Neurological Disorders: Molecules, Mechanisms and Therapeutics 433

The neuronal pathology of schizophrenia:

molecules and mechanisms
G.P. Reynolds and M.K. Harte1
Division of Psychiatry and Neuroscience, Queen’s University Belfast, Belfast BT9 7BL, Northern Ireland, U.K.

Abstract
There is an accumulation of evidence for abnormalities in schizophrenia of both the major neurotransmitter
systems of the brain – those of GABA (γ -aminobutyric acid) and glutamate. Initial studies have found deficits
in the putative neuronal marker, N-acetylaspartate, in a number of brain regions in schizophrenia. The animal
models have provided some interesting correlates and discrepancies with these findings. The deficit in
inhibitory interneurons within structures implicated in schizophrenic symptomatology may well have direct
functional relevance, and can be induced by animal models of the disease such as subchronic phencyclidine
administration or social isolation. Their association with these animal models suggests an environmental
involvement. A loss of glutamatergic function in schizophrenia is supported by decreases in markers for
the neuronal glutamate transporter in striatal structures that receive cortical glutamatergic projections.
Deficits in the VGluT1 (vesicular glutamate transporter-1) in both striatal and hippocampal regions support
this observation, and the association of VGluT1 density with a genetic risk factor for schizophrenia points
to genetic influences on these glutamatergic deficits. Further studies differentiating neuronal loss from
diminished activity and improved models allowing us to determine the temporal and causal relationships
between GABAergic and glutamatergic deficits will lead to a better understanding of the processes underlying
the neuronal pathology of schizophrenia.

Introduction (positron emission tomography) studies of D2 receptors con-

From the middle of the last century, research into the brain
in schizophrenia has attempted to implicate in the disorder
many different neurochemicals, particularly neurotransmit-
ters and other compounds with direct effects on neuronal
activity. Few of these theories have lasted long, although for
over 30 years the dopamine hypothesis of schizophrenia has
been an important stimulus to research. This was first based
on the ability of amphetamine and dopamine agonists to
induce a schizophreniform psychosis, and was strengthened
by the finding that almost all antipsychotic drugs are effective
antagonists of the dopamine D2 receptor subtype. Sub-
sequently, the higher density of dopamine D2 receptors found
in post-mortem brain from schizophrenic patients led to the
formulation of a modified dopamine hypothesis in which
elevated D2 receptors were proposed to underlie the positive
symptoms of schizophrenia. However, an up-regulation of
D2 receptors is seen in animals after chronic administration
of antipsychotic drugs, a treatment that most schizophrenic
patients will inevitably have received. It seemed likely,
therefore, that the elevation seen was a consequence of drug
treatment and unrelated to the disease process. Despite some
remaining inconsistencies, most post-mortem and PET
Key words: γ -aminobutyric acid (GABA), glutamate, isolation rearing, N-acetylaspartate,
phencyclidine, schizophrenia.
Abbreviations used: CB, calbindin; CBP, calcium-binding protein; CR, calretinin; GABA, γ -
aminobutyric acid; MRS, magnetic resonance spectroscopy; NAA, N-acetylaspartate; NAAG,
N-acetylaspartylglutamate; NMDA, N-methyl-D-aspartate; PCP, phencyclidine; PET, positron
emission tomography; PV, parvalbumin; VGluT1, vesicular glutamate transporter 1.
1
To whom correspondence should be addressed (email m.k.harte@qub.ac.uk).
clude that there is no elevation in drug-free schizophrenic
patients.
Nevertheless, dopamine dysfunction is still implicated in
schizophrenia, although with no evidence that it might repre-
sent a primary pathology in the disease. Post-mortem studies
show elevations of dopamine concentration in the amygdala
[1], and PET has demonstrated increases in stimulated dopa-
mine release in the striatum [2] in patients with schizophrenia.
These findings are likely, however, to be secondary changes
reflecting dysfunction and/or deficits in other neuronal
systems.
Certainly there is substantial evidence for neuronal deficits
in schizophrenia. Neuroimaging has identified increases in
ventricular size, reflecting volume decreases in brain regions,
particularly in the structures of the neocortex and medial
temporal lobe. Many functional imaging studies, including
PET determination of energy metabolism, have also impli-
cated these structures in the various symptoms of schizo-
phrenia. These macroscopic findings are reflected at the
neuronal level by substantial evidence for neuronal deficits
and dysfunction. This short review will concentrate on some
evidence from our own studies demonstrating changes in
neuronal systems containing the major inhibitory and excit-
atory neurotransmitters of the human brain: respectively
GABA (γ -aminobutyric acid) and glutamate. We shall also
draw on similarities and discrepancies with two animal
models of the disease, which provide some clues as to the
aetiological origins of the abnormalities observed in schizo-
phrenia.

C 2007 Biochemical Society
434 Biochemical Society Transactions (2007) Volume 35, part 2
GABAergic deficits
Post-mortem studies have provided evidence for abnormal-
ities of the GABAergic system in schizophrenia. Deficits in
GABA-containing neurons are consistently reported, parti-
cularly in the frontal cortex and hippocampus. Benes et al. [3]
found reduced density of interneurons in CA2 and CA3 re-
gions in the hippocampus of brains of schizophrenia patients,
including drug-free patients, suggesting that deficits in inter-
neurons may play a contributory role in the pathophysiology
of schizophrenia. Consistent with this interpretation, neuro-
chemical studies have demonstrated a deficit of GABAergic
uptake sites in the hippocampus [4], while other reports
showed a widespread compensatory up-regulation of post-
synaptic-specific GABAA receptor-binding activity through-
out most subfields of the hippocampal formation of schizo-
phrenia patients [5].
However, our understanding of the GABA system in
schizophrenia must take into account the fact that there are
several different types of non-pyramidal neuron that utilize
GABA as a neurotransmitter, and that each subset of neurons
may show a unique pattern of change. The CBPs (calcium-
binding proteins) PV (parvalbumin), CB (calbindin) and CR
(calretinin) can be used as markers for specific subpopulations
of GABAergic interneurons in the brain. It is now generally
accepted that the major neuronal role of these CBPs is the
buffering and transport of calcium ions and the regulation
of various enzyme systems. As cellular degeneration is often
accompanied by impaired calcium homoeostasis, a protective
role for CBPs has been postulated [6].
PV is particularly interesting in that it shows a late onset
of expression, occurring after GABAergic neurons have been
formed [7], imparting a period during which these cells may
be particularly susceptible to neurotoxic insult. We have pre-
viously reported deficits of PV-immunoreactive cells both in
the frontal cortex [8,9] and, more profoundly, in the hippo-
campus in schizophrenia [10]. These GABAergic deficits in
schizophrenia could well be an initial deficit, perhaps of
neurodevelopmental origin, that subsequently results in fur-
ther progressive neuronal dysfunction and the development
of schizophrenia in the second or third decade of life.
These GABAergic deficits are unrelated to antipsychotic drug
treatment, age or duration of illness [10,11].
Deficits in CB have also been reported in schizophrenia
[8,12]. In contrast, most of the studies using human post-
mortem tissue have reported that CR immunoreactive
neurons are unaffected in schizophrenia [10,12,13].
Although current animal models of schizophrenia are in-
evitably limited, behavioural and pharmacological studies
concentrate on two approaches with substantial validity.
These may be pharmacological [subchronic administration of
an NMDA (N-methyl-D-aspartate) antagonist such as PCP
(phencyclidine)] or non-pharmacological (isolation rearing
from weaning) animal models.
PCP has been shown to induce a psychosis in humans
that closely resembles schizophrenia in both positive and
negative symptoms and also exacerbates the symptoms in


C 2007 Biochemical Society
stabilized patients [14]. Thus it has been proposed that
NMDA receptor hypofunction may be viewed as a model for
the disease mechanism, which could explain the symptoms
and course of schizophrenia [15]. In reviewing the neuropsy-
chopharmacology of PCP, Jentsch and Roth [16] have con-
cluded that long-term repeated administration of PCP to both
humans and animals models the symptoms of schizophrenia
better than acute treatment. In animals long-term repeated
administration of PCP has been shown to mimic certain
aspects of the disorder. Utilizing a subchronic or chronic
intermittent treatment regime, our laboratory and others have
found that, compared with vehicle-treated controls, animals
receiving PCP exhibit reductions in PV-immunoreactive cells
in the frontal cortex and hippocampus [17,18].
Isolation rearing of rats is a non-lesion manipulation that
leads to deficits in PPI (prepulse inhibition) of the startle re-
flex and other behavioural and neurochemical alterations
reminiscent of schizophrenia [19].
Compared with socially housed rats, isolated rats exhibited
reductions in PV- and CB-immunoreactive cells in the hippo-
campus, with no significant change in CR [20]. These findings
demonstrate selective abnormalities of subpopulations of
GABAergic interneurons in the hippocampus of isolation
reared rats, which resemble the neuronal deficits seen in this
region in schizophrenia.
In conclusion, we find a deficit of PV-immunoreactive hip-
pocampal neurons in animals receiving subchronic PCP [18]
as well as in those reared in isolation [20]. These findings mi-
mic the pathology of the disease in their specificity; no defi-
cits of CR-containing neurons are observed in these animal
models.
NAA (N-acetylaspartate) and
glutamatergic deficits
NAA is an amino acid that is present in high concentrations
in the CNS (central nervous system), with intraneuronal con-
centrations between 10 and 14 mM. NAA is synthesized in
neuronal mitochondria from acetyl-coA and aspartate by
the enzyme NAA transferase. In the adult rat brain, NAA
is found almost exclusively in neurons, being absent from
mature glial cells [21]. Although it is 50 years since NAA was
first discovered [22] there is still no consensus on its principal
metabolic or neurochemical function. It has been shown to be
involved in myelin synthesis [23], osmoregulation [24], and a
recent study suggests a possible role as an anti-inflammatory
[25]. NAA is also a precursor for the biosynthesis of NAAG
(N-acetylaspartylglutamate), a neuropeptide with actions at
NMDA and metabotropic glutamate receptors [26]. NAA
correlates with neuronal density and as such it has been
widely used as a neuronal marker, with deficits in NAA
thought to reflect neuronal loss. However, previous studies
demonstrating the reversibility of NAA following acute brain
injury [27,28] indicate that NAA may be a marker of both
neuronal loss and cellular dysfunction [29].
The current focus on NAA dysfunction in neuropsychiat-
ric disorders stems from the prominence of the NAA signal
in MRS (magnetic resonance spectroscopy). This has led to
 Neurological Disorders: Molecules, Mechanisms and Therapeutics 435

Table 1 Deficits in NAA in schizophrenic post-mortem tissue, ing a glutamatergic hypofunction [40]. More recently, we
correlates and discrepancies with two animal models have found that striatal deficits in glutamatergic innervation
Isolation are reflected in losses of immunoreactivity of the neuronal
Schizophrenia PCP model rearing model excitatory amino acid transporter 3. We have gone on to
investigate a more specific and selective marker for glutama-
Frontal cortex – – – tergic terminals in the brain, namely VGluT1 (vesicular gluta-
Temporal cortex ↓ ↓ ↓ mate transporter 1). VGluTs are glutamate transporters local-
Striatum ↓ – – ized on the membrane of synaptic vesicles in the presynaptic
Hippocampus ↓ – – neuron. VGluT1, in particular, is exclusive to glutamatergic
terminals and undetectable in other cell types and neuronal
compartments [41–43], and as such provides a very selective
a wide number of MRS studies looking at NAA in brain re- marker for presynaptic glutamatergic terminals in the brain.
gions in neurodegenerative and psychiatric diseases, notably We find deficits in VGluT1 in both striatal and hippocampal
schizophrenia. Neuronal dysfunction, determined by MRS regions in schizophrenia [44].
measurement of NAA deficits in vivo, is seen in cortical Preliminary studies demonstrated an association between
and medial temporal structures in schizophrenia [30]. We and VGluT1 density and a genetic risk factor for schizophrenia
others have shown that these deficits are unlikely to be due to [45]. A single polymorphism in neuregulin1, identified as
treatment with antipsychotic medication [31–33] and may in a risk factor for the disease by Stefansson et al. [46], was
fact be related to the disease process. associated with VGluT1 in human post-mortem striatal and
In our laboratory we use a sensitive HPLC method to hippocampal tissue, with the at-risk allele predicting lower
determine levels of NAA in post-mortem tissue. In the first densities of the glutamatergic marker. These findings provide
of these studies we demonstrated a significant deficit in NAA a link between the genotype of a genetic risk factor for schizo-
in the temporal cortex in schizophrenia, with no change in phrenia and glutamatergic innervation. The results provide
the frontal cortex [34]. Interestingly, we also find selective further evidence that the glutamatergic deficits that we see in
deficits in NAA in the temporal cortex both in the PCP animal our human studies may, in part, be genetically determined,
model [35] and in the isolation rearing model [36]. There was and may indicate why these deficits are not seen in the
no change in NAA in any other region examined in these environmental animal models we studied.
models, namely the frontal cortex, striatum or hippocampus Further studies differentiating neuronal loss from dimin-
(Table 1). ished activity, and improved models allowing us to determine
Specific temporal cortex deficits in NAA can be induced the temporal and causal relationships between GABAergic
by both pharmacological and non-pharmacological effects in and glutamatergic deficits, as well as identifying the role
animals. This raises the question whether temporal cortical of specific genetic and environmental factors, will lead to a
NAA deficits might reflect an environmental aetiology of better understanding of the processes underlying the neuronal
schizophrenia. pathology of schizophrenia.
In our recent studies, we have found that in schizophrenia
the NAA deficits extend to regions of the medial temporal
lobe, the striatum, the hippocampus and amygdala (L.M. References
1 Reynolds, G.P. (1983) Nature 305, 527–529
Reynolds, personal communication). It is of interest to note 2 Laruelle, M. and Abi-Dargham, A. (1999) J. Psychopharmacol. 13,
that the NAA deficits that we see in the striatum and 358–371
hippocampus in the human post-mortem tissue are not seen in 3 Benes, F.M., Kwok, E.W., Vincent, S.L. and Todtenkopf, M.S. (1998)
Biol. Psychiatry 44, 88–97
either of the environmental animal models mentioned above 4 Reynolds, G.P., Czudek, C. and Andrews, H.B. (1990) Biol. Psychiatry 27,
(Table 1). Thus, in contrast with the findings in temporal 1038–1044
cortex, this leads us to speculate whether hippocampal (and 5 Benes, F.M., Khan, Y., Vincent, S.L. and Wickramasinghe, R. (1996)
Synapse 22, 338–349
subcortical) NAA deficits reflect a non-environmental, per- 6 Heizmann, C.W. (1992) Trends Neurosci. 15, 259–264
haps genetic, aetiology of schizophrenia. 7 Solbach, S. and Celio, M.R. (1991) Anat. Embryol. 184, 103–124
NAA has been suggested to reflect glutamate concen- 8 Beasley, C.L., Zhang, Z.J., Patten, I. and Reynolds, G.P. (2002)
Biol. Psychiatry 52, 708–715
trations and thus may provide a marker of glutamatergic 9 Beasley, C.L. and Reynolds, G.P. (1997) Schizophr. Res. 24, 349–355
neuronal function [37]. In support of this are several observ- 10 Zhang, Z.J. and Reynolds, G.P. (2002) Schizophr. Res. 55, 1–10
ations implicating abnormalities of glutamate in the brain 11 Cahir, M., Costello, I., King, D.J. and Reynolds, G.P. (2005) Neurosci. Lett.
373, 57–60
in schizophrenia. We have found a deficit in the number of 12 Cotter, D., Landau, S., Beasley, C., Stevenson, R., Chana, G., MacMillan, L.
glutamate uptake sites in the caudate nucleus, putamen and and Everall, I. (2002) Biol. Psychiatry 57, 377–386
nucleus accumbens in schizophrenia [38], which we interpret 13 Reynolds, G.P. and Beasley, C.L. (2001) J. Chem. Neuroanat. 22, 95–100
14 Javitt, D.C. and Zukin, S.R. (1991) Am. J. Psychiatry 148, 1301–1308
as a deficit of cortico-striatal glutamatergic innervation. 15 Olney, J.W., Newcomer, J.W. and Farber, N.B. (1999) J. Psychiatry Res.
Consistent with this is the increase in the density of NMDA 33, 523–533
receptors in the putamen from schizophrenic patients when 16 Jentsch, J.D. and Roth, R.H. (1999) Neuropsychopharmacology 20,
201–225
compared with controls [39]. Such increases in glutamate 17 Cochran, S.M., Kennedy, M., McKerchar, C.E., Steward, L.J., Pratt, J.A. and
receptors have been interpreted as compensatory and reflect- Morris, B.J. (2003) Neuropsychopharmacology 28, 265–275

C 2007 Biochemical Society
436 Biochemical Society Transactions (2007) Volume 35, part 2
18 Abdul-Monim, Z., Neill, J.C. and Reynolds, G.P. (2006)
J. Psychopharmacol., doi:10.1177/0269881106067097
19 Weiss, I.C. and Feldon, J. (2001) Psychopharmacology 156, 305–326
20 Harte, M.K., Powell, S.B., Swerdlow, N.R., Geyer, M.A. and Reynolds, G.P.
(2006) Schizophr. Res. 81, 210–211
21 Urenjak, J., Williams, S.R., Gadian, D.G. and Noble, M. (1993) J. Neurosci.
13, 981–989
22 Tallan, H.H., Moore, S. and Stein, W.H. (1956) J. Biol. Chem. 219,
257–264
23 Chakraborty, G., Mekala, P., Yahya, D., Wu, G. and Ledeen, R.W. (2001)
J. Neurochem. 78, 736–745
24 Baslow, M.H. (2003) Neurochem. Res. 28, 941–953
25 Rael, L.T., Thomas, G.W., Bar-Or, R., Craun, M.L. and Bar-Or, D. (2004)
Biophys. Res. Commun. 2, 847–853
26 Neale, J.H., Bzdega, T. and Wroblewska, B. (2000) J. Neurochem. 75,
443–452
27 De Stefano, N., Matthews, P.M. and Arnold, D.L. (1995)
Magn. Reson. Med. 34, 721–727
28 Kalra, S., Cashman, S.H., Genge, A. and Arnold, D.L. (1998) NeuroReport
9, 1757–1761
29 Demougeot, C., Garnier, P., Mossiat, C., Bertrand, N., Giroud, M.,
Beley, A. and Marie, C. (2001) J. Neurochem. 77, 408–415
30 Rowland, L., Bustillo, J.R. and Lauriello, J. (2001) Semin. Clin.
Neuropsychiatry 6, 121–130
31 Lindquist, D.M., Hawk, R.M., Karson, C.N. and Komoroski, R.A. (2000)
Magn. Reson. Med. 43, 355–358
32 Harte, M.K., Bachus, S.B. and Reynolds, G.P. (2005) Schizophr. Res. 75,
303–308
33 Bustillo, J., Barrow, R., Paz, R., Tang, J., Seraji-Bozorgzad, N., Moore, G.J.,
Bolognani, F., Lauriello, J., Perrone-Bizzozero, N. and Galloway, M.P.
(2006) Neuropsychopharmacology 31, 751–756


C 2007 Biochemical Society
34 Nudmamud, S., Reynolds, L.M. and Reynolds, G.P. (2003) Biol. Psychiatry
53, 1138–1141
35 Reynolds, L.M., Cochran, S.M., Morris, B.J., Pratt, J.A. and Reynolds, G.P.
(2005) Schizophr. Res. 73, 147–152
36 Harte, M.K., Powell, S.B., Reynolds, L.M., Swerdlow, N.R., Geyer, M.A.
and Reynolds, G.P. (2004) Biol. Psychiatry 56, 296–299
37 Petroff, O.A., Errante, L.D., Rothman, D.L., Kim, J.H. and Spencer, D.D.
(2002) Ann. Neurol. 52, 635–642
38 Aparicio-Legarza, M.I., Cutts, A.J., Davis, B. and Reynolds, G.P. (1997)
Neurosci. Lett. 232, 13–16
39 Aparicio-Legarza, M.I., Davis, B., Hutson, P.H. and Reynolds, G.P. (1998)
Neurosci. Lett. 241, 143–146
40 Deakin, J.F.W., Slater, P., Simpson, M.D.C., Gilchrist, A.C., Skon, W.J.,
Royston, M.C., Reynolds, G.P. and Cross, C.J. (1989) J. Neurochem. 52,
1781–1786
41 Bellocchio, E.E., Hu, H., Pohorille, A., Chan, J., Pickel, V.M. and Edwards,
R.H. (1998) J. Neurosci. 18, 8648–8459
42 Bellocchio, E.E., Reimer, R.J., Fremeau, Jr, R.T. and Edwards, R.H. (2000)
Science 289, 957–960
43 Fremeau, Jr, R.T., Troyer, M.D., Pahner, I., Nygaard, G.O., Tran, C.H.,
Reimer, R.J., Bellocchio, E.E., Fortin, D., Storm-Mathisen, J. and
Edwards, R.H. (2001) Neuron 31, 247–260
44 Piyabhan, P. and Reynolds, G.P. (2006) J. Psychopharm. 20, MB26
45 Reynolds, G.P., Piyabhan, P., Nudmamud-Thanoi, S., Dempster, E. and
Collier, D.A. (2006) Biol. Psychiatry 59, 123S
46 Stefansson, H., Sigurdsson, E., Inthorsdottir, V., Bjornsdottir, S.,
Sigmundsson, T., Ghosh, S., Brynjolfsson, J., Gunnarsdottir, S.,
Ivarsson, O., Chou, T.T. et al. (2002) Am. J. Hum. Genet. 71, 877–892
Received 31 August 2006