SUMMARY

1. The pathological process that precipitates schizophreniahas yet to be identified. However, many lines of evidence sug-gest that a change in the functioning of the frontal cortex is animportant abnormality that underlies schizophrenia.

2. Studies in Brodmann’s area 9, obtained post-mortem, haveshown changes in 5-hydroxytryptamine 5-HT2A, muscarinic M1 and GABAA receptors in tissue from subjects with schizo-phrenia.

3. Animal studies suggest a site in the cortex where therewould be an interaction between serotonergic and cholinergicinnervation and that this interaction would involve the 5-HT2A

and the M1 receptor. This site, in turn, would be a potent modu-lator of GABA activity and, hence, levels of GABAA receptors.

4. From combining these data, a theoretical site is proposedthat, if proven to exist in human cortex, is likely to be centralto the pathology of that illness.

Key words: acetylcholine, frontal cortex, GABA, 5-hydroxy-tryptamine, post-mortem, schizophrenia, serotonin.

THE INVOLVEMENT OF THE PREFRONTALCORTEX IN SCHIZOPHRENIA

Studies on cognitive function1 and those using functional neuro-imaging2 or neuroanatomy3 strongly suggest that abnormalities inthe frontal cortex have a role in the pathology of schizophrenia.Despite evidence from such studies, the locus of change that pre-cipitates the symptoms of schizophrenia has yet to be identified. Thishas increasingly led to the proposal that the pathology of schizo-phrenia involves changes in the molecular, rather than cellular, archi-tecture of the brain.4 To test this hypothesis, many studies haveattempted to identify abnormalities in the molecular organization

of the frontal cortex in tissue obtained post-mortem from subjectswith schizophrenia. However, a unifying hypothesis accounting fordata on the changes in the molecular cytoarchitecture of the frontalcortex in subjects with schizophrenia has yet to be formulated.

CHANGES IN NEUROCHEMICAL MARKERSIN THE FRONTAL CORTEX OF SUBJECTS

WITH SCHIZOPHRENIA

Many early studies using tissue obtained post-mortem involved themeasurement of neurochemical markers of the serotonergic systemin the frontal cortex from subjects with schizophrenia. This approachwas driven by observations that drugs acting as 5-hydroxytryptamine(5-HT) receptor agonists, such as lysergic acid (LSD), could induceschizophrenic-like symptoms or worsen existing symptoms inschizophrenic subjects.4 More recently, the focus on the serotonergicsystem in schizophrenia has been reinforced by the demonstrationthat drugs that have combined dopaminergic and serotonergic recep-tor antagonism appear to have improved clinical outcomes whenused to treat the illness.5 Together, the findings on the effects of drugsthat modulate serotonergic activity in the brain suggest at that leastsome of the symptoms of schizophrenia are caused by a hyper-serotonergic state.5

Studies on brain tissue obtained post-mortem have also producedevidence to support the argument that there is an altered serotonergicstate in schizophrenia. The most consistent finding supporting thisargument is that of a decreased density of the 5-HT2A receptor inBrodmann’s area 9 (BA 9) from subjects with schizophrenia.6–11

Only one study using BA 9 has failed to show a decrease in 5-HT2A

receptors in schizophrenia12 and this study was the only study touse an iodinated ligand ([125I]-LSD) to label the 5-HT2A receptor.This raises the possibility that the differing results from the latterstudy were due to methodological, rather than pathological,considerations.

The decrease in 5-HT2A receptors in BA 9 from subjects withschizophrenia does not appear to be part of a generalized change inserotonergic markers in this brain region. For example, studies onthe 5-HT2A receptor that have been limited to BA 9 have failed toshow a change in the density of this receptor in schizophrenia,12,13

but have found changes in neighbouring cortical regions. The 5-HT4

receptor and the 5-HT transporter are also reported not to be alteredin BA 9 from subjects with schizophrenia.13 In addition, the great-est downregulation in the 5-HT2A receptor in BA 9 from subjects

Proceedings of the Australian Neuroscience Society Symposium: Schizophrenia

A PREDICTED CORTICALSEROTONERGIC/CHOLINERGIC/GABAERGIC INTERFACE AS A

SITE OF PATHOLOGY IN SCHIZOPHRENIA

Brian Dean

The Rebecca L Cooper Research Laboratories, The Mental Health Research Institute of Victoria and The University of Melbourne Department of Psychiatry, Parkville, Victoria Australia

Correspondence: Dr Brian Dean, NARSAD Research Fellow, TheRebecca L Cooper Research Laboratories, The Mental Health ResearchInstitute of Victoria, Locked Bag 11, Parkville, Victoria 3052, Australia.Email: B.Dean@papyrus.mhri.edu.au

Presented at the Australian Neuroscience Society Symposium onSchizophrenia: The Search for a Neurobiology, Melbourne, January/February2000. The papers in these proceedings have been peer reviewed.

Received 16 June 2000; revision 28 June 2000; accepted 21 July 2000.

Clinical and Experimental Pharmacology and Physiology (2001) 28, 74–78

Frontal cortical pathology of schizophrenia 75

with schizophrenia is seen in subjects treated with haloperidol,whereas in rats only drugs with a high affinity for the 5-HT2A receptordecrease the levels of that receptor in the cortex.14This finding doesnot exclude the possibility that changes in the 5-HT2A receptor inBA 9 from subjects with schizophrenia are due to the effects of drugtreatment during life, but does show that data from rat studies onantipsychotic drugs cannot be readily transferred to humans. Thus,while currrent data do not exclude changes in the 5-HT2A receptorin BA 9 from subjects with schizophrenia being an effect of drugtreatment before death, the selectivity of the change favours it being,at least in part, due to the pathology of the illness.

One of the problems in the use of human brain tissue obtainedpost-mortem is to determine whether changes in radioligand bindingrepresent changes that would have functional consequences. Oneapproach currently being used to assess functional connectivity inthe human brain involves the use of positron emission tomography(PET) to test for the presence of theoretical connections suggestedby animal studies. In this approach, a drug is given that binds to aknown site in a system that should have a ‘flow-on’ effect on the‘downstream’marker in a theoretically interconnected pathway (i.e.a specific receptor).15 Some time after this pharmacological manipu-lation, PETis used to determine whether the remote site that is pro-posed to be affected by the applied pharmacological manipulationhas changed as predicted to confirm the presence of the functionalpathway. Unfortunately, such an approach is not possible using tissueobtained post-mortem. However, it is possible to use animal studiesto predict flow-on effects that should occur from changes in recep-tors measured in brain tissue from subjects with schizophrenia. Thepresence of the predicted changes can then be tested in brain tissuefrom subjects in whom the original receptor change was measured;this is a similar, but less direct, approach to that being used in PETstudies.

Having shown a decrease in 5-HT2A receptor density in the frontalcortex suggestive of altered serotonergic function, we examinedavailable animal data to determine the outcome of altered sero-tonergic systems in the frontal cortex. One consistent finding fromanimal studies and a study using human frontal cortex obtained atbiopsy, is that changes in serotonergic activity have marked effectson acetylcholine (ACh) release.16–18Moreover, the 5-HT2A receptorhas been shown to be one of the most important receptors in modu-lating the effects of 5-HTon ACh release.16,17 Thus, a functionaloutcome from a change in 5-HT2A receptors in the frontal cortex ofsubjects with schizophrenia should be a change in ACh release and,hence, the prevailing levels of ACh.

Post-mortem delay in collecting human tissue and the instabilityof neurotransmitters post-mortem19 make it difficult to accuratelymeasure levels of ACh in frontal cortex collected post-mortem. Incontrast, it is possible to accurately measure ACh followingmanipulation of the animal central nervous system. Such studieshave consistently shown that an effect of increasing levels of AChis to downregulate the muscarinic M1 receptor.20–22 Thus, if therewas a change in prevailing levels of ACh in the frontal cortex ofsubjects with schizophrenia, this should be accompanied by a changein M1 receptors, a component of brain tissue that is stable post-mortem.

To obtain data to support our hypothesis that changes in ACh lev-els have resulted from changes in serotonergic activity, we measuredthe density of [3H]-pirenzepine binding to M1/4 receptors in BA9from schizophrenic and control subjects,23 a brain region where

M1 receptors are much more abundant than M4 receptors.24–26

[3H]-Pirenzepine binding was significantly decreased in BA9 fromsubjects with schizophrenia (Fig.1). Given the magnitude of thedecrease in [3H]-pirenzepine binding in BA9 (where numbers ofreceptors are M1 »M4) from subjects with schizophrenia, it wouldseem extremely unlikely such a change could occur without somedecrease in M1 receptors. Therefore, these data support the hypoth-esis that there is an increase in the prevailing levels of ACh in BA 9 from subjects with schizophrenia.

In our original hypothesis, it was proposed that changes in sero-tonergic modulation of the cholinergic system in schizophreniawould involve a strong effect through the 5-HT2A receptor. Theincreased activation of the 5-HT2A receptor due to increased levelsof 5-HT resulted in that receptor being downregulated in BA9.However, for a ‘flow-on’effect to then cause changes in the cholin-ergic system, it would have to be postulated that this receptor down-regulation had not fully compensated for the increased activationby 5-HT. Therefore, it could be argued that the increased drive from5-HT on ACh release, which would be related to the levels of M1/4

receptors, would also be related to the level of downregulation ofthe 5-HT2A receptor. If this was a very strong effect, it would beexpected that the densities of 5-HT2A and M1/4 receptors, the twomarkers measured in this pathway, would be related. However, com-bining our data on 5-HT2A and M1 receptors from 20 schizophrenicand 20 control subjects, there were no strong correlation betweenthese two receptors in BA9 (Fig.2). Thus, these data do not supportthe hypothesis that there is a strong interaction between the 5-HT2A

receptor and M1 receptor in BA9, an argument central to thehypotheses being tested. It is of interest that the correlation betweenthe two receptors appears somewhat stronger in subjects withschizophrenia, suggestive of a weak interactivity. Importantly, thiswould still not support our original hypothesis that required a stronginteraction between the two receptors.

Further studies of neurochemical markers in BA9 have shownthat the density of the enzymes protein kinase C, adenylate cyclase27

and nitric oxide synthase11 are not altered in BA9 from subjects withschizophrenia. In addition, dopamine D1-like, D2-like and N-methyl-D-aspartate (NMDA) receptors and the dopamine transporter are not altered in that brain region from schizophrenic subjects who

Fig. 1 The binding (mean6SEM) of [3H]-pirenzepine in BA9 obtainedpost-mortem from schizophrenic () and control (h) subjects. ETE, esti-mated tissue equivalents. Reproduced with permission from Crook.23

76 B Dean

have a decreased density of 5-HT2A receptors.11 However, in the same subjects, there was a small but significant increase in GABAAreceptors, confirming a previous study on GABAA receptors in theprefrontal cortex.28 Importantly, current data do not suggest that largedifferences in mean density of receptors are common in BA9 inschizophrenia and that a unifying hypothesis must accommodate thechanges in 5-HT2A receptors, M1 receptors and GABAA receptorsin BA 9 from subjects with schizophrenia.

A UNIFYING SEROT ONERGIC/CHOLINERGICHYPOTHESIS OF SCHIZOPHRENIA

To formulate a hypothesis that could accommodate findings in BA 9 from subjects with schizophrenia, it is once again necessaryto examine data from animal studies. It has been shown that the5-HT2A receptor is located presynaptically on neurons innervatingthe cortex.29,30 Available data would also suggest that, when the5-HT2A receptor is located presynaptically, it is present on inner-vating cholinergic neurons in the cortex.31 This makes it possiblethat the 5-HT2A receptor could affect the levels of the M1 receptorby a number of presynaptic mechanisms, explaining why there is

not a simple relationship between the changes in 5-HT2A and M1

receptor numbers in human BA9. As an example of such a mechan-ism, electron microscopic examination of the rat cortex and otherbrain regions suggests the majority of 5-HT2A receptors are locatedin the cytosol.29These and other data have led to the suggestion thatthe 5-HT2A receptor may have a role as a retrograde messenger thatregulates gene expression; one such gene could be that for the M1

receptor.When located post-synaptically in the cortex, the 5-HT2A

receptor appears to be on GABAinterneurons.31 Thus, changes in post-synaptic 5-HT2A receptors could also result in changes inGABAergic function, which has been reported in subjects withschizophrenia.11,28At this point, it is not known whether the 5-HT2A

receptors lost in BA9 from subjects with schizophrenia are locatedpre-, post- or pre- and post-synaptically. This is an important delin-eation that must be made in future studies on this receptor inschizophrenia.

Like 5-HT2A receptors, M1 receptors have been shown to belocated pre- and post-synaptically in the cortex.32,33Moreover, pre-synaptically, the M1 receptor has been shown to modulate AChrelease34 and may, therefore, be the ACh autoreceptor in the cortex.This also adds to the complexity of the system because the 5-HT2A

receptor could influence M1 receptors directly on presynaptic inner-vating neurons and/or post-synaptically by controlling ACh release.Once again, further study will be needed to determine whether pre-or post-synaptic receptors are lost in schizophrenia.

As with the 5-HT2A receptor, the M1 receptor is important inmodulating GABArelease.35,36This provides a second mechanismby which GABAergic function could be altered in the frontal cortexof subjects with schizophrenia.11,28

In conclusion, data from studies on BA9 suggest that the pathol-ogy of schizophrenia involves a disruption of serotonergic/cholin-ergic/GABAergic interactions in the frontal cortex. At this point, thisdisruption seems most likely to be modulated through the 5-HT2A

receptor and would involve complex processes involving the modu-lation of ACh release. However, animal data predict that this com-plex process occurs at a single interactive site on innervatingcholinergic neurons. If this site is present in the human cortex, itwould seem likely that it would be involved in the pathology ofschizophrenia (Fig.3). The model predicts there would be an aber-rant cholinergic input into the cortex caused by a hyperserotonergicdrive onto the innervating cholinergic neuron. This abnormal drivewould be via the 5-HT2A receptor, results in changes in the M1 recep-tor and causes abnormalities in cholinergic activity that would thenchange cortical GABAergic activity. An added component of thishypothesis is that 5-HT2A receptor regulation of ACh release hasbeen shown to be modulated by dopaminergic activity,37 making thesite a prime target for all antipsychotic drugs. However, because oftheir affinities for the 5-HT2A and M1 receptors,38 atypical anti-psychotic drugs would be most potent at regulating the activity atthis site.

The challenge now is to confirm the presence of the hypothesizedinterface between the serotonergic/cholinergic/GABAergic systemsin human BA9. Lesion studies in animals will need to be performedto provide information on the synaptic location of the 5-HT2A andM1 receptors at this important site. If the presence of this proposedsite in the human cortex is confirmed, it will become a locus ofresearch directed at understanding cortical abnormalities inschizophrenia.

Fig. 2 The relationship between the binding of [3H]-ketanserin and [3H]-pirenzepine to 5-HT2A and M1 receptors in BA9 obtained post-mortem from(a) schizophrenic and (b) control subjects. For (a): r 5 0.34, P5 0.12; for(b), r 5 0.16, P5 0.49. ETE, estimated tissue equivalents.

Frontal cortical pathology of schizophrenia 77

ACKNOWLEDGEMENTS

BD is the holder of a National Alliance for Research on Schizophreniaand Depression (NARSAD) Young Investigator award. The hypo-theses proposed are a culmination of work performed by many pres-ent and past members of the Rebecca LCooper Research Laboratoriesand their efforts are greatly appreciated. Special thanks to DrElizabeth Scarr and Mr Geoffrey Pavey (Rebecca LCooper ResearchLaboratories) for their editorial assistance.

REFERENCES

1. Pantelis C, Barnes TR, Nelson HE et al. Frontal–striatal cognitivedeficits in patients with chronic schizophrenia. Brain 1997; 120:1823–43.

2. Kotrla KJ, Weinberger DR. Brain imaging in schizophrenia. Annu. Rev.Med.1995; 46: 113–22.

3. Goldman-Rakic PS, Selemon LD. Functional and anatomical aspectsof prefrontal pathology in schizophrenia. Schizophr. Bull. 1997; 23:437–58.

4. Meltzer HY. Biological studies in schizophrenia. Schizophr. Bull.1987;13: 77–107.

5. Huttunen M. The evolution of the serotonin–dopamine antagonistconcept. J. Clin. Psychopharmacol.1995; 15 (Suppl. 1): S4–10.

6. Bennett JP, Enna SJ, Bylund DB, Gillin JC, Wyatt RJ, Snyder SH.Neurotransmitter receptors in frontal cortex of schizophrenics. Arch.Gen. Psychiatry 1979; 36: 927–34.

7. Mita T, Hanada S, Nishino N et al.Decreased serotonin S2 and increaseddopamine D2 receptors in chronic schizophrenics. Biol. Psychiatry1986;21: 1407–14.

8. Hashimoto T, Kitamura N, Kajimoto Y et al. Differential changes inserotonin 5-HT1A and 5-HT2 receptor binding in patients with chronicschizophrenia. Psychopharmacology 1993; 112: S35–9.

9. Laruelle M, Abi-Dargham A, Casanova MF, Toti R, Weinberger DR,Kleinman JE. Selective abnormalities of prefrontal serotonergic recep-tors in schizophrenia: A postmortem study. Arch. Gen. Psychiatry 1993;50: 810–18.

10. Dean B, Hayes W. Decreased frontal cortical serotonin2A receptors insubjects with schizophrenia. Schizophr. Res.1996; 21: 133–9.

11. Dean B, Hussain T, Hayes W et al. Changes in serotonin2A andGABA(A) receptors in schizophrenia: Studies on the human dorsolateralprefrontal cortex. J. Neurochem.1999; 72: 1593–9.

12. Joyce JN, Shane A, Lexow N, Winokur A, Casanova MF, Kleinman JE.Serotonin uptake sites and serotonin receptors are altered in the limbic system of schizophrenics. Neuropsychopharmacology1993; 8:315–36.

13. Dean B, Tomaskovic-Crook E, Opeskin K, Keks N, Copolov D. No change in the density of the serotonin1A receptor, the serotonin4receptor or the serotonin transporter in the dorsolateral prefrontal cortex from subjects with schizophrenia. Neurochem. Int.1999; 34:109–15.

14. Dean B, Hayes W, Hill C, Copolov D. Decreased serotonin2A receptorsin Brodmann’s area 9 from schizophrenic subjects. A pathological orpharmacological phenomenon? Mol. Chem. Neuropathol. 1998; 34:133–45.

15. Smith GS, Schloesser R, Brodie JD et al. Glutamate modulation of dopamine measured in vivo with positron emission tomography (PET) and 11C-raclopride in normal human subjects. Neuropsycho-pharmacology1998; 18: 18–25.

16. Feuerstein TJ, Seeger W. Modulation of acetylcholine release in humancortical slices: Possible implications for Alzheimer’s disease.Pharmacol. Ther. 1997; 74: 333–47.

17. Hirano H, Day J, Fibiger HC. Serotonergic regulation of acetylcholinerelease in rat frontal cortex. J. Neurochem.1995; 65: 1139–45.

18. Cassel JC, Jeltsch H. Serotonergic modulation of cholinergic functionin the central nervous system: Cognitive implications. Neuroscience1995; 69: 1–41.

Fig. 3 A proposed model to unify current findings on 5-HT2A receptors, M1 receptors and GABAA receptors in BA9 from subjects with schizophrenia.This model predicts that increased serotonergic activity in the frontal cortex of subjects with schizophrenia will cause changes in the activity of both cholin-ergic and GABAergic neurons. These complex changes in multineurotransmitter systems are reflected by the decreases in 5-HT2A and M1/4 receptors, aswell as increases in GABAA receptors, in the frontal cortex of subjects with schizophrenia. ACh, acetylcholine; DA, dopamine; 5-HT, 5-hydroxytryptamine.

78 B Dean

19. Yoshimoto K, Irizawa Y, Komura S. Rapid postmortem changes of ratstriatum dopamine, serotonin, and their metabolites as monitored bybrain microdialysis. Forensic Sci. Int.1993; 60: 183–8.

20. Malatynska E, Waite S, Wei HB, Knapp RJ, Yamamura HI, Roeske WR.Structural correlates for down-regulation of m1 and m2 muscarinicreceptor subtypes. Brain Res. Bull.1998; 47: 285–90.

21. Wei HB, Yamamura HI, Roeske WR. Down-regulation and desensitiz-ation of the muscarinic M1 and M2 receptors in transfected fibroblastB82 cells. Eur. J. Pharmacol.1994; 268: 381–91.

22. Wang SZ, Hu JR, Long RM, Pou WS, Forray C, El Fakahany EE.Agonist-induced down-regulation of m1 muscarinic receptors and reduc-tion of their mRNAlevel in a transfected cell line. FEBS Lett.1990;276: 185–8.

23. Crook JM. A possible role for muscarinic cholinergic receptors in thepathology and treatment of schizophrenia. PhD thesis, University ofMelbourne, 1998.

24. Brann MR, Buckley NJ, Bonner TI. The striatum and cerebral cortexexpress different muscarinic receptor mRNAs. FEBS Lett.1988; 230:90–4.

25. Buckley NJ, Bonner TI, Brann MR. Localization of a family ofmuscarinic receptor mRNAs in rat brain. J. Neurosci. 1988; 8: 4646–52.

26. Levey AI, Kitt CA, Simonds WF, Price DL, Brann MR. Identificationand localization of muscarinic acetylcholine receptor proteins in brainwith subtype-specific antibodies. J. Neurosci.1991; 11: 3218–26.

27. Dean B, Opeskin K, Pavey G, Hill C, Keks N. Changes in the densityof protein kinase C and adenylate cyclase in the hippocampus of sub-jects with schizophrenia. J. Neural Transm. 2000 (in press).

28. Benes FM, Vincent SL, Marie A, Khan Y. Up-regulation of GABAAreceptor binding on neurons of the prefrontal cortex in schizophrenicsubjects. Neuroscience1996; 75: 1021–31.

29. Cornea-Hebert V, Riad M, Wu C, Singh SK, Descarries L. Cellular andsubcellular distribution of the serotonin 5-HT2A receptor in the centralnervous system of adult rat. J. Comp. Neurol. 1999; 409: 187–209.

30. Pootanakit K, Prior KJ, Hunter DD, Brunken WJ. 5-HT2A receptors inthe rabbit retina: Potential presynaptic modulators. Vis. Neurosci.1999;16: 221–30.

31. Morilak DA, Garlow SJ, Ciaranello RD. Immunocytochemical local-ization and description of neurons expressing serotonin2 receptors inthe rat brain. Neuroscience1993; 54: 701–17.

32. Rouse ST, Gilmor ML, Levey AI. Dif ferential presynaptic and post-synaptic expression of m1–m4 muscarinic acetylcholine receptors at theperforant pathway/granule cell synapse. Neuroscience1998; 86:221–32.

33. Vannucchi MG, Pepeu G. Muscarinic receptor modulation of acetyl-choline release from rat cerebral cortex and hippocampus. Neurosci. Lett.1995; 190: 53–6.

34. Suzuki T, Fujimoto K, Oohata H, Kawashima K. Presynaptic M1

muscarinic receptor modulates spontaneous release of acetylcholinefrom rat basal forebrain slices. Neurosci. Lett. 1988; 84: 209–12.

35. Sugita S, Uchimura N, Jiang ZG, North RA. Distinct muscarinic recep-tors inhibit release of gamma-aminobutyric acid and excitatory aminoacids in mammalian brain. Proc. Natl Acad. Sci. USA1991; 88: 2608–11.

36. Hashimoto T, Shu H, Kuriyama K. Muscarinic M1 receptor mediatedinhibition of GABA release from rat cerebral cortex. Neurochem. Int.1994; 24: 389–94.

37. Ramirez MJ, Cenarruzabeitia E, Lasheras B, Del Rio J. 5-HT2 receptorregulation of acetylcholine release induced by dopaminergic stimulationin rat striatal slices. Brain Res.1997; 757: 17–23.

38. Bymaster FP, Calligaro DO, Falcone JF et al. Radioreceptor binding profile of the atypical antipsychotic olanzapine. Neuropsycho-pharmacology1996; 14: 87–96.