The enzyme catechol-O-methyl transferase (COMT), identified in the 1950s, is involved in catabolism of monoamines that are influenced by psychotropic medications, including neuroleptics and antidepressants. The COMT gene lies in a chromosomal region of interest for psychosis and bipolar spectrum disorder and a common polymorphism within the gene alters the activity of the enzyme. As a consequence, COMT has been one of the most studied genes for psychosis. On the basis of prior probabilities it would seem surprising if functional variation at COMT did not have some influence either on susceptibility to psychiatric phenotypes, modification of the course of illness or moderation of response to treatment. There is now robust evidence that variation at COMT influences frontal lobe function. However, despite considerable research effort, it has not proved straightforward to demonstrate and characterise a clear relationship between genetic variation at COMT and psychiatric phenotypes. It is of course, possible that COMT will turn out to be an unusually intractable case but it seems more likely that the experiences with this gene will provide a foretaste of the complexity of genotype–phenotype relationships that will be found for psychiatric traits. In this review, we consider the current state of evidence and the implications both for further studies of COMT and more generally for studies of other genes.
The enzyme catechol-O-methyl transferase (COMT), identified in the 1950s,1 is involved in catabolism of monoamines that are influenced by psychotropic medications, including neuroleptics and antidepressants. The COMT gene lies in a chromosomal region of interest for psychosis and mood disorder and a common polymorphism within the gene alters the activity of the enzyme. As a consequence, COMT has been one of the most studied genes for psychosis. In this review we consider the current state of evidence and the implications both for further studies of COMT and more generally for studies of other genes.
Catechol-O-methyl transferase: enzyme and gene
COMT degrades catecholamines including dopamine. Two main COMT protein isoforms are known. In most assayed tissues, a soluble cytoplasmic (S-COMT) isoform predominates.2 In brain, a longer membrane-bound form (MB-COMT) is the major species.3 Although expressed widely, COMT appears to be a minor player in dopamine clearance compared with neuronal synaptic uptake by the dopamine transporter and subsequent monoamine oxidase (MAO) metabolism.4 However, in the prefrontal cortex (PFC) where dopamine transporter expression is low,5 the importance of COMT appears to be greater.6, 7
The structure of the COMT gene, which lies on chromosome 22q11, is shown in Figure 1. A common G>A polymorphism is present that produces a valine-to-methionine (Val/Met) substitution at codons 108 and 158 of S-COMT and MB-COMT, respectively,8 that results in a trimodal distribution of COMT activity in human populations.8, 9, 10 The polymorphism is usually referred to as the Val/Met locus, but is also known by the reference sequence identification code rs4680 (previously rs165688). Terminology varies and can be confusing: the Valine (Val) allele is also referred to as the high activity (H) allele or the G allele. We will refer to it as the Val allele.
A number of putative regulatory elements have been discovered in the COMT gene, which may explain the differential expression of the long and short transcripts in different tissues.3 These include numerous oestrogen response elements11 and oestradiol has been shown to downregulate COMT expression in cell culture.12 A recent report suggests that MB-COMT exists in two forms which may be differentially affected by the Val/Met genotype.13 Thus, it is to be expected that there will be a level of genetic complexity including possible gender-specific effects.
Polymorphism and haplotype frequencies at COMT have been shown to vary substantially across populations.14, 15 For example, the Val allele has been reported at frequencies varying between 0.99 and 0.48.14 Moreover, in certain Asian populations, a second functional variant, Ala72Ser, (MB COMT nomenclature) has been reported16 Hence, population origin of samples is a potentially important variable for interpreting genetic studies of COMT.
Positional studies of psychiatric phenotypes
The 22q11 chromosomal region in which COMT is located is involved in microdeletions that are present in most individuals with velocardiofacial syndrome (VCFS) and related clinical syndromes which we will collectively refer to as chromosome 22q11 deletion syndrome (22q11DS). In addition to characteristic core features of dysmorphology, abnormalities of the palate and congenital heart disease, cognitive impairments are common, and range in severity from minimal to severe.17 An increase in a broad spectrum of psychiatric disorders in children with 22q11DS has been reported including anxiety, mood disorders, obsessive-compulsive disorder (OCD), and attention deficit disorder.17, 18, 19, 20, 21 In adults with 22q11DS, high rates of psychosis have been reported,20, 22, 23, 24 with the majority of cases satisfying diagnostic criteria for schizophrenia25 and an estimated risk of around 25%.
Linkage studies have provided evidence for one or more loci in the 22q11 region in which COMT is located that influence susceptibility to several psychiatric phenotypes.
The region is supported by one of the two published meta-analyses of bipolar disorder.26
The only linkage study to use families selected on the basis of having at least one member with schizoaffective disorder, bipolar type provided genome-side suggestive evidence for linkage at 22q11.28
Many fewer linkage studies have been undertaken for other psychiatric phenotypes so there have not been the same opportunities for this chromosomal region to be implicated for these phenotypes. The region has not been implicated in genome scans of unipolar depression.29 Lod scores close to 3 have been reported at the 22q11 locus close to COMT in a study of 70 panic disorder pedigrees.30 This is of particular interest given that evidence has been reported to support a genetic contribution to the co-occurrence panic disorder in some bipolar disorder sufferers.31
In summary, on the basis of position, COMT must be considered as a strong candidate for involvement in psychiatric phenotypes, particularly psychosis and bipolar mood disorder.
Association studies of COMT and frontal lobe function
Several studies suggest that the COMT Val/Met locus influences performance on tests of frontal lobe function, with the Val allele and/or Val/Val genotype being associated with poorer performance. The Val allele was initially associated with poorer function as indexed by the Wisconsin Card Sorting Test and fMRI32 in patients with schizophrenia and controls. Subsequently, a fairly strong body of evidence has been reported for association between the Val allele and poorer performance in controls,33, 34, 35 patients with schizophrenia,34, 36, 37 the sibs of schizophrenics (although not the schizophrenics themselves),38 and a small sample of subjects with 22q11DS.39 One study of individuals with schizophrenia demonstrated a biphasic effect, with the Val allele being associated with poorer performance in some tests, but better performance in others.37 However, not all studies,40, 41 including the largest study comprising 543 Greek army conscripts,42 have supported association between frontal cognitive measures and COMT. Recently, the low activity Met allele has been reported as a risk factor for cognitive decline in 22q11DS.43
Most work to date has focussed on the Val/Met polymorphism and other variants reported to be associated with altered mRNA expression44, 45 have not been widely studied. However, on the current evidence, the mechanism for the cognitive effects at COMT is unlikely to be simple. Weinberger and colleagues have suggested that it is likely that the relationship between COMT activity and PFC function is more complex than simply ‘Met158 good, Val158 bad’46 and have argued in support of an inverted U-shaped relationship between dopamine levels and PFC function. The argument is that the precise effect of COMT activity on PFC function is likely to be dependent on where on the inverted-U curve the individual in question lies in any given environmental or genetic context. This is likely governed by multiple factors, including the nature of the measure being examined,37 state factors, for example the relative amount of stress that the individual is under, which is known to affect PFC dopamine levels47 and trait factors, such as the complex genetic background on which the COMT genotype is expressed, Under this model, 22q11DS can be considered as an example of the effect of genetic background, with the high activity Val allele being associated, in some studies, with better cognition48 and less decline43 because the higher Val activity rescues people with 22q11DS from the consequences of having only one gene copy.
In summary, there is strong evidence for an effect of COMT on cognitive function. Given that prefrontal cognitive function has been proposed as a trait marker for schizophrenia32 this offers the potential for explanatory mechanisms of how variation in COMT may influence abnormal brain function in psychiatric phenotypes.
Association studies of COMT in psychiatric phenotypes
Perhaps not surprisingly most psychiatric association studies of COMT to date have involved only the functional Val/Met polymorphism and have focussed mainly on the phenotypes of schizophrenia and, to a lesser extent, bipolar disorder. However, some studies have been undertaken in a range of other phenotypes. Further, there is a recent trend towards study of other polymorphisms across this locus. We will consider the evidence for the better studied phenotypes below.
Schizophrenia – Val/Met
The COMT Val/Met variant has been one of the most studied candidate polymorphisms for schizophrenia. The vast majority of case–control studies have failed to find evidence for association. In a meta-analysis of studies predating August 2002,49 only two50, 51 of 14 case–control studies yielded significant evidence for association (the Met allele in each case). However, when all studies were combined (total: 2205 cases; 2236 controls), the odds ratio (OR) for the Met allele was 1, indicative of no effect. Separate analysis of case–control samples by Asian or European origin also failed to provide evidence for association. Of five family studies included in the meta-analysis, two reported significant evidence for association, this time with the Val allele. In the meta-analysis,49 the authors concluded that overall there was some support for association of schizophrenia with the Val allele in European samples.
An updated meta-analysis of case–control literature published prior to December 200352 including eight Asian studies (2125 patients, 2504 controls) and 11 European studies (1350 patients, 1573 controls) found no significant effect but a trend for over-representation of the Val allele in cases (OR=1.09, confidence interval (CI) 0.94–1.26). A further meta-analysis using the December 2003 data but using a regression approach found significant evidence for association of the Val allele with schizophrenia if all studies were included but loss of significance if studies were excluded where the control sample showed departure from Hardy–Weinberg equilibrium (suggesting the possibility of genotyping error or population stratification).53 Studies published subsequently include our own study of two large association samples from the UK (709 cases, 710 controls) and Bulgaria (488 parent–proband trios),54 which found no support, and data from a Korean study of around 300 cases and 300 controls which was similarly negative.16 A recent study of Turkish cases (n=297) and controls (n=341) suggested the Met/Met genotype as a risk factor for schizophrenia55 but caution is required because controls departed very substantially from Hardy–Weinberg expectations.
In summary, despite the investigation of many thousands of schizophrenia cases and controls no consistent, significant evidence for association at the Val/Met locus has emerged. Meta-analyses have provided (nonsignificant) estimates of an over-representation of the Val allele with an effect size of approximately 1.1.
A number of groups have sought evidence for susceptibility variants elsewhere in the gene. Some of the strongest positive evidence was reported in a large study of Ashkenazi Jews56 comprising around 700 cases and approximately 3000 controls. Modest evidence was found for association between the Val allele and schizophrenia (P=0.024) but two other polymorphisms, rs737865 in intron 1 and rs165599 within the 3′UTR of some COMT mRNA species44 (see Figure 1), were more strongly associated. The haplotype carrying the G allele at all three loci (which at Val/Met encodes Val) was also strongly associated (P=9.5 × 10−8). Interestingly, the three other haplotypes carrying the Val allele were under-represented in cases. Gender effects were also reported although these are difficult to interpret because much of the effect was driven by differences in the controls, not the cases.
Several studies have subsequently examined a range of markers including those required to define the ‘Shifman haplotype’ (none were included in the meta-analyses discussed above). One based on 267 Irish multiplex families57 revealed modest evidence for excess transmission of the Val allele using a broad case definition including schizophrenia and mood-psychosis spectrum phenotypes (P=0.01). Haplotype analysis provided marginally stronger evidence than Val/Met alone. Only one of four relatively common haplotypes carrying the valine allele was significantly overtransmitted (A-G-A) to schizophrenics, while one, the G-G-G risk haplotype of Shifman, was significantly under-transmitted indicating a protective effect.
Sanders et al.58 examined eight markers spanning COMT and extending into the neighbouring gene, armadillo repeat gene deleted in velocardiofacial syndrome (ARVCF), in 136 families of mainly European American origin. Several haplotypes including markers reaching into ARVCF yielded significant evidence for association (best nominal global P∼0.002). The individual specific haplotype displaying the excess transmission was almost fully characterised by G-A at Val/Met-rs165599 and is, therefore, consistent with the associated A-G-A rs737865-Val/Met-rs165599 haplotype of Chen et al.57
In a sample of 50 white Australian affected sib-pairs evidence (permuted global P∼0.002) was obtained using haplotypes of the three Shifman markers.59 Much of the association appears to have been driven by under-transmission of the A-G-A (Val containing) haplotype (transmitted=1, nontransmitted=18). A number of gender-specific findings were observed but caution is required given the small sample sizes, and issues of multiple testing (although methods for accommodating this as well as controlling for multiple testing were still positive according to our communications with the authors).
Finally, in the largest single study of COMT including almost 1200 cases representing a case (N=709) control (N=710) sample from the UK and a family-based association sample of complete trios (N=488) from Bulgaria, our own group found no evidence for association in either sample to the Val/Met locus or to any of the Shifman markers or haplotypes.54 Analysis by gender also failed to identify any evidence for association to markers or haplotypes.
In summary, our own negative study nothwithstanding, the recent data based upon additional polymorphisms and haplotypes provide some encouragement that variation at COMT influences the schizophrenia phenotype. However, robust replication is required preferably using genewide tests of significance.60 Further, the data suggest a mechanism that involves substantial complexity over and above any effect of the Val/Met polymorphism itself and could involve variation in neighbouring ARVCF.
Studies of COMT in bipolar disorder
Although less studied than in schizophrenia, the Val/Met polymorphism has received substantial investigation in bipolar disorder and, like schizophrenia, the findings are inconclusive or negative. Meta-analysis of the seven case–control studies in the literature in 2001 (910 bipolar cases, 1069 controls) provided borderline significant evidence for association of the Met allele with susceptibility to bipolar disorder (OR=1.18, CI 1.02–1.35).61 However, an updated meta-analysis undertaken as part of the current review (2169 bipolar cases, 7804 controls; Figure 2) provides no evidence to support association, although a nonsignificant trend towards association with the Met allele remains (OR=1.08, CI 0.94–1.24; P=0.30). No evidence for association at COMT Val/Met has emerged from family-based association studies62, 63, 64 but only a few hundred families have been reported so this observation must be interpreted within the context of very limited power.
Analyses of bipolar samples have been undertaken with the Val/Met polymorphism to seek evidence for association with clinical subphenotypes. Positive reports include significant over-representation of the Met allele in rapid cycling.65, 66, 67 Although interesting, these reports must still be considered preliminary because the findings have not yet been subjected to widespread attempts at replication.
As for schizophrenia, some groups have sought evidence for susceptibility variants elsewhere in the gene. Little has been published and no compelling positive evidence has yet emerged but the modestly positive report by Shifman et al.68 is of interest and suggests the possibility that the same risk haplotype that confers risk of schizophrenia in the Ashkenazi population may also confer risk of bipolar disorder.
Studies of COMT in other phenotypes
Catechol-O-methyl transferase has received limited study in a range of other psychiatric phenotypes. Typically, studies are few in number, samples are small and only the Val/Met polymorphism has been considered. Many reports involved substantial multiple testing with examination of a variety of phenotypic definitions or traits, as well analyses in the total sample and the sample divided according to gender. These issues make it difficult to draw any firm conclusions. Within the scope of this review, we can only illustrate the current state of the literature by giving examples for some of the better studied phenotypes.
The group of Karayiorgou aroused great interest when they reported association of susceptibility to OCD with the COMT low activity Met allele in 73 cases and 148 controls69 with maximum effect in males. However, a meta-analysis of seven studies published up to 2003 (total: 144 cases; 337 controls; 269 trios) concluded that there was insufficient evidence to support association, with or without a gender effect.70 Two subsequent studies (total: 158 cases; 373 controls) also failed to provide significant evidence for association71, 72 although the second of these found a nonsignificant trend for excess Met allele in male OCD cases and the same effect in a sample of 113 schizophrenia patients with comorbid OCD.72
No consistent findings have emerged from the few studies of unipolar samples (most of 100 or less cases) that have been published. A recent study found evidence for association of the Val allele with early onset depression in a sample of 120 patients with onset below 25 years of age but not in the larger set of 378 cases unselected for age at onset.73 The other positive finding reported was association of the Met allele with unipolar depression in a small Japanese sample (75 cases, 135 controls).74
Attention deficit hyperactivity disorder
Borderline significant excess of the Met/Met genotype was reported for the phenotypes of panic disorder, trait anxiety and treatment response in a sample of 178 panic disorder cases and 182 controls.79 In contrast, excess Val allele was reported in 115 panic cases compared with matched controls,80 an effect present in females but not males. Interestingly, comorbid panic disorder has been suggested as a marker of a genetically homogeneous subphenotype of bipolar disorder31 so a possible effect of the COMT locus on both panic disorder and bipolar disorder could mediate this relationship.81 However, inconsistent with this attractive possibility, in a study of 111 bipolar disorder cases Rotondo et al.81 found evidence for association of the Met allele with bipolar disorder only in cases without comorbid panic disorder.
Other phenotypes studied include phobic anxiety (sample: 1234 females; Val allele associated),82 persistent anxiety (sample: 962 individuals; Met allele and Met/Met genotype associated in females),83 nicotine dependence (positive and negative studies reported)84 and anorexia nervosa (association reported in 66 restricting anorexics but not 19 with bingeing/purging subtype).85
To summarise, in general reports lack consistency and no clear pattern has emerged. For each phenotype domain (such as anxiety spectrum disorders) a formal meta-analysis is required that includes all available data (published and unpublished) and takes account of the relationships between the phenotypes examined and the level of multiple testing. Almost certainly, substantially larger and more systematic studies will be required before the true state will emerge.
It is expected, essentially by definition, that the pathogenesis of genetically complex disorders, including psychiatric phenotypes, will involve interaction between multiple genes and environmental factors. 86 If possible it is highly desirable to include interacting environmental variables within analyses because, under ideal circumstances, this optimises the power to identify both the genetic and environmental factors. However, in general this proves difficult because of lack of knowledge and data on environmental variables of relevance. Further, inclusion of irrelevant variables actually reduces power as a result of the requirement for correction for multiple testing and the introduction of random variation.87
Within this context, two recent reports support the possibility of gene–environment interactions involving the COMT Val/Met polymorphism. First, in a longitudinal study of the Dunedin birth cohort followed to adulthood carriers of the COMT Val allele were most likely to exhibit psychotic symptoms and to develop schizophreniform disorder if they used cannabis, whereas cannabis use had no such adverse influence on individuals with two copies of the Met allele.88 This observation is consistent with COMT modulating the psychotogenic effects of cannabis exposure. Second, in a family-based association study of 240 ADHD sufferers, bearers of the Val/Val genotype were substantially more susceptible to the adverse effects of prenatal risk (indexed by low birth weight) in influencing risk of early-onset antisocial behaviour.89 This observation is consistent with modulating effects of COMT and prenatal factors on the known relationships between prefrontal cortical dysfunction and antisocial behaviour.
These findings require replication but suggest the possibility that gene–environment interactions may contribute to the complexity genotype–phenotype relationships at COMT.
Interpretation of data for psychiatric phenotypes
It will be readily apparent to the reader that few unambiguous or simple findings have emerged from studies to date of variation at COMT for psychiatric phenotypes (Table 1). It is not appropriate to consider most phenotypes in detail because of the limited work undertaken. Indeed, for these phenotypes the requirement is for more studies, larger samples and systematic coverage of the gene (Table 1). However, it is useful to consider how to interpret the current body of work available for psychosis and bipolar mood disorder.
This is by far the best studied phenotype. The default hypothesis to consider must be the null hypothesis. The meta-analyses,49, 52, 53 and two of the three largest single studies52, 54 do not allow this to be rejected. However, in striking contrast, several recent studies employing multiple markers do suggest association that, at least on the face of it, appears to be at a level beyond that expected by chance alone. These data therefore suggest it is still fairly plausible, although not proven, that a susceptibility locus exists at or around COMT. However, there are several caveats. First, haplotype analysis is highly sensitive to genotyping errors and can result in false positive and negative associations. Second, haplotype analyses provide opportunities for extensive multiple testing and, in the absence of a clear and systematic analytic strategy specified a priori it can be difficult to interpret the true level of statistical significance. Third, in such indirect association analyses the aim is to detect polymorphisms or haplotypes in linkage disequilibrium with pathogenically relevant variants (rather than the variants themselves). Under this approach it is to be expected that the pattern of findings will vary between studies because of differences in population structure. However, this requires a genewide approach to analysis in order to minimise both type I and type II errors.
Most work on COMT has been predicated on the hypothesis that the Val/Met polymorphism is a direct risk factor for schizophrenia. The classic hypothesis that schizophrenia results from enhanced dopaminergic neurotransmission90, 91 predicts that the Met allele will be directly associated, a prediction also suggested by the observation that deletion of COMT is associated with the high rate of psychosis in 22q11DS. In contrast, the hypothesis that excess dopamine function in the mesolimbic system is secondary to low dopamine function in the prefrontal cortex (e.g.92, 93) predicts association to the Val allele as do reported associations between the Val allele and poor prefrontal function and between poor prefrontal function and schizophrenia.32
On the face of it, these competing hypotheses should be unambiguously resolved by analysis of Val/Met which should be straightforward because direct association analysis does not depend on linkage disequilibrium (LD) and therefore the results should be robust to the variable LD structure at COMT.15 The findings should also be relatively robust to ethnic variation because in populations showing any effect, the same allele should be associated, although the effect size may vary.
However, the overall evidence from published studies is not compatible with either simple hypothesis. In addition to the overall balance of the single locus data, it is notable that in all of the studies in which an effect at Val/Met has been detected, where additional markers have been typed, it has been possible to subdivide haplotypes carrying the risk allele into risk, neutral, and even in some cases, protective haplotypes. As four of these have been family-based studies, this cannot be attributed to stratification. Moreover, in the only study59 to present a formal analysis, the model including additional markers was significantly more significant than that restricted to the Val/Met locus alone, whereas the conditional analysis suggested two independent effects. These findings suggest that if function at the Val/Met locus is at all relevant to schizophrenia, the relative functional properties of the Val and Met alleles can be modified, even reversed, by at least one other relatively common cis-acting variant with an influence perhaps on COMT expression or splicing. This suggestion is broadly compatible with the demonstration of cis-acting loci that modify the expression of COMT mRNA independent of the Val/Met locus.44, 45 However, there are also data that argue strongly against the existence of more than one even moderately common functionally relevant polymorphism in COMT. Thus, while Chen et al.10 demonstrated that brain COMT enzyme activity is associated with the Val/Met locus, this did not correlate with the markers that we and others have associated with mRNA expression44, 45 or with schizophrenia. This finding is based on subjects of European and African ethnic origins10 although in certain Asian populations, a second functional Ala72Set variant has been reported.16 As enzyme activity can be reasonably assumed to be functionally more important than mRNA abundance, the findings of Chen et al.10 provides a strong refutation to our hypothesis44 that the associations with the Shifman haplotypes are likely to be attributable to major or minor alterations in COMT expression. Unfortunately, it is difficult to obtain an exact estimate of the proportion of the variance in brain COMT activity that can be attributed to the Val/Met locus from that study given confounders like post-mortem variance, age, and measurement error. However, earlier studies with respect to peripheral COMT activity suggest that all, or almost all, COMT activity can be attributed to Val/Met.8, 94
If it is correct that Val/Met is responsible for all almost all variance in COMT function, then if they indicate true association, the haplotype data must point to the involvement of another gene in LD with COMT. A strong candidate here is ARVCF58, 95 which shares common exonic sequence with COMT but on opposite strands. Moreover, one of the Shifman markers (rs165599) is located in the common exonic sequence and appears to influence ARVCF expression (Bray et al.44 and unpublished data). That evidence for association to schizophrenia at a locus as a priori highly plausible as COMT might be attributable to association to an adjacent gene might seem an unlikely turn of events. It is, however, more parsimonious than a model which specifies that the Val/Met contributes to risk, but a second polymorphism in an adjacent gene dictates whether the Val or the Met allele is associated and does so without influencing COMT activity, as is implied if the Val/Met is the only locus with an important effect on COMT activity. However, if as may be the case for some tests of executive functioning, the risk of schizophrenia shows an inverted U-shaped dose–response with dopamine activity, one can envisage circumstances where such a complex model might apply and therefore this latter model should not be entirely discounted.
Taken as a whole, the current evidence is suggestive of some involvement of variation at or near COMT in predisposition to schizophrenia but, despite greater study, the level of evidence is less convincing than that available for several of the other genes of current interest, including dysbindin, neuregulin 1, DISC1 and DAOA(G72).96, 97
Substantially less work has been reported for bipolar disorder than schizophrenia. However, as for schizophrenia, it is clear that there is no simple single Val/Met effect. The nonsignificant trend from meta-analysis is for over-representation of the Met allele in bipolar cases compared with controls. This trend is in the opposite direction to that for schizophrenia and of similar magnitude (OR∼1.1). Clearly this is consistent with the null hypothesis of no effect. It is, however, worth considering the possibility that Val/Met operates as a modifying locus for psychosis rather than a susceptibility locus, the Met allele increasing expression of positive and affective symptoms and the Val allele increasing expression of negative symptoms and cognitive dysfunction. This would be consistent with the mounting evidence linking the Val/Met variant with variation in cognitive and behavioural function (see above), but also some emerging but inconclusive findings suggesting it may influence response to medication (e.g.98, 99).
Insufficient work has been reported in bipolar disorder for variants other than Val/Met to allow useful comment. Large systematic studies are needed.
It has been conventional in studies of psychosis to recruit and investigate samples meeting criteria for either schizophrenia or bipolar disorder. Cases at the interface between these prototypical categories are common in clinical practice but have not been a focus of study. Few samples of Diagnostic and Statistical Manual of Mental Disorders (4th Edition) DSMIV Schizoaffective disorder are available although existing samples of DSMIV Bipolar disorder and schizophrenia include a diverse range of clinical picture and a variable proportion of cases that have a substantial mix of the features of bipolar and schizophrenia prototypes. Several pieces of evidence suggest that the COMT locus, rather than conferring some general susceptibility to schizophrenia and bipolar disorder, may specifically influence susceptibility to an intermediate form of mood-psychosis phenotype:
Genomewide suggestive linkage at 22q11 was found in a linkage study of families selected through a member with DSMIV schizoaffective disorder, bipolar type28 (but not in the larger samples of schizophrenia and bipolar disorder families from which these were drawn).
Linkage at 22q11 has been demonstrated within a subset of bipolar pedigrees in which individuals also experienced psychotic features.100
The schizophrenia association finding of Chen et al.57 at COMT was maximal when a broad spectrum of phenotype was used (including also mood-psychosis spectrum cases).
Funke et al.101 found evidence for association at COMT with individuals meeting criteria for schizoaffective disorder as well as those meeting criteria for schizophrenia and mood disorders.
In our own study of over 1575 mood-psychosis cases and 2309 controls we found no evidence for association at COMT for the categories of bipolar disorder or schizophrenia. However, in the set of 308 cases who had experienced manic episodes and prominent psychotic features in at least half of all episodes of illness, we identified significant evidence for association (unpublished, presented at World Congress of Psychiatric Genetics, Boston102).
These observations raise the possibility that clinical heterogeneity between samples may be an important, perhaps major, cause for the observed inconsistencies in findings between studies.
General issues: lessons of relevance to psychiatric genetics
Functional and positional evidence make COMT a strong candidate for involvement in multiple psychiatric phenotypes (particularly, but not restricted to, schizophrenia and bipolar disorder). There is a common functional variant that exerts substantial influence over enzyme activity. On the grounds of prior probability, it could be considered an almost ideal candidate for being able to demonstrate or refute genetic association.103 And yet, although it has been one of the most studied genes to date, no clear, unambiguous pattern of results has emerged. The experiences with COMT illustrate some general lessons that we can take forward for study of other genes (Table 2):
(1) Adequate sample sizes
Samples need to be adequately powered for plausible effect sizes.104Sample numbers need to be closer to 1000 than to 100. It is known that reliable replication of an effect typically takes samples several times larger than that in which the original observation was made.105 The true state of affairs for each phenotype is unlikely to be clear until several thousand samples have been studied. An example is that of the Pro12Ala polymorphism at the gene encoding peroxisome proliferative activated receptor, gamma (PPARG) in type 2 diabetes where the first attempts at replication failed but when several thousand cases had been studied a robust (but modest) effect was established.106
(2) Importance of meta-analysis
Given the large samples required, meta-analysis is a crucial technique for summarising available data. Access to all relevant findings including negative data is required. All publications should, therefore, include the basic allele and genotype counts for each of the studied groups. Unfortunately, some reports have failed to include these because they use, for example, regression methods (e.g.73) and this diminishes their utility to the scientific community.
(3) Gene-based approaches
The study of the COMT Val/Met polymorphism in isolation probably continues to be justified because of its clear and major functional significance. However, it is clear that if COMT variation is important in psychiatric phenotypes, variants elsewhere within or even outwith the COMT gene play an important, and possibly larger, role. Thus, the research focus should shift to study of the gene or locus as the unit, rather than a single polymorphism. Resources and methods are available that allow systematic coverage of a high proportion of the genetic variation at a locus and this gene-based approach offers many advantages including protection against genetic heterogeneity (i.e. different variants having maximal effects in different populations).60
(4) Closer attention to the clinical phenotype
Growing evidence suggests that genetic susceptibility will not respect current operational diagnostic boundaries.96, 97, 107, 108Most samples are defined according to DSMIV diagnostic categories. There is an implicit assumption that, for example, one schizophrenia sample should be pretty similar to another. However, the schizophrenia phenotype is so broad that it is possible for one sample to be composed of chronically disabled individuals with cognitive impairment, marked negative features and minimal affective or positive psychotic symptoms whereas another sample could include relatively well functioning individuals with an episodic course and marked affective and positive psychotic symptoms. At the very least, it is important that substantial additional clinical detail is provided about samples (perhaps as on-line supplements) in order to allow meaningful comparisons between studies.
(5) An iterative approach
It is entirely possible that a relatively large effect size within a subset of the data is hidden amongst an overly broad traditional diagnostic group. It is, therefore, important that when a signal is identified researchers explore their data to refine the associated phenotype (clearly in the first experiment the exploration must be treated with caution but it can be tested explicitly in attempts at replication). This type of iterative approach (Figure 3) is important for optimising power to replicate genetic findings and has a wider importance in terms of moving to a more valid psychiatric nosology.97
(6) Biology is complementary, not superior, to genetics
High impact journals typically require ‘biological support’ for genetic association findings, presumably under the presumption that this provides enhanced credibility. However, the findings with COMT demonstrate that compelling biological plausibility does not ensure that genetic findings can be easily replicated or validated. Given the limited understanding of the pathophysiology of psychiatric disorders and the complexity of brain biology it is important that an appeal to biological plausibility is not allowed to exaggerate the significance of weak genetic data. Further, it is important not to lose sight of the fact that a major attraction of genetic studies of psychiatric phenotypes is the potential to discover entirely novel mechanisms of pathogenesis. To this end, it should be hoped that there will be entirely robust genetic findings that cannot be readily explained at a functional level. In terms of eventual impact on patient care, such observations are likely to prove to be the most important findings that will emerge from psychiatric genetics.
(7) Closer attention to data quality
The most recent meta-analysis of the COMT Val/Met polymorphism in schizophrenia53 demonstrates that conclusions can be influenced substantially by consideration of the quality of the data included. Genotyping error can be a cause of both type I and type II errors109, 110, 111, 112, 113, 114, 115 in case control and family-based association studies. Haplotype analyses are particularly susceptible to genotyping errors116 because errors can create spurious evidence for rare haplotypes. Any systematic difference in error rate between cases and controls (as could occur as a result of differences in DNA quality or purity) has the potential to generate spurious case–control differences on haplotype analysis. It is obviously important that this is recognised and that error rates are minimised. However, even in the best laboratories error rates are not negligible. Estimated error rates should always be reported and their possible impact on findings considered. Particular caution is needed for findings where significance comes entirely from very rare haplotypes.
(8) The detection of modifier loci
The inconsistent data relating genetic variation in COMT to specific phenotypes might reflect the fact that it is a modifier rather than a risk locus. This gains plausibility from the relatively more convincing data relating COMT to cognitive function. Few if any association studies are undertaken with samples of incident cases and differing ascertainment procedures for prevalent cases is likely to lead to differences in factors such as severity, chronicity, and treatment that may be influenced by genes that do not increase risk of the disorder. The identification of modifier loci could have important implications for understanding pathogenesis and the design of novel therapies. However, it is perhaps important that researchers are able to distinguish modifier from risk effects relatively early on to avoid the plethora of inconsistent association studies with prevalent cases.
As a consequence of both its chromosomal location in a region of interest for psychosis and mood disorders and its function as an enzyme involved in catabolism of monoamines, COMT has been one of the most studied genes for psychosis. On the basis of prior probabilities it would seem surprising if variation at COMT did not have some influence either on susceptibility to psychiatric phenotypes, modification of the course of illness or moderation of response to treatment. There is now robust evidence that variation at COMT influences frontal lobe function. However, despite considerable research effort, it has not proved straightforward to demonstrate and characterise a clear relationship between genetic variation at COMT and psychiatric phenotypes. It is of course, possible that COMT will turn out to be a particularly challenging case but it seems more likely that the experiences with this gene will provide a foretaste of the complexity of genotype–phenotype relationships that will be found for psychiatric traits.
Axelrod J, Tomchick R . Enzymatic O-methylation of epinephrine and other catechols. J Biol Chem 1958; 233: 702–705.
Jeffery DR, Roth JA . Characterization of membrane-bound and soluble catechol-O-methyltransferase from human frontal cortex. J Neurochem 1984; 42: 826–832.
Tenhunen J, Salminen M, Lundstrom K, Kiviluoto T, Savolainen R, Ulmanen I . Genomic organization of the human catechol-O-methyltransferase gene and its expression from two distinct promoters. Eur J Biochem 1994; 223: 1049–1059.
Huotari M, Santha M, Lucas LR, Karayiorgou M, Gogos JA, Mannisto PT et al. Effect of dopamine uptake inhibition on brain catecholamine levels and locomotion in catechol-O-methyltransferase-disrupted mice. J Pharmacol Exp Ther 2002; 303: 1309–1316.
Sesack SR, Hawrylak VA, Matus C, Guido MA, Levey AI . Dopamine axon varicosities in the prelimbic division of the rat prefrontal cortex exhibit sparse immunoreactivity for the dopamine transporter. J Neurosci 1998; 18: 2697–2708.
Gogos JA, Morgan M, Luine V, Santha M, Ogawa S, Pfaff D et al. Catechol-O-methyltransferase-deficient mice exhibit sexually dimorphic changes in catecholamine levels and behavior. Proc Natl Acad Sci USA 1998; 95: 9991–9996.
Tunbridge EM, Bannerman DM, Sharp T, Harrison PJ . Catechol-O-methyltransferase inhibition improves set-shifting performance and elevates stimulated dopamine release in the rat prefrontal cortex. J Neurosci 2004; 24: 5331–5335.
Lachman HM, Papolos DF, Saito T, Yu YM, Szumlanski CL, Weinshilboum RM . Human catechol-O-methyltransferase pharmacogenetics: description of a functional polymorphism and its potential application to neuropsychiatric disorders. Pharmacogenetics 1996; 6: 243–250.
Floderus Y, Wetterberg L . The inheritance of human erythrocyte catechol-O-methyltransferase activity. Clin Genet 1981; 19: 392–395.
Chen J, Lipska BK, Halim N, Ma QD, Matsumoto M, Melhem S et al. Functional analysis of genetic variation in catechol-O-methyltransferase (COMT): effects on mRNA, protein, and enzyme activity in postmortem human brain. Am J Hum Genet 2004; 75: 807–821.
Xie T, Ho SL, Ramsden D . Characterization and implications of estrogenic down-regulation of human catechol-O-methyltransferase gene transcription. Mol Pharmacol 1999; 56: 31–38.
Jiang H, Xie T, Ramsden DB, Ho SL . Human catechol-O-methyltransferase down-regulation by estradiol. Neuropharmacology 2003; 45: 1011–1018.
Tunbridge EM, Weinberger DR, Harrison PJ . A novel protein isoform of catechol-O-methyltransferase (COMT): brain expression analysis in schizophrenia and bipolar disorder and effect of Val(158)Met genotype. Mol Psychiatry 2006; 11: 116–117.
Palmatier MA, Kang AM, Kidd KK . Global variation in the frequencies of functionally different catechol-O-methyltransferase alleles. Biol Psychiatry 1999; 46: 557–567.
Palmatier MA, Pakstis AJ, Speed W, Paschou P, Goldman D, Odunsi A et al. COMT haplotypes suggest P2 promoter region relevance for schizophrenia. Mol Psychiatry 2004; 9: 1359–4184.
Lee SG, Joo Y, Kim B, Chung S, Kim HL, Lee I et al. Association of Ala72Ser polymorphism with COMT enzyme activity and the risk of schizophrenia in Koreans. Hum Genet 2005; 116: 319–328.
Swillen A, Devriendt K, Legius E, Eyskens B, Dumoulin M, Gewillig M et al. Intelligence and psychosocial adjustment in velocardiofacial syndrome: a study of 37 children and adolescents with VCFS. J Med Genet 1997; 34: 453–458.
Gerdes M, Solot C, Wang PP, Moss E, LaRossa D, Randall P et al. Cognitive and behavior profile of preschool children with chromosome 22q11.2 deletion. Am J Med Genet 1999; 85: 127–133.
Golding-Kushner KJ, Weller G, Shprintzen RJ . Velo-cardio-facial syndrome: language and psychological profiles. J Craniofac Genet Dev Biol 1985; 5: 259–266.
Papolos DF, Faedda GL, Veit S, Goldberg R, Morrow B, Kucherlapati R et al. Bipolar spectrum disorders in patients diagnosed with velo-cardio-facial syndrome: does a hemizygous deletion of chromosome 22q11 result in bipolar affective disorder? Am J Psychiatry 1996; 153: 1541–1547.
Feinstein C, Eliez S, Blasey C, Reiss AL . Psychiatric disorders and behavioral problems in children with velocardiofacial syndrome: usefulness as phenotypic indicators of schizophrenia risk. Biol Psychiatry 2002; 51: 312–318.
Shprintzen RJ, Goldberg R, Golding-Kushner KJ, Marion RW . Late-onset psychosis in the velo-cardio-facial syndrome. Am J Med Genet 1992; 42: 141–142.
Pulver AE, Nestadt G, Goldberg R, Shprintzen RJ, Lamacz M, Wolyniec PS et al. Psychotic illness in patients diagnosed with velo-cardio-facial syndrome and their relatives. J Nerv Ment Dis 1994; 182: 476–478.
Bassett AS, Hodgkinson K, Chow EW, Correia S, Scutt LE, Weksberg R . 22q11 deletion syndrome in adults with schizophrenia. Am J Med Genet 1998; 81: 328–337.
Murphy KC, Jones LA, Owen MJ . High rates of schizophrenia in adults with velo-cardio-facial syndrome. Arch Gen Psychiatry 1999; 56: 940–945.
Badner JA, Gershon ES . Meta-analysis of whole-genome linkage scans of bipolar disorder and schizophrenia. Mol Psychiatry 2002; 7: 405–411.
Lewis CM, Levinson DF, Wise LH, DeLisi LE, Straub RE, Hovatta I et al. Genome scan meta-analysis of schizophrenia and bipolar disorder, part II: Schizophrenia. Am J Hum Genet 2003; 73: 34–48.
Hamshere ML, Bennett P, Williams N, Segurado R, Cardno A, Norton N et al. Genome-wide linkage scan in schizoaffective disorder: significant evidence for linkage (LOD=3.54) at 1q42 close to DISC1, and suggestive evidence at 22q11 and 19q13. Archives of General Psychiatry 2005; 62: 1081–1088.
Craddock N, Forty E . Genetics of mood disorders. Eur J Hum Genet, in press.
Hamilton SP, Slager SL, Heiman GA, Deng Z, Haghighi F, Klein DF et al. Evidence for a susceptibility locus for panic disorder near the catechol-O-methyltransferase gene on chromosome 22. Biol Psychiatry 2002; 51: 591–601.
MacKinnon DF, Zandi PP, Cooper J, Potash JB, Simpson SG, Gershon E et al. Comorbid bipolar disorder and panic disorder in families with a high prevalence of bipolar disorder. Am J Psychiatry 2002; 159: 30–35.
Egan MF, Goldberg TE, Kolachana BS, Callicott JH, Mazzanti CM, Straub RE et al. Effect of Comt Val(108/158) Met genotype on frontal lobe function and risk for schizophrenia. Proc Natl Acad Sci USA 2001; 98: 6917–6922.
Malhotra AK, Kestler LJ, Mazzanti C, Bates JA, Goldberg T, Goldman D et al. A functional polymorphism in the COMT gene and performance on a test of prefrontal cognition. Am J Psychiatry 2002; 59: 652–654.
Goldberg TE, Egan MF, Gscheidle T, Coppola R, Weickert T, Kolachana BS et al. Executive subprocesses in working memory: relationship to catechol-O-methyltransferase Val158Met genotype and schizophrenia. Arch Gen Psychiatry 2003; 60: 889–896.
Diamond A, Briand L, Fossella J, Gehlbach L . Genetic and neurochemical modulation of prefrontal cognitive functions in children. Am J Psychiatry 2004; 161: 125–132.
Bilder RM, Volavka J, Czobor P, Malhotra AK, Kennedy JL, Ni X et al. Neurocognitive correlates of the COMT Val(158)Met polymorphism in chronic schizophrenia. Biol Psychiatry 2002; 52: 701–707.
Nolan KA, Bilder RM, Lachman HM, Volavka J . Catechol-O-methyltransferase Val158Met polymorphism in schizophrenia: differential effects of Val and Met alleles on cognitive stability and flexibility. Am J Psychiatry 2004; 161: 359–361.
Rosa A, Peralta V, Cuesta M, Zarzuela A, Serrano F, Martinez-Larrea A et al. New evidence of association between COMT gene and prefrontal neurocognitive function in healthy individuals from sibling pairs discordant for psychosis. Am J Psychiatry 2004; 161: 1110–1112.
Bearden CE, Jawad AF, Lynch DR, Sokol S, Kanes SJ, McDonald-McGinn DM . Effects of a functional COMT polymorphism on prefrontal cognitive function in patients with 22q11.2 deletion syndrome. Am J Psychiatry 2004; 161: 1700–1702.
Tsai SJ, Yu YW, Chen TJ, Chen JY, Liou YJ, Chen MC et al. Association study of a functional catechol-O-methyltransferase-gene polymorphism and cognitive function in healthy females. Neurosci Lett 2003; 8: 123–126.
Ho BC, Wassink TH, O'leary DS, Sheffield VC, Andreasen NC . Catechol-O-methyl transferase Val(158)Met gene polymorphism in schizophrenia: working memory, frontal lobe MRI morphology and frontal cerebral blood flow. Mol Psychiatry 2005; 10: 287–298.
Stefanis NC, Van Os J, Avramopoulos D, Smyrnis N, Evdokimidis I, Hantoumi I et al. Variation in catechol-O-methyltransferase val158 met genotype associated with schizotypy but not cognition: a population study in 543 young men. Biol Psychiatry 2004; 56: 510–515.
Gothelf D, Eliez S, Thompson T, Hinard C, Penniman L, Feinstein C et al. COMT genotype predicts longitudinal cognitive decline and psychosis in 22q11.2 deletion syndrome. Nat Neurosci 2005; 8: 1500–1502.
Bray NJ, Buckland PR, Williams NM, Williams HJ, Norton N, Owen MJ et al. A haplotype implicated in schizophrenia susceptibility is associated with reduced COMT expression in human brain. Am J Hum Genet 2003; 73: 152–161.
Zhu G, Lipsky RH, Xu K, Ali S, Hyde T, Kleinman J et al. Differential expression of human COMT alleles in brain and lymphoblasts detected by RT-coupled 5′ nuclease assay. Psychopharmacology (Berl) 2004; 177: 178–184.
Winterer G, Weinberger DR . Genes, dopamine and cortical signal-to-noise ratio in schizophrenia. Trends Neurosci 2004; 27: 683–690.
Thierry AM, Tassin JP, Blanc G, Glowinski J . Selective activation of mesocortical DA system by stress. Nature 1976; 263: 242–244.
Baker K, Baldeweg T, Sivagnanasundaram S, Scambler P, Skuse D . COMT Val108/158 Met modifies mismatch negativity and cognitive function in 22q11 deletion syndrome. Biol Psychiatry 2005; 58: 23–31.
Glatt SJ, Faraone SV, Tsung MT . Association between a functional catechol-O-methyltransferase gene polymorphism and schizophrenia: meta-analysis of case–control and family-based studies. Am J Psychiatry 2003; 160: 469–476.
Ohmori O, Shinkai T, Kojima H, Terao T, Suzuki T, Mita T et al. Association study of a functional catechol-O-methyltransferase gene polymorphism in Japanese schizophrenics. Neurosci Lett 1998; 243: 109–112.
Kotler M, Barak P, Cohen H, Averbuch IE, Grinshpoon A, Gritsenko I et al. Homicidal behavior in schizophrenia associated with a genetic polymorphism determining low catechol-O-methyltransferase (COMT) activity. Am J Med Genet 1999; 88: 628–633.
Fan JB, Zhang CS, Gu NF, Li XW, Sun WW, Wang HY et al. Catechol-O-methyltransferase gene Val/Met functional polymorphism and risk of schizophrenia: a large-scale association study plus meta-analysis. Biol Psychiatry 2005; 57: 139–144.
Munafo MR, Bowes L, Clark TG, Flint J . Lack of association of the COMT (Val158/108 Met) gene and schizophrenia: a meta-analysis of case–control studies. Mol Psychiatry 2005; 10: 765–770.
Williams HJ, Glaser B, Williams NM, Norton N, Zammit S, MacGregor S et al. No association between schizophrenia and polymorphisms in COMT in two large samples. Am J Psychiatry 2005; 162: 1736–1738.
Sazci A, Ergul E, Kucukali I, Kilic G, Kaya G, Kara I . Catechol-O-methyltransferase gene Val108/158Met polymorphism, and susceptibility to schizophrenia: association is more significant in women. Brain Res Mol Brain Res 2004; 132: 51–56.
Shifman S, Bronstein M, Sternfeld M, Pisanté-Shalom A, Lev-Lehman E, Weizman A et al. A highly significant association between a COMT haplotype and schizophrenia. Am J Hum Genet 2002; 71: 1296–1302.
Chen X, Wang X, O'Neill AF, Walsh D, Kendler KS . Variants in the catechol-O-methyltransferase (COMT) gene are associated with schizophrenia in Irish high-density families. Mol Psychiatry 2004; 9: 962–967.
Sanders AR, Rusu I, Duan J, Molen JE, Hou C, Schwab SG et al. Haplotypic association spanning the 22q11.21 genes COMT and ARVCF with schizophrenia. Mol Psychiatry 2005; 10: 353–365.
Handoko HY, Nyholt DR, Haywood NK, Nertney DA, Hannah DE, Windus LC et al. Separate and interacting effects within the catechol-O-methyltransferase (COMT) are associated with schizophrenia. Mol Psychiatry 2005; 10: 589–597.
Neale BM, Sham PC . The future of association studies: gene-based analysis and replication. Am J Hum Genet 2004; 75: 353–362.
Craddock N, Dave S, Greening J . Association studies of bipolar disorder. Bipolar Disord 2001; 3: 284–298.
Mynett-Johnson LA, Murphy VE, Claffey E, Shields DC, McKeon P . Preliminary evidence of an association between bipolar disorder in females and the catechol-O-methyltransferase gene. Psychiatr Genet 1998; 8: 221–225.
Kirov G, Jones I, McCandless F, Craddock N, Owen MJ . Family-based association studies of bipolar disorder with candidate genes involved in dopamine neurotransmission: DBH, DAT1, COMT, DRD2, DRD3 and DRD5. Mol Psychiatry 1999; 4: 558–565.
Serretti A, Cusin C, Cristina S, Lorenzi C, Lilli R, Lattuada E et al. Multicentre Italian family-based association study on tyrosine hydroxylase, catechol-O-methyl transferase and Wolfram syndrome 1 polymorphisms in mood disorders. Psychiatr Genet 2003; 13: 121–126.
Lachman HM, Morrow B, Shprintzen R, Veit S, Parsia SS, Faedda G et al. Association of codon 108/158 catechol-O-methyltransferase gene polymorphism with the psychiatric manifestations of velo-cardio-facial syndrome. Am J Med Genet 1996; 67: 468–472.
Kirov G, Murphy KC, Arranz MJ, Jones I, McCandles F, Kunugi H et al. Low activity allele of catechol-O-methyltransferase gene associated with rapid cycling bipolar disorder. Mol Psychiatry 1998; 3: 342–345.
Papolos DF, Veit S, Faedda GL, Saito T, Lachman HM . Ultra-ultra rapid cycling bipolar disorder is associated with the low activity catecholamine-O-methyltransferase allele. Mol Psychiatry 1998; 3: 346–349.
Shifman S, Bronstein M, Sternfeld M, Pisante A, Weizman A, Reznik I et al. COMT: a common susceptibility gene in bipolar disorder and schizophrenia. Am J Med Genet B Neuropsychiatr Genet 2004; 128: 61–64.
Karayiorgou M, Altemus M, Galke BL, Goldman D, Murphy DL, Ott J et al. Genotype determining low catechol-O-methyltransferase activity as a risk factor for obsessive-compulsive disorder. Proc Natl Acad Sci USA 1997; 94: 4572–4575.
Azzam A, Mathews CA . Meta-analysis of the association between the catecholamine-O-methyl-transferase gene and obsessive-compulsive disorder. Am J Med Genet B Neuropsychiatr Genet 2003; 123: 64–69.
Meira-Lima I, Shavitt RG, Miguita K, Ikenaga E, Miguel EC, Vallada H . Association analysis of the catechol-O-methyltransferase (COMT), serotonin transporter (5-HTT) and serotonin 2A receptor (5HT2A) gene polymorphisms with obsessive-compulsive disorder. Genes Brain Behav 2004; 3: 75–79.
Poyurovsky M, Michaelovsky E, Frisch A, Knoll G, Amir I, Finkel B et al. COMT Val158Met polymorphism in schizophrenia with obsessive-compulsive disorder: a case–control study. Neurosci Lett 2005; 389: 21–24.
Massat I, Souery D, Del-Favero J, Nothen M, Blackwood D, Muir W et al. Association between COMT (Val158Met) functional polymorphism and early onset in patients with major depressive disorder in a European multicenter genetic association study. Mol Psychiatry 2005; 10: 598–605.
Ohara K, Nagai M, Suzuki Y, Ohara K . Low activity allele of catechol-O-methyltransferase gene and Japanese unipolar depression. Neuroreport 1998; 11: 1305–1308.
Turic D, Williams H, Langley K, Owen M, Thapar A, O'Donovan MC . A family based study of catechol-O-methyltransferase (COMT) and attention deficit hyperactivity disorder (ADHD). Am J Med Genet B Neuropsychiatr Genet 2005; 133: 64–67.
Qian Q, Wang Y, Zhou R, Li J, Wang B, Glatt S et al. Family-based and case–control association studies of catechol-O-methyltransferase in attention deficit hyperactivity disorder suggest genetic sexual dimorphism. Am J Med Genet B Neuropsychiatr Genet 2003; 118: 103–109.
Bellgrove MA, Domschke K, Hawi Z, Kirley A, Mullins C, Robertson IH et al. The methionine allele of the COMT polymorphism impairs prefrontal cognition in children and adolescents with ADHD. Exp Brain Res 2005; 163: 352–360.
Jiang SD, Wu XD, Zhang Y, Tang GM, Qian YP, Wang DX . No association between attention-deficit hyperactivity disorder and catechol-O-methyltransferase gene in Chinese. Yi Chuan Xue Bao 2005; 32: 784–788.
Woo JM, Yoon KS, Choi YH, Oh KS, Lee YS, Yu BH . The association between panic disorder and the L/L genotype of catechol-O-methyltransferase. J Psychiatr Res 2004; 38: 365–370.
Domschke K, Freitag CM, Kuhlenbaumer G, Schirmacher A, Sand P, Nyhuis P et al. Association of the functional V158M catechol-O-methyl-transferase polymorphism with panic disorder in women. Int J Neuropsychopharmacol 2004; 7: 183–188.
Rotondo A, Mazzanti C, Dell'Osso L, Rucci P, Sullivan P, Bouanani S et al. Catechol-O-methyltransferase, serotonin transporter, and tryptophan hydroxylase gene polymorphisms in bipolar disorder patients with and without comorbid panic disorder. Am J Psychiatry 2002; 159: 23–29.
McGrath M, Kawachi I, Ascherio A, Colditz GA, Hunter DJ, De Vivo I . Association between catechol-O-methyltransferase and phobic anxiety. Am J Psychiatry 2004; 161: 1703–1705.
Olsson CA, Anney RJ, Lotfi-Miri M, Byrnes GB, Williamson R, Patton GC . Association between the COMT Val158Met polymorphism and propensity to anxiety in an Australian population-based longitudinal study of adolescent health. Psychiatr Genet 2005; 15: 109–115.
Redden DT, Shields PG, Epstein L, Wileyto EP, Zakharkin SO, Allison DB et al. Catechol-O-methyl-transferase functional polymorphism and nicotine dependence: an evaluation of nonreplicated results. Cancer Epidemiol Biomarkers Prev 2005; 14: 1384–1389.
Michaelovsky E, Frisch A, Leor S, Stein D, Danziger Y, Carel C et al. Haplotype analysis of the COMT-ARVCF gene region in Israeli anorexia nervosa family trios. Am J Med Genet B Neuropsychiatr Genet 2005; 139: 45–50.
Moffitt TE, Caspi A, Rutter M . Strategy for investigating interactions between measured genes and measured environments. Arch Gen Psychiatry 2005; 62: 473–481.
Zammit S, Owen MJ . Stressful life events, 5-HTT genotype, and risk of depression. Br J Psychiatry, in press.
Caspi A, Moffitt TE, Cannon M, McClay J, Murray R, Harrington H et al. Moderation of the effect of adolescent-onset cannabis use on adult psychosis by a functional polymorphism in the catechol-O-methyltransferase gene: longitudinal evidence of a gene X environment interaction. Biol Psychiatry 2005; 57: 1117–1127.
Thapar A, Langley K, Fowler T, Rice F, Turic D, Whittinger N et al. Catechol-O-methyltransferase gene variant and birth weight predict early-onset antisocial behavior in children with attention-deficit/hyperactivity disorder. Arch Gen Psychiatry 2005; 62: 1275–1278.
van Rossum JM . The significance of dopamine receptor blockade for the mechanism of action of neuroleptic drugs. Archives Internationales de Pharmacodynamic et de Therapie 1966; 160: 492–494.
Carlsson A . Mechanism of action of neuroleptic drugs. In: Lipton MA, DiMascio A, Killam KF (eds). Psychopharmacology. A Generation of Progress. Raven Press: New York, 1978, pp 1057–1070.
Daniel DG, Berman KF, Weinberger DR . The effect of apomorphine on regional cerebral blood flow in schizophrenia. J Neuropsychiatry Clin Neurosci 1989; 1: 377–384.
Davis KL, Kahn RS, Ko G, Davidson M . Dopamine in schizophrenia: a review and reconceptualization. Am J Psychiatry 1991; 148: 1474–1486.
Weinshilboum R, Dunnette J . Thermal stability and the biochemical genetics of erythrocyte catechol-O-methyl-transferase and plasma dopamine-beta-hydroxylase. Clin Genet 1981; 19: 426–437.
Li T, Ball D, Zhao J, Murray RM, Liu X, Sham PC et al. Family-based linkage disequilibrium mapping using SNP marker haplotypes: application to a potential locus for schizophrenia at chromosome 22q11. Mol Psychiatry 2000; 5: 77–84. Erratum in: Mol Psychiatry 2000; 5(4):452.
Craddock N, O'Donovan MC, Owen MJ . The genetics of schizophrenia and bipolar disorder: dissecting psychosis. J Med Genet 2005; 42: 193–204.
Craddock N, O'donovan MC, Owen MJ . Genes for schizophrenia and bipolar disorder? Implications for psychiatric nosology. Schizophr Bull 2006; 32: 9–16.
Bertolino A, Caforio G, Blasi G, De Candia M, Latorre V, Petruzzella V et al. Interaction of COMT (Val(108/158)Met) genotype and olanzapine treatment on prefrontal cortical function in patients with schizophrenia. Am J Psychiatry 2004; 161: 1798–1805.
Weickert TW, Goldberg TE, Mishara A, Apud JA, Kolachana BS, Egan MF et al. Catechol-O-methyltransferase val108/158met genotype predicts working memory response to antipsychotic medications. Biol Psychiatry 2004; 56: 677–682.
Potash JB, Zandi PP, Willour VL, Lan TH, Huo Y, Avramopoulos D et al. Suggestive linkage to chromosomal regions 13q31 and 22q12 in families with psychotic bipolar disorder. Am J Psychiatry 2003; 160: 680–686.
Funke B, Malhotra AK, Finn CT, Plocik AM, Lake SL, Lencz T et al. COMT genetic variation confers risk for psychotic and affective disorders: a case control study. Behav Brain Funct 2005; 18: 1:19.
Craddock N, Raybould R, Green E, Macgregor S, Grozeva D, Williams H et al. Genetic variation at or near COMT influences susceptibility to a phenotype characterized by the co-existence of marked features of mania and psychosis. Am J Med Genet B Neuropsychiatr Genet 2005; 138B: 23–24, (abstract).
Green E, Craddock N . Brain-derived neurotrophic factor as a potential risk locus for bipolar disorder: evidence, limitations, and implications. Curr Psychiatry Rep 2003; 5: 469–476.
Wang WY, Barratt BJ, Clayton DG, Todd JA . Genome-wide association studies: theoretical and practical concerns. Nat Rev Genet 2005; 6: 109–118.
Suarez BK, Hampe CL, Van Eerdewegh P . Problems of replicating linkage claims in psychiatry. In: Gerson ES, Cloninger CR (eds). Genetic Approaches to Mental Disorders. American Psychiatric Press, Inc.: Washington, DC, 1994, pp 23–46.
Altshuler D, Hirschhorn JN, Klannemark M, Lindgren CM, Vohl MC, Nemesh J . The common PPARgamma Pro12Ala polymorphism is associated with decreased risk of type 2 diabetes. Nat Genet 2000; 26: 76–80.
Berrettini W . Evidence for shared susceptibility in bipolar disorder and schizophrenia. Am J Med Genet C Semin Med Genet 2003; 123: 59–64.
Craddock N, Owen MJ . The beginning of the end for the Kraepelinian dichotomy. Br J Psychiatry 2005; 186: 364–366.
Gordon D, Finch SJ, Nothnagel M, Ott J . Power and sample size calculations for case–control genetic association tests when errors are present: application to single nucleotide polymorphisms. Hum Hered 2002; 54: 22–33.
Kang SJ, Finch SJ, Haynes C, Gordon D . Quantifying the percent increase in minimum sample size for SNP genotyping errors in genetic model-based association studies. Hum Hered 2004; 58: 139–144.
Kang SJ, Gordon D, Finch SJ . What SNP genotyping errors are most costly for genetic association studies? Genet Epidemiol 2004; 26: 132–141.
Rice KM, Holmans P . Allowing for genotyping error in analysis of unmatched case–control studies. Ann Hum Genet 2003; 67: 165–174.
Knapp M, Becker T . Impact of genotyping errors on type I error rate of the haplotype-sharing Transmission/Disequilibrium test (HS-TDT). American Journal of Human Genetics 2004; 74: 589–591.
Kirk KM, Cardon LR . The impact of genotyping error on haplotype reconstruction and frequency estimation. Eur J Hum Genet 2002; 10: 616–622.
Clayton D, Walker NM, Smyth DJ, Pask R, Cooper JD, Maier LM . Population structure, differential bias and genomic control in a large-scale, case–control association study. Nat Genet 2005; 37: 1243–1246.
Moskvina V, Craddock N, Holmans P, Owen M, O'Donovan M . Minor genotyping error can result in substantial elevation in type I error rate in haplotype based case control studies. Am J Med Genet (Neuropsychiatric Genetics) 2005; 138B: 19, (abstract).
Cucherat M, Boissel JP, Leizorovicz A, Haugh MC . EasyMA: a program for the meta-analysis of clinical trials. Comput Methods Programs Biomed 1997; 53: 187–190.
BIOMED. [No authors listed]. No association between bipolar disorder and alleles at a functional polymorphism in the COMT gene. Biomed European Bipolar Collaborative Group. Br J Psychiatry 1997; 170: 526–528.
Gutierrez B, Bertranpetit J, Guillamat R, Valles V, Arranz MJ, Kerwin R et al. Association analysis of the catechol-O-methyltransferase gene and bipolar affective disorder. Am J Psychiatry 1997; 154: 113–115.
Lachman HM, Kelsoe J, Moreno L, Katz S, Papolos DF . Lack of association of catechol-O-methyltransferase (COMT) functional polymorphism in bipolar affective disorder. Psychiatr Genet 1997; 7: 13–17.
Kunugi H, Vallada HP, Hoda F, Kirov G, Gill M, Aitchison KJ et al. No evidence for an association of affective disorders with high- or low-activity allele of catechol-O-methyltransferase gene. Biol Psychiatry 1997; 42: 282–285.
Li T, Vallada H, Curtis D, Arranz M, Xu K, Cai G . Catechol-O-methyltransferase Val158Met polymorphism: frequency analysis in Han Chinese subjects and allelic association of the low activity allele with bipolar affective disorder. Pharmacogenetics 1997; 7: 349–353.
We are grateful to the Medical Research Council (UK) who support our work on schizophrenia, and to the Wellcome Trust who fund our studies on bipolar spectrum disorders and our psychosis work on chromosome 22.
About this article
Cite this article
Craddock, N., Owen, M. & O'Donovan, M. The catechol-O-methyl transferase (COMT) gene as a candidate for psychiatric phenotypes: evidence and lessons. Mol Psychiatry 11, 446–458 (2006). https://doi.org/10.1038/sj.mp.4001808
- bipolar disorder
- schizoaffective disorder
Molecular Neurobiology (2020)
The Pharmacogenomics Journal (2016)
Eating and Weight Disorders - Studies on Anorexia, Bulimia and Obesity (2016)
Potential relationship between single nucleotide polymorphisms used in forensic genetics and diseases or other traits in European population
International Journal of Legal Medicine (2015)
Brain Structure and Function (2014)