The high incidence and serious nature of schizo-phrenia has led to extensive efforts to discover macro-or microscopic lesions in the brain that could cause theillness. Failure to identify such lesions [1] has lentsupport to a number of hypotheses that proposechanges in the molecular architecture of the brain as the basis of schizophrenia. Both neuroimaging

techniques and brain tissue obtained post-mortemhave been used to gain direct evidence to support suchhypotheses. It is now pertinent to review the outcomesof key studies using these two techniques and howthey are influencing current thinking on the neuro-biology of schizophrenia. This review will not addressthe basic biochemistry of neurochemical systemswhich are extensively reviewed elsewhere [2].

The dopamine hypothesis of schizophrenia

The dopamine (DA) hypothesis of schizophrenia is along-standing hypothesis that proposes that over-active

Signal transmission, rather than reception, isthe underlying neurochemical abnormality inschizophrenia

Brian Dean

Objective: This review aims to summarise the outcome of studies on changes in the molecular architecture of the brain of subjects with schizophrenia and formulate ahypothesis on mechanisms involved in the pathology of the illness.Method: The outcomes from key studies using neuroimaging techniques and tissueobtained post-mortem that have been directed toward identifying abnormalities in themolecular architecture of the brain in subjects with schizophrenia were summarised.Using the results from these studies hypotheses were formulated on the underlyingpathological process that precipitate schizophrenia.Results: Studies using neuroimaging techniques or tissue obtained post-mortemhave revealed changes in the dopaminergic, serotoninergic, glutamatergic,GABAergic and cholinergic systems of the brain in schizophrenia. Some of thesestudies have identified abnormalities in presynaptic proteins or functioning that maybe central to the pathology of schizophrenia.Conclusions: There appears to be diverse changes in the molecular cytoarchitec-ture of the brains from subjects with schizophrenia. It could be that it is by affectingthese multiple systems that the atypical antipsychotic drugs produce their improvedclinical outcomes. Abnormal functioning of presynaptic processes could be central to the pathology of schizophrenia. If the ‘presynaptic’ hypothesis is proven, futureantipsychotic drug design should be directed away from post-synaptic receptorantagonism toward the modulating the functions of presynaptic neurones.Key words: neuropathology, schizophrenia, signal transmission.

Australian and New Zealand Journal of Psychiatry 2000; 34:560–569

Brian Dean, Research Fellow

The Rebecca L. Cooper Research Laboratories, The MentalHealth Research Institute of Victoria, Locked Bag 11, Parkville,Victoria 3052, Australia. Email: B.Dean@papyrus.mhri.edu.au

Received 13 October 1999; revised 6 January 2000; accepted12 January 2000.

dopaminergic neurones cause the psychoses associ-ated with schizophrenia [3]. Support for the DAhypothesis comes from the effects of drugs that targetthe dopaminergic systems of the brain [4]. In particu-lar, it has been observed that drugs which activatedopaminergic neurones may induce or exacerbatepsychoses while those that reduce the activity of theseneurones can lessen psychoses. Since the formulationof the DA hypothesis of schizophrenia there has beena concerted effort to obtain direct experimental evi-dence to prove its validity.

Initial attempts to confirm a role for DA in thepathology of schizophrenia were encouraging, withincreased levels of DA being reported in the nucleusaccumbens [5] and amygdala [6] from subjects withschizophrenia. Subsequently, findings on DA [7] andother neurotransmitters, as well as enzymes involvedin the synthesis and degradation of neurotransmitters,have been inconsistent. It is now widely accepted thatneurotransmitters and enzyme activity are not stablein the brain post-mortem [8]. Hence, it is likely the apparent variability in post-mortem studies mea-suring these variables could have been due to vari-ability in tissue collection. This makes it difficult tointerpret the early results on DA levels in the brainsof subjects with schizophrenia.

The inconsistencies in results from measuringneurotransmitters and/or enzyme activity haveincreasingly led to the investigation of more stablecomponents of the dopaminergic system, in particu-lar the levels of DA receptors in the brain of subjectswith schizophrenia. Initially it was reported that DA2

receptor was increased in the caudate nucleus fromsubjects with schizophrenia [9]. These results led tothe proposal that there was an overabundance of DA2

receptors in the brain of subjects with schizophreniaand, in the presence of normal or increased levels of DA, resulted in a super-sensitive state and over-activation of DA responsive neurones. However, ithas been reported that DA2 receptors are not alteredin schizophrenic subjects never treated with antipsy-chotic drugs [10]. This suggested that the increase indensity of DA2 receptors in tissue from subjects withschizophrenia was due to antipsychotic drug treat-ment during life.

Positron emission tomography (PET) has also beenused to measure DA2 receptors in the human brain. As with post-mortem studies, some studies using PET have reported an increase in the density of DA2

receptors in the caudate nucleus of subjects withschizophrenia, whether or not they were naive to anti-psychotic drugs [9]. These data appeared to support the

argument that an increase in DA2 receptors in thecaudate-putamen is involved in the pathology ofschizophrenia. However, later PET studies could notconfirm an increase in DA2 receptors in drug naivesubjects with schizophrenia. Hence, the use of PEThas not fully resolved whether there is an increase in DA2 receptors in the brain of subjects with schizophrenia [9].

Early studies on DA receptors based their conclu-sions on the hypothesis that there were two DAreceptors [11]. One of these receptors, the DA2 recep-tor, was partly defined as the site that bound antipsy-chotic drugs with high affinity. Cloning studies havenow shown that the pharmacologically defined DA2

receptor encompasses a family of receptors, nowtermed the DA2-like receptors [12]. This family ofreceptors is made up of the D2, D3 and D4 receptors,each being a separate gene product. Moreover, the D2

receptor undergoes post-transcriptional modificationproducing a long (D2L) and short (D2S) form of thereceptor. Pharmacological agents that are totally spe-cific to a single DA receptor have yet to be devel-oped. Thus, the study of each individual DA receptorin schizophrenia has involved measuring levels ofmessenger RNA (mRNA) encoding each receptor inbrain tissue obtained postmortem. One such studyreported increased levels of mRNA for the D2 recep-tor in the ventral orbital gyrus, inferior orbital gyrusand caudate nucleus from subjects with schizophre-nia [13]. This would, at least, suggest that anyincrease in DA2 receptor protein in the brain fromsubjects with schizophrenia has resulted from anincrease in the rate of expression of that receptor. Thedevelopment of radioactive ligands and/or effectivereceptor antibodies that are selective for the D2 recep-tor will be required to pursue this hypothesis.

Using an indirect pharmacological subtractiontechnique, a number of studies have reported a rela-tively high concentration of the D4 receptor in thehuman caudate-putamen and an increase of thatreceptor in that brain region in schizophrenia [9].This was taken as being highly significant as it hadbeen suggested that clozapine binds with high affinity to the D4 receptor [14] and may thereforehave some of its unique therapeutic effects [15] byselectively blocking that receptor. However, con-cerns about the validity of these results have arisenbecause there appears to be no mRNA for the D4

receptor in the caudate-putamen [16] making itunlikely this region would contain high levels of D4

receptor, as suggested by the subtraction technique.In addition, more direct pharmacological approaches

B. DEAN 561

to measuring the D4 receptor failed to show differ-ences in schizophrenia [9] and methodological problems associated with the use of the subtractiontechnique suggest that method would not accuratelymeasure the density of D4 receptors in schizophrenia[17]. Significantly, mRNA encoding the D4 receptoris present in the frontal cortex and has been reportedas decreased [16,18] or unchanged [19] in schizo-phrenia. Hence, current data on mRNA for the D4

receptor favour a change in that receptor in thefrontal cortex in schizophrenia.

Decreases in mRNA for the D3 receptor have beenreported in Brodmann’s areas 1 through 5 [20] and 11[16] of the frontal cortex, but not in the striatum ornucleus accumbens [16], from subjects with schizo-phrenia. This is a particularly intriguing finding asthere is growing evidence that suggests the D3 recep-tor is the presynaptic autoreceptor in certain areas ofthe brain [21,22]. The role of the DA autoreceptor is to regulate DA synthesis and release [23]. Hence,alterations in the D3 autoreceptor in the brain of subjects with schizophrenia is likely to result in an aberrant control of DA release. Significantly, twofunctional neuroimaging studies have reported anincrease in the release of DA following a standard-ised amphetamine challenge [24,25]. These datafurther support the hypothesis that there is anincrease in DA release in subjects with schizophreniawhich results in an increase in levels of extra-neuronal DA, a finding consistent with some earlystudies using tissue obtained post-mortem [5,6].

While the DA hypothesis of schizophrenia stillrequires total validation, studies using brain tissueobtained at autopsy and PET suggest there arechanges in the functioning of the dopaminergicsystems in subjects with schizophrenia. Most intri-guingly, recent data could indicate that the majorproblem in subjects with schizophrenia is presynap-tic and involves the control of DA release. If proven,this would represent a major shift in the focus ofresearch into abnormalities in dopaminergic functionin schizophrenia.

The dopamine/serotonin hypothesis ofschizophrenia

Changes in the serotonergic systems of the brain inschizophrenia are also supported by neuropharma-cological observations [5]. At the centre of this hypo-thesis is the tenet that certain indoles, such as theserotonin (5HT) receptor agonist lysergic acid (LSD),induce or exacerbate hallucinations. In addition, it is

increasingly accepted that antipsychotic drugs whichantagonise both DA and 5HT receptors have a thera-peutic advantage over drugs that more selectivelytarget DA receptors [26]. These data suggest it isover-active serotonergic neurones that are involved inschizophrenia.

In addition to neuropharmacological evidence,there has been a significant amount of data obtainedfrom studies using tissue obtained post-mortem tosupport a role for changes in the serotonergic systemin the pathology of schizophrenia. An early studyreported a decrease in [3H]-LSD binding to 5HTreceptors in the frontal cortex from subjects withschizophrenia [27]. However, in a follow-up study,[3H]-LSD binding was reported to be increased intissue from subjects who had not received anti-psychotic drugs close to death [28]. These authorssuggested the decrease in 5HT receptors in the earlierstudy was an effect of antipsychotic drug treatmentprior to death.

[3H]-LSD binds to a number of 5HT receptors.Thus, in an attempt to better understand the nature ofthe changes in 5HT receptors in schizophrenia, morerecent studies used radioactive ligands that weremore selective for specific 5HT receptors. Several ofthese studies have now reported a decrease in5HT2A/2C receptors in the frontal cortex of subjectswith schizophrenia [29–32]. Moreover, the changesin the density of 5HT2A/2C receptors were independentof the antipsychotic drug received by the schizo-phrenic subjects close to death [30,31] and may beassociated with decreases in levels of mRNA encod-ing for the 5HT2A receptor [32]. It is now widelyaccepted that the 5HT2A/2C receptor is decreased inregions of the cortex from subjects with schizo-phrenia, but it is remains to be established whetherthis is due to a pathological process.

Positron emission tomography has also been usedto examine the density of 5HT2 receptors in thecortex. These studies did not show a difference in thedensity of cortical 5HT2 receptors in schizophrenia[33,34], results which differed from the majority of those obtained using post-mortem tissue. It is important to note that both PET studies used [18F]-setoperone as the radioligand to measure the 5HT2

receptors. Given that the use of [3H]-LSD in post-mortem tissue led to an inconsistency of results, itcould be that the use of a different radioligand in PETand post-mortem studies has led to inconsistencies in results on 5HT2 receptors in schizophrenia.

An increase in the density of the 5HT1A receptorhas been reported in the prefrontal cortex [32,35–37],

SIGNAL TRANSMISSION IN SCHIZOPHRENIA562

B. DEAN 563

orbital frontal cortex [37,38], temporal cortex [34],posterior cingulate, motor cortex and hippocampus[39] from subjects with schizophrenia. This increasein 5HT1A receptor density does not appear to be asso-ciated with a change in mRNA encoding for thereceptor [32]. Significantly, the 5HT2A/2C and the5HT1A receptor are differentially distributed in the frontal cortex [31,40] (Fig. 1). This has led to thesuggestion that the changes in these two receptors inthe frontal cortex were ‘normalising’ some abnormaldrive in the cortex of subjects with schizophrenia[41]. It would seem most likely that such an abnor-mal drive would involve changes in innervating serotonergic neurones.

There have been two reports that have suggestedthere is a decrease in the density of the serotonintransporter in the prefrontal cortex of subjects withschizophrenia [36,39]. Such differences would beexpected to change 5HT levels in that brain region.Moreover, this change in levels of 5HT could be thecause of the changes in 5HT receptors reported in anumber of studies. It has also been suggested thatthere is a conformational change in the 5HT trans-porter in the hippocampus from subjects with schiz-ophrenia [42,43]. This conformational change is notassociated with any particular polymorphism in the5HT transporter gene promoter region [44], but man-ifests as a change in the affinity of [3H]-paroxetinebinding [43]. Thus, changes in the 5HT transporter in schizophrenia could also be important in the hippocampus as well as in the frontal cortex.

Moreover, in both brain regions, it could be by antag-onising 5HT receptors that atypical drugs counteractthe ‘flow-on’ effects of changes in the serotonintransporter and hence give improved clinical out-comes [26].

The glutamate hypothesis of schizophrenia

The glutamate hypothesis of schizophrenia ismainly based on the observation that drugs such asphencyclidine and SKF 10,047, which block the ionchannel of the N-methyl-D-aspartate (NMDA) recep-tor, precipitate psychoses [45]. Moreover, it has beenproposed that these NMDA receptor ion channelblockers are the only pharmacological agents thatcause the deficit symptoms associated with schizo-phrenia [46]. These observations have led to the pro-posal that deficiencies in glutamate systems providethe best model of the symptoms of schizophrenia.

Studies using tissue obtained post-mortem havesubsequently examined the density of the binding sitefor phencyclidine on the NMDA receptor. One ofthese studies reported an increase in the binding of N-(1-[2-thienyl] cyclohexyl) 3,4-[3H] piperidine([3H]-TCP), a phencyclidine derivative, to orbitalfrontal cortex from subjects with schizophrenia [47].[3H]-TCP binding has also been reported to bedecreased in the Cornu Ammonis (CA3) region of thehippocampus from subjects with schizophrenia [48].Other radioactive drugs that bind to the NMDAreceptor have been used to study these receptors in

Figure 1: The distribution of serotonin2A/2C (A) andserotonin1A (B) receptors inBrodmann’s area 9 of thefrontal cortex. In theseimages, the density ofradioligand binding andhence receptor isproportional to the ‘greyness’of the image: the darker theimage, the more receptor.

brain tissue obtained post-mortem. These studieshave shown an increase in the density of the NMDAreceptors in the putamen [49,50], caudate nucleus[51] and superior temporal cortex [52] from subjectswith schizophrenia. Hence, many studies on theNMDA receptor suggest this receptor is changed inthe brain of subjects with schizophrenia, but thatthese changes are not consistent throughout the brain.

Glutamate has also been shown to have effectsthrough a number of non-NMDA receptors includingkainate and amino-3-hydroxy-5-methyl-4-isoxazolepropionate (AMPA) receptors [53]. The density ofkainate receptors has been reported to be increased inthe frontal [54] and orbital [55] cortex, as well as inthe striatum [56] from subjects with schizophrenia.By contrast, the level of kainate receptors has beenreported to be decreased in the CA4/CA3 regions ofthe hippocampus from subjects with schizophrenia[57]. These data, added to those on the NMDAreceptor, suggest that changes in glutamate functionin the hippocampus from subjects with schizophre-nia differ to those in other brain regions. Finally, ithas been shown that there are no significant changesin AMPA receptors in the cortex of subjects withschizophrenia [58].

The NMDA, kainate and AMPA receptors havebeen shown to be made up of subunits, all of whichare separate gene products [59]. Using data fromcloning studies it has been possible to constructprobes to measure levels of mRNA for the differentsubunits of the glutamate receptors in tissue fromsubjects with schizophrenia. In addition, by translat-ing the genetic codes for the subunits into amino acidsequences it has been possible to produce subunitspecific antibodies that have been used to measurelevels and distribution of receptor subunits in humanbrain tissue. Studies on levels of mRNA showed areduction in the expression of subunits of the NMDAreceptor in the superior temporal cortex [60] as wellas frontal cortex [61] of ‘neuroleptic-free’ subjectswith schizophrenia. By contrast, it has been reportedthat only levels of mRNA for the NR2 subunit familyof the NMDA receptors is increased in the frontalcortex of subjects with schizophrenia [62].

Messenger RNA for the non-NMDA kainate/AMPA receptor was initially reported to be reducedin the CA3 region of the hippocampus from subjectswith schizophrenia [63]. More recently, reducedexpression of a subunit of the kainate receptor (KA2)has been measured in the dentate gyrus, CA1 andCA2 regions of the hippocampus from subjects withschizophrenia [64]. In addition, in contrast to the lack

of change observed in radioligand studies, low levelsof mRNA for the two spliced variants of the GluR2subunit of the AMPA receptor were observed in thehippocampal formation from subjects with schizo-phrenia [65]. These data were supported by a immuno-autoradiographic study which showed a decrease inGluR1 subunit in parahippocampal gyrus and a de-crease in GluR2/3 subunits in the CA4 region of thehippocampus from subjects with schizophrenia [66].

Data from studies using brain tissue obtained post-mortem seem to indicate that there is disruption tothe cortical and hippocampal glutamatergic systemsin subjects with schizophrenia. This hypothesis hasbeen tested in a nuclear magnetic resonance (NMR)study which reported an increase in glutamine levelswithout a change in glutamate levels in the frontalcortex from subjects with schizophrenia [67]. Thisresult was taken as revealing an abnormality in theconversion of glutamine to glutamate by glutamater-gic neurones. This raises the possibility that, as fordopamine and serotonin, aberrant control of thelevels of available neurotransmitter by presynapticneurones maybe involved in the pathology of schizo-phrenia. In addition, a PET study has shown that ketamine, a non-competitive NMDA receptor anta-gonist, increases cerebral blood flow in the anteriorcingulate cortex in subjects with schizophrenia butreduces blood flow in the hippocampus and primaryvisual cortex [68]. This is further evidence that thechanges in glutamate function in the hippocampusdiffer from that in certain cortical regions of subjectswith schizophrenia.

Finally, two studies have reported that glutamateuptake sites are reduced in the striatum from subjectswith schizophrenia [69,70]. Intuitively, a reduction ofglutamate uptake sites would most likely to be asso-ciated with a reduction in glutamate uptake leading toincreased levels of glutamate and hence increasedglutamate activity. However, the reduction in uptakesites could also be an appropriate physiologicalresponse in an attempt to rectify the decrease in glu-tamate synthesis which has been suggested by theNMR finding on cortical glutamate in schizophrenia.

The g-aminobutyric acid/dopaminehypothesis of schizophrenia

Unlike other hypotheses of schizophrenia, the g-aminobutyric acid (GABA) hypothesis is not pri-marily based on drug effects in humans, but on datafrom studies using animals. Such studies suggest thatif the GABA system, which has an inhibitory drive

SIGNAL TRANSMISSION IN SCHIZOPHRENIA564

on dopaminergic systems of the brain, is under-active, this would result in an over-active dopaminer-gic system and hence induce psychoses [71].

Studies using tissue obtained post-mortem haveattempted to confirm a role for a decrease in GABAfunctioning in the pathology of schizophrenia.Initially, the finding that GABA was decreased in thenucleus accumbens and thalamus of subjects withschizophrenia [72] seemed to lend support for adecreased GABA inhibitory drive onto the dopamin-ergic system in schizophrenia. However, in a similarparadigm to that which occurred with findings onDA, the inability to replicate this result led to thestudy of GABA receptors in the brain of subjectswith schizophrenia. The most abundant GABA recep-tor in the brain is the GABAA receptor and it has beenshown that levels of GABAA receptors are inverselyrelated to levels of GABA [73]. Significantly, increasesin the density of the GABAA receptor, as measured by[3H]-muscimol binding, have been reported in thesuperficial layers of the cingulate cortex [74] as wellas in the prefrontal cortex [75,76], the dentata, CA4,CA3 of the hippocampal formation, the subiculumand presubiculum [77] from subjects with schizo-phrenia. Importantly, the increase in the density of theGABAA receptor is thought to be a response to adecrease in activation by GABA rather than a primarydeficit relating to the pathology of schizophrenia.

The GABAA receptor contains a number of bindingsites for various centrally active compounds. [3H]-muscimol is a GABAA receptor agonist that binds tothe GABA binding site of the receptor. Anotherbinding site on the GABAA receptor is the benzodi-azepine binding site, which is where benzodiazepinesexert their anxiolytic effects. The presence of this siteon the GABAA receptor means that radioactive benzo-diazepines, like [3H]-muscimol, can be used tomeasure the density of GABAA receptors in the brain.Such studies have reported an increase in benzo-diazepine binding in the superior temporal gyrus, butnot medial frontal cortex, orbitofrontal cortex, medialand inferior temporal gyri, CA1–3 or putamen fromsubjects with schizophrenia [78]. By contrast, otherstudies have reported that benzodiazepine binding isnot changed in the hippocampal formation [79] orfrontal cortex [80] from subjects with schizophrenia.Finally, one study has reported a decrease in benzo-diazepine binding in the anterior cingulate cortex,hippocampus, somatomotor cortex, cerebellar cortexand globus pallidus from subjects with schizophrenia[81]. These studies, plus the finding that mRNA forthe subunits of the GABAA receptor is not altered in

prefrontal cortex from subjects [82,83], make the casefor an increase in GABAA receptors in the brain ofsubjects with schizophrenia less compelling.

Benzodiazepine binding in schizophrenia has alsobeen studied using single photon emission computedtomography (SPECT). These studies were unable tofind changes in benzodiazepine binding in subjectswith schizophrenia [84–86] and therefore do notsupport a role for changes in GABAA receptors inschizophrenia. A study on the binding of [3H]-muscimol and radioactive benzodiazepine in tissuefrom the same individuals would seem essential tohelp reconcile differences in existing data on GABAA

receptors in schizophrenia.As with other neurotransmitter systems, the uptake

system for GABA has been studied in schizophrenia.These studies have reported a decrease in GABAuptake sites in the hippocampus [87,88], striatum[69] and frontal cortex [89] of subjects with schizo-phrenia. Finally, a decrease in glutamic acid decarbox-ylase (GAD) has been reported in the frontal cortexof subjects with schizophrenia [90]. Glutamic aciddecarboxylase is the rate-limiting enzyme in the syn-thesis of GABA and is present in GABA-producingneurones. These data could indicate either a decreasein GABA-producing neurones in subjects withschizophrenia, or a decrease in GABA uptake siteson an unchanged number of neurones. This is animportant issue that needs to be resolved.

The cholinergic hypothesis ofschizophrenia

The cholinergic hypothesis of schizophrenia hasbeen developed from many lines of evidence [91].These include evidence that the modulation of thecholinergic system can affect both positive and nega-tive symptoms of schizophrenia [91]. In addition, it is becoming apparent that the cholinergic system iscentral to processing of sensory stimuli and in ensuring that the higher functions of the brain operateoptimally [92]. A derangement of these functionscould well result in the symptoms of schizophrenia.In addition, atypical antipsychotic drugs such as cloza-pine and olanzapine, that appear to have improvedtherapeutic outcomes [15,93], bind with high affinityto muscarinic receptors [94]. Finally, anticholinergicagents have proven useful in reducing extrapyramidalside-effects of typical antipsychotic drugs.

Direct evidence to support a role for changes inthe muscarinic system in schizophrenia came fromthe study of a critical enzyme in the synthesis of

B. DEAN 565

acetylcholine. The first evidence to suggest that thecholinergic system was altered in the brains of sub-jects with schizophrenia was the demonstration thatthe activity of choline acetyltransferase (ChAT) wasincreased in the hippocampus, striatum and nucleusaccumbens from subjects with schizophrenia [95].Differences in the activity of ChAT were not detectedin any of the other 47 regions studied. These findingswere challenged by another study that showed thatChAT was decreased in the septal area and caudatehead from subjects with schizophrenia [96]. ChATwas not altered in any of the other 13 regions studied,including the pons. Finally, a study looking at levelsof protein rather than enzyme activity showed thatlevels of ChAT was reduced in the pons of subjectswith schizophrenia [97]. As with other neurotransmit-ter-related enzymes, the differences between thesestudies could be due to problems in the stability ofenzymatic activity post-mortem.

Studies on muscarinic receptors are limited andcontradictory. A radioligand-binding study that mea-sured the summed density of all muscarinic receptorsshowed an increase in receptor density in the orbito-frontal cortex and caudate from subjects treated withantipsychotics up until death [98]. However, using amore selective radioligand, reduction in density ofmuscarinic receptors M1 + M4 has been reported in thecaudate-putamen not related to antipsychotics [99].Moreover, this was not accompanied by a change inthe levels of M1 receptor mRNA in the caudate-putamen [100]. The availability of increasingly spe-cific radioligands for muscarinic receptors, receptorspecific antibodies and measurement of levels ofmRNA for each receptor will make it possible todetermine which receptors are altered in schizo-phrenia. This will allow the precise nature of changesin muscarinic receptors to be determined.

Conclusions

Studies using neuroimaging techniques or tissueobtained post-mortem have revealed extensivechanges in neurochemical markers in differentregions of schizophrenic patients. Moreover, it isclear these changes would affect many neurotrans-mitter systems and could, therefore, precipitatediverse symptoms [101]. It would seem logical thatthe atypical antipsychotics affect a greater number ofthe systems altered in schizophrenia [94].

Interpreting data from post-mortem and neuro-imaging studies is difficult. Growing evidence supports

the hypothesis that changes in functioning of dopa-minergic neurones is central to the genesis of schizo-phrenic psychoses and that changes in serotonergic andglutamatergic neurones are important in the pathol-ogy. Further evidence is required to implicatechanges in GABAergic and cholinergic systems.

Data from both neuroimaging and post-mortemstudies have identified changes in the presynaptic pro-teins and/or function in the brains of schizophrenicpatients: in dopamine release and glutamate syn-thesis, in the dopamine autoreceptor, and the sero-tonin, glutamate and GABA transporters. Changes in functioning of presynaptic processes could be acomponent of the pathology. As molecules in theseprocesses are isolated and their function understood[102], it should be possible to determine if changes ina single protein common to all affected presynapticprocesses is the cause of schizophrenia. Alternatively,it results from changes in different proteins such asthe serotonin, glutamate and GABA transporters.Whichever proves to be the case, if the ‘presynaptichypotheses’ is proven, then current antipsychoticsprobably act by blocking the ‘flow-on’ effects ofabnormal neurotransmitter levels by antagonisingpost-synaptic receptors. It would, therefore, seem thatthe focus of future drugs should change from block-ade of the post-synaptic receptor to modulation ofpresynaptic function, that is, ‘modulate signal trans-mission’ rather than ‘reduce signal reception’.

Acknowledgements

I thank Elizabeth Scarr and Geoffrey Pavey fortheir suggestions, scientists in the Rebecca L. CooperResearch Laboratories for their commitent to ourresearch, and the National Alliance for Research into Schizophrenia and Depression, The Rebecca L.Cooper Medical Research Foundation and the StanleyFoundation for supporting that research.

References

1. Chua SE, McKenna PJ. Schizophrenia – a brain disease? A critical review of structural and functional cerebral abnormality in the disorder. British Journal of Psychiatry1995; 166:563–582.

2. Stahl S. Psychopharmacology of antipsychotics. London:Marin Dunitz, 1999.

3. Carlsson A, Linquist M. Effect of chlorpromazine orhaloperidol on formation of 3-methoxy-tyramine andnormetanephrine in mouse brain. Acta Pharmacologia etToxicologia 1963; 20:140–144.

4. Meltzer HY. Biological studies in schizophrenia.Schizophrenia Bulletin 1987; 13:77–107.

SIGNAL TRANSMISSION IN SCHIZOPHRENIA566

5. Bird ED, Spokes EG, Barnes J, Mackay AVP, Iversen LL,Shepherd M. Increased brain dopamine and reduced glutamic acid decarboxylase and choline acetyl transferaseactivity in schizophrenia and related psychoses. Lancet1977; ii:1157–1159.

6. Reynolds GP. Increased concentrations and lateral asymmetries of amygdala dopamine in schizophrenia.Nature 1983; 305:527–529.

7. Toru M, Nishikawa T, Mataga N, Tashima M. Dopaminemetabolism increases in post-mortem schizophrenic basalganglia. Journal of Neural Transmission 1982; 54:181–191.

8. Beskow J, Gottfries CG, Roos BE, Winblad B.Determination of monoamine and monoamine metabolitesin the human brain: post mortem studies in a group of suicides and in a control group. Acta PsychiatricaScandinavica 1976; 53:7–20.

9. Bachus SE, Kleinman JE. The neuropathology of schizophrenia. Journal of Clinical Psychiatry 1996; 57(Suppl. 11):72–83.

10. Reynolds GP, Reynolds LM, Riederer P, Jellinger K, GabrielE. Dopamine receptors and schizophrenia: drug effect orillness. Lancet 1980; i:1251.

11. Kebabian JW, Calne DB. Multiple receptors for dopamine.Nature 1979; 277:93–96.

12. Sibley DR, Monsma FJ. Molecular biology of dopaminereceptors. Trends in Pharmacological Sciences 1992;13:61–69.

13. Roberts DA, Balderson D, Pickering-Brown SM, Deakin JFW,Owen F. The abundance of mRNA for dopamine D2 receptorisoforms in brain tissue from controls and schizophrenics.Brain Research: Molecular Brain Research 1994; 25:173–175.

14. Van Tol HHM, Bunzow JR, Guan H-C et al. Cloning of thegene for a human dopamine D4 receptor with high affinityfor the antipsychotic clozapine. Nature 1991; 350:610–614.

15. Baldessarini RJ, Frankenburg FR. Clozapine: a novelantipsychotic agent. New England Medical Journal 1991;324:746–754.

16. Meador-Woodruff JH, Haroutunian V, Powchik P, DavidsonM, Davis KL, Watson SJ. Dopamine receptor transcriptexpression in striatum and prefrontal and occipital cortex –focal abnormalities in orbitofrontal cortex in schizophrenia.Archives of General Psychiatry 1997; 54:1089–1095.

17. Dean B, Pavey G, Opeskin K. [3H]raclopride binding to braintissue from subjects with schizophrenia: methodologicalaspects. Neuropharmacology 1997; 36:779–786.

18. Stefanis NC, Bresnick JN, Kerwin RW, Schofield WN,McAllister G. Elevation of D4 dopamine receptor mRNA inpostmortem schizophrenic brain. Brain Research: MolecularBrain Research 1998; 53:112–119.

19. Mulcrone J, Kerwin RW. No difference in the expression ofthe D4 gene in post-mortem frontal cortex from controls andschizophrenics. Neuroscience Letters 1996; 219:163–166.

20. Schmauss C, Haroutunian V, Davis KL, Davidson M.Selective loss of dopamine D3-type receptor mRNA expression in parietal and motor cortices of patients withchronic schizophrenia. Proceedings of the NationalAcademy of Sciences of the United States of America 1993;90:8942–8946.

21. Gilbert DB, Millar J, Cooper SJ. The putative dopamine D3agonist, 7-OH-DPAT, reduces dopamine release in thenucleus accumbens and electrical self-stimulation to theventral tegmentum. Brain Research 1995; 681:1–7.

22. Meller E, Bohmaker K, Goldstein M, Basham DA. Evidencethat striatal synthesis-inhibiting autoreceptors are dopamineD3 receptors. European Journal of Pharmacology 1993;249:R5–R6.

23. Clark D, Hjorth S, Carlsson A. Dopamine receptor agonists:mechanisms underlying autoreceptor selectivity. II:theoretical considerations. Journal of Neural Transmission1985; 62:171–207.

24. Laruelle M, Abi-Dargham A, van Dyck CH et al. Singlephoton emission computerized tomography imaging ofamphetamine-induced dopamine release in drug-free schizophrenic subjects. Proceedings of the NationalAcademy of Sciences of the United States of America 1996;93:9235–9240.

25. Breier A, Su TP, Saunders R, Carson RE et al.Schizophrenia is associated with elevated amphetamine-induced synaptic dopamine concentrations: evidence from anovel positron emission tomography method. Proceedings ofthe National Academy of Sciences of the United States ofAmerica 1997; 94:2569–2574.

26. Huttunen M. The evolution of the serotonin-dopamineantagonist concept. Journal of ClinicalPsychopharmacology 1995; 15 (Suppl. 1):4S–10S.

27. Bennett JP, Enna SJ, Bylund DB, Gillin JC, Wyatt RJ,Snyder SH. Neurotransmitter receptors in frontal cortex ofschizophrenics. Archives of General Psychiatry 1979;36:927–934.

28. Whitaker PM, Crow TJ, Ferrier IN. Tritiated LSD binding infrontal cortex in schizophrenia. Archives of GeneralPsychiatry 1981; 38:278–280.

29. Mita T, Hanada S, Nishino N et al. Decreased serotonin S2and increased dopamine D2 receptors in chronic schizophrenics. Biological Psychiatry 1986; 21:1407–1414.

30. Gurevich EV, Joyce JN. Alterations in the cortical serotonergic system in schizophrenia: a postmortem study.Biological Psychiatry 1997; 42:529–545.

31. Dean B, Hayes W, Hill C, Copolov D. Decreased serotonin2A receptors in Brodmann’s area 9 from schizophrenic subjects. A pathological or pharmacologicalphenomenon? Molecular and Chemical Neuropathology1998; 34:133–145.

32. Burnet PW, Eastwood SL, Harrison PJ. 5-HT1A and 5-HT2A receptor mRNA and binding site densities are differentially altered in schizophrenia.Neuropsychopharmacology 1996; 15:442–455.

33. Lewis R, Kapur S, Jones C et al. Serotonin 5–HT2 receptorsin schizophrenia: a PET study using [18F]setoperone in neuroleptic-naive patients and normal subjects. AmericanJournal of Psychiatry 1999; 156:72–78.

34. Trichard C, Paillere-Martinot ML, Attar-Levy D, Blin J,Feline A, Martinot JL. No serotonin 5-HT2A receptordensity abnormality in the cortex of schizophrenic patientsstudied with PET. Schizophrenia Research 1998; 31:13–17.

35. Hashimoto T, Nishino N, Nakai H, Tanaka C. Increase inserotonin 5-HT1A receptors in prefrontal and temporal cortices of brains from patients with chronic schizophrenia.Life Sciences 1991; 48:355–363.

36. Laruelle M, Abi-Dargham A, Casanova MF, Toti R,Weinberger DR, Kleinman JE. Selective abnormalities of pre-frontal serotonergic receptors in schizophrenia: a postmortemstudy. Archives of General Psychiatry 1993; 50:810–818.

37. Sumiyoshi T, Stockmeier CA, Overholser JC, Dilley GE,Meltzer HY. Serotonin1A receptors are increased in post-mortem prefrontal cortex in schizophrenia. Brain Research1996; 708:209–214.

38. Simpson MDC, Lubman DI, Slater P, Deakin JFW.Autoradiography with [3H]8-OH-DPAT reveals increases in5-HT1A receptors in ventral prefrontal cortex in schizophrenia. Biological Psychiatry 1996; 39:919–928.

B. DEAN 567

39. Joyce JN, Shane A, Lexow N, Winokur A, Casanova MF,Kleinman JE. Serotonin uptake sites and serotonin receptorsare altered in the limbic system of schizophrenics.Neuropsychopharmacology 1993; 8:315–336.

40. Dean B, Tomaskovic-Crook E, Opeskin K, Keks N, CopolovD. No change in the density of the serotonin1A receptor, the serotonin4 receptor or the serotonin transporterin the dorsolateral prefrontal cortex from subjects withschizophrenia. Neurochemistry International 1999;34:109–115.

41. Hashimoto T, Kitamura N, Kajimoto Y et al. Differentialchanges in serotonin 5-HT1A and 5–HT2 receptor bindingin patients with chronic schizophrenia. Psychopharmacology1993; 112:S35–S39.

42. Dean B, Opeskin K, Pavey G et al. [3H]paroxetine bindingis altered in the hippocampus but not the frontal cortex orcaudate nucleus from subjects with schizophrenia. Journalof Neurochemistry 1995; 64:1197–1202.

43. Naylor L, Dean B, Opeskin K et al. Changes in the serotonin transporter in the hippocampus of subjects withschizophrenia identified using [3H]paroxetine. Journal ofNeural Transmission 1996; 103:749–757.

44. Naylor L, Dean B, Pereira A, Mackinnon A, Kouzmenko A,Copolov D. No association between the serotonin transporter-linked promoter region polymorphism and eitherschizophrenia or density of the serotonin transporter inhuman hippocampus. Molecular Medicine 1998; 4:671–674.

45. Deutsch SI, Mastropaolo J, Schwartz BL, Rosse RB,Morihisa JM. A ‘glutamate hypothesis’ of schizophrenia.Clinical Neuropharmacology 1989; 12:1–13.

46. Javitt DC. Negative symptoms and the PCP (phencyclidine)model of schizophrenia. Hillside Journal of ClinicalPsychiatry 1987; 9:12–35.

47. Simpson MDC, Slater P, Royston MC, Deakin JFW.Alterations in phencyclidine and sigma binding sites inschizophrenic brain. Schizophrenia Research 1992; 6:41–48.

48. Dean B, Scarr E, Bradbury R, Copolov D. Decreased hippocampal (CA3) NMDA receptors in schizophrenia.Synapse 1999; 32:67–69.

49. Aparicio-Legarza MI, Davis B, Hutson PH, Reynolds GP.Increased density of glutamate/N-methyl-D-aspartate receptors in putamen from schizophrenic patients.Neuroscience Letters 1998; 241:143–146.

50. Kornhuber J, Mack-Burkhardt F, Riederer P et al.[3H]MK-801 binding sites in postmortem brain regions of schizophrenic patients. Journal of Neural Transmission1989; 77:231–236.

51. Noga JT, Hyde TM, Herman MM et al. Glutamate receptorsin the postmortem striatum of schizophrenic, suicide, andcontrol brains. Synapse 1997; 27:168–176.

52. Grimwood S, Slater P, Deakin JF, Hutson PH. NR2B-containing NMDA receptors are up-regulated in temporalcortex in schizophrenia. Neuroreport 1999; 10:461–465.

53. Collingridge GL, Lester RAJ. Excitatory amino acid receptors in the vertebrate central nervous system.Pharmacological Reviews 1989; 40:145–210.

54. Nishikawa T, Takashima M, Toru M. Increased [3H]kainicacid binding in the prefrontal cortex in schizophrenia.Neuroscience Letters 1983; 40:245–250.

55. Deakin JFW, Slater P, Simpson MDC et al. Frontal corticaland left temporal glutamatergic dysfunction in schizophrenia.Journal of Neurochemistry 1989; 52:1781–1786.

56. Toru M, Watanabe S, Shibuya H et al. Neurotransmitter,receptors and neuropeptides in post-mortem brains ofchronic schizophrenic patients. Acta PsychiatricaScandinavica 1988; 78:121–137.

57. Kerwin RW, Patel S, Meldrum B. Quantitative autoradiographic analysis of glutamate binding sites in thehippocampal formation in normal and schizophrenic brainpost mortem. Neuroscience 1990; 39:25–32.

58. Kurumaji A, Ishimaru M, Toru M. α-[3H]amino-3-hydroxy-5-methylisoxazole-4-propionic acid binding tohuman cerebral cortical membranes: minimal changes inpostmortem brains of chronic schizophrenics. Journal ofNeurochemistry 1992; 59:829–837.

59. Seeburg PH. The TINS/TiPS lecture: The molecular biologyof mammalian glutamate receptor channels. Trends inPharmacological Sciences 1993; 16:359–365.

60. Humphries C, Mortimer A, Hirsch S, De Belleroche J.NMDA receptor mRNA correlation with antemortem cognitive impairment in schizophrenia. Neuroreport 1996;7:2051–2055.

61. Sokolov BP. Expression of NMDAR1, GluR1, GluR7, andKA1 glutamate receptor mRNAs is decreased in frontalcortex of ‘neuroleptic-free’ schizophrenics: evidence onreversible up-regulation by typical neuroleptics. Journal ofNeurochemistry 1998; 71:2454–2464.

62. Akbarian S, Sucher NJ, Bradley D et al. Selective alterations in gene expression for NMDA receptor subunitsin prefrontal cortex of schizophrenics. Journal ofNeuroscience 1996; 16:19–30.

63. Harrison PJ, McLaughlin D, Kerwin RW. Decreased hippocampal expression of a glutamate receptor gene inschizophrenia. Lancet 1991; 337:450–452.

64. Porter RHP, Eastwood SL, Harrison PJ. Distribution ofkainate receptor subunit mRNAs in human hippocampus,neocortex and cerebellum, and bilateral reduction of hippocampal GluR6 and KA2 transcripts in schizophrenia.Brain Research 1997; 751:217–231.

65. Eastwood SL, Burnet PW, Harrison PJ. GluR2 glutamatereceptor subunit flip and flop isoforms are decreased in thehippocampal formation in schizophrenia: a reverse transcriptase-polymerase chain reaction (RT-PCR) study.Brain Research: Molecular Brain Research 1997; 44:92–98.

66. Eastwood SL, Kerwin RW, Harrison PJ. Immunoautoradio-graphic evidence for a loss of alpha-amino-3-hydroxy-5-methyl-4-isoxazole propionate-preferring non-N-methyl-D-aspartate glutamate receptors within the medial temporallobe in schizophrenia. Biological Psychiatry 1997;41:636–643.

67. Bartha R, Williamson PC, Drost DJ et al. Measurement ofglutamate and glutamine in the medial prefrontal cortex ofnever-treated schizophrenic patients and healthy controls byproton magnetic resonance spectroscopy. Archives ofGeneral Psychiatry 1997; 54:959–965.

68. Lahti AC, Holcomb HH, Medoff DR, Tamminga CA.Ketamine activates psychosis and alters limbic blood flow inschizophrenia. Neuroreport 1995; 6:869–872.

69. Simpson MDC, Slater P, Royston MC, Deakin JFW.Regionally selective deficits in uptake sites for glutamateand gamma-aminobutyric acid in the basal ganglia of schizophrenics. Psychiatry Research 1992; 42:273–282.

70. Aparicio-Legarza MI, Cutts AJ, Davis B, Reynolds GP.Deficits of [3H]D-aspartate binding to glutamate uptake sitesin striatal and accumbens tissue in patients with schizophrenia. Neuroscience Letters 1997; 232:13–16.

71. van Kammen DP. γ-Aminobutyric acid (GABA) and thedopamine hypothesis of schizophrenia. American Journal ofPsychiatry 1977; 134:138–143.

72. Perry TL, Buchanan J, Kish SJ, Hansen S. γ-aminobutyric-acid deficiency in brain of schizophrenic patients. Lancet1979; i:237–239.

SIGNAL TRANSMISSION IN SCHIZOPHRENIA568

B. DEAN 569

73. Mhatre MC, Ticku MK. Chronic GABA treatment downregulates the GABAA receptor alpha 2 and alpha 3subunit mRNAS as well as polypeptide expression inprimary cultured cerebral cortical neurons. Brain ResearchMolecular Brain Research 1994; 24:159–165.

74. Benes FM, Vincent SL, Alsterberg G, Bird ED, SanGiovanniJP. Increased GABAA receptor binding in superficial layersof cingulate cortex in schizophrenia. Journal ofNeuroscience 1992; 12:924–929.

75. Benes FM, Vincent SL, Marie A, Khan Y. Up-regulation ofGABAA receptor binding on neurons of the prefrontal cortexin schizophrenic subjects. Neuroscience 1996;75:1021–1031.

76. Dean B, Hussain T, Hayes W et al. Changes in serotonin2Aand GABA (A) receptors in schizophrenia: studies on thehuman dorsolateral prefrontal cortex. Journal ofNeurochemistry 1999; 72:1593–1599.

77. Benes FM, Khan Y, Vincent SL, Wickramasinghe R.Differences in the subregional and cellular distribution ofGABAA receptor binding in the hippocampal formation ofschizophrenic brain. Synapse 1996; 22:338–349.

78. Kiuchi Y, Kobayashi T, Takeuchi J, Shimizu H, Ogata H,Toru M. Benzodiazepine receptors increase in post-mortembrain of chronic schizophrenics. European Archives ofPsychiatry and Neurological Science 1989; 239:71–78.

79. Reynolds GP, Stroud D. Hippocampal benzodiazepinereceptors in schizophrenia. Journal of Neural Transmission(General Section) 1993; 93:151–155.

80. Pandey GN, Conley RR, Pandey SC et al. Benzodiazepinereceptors in the post-mortem brain of suicide victims andschizophrenic subjects. Psychiatry Research 1997; 71:137–149.

81. Squires RF, Lajtha A, Saederup E, Palkovits M. Reduced[3H]flunitrazepam binding in cingulate cortex and hippocampus of postmortem schizophrenic brains: is selective glutamatergic neurons associated with major psychosis? Neurochemical Research 1993; 18:219–223.

82. Akbarian S, Huntsman MM, Kim JJ et al. GABAA receptorsubunit gene expression in human prefrontal cortex.Comparison of schizophrenics and controls. Cerebral Cortex1995; 5:550–560.

83. Huntsman MM, Tran BV, Potkin SG, Bunney Jr WE, JonesEG. Altered ratios of alternatively spliced long and shortgamma2 subunit mRNAs of the gamma-amino butyrate typeA receptor in prefrontal cortex of schizophrenics.Proceedings of the National Academy of Sciences of theUnited States of America. 1998; 95:15066–15071.

84. Busatto GF, Pilowsky LS, Costa DC et al. Correlationbetween reduced in vivo benzodiazepine receptor bindingand severity of psychotic symptoms in schizophrenia.American Journal of Psychiatry 1997; 154:56–63.

85. Ball S, Busatto GF, David AS et al. Cognitive functioningand GABAA/benzodiazepine receptor binding in schizophrenia: a 123I-Iomazenil SPET study. BiologicalPsychiatry 1998; 43:107–117.

86. Schroder J, Bubeck B, Demish S, Sauer H. Benzodiazepinereceptor distribution and diazepam binding in schizophrenia:an exploratory study. Psychiatry Research 1997; 68:125–131.

87. Simpson MDC, Slater P, Deakin JFW, Royston MC,Skan WJ. Reduced GABA uptake sites in the temporal lobe in schizophrenia. Neuroscience Letters 1989;107:211–215.

88. Reynolds GP, Czudek C, Andrews HB. Deficit and hemispheric asymmetry of GABA uptake sites in hippocampus in schizophrenia. Biological Psychiatry 1990;27:1038–1044.

89. Woo T-U, Miller JL, Lewis DA. Schizophrenia and the parvalbumin-containing class of cortical local circuitneurons. American Journal of Psychiatry 1997;154:1013–1015.

90. Akbarian S, Kim JJ, Potkin SG et al. Gene expression forglutamic acid decarboxylase is reduced without loss ofneurons in prefrontal cortex of schizophrenics. Archives ofGeneral Psychiatry 1995; 52:258–266.

91. Tandon R. Cholinergic aspects of schizophrenia. BritishJournal of Psychiatry (Suppl.) 1999; 37:7–11.

92. Sarter M, Bruno JP. Abnormal regulation of corticopetalcholinergic neurons and impaired information processing inneuropsychiatric disorders. Trends in Neuroscience 1999;22:67–74.

93. Tollefson GD, Beasley Jr CM, Tran PV et al. Olanzapineversus haloperidol in the treatment of schizophrenia andschizoaffective and schizophreniform disorders: results of aninternational collaborative trial. American Journal ofPsychiatry 1997; 154:457–465.

94. Bymaster FP, Calligaro DO, Falcone JF et al. Radioreceptorbinding profile of the atypical antipsychotic olanzapine.Neuropsychopharmacology 1996; 14:87–96.

95. McGeer PL, McGeer EG. Possible changes in striatal andlimbic cholinergic systems in schizophrenia. Archives ofGeneral Psychiatry 1977; 34:1319–1323.

96. Domino EF, Krause RR, Bowers J. Various enzymesinvolved with putative neurotransmitters. Archives ofGeneral Psychiatry 1973; 29:195–201.

97. Karson GN, Casanova MF, Kleinman JE, Griffin WST.Choline acetyltransferase in schizophrenia. AmericanJournal of Psychiatry 1993; 150:454–459.

98. Watanabe S, Nishikawa T, Takashima M, Toru M. Increasedmuscarinic cholinergic receptors in prefrontal cortices ofmedicated schizophrenics. Life Sciences 1983;33:2187–2196.

99. Dean B, Crook JM, Opeskin K, Hill C, Keks N, CopolovDL. The density of muscarinic M1 receptors is decreased inthe caudate- putamen of subjects with schizophrenia.Molecular Psychiatry 1996; 1:54–58.

100. Dean B, Crook JM, Pavey G, Opeskin K, Copolov DL.Studies on M1 and M2 muscarinic receptor mRNA in thehuman caudate-putamen: no change in M1 receptor mRNA in schizophrenia. Molecular Psychiatry 2000; 5:203–207.

101. American Psychiatric Association. Diagnostic and statisticalmanual of mental disorders. 4th edn. Washington: AmericanPsychiatric Association, 1994.

102. Marsh M, McMahon HT. The structural era of endocytosis.Science 1999; 285:215–220.