophrenia. The mode of inheritance, however, is complex and non-Mendelian. Although the aetiology of
schizophrenia is unknown, it has been hypothesized that the necessary conditions for developing the disease
are environmental stress and a vulnerability to psychosis. The implication of dopamine receptors to
schizophrenia has been greatly studied. Several linkage and association studies have been performed in an
attempt to establish the involvement of dopamine receptors in schizophrenia. However, although no
conclusive evidence of linkage or association to any gene has been established, some results, suggestive of
linkage for chromosomes 6, 22 and 13, await confirmation from other studies. Concerning association studies,
it is also of interest that some studies support an association between schizophrenia and homozygosity at D
$
.
More work in larger samples is required before conclusive linkage hypothesis or association to a dopamine
receptor may be established.
Schizophrenic patients have been shown to have significant deficits in a wide range of cognitive processes,
including memory, attention, reasoning ability and language. Since cognitive deficits are significant symptoms
of schizophrenia which require effective treatment, their assessment in schizophrenic patients and during
clinical trials of new potential antipsychotics is highlighted. Cognitive impairment in schizophrenia impedes
psychosocial performance and is therefore an especially relevant target variable in the development of new
therapeutic approaches. It is most prominent in tasks involving attention, memory and executive functions
which are thought to reflect involvement of prefrontal and left-temporal brain areas. Semantic networks in
schizophrenic patients with a younger age of onset are observed to be more disorganized and differ
significantly to those of control subjects. The need to use broader approaches such as neuropsychological-
related measures to identify pertinent phenotypes in non-affected subjects carrying vulnerability genes is also
emphasized.
Since dopamine receptors are the primary targets in the treatment of schizophrenia, improved therapy may
be obtained by drugs that selectively target a particular subtype of dopamine receptor. In the development of
novel antipsychotics, D
$
and D
%
receptors have received much attention and this is partly related to the fact
that these receptors have a high abundance in brain areas associated with cognitive and emotional functions,
such as parts of the limbic system and cortex. Recent studies suggest that atypical neuroleptics may
significantly improve the cognitive deficits observed in schizophrenic patients and that atypical neuroleptics
such as risperidone appear to improve memory and alertness suggesting that further clinical studies are needed
to determine the precise influence of antipsychotics on the cognitive system of schizophrenic patients. Such
studies could lead to useful insights as to the potential advantages of the newer antipsychotics which appear
to have a sparing or beneficial effect on various components of cognitive function. However, the observation
that cortical D
#
receptors are important sites of action for antipsychotics, that the cerebral cortex may harbour
the common sites of actions of antipsychotics and that the balancing of the opposing actions of D
"
and D
#
receptor regulation may be an appropriate drug treatment suggests that the adjustment of D
"
receptor levels
in the cortex may become an important goal of future antipsychotic generation. Such antipsychotics will be
able to treat the positive, negative and cognitive deficits of schizophrenia.
Received 1 October 1998 ; Reviewed 29 November 1998 ; Revised 3 February 1999 ; Accepted 10 February 1999
Key words : Schizophrenia, dopamine receptors, clinical neuropsychology, neuroleptics, antipsychotics,
molecular genetics, cognition, memory.
Address for correspondence : Dr Ge! rard Emilien, 127 rue Henri Prou, 78340 Les Clayes Sous Bois, France.
Tel. : 33 1 41 02 7464 Fax : 33 1 30 54 02 47 E-mail : GEmilien!aol.Com
198 G. Emilien et al.
Contents
1. Introduction 198
2. Role of dopamine in schizophrenia 199
2.1. Structure and characteristics of dopamine
receptors 199
2.2. Symptoms of schizophrenia 200
2.3. Implications of dopamine receptors 200
2.4. Dopamine receptors and cognitive function 202
2.5. Issues and criticisms 202
3. Molecular genetics of schizophrenia 203
3.1. Linkage studies 203
3.2. Genome scanning studies 203
3.3. Association studies 206
3.4. Criticisms and perspectives 208
4. Cognitive impairment in schizophrenia 208
4.1. Attention 209
4.2. Language and information processing 209
4.3. Executive function and memory 212
4.4. Cognitive deficits and negative symptoms 213
4.5. Neuropsychological deficits as markers of
vulnerability 214
5. Methodological issues and testing 215
5.1. Computerized neuropsychological test battery 215
5.2. Mini-mental state examination 215
6. Effects of neuroleptics on cognitive deficits 215
6.1. Clozapine 216
6.2. Risperidone 217
6.3. New generation of antipsychotic drugs 217
7. Discussion and conclusion 218
8. Acknowledgements 220
9. References 220
1. Introduction
Schizophrenia, the most severe of the mental illnesses, is
a psychotic disorder characterized by late-adolescent or
early adult onset and has, in most cases, a chronic course.
While the average lifetime prevalence, regardless of race
or country, is about 1%, in the siblings of patients it is
8–10%, and in the children of patients it is 12–15%
(Regier et al., 1988 ; Gottesman, 1993) ; 10% of patients
die from suicide (Caldwell and Gottesman, 1992). A
recent study of the genetic epidemiology of schizophrenia
in a Finnish twin cohort suggests that there is a marginally
higher prevalence of schizophrenia in men (2–2%) than
women (1–8%) (Cannon et al., 1998 ; Kringlen, 1990). The
aetiology of schizophrenia is still unknown, but current
research suggests that the necessary conditions for
developing psychosis are environmental stress and a
vulnerability to psychosis (McGue et al., 1983 ; McGuffin
et al., 1994). It has been suggested that the transmission of
liability to schizophrenia could be accounted for by
genetic factors only and that environmental factors may
be idiosyncratic and random. These important idiosyn-
cratic effects could contribute as much as 26% to the
variance in total liability to definite schizophrenia. In
another study, when the population risk of schizophrenia
was fixed at 0–6%, model fitting indicated that the
maximum-likelihood of heritability was 87–5% (McGuffin
et al., 1994). The use of a second model to attempt to
estimate simultaneously the genetic and residual com-
ponents plus the population risk increased the heritability
to 89%, with a population risk of 0–3%.
Endophenotypes are traits that are associated with the
expression of an illness and are believed to represent the
genetic liability of the disorder among non-affected
subjects. Endophenotypescan include neurophysiological
or neuropsychological and cognitive measures. To meet
the criteria for a marker trait, an endophenotype must
occur before the onset of illness and must be heritable. If
a marker is a vulnerability trait for an illness then the
genes that are important for the expression of this
endophenotype will allow the identification of genes that
increase the susceptibility for the illness. When similar
perturbations at a given test are observed both in clinically
stable schizophrenics and their non-schizophrenic first-
degree relatives, this test could be qualified as an indicator
of the vulnerability to schizophrenia. This appears to be
the case for several neuropsychological tasks, exploring
attentional abilities such as the Continuous Performance
Task (CPT), Span of Apprehension Task (SAT) and
Wisconsin Card Sorting Test (WCST) (Franke et al., 1992 ;
Kremen et al., 1994 ; Nuechterlein et al., 1994). This
vulnerability is probably genetic but is not usually active
or expressed until the late-adolescent developmental
phase. The deficient processes involved in this order are
clinically silent until the onset of prodromal or psychotic
symptoms, at which time neuropsychological testing
demonstrates the presence of cognitive deficits that are
often chronic and apparently irreversible (Goldberg et al.,
1993). Thus, cognitive impairment is a central mani-
festation of the schizophrenic illness that impacts on the
quality of life of the patient and the cost of the illness to
society. Apart from the potential utility of indicating the
degree or severity of neuropsychological deficiency, the
measurement of cognitive processes and its impairment
may also serve as an indicator or marker of vulnerability
to schizophrenia in normal individuals at high risk of
schizophrenia.
Drug development for schizophrenia has previously
relied almost exclusively on laboratory animal models.
Since pathophysiological mechanisms remain obscure,
this strategy has led to many ‘me-too ’ drugs but no new
treatment strategies. Advances in molecular genetics offer
the possibility of more rational approaches to the
199Dopamine receptors and schizophrenia
treatment of schizophrenia by leading to insights into the
pathophysiology of the disorder, helping to identify
patients at high risk of developing the disease to whom
early treatment could be targeted, and allowing the
identification of subgroups of patients who may be
particularly responsive to a particular type}class of drug
treatment.
A recent comprehensive survey of the content and
quality of 2000 controlled trials in schizophrenia over the
last 50 yr indicates several important issues which suggest
that half a century of studies of limited quality, and clinical
utility leave much scope for well planned and well
conducted clinical trials (Thornley & Adams, 1998).
Among the objectives for a novel antipsychotic drug is an
expanded efficacy profile. More often, the primary
objective of treatment is the improvement of negative
symptoms which respond poorly to most current anti-
psychotic drugs. The ability to either prevent further
cognitive deterioration or improve performance are
equally important targets. An important criticism in the
above report was that since it is important to assess
normal cognitive functioning in schizophrenia studies,
proper measurement of this aspect was largely neglected.
Often, the lack of statistical power was reflected in the use
of an extraordinary number of non-validated tests and
rating scales and that it is often possible to achieve
significance on these measures with a small number of
patients. As low-quality scores were associated with an
increased estimate of benefit, these schizophrenia trials
may well have consistently overestimated the effects of
experimental interventions. Distinctive research progress
in molecular genetics and neuropsychology has brought
significant contributions to the area of neuropsychiatric
disease and the effect of neuropharmacological inter-
ventions, particularly in pharmacological treatment of
schizophrenia. Considering research advances during the
last few years, it may be suggested that significant and
continuous advancement in these two scientific areas
which, though at first instance seem unrelated, has
considerably increased our understanding of the genetics
and cognitive processes involved in the aetiology and
treatment of schizophrenia. Since neuropsychological
deficits are significant variables to be considered and
assessed in schizophrenia, there is some evidence that
atypical neuroleptics may, owing to their novel mechan-
isms of action, have the capacity to remediate cognitive
impairment in schizophrenia. This paper critically reviews
the importance and significance of neuropsychological
deficits in schizophrenia and argues for the inclusion of the
assessment of cognitive parameters in clinical trials of new
investigational drugs for the treatment of schizophrenia.
It is hypothesized that the assessment of neuro-
psychological variables in vulnerable schizophrenic popu-
lations will further increase our understanding of the
genetics of schizophrenia. Further advances in our under-
standing of the dopaminergic receptors and the genetics
of schizophrenia are also included in this discussion ;
outlining the potential, limits and future perspectives of
our comprehension of the pharmacological possibilities
for the effective treatment of schizophrenia. Finally, this
paper assembles two areas of research, which are often
published separately in various journals, to carefully
consider the possibility and perspective of producing a
superior future pharmacological treatment of schizo-
phrenia.
2. Role of dopamine in schizophrenia
2.1. Structure and characteristics of dopamine
receptors
At present, five subtypes of dopamine receptors have
been characterized (Seeman, 1987 ; Seeman and Van Tol,
1993). The human genes corresponding to these different
subtypes were cloned and assigned to chromosomes
5q35 .1 (D
"
), 11q22 .23 (D
#
), 3q13 .3 (D
$
), 11p15 .5 (D
%
),
and 4p16 .1 (D
&
), respectively (Civelli, 1995). Two D
"
-like
receptor subtypes (D
"
, D
&
) couple to the G protein G
s
and
activate adenylyl cyclase. The other receptor subtypes
belong to the D
#
-like subfamily (D
#
, D
$
, D
%
) and are
prototypic of G-protein-coupled receptors that inhibit
adenylyl cyclase and activate K+ channels.
The genomic organization of the dopamine receptors
supports the concept that they derive from the divergence
of two gene families that mainly differ in the absence or
the presence of introns in their coding sequences. The D
"
and D
&
receptor genes do not contain introns in their
coding regions, a characteristic shared with most G-
protein-coupled receptors (Dohlman et al., 1987). In
contrast, the genes encoding the D
#
-like receptors are
interrupted by introns so that the D
#
receptor coding
region contains 6 introns, the D
$
receptor coding region
5 and the D
%
receptor 3 (Monsma et al., 1989 ; Sokoloff et
al., 1990 ; Van Tol et al., 1991). The presence of introns
within the coding region of D
#
-like receptors allows the
generation of splicing variants. Indeed, the D
#
receptor
has two main variants, called D
#s
and D
#L
, which are
generated by alternative splicing of a 87 bp exon between
introns 4 and 5. Both variants share the same distribution
pattern, with the shorter form less abundantly transcribed
and both isoforms revealing the same pharmacological
profile, even if a marginal difference in the affinity of somesubstituted benzamides has been reported (Castro and
Strange, 1993 ; Malmberg et al., 1993 ; Neve et al., 1991).
Splice variants of the D
$
receptor encoding non-functional
proteins have also been identified (Fishburn et al., 1993).
200 G. Emilien et al.
A recent study which examined the anatomical dis-
tribution of D
$
receptor mRNA expression at different
levels of the human brain showed that the most abundant
D
$
mRNA expression levels were found in the islands of
Calleja and discrete cell cluster populations within the
striatum}nucleus accumbens region (Suzuki et al., 1998).
High levels were also evident within the dentate gyrus
and striate cortex. This study also confirmed previous
reports that D
#
receptors may play a significant role in
limbic-related functions such as cognition and emotion.
The study of the physiological functions mediated by
the D
%
receptor indicate that several mutations and
polymorphisms change the D
%
receptor structure (Seeman
and Van Tol, 1994 ; Van Tol and Seeman, 1995). These
include an insertion-deletion of a 4-amino-acid sequence
immediately upstream from transmembrane 1 (TM1), a
frame-shift mutation in TM2, a single nucleotide sub-
stitution that converts Val194 into Gly194, and a variable
number of a 48-bp tandem repeat (VNTR) in the third
cytoplasmic loop (Lichter et al., 1993 ; Van Tol et al.,
1992). In addition, it has been observed that translation
initiation of the D
%
receptor can occur within the TM
region and individuals with 2–10 tandem repeat units
(called D
%.#
, D
%.$
, D
%.%
, etc.) have been described. These
different units can be found at various positions and
frequencies within the VNTR and to date more than 27
polymorphic variants of the D
%
receptor have been noted
(Asghari et al., 1994 ; Lichter et al., 1993 ; Van Tol et al.,
1992). At the amino acid level, this has resulted in the
identification of at least 20 different polymorphic forms of
the D
%
receptor with respect to this sequence. Unlike the
unstable trinucleotide repeat polymorphisms that underlie
various genetic disorders and confer genetic anticipation,
this VNTR is transmitted in a normal Mendelian pattern.
This polymorphism appears to be primate-specific and has
not been observed in rodents. The 4-repeat form (D
%.%
) is
predominant in the human population (60%). The D
%.(
variant is present in 14% of the population and the D
%.#
in 10% (Seeman and Van Tol, 1994 ; Van Tol and Seeman,
1995). The functional significance of these variants has
not been elucidated. They display a slightly different
affinity for the neuroleptic clozapine, but none of them has
been related to an increased incidence of schizophrenia
(Seeman and Van Tol, 1994 ; Van Tol and Seeman, 1995).
2.2. Symptoms of schizophrenia
Schizophrenia tends to present two clusters of symptoms,
positive and negative. Positive symptoms include de-
lusions, hallucinations, excitement, grandiosity, sus-
piciousness and hostility. Negative symptoms include
inability to focus on relevant issues, paucity in speech,
distractibility, emotional flattening, lack of spontaneity,
and stereotyped thinking. Patients often manifest one
cluster more than another, though relative predominance
of the two frequently changes over time. This two-
syndrome concept of schizophrenia with Type 1 patients
having mostly positive symptoms and Type 2 negative
symptoms is nowwidely regarded as being very simplistic
(Crow, 1980 ; McGlashan and Fenton, 1992). Indeed,
recent data from factor analyses suggests that at least
three orthogonal factors underlie the symptoms of
schizophrenia which can be described as follows : psycho-
motor impoverishment (mostly negative symptoms),
disorganization (thought disorder), and reality distortion
(delusions and hallucinations) (Lenzenweger andDworkin,
1996 ; Rowe and Shean, 1997).
Cognitive deficits in abstract reasoning and attention,
together with structural brain abnormalities such as
enlargement of the ventricular system and reductions in
cortical (prefrontal) and mesotemporal volume are also
common findings (Hirsch and Weinberger, 1994). Dis-
orders of attention in schizophrenia were reported as
early as 1919 by Kraepelin who noted ‘a certain
unsteadiness of attention ’ as well as ‘ a rigid attachment of
attention ’. Subsequent studies suggest that schizophrenic
patients appear incapable of focusing their attention
(Goldberg and Gold, 1995). The difficulties of schizo-
phrenic patients are characterized by difficulties in main-
taining attention to relevant information while dis-
regarding unimportant material. Furthermore, the speed
with which they process information is compromised. A
potential approach to clarifying such constructs is through
neuropsychological indices and cognitive assessment.
To further understand and categorize the symptoms of
schizophrenia, several symptom-rating scales, such as the
Positive and Negative Syndrome Scale (PANSS) have
been developed (Kay et al., 1987). PANSS consists of 30
items (7 items for positive subscale, 7 items for negative
subscale and 16 items for general psychopathology
subscale) with strict criteria for definition and evaluation.
PANSS is currently widely used in clinical and therapeutic
trials in its three-dimensional form. Recently, PANSS has
been revised and 5 subscales have been proposed
(negative, positive, excited, depressive and cognitive
factors) which need to be validated (Kay and Sevy, 1990).
Due to the various cognitive difficulties encountered by
schizophrenic patients, it has become important to include
a cognitive subscale in PANSS.
2.3. Implications of dopamine receptors
In recent years, a modified dopamine hypothesis of
schizophrenia has been introduced. While the standard
dopamine hypothesis attributes schizophrenic psycho-
pathology to elevated dopamine levels, several authors
201Dopamine receptors and schizophrenia
have proposed that negative symptoms in schizophrenia
are due to a decrease in dopaminergic activity (Davis et
al., 1991 ; Heritch, 1990). It was suggested that negative
symptoms are caused by low prefrontal dopamine
activity, which leads to excessive dopamine activity in
mesolimbic dopaminergic neurons, and may eventually
lead to positive symptoms (Davis et al., 1991). It was
argued that schizophrenia patients suffer from a dimin-
ished ‘ tonic ’ striatal dopamine release, consecutive up-
regulation of striatal postsynaptic dopamine receptors
and, hence, increased responses to ‘phasic ’ striatal dopa-
minergic activation due to environmental stress (Grace,
1991). This would result in both low dopaminergic-
negative symptoms and stress-related hyperdopamin-
ergic-positive symptoms. The hypothesis that negative
symptoms in schizophrenia are due to a decrease in
dopaminergic activity, rather than an increase, was also
supported by various other authors (Crow, 1980 ; Davis et
al., 1991 ; Grace, 1991).
A study in which personality traits (Karolinska scales of
personality) and D
#
receptor density (PET) in the putamen
were measured in normal subjects indicated that the
density of D
#
receptors strongly correlated with a
detached personality (rflfi0–68, p! 0–001) (Farde et al.,
1997). It was pointed out that detachment can encompass
social isolation, indifference to other individuals and lack
of intimate friendships, traits which are included among
the category of ‘negative symptoms ’ that commonly
characterize patients with schizophrenia. It was proposed
that D
#
receptor density may be a useful neurochemical
measure for relating the genetic endowment to human
personality traits.Anatomical substrates for the clinical efficacy of D
#
receptor antagonism in improving positive symptoms,
including auditory hallucinations, in schizophrenia were
investigated in tissues obtained post-mortem from
schizophrenics (Goldsmith et al., 1997). A modular
organization of D
#
receptors unique to the temporal lobe
was reported ; the dense bands of D
#
receptors showed
highest frequency in auditory and speech association
cortices (Brodmann areas 22, 39, 42) and auditory–visual
association areas (Brodmann areas 20, 37). It was hyp-
othesized that blockade of D
#
receptors in auditory and
auditory–visual association cortices is a likely mechanism
for the clinical efficacy of D
#
antagonists in reducing
hallucinations. An in vivo evidence for a dysregulation of
striatal dopamine release in schizophrenia has also been
demonstrated (Abi-Dargham et al., 1998). Patients with
schizophrenia exhibited a significantly larger reduction in
D
#
receptor availability following acute amphetamine
challenge than the comparison group. While the mech-
anism, specificity and significance of this increased
dopaminergic response in schizophrenia remains to be
clarified, this observation is consistent in documenting an
excessive neurochemical response and provides evidence
to support the hypothesis of a dysregulation of central
dopamine transmission in schizophrenia.
The examination of the expression of the transcripts
encoding the dopamine receptors in cortical and striatal
regions of post-mortem schizophrenic brains shows a
focal abnormality of dopaminergic circuitry in the schizo-
phrenic prefrontal cortex (PFC) suggesting that prefrontal
cortical dopaminergic activity is diminished in this illness
(Meadow-Woodruff et al., 1997). These data indicate that
cortical dopaminergic neurotransmission may be dis-
rupted in schizophrenia at the level of receptor expression.
These changes were restricted to the D
$
and D
%
receptors
and localized to Brodmann area 11 (orbitofrontal cortex).
It also appears that the down-regulation of D
$
and D
%
mRNA in Brodmann area 11 is probably related to
schizophrenia itself and not to the medication. Schmauss
et al. (1993) found a selective loss of D
$
mRNA expression
in the parietal and motor cortices of post-mortem,
schizophrenic brains. This phenomena may be due to
either the course of the disease or the therapy given to the
patient during the course of the disease.
Of the other dopamine receptor gene products, D
%
demanded immediate attention because of its high affinity
for clozapine, its remarkable allelic variability and its
cortical and limbic distribution (Seeman & Van Tol, 1994 ;
Van Tol et al., 1991 ; Van Tol & Seeman, 1995). Seeman et
al. (1993) measured indirectly the density of D
%
receptor
binding sites in striatal homogenates using two ligands
[$H]emonapride which detects D
#
, D
$
and D
%
receptors
and [$H]raclopride in the presence of guanine nucleotide
which detects D
#
and D
$
receptors. The difference in
binding between the two ligands was used as an estimate
of D
%
receptor density. A 6-fold increase in D
%
receptors
in the basal ganglia from deceased schizophrenic patients
was noted (Seeman et al., 1993). However, Reynolds and
Mason (1994) using a competitive (rather than sub-
tractive) method were unable to confirm the existence of
D
%
receptor in human striatal tissue from schizophrenic
patients. Regardless of the methodology employed, it
remains to be established whether elevations in D
%
-like
receptors in schizophrenia are partially or wholly an effect
of neuroleptic treatment. Overall, the importance of D
%
receptor involvement in schizophrenia remains unclear.
The fact that different dopamine ligands and selective
antagonists are currently available from many sources
means that further investigations with these tools will
probably help to further clarify the role of these dopamine
receptors in schizophrenia (Lie! geois et al., 1988). Never-
theless, from the recent observation that a selective D
%
receptor antagonist (L745,870) was ineffective as an
antipsychotic for the treatment of neuroleptic-responsive
202 G. Emilien et al.
in-patients with acute schizophrenia may suggest the
limited implications of D
%
receptor in schizophrenia
(Kramer et al., 1997). In fact, in this 4-wk, placebo-
controlled, double-blind study the patients became worse
after treatment with L745,870 and a greater percentage
receiving the drug compared to placebo discontinued the
treatment due to insufficient therapeutic response (32 vs.
16%).
2.4. Dopamine receptors and cognitive function
Preclinical studies provide some useful hints about the
implications of dopamine receptors in cognitive processes.
Dopamine plays an important role in both working and
long-term memory (LTM) (Goldman-Rakic, 1995). In
LTM, dopamine is involved specifically in the mechanisms
of reinforcement (Schultz et al., 1993). A study investi-
gating the role of dopamine on short-term memory
(STM) and LTM in rats with cannulae implanted in the
dorsal CA1 region of the hippocampus or the entorrhinal
cortex, trained in one-trial step-down inhibitory avoid-
ance, and tested 1–5 or 24 h later, indicated that STM and
LTM are differentially modulated by D
"
receptors in the
CA1 and entorrhinal cortex (Izquierdo et al., 1998). The
D
"
antagonist SCH23390 (0–5 lg) enhanced STM with-
out affecting LTM when implanted in CA1, and blocked
LTM without affecting STM when implanted in the
entorrhinal cortex. D
"
receptors in the PFC are involved in
working memory processes other than just the short-term
active retention of information and also provide direct
evidence for dopamine modulation of limbic-PFC circuits
during behaviour (Seamans et al., 1998). Further animal
studies also suggest that supranormal D
"
receptor stimu-
lation in the PFC is sufficient to impair PFC working
memory function and that the impairment may be
reversed by pretreatment with a D
"
receptor antagonist,
SCH23390, consistent with drug actions at D
"
receptors
(Zahrt et al., 1997). Since D
"
receptor which is highly
expressed in the PFC has been implicated in the control of
workingmemory, andmemory dysfunction is a prominent
feature of schizophrenia, it is therefore, important to
understand how dopamine affects cognition in schizo-
phrenia. The hypothesis that the dependence of working
memory on D
"
receptor activity can be described as an
inverted U-shaped function in which there is an optimal
range of dopamine concentration and cortical D
"
receptor
for normal cognitive performance was suggested (Lidow
et al., 1998). Too little or too much D
"
receptor activation
leads to a deficient operation of the neural systems
necessary for working memory thus resulting in deficient
cognitive performance. Therefore, depending on how
antipsychotics regulate cortical D
"
receptors in relation to
the optimal dose range, they may either have a beneficial,
detrimental or absence of effect on cognitive function in
schizophrenia (Lidow et al., 1998).
Regarding the role of other dopamine receptors in
cognition, it was noted that the D
$
receptor has a
restricted expression in brain limbic areas, associated with
cognitive functions and motivated behaviour (Griffon
et al., 1995). A recent study in the rat showed that the
D
$
agonist R(›)-7-hydroxy-N,N-di-n-propyl-2-amino-
tetralin (7-OH-DPAT, 0–1–100 lg}kg, s.c.) administered
before training, immediately after training, and before
retention significantly shortened step-down latency of
passive avoidance learning, indicatingthe amnesic effects
of 7-OH-DPAT (Ukai et al., 1997). Neither the D
"
receptor antagonist SCH23390 (2–5, 5 lg}kg, i.p.) nor the
D
#
receptor antagonist sulpiride (10, 100 mg}kg, i.p.)
markedly influenced the 7-OH-DPAT-induced amnesia. It
was hypothesized that the amnesic affects of the D
$
receptor agonist 7-OH-DPAT are not mediated via D
"
or
D
#
receptors in the brain.
A recent study using positron emission tomography
(PET) to examine the distribution of D
"
and D
#
receptors
in brains of drug-naive and drug-free schizophrenic
patients reported that binding of radioligand to D
"
receptor was reduced in the PFC of schizophrenics (Okubo
et al., 1997). This decrease was related to the severity of
the negative symptoms such as emotional withdrawal and
to poor performance in the WCST. It was suggested that
dysfunction of D
"
receptor signalling in the PFC may
contribute to the negative symptoms and cognitive
deficits seen in schizophrenia. In another clinical investi-
gation, the association between dopamine activity with
cognitive and motor impairment was studied in healthy
volunteers (Volkow et al., 1998). All subjects underwent a
neuropsychological test battery. Correlations between D
#
receptors and neuropsychological test performance were
strongest for the motor task (finger tapping test) and were
also significant for most tasks involving frontal brain
regions, including measures of abstraction and mental
flexibility (WCST) and attention and response inhibition
(Stroop Colour-Word Test, interference score). These
relationships remained significant after control for age
effects. It was concluded that dopamine activity may
influence motor and cognitive performance irrespective of
age and that interventions that enhance dopamine activity
may improve cognitive performance and quality of life in
individuals such as elderly subjects or schizophrenic
patients.
2.5. Issues and criticisms
Dopamine may be critical for working memory and other
cognitive functions (Gabrielli, 1998). Associations be-
tween working memory, reasoning, and strategic memory
203Dopamine receptors and schizophrenia
occur in many clinical studies and all three capacities
appear to depend on dopaminergic fronto-striatal
systems. The extent to which these associations reflect
shared vs. neighbouring processes in both normal indi-
viduals and schizophrenic patients needs to be assessed.
Another important question of whether reductions in one
capacity are causal or merely correlated with changes in
other capacities remains also to be determined.
3. Molecular genetics of schizophrenia
Two methods often employed in molecular genetic
studies are the linkage and association approaches.
Currently, five chromosomal regions 5q, 6p, 8p, 13q and
22q have been intensely investigated by the linkage
analysis methodology.
3.1. Linkage studies
Linkage in molecular genetic studies refers to the fact that
a gene is located near a specific DNA marker on the
chromosome. Linkage analyses are based on family and
pedigree studies in which families with ill members are
tested by genotyping the variable sequences at DNA
marker loci, which are unrelated to the disease. Linkage
analyses determine whether ill relatives have inherited the
same marker allele more often than expected by chance.
Genetic linkage to illness is present when ill relatives share
alleles at one or several genetic loci ; however, at any
linked locus, the shared allele need not be the same in
different families. Detectability depends on the magnitude
of the risk imparted by a given locus and the size of the
sample of families studied. Genetic markers are poly-
morphic, with multiple genetic variants and with known
location on the genome. Maximum lod scores (MLS) are
a measure of the strength of an association with values
between 1–9 and 3–3 being suggestive of linkage and
values above 3.3 being evidence for linkage. A lod score
of fi2 or lower excludes linkage.
Linkage studies have identified several chromosomal
regions as candidates for containing a schizophrenia
susceptibility locus. The strongest support is for 6p24–22,
where at least four groups have reported results sugges-
tive of positive linkage (Antonarakis et al., 1995 ; Moises
et al., 1995 ; Schwab et al., 1995 ; Straub et al., 1995 ; Wang
et al., 1995) (see Table 1). A linkage analysis was
performed in 186 families (567 individuals classified as
affected which included typical schizophrenia, simple
schizophrenia, schizoaffective disorder, schizotypal per-
sonality disorder, schizophreniform disorder, delusional
disorder, atypical psychosis and mood incongruent psy-
chotic affective disorder) and under a model with partially
dominant inheritance, moderately broad disease definition
and assuming locus homogeneity, a lod score of 3–2 was
reported for D6S260 on chromosome 6p23 (Wang et al.,
1995). After combinations of two marker loci simul-
taneously, a maximum multipoint lod score of 3–9 was
noted, after allowing for locus heterogeneity at a map
position of 5 cM distal to D6S260, in an analysis using
this marker and F13A1. These positive results do not
include HLA region (6p21\3) which was shown pre-
viously to be possibly linked or associated with schizo-
phrenia (Kendler and Diehl, 1993). In a large study (14
research groups using 14 markers in a new sample of
403–567 pedigrees per marker), no evidence for linkage
on chromosome 3 was noted (Schizophrenia Linkage
Collaborative Group for Chromosomes 3, 6 and 8, 1996).
However, the results, although inconclusive, were
suggestive of linkage for chromosome 6 [MLS of 2–19
(new sample) and 2–68 (combined sample)] and chromo-
some 8 [MLS of 2–22 (new sample) and 3–06 (combined
sample)]. Although others have failed to find linkage to
schizophrenia in regions of chromosome 6 (Daniels et al.,
1997 ; Gurling et al., 1995 ; Mowry et al., 1995 ; Riley et
al., 1996), these studies do not necessarily constitute a
refutation as only 15–30% of the schizophrenia families in
the positive linkage studies were estimated to carry a
vulnerability locus (Schizophrenia Linkage Collaborative
Group for Chromosomes 3, 6 and 8, 1996).
Another chromosome of strong interest is 22q, where
several groups have reported support for linkage [Coon et
al., 1994 ; Gill et al., 1996 ; Moises et al., 1995 ; Pulver et
al., 1994 ; Schizophrenia Collaborative Linkage Group
(Chromosome 22), 1996]. Other regions for which there is
some suggestions for linkage include 8p (Kendler et al.,
1996 ; Pulver et al., 1995 ; Schizophrenia Linkage Collab-
orative Group for Chromosome 3, 6 and 9, 1996), 13q (Lin
et al., 1995 ; Pulver et al., 1995) and 5q (Schwab et al.,
1997 ; Straub et al., 1997).
3.2. Genome scanning studies
Genome scanning has become a sensible strategy for
detecting biologically important genetic derangements.
The major strength of genome scanning is that it covers
all possible biological mechanisms of inherited disease,
including possibilities that the investigator has not
thought of. If a mutation is present, the biology of a
disease can be elucidated rapidly. In a large study,
genome-wide scanning for a schizophrenia locus was
carried out in three phases (Moises et al., 1995). In the first
stage, only five family lines with a total of 37 affected
individuals from Iceland were used. Screening with 413
markers produced 26 potential loci, of which 10 were
2
0
4
G
.E
m
ilien
et
al.Table 1. Some linkage studies of chromosome implicated in schizophrenia.
Chr. MLS
No. of
families
sample Markers Comments Ref.
5q 1–8 44 D5S642, 666, 393, 399, 500,
658, 438, 210, 434,IL-9,
CSF1R
In sample I (14 families), a lod score of 1–8 by two-point lod score analysis was noted
for the marker IL-9. In sample II (44 families), a lod score of 1–8 around the marker
D5S399 was obtained by multipoint analysis.
Schwab et al.
(1997)
3–35 265 (Irish) D5S815, 1467, 421, 489,
2055, 818, 804, 642, 666,
393, 399, 500, 658, IL-9
Strongest evidence for linkage occurred under the narrow phenotypic definition and
recessive genetic model, with a peak at marker D5S804 (pfl 0–0002).
Straub et al.
(1997)
6p 3–51 265 (Irish) D6S477, 296, 277, 470, 443,
259, 260, 274, 285, 422,
299, 105, 276, 273, 291,
F13A1
MLS was 3–51 (pfl 0–0002), assuming locus heterogeneity, with D6S296. C test also
supported linkage with the strongest results obtained with D6S296 (pfl 0–00001),
D6S274 (pfl 0–004) and D6S285 (pfl 0–006).
Straub et al.
(1995)
1–17 57 D6S477, 296, 277, 259, 260,
285, 276, 1011
The most significant results were obtained for D6S296 using the recessive model,
giving a lod score of 1–17. However, this linkage does not extend to the HLA region.
Antonarakis et al.
(1995)
0–05–1–0 23 D6S296, 285 No evidence of linkage. Failed to confirm Straub et al. study. Gurling et al.
(1995)
0–23–0–34 45 D6S296, 470, 259, 285 No evidence of linkage. Slightly positive lod scores occurred near D6S259. Mowry et al.
(1995)
1–0–1–1 211 GATA23E10, GAAT12F07,
SCAI, D6S477, 309, 277,
296, 470, 1058, 259, 469,
260, 289, 1676, 288, 285,
422
Despite the use of 17 polymorphic markers spanning a 37 cM region, neither two-point
nor multipoint non-parametric analyses reached significance at a level less than 0–01
for any markers examined and lod score analyses were not suggestive of linkage.
Therefore, there was no evidence of linkage to chromosome 6.
Daniels et al.
(1997)
0–0–261 19
(African*)
D6S296, 277, 470, 259, 285 No evidence to support linkage in this region of chromosome 6. Riley et al.
(1996)
2
0
5
D
opam
ine
receptors
and
schizophrenia
2–19, 2–68 403–567 D3S1293, D3S1283,
D3S1266, D3S1298,
D6S296, D6S277, D6S470,
D6S259, D6S285, D8S261,
D8S258, D8S133, D8S136,
D8S283
Fourteen collaborative groups studying chromosomes 3, 6 and 8 in independent
samples. No evidence for linkage on chromosome 3. Results were interpreted as
inconclusive but suggestive of linkage in the latter two regions.
SLCG (1996)
0–192 86 D6S309, 296, 470, 259, 260,
285, 461, 276, 291
No evidence for linkage}MLS of 0–192 was for D6S309. Daniels et al.
(1997)
3–2 186 D6S277, 259, 260, 285, 273,
F13A1
A lod score of 3–2 was obtained for D6S260 on chromosome 6p23. A multipoint score
of 3–9 was achieved when the F13A1 and D6S260 loci were analysed, allowing for
locus heterogeneity.
Wang et al.
(1995)
8p 2–35 57 Genome-wide search}520
markers
MLS occurred using the ‘ affected only ’ models for markers D8S136 (Z
max
, dominant
fl 2–35 ; recessivefl 2–20) and for D8S133 (Z
max
, dominantfl 1–34 ; recessivefl 2–02).
Pulver et al.
(1995)
2–00, 2–52,
2–08
265 D8S552, 511, 1731, 261,
1715, 258, 282, 298, 133,
1733, 136, 1752, 1739,
137, 283
According to two-point heterogeneity lod scores, the strongest evidence for linkage
was for markers D8S1731 (2–08), D8S1715 (2–52) and D8S133 (2–08)
Kendler et al.
(1996)
2–22, 3–06 403–567 D8S261, 258, 133, 136, 283 Fourteen collaborative groups studying chromosomes 3, 6 and 8 in independent
samples.
SLCG (1996)
13q 1–62 13 D13S175, 232, 192, 120,
260, 263, 126, 119, 144,
160, 121, 122, 128, 64,
173, 285, 71, HTR2A
Under the assumption of homogeneity, most of the markers gave negative total lod
scores, although with a narrow model marker D13S119 gave a total lod score of
1–62 and marker D13S144 gave a total lod score of 1–48.
Lin et al.
(1995)
22q 2–82–1–54 39 Genome scan}240
randomly distributed
markers
Pairwise linkage analyses suggest a linkage (MLSfl 1–54) for region 22q12–q13–1.
Reanalyses, varying parameters in the dominant model, maximized the MLS to 2–82.
Pulver et al.
(1994)
na 574 D22S278 Results are suggestive of a susceptibility locus for schizophrenia near to the D22S278
locus on chromosome 22.
SCLG (1996)
MLS, Maximal lod score ; SLCG, Schizophrenia Linkage Collaborative Group ; * Southern African Bantu-speaking black population ; na, not available ; SCLG, Schizophrenia
Collaborative Linkage Group for Chromosomes 3, 6 and 8.
206 G. Emilien et al.
selected for the second stage and tested in a total of 65
families from numerous populations. This material re-
vealed some evidence for linkage to four loci. When
results from the first and second stage were combined and
analysed, the statistical evidence for linkage increased
slightly for loci on chromosomes 6p, 8p and 20. The most
stringent significance level could be reached only for 6p.
A genome scan of cortical-evoked potential (P50)
abnormality was performed in members of schizophrenia
pedigrees, without respect to affection status with
schizophrenia (Freedman et al., 1997). Linkage was
observed to be present, using a single-locus model of the
trait, to markers on chromosome 15, very close to the
gene for the a
(
-subunit of the nicotinic cholinergic
receptor. This work is an innovative approach in choosing,
as a phenotype, not the illness but a component of the
illness that can also be found in clinically unaffected first-
degree relatives. Although the relationship of P50 and the
chromosome 15 region is indicated by the results of this
study, a relationship of the genetic finding to schizo-
phrenia and to the nicotinic cholinergic receptor is only
suggested.
Recent study on genome scan of schizophrenia does
not support the hypothesis that a single gene causes a
large increase in the risk of schizophrenia (Levinson et al.,
1998). A genome scan of 43 schizophrenia pedigrees
including 126 patients with schizophrenia-related psy-
choses did not show any genome-wide, statistically
significant or suggestive linkage result. However, nom-
inally significant results were observed in five regions.
There were p values less than 0–01 at chromosomes 2q
and 10q, and there were p values less than 0–05 at
chromosomes 4q, 9q and 11q. This study is the largest
schizophrenia genome scan published to date and the
results failed to produce significant evidence for linkage.
However, a major gene could exist in certain populations,
in a chromosomal region not well covered in the map used
in this study or in an undetermined subgroup of families.
Detection of these genetic effects could require larger
samples than has been currently used (e.g. 500–1000
pedigrees). At this time there are no strong candidate
genes for schizophrenia within the regions identified by
linkage studies. Contradicting results in this area may
often simply be due to the possible genetic heterogeneity
of the sample selected for study. Controlling for the
effects of medication usage and other factors such as sex
and age, all present significant challenges to the in-
vestigator.
3.3. Association studies
Association studies attempt to determine whether a
genetic variant is more common among affected than
among non-affected individuals. Therefore, association
studies are case-control studies. In association, a particular
allele or mutation in the candidate gene is found more
frequently in patients than controls. Association studies
have also their limitations. Given a gene with a major or
a modest effect size, genetic markers have substantially
less coverage in association studies than in linkage studies.
Associations between schizophrenia and the
Ser311Cys polymorphism in exon 7 of the D
#
receptor
gene as well as a polymorphismSer9Gly in exon 1 of the
D
$
receptor gene have been reported (Crocq et al., 1992).
However, recent investigation by a European multicentre
association study of schizophrenia suggest that there is no
evidence for allelic association between schizophrenia and
the Cys311 variant of the D
#
receptor gene (Spurlock et
al., 1998). Therefore, it appears that the rare Cys311
variant in exon 7 of the D
#
receptor gene does not play a
role in the pathogenesis of schizophrenia in European
populations. This finding was not confirmed in Japanese
population either (Fujiwara et al., 1997).
Evidence of an association between schizophrenia and
homozygosity for a Ser to Gly polymorphism in exon 1
of the dopamine D
$
receptor gene has been reported
(Crocq et al., 1992). This polymorphism creates a BalI
restriction endonuclease site and brings about an amino-
acid change (Gly-Ser) in the N-terminal extracellular
domain of the receptor. Several studies have subsequently
confirmed this finding in independent samples and using
a family-based design (Asherton et al., 1996 ; Mant et al.,
1994 ; Williams et al., 1998). A number of negative studies
have also been reported (Durany et al., 1996). However,
a meta-analysis of all available results, now comprising
over 5000 individuals showed a small (ORfl 1–23) but
significant (pfl 0–0002) association between homo-
zygosity and schizophrenia, which is unlikely to be the
result of publication bias, has also been reported (Williams
et al., 1998). We have also screened all six exons that
make up the coding region of the gene using single-
stranded conformational polymorphism analysis (SSCP)
(Asherton et al., 1996). No other mutations were found
that altered protein structure in a total of 36 schizo-
phrenics and the same number of controls. We are
currently following up these results with cell culture
studies on the functional significance of the D
$
poly-
morphism and by sequencing the D
$
promoter and
screening it for polymorphisms. A second meta-analysis
(29 independent samples, from 24 different association
studies so far published ; 2619 schizophrenic patients and
2517 controls) also reported an excess of homozygosity
and 1–1 genotype in schizophrenic patients in African and
Caucasian groups (p! 0–05) (Dubertret et al., 1998).
Clearly more work is needed to establish the relevance of
homozygosity of the D
$
receptor to schizophrenia.
2
0
7
D
opam
ine
receptors
and
schizophrenia
Table 2. Examples of some association studies
Populations (sample size) Hypothesis and method Results Ref.
Schizophrenic Spanish patients, ICD-10 F20 (107)
Healthy matched controls (100)
To test an association between D
$
receptor gene
mutation and the liability to develop schizophrenia
by comparing allele frequencies and genotype
distribution in patients and controls. For genotyping,
genomic DNA was extracted from whole blood
of each individual. PCR amplification of DNA was
conducted with primers flanking the gene region
containing the polymorphism of interest.
No statistically significant differences between the
patients and control group were detected with
respect to either allele frequencies or genotype
distribution. However, if not corrected for
multiple testing a correlation was noted between
homozygosity and early age of onset of
schizophrenia and between A1 allele frequency
and disorganized and undifferentiated
schizophrenia.
Durany et al.
(1996)
Schizophrenia DNA samples}DSM-III-R (51 from
living patients, 7 from post-mortem tissues of
patients who died with schizophrenia)
Controls (296)
To test for possible abnormalities in the coding
region of the genomic DNA sequence for the
dopamine D
%
receptor in control and schizophrenia
tissues. The genomic DNA was extracted from
blood samples and in some cases from post-
mortem brain tissues. Genomic DNA was
amplified in vitro by the PCR using Taq
polymerase.
Twenty-three out of 183 control blacks (12–6%) and
3 out of 24 (12–5%) schizophrenic blacks revealed a
replacement of T by G, suggesting a substitution
of valine by glycine at amino-acid position 194
(variant D
%
Gly194). The identical prevalence in
the two groups indicate that the variant is not
associated with schizophrenia. However, none of
the caucasians in the study (113 controls, 34
schizophrenics) revealed the D
%
Gly194 variant.
Seeman et al.
(1994)
Schizophrenic patients, DSM-III-R (78 unrelated
schizophrenic individuals were selected for an
initial detailed examination of D5 gene)
To determine whether mutations in the D
&
receptor
gene are associated with schizophrenia, the gene
(78 patients}156 D
&
alleles) was first examined to
identify sequence variations affecting protein
structure or expression. Sequence changes of likely
functional significance then were tested for an
association with disease via case-control and
family-based analyses.
Five different sequence changes (C335X, N351D,
A269V, S453C, P330Q) that would result in
protein alterations were identified. However, no
statistically significant associations were noted
with schizophrenia or other neuropsychiatric
diseases. There also were no significant
associations between any one measure of
neuropsychological function. However, a post-hoc
analysis of combined measures of frontal lobe
function hinted that heterozygotes for C335X
may have a vulnerability to mild impairment.
Sobell et al.
(1995)
115 Japanese patients and controls (52 patients
met the ICD-10 and DSM-III-R criteria for
schizophrenia)
Using four loci (D2, D3, D4 and DAT) as candidate
genes, the association between these markers and
neuropsychiatric diseases was tested. Genomic
DNA was isolated from lymphocytes with a DNA
extractor WB kit (Wako, Osaka) and analysed by
PCR with oligonucleotide primers specific for part
of each dopamine receptor or DAT sequence.
The frequency of each variant was not significantly
greater in the patient group than in the control
group, and there was no evidence for an
association of each variant with subtypes of
schizophrenia and reactivity to pharmacotherapy
with dopamine antagonists.
Fujiwara et al.
(1997)
PCR, polymerase chain reaction ; DAT, dopaminergic transporter.
208 G. Emilien et al.
The implication of D
%
receptor in schizophrenia has
also been investigated (Seeman et al., 1994). This study
led to the identification of a single base substitution
occurring in exon 3 of the dopamine D
%
receptor gene of
12–5% of those of African descent in both control and
schizophrenic individuals. This substitution suggests a
replacement of Val by Gly at amino-acid position 194 and
this variant was termedD
%
Gly194. Because the prevalence
of the D
%
Gly194 was the same in controls and in
schizophrenics, this variant is probably not associated
with schizophrenia.
The role of the D
&
gene as a candidate for involvement
in schizophrenia was studied (Sobell et al., 1995). The
gene was examined in a group of unrelated schizophrenic
patients to identify sequence variations affecting protein
structure or expression. Sequence changes of likely
functional significance were then tested for an association
with schizophrenia via case-control and family-based
analyses. No significant associations with schizophrenia
were detected. Considering results of neuropsychological
tests, among the heterozygotes for the C335X allele,
neuropsychological testing revealed a high rate of poor
performance on tests sensitive to the frontal lobe
impairment, but no individual measure was associated at
a statistically significant level with heterozygosity for the
variant allele. However, a post-hoc analysis of combined
measures of frontal lobe function (the ControlledOral
Word Association Test and the Retroactive Interference
Index from the Rey Auditory Verbal Learning Test)
revealed a trend toward greater impairment among
C335X heterozygotes. However, the data must be
interpreted with caution given the post-hoc analysis and
the small size of the sample examined.
3.4. Criticisms and perspectives
Whether an alteration of dopaminergic function exists in
schizophrenia or not, it is tempting to speculate that a
genetic abnormality of dopamine receptors, either struc-
tural or affecting gene expression, might play a causative
role in the pathogenesis of schizophrenia. To date, studies
published have not found any unequivocal association of
schizophrenia with D
"
–D
&
receptor genes, although the
data on D
$
look promising (Kalsi et al., 1995 ; Liu et al.,
1995 ; Yang et al., 1993). Linkage strategy has not
produced replicable results of schizophrenia. Increased
understanding of complex inheritance has led to an
appreciation of the frequency of false positive results and
of the large sample size required to detect genes of small
effect (Hauser et al., 1996 ; Kruglyak and Lander, 1995).
Labelling patients as ‘unaffected ’ when they are affected
can reduce the apparent penetrance, thus a larger sample
size is required to obtain significant results. Conversely,
labelling patients as ‘ affected ’ when they are unaffected
can mask the presence of linkage because it appears that
there is a recombination when there is none. This makes
detection of linkage more difficult and specific localization
of susceptibility genes almost impossible. In the light of
the numerous molecular genetic studies, many geneticists
favour an oligogenic hypothesis, according to which 2 or
3 genes together lead to a predisposition to schizophrenia
(Owen and McGuffin, 1993). Their effects may be
dependent on interaction with physical and psychosocial
environmental factors. Whatever the explanation, the
identification of a genetic association with schizophrenia
has several consequences. It provides a new and much
needed clue that may help unravel the pathophysiology
of the disease. Genetic variation may also identify
prognostic or therapeutic subgroups of patients. These
results also suggest a research strategy for an integrative
approach to the study of schizophrenia integrating
neuropsychiatric, genetic and neuropsychological studies
in the search to establish the origins, nature and treatment
of schizophrenia.
4. Cognitive impairment in schizophrenia
Cognitive impairment in schizophrenia is mostly defined
in terms of performance deficits on neuropsychological
tests (see Table 3). This deficit is present at the onset of the
illness, persists for most of the patient’s life without
periods of spontaneous remission and may precede the
development of psychotic symptoms (Hoff et al., 1992). In
terms of social adjustment, neuropsychological test
deficits reflect inabilities to carry out everyday tasks and
obtain employment. For many elderly schizophrenics,
cognitive deficits may lead to early admission in nursing
homes or long-term psychiatric institutions. Neuro-
psychological tests such as those measuring attention,
executive functions and information processing speed are
only a few of the neuropsychological tests on which
schizophrenic patients perform poorly. Although some
test results are more abnormal than others, cognitive
deficits in schizophrenia are generally believed to be
diffuse and heterogeneous and not limited to one discrete
function (Gold and Harvey, 1993). Interest in sex
differences in neuropsychological functions suggest that
women with schizophrenia may be less vulnerable to
particular cognitive deficits, especially those involving
verbal processing, than schizophrenic men (Goldstein et
al., 1998). Further observations on the cognitive dys-
functions of schizophrenia also point out that poor social
functioning in schizophrenia might be related to
attentional and cognitive deficits in combination with
209Dopamine receptors and schizophrenia
dysfunctional reactions to social or environmental
stressors, and further highlight the usefulness of neuro-
psychological assessment in the study of schizophrenia
(Bellack, 1992 ; Brenner et al., 1992).
In a recent study investigating which variables best
identify geriatric patients with chronic schizophrenic
illness and which variables are more typical of those who
are community residents, it was noted that in patients
with either persistent or episodic symptoms of schizo-
phrenia, the severity of cognitive impairment was a
stronger predictor of adaptive deficits than was the
severity of either positive or negative schizophrenia
symptoms (Harvey et al., 1998). The results suggest that
adaptive-functioning deficit is correlated with overall
functional status and that cognitive impairment is an
important correlate of adaptive skills. Therefore, this
study underscored the importance of cognitive func-
tioning in the outcome of schizophrenia and recom-
mended that treatment of cognitive impairment in
schizophrenia should be a focus of intervention.
4.1. Attention
Taylor et al. (1996) examined the procedure of the well-
known Stroop Colour-Word Test (colour-words written
in various colours are presented to subjects, who must
name the colour in which the words are written ; the
colours and the colour-words may be the same or
different), a test usually interpreted as showing an
individual’s ability to use selective attention. This function
is believed to be impaired in schizophrenic individuals,
showing the greater interference effects in patients,
relative to normal individuals. After correction for
generalized slowing in the schizophrenia group, it was
noted that the two groups did not differ in the degree of
interference on the incongruent condition although the
patients exhibited significantly greater facilitation on the
Stroop test compared to the normal controls (81 vs.
12 ms). It was concluded that attentional dysfunction
theories did not account sufficiently for the Stroop
performance in patients with schizophrenia.
4.2. Language and information processing
Language disturbances in schizophrenia can be grouped
into a number of broad categories, two of which are the
most important. The first may be referred to as negative
thought disorder, consisting of reduced verbosity, re-
duced syntactic complexity and increased pausing (Alpert
et al., 1994). The second dimension which may be
considered as discourse coherence disturbances includes a
subset of the classic subtypes of formal thought disorder
(e.g. tangential responses, loss of goal, derailments, non-
sequitur responses, distractible speech) as well as vague or
ambiguous word meaning references (Berenbaum and
Barch, 1995). Numerous studies have demonstrated that
these types of language disturbances tend to covary
among schizophrenic patients (Andreasen, 1979 ; Harvey
et al., 1992). A reduction in the syntax complexity of
comprehended language and expressed langauge has
been observed in schizophrenia patients, though whether
this phenomenon was a failure to acquire, or a loss after
acquisition, has not been determined (Morice and
McNicol, 1985, 1986). Negative thought disorder appears
to reflect a disturbance in generating a discourse plan
which requires selecting ideas to be expressed, retrieving
concepts or ideas from LTM, and connecting such ideas to
a logical format (Levelt, 1989). Generating a discourse
plan is thought to involve the retrieval of conceptual
information from working memory and}or LTM (Levelt,
1989). Thus, one factor that could influence the generationof a discourse plan is the ease of retrieving conceptual
information to be expressed in speech. Discourse co-
herence disturbances may reflect deficits in the ability to
maintain a discourse plan and to monitor the ongoing
content of speech. Both negative thought disorder and
disturbances in discourse coherence among schizophrenic
patients reflect deficits in specific components of language
production (Barch and Berenbaum, 1997). The presence of
structure improves the maintenance of a discourse plan
such as fewer discourse coherence disturbances but does
not improve the initiation of a discourse plan (e.g. no
decrease in negative thought disorder).
In a recent study designed to investigate what type of
cognition might be most strongly associated with thought
disorder in schizophrenia, the authors questioned whether
thought disorder resided in the semantic system or
elsewhere (Goldberg et al., 1998). All patients and normal
controls received tests of executive function and working
memory, including WCST and the Letter-Number Span
Test, a test of deployment of attentional resources, and
tests of semantic processing and language comprehension,
including the Peabody Vocabulary Test, the Speed and
Capacity of Language Processing Test, the Boston
Naming Test, and tests of semantic verbal fluency and
phonologic verbal fluency. The normal subjects were
compared with the schizophrenic patients who were rated
as high, having mild thought disorder or moderate to
severe thought disorder. While differences between the
schizophrenic subgroups and the comparison subjects
were observed on nearly all tests, a large difference in
effect size between the two schizophrenic subgroups was
apparent only in the verbal fluency difference score. Since
the fluency measure discriminated between the groups
with high and low levels of thought disorder, it was
hypothesized that clinically rated thought disorder in
2
1
0
G
.E
m
ilien
et
al.
Table 3. Some studies indicating the assessment of cognitive performance in schizophrenic patients
Populations (sample size) Duration of illness Neuroleptic treatment Results Ref.
Monozygotic pairs of
individuals discordant for
schizophrenia, DSM-III-R
(24)
Normal pairs of monozygotic
twins (7)
Mean of 10–3 yr
(range : 1–24 yr)
Nineteen of the affected twins
were receiving neuroleptic
treatment.
On the declarative memory tasks, the affected group
performed significantly worse than the discordant
unaffected group on story recall, paired associate learning
and visual recall of designs. Effortful, volitional retrieval
from the lexicon, measured by verbal fluency, was also
compromised in the affected group. Comparisons of the
normal group and unaffected group indicated that the
latter group had very mild impairments in some aspects
of episodic memory. Significant correlations between
measures of memory and global level of social and
vocational functioning within the discordant group were
also observed.
Goldberg et al.
(1993)
DSM-II-R schizophrenic (12)
Age-matched controls (12)
na Ten patients were taking
neuroleptics (5 on
risperidone, 1 on clozapine,
4 on conventional
neuroleptics) and 2 had been
off antipsychotic medication
for 1 wk.
In the Stroop task, the schizophrenic patients had slower
reaction times in all conditions. There was a significant
difference for facilitation between the schizophrenic and
control groups with the schizophrenics showing greater
facilitation (81 vs. 12 ms). Interference effects were not
significantly different.
Taylor et al.
(1996)
DSM-III-R schizophrenic (12)
Age-matched controls (7)
" 2 yr Average dosage in
chlorpromazine equivalents
was 450 mg (100–2000 mg).
Patients showed essentially perfect recall with word lists of
up to 4 words. Beyond this, performance declined, with
the steepest fall in the impaired schizophrenic patients.
Patients with schizophrenia showed a failure in DLPFC
activation only in the face of diminished performance
measures, suggesting that a full characterization of task-
related changes in DLPFC activation must consider
performance levels.
Fletcher et al.
(1998)
DSM-III-R schizophrenic (38)
Age-matched controls (39)
na All patients were receiving
medication.
One subgroup of patients has a selective deficit in verbal
memory despite normal motivation, attention and general
perceptual function. A second group of patients has
deficits in multiple aspects of cognitive function suggestive
of deficiencies in early stages of information processing.
Wexler et al.
(1998)
2
1
1
D
opam
ine
receptors
and
schizophrenia
DSM-III-R geriatric
schizophrenic patients as
follows
Chronically hospitalized
residents (97)
Nursing home residents
(37)
Acute admissions
" 19 yr
(diagnosed when
less than 45 yr
old at onset)
All patients were treated with
either typical neuroleptic or
risperidone.
Cognitive impairment was the strongest predictor of
adaptive deficits for all 3 groups The data suggest that
interventions aimed at cognitive impairment may have
an impact on overall functional status.
Harvey et al.
(1998)
DSM-IV schizophrenic (6)
(patients served as their own
control in a within-subject
design)
Mean of 12 yr
(range : 3–19 yr)
Had been receiving atypical
neuroleptic treatment for at
least 1 month (mean dose
450 mg chlorpromazine
equivalents ; range, 250–750).
The severity of positive thought disorder was inversely
correlated with activity in areas implicated in the
regulation and monitoring of speech production.
Decreased activity in these regions may contribute to the
articulation of the linguistic anomalies that characterize
positive thought disorder.
McGuire et al.
(1998)
DSM-II-R schizophrenic (23)
Normal controls (23)
Mean of 17–1 yr
(s.d.fl 8–6)
All patients were receiving
neuroleptics during study
[clozapine, 9 ; risperidone, 4 ;
others (haloperidol,
fluphenazine or loxapine),
10].
Thought disorder may be associated with semantic
processing abnormalities. In particular, patients with more
severe thought disorder may have difficulty accessing
semantic items because of disorganization of the semantic
systems or possible lack of a semantic knowledge
base.
Goldberg et al.
(1998)
DSM-III-R schizophrenic (73)
First-degree relatives (33
siblings, 28 parents)
Age-matched controls (35)
na}consecutively
admitted in-patients
No neuroleptics or
antidepressants or
benzodiazepines within the
last 2 wk or depot
neuroleptics within the last
2 months.
Performance of schizophrenic patients was worse than
those of controls in all variables of WCST (including
perseverative and non-perseverative responses). Healthy
siblings of schizophrenic probands showed more
perseverative responses than controls, but did not show
any difference with respect to the non-perseverative
responses.
Franke et al.
(1992)
DLPFC, dorsolateral region of prefrontal cortex.
212 G. Emilien et al.
schizophrenia may result from semantic processing ab-
normalities and that language disorder present in schizo-
phrenia may occupy an important place in the treatment
of cognitive processes of schizophrenic patients.
In a within-subject design study performed to further
understand the pathophysiology of thought disorder
using positron emission tomography (PET), regional
cerebral blood flow was measured while six schizophrenic
patients described a series of 12 ambiguous pictures
[drawn from the Thematic Apperception Test (Murray,
1943) and comprised grey scale whole-body repre-
sentations ofpeople in scenes whose interpretation was
ambiguous] which elicited different degrees of thought-
disordered speech (McGuire et al., 1998). The total score
for positive ‘Thought Language and Communication
Index ’ (TLCI) (Liddle, 1998) items (e.g. looseness, peculiar
word usage, peculiar sentence construction, peculiar logic
and distractibility) was used as an index of verbal
disorganization or positive thought disorder. The results
showed that the severity of positive thought disorder was
inversely correlated with activity in the inferior frontal
and cingulate cortices. The negative correlations in these
regions may reflect a failure to engage areas which
normally control the production of speech. Patho-
physiological changes in these areas might also account
for the association between positive thought disorder and
the disorganization of emotion and cognitive behaviour
in schizophrenia (Liddle and Morris, 1991).
Recent findings suggest that early age of onset in
schizophrenia may be associated with more impaired
cognitive functions (Hoff et al., 1996). The later the age of
onset, the better the performance. Earlier age-of-onset
patients have more impaired language function. Per-
formance in receptive and expressive speech is relatively
poor. The poor performance of the early onset patients
may simply reflect a deterioration of cognitive function
with disease progression ; the longer one remains psy-
chotic, the more disorganized one’s semantic network
becomes. Longitudinal studies would be needed to
address this hypothesis. A second hypothesis is that
disorganization of semantic knowledge may reflect pre-
morbid cerebral vulnerability which predisposes indi-
viduals to both disorganized thought organization and
illness manifestation at an earlier age. However, the
superior performance of late-onset patients may simply be
due to their longer previous opportunity for social
interaction. There is evidence to suggest these language
disturbances may be viewed as possible genetic vul-
nerability markers since communication impairment in the
relatives of schizophrenic patients appear to be het-
erogeneous in both form and origin (Docherty et al.,
1997). Similar to the schizophrenic patients, their parents
showed frequent instances of communication failures in
their speech characterized by structural lack of clarity,
vague references and ambiguous word meanings.
4.3. Executive function and memory
Frontal}executive impairments have been observed in
schizophrenia (Morice and Delahunty, 1996). The term
‘executive ’ refers generally to the ability to maintain or
shift a mental set, to establish goals, and to plan –
elements that can be measured by neuropsychological
tests such as WCST and the Tower of London. Executive
function includes the capacity to both devise and carry
out solutions to problems whose solutions are not
immediately obvious (e.g. problems that may require
abstract reasoning). Of the three functions gaining greater
recognition as executive functions – cognitive shift (or
flexibility), forward planning, and working memory, much
work is being performed to better understand their
implications in schizophrenia. Working memory has been
associated with complex functions ranging from mental
arithmetic, syntactic processing, comprehension and read-
ing to the acquisition of complex cognitive skills and
development of procedural skills (Baddeley et al., 1985 ;
Carlson et al., 1989 ; Logie et al., 1989). Although an
executive function test such as the WCST purportedly
measures reasoning ability, it also taps short- and long-
term memory, distractibility, sustained attention and
learning ability (Heaton, 1981). The increased per-
severative error scores by schizophrenia patients per-
forming the WCST has been of particular interest. Poor
WCST performance has been linked both to cognitive
inflexibility and to left dorsolateral PFC dysfunction
(Weinberger et al., 1986).
A study of frontal executive impairment in schizo-
phrenic patients indicated that 64–7% schizophrenia
patients and no controls were impaired on WCST with
respect to perseverative errors (cognitive flexibility) and
76–5% schizophrenia patients and no controls were
impaired on the Tower of London Test with respect to
targets achieved in minimum moves (forward planning)
(Morice and Delahunty, 1996). Further cognitive as-
sessment by different tests indicated that the schizo-
phrenics were significantly impaired compared to controls
on two tests of working memory, Alphabet Span and
Sentence Span. Using a cutoff derived from the mean
score for the controls, 65% of schizophrenia patients
proved to be impaired on Sentence Span and 94% were
impaired on one or more of the three tests of executive
functioning used. It was hypothesized that schizophrenia
represents a loss of, or a failure to acquire, the ability to
process complex information.
Memory deficits observed in schizophrenia are not
restricted to a single element of memory but strike
213Dopamine receptors and schizophrenia
different systems, such as declarative memory, procedural
memory and working memory (Goldberg et al., 1993 ;
McKenna et al., 1990 ; Tamlyn et al., 1992). McKenna et
al. (1990) and Tamlyn et al. (1992) reported significant
correlations between declarative memory deficits and
both formal thought disorders and negative symptoms,
while Goldberg et al. (1993) found that negative
symptoms also correlate with procedural memory im-
pairment. In a study investigating learning and memory
using a wide number of neuropsychological measures in
monozygotic pairs of individuals discordant for schizo-
phrenia, compared to normal pairs of monozygotic twins,
significant correlations between many measures of mem-
ory and global level of social and vocational functioning
within the discordant group were observed (Goldberg et
al., 1993).
The assessment of a graded memory task on brain
activation of schizophrenic patients is an important subject
to study. In an experiment in which schizophrenic patients
and volunteers underwent scanning while learning and
recalling word lists of variable length, it was observed that
all patients showed perfect recall with word lists of up to
4 words (Fletcher et al., 1998). Beyond this, performance
declined, with the steepest fall in the impaired schizo-
phrenic patients. As task demands increased and per-
formance deteriorated, schizophrenic patients failed to
show an increasing frontal response. The more demanding
tasks may have engaged frontally mediated strategies that
the schizophrenic patients did or could not adopt.
Therefore, the abnormal PFC activation under more
demanding conditions may reflect a motivation deficit
occurring as the task becomes too difficult for a patient. It
is also interesting to note the difference of performance
between STM and LTM. Verbal STM function has been
noted to be relatively preserved in schizophrenia, whereas
the deficit in LTM is significant (Goldberg et al., 1993 ;
Tamlyn et al., 1992). A striking observation in the study
performed by Fletcher et al. (1998) was that a region
showing impaired activation, the posterior parietal region
(Brodmann area 40) was specific to the impaired schizo-
phrenic group. Posterior parietal activations are a common
finding in functional imaging studies of memory retrieval
and the significant decrease of activation in the impaired
schizophrenic group may reflect an impairment in LTM
retrieval processes in schizophrenia (Shallice et al., 1994).
Several studies have considered recall and recognition
memory impairment in schizophrenic patients. Recall
refers to the ability