Schizophrenia is most closely linked with excess receptor activity for the neurotransmitter

The Dopamine Hypothesis

The dopamine hypothesis of schizophrenia postulates that hyperactivity of dopamine D2 receptor neurotransmission in subcortical and limbic brain regions contributes to positive symptoms of schizophrenia, whereas negative and cognitive symptoms of the disorder can be attributed to hypofunctionality of dopamine D1 receptor neurotransmission in the prefrontal cortex (Toda & Abi-Dargham, 2007). In support of this, studies have shown an increased density of the dopamine D2 receptor in postmortem brain tissue of schizophrenia sufferers (Seeman et al., 2000). It is also reported that upregulation of D2 receptors in the caudate nucleus of patients with schizophrenia directly correlates with poorer performance in cognitive tasks involving corticostriatal pathways (Hirvonen et al., 2004). That dopamine-releasing drugs, such as amphetamine, possess psychotomimetic properties in addition to the D2-antagonist property common to many of the currently prescribed antipsychotic treatments, giving credence to the dopamine hypothesis of schizophrenia.

Abstract

The dopamine hypothesis of schizophrenia suggests that a dysregulated dopamine system contributes to positive, negative, and cognitive symptoms of the disease. Molecular imaging techniques allow accurate measurement of neuroreceptors binding with high sensitivity in the human brain, and these techniques have been abundantly used in the past three decades to examine dopaminergic abnormalities in brain in patients with schizophrenia. The purpose of this chapter is to review the currently available literature on imaging dopamine receptors in patients with schizophrenia. Because schizophrenia is highly heritable, and because the exact neurobiological correlates of genetic risk are largely unknown, we will focus on studies examining dopaminergic abnormalities in individuals at genetic risk. These studies suggest that dopaminergic abnormalities in schizophrenia are shared by individuals at genetic risk who do not express the illness, suggesting a “dopamine hypothesis of schizophrenia vulnerability”. This hypothesis should allow us to better understand the dopaminergic dysfunction in the context of the complex pathophysiological process leading to schizophrenia.

Abstract

The dopamine hypothesis of schizophrenia, which was formulated in the 1960s after the discovery of the antipsychotic actions of chlorpromazine, was extremely successful as a heuristic principle for interpreting aspects of the phenomenology of schizophrenia. The development of improved antipsychotic medications was guided by a search for dopamine blockers based on the concept that schizophrenia is, in part, a hyperdopaminergic state. Molecular imaging studies performed over the past 25 years strongly support an association of increased subcortical dopamine transmission with the positive symptoms of schizophrenia, with the caveat that this finding is not pathognomonic due to neurochemical heterogeneity of populations of schizophrenia patients. Although subcortical hyperdopaminergia contributes importantly to aberrant salience (manifesting in positive symptoms), the original dopamine hypothesis must be extended to include contributions of other neurotransmitter systems, with glutamate being particularly implicated in the pathophysiology of schizophrenia.

The dopamine hypothesis of schizophrenia

The dopamine hypothesis of schizophrenia has so far been the most influential hypothesis about schizophrenia (Howes and Kapur, 2009). In 1966 Jacques Van Rossum proposed that “overstimulation of dopamine receptors could be part of the etiology” of schizophrenia (for a historical review: (Baumeister and Francis, 2002)). The hypothesis was originally based on the observation that known psycho-stimulants, such as amphetamine, induce stereotypic motor behaviors. These behaviors could be blocked by antipsychotic medication, such as chlorpromazine, which by interfering with dopamine function was known to lead to parkinsonian-like movement disorders. Arguably, the strongest support for the dopamine hypothesis was provided in the 1970th by Solomon Snyder and Philip Seeman who found that the efficacy of antipsychotic medication correlated directly with its occupancy of dopamine receptors.

The identification of an effective drug target for psychosis does however not necessarily imply that this target needs to be involved in the pathophysiology or even the etiology of schizophrenia. One initial criticism of the dopamine hypothesis has therefore been that it is not based on measurable physiological alterations in the dopamine system. Because overwhelming evidence for alterations in the brain dopamine system has been found in the last two decades, a role of dopamine in the pathophysiology of the disease is not questioned any longer by most scientists. Whether dopamine is also involved in the etiology of the disease is still unknown. This question is much harder to address because schizophrenia is considered a neuro-developmental disease, consequently patients are diagnosed long after the disease has started its course.

INTRODUCTION

The dopamine hypothesis of schizophrenia is based on a wide variety of circumstantial evidence, as follows:

High doses of dopamine-mimetics elicit hallucinations (Angrist et al., 1974; Snyder, 1976).

Neuroleptics accelerate the turnover of brain dopamine (Da Prada and Pletscher, 1966; Rollema et al., 1976).

Neuroleptics block the action of dopamine-mimetics (Van Rossum, 1966; Niemegeers and Janssen, 1979). Thus, the neuroleptics elicit catalepsy, tremor, akinesia and rigidity, a syndrome similar to Parkinson's disease where the brain is grossly deficient in dopamine (Hornykiewicz, 1975).

Neuroleptics also block dopamine receptors in the pituitary, resulting in excessive release of prolactin (Caron et al., 1978).

There is an excellent structural fit between the neuroleptics and dopamine (Horn et al., 1975; Philipp et al., 1979).

Thus, the overactive dopamine transmission which has been postulated to occur in schizophrenia may stem from:

increased content and release of dopamine;

increased number or sensitivity of post-synaptic dopamine receptors or;

decreased number of presynaptic dopamine receptors.

B Elevation of Dopamine D2 Receptors in Schizophrenic Brain:

Our current classification of brain dopamine receptors is shown in Table 1. The term D1 simply refers to the site for dopamine-sensitive adenylate cyclase, without implying whether or not this enzyme is linked to any other cell component. The entire classification is based on the molarities of dopamine and neuroleptics (e.g. spiperone) to which the site is sensitive. A thorough discussion and analysis is presented elsewhere (Seeman, 1980).

TABLE 1. CLASSIFICATION OF BRAIN DOPAMINE RECEPTORS

	D1 SITE*	D2 RECEPTOR	D3 SITE	D4 SITE#
DOPAMINE IC50:	3000 nM	5000 nM	3 nM	20 nM
SPIPERONE IC50:	2000 nM	0.3 nM	1500 nM	0.1 nM
Density (Bmax):	2000 fm/mg	300 fm/mg	80 fm/mg	100 fm/mg
Synaptic locus:	post	post-	50% pre-	post?
In Pituitary?	no	yes	no	Int.lobe
Thermal stability:		good	poor
Best 3H-ligands:	flupen-thixol	haloperidol domperidone	dopamine	ADTN;Apo (±spip.)
% solubilizible:		18%	36%
Alkylated by:		phenoxybenz.	NCA
Nigral lesion:	up**	up	down**
Parkinson's		up	down
Schizophrenia:		up	normal
Related to:
Neuroleptic doses:	no	yes	no
Locomotor doses:	no	yes	no	no
Rotation doses:	no	yes	no	no
Emetic doses:	no	yes	no	no
Prolactin release:	no	fair	no	good

NCA = (-)-N-2-chlorethyl-norapomorphine.

Apo = R (-)-apomorphine.

ADTN = 6, 7-dihydroxy-2-aminotetralin.

Spip. = 30 nM spiperone, which is used to define the displaceable binding of 3H-ADTN or 3H-apomorphine to the D4 site in the striatum; this site has not yet been detected by the 3H-ligand-binding method in the pituitary.

*dopamine-stimulated adenlylate cyclase.#dopamine-inhibited adenylate cyclase (Meunier; Labrie; De Camillo).**up or down refers to change in density in the striatum.

After first developing the radioreceptor assay for dopamine receptors using 3H-haloperidol (Seeman, et al., 1974, 1975a, 1975b), we later found that the dopamine containing regions in schizophrenic brain had more D2 receptors than control tissues (Lee and Seeman, 1977, 1978a, 1980a,b). These data were confirmed by Crow et al. (1978), Owen et al. (1978) and by Reisine et al. (1980). Although the first report of Mackay et al. (1978) could not confirm our findings, some of their recent work does (Mackay et al., 1980). A summary of all these findings is given in Table 2 and Fig. 1.

TABLE 2. DOPAMINE RECEPTORS IN POST-MORTEM SCHIZOPHRENIC BRAINS [up to Jan. 1981]

3H-ligand	nM baseline	NORMAL fmol/mg protein	SCHIZOPHRENIC fmol/mg	% difference	Reference*
D2 RECEPTORS:
CAUDATE NUCLEUS
3H-HALO(1.7 nM)	100 B	45 ± 2 (N = 21)	78 ± 3 (N = 18)	+74%	Lee & Seeman, 1977, 1978.
3H-SPIP(1 nM)	1000 B	99 ± 15 (N = 5)	159 ± 18 (N = 9)	+59%	Lee & Seeman, 1977, 1978.
3H-SPIP(0.8 nM)	100 B	95 (N = 18)	150 (N = 19)	+58%	Crow et al., 1978.
3H-SPIP(Bmax)	100 B	167 ± 50 (N = 15)	340 ± 120 (N = 15)	+103%	Owen et al., 1978.
			256 ± 41 (N = 2)°°	+53%	Owen et al., 1978.
			250 ± 68 (N = 5)°	+50%	Owen et al., 1978.
3H-HALO(2 nM)	100 B	45 ± 1 (N = 39)	{86±4(N=29)75±9(N=5)°	+93%	Lee & Seeman, 1980.
			+67%	Lee & Seeman, 1980.
3H-SPIP(1 nM)	1000 B	101 ± 5 (N = 27)	{149±7(N=29)132±8 (N=8)	+47%	Lee & Seeman, 1980.
			+31%	Lee & Seeman, 1980.
3H-SPIP(0.1nM)	100 B	90 ± 9 (N = 11)	124 ± 11 (N = 11)	+38%	Reisine et al., 1980.
3H-SPIP(Bmax)	10000 Su	127 ± 15 (N = 5)	204 ± 47 (N = 3)	+61%	Lee & Seeman, this work.
3H-SPIP(Bmax)	1000 B	39 ± 6 (N = 17)*	{65±6(N=21)35±4(N=7)**	+67%	Mackay et al., 1980.
			−10%	Mackay et al., 1980.
PUTAMEN
3H-HALO(1.7nM)	100 B	50 ± 3 (N = 21)	75 ± 4 (N = 18)	+15%	Lee & Seeman, 1977, 1978.
3H-SPIP(1 nM)	1000 B	84 ± 6 (N = 5)	154 ± 17 (N = 9)	+75%	Lee & Seeman, 1977, 1978.
3H-SPIP(0.8nM)	100 B	90−(N = 19)	136− (N = 19)	+51%	Crow et al., 1978.
3H-HALO(2 nM)	100 B	47 ± 2 (N = 39)	{77±3(N=30)73±8(N=7)°	+64%	Lee & Seeman, 1980.
			+55%	Lee & Seeman. 1980.
3H-SPIP(1 nM)	1000 B	104 ± 4 (N = 33)	{150±8(N=37)129±17 (N=11)°	+45%	Lee & Seeman, 1980.
			+24%	Lee & Seeman, 1980.
3H-SPIP(Bmax)	1000 B	66 ± 9(N = 6)	136 ± 14 (N = 8)	+105%	Lee & Seeman, 1980.
3H-SPIP(0.1nM)	100 B	75 ± 13 (N = 11)	114 ± 8 (N = 11)	+52%	Reisine et al., 1980.
3H-SPIP(Bmax)	100 B	213 ± 21 (N = 3)	308 ± 11 (N = 3)	+45%	Reisine et al., 1980.
3H-FPT				+75%	Cross et al., 1980
3H-5PIP(Bmax)	10000 Su	202 ± 10 (N = 23)	297 ± 35 (N = 23)	+47%	Lee & Seeman, this work.
NUCLEUS ACCUMBENS
3H-HALO(1.7nM)	100 B	37 (N = 2)	63 (N = 2)	+71%	Lee & Seeman, 1977, 1978.
3H-HALO(2 nM)	100 B	38 (N = 2)	79 ± 7 (N = 9)	+110%	Lee & Seeman, 1980.
3H-SPIP(1 nM)	1000 B	55 (N = 2)	118 ± 18 (N = 4)	+115%	Lee & Seeman. 1980.
3H-SPIP(0.8nM)	100 B	79 (N = 17)	108 (N = 17)	+37%	Crow et al., 1978.
3H-SPIP(0.5nM)	10000 A	54 (N = 16)	64 (N = 26)	[+19%ns]	Mackay et al., 1978.
3H-SPIP(Bmax)	10000 Su	124 (N = 5)	161 (N = 3)	+30%	Lee & Seeman, this work.
3H-SPIP(Bmax)	10000 A	27 ± 3 (N = 16)*	{44±4(N=13)* *26±8(N=4)∘	+63%	Mackay et al., 1980.
			−4%	Mackay et al., 1980.
D3 SITES:
CAUDATE NUCLEUS
3H-APO (3.2nM)	1000 B	28 ±3 (N = 13)	27 ± 4 (N = 7)	0%	Lee & Seeman, 1977, 1978.
3H-D (Bmax)	100 Apo	68 ±8 (N = 4)	61 ± 5 (N = 3)	−10%ns	Lee & Seeman, this work.
PUTAMEN
3H-APO (3.2nM)	1000 B	25 ± 3 (N = 13)	26 ± 2 (N = 7)	0%	Lee & Seeman, 1977, 1978
3H-ADTN (7.5nM)	1000 D	42 ± 6 (N = 17)	49 ± 4 (N = 19)	+17%ns	Cross et al., 1979.
3H-D (Bmax)	100 Apo	70 ± 7 (N = 17)	66 ± 6 (N = 16)	−6%ns	Lee & Seeman, 1980b.
NUCLEUS ACCUMBENS
3H-APO (3.2nM)	100C B	23 ± 3 (N = 4)	19 ± 4 (N = 3)	0%	Lee & Seeman, 1977, 1978.

Abbreviations:

A: ADTN or (1)-6,7-dihydroxy-2-aminotetralin.

∘: Patients off neuroleptics for 1 mo. or more.

B: (+)-butaclamol

ns: Not statistically significant.

∘∘: Never had medication.

D: Dopamine

FPT: Cis-flupenthixol

*: Assuming 15% of wet tissue is protein.

Su: Sulpiride

**: All patients on neuroleptics.

Apo: Apomorphine

HALO: Haloperidol

SPLP: Spiperone

* 1978 in each case reters to reference 1978a; 1980 refers to reference 1980a and 1930b.

Fig. 1. Densities of D2 receptors as revealed by Scatchard analysis of 3H-spiperone binding in normal and schizophrenic caudate, putamen and nucleus accumbens. Each solid bar represents average maximum density determined from the number of separate brains assayed as indicated. C stands for control normal brains and s stands for schizophrenic brains.

“Lee and Seeman 1981” refers to this report.

An elevated density of D2 receptors in the schizophrenic brain is also found in post-mortem tissues wherein the schizophrenic had taken little or no neuroleptic medication (see Fig. 1). Mackay et al. (1980) state, however, that the elevated density of D2 receptors is only found in tissues from medicated patients. It is difficult to accept this finding of Mackay et al., not only because it is out of line with all other data, but also because their absolute densities of D2 receptors is so extremely low (see Fig. 1.); these low values suggest that considerable degradation of the D2 receptors had taken place under the condition used by those workers.

Reynolds et al. (1980) found normal densities for 3H-spiperone binding to the post-mortem putamens from 12 schizophrenic patients, four of whom had not received neuroleptics. Since the KD values for the schizophrenic tissues (0.9 nM) were much higher than those for the control tissues (0.13 nM), this indicates that considerable residual neuroleptic had remained in the tissue during the assay.

Fig. 2. Elevation of D2 receptors in 22 schizophrenic putamens compared to 23 normal human putamens. Maximum densities were determined by Scatchard analysis using 3H-spiperone in the presence or absence of 10 μM sulpiride. The horizontal lines, interrupted and solid, indicate the mean and S.E.M., respectively, for each group.

Such differences in KD can mask potential differences in receptor density (Bmax). This is why we recommend that the tissue be washed at least four times, and that 10,000 nM sulpiride be used to define displaceable binding (to preclude measurement of S2 serotoninergic sites).

C Schizophrenic Brain Has a Normal Density of Pre-synaptic Dopamine Receptors (D3 Sites):

It is conceivable, of course, that any possible hyperdopaminergic transmission in schizophrenia could also arise from a deficiency in pre-synaptic dopamine receptors (i.e. autoreceptors). We explored this possibility after first determining that approximately 50% of the D3 sites (as labelled by 3H-dopamine or 3H-apomorphine) are located on pre-synaptic nerve terminals, as based on the following findings:

1.

The striata of nigral-lesioned rats revealed a marked reduction in the density of D3 sites (Nagy et al., 1978; Sokoloff et al., 1980).

2.

The Parkinson's diseased striatum reveals a marked reduction in the density of D3 sites (Lee et al., 1978b, 1981).

3.

Rat pups receiving 6-hydroxy-dopamine (intracisternally) at day 5 of age subsequently have only 50% of the normal density of D3 sites in the striatum (Watanabe, Seeman and Shaywitz, to be published).

We found that the schizophrenic brain tissues contained normal densities of D3 sites. This is illustrated in Fig. 3 and Table 2. This observation concurs with the findings of Cross et al. (1979) who found normal amounts of 3H-ADTN sites in schizophrenic brains. It had previously been established that 3H-apomorphine and 3H-ADTN label the same types of dopaminergic sites (Seeman et al., 1979).

Fig. 3. Densities of D3 site as determined by 3H-dopamine in 23 normal human putamens and 22 schizophrenic putamens. Specific 3H-dopamine binding was defined by 100 nM apomorphine. The horizontal lines, interrupted and solid, indicate the mean and S.E.M., respectively, for each group.

F Schizophrenia

The dopamine hypothesis of schizophrenia has already been established in the 1960s, but it still forms the most prominent assumption about the pathophysiological mechanisms that underlie the development of this severe mental disease (Keshavan, Tandon, Boutros, & Nasrallah, 2008). It originated from the observation that typical antipsychotics are able to alleviate schizophrenic, particularly positive symptoms by blocking dopamine D2 receptors (Seeman, Lee, Chau-Wong, & Wong, 1976). However, other neurotransmitter systems appear to contribute to the pathophysiology of schizophrenia as well. In this context, serotonergic neurotransmission is of particular scientific interest given its role in the therapeutic effects of atypical antipsychotics in the treatment of schizophrenic negative symptoms (Meltzer, 1999). Further evidence for an increased serotonergic neurotransmission has been provided by several postmortem studies, CSF studies of 5-hydroxyindoleacetic acid, genetic as well as neuroimaging studies (Sawa & Snyder, 2002).

In line with the hypothesis of an increased central serotonergic activity in schizophrenia, a decreased LDAEP could be demonstrated in unmedicated schizophrenia patients compared to healthy controls (Juckel et al., 2003). This finding could be replicated in a second study reporting a weaker LDAEP, indicating high serotonergic activity, in schizophrenic patients compared to healthy controls (Juckel et al., 2008b). Interestingly, this finding was also valid for patients under antipsychotic medication supporting the assumption that enhanced serotonergic neurotransmission, as shown by a weak LDAEP, could be a trait characteristic in schizophrenia. Similar results have been reported by Park et al. (2010) for schizophrenia patients under ongoing antipsychotic medication.

The outbreak of manifest schizophrenia is typically preceded by a characteristic prodromal phase. This initial prodrome of schizophrenia is defined as a period of time which begins with the first behavioral, cognitive and perceptual changes in a person and extends up to the development of the first psychotic symptoms (Fusar-Poli et al., 2013; Yung & McGorry, 1996). The so-called prodromal symptoms include cognitive basic symptoms (disturbances of thought, language, perception and attention), attenuated psychotic symptoms (delusional ideas, perceptual abnormalities, disorganized speech) and brief limited intermittent psychotic symptoms (hallucinations and delusions for less than 7 days) (Schultze-Lutter et al., 2015). Early recognition of the prodromal phase and early therapeutic interventions might be helpful to prevent or to delay the transition to schizophrenia in vulnerable persons. Therefore, it is of great importance to identify the potential neurobiological markers and to elucidate the pathophysiological mechanisms that underlie the transition to schizophrenia. In this context, early and late neurodevelopmental disturbances in schizophrenia have been studied and results show that structural und functional brain abnormalities are already apparent in the premorbid stage (Juckel et al., 2012; Pantelis et al., 2005; Witthaus et al., 2010). However, changes in the action of brain neurotransmitters, especially serotonin, are more difficult to measure in humans than structural abnormalities, although it is of great interest for early recognition to know whether altered serotonergic activity can be interpreted as an indicator for vulnerability of schizophrenia or for the progression of the illness.

In this context, a recent study examined the LDAEP in subjects at high-risk and ultrahigh-risk for developing schizophrenia, but did not find any differences compared to healthy controls (Hagenmuller et al., 2016). Another study addressed the same question of whether dysfunctional serotonergic neurotransmission is already present in subjects of at-risk mental state for schizophrenia before the onset of psychotic symptoms, and whether serotonergic activity further increases during the development of schizophrenia and the chronic course (Gudlowski et al., 2009). Interestingly, prodromal subjects showed significantly weaker LDAEP compared to healthy controls, but no differences compared to first-episode and chronic patients. Moreover, this result was not influenced by covariates such as age, gender, medication, age of onset, or psychopathology. A further evaluation of a subsample of the prodromal subjects 10 months later revealed that LDAEP values remained unchanged in comparison to the first testing. In addition, there was a close relationship between LDAEP and prodromal negative symptoms. This result suggests that serotonergic neurotransmission had already increased before the onset of the full-blown psychosis of schizophrenia and remained enhanced in the further course of the disease. In conclusion, a weak LDAEP may therefore represent a biological vulnerability marker of schizophrenia rather than an expression of illness progression.

Although the crucial role of serotonergic neurotransmission in the pathophysiology of negative symptoms in schizophrenia and the efficacy of atypical antipsychotics and some antidepressants in the treatment of this disabling syndrome via serotonergic action are, well documented, the association between serotonergic functioning and the extent of negative symptoms is still not clear. A recent study investigated the status of central serotonergic activity in negative symptoms in schizophrenia by using the LDAEP (Wyss et al., 2013). Interestingly, schizophrenia patients with predominant negative symptoms showed significantly higher LDAEP values in both hemispheres compared to healthy controls. Moreover, the LDAEP in the right hemisphere in patients was related to higher scores of negative symptoms. A relationship with positive symptoms was not found. These results might suggest a decreased central serotonergic activity in patients with predominant negative symptoms. However, a subsequent study with a larger sample size revealed a strong opposite result (Uhl et al., 2018). In this study, predominant negative symptoms in patients with schizophrenia were associated with reduced LDAEP values, indicating an enhanced serotonergic neurotransmission. As schizophrenic disorders characteristically begin with prodromal negative symptoms, and a weak LDAEP appears to be a biological trait marker of this disease, an upregulated serotonergic system might be closely related to the origin of negative symptoms and schizophrenia.

CONCLUSIONS

The DA hypothesis of schizophrenia has been a useful paradigm for investigation as evidenced by the many studies reported here. However, the available biochemical approaches have not confirmed a DA disturbance as the primary etiology in schizophrenia. Indeed, indirect pharmacologic studies are still the major support for the hypothesis despite the extensive biochemical investigation of schizophrenic patients. Furthermore, pharmacologic evidence does not necessarily indicate the primary locus of the defect. Nearly all pharmacologic agents active on DA systems also notably affect other neurotransmitter systems. The relationships between central NE [83], serotonin (5-hydroxytryptamine; 5-HT) [61], γ-aminobutyric acid (GABA) [88], substance P [19], endorphins [1], and other neurotransmitter systems and DA activity in schizophrenia require further study.

The hypothesis that schizophrenics have supersensitive brain DA receptors is supported by the postmortem brain DA receptor binding studies of four of five laboratories and the neuroendocrine reports of an exaggerated GH response to apomorphine in acute patients. However, the hypothesized DA receptor supersensitivity is challenged by the suggestions that the elevated receptor binding is related to neuroleptic treatment [46] and by the reported lack of an enhanced sensitivity in schizophrenics to amphetamine-induced psychosis following abrupt withdrawal of neuroleptic treatment [87].

Modifications in the hypothesis which might lead to further understanding of the syndrome's pathogenesis include (1) the DA abnormality may only occur in a very specific brain area (e.g., prefrontal cortex) and (2) the primary disturbance in schizophrenia may occur in another neurotransmitter system that interacts with DA neurons. Neuroleptics may thus be operating on a “secondary” DA system. Similarly, although anticholinergic drugs are of clinical benefit in Parkinson's disease, the primary defect in parkinsonism lies in the nigrostriatal DA system rather than in a cholinergic system. Furthermore, although neuroleptics rapidly produce DA receptor blockade, as evidenced by the rapid neuroleptic-induced rise in plasma PRL [49], the full clinical antipsychotic response to them requires a number of weeks. Thus, whereas DA receptor blockade does appear necessary for the antipsychotic effects of neuroleptic medication, that blockade may allow other slower processes to take place which are more directly responsible for the therapeutic change. (3) Several biochemical factors involved in central DA function (e.g., low MAO, low DBH, DA receptor supersensitivity) may each be a vulnerability factor toward the illness. That is, each abnormality may be a necessary but not sufficient element for the development of schizophrenia. (4) The heterogeneity of the clinical syndrome of schizophrenia itself may be responsible for the inconclusive results. As stated earlier, schizophrenia probably represents a variety of disease entities, each having a different biologic dysfunction [16]. Some or all of these may entail a defect in DA systems. Thus, Crow [25] has attempted to draw a neurobiologic distinction between schizophrenic patients who have good antipsychotic responses to neuroleptic treatment and patients who remain psychotic during such treatment. He proposed that there are two syndromes with distinct disease processes: (1) an acute episodic schizophrenic syndrome with positive symptoms reversed by neuroleptic treatment, the illness thus being associated with increased DA neurotransmission (type I syndrome); and (2) a chronic deteriorating syndrome with negative symptoms not reversed by neuroleptic treatment, the illness thus being unrelated to DA transmission, but possibly related to structural brain changes (type II syndrome). Recent pharmacologic [54], neuroendocrinologic [40], and neuroradiologic [90] reports have provided preliminary support for this hypothesized distinction. The DA hypothesis may then only apply to the type I subgroup.

Because of the clinical heterogeneity of people diagnosed as schizophrenic and the complex relationships among neurobiologic systems, rather than attempting to find a single “cause” for the entire spectrum of schizophrenia, we suggest that studies concentrate on two more modest goals.

First, a finer delineation of diagnostic and biologic heterogeneity would be obtained by identifying the following: (1) clinical (paranoid versus catatonic, early versus late onset); (2) pharmacologic (neuroleptic responders versus partial responders versus nonresponders); and (3) biochemical (high versus low CSF substance levels) subgroups in large populations of schizophrenic patients. The next step would be to identify patterns in these subgroups. The presumed heterogeneity of the disorder poses special problems for the clinical investigator. Statistically significant findings in a large group of patients are very likely to be secondary to the previously discussed nonspecific factors and to artifacts such as drug treatment (past or present). On the other hand, studies with a small patient sample are not likely to recognize an abnormality that may occur in only a small proportion of patients diagnosed as schizophrenic. A fruitful approach to finding this subgroup would be to focus on those patients with extremely aberrant values, even though they may not affect the statistical significance of the entire study population. In this way, the biologic value can be used as an independent variable to identify a subgroup of schizophrenic patients with consequences for etiology, course, and treatment response.

The second goal should be to relate biologic factors to specific component behaviors that make up the schizophrenic disorders: one can classify the behavioral components into separate groups to examine whether specific biologic variables relate more to one of these component groups than to the variety of behavioral disorders grouped under the diagnosis schizophrenia. A distinction that we think especially useful in conceptualizing schizophrenia is that of “state components” and “trait components.” State components refer to aspects of the psychotic state itself, such as behavioral disorganization, hallucinations, and delusions. Specific state-related biologic concomitants may relate primarily to the psychotic state and would be less evident during periods of remission. Trait components would be those aspects evident in the prepsychotic or postpsychotic period, such as social isolation, affective blunting, impaired role functioning, impaired eye tracking, CAT scan abnormalities, or other as yet unknown behaviors. Trait-related biologic concomitants would relate to behaviors of the nonpsychotic state, would not change over time, and thus could reflect a genetic vulnerability to psychotic decompensation. Further delineation of biologic measures that are state-related or trait-related would provide an approach to understanding those aspects of the illness that are present in a range of people, including nonschizophrenics, as well as to understanding those aspects that are illness specific.

The dopamine hypothesis of schizophrenia

The ‘dopamine hypothesis of schizophrenia’, simply stated, postulates that certain dopaminergic pathways are overactive in schizophrenia and so cause the symptoms of an acute schizophrenic episode. Clinical studies indicate that drugs like L-dopa or amphetamine, which potentiate dopaminergic activity, may induce or exacerbate schizophrenic symptoms.

When the antipsychotic drugs were first introduced, their mode of action was unknown. At first, studies in the peripheral nervous system suggested that the anti-adrenergic effects of chlorpromazine probably explained its antipsychotic action, perhaps by reducing arousal. However, the fact that potent anti-adrenergic agents had no antipsychotic benefit did not support this hypothesis. Carlsson & Lindqvist (1963) first suggested that DA receptor blockade was the basis of antipsychotic effects. The low activity of butyrophenone antipsychotics at DA receptor sites linked to adenylate cyclase stimulation was seen as evidence against this idea. It was supported, however, by the recognition of two types of DA receptor. One (called D1) was linked to adenylate cyclase stimulation, and another, higher affinity one (called D2) was sometimes associated with adenylate cyclase inhibition and exhibited preferential binding of butyrophenones.

Neuropharmacological studies provide virtually all the evidence to support the ‘dopamine hypothesis of schizophrenia’. Although some of the newer so-called ‘atypical’ antipsychotic agents are weak DA receptor antagonists, all effective antipsychotics are believed to share the ability to impair dopaminergic neurotransmission. Postmortem studies of schizophrenic brains have demonstrated increased DA receptor (D2) densities, but these densities are probably considerably influenced by ante-mortem drug treatments. Positron emission tomographic studies of D2 receptor binding in antipsychotic-naive schizophrenic patients have provided conflicting results.

The CNS location of the site of antipsychotic drug action is unknown and subject to much debate. DA receptors are present in the basal ganglia, the mesolimbic system, the tuberoinfundibular region and, to a much lesser extent, in the cerebral cortex. Studies on the effects of dopaminergic transmission of psychotomimetic agents such as amphetamine, PCP and benzmorphan point to a possible common mechanism of psychotic action. Carlsson (1988) proposed that ‘information overload’ and ‘hyper-arousal’ are integral features of many psychotic illnesses. He postulated that these features arise because of impairment in the mesolimbic system's protective effects on cortical function. In health, Carlsson argued, mesolimbic glutamate-releasing neurons oppose mesolimbic dopaminergic pathways and maintain this protective function. In this model, mesolimbic DA dysregulation is considered secondary to frontal dysfunction. A further recent elaboration on the DA hypothesis of schizophrenia considers the function of the mesolimbic DA system in assigning importance, or salience, to stimuli or ideas (Kapur 2003; Murray et al 2008). It is proposed that DA attaches a label (e.g.‘dangerous’, ‘pleasant’, etc.) to stimuli and ideas and that with the labels in place, motivation and goal-directed behaviour easily follows. In schizophrenia, excess DA leads to the assignment of ‘labels’ or salience to irrelevant or insignificant thoughts or events, creating a psychotic state.

1 Introduction

The classical dopamine hypothesis of schizophrenia proposed that hyperactivity of dopamine transmission is responsible for the positive symptoms (hallucination, delusions) observed in this disorder. This hypothesis was supported by the strong correlation between the clinical effectiveness and the ability to block dopamine D2 (D2R) receptors.1–3 It has also been demonstrated that deficit in dopamine transmission in the prefrontal cortex and lack of stimulation of D1 (D1R) receptors induces cognitive impairments reminiscent of those observed in schizophrenia patients.4 These two receptors appear to be important also because of their physical interaction. Dopamine D1 and D2 receptors exist as receptor homo or heterodimers complexes that can exhibit pharmacological and functional properties distinct from their constituent receptors. Dopamine D1R-D2R heterodimers were first identified in vivo by co-immunoprecipitation in the rat striatum5 and soon thereafter confirmed by studies on HEK 293 cells line,6 and subsequently in primary striatal neuronal culture by fluorescence resonance transfer (FRET) studies using confocal microscopy.7 The D1R-D2R heterodimers have been shown to exhibit pharmacological and cell signaling properties distinct from its constituent receptors5,7–9 and the expression of dopamine D1R-D2R heterodimers in the mesocorticolimbic system and basal ganglia nuclei10–12 suggest that this receptor complex may have etiological significance in disorders characterized by abnormal dopamine signaling. For example, the calcium signaling elicited by the dopamine receptors heterodimers, through activation of Gq/11 protein and phospholipase C (PLC), resulted in the activation of calcium calmodulin kinase II13,14 and consequently increased expression of brain-derived neurotrophic factor (BDNF) in the nucleus accumbens (Nac) and ventral tegmental area (VTA).7,14 Moreover, a potential role of D1R-D2R heterodimers in schizophrenia has been suggested since altering of Ca2 +-signaling has been reported in post-mortem studies of brains obtained from schizophrenic patients.15 The affinity of ligands to the dopamine receptors has been shown to be highly dependent on co-existing homo- or heterodimers.16–20 In addition to the interaction described above, dopamine D2R can form dimers with other dopamine receptors: D3R,21 D4R22 and D5R23 as well as with other GPCRs (http://www.gpcr-hetnet.com). The pharmacological profile of atypical antipsychotic drugs action indicates that the interaction between dopamine D2 and serotonin 5-HT1A or 5-HT2A receptors seems to be important.24,25 It is well established that dopamine D2 and serotonin 5-HT1A and 5-HT2A receptors play an important role in neurotransmission and that alterations in their functioning have been implicated in many human neurological and psychiatric disorders, including schizophrenia. All of these receptors are co-localized on the same neuron and form heterodimers.18,25–29 For a long time, the main focus has been attributed to the antipsychotic effects of D2R blockade (in the extended striatum), what alleviated the positive symptoms of schizophrenia. Further studies have shown the significant effects of antipsychotic drugs that may be beneficial for negative (cognitive) symptoms via a combined effect on dopamine and serotonin and other classes of neuroreceptors.30 Significant evidence indicates that, in fact, the balance between the properties of dopamine D2R and serotonin receptors has a profound effect on the profile of these drugs in preclinical models.31 However, in studies related to the potential role of GPCR receptor dimers, it appears that the key role of these interactions plays dopamine D2R. This receptor is also a crucial player in schizophrenia, so naturally gene encoding this receptor is considered a candidate risk gene of this disease.31a Recently, the data obtained from genome-wide associated study (GWAS) further confirmed the involvement of this receptor in schizophrenia.31b However, the results of the research indicate that the role of D2R polymorphisms in schizophrenia still remains controversial.

The first association between schizophrenia and the genetic variant of D2R was observed for missense nucleotide (SNP, single nucleotide polymorphism) change causing an amino acid substitution of serine into cysteine at codon 311: rs1801028 (Ser311Cys) in the Japanese population, which was described by Arinami et al.32 Soon after, other studies appeared, confirming33–38 or negating39–43 correlations of this polymorphism with schizophrenia. It seems that origin and heredity are responsible for the lack of homogeneity of the described results.44,45 There are a number of polymorphisms within the D2R gene, although only a few appear to be potentially important for schizophrenia. One of them is polymorphism that also occurs within exon 7 is rs1800496 (Pro310Ser) for which family-based association studies strongly implicated a risk for schizophrenia in Han Chinese from Taiwan.46 However, this genetic variant is not widely studied in the context of association with schizophrenia. In other in vitro studies it has been shown that C957T polymorphism rs6277 (Pro319Pro/C > T) has marked functional consequences for D2R mRNA stability and dopamine-regulated D2R expression.47 Moreover, the C957T affected striatal D2R binding in healthy humans48 and this SNP also correlated with schizophrenia.47,49,50 Another genetic variant potentially relevant in schizophrenia is polymorphism rs1799732 − 141C Ins/Del, which is located in the promoter sequence, downstream of D2R coding sequence. It has been demonstrated to alter gene expression in vitro and striatal receptor density in vivo, and it has a replicable effect on antipsychotic treatment response.51 On the other hand, in a meta-analysis performed by Glatt et al.,51a this genetic variant was not associated with schizophrenia. Recent study also indicated a lack of association of this genetic variant with schizophrenia.38 Besides the association between − 141C Ins/Del mutation and clozapine response52 another SNP rs1800497, also known as the TaqIA (or Taq1A) polymorphism of the dopamine D2R, has been identified within exon 8 of the ankyrin repeat and kinase domain containing 1 (ANKK1) gene, 10 kb away from D2R gene in the 3′ untranslated region. This polymorphism, which leads to a substitution of glutamic acid for a basic lysine (Glu713Lys) that may alter substrate-binding specificity,53 likely modulates the function and expression of D2R due to its close proximity. Despite the fact that this SNP is localized to the ANKK1 gene, it seems to be in linkage disequilibrium with several D2R genetic variants, which could potentially explain a dopaminergic role in the etiopathogenesis of schizophrenia.54 However, recent meta-analysis showed a lack of correlation of this polymorphism with schizophrenia.38

Among the genes encoding other dopamine receptors, equally important in the context of schizophrenia, is the gene encoding for D3R dopamine receptor. Especially one polymorphism, rs6280, also known as Ser9Gly, has its influence: it has been demonstrated that binding affinity of selective ligands for D3R is higher, as has been shown for mutant homozygote.55 However, meta-analysis studies are also inconsistent.55a–58 On the other hand in a study of patients treated with olanzapine, those who were rs6280(C;C) homozygotes had greater positive symptom remission,59 while having a minor role or lack of association in clozapine response.60 This result might eventually become more clearly defined with increased sample size.61

As stated by Scharfetter,62 although a single polymorphism may not play a significant role or be marginal, the accumulation of these polymorphisms may already play a crucial role in both the development of schizophrenia as well as in response to treatments. From this point of view, it is interesting to analyze the effect of SNP on the interaction between receptors for which they have been shown to form functional dimers relevant to this disease.

4.4 Implications for the dopamine hypothesis of schizophrenia

It is important to distinguish between the dopamine hypothesis of schizophrenia and the dopamine hypothesis of antipsychotic drug action. The latter proposes that reduction of dopaminergic function via either blockade of postsynaptic receptors or attenuation of presynaptic neuronal activity underlays the therapeutic effect of most known antipsychotic agents. Conversely, the former takes this concept a stage further and from it proposes that dopaminergic hyperfunction, via either supersensitivity of postsynaptic receptors or elevated activity of presynaptic neuronal activity, is an important element in the pathophysiology of schizophrenia (Carlsson, 1988).

The notions discussed in this chapter concern variants of this long-standing dopamine hypothesis of antipsychotic drug action, in terms of differing roles for distinct receptor subtypes in regulating dopamine-mediated function. In themselves, they do not yet demand any fundamental revision to the dopamine hypothesis of schizophrenia, pending more extensive feedback from clinical trials, but there are other reasons for contemplating such revision. Philosophically, the one hypothesis need not follow necessarily from the other. Indeed the search for neurochemical correlates of putative dopaminergic hyperfunction, either in post-mortem brain tissue (see Reynolds, 1989) or in vivo by positron emission tomography (PET; see Waddington, 1989d, and Chapter 5), has produced insubstantial or contradictory findings. Furthermore, much current theory considers schizophrenia to be a neurodevelopmental disorder of early origin (Weinberger, 1987; Murray & Lewis, 1987; Waddington & Torrey, 1991), with an emerging focus from recent neuropathological and magnetic resonance imaging studies on dysplasia of temporal lobe and related structures (Roberts, 1990; Waddington et al., 1990; Waddington & Torrey, 1991). Curiously, one of the temporal lobe regions implicated more consistently in these processes, the parahippocampal gyrus/entorhinal cortex, shows in animals not only the high ratio of D1 to D2 receptor densities characteristic of several cortical regions but also an unusually high endogenous dopamine content (Dewar & Reader, 1989); the significance of these associations is unclear, but may repay further study.

Regarding the dopamine hypothesis of antipsychotic drug action at D1 versus D2 receptors, new insights have been suggested by several recent findings. Fundamentally, there is a widely perceived discrepancy between the acute dopamine receptor-blocking activity of neuroleptics and their delayed therapeutic effects; this has been approached by considering secondary effects consequent to primary dopamine receptor blockade (Pickar, 1988), or by questioning the substance of the perceived discrepancy (Keck et al., 1989). In relation to the problems of extrapyramidal side-effects and/or lack of therapeutic efficacy, direct studies of neuroleptic action have been made possible by PET techniques: patients with parkinsonism or akathisia tend to have higher neuroleptic occupancies of D2 receptors (Farde et al., 1989), suggesting the possibility of defining on an individual basis a threshold occupancy for therapeutic efficacy with versus without such side-effects. Patients who show little or no therapeutic response have neuroleptic occupancies of D2 receptors indistinguishable from those of responders (Wolkin et al., 1989), suggesting that non-responders and responders might differ in pathophysiology. Provocatively, the atypical neuroleptic clozapine, which not only appears to induce fewer extrapyramidal side-effects but may be efficacious in some patients unresponsive to typical neuroleptics (Kane et al., 1988), shows the highest occupancy of D1 and lowest occupancy of D2 receptors among all neuroleptics examined so far (Farde et al., 1989). Indeed, in animal studies, clozapine appears to exert preferential attenuation of D1 receptor-mediated function (see Murray & Waddington, 1990; also Chapter 2).

Though it remains important not to overlook the possible role(s) of non-dopaminergic systems in the pathophysiology of schizophrenia and in antipsychotic drug action (Reynolds, 1989; Kerwin, 1989; Carlsson & Carlsson, 1990; Itzhak & Stein, 1990; Waddington & Torrey, 1991), it should be emphasized that clarification of whether selective D1 antagonists do or do not show therapeutic efficacy in this disorder will be a watershed in the evolution of these concepts (Waddington & Daly, 1992). Furthermore, the putative roles of individual, molecular biologically defined members of the ‘D1-like’ (D1a, D1b/D5) and of the ‘D2-like’ (D2long, D2short, D3 and D4) families of dopamine receptor in mediating antipsychotic activity remain enigmatic, but may in the future challenge further our present perspectives.