Main

Catechol-O-methylransferase (COMT) had an illustrious beginning. The enzyme, which metabolizes catecholamines and catechol-estrogens in the CNS and periphery, was discovered by Julius Axelrod in the late 1950s, contributing to his share of the 1970 Nobel Prize for Physiology or Medicine (Axelrod and Tomchick, 1958). COMT soon captured the attention of pioneering biological psychiatrists, largely as a consequence of the catecholamine hypothesis, the prevailing early biological model for affective disorders (Schildkraut, 1965). Analysis of red blood cell COMT activity and norepinephrine metabolites in the CNS in psychiatric disorders developed into a cottage industry. These studies, in retrospect, seem quaint (even naive) in terms of the miniscule number of subjects analyzed, the lack of attention to confounders, such as population stratification, and the analytical tools used (compare measuring red cell COMT enzyme activity in a few dozen subjects with the current craze of genotyping one million SNPs in thousands of carefully matched cases and controls). Foretelling the experience of more ‘sophisticated’ future generations of psychiatric geneticists, the early COMT studies were characterized by a grab bag of findings; increased, decreased or no change in activity in depression, bipolar disorder, and schizophrenia were all described.

Interest in COMT surged in the 1990s following the discovery of a common functional genetic variant at codon 158 (val158met), which leads to substantial differences in enzyme activity (Lachman et al, 1996; Lotta et al, 1995). Genetic analysis of this functional SNP provided a simple, and biologically meaningful way by which psychiatric geneticists could reexamine COMT as a candidate gene for psychiatric disorders in large numbers of subjects. Since then, hundreds of studies have been published in which COMT val158met has been examined for essentially every neuropsychiatric disorder. Given the recalcitrant nature of complex traits, however, the modern generation of researchers has fared only a little better than its predecessors, and the role of COMT in psychiatric disorders remains largely an open question (Craddock et al, 2006). However, as reported in the perspective by Harrison and Tunbridge (2008), evidence for sexual dimorphism with respect to COMT val158met in OCD is arguably one of the more robust associations. In addition, based on the original work by Egan et al and a number of follow-up studies COMT val158met has become a leading contender to explain part of the genetic variance underlying interindividual differences in prefrontal executive function, with val158 being associated with worse performance, albeit with a small effect size (Egan et al, 2001). However, a recent meta-analysis questions the validity of those findings (Barnett et al, 2008).

This issue of Neuropsychopharmacology with five original papers and the Harrison and Tunbridge perspective features the analysis of val158met in a variety of behavioral and pharmacological paradigms adding to the everincreasing COMT literature, with some interesting new twists and experimental designs. One of the most intriguing is the study by Giakoumaki et al (2008). The investigators carried out a double blind crossover assessment of executive function and prepulse inhibition using the centrally acting COMT inhibitor tolcapone in 24 healthy subjects homozygous for one or the other codon 158 variant. They found that tolcapone improved executive function and prepulse inhibition, but only in val/val homozygotes, a finding consistent with the lower levels of cortical dopamine expected of individuals with this genotype. The small sample size and preliminary nature of the observations prevents one from extrapolating too quickly to treating patients with cognitive dysfunction with tolcapone or similarly acting drugs, as suggested previously by Apud and Weinberger (2007), perhaps in a COMT genotype dependent manner. However, the results are sufficiently interesting and potentially of such clinical importance that further investigation must be carried out.

Sengupta et al (2008) report another ‘pharmacogenomic’ type of analysis related to COMT genotype in which task-oriented behaviors and response to methylphenidate in ADHD were assessed. Children with ADHD (n=188) were subjected to a double blind, placebo controlled cross over study and evaluated by RASS (Restricted Academic Situation Scale), which records fidgety and distracting behavior in a simulated academic setting (lower scores mean greater attention). Although no significant association between COMT genotype and ADHD susceptibility had been described previously, Sengupta et al found that children with met/met and met/val had lower RASS scores compared with val/val. However, although methyphenidate significantly improved RASS scores, the response was independent of genotype. If the findings are confirmed, it would support the idea that an ADHD endophenotype is associated with val158met. Although this understanding may not be clinically relevant with respect to predicting methyphenidate responsiveness, the findings suggest that a COMT inhibitor might benefit children with ADHD who carry the val/val genotype.

Two of the studies in this issue are related in their assessment of val158met as a candidate for addiction or addiction-related phenotypes. Lohoff et al (2008) provide evidence for an association between met158 and cocaine dependent African Americans. This is in contrast to positive associations made in methamphetamine abusers, nicotine addiction, and polysubstance abusers to the val158 allele in previously published studies (Beuten et al, 2006; Li et al, 2004; Vandenberg et al, 1997). If the findings are replicated, it could suggest that COMT may play a role in addiction vulnerability in a genotype/drug-dependent manner. However, this is unlikely considering that many cocaine addicts are polysubstance abusers. Whether the Lohoff et al study and the contrary findings in other addiction studies stand the test of time (and replication) remains to be seen; most studies examining the role of COMT in addiction have been negative.

Consistent with the Beuten et al, Li et al, and Vandenberg et al findings, the paper reported here by Wichers et al (2008) also supports the idea that val158 could be an addiction susceptibility allele. Test subjects (all women to reduce COMT-related sex effects as a confounder) were asked to assess their daily living environments using ESM (experience sampling method), a structured diary technique. Events were rated as very unpleasant, neutral, and very pleasant. ‘Positive affect’ was also assessed using four mood adjectives (cheerful, content, energetic, enthusiastic) rated on a seven-point Likert scale, as was ‘negative affect’ (by six adjectives; feeling insecure, lonely, anxious, low, guilty, suspicious). Ability to experience everyday reward was associated with met158, whereas subjects with the val/val genotype experienced significantly less reward. This finding is compatible with the ‘reward deficiency’ hypothesis, which posits that individuals who do not experience sufficient reward from everyday pleasures (food, sex, social interaction, work, and school, for example) are prone to habitually seek the intensity of brain reward region activation induced by addicting substances. Wichers et al suggest that individuals with val/val genotypes who have suboptimal reward experiences are more at risk for depression and addiction. Longitudinal studies to follow the progression from lack of reward sensation to addiction, in the context of COMT genotype, would be especially interesting.

Finally, the paper by Ettinger et al (2008) contributes to the growing literature involving the analysis of brain functional responses as plausible endophenotypes that can be used to identify genes involved in complex psychiatric disorders more effectively, as suggested by Meyer-Lindenberg and Weinberger (2006). Ettinger et al evaluated the effect of val158met on fMRI BOLD response during prosaccade and antisaccade task performance and found differential effects of val158 and met158. Although val158 carriers (homozygotes and heterozygotes) showed lower BOLD response in the prefrontal cortex during antisaccades, met158 homozygotes showed lower BOLD response in postsaccades in the posterior cingulate and precuneus. This differential response is compatible with the hypothesis proposed by Bilder et al (2004) who suggested that val158 and met158 have opposite effects on phasic and tonic dopamine transmission in cortex and subcortical regions, influencing cognitive function and behavior. Although intriguing, the Ettinger study is limited by small sample size (n=36).

Fortunately we are beyond the stage where we must measure levels of enzymes and substrates in every conceivable bodily fluid to piece together the role played by COMT in a biological system of interest. The COMT gene provides a relatively rare example where there is a common functional polymorphism that can be used to track differences between individuals in enzyme activity important for neurotransmission. Although our ability to study this variant has opened a proverbial window into the brain through which we can begin to observe individual differences in catecholaminergic metabolism, there are few ‘eureka’ moments in complex traits genetics, and none of the original studies reported here fulfill this lofty ambition. However, if one is content presenting novel and interesting first-stage findings, then the investigators contributing to the COMT papers in this volume have successfully accomplished this goal.