Download The GABAergic system in schizophrenia

The GABAergic system in schizophrenia
Brian Paul Blum and J. John Mann
R E V I E W A RT I C LE
International Journal of Neuropsychopharmacology (2002), 5, 159–179. Copyright # 2002 CINP
DOI : 10.1017\S1461145702002894
Columbia University College of Physicians and Surgeons, Department of Psychiatry
New York State Psychiatric Institute, Department of Neuroscience, New York, USA
Abstract
A defect in neurotransmission involving γ-amino butyric acid (GABA) in schizophrenia was first proposed in
the early 1970s. Since that time, a considerable effort has been made to find such a defect in components of
the GABAergic system. After a brief introduction focusing on historical perspectives, this paper reviews postmortem and other biological studies examining the following components of the GABAergic system in
schizophrenic subjects : the GABA biosynthetic enzyme, glutamate decarboxylase ; free GABA ; the GABA
transporter ; the GABAA, GABAB and benzodiazepine receptors ; and the catabolic enzyme GABA
transaminase. Additionally, post-mortem studies using morphology or calcium-binding protein to identify
GABAergic neurons are also reviewed. Substantial evidence argues for a defect in the GABAergic system of
the frontal cortex in schizophrenia which is limited to the parvalbumin-class of GABAergic interneurons.
Received 18 July 2001 ; Reviewed 11 November 2001 ; Revised 28 January 2002 ; Accepted 30 January 2002
Key words : CSF, GABA, post-mortem, schizophrenia.
Historical perspectives
The physiologists Ernst and Friedrich Weber (1845), Ivan
Pavlov (1885) and Wilhelm Biedermann (1887) established the concept of inhibition in the nervous system and
lead to the identification of inhibitory neurons by Cornelis
Wiersma (1933) (Florey, 1991). Roberts and Frankel
(1950), Awapara et al. (1950), and Udenfriend (1950)
reported the isolation of γ-aminobutyric acid (GABA)
from animal brain material. Working on crayfish stretch
receptors initially and later with the monosynaptic kneejerk reflex in cats, Ernst Florey reported that a Factor I had
inhibitory effects in these systems (Florey and McLennan,
1955). Factor I was later purified from beef brain and
shown to be identical to GABA (Bazemore et al., 1957).
The status of GABA as neurotransmitter was not widely
accepted until further studies in crustaceans established
that GABA was the most common inhibitory substance in
the CNS (Dudel, 1963), that peripheral inhibitory but not
excitatory neurons contained high GABA concentrations
(Kravitz et al., 1963), that inhibitory motor neurons
released GABA (Otsuka et al., 1966), and that GABA is
removed from the postsynaptic cleft by an uptake process
(Orkand and Kravitz, 1971).
GABA is the major inhibitory neurotransmitter in the
mammalian brain ; up to 30 % of cortical neurons in rats
Address for correspondence : Dr B. P. Blum, New York State
Psychiatric Institute, Department of Neuroscience, 1051 Riverside
Drive, Unit 42, New York, NY, 10032, USA.
Tel. : 212-543-6223 Fax : 212-543-6017
E-mail : bb453!columbia.edu
use GABA in neurotransmission (Bloom and Iversen,
1971). In the monkey cortex approx. 25 % of the neurons
in most regions are GABAergic ( Jones, 1990). In addition
to the neocortex, significant populations of glutamate
decarboxylase (GAD)- or GABA-immunoreactive (IR) cell
bodies or axon terminals have also been identified in
primate brain regions including the midbrain (Holstein et
al., 1986 ; Okada et al., 1971), hippocampal formation
(Schlander et al., 1987), the thalamus (Smith et al., 1987b),
the basal ganglia (Smith et al., 1987a), and the amygdala
(Sorvari et al., 1995).
GABA was first implicated in the pathophysiology of
schizophrenia by Eugene Roberts in 1972. He proposed
that a susceptibility to schizophrenia might be due to a
defect in the inhibitory GABAergic neurons control of
neural circuits governing behavioral responses. This
defect would be exacerbated under stressful conditions in
which increased monoaminergic drive would increase
disinhibitory input onto those GABA neurons, producing
abnormalities of perceptual and cognitive integration
(Roberts, 1972). Since this initial proposal, a role for
GABA in the pathophysiology of schizophrenia continues
to be formulated in the context of complex interactions
between GABA and other neurotransmitter systems.
Carlsson (1988) proposed a model of psychosis that
involves multiple neurotransmitters including a defective
GABA-mediated inhibition of glutamatergic feedback
inhibition of mesolimbic dopamine function and describes
a defect in thalamic filtering of sensory and arousal input
to the cortex (see also Carlsson et al., 2001). An
abnormality in GABAergic regulation of dopamine cell
160
B. P. Blum and J. J. Mann
burst firing has been postulated to underlie the symptoms
of schizophrenia (Grace, 1991 ; Moore et al., 1999).
Others have noted, in the direct modulation of the
dopaminergic system by GABAergic neurons, a potential
mechanism whereby an abnormality in the GABAergic
system could be involved in the dopaminergic dysfunction of schizophrenia (Carlsson, 1988 ; Fuxe et al.,
1977 ; Garbutt and van Kammen, 1983 ; Stevens et al.,
1974 ; van Kammen, 1979). Squires and Saederup (1991)
postulated that schizophrenia involved a GABAergic
predominance caused by either hyperactive GABAA
receptors or hypoactive glutamate receptors and\or
destruction of counterbalancing glutamatergic neurons by
neurotropic pathogens. This last model has received little
ongoing support.
Olney and Farber (1995) developed a model of
schizophrenia in which a state of ‘ NMDA receptor
hypofunction ’ is caused by either intrinsically hypofunctioning NMDA receptors or through excitotoxic loss of
NMDA receptor-bearing GABAergic neurons. This state
results in excessive dopaminergic input into corticolimbic
regions (also see Carlsson et al., 2001) with resultant
further hypofunctioning of the glutamatergic system
through feedback mechanisms. Several classes of compounds, including benzodiazepines (BZD), muscurinic
receptor antagonist and haloperidol, blocked NMDAinduced neurotoxicity in the posterior cingulate and
retrospenial regions of experimental animals. Loss of
GABAergic interneurons in the hippocampal formation,
possible secondary to excitotoxic injury (Benes, 1999) or
to loss of glutamatergic neurons has also been hypothesized (Deakin and Simpson, 1997). Similarly, Deutsch
et al. (2001) postulates a failure of GABAergic inhibition
of the AMPA\kainite class of glutamatergic receptor with
a resultant cascade of excitotoxic events.
A dysfunction of α7-nicotinic acetylcholine receptor
on GABA interneurons in the hippocampus (Adler et al.,
1998) or disruption of interactions between the cholinergic system and 5-HT A receptor on GABAergic inter#
neurons in the frontal cortex (Dean, 2001) has been
proposed as sites of pathophysiology in schizophrenia.
Relationships between epilepsy, schizophrenia and the
GABAergic system have been proposed (Keverne, 1999 ;
Stevens, 1999). Effects of the GABAergic system in
neuro- and in particular cortico-developmental processes
have been integrated into developmental hypotheses of
psychosis and schizophrenia. GABAergic interneurons
form the substrate for the gamma frequency oscillations
postulated to synchronize brain activity in disparate
regions of the brain and an abnormality in such may cause
psychosis (for a review see Keverne, 1999).
Although there is some evidence for a role for BZD and
valproate in the treatment of schizophrenia, GABAergic
agents have generally not been demonstrated to produce
antipsychotic effects in of themselves (Wassef et al.,
1999). In-vivo pharmacological manipulation of the
GABAergic system indicates that GABAergic function is
potentially relevant to the pathophysiology of schizophrenia. For example, blockade of GABA receptors with
picrotoxin in the prefrontal cortex of rats impairs
sensorimotor gating, an effect that is reversed by
haloperidol (Japha and Koch, 1999). Conversely, enhancement of GABAergic activity by either γ-vinylGABA (GVG) or lorazepam in baboons inhibits dopamine
transmission in the striatum as indicated by increased
[""C]raclopride binding (Dewey et al., 1992). Furthermore,
GVG treatment has been shown to increases phencyclidine-induced release of dopamine in a dose-dependent
manner in the rat prefrontal cortex but not in the striatum
(Schiffer et al., 2001). Hypofunctioning of the GABAergic
system may be responsible for the striatal dopamine
overactivity and behavioural changes noted in schizophrenic subjects (Breier et al., 1997).
The purpose of this paper is to review the data
from post-mortem and other biological studies of the
GABAergic system in schizophrenia in order to provide a
synthesis of what is known.
GAD
GAD, the rate-limiting biosynthetic enzyme of GABA,
catalyses the decarboxylation of glutamic acid to yield
GABA. Two major isoenzymes of GAD, named GAD
'&
and GAD , based on their approximate molecular weight
'(
of 65n4 and 66n6 kDa respectively, have been identified in
human brain (Bu et al., 1992). GAD is preferentially
'&
localized in axon terminals (Esclapez et al., 1994), more
tightly membrane associated and more often exists in an
inactive apoGAD form (lacking the cofactor pyridoxal
phosphate) compared with the GAD isoenzyme (Kauf'(
man et al., 1991). In rat hippocampus, most cells express
transcripts of both GAD isoenzymes (Stone et al., 1999).
It has been suggested that GAD might preferentially
'&
synthesize GABA for vesicular release and that GAD
'(
may be preferentially involved in synthesis of cytoplasmic
GABA (Erlander and Tobin 1991 ; Esclapez et al., 1994 ;
Feldblum et al., 1993).
Early efforts to detect an abnormality in the GABAergic
system focused on determining activity levels of GAD.
Seven such studies were performed on cortical tissue
homogenates from parts of the temporal and frontal
lobes : all but two of which reported no significant
difference between controls and patients with schizophrenia in these regions (see Table 1a) (Bennett et al.,
1979 ; Bird et al., 1977 ; Cross and Owen, 1979 ; Crow et
The GABAergic system in schizophrenia
161
Table 1a. GABAergic presynaptic markers in schizophrenia
Marker
Area
Finding
GAD activity
Hippocampus
BA 11, 37, 38
BA 7
BA 18
BA 34
Hippocampus
N. accumbens
Putamen
Amygdala
Hippocampus
BA 21
BA 10
Frontal cortex
H 35 %*
H 43 %*
H 39 %*
H 48n2 %*
H 44 %*
H 27 %*
H 46 %*
Frontal cortex
Amygdala
Multiple brain regions
Frontal cortex
BA 9
BA 9
BA 22
H 40 %*
H 48 %*
H 30 %*
H 70 %*
BA 22
H ns
BA 9
H 25–35 %*
Layers III–V
Volk et al. (2000)
BA 9
H 68 %*
nl6
Guidotti et al. (2000)
BA 9
BA 9
Hippocampus
BA 10
BA 24
H 54 %*
n l 15
n l 15
GAD activity
GAD activity
GAD activity
GAD activity
GAD activity
GAD activity
GAD activity
GAD mRNA
'(
GAD \β-actin
'(
immunoreactivity
GAD \β-actin
'&
immunoreactivity
GAD mRNA-positive
'(
neuron densities
Ratio of GAD
'(
mRNA to neuron-specific
enolase mRNA level
GAD protein level
'(
GAD protein level
'&
GAD -IR puncta
'&
GAD -IR puncta
'&
Comment
Author
McGeer and McGeer (1977)
Bird et al. (1977)
Perry et al. (1978)
Found H GAD
with I PMI
Crow et al. (1978)
Cross and Owen (1979)
Layer I
Layer II
Layers III–V
Spokes (1980)
Bennett et al. (1979)
Hanada et al. (1987)
Akbarian et al. (1995b)
Impagnatiello et al. (1998)
Todtenkopf and Benes (1998)
Benes et al. (2000)
* Indicates findings reported as significant by original authors (and throughout all tables).
ns, not significant.
al., 1978 ; McGeer and McGeer, 1977 ; Perry et al., 1978 ;
Spokes, 1980). One of the divergent studies found
significantly lower GAD activity in the sensory association, calcarine fissure and insular cortex in the
schizophrenic group compared with controls but not in
many other cortical and subcortical regions (McGeer and
McGeer, 1977). The second divergent study found
significantly lower GAD activity in patients with schizophrenia in all areas examined : amygdala, hippocampus,
nucleus accumbens, and putamen (Bird et al., 1977).
Whereas post-mortem interval (PMI) and age were
controlled for in this study, other possible confounds such
as comorbidity, medication and smoking history, diagnostic heterogeneity and cause-of-death effects were
not clearly addressed. Although some studies have
indicated that GAD is stable in human brain during
routine post-mortem handling (Spokes, 1979 ; Spokes et
al., 1979), Crow et al. (1978) found a significant negative
correlation between GAD activity levels and PMI.
Although GAD activity levels is significantly decreased in
brain material obtained from patients dying of protracted
illness (McGeer et al., 1971 ; McGeer and McGeer 1976 ;
Spokes 1979 ; Spokes et al., 1979), most of the studies
mentioned above did not control for this potential effect.
162
B. P. Blum and J. J. Mann
A later study of GAD activity in frontal cortex [Brodmann
area (BA) 9] and caudate from chronic schizophrenics also
revealed no abnormality (Hanada et al., 1987). This study
also reported lower GAD activity in the subgroup of
controls that had died after a prolonged terminal illness
(PTI). The control and chronic schizophrenic group were
fairly well matched for age, length of PMI and sex ratios.
Before interpreting studies examining GAD gene
expression by in-situ hybridization, it is important to note
that GAD protein levels may not match GAD mRNA
levels because of a variety of transcriptional, translational
and post-translational modifications. For example, increases in GABA levels decreases GAD activity but
'(
does not alter GAD mRNA levels nor GAD activity
'(
'&
in rats (Rimvall et al., 1993 ; Rimvall and Martin, 1994).
Also, elevation of GABA by vigabatrin treatment affects
GAD protein levels differently in various brain regions
'(
of the rat (Sheikh and Martin, 1998).
A study measured GAD mRNA in human prefrontal
'(
cortex by in-situ hybridization in order to determine if a
postulated decrease in GABA in this region was due
either to decreased gene expression or to a decrease in the
number of GABAergic cells (Akbarian et al., 1995b). Ten
schizophrenic subjects were compared to ten age, gender
and autolysis-time matched controls. Subjects with incomplete medical records, substance abuse histories or
prolonged agonal states were excluded. Fewer GAD
'(
mRNA-expressing neurons with no significant overall
loss of neurons were found in cortical layers I–V of the
schizophrenic subjects compared with controls. The
GAD mRNA levels, as measured by optical densities of
'(
film autoradiographs, were significantly lower in cortical
layers II, III, IV and V of the schizophrenic subjects
compared with controls. The authors expressed doubt
that the lower mRNA levels of the schizophrenic subjects
were secondary to neuroleptic treatment by noting that
the single neuroleptic-naive schizophrenic subject had the
lowest mRNA values.
A second study examined GAD mRNA expression in
'(
the prefrontal cortex of 10 schizophrenic subjects and 10
sex-matched controls (Volk et al., 2000). The schizophrenic group did not significantly differ from the control
group with respect to age, PMI, brain pH or storage time.
One control and four schizophrenic subjects had lifetime
diagnoses of alcohol or other substance abuse and one
control subject had a lifetime diagnosis of depressive
disorder not otherwise specified. Seven schizophrenic and
nine control subjects had sudden deaths occurring outside
of a hospital ; one schizophrenic subject was a suicide
victim. This study used in-situ hybridization followed by
counts of silver grains within neuronal soma from
randomly selected cortical sites within specific laminar
levels. Significantly (25–35 %) lower density of GAD
'(
mRNA-labelled neurons in cortical layers III–V was found
in the schizophrenic group compared with the controls.
However, mean grain density per neuron did not
significantly differ across the two groups. Additionally
comparison of chronic haloperidol- and benztropine
mesylate-treated Cynomolgus monkeys with untreated
controls indicated that this medication treatment does not
affect GAD mRNA expression.
'(
Reelin is a protein which regulates cortical cell
positioning and\or movement during development and
which appears to be expressed preferentially in
GABAergic interneurons in the adult human neocortex
(Curran and D ’Arcangelo, 1998 ; Impagnatiello et al.,
1998). Reelin protein and mRNA levels were found to be
40–50 % lower in schizophrenic subjects compared with
controls in the prefrontal cortex (BA 10 and BA 46),
temporal cortex (BA 22), hippocampus, caudate and
cerebellum (Impagnatiello et al., 1998). This study also
reported significantly ($ 70 %) lower GAD \β-actin but
'(
not GAD \β-actin IR optical densities in the schizo'&
phrenic subjects compared with controls.
A recent study measured GAD , GAD , and reelin
'&
'(
mRNA levels by quantitative reverse transcriptasepolymerase chain reaction and GAD and GAD protein
'&
'(
levels in brains from schizophrenic, bipolar and depressed
subjects. Reelin mRNA, GAD protein and mRNA were
'(
significantly lower in the prefrontal cortex and cerebellum
in schizophrenic and psychotic bipolar but not in unipolar
depressed subjects without psychosis compared to normal
controls (Guidotti et al., 2000). GAD mRNA levels did
'&
not differ across the diagnostic groups. Reelin and GAD
'(
levels were found to be unrelated to PMI or neuroleptic
treatment history.
The distribution of GAD -immunoreactive (GAD '&
'&
IR) puncta in the hippocampus was examined in a group
of 13 schizophrenic subjects and 13 age-, gender- and
PMI-matched controls (Todtenkopf and Benes, 1998). No
significant difference was found in the density of GAD '&
IR puncta in contact with pyramidal or non-pyramidal
cells or dispersed within the neuropil of the layers CA1–4.
However, a significant positive correlation was found
between the density of GAD -IR puncta in contact with
'&
pyramidal and non-pyramidal cells and neuroleptic exposure in the schizophrenic subjects. This finding and the
fact that the two neuroleptic-naive schizophrenic subjects
had the lowest density of GAD -IR puncta led the
'&
authors to speculate that schizophrenics might inherently
have lowered density of GABAergic terminals in certain
regions of the hippocampus.
The density of GAD -IR terminals in layers II–VI of
'&
the cingulate and prefrontal cortices did not differ between
the groups (Todtenkopf and Benes, 1998) ; however, the
density of GAD -IR terminals was significantly lower in
'&
The GABAergic system in schizophrenia
layers II–IV of five bipolar subjects (added in this study)
compared with the normal controls (Benes et al., 2000).
Overall, neuroleptic treatment history did not appear to
correlate with terminal densities in the schizophrenic
group. A two-dimensional counting method was used.
Three studies report lower GAD mRNA expression
'(
and two studies report lower GAD protein levels
'(
indicating that schizophrenia may be associated with less
GAD gene expression in the prefrontal cortex. Total
'(
GAD activity and GAD -immunoreactivity do not
'&
appear to be altered in schizophrenia. However, it should
be noted that the amount of GAD protein is 3- to 8-fold
'&
greater than the amount of GAD protein in most rat
'(
brain regions (Sheikh et al., 1999). One possibility is that
the abnormality in schizophrenia is restricted to GAD
'(
and is not detectable by measurement of total GAD
enzyme activity. Moreover, there is evidence that the
GAD deficit is limited to subset of neurons in the
'(
prefrontal cortex (Volk et al., 2000).
GABA concentrations
The search for a GABAergic defect in schizophrenia also
stimulated examination of GABA concentrations, both in
brain tissue (see Table 1b) and in cerebrospinal fluid (CSF).
GABA concentrations in brain tissue are unaffected by
agonal status (Spokes et al., 1979) but rise rapidly 1–2 h
after death. The rise in GABA levels may continue even
24 h post-mortem (Perry et al., 1981). Free CSF GABA
levels are unaffected by agonal status but decline
significantly with age (Perry et al., 1979 ; Spokes et al.,
1979). Most studies controlled for age and post-mortem
processing, however drug history was not consistently
controlled for and the study of Perry et al. (1979) included
a number of controls with various neurological illnesses.
Using a single-cation exchange column method lower
GABA concentrations was found in the nucleus accumbens and thalamus from schizophrenics compared with
control (Perry et al., 1979). Another study used the same
method and did not find lower GABA concentrations in
the nucleus accumbens, medial dorsal thalamus, frontal
cortex or caudate of the schizophrenic subjects compared
with controls (Perry et al., 1989). The authors suggested
that the first study was flawed by lack of anatomical
accuracy with respect to dissection of the nucleus
accumbens and the thalamus. Korpi et al. (1987) found no
effect of diagnosis on GABA levels in the nucleus
accumbens, frontal cortex, and caudate, but GABA was
37n5 % lower in the amygdala in the schizophrenic group
compared with controls. Kutay et al. (1989) found lower
GABA levels in multiple brain regions including the
amygdala, the hippocampus, frontal pole, superior temporal cortex and thalamus. Toru et al. (1988) (see GAD
163
section above) examined multiple cortical and subcortical
regions and noted lower levels only in the posterior
portion of the hippocampus of the schizophrenic group.
Spokes et al. (1980) found lower GABA concentrations in
the nucleus accumbens and the amygdala of the schizophrenic group compared with the controls. Absolute
GABA levels in this study were in agreement with a study
by Cross et al. (1979) but tended to be an order of
magnitude higher than the other studies. In contrast to the
study of Perry et al. (1979), Cross and colleagues found no
difference in GABA levels in the nucleus accumbens and
thalamus between the study groups. Ohnuma et al.
(1999), using a more specific brain region definition than
the studies of Korpi et al. (1987), Perry et al. (1989), or
Kutay et al. (1989), reported lower GABA levels in BA 9
and 10, but not 11. The PMI was longer in the control
group than in the six schizophrenic subjects and possible
group sex and medication effects were not ruled out.
Nevertheless, this remains an interesting study as it
reported a specific regional GABA level abnormality that
was paralleled by increase in GABAA receptor α subunit
"
mRNA and decrease in GAT-1 mRNA (see below).
While most studies report low GABA levels in at least
some brain regions in schizophrenia, there is no clear
consensus on the affected brain regions except for a
consistent finding of lower GABA in the amygdala (3 out
of 3 studies). Measurement of total GABA levels may be
insufficiently sensitive to consistently detect a GABAergic
defect affecting only a subpopulation of GABAergic cells.
The findings of lower GABA in the amygdala in
schizophrenia is interesting in light of recently reported
rat model in which a experimentally induced GABAergic
dysfunction in the amygdala induces changes in the
GABAergic system of the hippocampus. The subregional
distribution of these changes is similar to findings in
previous post-mortem studies of schizophrenia (Benes,
1999 ; Berretta et al., 2001).
CSF studies
Nine published studies examined GABA concentrations
in CSF. The majority reported no difference between
controls and schizophrenic patients (see Table 2). One
study found lower baseline CSF levels in a group of
schizophrenics compared with controls ; however it is not
clear if a number of schizoaffective patients (previously
mentioned in the report) were included in this group
(Sternberg, 1980). If so, perhaps depression explained the
lower CSF GABA levels. All patients were drug free for
2 wk prior to the baseline lumbar puncture and GABA
levels showed no relationship with age, sex, or degree of
psychosis. This study also reported that a trial of pimozide
increased GABA levels in the patients. Van Kammen et al.
164
B. P. Blum and J. J. Mann
Table 1b. Presynaptic markers
Marker
Area
Finding
GABA concentration
Frontal cortex
N. accumbens
Amygdala
Frontal pole
Hippocampus
Sup. temp. cortex
Inf. temp. cortex
Amygdala
Dorsal thalamus
Frontal cortex
N. accumbens
Mediodorsal thalamus
Thalamus
N. accumbens
Thalamus
N. accumbens
Hippocampus
Amygdala
N. accumbens
Ventrolateral thalamus
Dentate gyrus
CA 1–3
Subiculum
Post. hippocampus
Sup., inf., and med.
temporal gyri
Medial frontal and
orbitofrontal cortex
BA 9
BA 10
BA 11
Amygdala
Hippocampus
H 37n5 %*
H 40 %
H 45 %
H 48 %
H 61 %
H 72 %
H 21 %*
H 35 %*
H 31n9 %*
H 13 %
H 25 %*
GABA concentration
GABA concentration
GABA concentration
GABA concentration
GABA concentration
GABA concentration
GABA concentration
GABA-transaminase
Serum GABAtransaminase
Serum GABAtransaminase
GABA release
GABA release
(veratridine induced)
GABA uptake sites
([$H]nipecotic acid)
GABA uptake sites
([$H]nipecotic acid)
GABA uptake sites
([$H]nipecotic acid)
Comment
Korpi et al. (1987)
n l 7 SCZ,
4 controls
Kutay et al. (1989)
Perry et al. (1989)
Cross et al. (1979)
Perry et al. (1979)
Spokes et al. (1980)
H 25 %* early
onset cases only
nl7
Toru et al. (1988)
H 44 %*
H 25 %*
PMI controls PMI SCZ
BA 34
BA 8
H 77n3 %*
H 68n9 %*
BA 11
BA 38
Hippocampus
Amygdala
Hippocampus
H 18n9 %* left side
H bilat.**
H bilat.**
H 24 % left side
H 29 %* left side
H 21 % right side
H left side 15 %
I right side 15n7 %
Ohnuma et al. (1999)
Sherif et al. (1992)
White et al. (1980)
Amygdala
BA 11
BA 38
Author
Reveley et al. (1980)
nl5
Homogenate
Sherman et al. (1991)
Sherman et al. (1991)
Simpson et al. (1989)
Reynolds et al. (1990)
Sudden death cases
Simpson et al. (1992a)
Males
The GABAergic system in schizophrenia
165
Table 1b (cont.)
Marker
Area
Finding
GABA uptake sites
([$H]nipecotic acid)
Putamen
caudate
N. accumbens
Globus pallidus
Ant. cingulate ;
Ant. precentral
gyrus
Putamen (head)
Putamen (tail)
Caudate
H bilat. $ 50 %*
I*
I*
Globus pallidus
BA 9
BA 10
BA 11
H 18n1 %*
GABA uptake sites
([$H]nipecotic acid)
GABA uptake sites
([$H]nipecotic acid)
GAT-1 mRNA
Comment
Author
Simpson et al. (1992b)
Simpson et al. (1998a)
Manchester collection
Manchester and
Gothenburg collection
Simpson et al. (1998b)
Tissue sections
Ohnuma et al. (1999)
Table 2. CSF and serum GABA findings in schizophrenia
Measure
Finding
Comment
Author
GABA, CSF
concentration
GABA, CSF
concentration
Lichtshtein et al. (1978)
GABA, CSF
concentration
GABA, CSF
concentration
GABA, CSF
concentration
H*
n l 17 schizophrenics,
9 control
n l 7 schizophrenic,
5 schizoaffective, 2 other
psychosis, compared
with neurological control group
n l 17 schizoaffective
and schizophrenic
n l 11 schizophrenic,
29 controls
n l 17 controls,
9 untreated\7 treated
schizophrenics
McCarthy et al. (1981)
all schizophrenic
H 26 %* female pts only
In chronic schizophrenic
subset only
n l 25 drug-free schizophrenic and
5 schizoaffective
n l 20 chronic schizophenics
van Kammen et al. (1982)
n l 19 schizophrenic
Perry et al. (1989)
H plasma but not
CSF GABA associated
with prefrontal but
not global sulcal widening
n l 62 chronic
schizophenics
van Kammen et al. (1998)
n l 15 schizophrenics
n l 22 schizophrenics
Petty and Sherman (1984)
Reveley et al. (1980)
n l 14 schizophrenics
White et al. (1980)
GABA, CSF
concentration
GABA, CSF
concentration
GABA, CSF
concentration
GABA, CSF
concentration
GABA, CSF and
plasma levels
Plasma GABA
Platelet GABAtransaminase level
Platelet GABA
transaminase level
between untreated and
controls
I GABA in CSF
with long-term neuroleptic tx.
I 45n7 %*
Gold et al. (1980)
Sternberg (1980)
Gerner and Hare (1981)
Zimmer et al. (1981)
Gerner et al. (1984)
166
B. P. Blum and J. J. Mann
(1982) noted a significant decrease in the female schizophrenic sub-population compared to female controls while
also reporting a tendency towards increased GABA levels
with increased length of illness. That elevation of CSF
GABA levels may be correlated with length of schizophrenic illness finds support in a study by McCarthy et al.
(1981) in which a sub-population of chronic schizophrenics had higher GABA levels compared to controls.
However, Gerner et al. (1984) did not corroborate this
suggestion. An increase in CSF GABA levels has been
found after 30 d treatment with sulpride and to be
correlated with long-term neuroleptic treatment (Zimmer
et al., 1981). However, Lichtshtein et al. (1978) noted a
small but significant decrease in CSF GABA levels after 2
months of neuroleptic treatment while Gattaz et al. (1986)
observed no change in free CSF GABA levels in
schizophrenic patients after 3 months of haloperidol
treatment. Lastly, Zander et al. (1981) reported that
stopping chronic anti-psychotic medication treatment
produced no change in CSF GABA levels. CSF GABA is
lowered in depressed patients and therefore comorbidity
must be considered in interpretation of these studies
(Gerner and Hare, 1981 ; Gold et al., 1980). A later study
by Van Kammen et al. (1998) found that plasma GABA
levels showed a significant negative correlation with both
prefrontal sulcal widening and ventricle\brain ratio on CT
scans but not to global sulcal widening in patients with
schizophrenia. CSF GABA levels did not correlate with
these CT measures but did show a negative correlation
with age and age of onset. The disassociation between
CSF and serum GABA level is puzzling. One would
expect CSF GABA to reflect brain pathology better than
plasma. Moreover, the plasma GABA finding may not be
correct since plasma GABA levels were not found to be
lower in patients with schizophrenia (Petty and Sherman,
1984).
CSF GABA is not clearly lower in schizophrenia. There
are insufficient studies in which the possible confounds of
anti-psychotic medication treatment, comorbidity (in
particular affective disorders), length of illness and sex are
all controlled. Additionally, while pharmacological studies
in animals suggest that total CSF GABA concentrations
are mostly related to brain GABA (Bo$ hlen et al., 1979 ;
Ferkany et al., 1979), it remains unknown to what degree
this is true in humans. A defect limited to a specific
subtype of GABAergic neurons, such as the chandelier
subtype (see below), may not be reflected in CSF GABA
levels.
GABA-transaminase
Two studies measuring GABA-transaminase, the principal
catabolic enzyme for GABA in the mammalian brain,
failed to find any significant difference in platelet GABAtransaminase levels between schizophrenics and controls
(Reveley et al., 1980 ; White et al., 1980). No significant
effect of sex, psychotic state, length of illness or
medication treatment was noted in the study by White et
al. (1980) ; Reveley et al. (1980) reported no correlation
between age or sex and GABA-transaminase levels. Sherif
et al. (1992) measured GABA-transaminase in brain
homogenates from various regions including hippocampus, amygdala, cingulate and frontal gyrus and found no
significant difference between controls and undifferentiated schizophrenics. Thus, there is no evidence of altered
catabolism of GABA in schizophrenia.
GABA release and uptake
Synaptosomal preparations are used in the study of
synaptic function and neurotransmission in animal models.
Synaptosomal preparations obtained up to 24 h postmortem from human brains are metabolically active and
can release various neurotransmitters after veratrine
stimulation (Hardy et al., 1982). Using this model,
Sherman et al. (1991) compared schizophrenics and
controls, with PMI of 20p7 h (meanp..) and 23p7 h,
respectively, and reported a significantly lower veratridine-induced release of glutamate and GABA but not
aspartate in synaptosomes from temporal and frontal
cortex of the schizophrenic group.
The concentration and duration of a neurotransmitter
in the synaptic cleft is mostly regulated by rate of uptake
by transporter proteins. Four GABA transporters have so
far been described (GAT-1, GAT-2, GAT-3 and BGT-1),
each varying in its localization pattern and pharmacological profile (Borden, 1996). Simpson et al. (1989)
reported significantly lower binding of [$H]nipecotic acid
to GABA-uptake sites in the left BA 38 (polar temporal),
bilaterally in the amygdala and hippocampus in schizophrenic subjects compared to controls (see Table 1). This
study did have sufficient numbers of age-matched subjects
and controls with similar PMI ; agonal state effects were
said to have been minimized by selection of subjects who
had died acutely and were matched for cause of death.
Possible medication influence could not be entirely ruled
out, although the authors reported that binding data from
subjects that were drug-free were indistinguishable from
those treated with neuroleptics.
Comparing schizophrenic subjects to age- and PMImatched controls, Reynolds et al. (1990) found lower
[$H]nipecotic acid binding in both groups in the left
hippocampus compared to the right side but not so in the
amygdala. The schizophrenic group tended towards lower
binding values in the left hippocampus compared with
controls (p l 0n08) ; this tendency became statistically
The GABAergic system in schizophrenia
significant when subgroups of sudden-death cases were
compared. The rationale for this distinction was that
nipecotic acid binding may be reduced in patients with
chronic respiratory illnesses (Czudek and Reynolds, 1990).
Simpson et al. (1992a) used [$H]nipecotic acid to
measure GABA-uptake sites in a series of schizophrenic
and control brains in which cerebral atrophy had been
previously established and reported higher [$H]nipecotic
acid binding in both groups in the right compared with
left BA 38, lower left BA 38 binding and increased
putaminal binding in the schizophrenic brains compared
with controls. Subcortical GABA-uptake sites were further
studied by Simpson et al. (1992b) in brains of 19
schizophrenic subjects who had died with the diagnosis of
schizophrenia along with 22 neuropsychiatrically normal
controls, matched for age, gender ratio and PMI. In
contradiction to the above study, an approx. 50 % lower
[$H]nipecotic acid binding was seen in the putamen
bilaterally in the schizophrenic group. The binding of this
ligand did not differ between the two groups in the
caudate, globus pallidum or nucleus accumbens. This
study also reported no correlation between binding to
GABA-uptake sites and length of the neuroleptic-free
period in a subgroup of medication-free schizophrenics.
Simpson et al. (1998a) measured [$H]nipecotic acid
binding to GABA-uptake sites in 11 brain regions from a
group of 12 neuroleptic-treated chronic schizophrenia (9
males, 3 females) and a group of normal controls (14
males, 5 females). No significant overall difference was
noted between the two groups in any of the 11 temporal
and frontal lobe areas. The authors suggested these
findings might be influenced by low uptake measurements
in three of the female controls and large variance in the
hippocampal data.
A recent study examined [$H]nipecotic acid binding in
the basal ganglia from three brain collections (Manchester,
Gothenburg and Runwell) (Simpson et al., 1998b). The
schizophrenic (n l 12–18) and the control subjects (n l
19–22) did not differ with respect to age, PMI or storage
time. [$H]Nipecotic acid binding was higher in the
schizophrenic groups compared to controls in the heads
of the caudate and putamen of the Manchester collection.
Higher binding was also noted in the caudate of the
female schizophrenic subjects of the Gothenburg collection ; the caudate-binding values obtained from this
collection were 2- to 3-fold greater than those seen in the
Manchester collection. Caudate-binding values were not
reported for the Runwell collection.
Several studies indicate that GABA uptake may be
moderately lower in both the hippocampus ([$H]nipecotic
acid binding studies) and in BA 9 (GAT-1 mRNA studies,
see below) of schizophrenic subjects compared with
controls. GABA-uptake sites may reflect GABA terminal
167
distribution and appear to be sufficiently sensitive to
detect impaired input. There are approx. 100 times more
GABA terminals on the apical dendrites than on the
proximal axon segment. Lewis et al. (2000) reported a
deficit of the chandelier GABAergic neurons, which
specifically target the proximal axon segment. Such a
localized GABAergic input defect may not be equally
detectable by assays of GAT, GAD, brain GABA or CSF
GABA.
GABA receptors
Two types of GABA receptors have been identified in the
human brain : the GABAA receptor, which is associated
with a chloride channel and mediates fast inhibitory
synaptic transmission and the GABAB receptor which is
associated with potassium and calcium channels and is a
G protein-linked metabotrobic receptor (Bowery, 2000 ;
Olsen and Homanics, 2000). The GABAA receptor is
thought to be a heteropentameric glycoprotein composed
of subunits of six distinct subclasses : α, β, γ, δ, ε and ρ, the
largest being the α subclass which includes six known
members (α – ). In the adult mammalian brain, the subunit
"'
combination of α β γ is thought to be the most common
" # #
(Olsen and Homanics, 2000).
Bennett et al. (1979) used tritiated GABA as a ligand
(see Table 3) and reported that post-mortem binding in
frontal cortex homogenates of schizophrenics was not
significantly different from controls. The study did report
alterations in serotonergic receptor binding. Control and
schizophrenia groups were not well matched for age or
sex ratio. The authors reported no correlation between
PMI or time frozen and receptor-binding results ; however
agonal and possible drug effects could not be excluded.
Hanada et al. (1987) measured GABA receptor binding
using the GABA agonist [$H]muscimol and observed
significantly higher binding (Bmax) in both caudate and BA
9 in the chronic schizophrenic group as a whole compared
with controls. This finding survived subdivision into
sudden death and PTI subgroups in both regions of the
sudden-death subgroup but not in the caudate of the PTI
subgroup.
Benes et al. (1992) examined GABAA receptor binding
in the anterior cingulate gyrus in order to test a hypothesis
that upregulation of these receptors would follow the loss
of cortical interneurons reported to occur in this region
and the prefrontal cortex of chronically psychotic patients
(Benes et al., 1991). By using a bicuculline-sensitive
[$H]muscimol binding assay and a nuclear-track, coverslipemulsion technique, they counted autoradiographic grains
per neuron and per 200 µm# of neuropil. [$H]Muscimol
binding on neuronal cell bodies is 84 % higher in layer II
and 74 % higher in layer III in the schizophrenic group
168
B. P. Blum and J. J. Mann
Table 3. Postsynaptic markers
Marker
Area
Finding
Comment
Author
GABAA receptor
binding ([$H]GABA)
GABAA receptor
binding ([$H]muscimol)
GABAA receptor
binding (bicucullinesensitive [$H]muscimol)
GABAA receptor
binding (bicucullinesensitive [$H]muscimol)
Frontal cortex
Homogenate
Bennett et al. (1979)
BA 9
I 32 %*
Homogenate
Hanada et al. (1987)
Cingulate cortex
I 84 %* L II
I 74 %* L III
I 43 %* L II
I 70 %* L II
I 44 %* L III
I 48 %* L V
Tissue sections
Benes et al. (1992)
BA10
I 66 %* L VI
GABAA receptor
binding ([$H]muscimol)
GABAA receptor
binding ([$H]muscimol)
GABAA receptor
subunit mRNAs
GABAA receptor
subunit mRNAs
(γ S and γ L)
#
#
GABAA receptor
subunit mRNA
GABAB receptor
immunoreactivity
BZD receptor sites
[$H]flunitrazepam
BZD receptor sites
[$H]flunitrazepam
BZD receptor sites
[$H]flunitrazepam
BZD receptor sites
[$H]RO15-1788
BZD receptor sites
[$H]-flunitrazepam
Benes et al. (1996a)
I 90 %* L II
large neurons
I 135 %* L VI
sm. non-pyram.
Area dentate
Molecular
Granular
CA4, subiculum
presubiculum
CA3
I 20–40 %
I 40–60 %*
I 60–80 %*
I 60–80 %*
I 74–90 %*
CA1
BA 9
I 22–36 %
I 18n5 %*
BA 46
Akbarian et al. (1995a)
BA 46
H 28 % (both isoforms)
H 51n7 %* (γ S)
#
H 16n9 % (γ L)
#
I 49n1 %*
I 32n5 % (p l 0n051)
I 36n7 % (p l 0n0)
Not quantitated
nl5
H (p 0n01)
Huntsman et al. (1998)
BA 9
BA 10
BA 11
Dentate gyrus
CA1-4
Medial, inferior
and superior
temporal gyri
CA1-3
Dentate gyrus
BA 9, 10, 46
BA 45 and 47
BA 11 and 12
Hippocampus
Frontal cortex
Hippocampus
BA 10
Benes et al. (1996b)
I non-pyramidal
cells 3x pyram.
Dean et al. (1999)
nl5
nl6
Ohnuma et al. (1999)
Mizukami et al. (2000)
Homogenates
Kiuchi et al. (1989)
H (p 0n05)
I 25 %
I (p l 0n05 %)
I (p l 0n01 %)
H 29n0 %*
Homogenates
nl3
Homogenates
Squires et al. (1993)
Reynolds and Stroud (1993)
Homogenates
Pandey et al. (1997)
Tissue sections
Benes et al. (1997)
Area dentate
Subiculum
I 20–30 %*
Presubiculum
I 15–20 %*
CA1
CA2
CA3 (s. oriens only) I 30 %*
CA4
The GABAergic system in schizophrenia
compared to normal controls. In layer I neuropil [$H]
muscimol was increased in the schizophrenic group. PMIs
were similar in the two groups ; however, group sex ratios
and cause of death were not mentioned. The schizophrenic
group was significantly younger than the control group
but the authors discounted the possibility of a confounding effect as both younger and older schizophrenics had
elevated numbers of receptor sites compared to controls.
The authors believed that elevation of [$H]muscimol
binding was not secondary to neuroleptic treatment, as a
neuroleptic-naive and a minimally exposed patient both
had elevated binding.
Benes et al. (1996b) also used a bicuculline-sensitive
[$H]muscimol-binding assay to examine GABAA receptor
levels in the prefrontal cortex (BA 10) of 7 schizophrenic
subjects and 16 normal controls. No difference in average
size of neuronal cell bodies was observed between the
two groups ; however, more grains per cell were found on
the large (pyramidal) neurons of layers II–VI (greatest in
layer II) and on the small (non-pyramidal) neurons of layer
VI in the schizophrenic subjects compared with controls.
Although the control group was significantly older and
had a significantly shorter mean PMI compared to the
schizophrenic group, no correlation was found between
these potential confounds and GABAA binding. Two
schizophrenic subjects without history of neuroleptic
exposure had binding values that were lower than the
neuroleptic-treated schizophrenics and similar to the
average of the control group. These same two neurolepticfree schizophrenic subjects had exhibited a higher layer II
GABAA receptor-binding value in a previous study of the
anterior cingulate gyrus (Benes, 1992) compared with the
schizophrenic group as a whole indicating that the higher
GABAA receptor binding in the prefrontal cortex of the
schizophrenic group may not be simply a medication
effect.
Benes et al. (1996a), using brain tissue from the same
subjects in the above study (with the addition of one
subject to the schizophrenia group), reported higher
[$H]muscimol in subregions of the hippocampus of the
schizophrenic group compared with controls. Increases of
90 % (stratum oriens of CA3), of 74 % (stratum pyramidales of CA3), of 60–73 % (subiculum and presubiculum)
and of 22–36 % were seen in the CA1 subregion (ranges
indicating layer differences within a subregion). Increased
GABAA binding in the subregion CA3 was limited to
non-pyramidal cells while binding increases in the CA1
subregion were noted only on pyramidal cells. The author
postulated that these subregional increases in GABAA
receptor binding might reflect increased vulnerability of
certain subpopulation of GABAergic neurons to injury
during development.
Dean et al. (1999) reported increased binding of
169
[$H]muscimol to GABAA receptors, as well as decreased
[$H]ketanserin binding to 5-HT A receptors in BA 9 of
#
schizophrenic subjects compared to controls. The groups
were well matched for donor age, PMI, tissue pH and time
frozen ; analysis of covariance showed that these potential
confounds as well as final neuroleptic dose did not effect
the comparison of ligand binding between the groups.
Potential effects of agonal states were not addressed.
In animal experiments, reduced neuronal activity can
lead to decreased gene expression for a number of
GABAA receptor subunits (Hendry et al., 1990, 1994 ;
Huntsman et al., 1994). Akbarian et al. (1995a) used in-situ
hybridization histochemisty to quantitate mRNA of the
GABAA receptor subunits α , α , α , β , β and γ in the
" # & " #
#
prefrontal cortex. The schizophrenic and control groups
showed similar laminar gene expression patterns with
highest α , β , and γ expression in layers III and IV,
" #
#
highest α and β expression in layer II, and higher α
#
"
&
expression in layers IV–VI with peak expression in layer
IV. No significant difference in expression of any of the
subunit genes was noted between the two groups. The 12
schizophrenic and 12 control subjects were matched for
age, sex and PMI.
Huntsman et al. (1998) used in-situ hybridization
histochemisty and semi-quantitative reverse transcription–PCR to measure the relative abundance of two
species of mRNA of the γ subunit of the GABAA
#
receptor in the prefrontal cortex of five matched pairs of
schizophrenics and controls. The γ subunit, which is
#
necessary for high-affinity BZD binding, exists in two
forms : short (γ S) and long (γ L), which differ by a
#
#
functionally significant 8-amino-acid insert. The laminar
pattern of γ subunit mRNA labelling was consistent with
#
past reports for both schizophrenics and controls. Although the schizophrenic group was found to have lower
γ message labelling in each of the six cortical levels, this
#
difference reached statistical significance in only layers II
and III. The authors reported a lower level (average
51n7 %, p 0n001) of short (γ S) mRNA (but only 16n9 %
#
lower long (γ L) mRNA) in the prefrontal cortex of the
#
schizophrenic group compared with their matched controls. The authors speculated that this relative overabundance of the long (γ L) mRNA in the prefrontal cortex of
#
schizophrenics would result in GABAA receptors of
decreased functionality. Agonal effects were not discussed
but the authors expressed concern about possible medication effects.
Ohnuma et al. (1999) measured α subunit mRNA
"
expression in BA 9, 10, and 11 of 6 schizophrenics and 12
controls and found a general increase in the schizophrenic
group which attained statistical significance in the large
cells of layer V of BA 9 and in layer III of BA 10. The
patient group was comprised of neuroleptic-treated
170
B. P. Blum and J. J. Mann
chronic schizophrenics who had a shorter averaged PMI
than controls.
One study examined the anatomical distribution of
immunolabelled GABAB receptors in the hippocampus of
5 chronic schizophrenics and 3 controls matched for age
and PMI (Mizukami et al., 2000). Schizophrenic subjects
were reported to be resistant to neuroleptic treatment ;
however cause of death and treatment status at time of
death were not reported. The authors found less immunolabelling of the mossy cells in CA4 and the pyramidal cells
in CA1–3 in the schizophrenic subjects compared to the
controls. The granule cells of the dentate gyrus appeared
unstained in the schizophrenic subjects whereas staining
in controls was reported as moderate. In all regions the
degree of staining of interneurons was similar in both
subject types.
In summary, GABAA receptor binding is higher in
schizophrenia in cortical regions generally regarded as
important in the pathophysiology of schizophrenia.
Somewhat at odds with this observation is the tendency
towards less subunit mRNA in the prefrontal cortex of
schizophrenic subjects in two studies. These receptor
changes may represent upregulation in response to
reduced GABAergic input. What remains unclear is the
functional significance of alterations in binding or gene
expression. The functional response mediated by these
receptors may be impaired and counteract the benefits of
up-regulation. Studies of receptor coupling and signal
transduction are needed.
BZD binding studies
The therapeutic efficacy of BZDs as anxiolytic agents is
attributed to their ability to potentiate GABAA receptormediated inhibition by increasing the receptors affinity
for GABA. Selectivity of BZD binding to the GABAA
receptor is determined by specific amino-acid residues
in the γ and the α subunits (Mo$ hler et al., 2000).
Kiuchi et al. (1989) assayed [$H]flunitrazepam binding
in homogenates from multiple cortical regions of brains
from schizophrenic and control subjects and reported
significantly higher binding in the medial frontal cortex
orbitofrontal cortex, orbital cortex, medial and inferior
temporal gyri, cornu Ammonis 1–3 of the hippocampus
and putamen of the schizophrenic subjects compared with
controls (see Table 3). No significant differences in binding
were found in other areas. Medication effects might be a
confound in this study and agonal state issues were
unaddressed.
Reynolds and Stroud (1993) found no difference in
[$H]flunitrazepam binding in hippocampal homogenates
between a group of 15 schizophrenic subjects and normal
controls with matching sex compositions and ages.
Medication history was not reported.
Squires et al. (1993) found lower [$H]flunitrazepam
binding in a schizophrenic group compared with controls,
with differences reaching statistical significance in the
somatomotor and cingulate cortex but not in other
cortical regions such as the frontal cortex. Lower binding
in the schizophrenic group was also noted in the globus
pallidus, hippocampus and cerebellar cortex (vermis) but
not in the putamen. The authors speculated that these
reductions in binding might represent the loss of glutamatergic (pyramidal) cells. Four of 15 schizophrenic
subjects were suicide victims whereas the nine controls
were victims of traffic accidents. Past studies of BZD
receptors in suicide victims found altered (Cheetham et al.,
1988) and unaltered binding (Manchon et al., 1987 ;
Rochet et al., 1992 ; Stocks et al., 1990) ; therefore the use
of suicide victims may be a confound. The schizophrenic
subjects were reported to be drug-free for months prior to
death ; the average PMI appears to have been significantly
longer for the schizophrenic group (Squires et al., 1993).
To further study the relationships between suicide,
schizophrenia and BZD receptor binding, Pandey et al.
(1997) examined binding of the selective, high-affinity
radioligand [$H]RO15-1788 in prefrontal cortex Bmax
values (BA 10) homogenates from 13 suicide victims
without schizophrenia, 8 schizophrenic suicide victims, 5
non-suicide schizophrenic subjects and 15 normal controls. The Bmax of BZD receptors in the prefrontal cortex
was higher in suicide victims, largely due to increased
Bmax in the suicide victims who had died by violent
means. Overall, the Bmax of the schizophrenic subjects did
not differ from controls ; however, the sample size was
small.
Benes et al. (1997) assayed BZD binding with [$H]flunitrazepam in hippocampal tissue sections from the same
schizophrenic and control subjects used in a previous
study (Benes et al., 1996a). After normalization of the
data, the ratios of BZD binding to GABAA binding in
controls was similar throughout most of the hippocampal
region except in the inner and outer molecular layers of
the area dentata where higher BZD binding was observed.
[$H]Flunitrazepam binding was found to be only modestly
higher in the stratum oriens of the CA3, the subiculum
and the presubiculum of the schizophrenic subjects compared with controls. As the magnitude of these increases
did not match the increases in GABAA binding in these
regions, the authors speculated that the regulation of the
BZD-binding elements might be uncoupled from the regulation of the GABAA receptor. The authors noted that
such an uncoupling phenomena was reported in the cerebellum of the stagger mouse (Luntz-Leybman et al., 1995).
Taken as a group, these papers on BZD binding in schizo-
The GABAergic system in schizophrenia
phrenia do not provide a consensus about BZD binding
in examined regions of the frontal or temporal lobes.
Calcium-binding proteins as markers of GABAergic
neurons
In the prefrontal cortex of primates, sub-populations of
GABAergic interneurons can be classified based on
morphological characteristics, synaptic targets or the
presence of different calcium-binding proteins (Conde! et
al., 1994 ; Lund and Lewis, 1993). The calcium-binding
protein parvalbumin is found primarily in the wide-basket
and chandelier subclasses of GABA neurons. The axon
terminals of the chandelier neurons synapse on the initial
axon segments of pyramidal cell. The axon terminals of
the wide- basket neurons synapse on the cell bodies and
dendrites of pyramidal cells. GABA neurons in the doublebouquet subclass contain calretinin (CR) and have terminal
axons that synapse onto the dendritic shafts of both
pyramidal and non-pyramidal neurons. The parvalbumincontaining GABA neurons of the chandelier subclass have
attracted the most scrutiny in studies of schizophrenia
because their synaptic targeting of the axon initial
segment of pyramidal cells suggest a strong influence on
171
pyramidal cell output ; they also appear to receive direct
synaptic input from mesocortical dopamine and thalamocortical glutamatergic projection (Muly III et al., 1998 ;
Sesack et al., 1995, 1998).
An early study of calcium-binding proteins in schizophrenia used CR and calbindin (CB) immunohistochemical
labelling of tissue from prefrontal cortical areas 9 and 46
obtained from 1 schizoaffective and 4 schizophrenic
subjects and 5 controls matched for age, sex and PMI (see
Table 4) (Daviss and Lewis, 1995). One of the schizophrenic subjects died by suicide ; the cause of death listed
for the remainder of subjects are consistent with short
agonal periods. The authors found a 50–70 % greater
density of the CB-immunoreactive (CB-IR) and a 10–20 %
(non-significant) greater density of the CR-immunoreactive (CR-IR) non-pyramidal neurons of both cortical areas
in the schizophrenic group compared with the controls.
The authors noted small sample size, lack of stereological
methodology and potential medication effects as caveats.
Beasley and Reynolds (1997) used a monoclonal
antibody against parvalbumin to quantitate parvalbumincontaining chandelier and wide-basket GABA neurons in
tissue sections from BA 10 obtained from schizophrenic
and control subjects. The authors reported fewer parval-
Table 4. Calcium protein markers of non-pyramidal neurons in schizophrenia
Marker
Area
Non-pyramidal neurons
Calbindin-IR
Calretinin-IR
Parvalbumin-IR neurons
BA 9 and 46
Parvalbumin-IR neurons
Parvalbumin-IR neurons
GAT-1-IR cartridges
(chandelier cells)
GAT-1-IR cartridges
(chandelier cells)
BA 10
BA 9
BA 46
BA 17
Ant. cingulate
cortex
BA 9
BA 46
BA 46
Finding
Author
Daviss and Lewis (1995)
I 50–70 %*
I 10–20 %
H* layer III
H* layer IV
I layers Va–Vb
H 40 %*
H 40 %*
H 27n7 %* layer II–IIIb
H 31n5 %* layer IIIb–IV
layer VI
Parvalbumin-IR neurons
BA 9 and
BA 46
GAT-1-IR cartridges
(chandelier cells)
BA 9 and
BA 46
H 40 %*
Parvalbumin-IR neurons
BA 9 and
BA 46
H*
Calbindin-IR neurons
Calretinin-IR neurons
Comment
H*
Beasley and Reynolds (1997)
Woo et al. (1997)
Kalus et al. (1997)
Woo et al. (1998)
Pierri et al. (1999)
Results previously
reported in Woo et al.
(1997)
Results previously
reported inWoo et al.
(1998)
Lewis (2000)
Lewis (2000)
Reynolds and Beasley (2001)
172
B. P. Blum and J. J. Mann
bumin-positive cells in the schizophrenic subjects compared with the normal controls. Differences reached
statistical significance only in layers III and IV. No group
difference was found in cortical thickness. Age, sex, and
duration of illness did not have an effect on cell counts.
The authors did not use a stereological method. The
question of an effect of neuroleptics on parvalbumin
expression and cell counts was left open.
Woo et al. (1997) examined parvalbumin-IR local
circuit GABAergic neurons in tissue sections from BA 9,
46 (prefrontal) and 17 (visual) obtained from 15 schizophrenic subjects and sex-matched controls and detected
no significant difference in their densities between the
schizophrenics and controls. As the authors found no
somal size differences between the two subject groups,
the inability to perform absolute cell counts was not
thought to be a confound. However, differences in
neuropil or tissue shrinkage could be critical for density
measures. Cause of death and medication histories of the
subjects were not reported.
More parvalbumin-IR GABA interneurons in layers Va
and Vb of the anterior cingulate cortex was found in
schizophrenics compared with controls, but the density of
Nissl-stained neuron profiles did not differ in any of the
layers (Kalus et al., 1997). Stereological methods were not
employed. The two groups differed significantly in
average PMI ; disparities in tissue shrinkage, medication
histories and agonal states may have confounded the
results.
Woo et al. (1998) used an antibody against the GABA
transporter GAT-1 to identify the distinctive vertical
arrays of chandelier axons known as cartridges. This
study included 15 schizophrenic subjects matched by age,
sex and PMI to both a normal control and a nonschizophrenic psychiatric group. The relative density of
GAT-1-IR cartridges, assessed by stereological methods,
was lower in the schizophrenic subjects across layers II-VI
in both BA 9 and 46 compared with both the psychiatric
and normal controls. Density of CR-IR axon boutons in
layers II–IIIa did not differ between the schizophrenic and
normal control subjects. The schizophrenic group included
two suicide victims, the psychiatric group included 12
suicide victims and normal control group had no suicide
cases ; cause of death in the remaining subjects was not
reported. The majority of the schizophrenic group had
been treated with neuroleptic medication ; however, the
authors noted that two of the schizophrenic subjects who
had been off medications for a significant time before
death also had GAT-1 cartridge densities that were lower
than control densities.
Pierri et al. (1999) also examined the laminar densities
of GAT-1-IR cartridges and found that in comparison
with a psychiatric and a normal control group, a group of
30 schizophrenics [15 of the comparison triads had been
used in a previous study (Woo et al., 1998)] had
significantly lower GAT-1-IR cartridge density in layers
II–IIIa and IIIb–IV. The schizophrenic subjects were
matched to controls by sex, age and PMI. Significant
numbers of subjects in both the schizophrenia group and
psychiatric control group but not the normal control
group had histories of substance abuse or were suicide
victims. Medication effects were not apparent on GAT-1IR cartridge density in prefrontal cortex of male Cynomolgus monkeys were treated for 9–12 months with
haloperidol decanoate and benztropine mesylate.
GAT-1 mRNA levels were quantitated in 10 pairs of
schizophrenic subjects and controls (see Volk et al., 2000)
in a study which sought to determine if lower GAT-1
density in the prefrontal cortex was accompanied by
lower GAT-1 gene expression (Volk et al., 2001). A
threshold of 2-fold background was used to exclude nonspecific labelling and a somal size criterion of greater than
50 µm# was used to exclude glial cells. GAT-1 mRNApositive neuron density was lower (21–33 %) in layer I,
layer II, the superficial portion of layer III, and at the
boundary of layers III–IV in the schizophrenic group
compared with controls. Grain density per neuron was
also significantly decreased (11 %, p l 0n009) in the
schizophrenic group only at the layers III–IV border and
cross-sectional size did not differ significantly between
the two groups. This study also compared GAT-1 mRNA
labelling between 4 haloperidol- and benztropine mesylate-treated Cynomolgus monkeys and 4 untreated controls and reported that after 9–12 months of treatment
there were no significant differences. The authors concluded that GAT-1 expression in the prefrontal cortex of
schizophrenics is unaltered overall, but that it may be
lower in the chandelier class of GABAergic cells. The
authors noted that lower density of GAT-1 mRNApositive neurons is congruent with a previous finding of
laminar-specific decreases in GAD -positive neuronal
'(
but not synaptophysin-mRNA-positive neuronal densities (Volk et al., 2000).
Overall, it appears that there may be fewer GAT-1-IR
axon cartridges consistent with less GABAergic inhibition
at the proximal axon segment of pyramidal cells by
parvalbumin-positive chandelier cells. Of the calciumbinding proteins, parvalbumin alone is expressed later in
foetal development in GABAergic interneurons. Late
expression of parvalbumin is hypothesized to lead to a
‘ window of vulnerability ’ in which an insult to the foetus
leads to glutamate receptor stimulation and cytotoxic
calcium influx (Reynolds and Beasley, 2001). An important
question to be resolved is whether or not the abnormality
in the parvalbumin-class interneurons involves a loss of
such cells (Beasley and Reynolds, 1997 ; Reynolds and
The GABAergic system in schizophrenia
Beasley, 2001) or is limited to a decrease in the number of
axon cartridges (Pierri et al., 1999). Of the calcium-protein
positive GABAergic interneurons in the adult mouse
cortex, parvalbumin-class interneurons alone do not
express reelin protein, suggesting that the abnormality in
this class of interneurons is not related to the deficits in
reelin reported in schizophrenia (Alca! ntara et al., 1998 ;
Guidotti et al., 2000 ; Impagnatiello et al., 1998).
Non-pyramidal cell counts
A non-stereological study of 9 chronic schizophrenic, 9
schizoaffective and 12 control subjects reported fewer
small neurons in layers I and II of the prefrontal cortex
(BA 10) and in layers II–VI of the anterior cingulate (BA
24) in the two patient groups compared with controls
(Benes et al., 1991) These decreases tended to be greater
in the schizoaffective subgroup. Glial cell numbers did not
differ between the groups nor did pyramidal cell numbers
except in layer V of the patient group in which had
significantly higher counts were observed.
A recent stereological study of post-mortem tissue
from the hippocampus obtained from 10 schizophrenic
and 10 age- and PMI-matched controls found fewer
numbers of non-pyramidal cells in CA2 sector of the
schizophrenics compared with controls (Benes et al.,
1998). A similar finding was reported in a group of four
bipolar patient also included in this study. Numbers of
pyramidal cells did not differ between the groups. Three
of the schizophrenic subjects were suicide victims and
both groups may have been partly composed of subjects
with prolonged agonal intervals. Two schizophrenic
subjects who were neuroleptic-free for at least 1 yr also
had decreased non-pyramidal counts in the sector CA2.
There may be fewer non-pyramidal neurons in the
prefrontal cortex in schizophrenia, further evidence of a
GABAergic deficit.
Conclusion
Substantial evidence argues for a defect in the GABAergic
system of the frontal cortex in schizophrenia, particularly
in the prefrontal region and to a lesser degree in the
anterior cingulate gyrus. A coherent pattern can be
described : lower GAD mRNA and protein (Akbarian et
'(
al., 1995b ; Guidotti et al., 2000 ; Impagnatiello et al.,
1998 ; Volk et al., 2000) is possibly paralleled by lower
GABA concentrations (Kutay et al., 1989), less release of
GABA (Sherman et al., 1991), lower GAT-1 mRNA
(Ohnuma et al., 1999 ; Volk et al., 2001) and up-regulation
of GABAA sites (Benes et al., 1992, 1996b ; Dean et al.,
1999 ; Hanada et al., 1987).
Use of calcium-binding proteins as markers indicates
173
that a GABAergic defect may be specific for the chandelier
class interneurons (Beasley and Reynolds, 1997 ; Lewis,
2000 ; Reynolds et al., 2000). Fewer chandelier-class
GABAergic synapsing onto cortical pyramidal cells may
contribute to impaired ability to perform dopaminedependent functions such as working memory (GoldmanRakic, 1996 ; Lewis et al., 1999). Decreases in dopamine
input into the prefrontal cortex may also lead to decreased
cortical glutamatergic input to the ventral striatum\
ventral pallidum. This may lead to a decrease in tonic
dopamine release resulting in a decreased ability to
regulate phasic dopamine release in mesolimbic circuits
leading to positive symptoms (Grace, 1991 ; Moore et al.,
1999). Alternatively, a decrease in cortical glutamatergic
activity onto striatal GABAergic projection neurons may
lead to a decrease in the inhibitory effects of the indirect
striatothalmic pathway on the thalamus. Such an effect
may decrease the ability of the thalamus to filter off
excessive or irrelevant stimuli (Carlsson et al., 2001).
Some evidence is also presented for the existence of
GABAergic defect in regions of the temporal lobe, in
particular in the hippocampus. In this region, there also
appears to be deficits in GABA uptake (Reynolds et al.,
1990 ; Simpson et al., 1989) with increased (and possibly
compensatory) GABAA receptor binding (Benes et al.,
1996a). Studies of BZD receptor binding have generated
conflicting results in both frontal cortical regions and in
the hippocampus. The only study employing the use of
tissue sections reported modest regional increases in
hippocampal BZD binding in the schizophrenic group
that suggest an uncoupling of the BZD and GABAA
receptors (Benes et al., 1997). Whether or not this reflects
a true uncoupling of the BZD and GABAA receptors and
is related to an abnormality in γ-subunit processing
(Huntsman et al., 1998) remains to be determined.
The full anatomical distribution of post-mortem findings in the GABAergic system in schizophrenia is not
known with certainty because most studies have selectively examined certain regions. Systematic mapping
studies of the human neocortex are lacking for most
GABAergic markers. The authors also wish to emphasize
that many of the findings reviewed in this article remain
unreplicated. Interpretation of abnormal findings in the
GABAergic system in schizophrenia should be tempered
by lack of information on functional changes in
GABAergic transmission, the awareness that schizophrenia may be a heterogeneous set of disorders and that
multiple defects may cause the same basic illness. One
also needs to keep in mind the likely complexity of the
GABAergic system ; a system in which the major synthetic
enzyme occurs in two distinct forms at the genomic level,
the number of recognized receptor subtypes is approx. 20
(Olsen and Homanics, 2000). Complexity is also added by
174
B. P. Blum and J. J. Mann
the fact that GABAergic neurons interact with multiple
neurotransmitters systems, exist in at least 14 distinct
electrophysiological subtypes (Gupta et al., 2000) and are
involved in virtually every brain circuit. Additionally, one
needs to use caution in interpreting findings from any
study that has not controlled for medication history. For
example, chronic haloperidol treatment increases the size
of GABA-IR axosomatic terminals in the medial prefrontal
cortex of rats (Vincent et al., 1994) and increases GABA
receptor binding in the substantia nigra, the latter effect
being partially reversed after 8 d of treatment cessation
(Huffman and Ticku, 1983). Part of the antipsychotic
effects of medications such as haloperidol may be due to
such secondary changes in the GABAergic system.
References
Adler LE, Olincy A, Waldo M, Harris JG, Griffith J, Stevens
K, Flach K, Nagamoto H, Bickford P, Leonard S, Freedman
R (1998). Schizophrenia, sensory gating, and nicotinic
receptors. Schizophrenia Bulletin 24, 189–202.
Akbarian S, Huntsman MM, Kim JJ, Tafazzoli A, Potkin SG,
Bunney Jr. WE, Jones EG (1995a). GABAA receptor subunit
gene expression in human prefrontal cortex : comparison of
schizophrenics and controls. Cerebral Cortex 5, 550–560.
Akbarian S, Kim JJ, Potkin SG, Hagman JO, Tafazzoli A,
Bunney Jr. WE, Jones EG (1995b). Gene expression for
glutamic acid decarboxylase is reduced without loss of
neurons in prefrontal cortex of schizophrenics. Archives of
General Psychiatry 52, 258–266.
Alca! ntara S, Ruiz M, D’Arcangelo G, Ezan F, de Lecea L,
Curran T, Sotelo C, Soriano E (1998). Regional and cellular
patterns of reelin mRNA expression in the forebrain of the
developing and adult mouse. Journal of Neuroscience 18,
7779–7799.
Awapara J, Landua AJ, Fuerst R, Seale B (1950). Free γaminobutyric acid in the brain. Journal of Biological
Chemistry 187, 35–39.
Bazemore AW, Elliott KAC, Florey E (1957). Isolation of
Factor I. Journal of Neurochemistry 1, 334–339.
Beasley CL, Reynolds GP (1997). Parvalbuminimmunoreactive neurons are reduced in the prefrontal
cortex of schizophrenics. Schizophrenia Research 24,
349–355.
Benes FM (1999). Evidence for altered trisynaptic circuitry in
schizophrenic hippocampus. Biological Psychiatry 46,
589–599.
Benes FM, Khan Y, Vincent SL, Wickramasinghe R (1996a).
Differences in the subregional and cellular distribution of
GABAA receptor binding in the hippocampal formation of
schizophrenic brain. Synapse 22, 338–349.
Benes FM, Kwok EW, Vincent SL, Todtenkopf MS (1998). A
reduction of nonpyramidal cells in sector CA2 of
schizophrenics and manic depressives. Biological Psychiatry
44, 88–97.
Benes FM, McSparren J, Bird ED, SanGiovanni JP, Vincent SL
(1991). Deficits in small interneurons in prefrontal and
cingulate cortices of schizophrenic and schizoaffective
patients. Archives of General Psychiatry 48, 996–1001.
Benes FM, Todtenkopf MS, Logiotatos P, Williams M (2000).
Glutamate decarboxylase -immunoreactive terminals in
'&
cingulate and prefrontal cortices of schizophrenic and
bipolar brain. Journal of Chemical Neuroanatomy 20,
259–269.
Benes FM, Vincent SL, Alsterberg G, Bird ED, SanGiovanni JP
(1992). Increased GABAA receptor binding in superficial
layers of cingulate cortex in schizophrenics. Journal of
Neuroscience 12, 924–929.
Benes FM, Vincent SL, Marie A, Khan Y (1996b). Upregulation of GABAA receptor binding on neurons of the
prefrontal cortex in schizophrenic subjects. Neuroscience 75,
1021–1031.
Benes FM, Wickramasinghe R, Vincent SL, Khan Y,
Todtenkopf M (1997). Uncoupling of GABAA and
benzodiazepine receptor binding activity in the
hippocampal formation of schizophrenic brain. Brain
Research 755, 121–129.
Bennett Jr. JP, Enna SJ, Bylund DB, Gillin JC, Wyatt RJ,
Snyder SH (1979). Neurotransmitter receptors in frontal
cortex of schizophrenics. Archives of General Psychiatry 36,
927–934.
Berretta S, Munno DW, Benes FM (2001). Amygdalar
activation alters the hippocampal GABA system, ‘ partial ’
modelling for postmortem changes in schizophrenia. Journal
of Comparative Neurology 431, 129–138.
Bird ED, Spokes EG, Barnes J, MacKay AV, Iversen LL,
Shepherd M (1977). Increased brain dopamine and reduced
glutamic acid decarboxylase and choline acetyl transferase
activity in schizophrenia and related psychoses. Lancet 2,
1157–1158.
Bloom FE, Iversen LL (1971). Localizing $H-GABA in nerve
terminals of rat cerebral cortex by electron microscopic
autoradiography. Nature 229, 628–630.
Bo$ hlen P, Huot S, Palfreyman MG (1979). The relationship
between GABA concentrations in brain and cerebrospinal
fluid. Brain Research 167, 297–305.
Borden LA (1996). GABA transporter heterogeneity,
pharmacology and cellular localization. Neurochemistry
International 29, 335–356.
Bowery N (2000). GABAB Receptors : structure and function.
In : Martin D, Olsen R (Eds.), GABA in the Nervous System,
The View at Fifty Years (pp. 233–244). Philadelphia :
Lippincott Williams & Wilkins.
Breier A, Su TP, Saunder R, Carson RE, Kolachana BS, De
Bartolomeis A, Weinberger DR, Weisenfeld N, Malhotra
AK, Eckelman WC, Pickar D (1997). Schizophrenia is
associated with elevated amphetamine-induced synaptic
dopamine concentrations : evidence form a novel positron
emission tomography method. Proceedings of the National
Academy of Sciences USA 94, 2569–2574.
Bu DF, Erlander MG, Hitz BC, Tillakaratne NJ, Kaufman DL,
Wagner-McPherson CB, Evans GA, Tobin AJ (1992). Two
The GABAergic system in schizophrenia
human glutamate decarboxylases, 65-kDa GAD and 67kDa GAD, are each encoded by a single gene. Proceedings
of the National Academy of Sciences USA 89, 2115–2119.
Carlsson A (1988). The current status of the dopamine
hypothesis of schizophrenia. Neuropsychopharmacology 1,
179–186.
Carlsson A, Waters N, Holm-Waters S, Tedroff J, Nilsson M,
Carlsson ML (2001). Interactions between monoamines,
glutamate, and GABA in schizophrenia, new evidence.
Annual Review of Pharmacology and Toxicology 41, 237–260.
Cheetham SC, Crompton MR, Katona CLE, Parker SJ, Horton
RW (1988). Brain GABAA\benzodiazepine binding sites
and glutamic acid decarboxylase activity in depressed
suicide victims. Brain Research 460, 114–123.
Conde! F, Lund JS, Jacobowitz DM, Baimbridge KG, Lewis
DA (1994). Local circuit neurons immunoreactive for
calretinin, calbindin D-28k or parvalbumin in monkey
prefrontal cortex : distribution and morphology. Journal of
Comparative Neurology 341, 95–116.
Cross AJ, Crow TJ, Owen F (1979). Gamma-aminobutyric
acid in the brain in schizophrenia. Lancet 1, 560–561.
Cross AJ, Owen F (1979). The activities of glutamic acid
decarboxylase and choline acetyltransferase in post-mortem
brains of schizophrenics and controls. Biochemical Society
Transactions 7, 145–146.
Crow TJ, Owen F, Cross AJ, Lofthouse R, Longden A (1978).
Brain biochemistry in schizophrenia. Lancet 1, 36–37.
Curran T, D’Arcangelo G (1998). Role of reelin in the control
of brain development. Brain Research Brain Research Reviews
26, 285–294.
Czudek C, Reynolds GP (1990). [$H]nipecotic acid binding to
gamma-aminobutyric acid uptake sites in postmortem
human brain. Journal of Neurochemistry 55, 165–168.
Daviss SR, Lewis DA (1995). Local circuit neurons of the
prefrontal cortex in schizophrenia, selective increase in the
density of calbindin-immunoreactive neurons. Psychiatry
Research 59, 81–96.
Deakin JFW, Simpson MDC (1997). A two-process theory of
schizophrenia, evidence from studies in post-mortem brain.
Journal of Psychiatric Research 31, 277–295.
Dean B (2001). A predicted cortical serotonergic\
cholinergic\GABAergic interface as a site of pathology
in schizophrenia. Clinical and Experimental Pharmacology
and Physiology 28, 74–78.
Dean B, Hussain T, Hayes W, Scarr E, Kitsoulis S, Hill C,
Opeskin K, Copolov DL (1999). Changes in serotonin A
#
and GABAA receptors in schizophrenia, studies on the
human dorsolateral prefrontal cortex. Journal of
Neurochemistry 72, 1593–1599.
Deutsch SI, Rosse RB, Schwartz BL, Mastropaolo J (2001). A
revised excitotoxic hypothesis of schizophrenia, therapeutic
implications. Clinical Neuropharmacology 24, 43–49.
Dewey SL, Smith GS, Logan J, Brodie JD, Yu DW, Ferrieri
RA, King PT, MacGregor RR, Martin TP, Wolf AP (1992).
GABAergic inhibition of endogenous dopamine release
measured in vivo with 11C-raclopride and positron
emission tomography. Journal of Neuroscience 12,
3773–3780.
175
Dudel J, Gryder R, Kaji A, Kuffler SW, Potter DD (1963).
Gamma-aminobutyric acid and other blocking compounds
in crustacea I. Central nervous system. Journal of
Neurophysiology 26, 721–728.
Erlander MG, Tobin AJ (1991). The structural and functional
heterogeneity of glutamic acid decarboxylase, a review.
Neurochemical Research 16, 215–226.
Esclapez M, Tillakaratne NJK, Kaufman DL, Tobin AJ, Houser
CR (1994). Comparative localization of two forms of
glutamic acid decarboxylase and their mRNAs in rat brain
supports the concept of functional differences between the
forms. Journal of Neuroscience 14, 1834–1855.
Feldblum S, Erlander MG, Tobin AJ (1993). Different
distributions of GAD65 and GAD67 mRNAs suggest that
the two glutamate decarboxylases play distinctive
functional roles. Journal of Neuroscience Research 34,
689–706.
Ferkany JW, Butler IJ, Enna SJ (1979). Effect of drugs on rat
brain, cerebrospinal fluid and blood GABA content. Journal
of Neurochemistry 33, 29–33.
Florey E (1991). GABA : history and perspectives. Canadian
Journal of Physiology and Pharmacology 69, 1049–1056.
Florey E, McLennan H (1955). The release of an inhibitory
substance from mammalian brain and its action on
peripheral synaptic transmission. Journal of Physiology
(London) 129, 384–392.
Fuxe K, Perez de la Mora M, Ho$ kfelt T (1977). GABA–DA
interactions and their possible relation to schizophrenia. In :
Shagass C, Gershon S, Friedhoff AJ (Eds.), Psychopathology
and Brain Pathology (pp. 97–111). New York : Raven Press.
Garbutt JC, van Kammen DP (1983). The interaction between
GABA and dopamine : implications for schizophrenia.
Schizophrenia Bulletin 9, 336–353.
Gattaz WF, Roberts E, Beckmann H (1986). Cerebrospinal
fluid concentrations of free GABA in schizophrenia, no
changes after haloperidol treatment. Journal of Neural
Transmission 66, 69–73.
Gerner RH, Fairbanks L, Anderson GM, Young JG, Scheinin
M, Linnoila M, Hare TA, Shaywitz BA, Cohen DJ (1984).
CSF neurochemistry in depressed, manic, and schizophrenic
patients compared with that of normal controls. American
Journal of Psychiatry 141, 1533–1540.
Gerner RH, Hare TA (1981). CSF GABA in normal subjects
and patients with depression, schizophrenia, mania, and
anorexia nervosa. American Journal of Psychiatry 138,
1098–1101.
Gold BI, Bowers Jr. MB, Roth RH, Sweeney DW (1980).
GABA levels in CSF of patients with psychiatric disorders.
American Journal of Psychiatry 137, 362–364.
Goldman-Rakic PS (1996). Regional and cellular fractionation
of working memory. Proceedings of the National Academy of
Sciences USA 93, 13473–13480.
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.
176
B. P. Blum and J. J. Mann
Guidotti A, Auta J, Davis JM, Gerevini VD, Dwivedi Y,
Grayson DR, Impagnatiello F, Pandey G, Pesold C, Sharma
R, Uzunov D, Costa E (2000). Decrease in reelin and
glutamic acid decarboxylase (GAD ) expression in
'(
'(
schizophrenia and bipolar disorder, a postmortem brain
study. Archives of General Psychiatry 57, 1061–1069.
Gupta A, Wang Y, Markram H (2000). Organizing principles
for a diversity of GABAergic interneurons and synapses in
the neocortex. Science 287, 273–278.
Hanada S, Mita T, Nishino N, Tanaka C (1987). [$H]muscimol
binding sites increased in autopsied brains of chronic
schizophrenics. Life Sciences 40, 259–266.
Hardy JA, Dodd PR, Oakley AE, Kidd AM, Perry RH,
Edwardson JA (1982). Use of post-mortem human
synaptosomes for studies of metabolism and transmitter
amino acid release. Neuroscience Letters 33, 317–322.
Hendry SHC, Fuchs J, deBlas AL, Jones EG (1990).
Distribution and plasticity of immunocytochemically
localized GABAA receptors in adult monkey cortex. Journal
of Neuroscience 10, 2438–2450.
Hendry SHC, Huntsman MM, Vin4 uela A, Mo$ hler H, de Blas
AL, Jones EG (1994). GABAA receptor subunit
immunoreactivity in primate visual cortex, distribution in
macaque and humans and regulation by visual input in
adults. Journal of Neuroscience 14, 2383–2401.
Holstein GR, Pasik P, Ha! mori J (1986). Synapses between
GABA-immunoreactive axonal and dendritic elements in
monkey substantia nigra. Neuroscience Letters 66, 316–322.
Huffman RD, Ticku MK (1983). The effects of chronic
haloperidol administration on GABA receptor binding.
Pharmacology, Biochemistry and Behavior 19, 199–204.
Huntsman MM, Isackson PJ, Jones EG (1994). Lamina-specific
expression and activity-dependent regulation of seven
GABAA receptor subunit mRNA in monkey visual cortex.
Journal of Neuroscience 14, 2236–2259.
Huntsman MM, Tran BV, Potkin SG, Bunney Jr. WE, Jones
EG (1998). Altered ratios of alternatively spliced long and
short γ2 subunit mRNAs of the γ-amino butyrate type A
receptor in prefrontal cortex of schizophrenics. Proceedings
of the National Academy of Sciences USA 95, 15066–15071.
Impagnatiello F, Guidotti AR, Pesold C, Dwivedi Y, Caruncho
H, Pisu MG, Uzunov DP, Smalheiser NR, Davis JM,
Pandey GN, Pappas GD, Tueting P, Sharma RP, Costa E
(1998). A decrease of reelin expression as a putative
vulnerability factor in schizophrenia. Proceedings of the
National Academy of Sciences USA 95, 15718–15723.
Japha K, Koch M (1999). Picrotoxin in the medial prefrontal
cortex impairs sensorimotor gating in rats, reversal by
haloperidol. Psychopharmacology 144, 347–354.
Jones EG (1990). GABA-peptide neurons in the neocortex
(‘ Inhibition in the Brain ’ Symposium, November 1986,
Washington, DC). In: Paxinos G (Ed.), The Human Brain (p.
1116). San Diego : Academic Press.
Kalus P, Senitz D, Beckmann H (1997). Altered distribution of
parvalbumin-immunoreactive local circuit neurons in the
anterior cingulate cortex of schizophrenic patients.
Psychiatry Research, Neuroimaging Section 75, 49–59.
Kaufman DL, Houser CR, Tobin AJ (1991). Two forms of the
γ-aminobutyric acid synthetic enzyme glutamate
decarboxylase have distinct intraneuronal distributions and
cofactor interactions. Journal of Neurochemistry 56, 720–723.
Keverne EB (1999). GABA-ergic neurons and the
neurobiology of schizophrenia and other psychoses. Brain
Research Bulletin 48, 467–473.
Kiuchi Y, Kobayashi T, Takeuchi J, Shimizu H, Ogata H, Toru
M (1989). Benzodiazepine receptors increase in postmortem brain of chronic schizophrenics. European Archives
of Psychiatry and Neurological Sciences 239, 71–78.
Korpi ER, Kleinman JE, Goodman SI, Wyatt RJ (1987).
Neurotransmitter amino acids in post-mortem brains of
chronic schizophrenic patients. Psychiatry Research 22,
291–301.
Kravitz EA, Kuffler SW, Potter DD (1963). Gammaaminobutyric acid and other blocking compounds in
crustaceans. III. Their relative concentrations in separated
motor and inhibitory axons. Journal of Neurophysiology 26,
739–751.
Kutay FZ, Po$ g) u$ n SS , Hariri NI, Peker G, Erlac: in S (1989). Free
amino acid level determinations in normal and
schizophrenic brain. Progress in Neuro-Psychopharmacology
and Biological Psychiatry 13, 119–126.
Lewis DA, Pierri JN, Volk DW, Melchitzky DS, Woo TU
(1999). Altered GABA neurotransmission and prefrontal
cortical dysfunction in schizophrenia. Biological Psychiatry
46, 616–626.
Lewis DA (2000). GABAergic local circuit neurons and
prefrontal cortical dysfunction in schizophrenia. Brain
Research Brain Research Reviews 31, 270–276.
Lichtshtein D, Dobkin J, Ebstein RP, Biederman J, Rimon R,
Belmaker RH (1978). Gamma-aminobutyric acid (GABA) in
the CSF of schizophrenic patients before and after
neuroleptic treatment. British Journal of Psychiatry 132,
145–148.
Lund JS, Lewis DA (1993). Local circuit neurons of
developing and mature macaque prefrontal cortex, Golgi
and immunocytochemical characteristics. Journal of
Comparative Neurology 328, 282–312.
Luntz-Leybman V, Rotter A, Zdilar D, Frostholm A (1995).
Uncoupling of GABAA\benzodiazepine receptor α , β ,
" #
and γ subunit mRNA expression in cerebellar Purkinje
#
cells of staggerer mutant mice. Journal of Neuroscience 15,
8121–8130.
Manchon M, Kopp N, Rouzioux JJ, Lecestre D, Deluermoz S,
Miachon S (1987). Benzodiazepine receptor and
neurotransmitter studies in the brain of suicides. Life
Sciences 41, 2623–2630.
McCarthy BW, Gomes UR, Neethling AC, Shanley BC,
Taljaard JJ, Potgieter L, Roux JT (1981). γ-aminobutyric
acid concentration in cerebrospinal fluid in schizophrenia.
Journal of Neurochemistry 36, 1406–1408.
McGeer PL, McGeer EG (1976). Enzymes associated with the
metabolism of catecholamines, acetylcholine and GABA in
human controls and patients with Parkinson’s disease and
Huntington’s chorea. Journal of Neurochemistry 26, 65–76.
The GABAergic system in schizophrenia
McGeer PL, McGeer EG (1977). Possible changes in striatal
and limbic cholinergic systems in schizophrenia. Archives of
General Psychiatry 34, 1319–1323.
McGeer PL, McGeer EG, Wada JA (1971). Glutamic acid
decarboxylase in Parkinson’s disease and epilepsy.
Neurology 21, 1000–1007.
Mizukami K, Sasaki M, Ishikawa M, Iwakiri M, Hidaka S,
Shiraishi H, Iritani S (2000). Immunohistochemical
localization of γ-aminobutyric acidB receptor in the
hippocampus of subjects with schizophrenia. Neuroscience
Letters 283, 101–104.
Mo$ hler H, Benke D, Fritschy JM, Benson J (2000). The
benzodiazepine site of GABAA receptors. In : Martin D,
Olsen R (Eds.), GABA in the Nervous System, The View at
Fifty Years (pp. 97–112). Philadelphia : Lippincott Williams
& Wilkins.
Moore H, West AR, Grace AA (1999). The regulation of
forebrain dopamine transmission, relevance to the
pathophysiology and psychopathology of schizophrenia.
Biological Psychiatry 46, 40–55.
Muly III EC, Szigeti K, Goldman-Rakic PS (1998). D1 receptor
in interneurons of Macaque prefrontal cortex, distribution
and subcellular distribution. Journal of Neuroscience 18,
10553–10565.
Ohnuma T, Augood SJ, Arai H, McKenna PJ, Emson PC
(1999). Measurement of GABAergic parameters in the
prefrontal cortex in schizophrenia, focus on GABA content,
GABAA receptor α-1 subunit messenger RNA and human
GABA transporter-1 (HGAT-1) messenger RNA
expression. Neuroscience 93, 441–448.
Okada Y, Nitsch-Hassler C, Kim JS, Bak IJ, Hassler R (1971).
Role of γ-aminobutyric acid (GABA) in the extrapyramidal
motor system. 1. Regional distribution of GABA in rabbit,
rat, guinea pig and baboon CNS. Experimental Brain
Research 13, 514–518.
Olney JW, Farber NB (1995). Glutamate receptor dysfunction
and schizophrenia. Archives of General Psychiatry 52,
998–1007.
Olsen R, Homanics G (2000). Function of GABAA receptors ;
insights from mutant and knockout mice. In : Martin D,
Olsen R (Eds.), GABA in the Nervous System, The View at
Fifty Years (pp. 81–96). Philadelphia : Lippincott Williams &
Wilkins.
Orkand PM, Kravitz EA (1971). Localization of the sites of γaminobutyric acid (GABA) uptake in lobster nerve-muscle
preparations. Journal of Cell Biology 49, 75–89.
Otsuka M, Iversen LL, Hall ZW, Kravitz EA (1966). Release
of gamma-aminobutyric acid from inhibitory nerves of
lobster. Proceedings of the National Academy of Sciences USA
56, 1110–1115.
Pandey GN, Conley RR, Pandey SC, Goel S, Roberts RC,
Tamminga CA, Chute D, Smialek J (1997). Benzodiazepine
receptors in the post-mortem brain of suicide victims and
schizophrenic subjects. Psychiatry Research 71, 137–149.
Perry EK, Blessed G, Perry RH, Tomlinson BE (1978). Brain
biochemistry in schizophrenia. Lancet 1, 35–36.
177
Perry TL, Hansen S, Gandham SS (1981). Postmortem
changes of amino compounds in human and rat brain.
Journal of Neurochemistry 36, 406–412.
Perry TL, Hansen S, Jones K (1989). Schizophrenia, tardive
dyskinesia, and brain GABA. Biological Psychiatry 25,
200–206.
Perry TL, Kish SJ, Buchanan J, Hansen S (1979). γaminobutyric-acid deficiency in brain of schizophrenic
patients. Lancet 1, 237–239.
Petty F, Sherman AD (1984). Plasma GABA levels in
psychiatric illness. Journal of Affective Disorders 6, 131–138.
Pierri JN, Chaudry AS, Woo TU, Lewis DA (1999).
Alterations in chandelier neuron axon terminals in the
prefrontal cortex of schizophrenic subjects. American Journal
of Psychiatry 156, 1709–1719.
Reveley MA, Gurling HMD, Glass I, Glover V, Sandler M
(1980). Platelet γ-aminobutyric acid-aminotransferase and
monoamine oxidase in schizophrenia. Neuropharmacology
19, 1249–1250.
Reynolds GP, Beasley CL (2001). GABAergic neuronal
subtypes in the human frontal cortex – development and
deficits in schizophrenia. Journal of Chemical Neuroanatomy
22, 95–100.
Reynolds GP, Czudek C, Andrews HB (1990). Deficit and
hemispheric asymmetry of GABA uptake sites in the
hippocampus in schizophrenia. Biological Psychiatry 27,
1038–1044.
Reynolds GP, Stroud D (1993). Hippocampal benzodiazepine
receptors in schizophrenia. Journal of Neural Transmission
(General Section) 93, 151–155.
Rimvall K, Martin DL (1994). The level of GAD67 protein is
highly sensitive to small increases in intraneuronal gammaaminobutyric acid levels. Journal of Neurochemistry 62,
1375–1381.
Rimvall K, Sheikh SN, Martin DL (1993). Effects of increased
gamma-aminobutyric acid levels on GAD67 protein and
mRNA levels in rat cerebral cortex. Journal of
Neurochemistry 60, 714–720.
Roberts E (1972). An hypothesis suggesting that there is a
defect in the GABA system in schizophrenia. Neurosciences
Research Program Bulletin 10, 468–481.
Roberts E, Frankel S (1950). γ-aminobutyric acid in brain, its
formation from glutamic acid. Journal of Biological Chemistry
187, 55–63.
Rochet T, Kopp N, Vedrinne J, Deluermoz S, Debilly G,
Miachon S (1992). Benzodiazepine binding sites and their
modulators in hippocampus of violent suicide victims.
Biological Psychiatry 32, 922–931.
Schiffer WK, Gerasimov M, Hofmann L, Marsteller D, Ashby
CR, Brodie JD, Alexoff DL, Dewey SL (2001). Gamma
vinyl-GABA differentially modulates NMDA antagonistinduced increases in mesocortical versus mesolimbic DA
transmission. Neuropsychopharmacology 25, 704–712.
Schlander M, Thomalske G, Frotscher M (1987). Fine
structure of GABAergic neurons and synapses in the
human dentate gyrus. Brain Research 401, 185–189.
178
B. P. Blum and J. J. Mann
Sesack SR, Hawrylak VA, Melchitzky DS, Lewis DA (1998).
Dopamine innervation of a subclass of local circuit neurons
in monkey prefrontal cortex : ultrastructural analysis of
tyrosine hydroxylase and parvalbumin immunoreactive
structures. Cerebral Cortex 8, 614–622.
Sesack SR, Snyder CL, Lewis DA (1995). Axon terminals
immunolabeled for dopamine or tyrosine hydroxylase
synapse on GABA-immunoreactive dendrites in rat and
monkey cortex. Journal of Comparative Neurology 363,
264–280.
Sheikh SN, Martin DL (1998). Elevation of brain GABA levels
with vigabatrin (gamma-vinylGABA) differentially affects
GAD65 and GAD67 expression in various regions of rat
brain. Journal of Neuroscience Research 52, 736–741.
Sheikh SN, Martin SB, Martin DL (1999). Regional
distribution and relative amounts of glutamate
decarboxylase isoforms in rat and mouse brain.
Neurochemistry International 35, 73–80.
Sherif F, Eriksson L, Oreland L (1992). Gamma-aminobutyrate
aminotransferase activity in brains of schizophrenic
patients. Journal of Neural Transmission (General Section) 90,
231–240.
Sherman AD, Davidson AT, Baruah S, Hegwood TS, Waziri
R (1991). Evidence of glutamatergic deficiency in
schizophrenia. Neuroscience Letters 121, 77–80.
Simpson MDC, Royston MC, Slater P, Deakin JFW (1992a).
Neurochemical abnormalities of the cerebral cortex in
schizophrenia. Schizophrenia Research 6, 133–134.
Simpson MDC, Slater P, Deakin JFW (1998a). Comparison of
glutamate and gamma-aminobutyric acid uptake binding
sites in frontal and temporal lobes in schizophrenia.
Biological Psychiatry 44, 423–427.
Simpson MDC, Slater P, Deak JFW, Gottfries CG, Karlsson I,
Grenfeldt B, Crow TJ (1998b). Absence of basal ganglia
amino acid neuron deficits in schizophrenia in three
collections of brains. Schizophrenia Research 31, 167–175.
Simpson MDC, Slater P, Deakin JFW, Royston MC, Skan WJ
(1989). Reduced GABA uptake sites in the temporal lobe in
schizophrenia. Neuroscience Letter 107, 211–215.
Simpson MDC, Slater P, Royston MC, Deakin JFW (1992b).
Regionally selective deficits in uptake sites for glutamate
and gamma-aminobutyric acid in the basal ganglia in
schizophrenia. Psychiatry Research 42, 273–282.
Smith Y, Parent A, Seguela P, Descarries L (1987a).
Distribution of GABA-immunoreactive neurons in the basal
ganglia of the squirrel monkey (Saimiri sciureus). Journal of
Comparative Neurology 259, 50–64.
Smith Y, Seguela P, Parent A (1987b). Distribution of GABAimmunoreactive neurons in the thalamus of the squirrel
monkey (Saimiri sciureus). Neuroscience 22, 579–591.
Sorvari H, Soininen H, Paljarvi L, Karkola K, Pitkanen A
(1995). Distribution of parvalbumin-immunoreactive cells
and fibers in the human amygdaloid complex. Journal of
Comparative Neurology 360, 185–212.
Spokes EGS (1979). An analysis of factors influencing
measurements of dopamine, noradrenaline, glutamate
decarboxylase and choline acetylase in human post-mortem
brain tissue. Brain 102, 333–346.
Spokes EG (1980). Neurochemical alterations in Huntington’s
chorea, a study of post-mortem brain tissue. Brain 103,
179–210.
Spokes EGS, Garrett NJ, Iversen LL (1979). Differential effects
of agonal status on measurements of GABA and glutamate
decarboxylase in human post-mortem brain tissue from
control and Huntington’s chorea subjects. Journal of
Neurochemistry 33, 773–778.
Spokes EGS, Garrett NJ, Rossor MN, Iversen LL (1980).
Distribution of GABA in post-mortem brain tissue from
control, psychotic and Huntington’s chorea subjects. Journal
of the Neurological Sciences 48, 303–313.
Squires RF, Lajtha A, Saederup E, Palkovits M (1993).
Reduced [$H]flunitrazepam binding in cingulate cortex and
hippocampus of postmortem schizophrenic brains : is
selective loss of glutamatergic neurons associated with
major psychoses? Neurochemical Research 18, 219–223.
Squires RF, Saederup E (1991). A review of evidence for
GABergic predominance\glutamatergic deficit as a
common etiological factor in both schizophrenia and
affective psychoses, more support for a continuum
hypothesis of ‘ functional ’ psychosis. Neurochemical Research
16, 1099–1111.
Sternberg DE (1980). CSF γ-aminobutyric acid (GABA) in
schizophrenia, Proceedings of the 133rd American Psychiatric
Association, pp. 80–81.
Stevens J, Wilson K, Foote W (1974). GABA blockade,
dopamine and schizophrenia, experimental studies in the
cat. Psychopharmacologia (Berlin) 39, 105–119.
Stevens JR (1999). Epilepsy, schizophrenia, and the extended
amygdala. Annals of the New York Academy of Sciences 877,
548–561.
Stocks GM, Cheetham SC, Crompton MR, Katona CL,
Horton RW (1990). Benzodiazepine binding sites in
amygdala and hippocampus of depressed suicide victims.
Journal of Affective Disorders 18, 11–15.
Stone DJ, Walsh J, Benes FM (1999). Localization of cells
preferentially expressing GAD(67) with negligible
GAD(65) transcripts in the rat hippocampus. A double in
situ hybridization study. Brain Research Molecular Brain
Research 71, 201–209.
Todtenkopf MS, Benes FM (1998). Distribution of glutamate
decarboxylase immunoreactive puncta on pyramidal and
'&
nonpyramidal neurons in hippocampus of schizophrenic
brain. Synapse 29, 323–332.
Toru M, Watanabe S, Shibuya H, Nishikawa T, Noda K,
Mitsushio H, Ichikawa H, Kurumaji A, Takashima M,
Mataga N, Ogawa A (1988). Neurotransmitters, receptors
and neuropeptides in post-mortem brains of chronic
schizophrenic patients. Acta Psychiatrica Scandinavica 78,
121–137.
Udenfriend S (1950). Identification of γ-aminobutyric acid in
brain by the isotope derivative method. Journal of Biological
Chemistry 187, 65–69.
van Kammen DP (1979). The dopamine hypothesis of
schizophrenia revisited. Psychoneuroendocrinology 4, 37–46.
The GABAergic system in schizophrenia
van Kammen DP, Petty F, Kelley ME, Kramer GL, Barry EJ,
Yao JK, Gurklis JA, Peters JL (1998). GABA and brain
abnormalities in schizophrenia. Psychiatry Research,
Neuroimaging Section 82, 25–35.
van Kammen DP, Sternberg DE, Hare TA, Waters RN,
Bunney Jr. WE (1982). CSF levels of γ-aminobutyric acid in
schizophrenia. Low values in recently ill patients. Archives
of General Psychiatry 39, 91–97.
Vincent SL, Adamec E, Sorensen I, Benes FM (1994). The
effects of chronic haloperidol administration on GABAimmunoreactive axon terminals in rat medial prefrontal
cortex. Synapse 17, 26–35.
Volk DW, Austin MC, Pierri JN, Sampson AR, Lewis DA
(2000). Decreased glutamic acid decarboxylase messenger
'(
RNA expression in a subset of prefrontal cortical γaminobutyric acid neurons in subjects with schizophrenia.
Archives of General Psychiatry 57, 237–245.
Volk DW, Austin MC, Pierri JN, Sampson AR, Lewis DA
(2001). GABA transporter-1 mRNA in the prefrontal cortex
in schizophrenia, decreased expression in a subset of
neurons. American Journal of Psychiatry 158, 256–265.
Wassef AA, Dott SG, Harris A, Brown A, O’Boyle M, Meyer
179
III WJ, Rose RM (1999). Critical review of GABA-ergic
drugs in the treatment of schizophrenia. Journal of Clinical
Psychopharmacology 19, 222–232.
White HL, Davidson JR, Miller RD, Faison LD (1980). Platelet
γ-aminobutyrate-α-ketoglutarate transaminase (GABA-T) in
schizophrenia. American Journal of Psychiatry 137, 733–734.
Woo TU, Miller JL, Lewis DA (1997). Schizophrenia and the
parvalbumin-containing class of cortical local circuit
neurons. American Journal of Psychiatry 154, 1013–1015.
Woo TU, Whitehead RE, Melchitzky DS, Lewis DA (1998). A
subclass of prefrontal γ-aminobutyric acid axon terminals
are selectively altered in schizophrenia. Proceedings of the
National Academy of Sciences USA 95, 5341–5346.
Zander KJ, Fischer B, Zimmer R, Ackenheil M (1981). Longterm neuroleptic treatment of chronic schizophrenic
patients, clinical and biochemical effects of withdrawal.
Psychopharmacology 73, 43–47.
Zimmer R, Teelken AW, Meier KD, Ackenheil M, Zander KJ
(1981). Preliminary studies on CSF gamma-aminobutyric
acid levels in psychiatric patients before and during
treatment with different psychotropic drugs. Progress in
Neuro-Psychopharmacology 4, 613–620.