Association studies on COMT gene polymorphism

2.4.4.1 COMT enzyme and COMT polymorphism

Catechol-O-methyltransferase enzyme inactivates catechol hormones, drugs containing catechol (e.g. L-dopa for Parkinson’s disease) and catecholamines, including those acting as neurotransmitters such as dopamine. COMT enzyme occurs in mammals as two distinct forms: a cytoplasm-soluble protein (S-COMT) and a membrane-bound protein (MB-(S-COMT). S-COMT activity is the most prevalent in all tissues except the human brain, where most of the enzyme activity (70%) comes from MB-COMT (Tenhunen et al., 1994; Syvänen et al., 1997). The level of COMT activity in humans has been shown to vary between individuals in trimodal distribution of low, intermediate and high levels, and was known to be genetically polymorphic already in the 1970s (Lachman et al., 1996).

Both forms of COMT enzyme are coded by a single gene, which is expressed in all tissues and located on chromosome 22q11. No brain specific variants have been detected (Syvänen et al., 1997; Lachman et al., 1996). A common polymorphism at codon 108/158 (S-COMT / MB-COMT) of the COMT gene, resulting from a nucleotide transition (G to A) and causing a valine to methionine substitution in the enzyme, was found to be functional.

Homozygosity for the valine (G) allele was associated with three- to four-fold enzyme activity compared with the homozygous methionine (A) genotype (and thus possibly leading to less sustained dopamine effect in the brain). The

heterozygous genotype had an intermediate level of enzyme activity. At least some of the variation in activity may be caused by the thermolability of the COMT-methionine form. These results obtained from human liver tissue or red blood cells will very probably be expressed in brain tissue as well (Lachman et al., 1996). The frequency of the valine/methionine alleles in white population is found to be 0.51/0.49 (Bevilacqua and Goldman, 2009). The COMT valine and methionine alleles (val and met) are later also referred to as H (for high activity) and L (for low activity) alleles, respectively.

The sex- and brain-region-specific contribution of COMT enzyme activity (of genetic origin) in the maintenance of dopamine levels and in the regulation of dopamine neurotransmission and consequent changes in behaviour was shown in a study on COMT knock-out mice (Gogos et al., 1998). COMT-deficient homozygous male mice showed a 2- to 3-fold higher level of dopamine in the frontal cortex, but no changes of dopamine content in either the striatum or hypothalamus were noted, nor were there any locomotor deficits present.

COMT-deficient homozygous female mice showed higher anxiety. Aggressive behaviour was markedly higher in heterozygous COMT-deficient male mice.

The authors speculated that the COMT activity of a heterozygous COMT-deficient mouse is comparable with that in a human subject homozygous for the L allele, although in general experimental results on knock-out mice cannot readily be used as a model for complex human behaviours.

It has been detected that COMT valine and methionine alleles reside on haplotypes (i.e. in patterns of alleles at nearby linked loci, transmitted and inherited together rather than segregated by meiotic recombination), and these haplotypes are common to different populations. These aspects reflect an ancient origin of both alleles in the human genome, and a potential role of selection in their maintenance (Oroszi and Goldman, 2004; Goldman et al., 2005a).

2.4.4.2 COMT polymorphism and cognitive function, and personality traits The abundant literature concerning COMT polymorphism and major mental disorders such as schizophrenia or bipolar disorder is not discussed here because it is irrelevant. The effect of COMT polymorphism on personality and cognition is closely related to alcohol use disorders, and it is therefore briefly discussed below.

The inherited differences in COMT activity are likely to predict variation in dopaminergic neurotransmission in the frontal cortex of the brain (prefrontal area), where the levels of another crucial regulator, dopamine transporters, are low. This was shown in the study on COMT knock-out male mice described above (Gogos et al., 1998). Solid data indicate that dopamine enhances prefrontal neuronal function during working memory tasks and thus also enhances cognitive performance (Egan et al., 2001; Goldman et al., 2005a).

A highly sophisticated study among 175 patients with schizophrenia, their 219 unaffected siblings and 55 healthy controls showed that low activity L allele of COMT polymorphism predicted better cognitive performance during neuropsychological testing of executive functions. When the normal subjects were assessed with functional magnetic resonance imaging (fMRI) during a working memory task, the L allele again predicted a more efficient physiological response in the prefrontal cortex. The effect was allele dose dependent: the individuals homozygous for L allele showed the most efficient results, and heterozygotes were intermediate when compared with the least efficient group of subjects homozygous for high activity H allele (with faster dopamine inactivation and less sustained transmission). In a family-based analysis there was also a significant hereditary transmission of the H allele to schizophrenic offspring. This indicates that the COMT H allele might also be if not an actual risk factor for then a modulator of the prognosis in schizophrenia, because it impairs prefrontal function (Egan et al., 2001). The COMT H allele and its haplotypes have been linked to prefrontal lobe function in other studies, as well (Malhotra et al., 2002).

On the other hand, the low activity L allele has also been linked to higher levels of anxiety and a lower pain threshold. A study among two community samples of 149 white and 252 Plains American Indians subjects (predominantly females, without lifetime psychiatric diagnosis) found an association between COMT L/L homozygous genotype and higher anxiety levels measured by the Tridimensional Personality Questionnaire (TPQ, Cloninger et al., 1991) and low-voltage alpha resting EEG among the female subjects (Enoch et al., 2006a).

These EEG findings had previously been shown to be associated with anxiety and alcoholism by the authors, potentially reflecting intermediate mechanisms (Enoch et al., 1999). A lower pain threshold, stronger affective response to pain, and inability to activate the endogenous brain opioid system following pain was associated with the L allele in a brain imaging study by Zubieta et al. (2003).

Lower pain threshold was associated with L allele haplotypes in a study of a large cohort of women by Diatchenko et al. (2005), as well.

Women generally are considered to be more prone to anxiety than men.

However, the findings above might indicate that the carriers of the low activity L allele tend to "worry" or suffer from anxiety because of the lower resiliency to stress and pain, but they also gain better cognitive functions in change (with more sustained frontal dopamine transmission). This balance of advantages forms the basis for the "Warrior/Worrier" model (Goldman et al., 2005a): the warrior has better stress resiliency, and a higher pain threshold with a more functional endogenous opioid system, but also less effective executive cognitive performance (which probably reduces stress before the “battle”). This dichotomy of advantages/losses may be responsible for the selection to conserve both of these alleles (or haplotypes containing the alleles) in the human genome across populations.

There are also differences in COMT activity based on sex. At least in vitro estrogen can inhibit COMT gene transcription, and women have been shown to have significantly lower COMT activity than men (Enoch et al., 2006a). This is consistent with the data mentioned above on the association between anxiety and low COMT activity and, on the other hand, between female sex and

predisposition to anxiety. Sexually dimorphic behavioural changes in mice associated with low COMT activity are described above (Gogos et al., 1998).

Altogether, this indicates that sex may modify genetically determined behaviour, the same genotype resulting in different behavioural consequences in males and females.

It has also been suggested that the COMT genotype and associated variation in prefrontal cognitive function act as a modifier against a background of other aetiological factors to affect behaviour. In a recent study among almost 500 children from three cohorts, Caspi et al. (2008) showed a conditional association between antisocial conduct disorder and the COMT H/H (Val/Val) genotype, presumably with less effective prefrontal cognitive processing. The association was observed only in children with ADHD, not in those without. It is generally accepted that about 50% of the children and adolescents with ADHD show antisocial tendencies (Caspi et al., 2008).

2.4.4.3 COMT polymorphism and alcoholism or substance abuse

Dopamine release in the human brain has been suggested to play a major role in euphoria induced by substances of abuse and thus possibly in the development of alcoholism. Consequently, it would be logical to search for an association between the functional COMT L/H polymorphism and alcoholism or other substance abuse. A study among 175 Japanese alcoholics (the majority males) and 350 age and gender matched controls found no difference in COMT allele frequencies or genotypes between subjects and controls. No association between genotypes and age of onset of alcoholism or antisocial behaviour were detected, either (Ishiguro et al., 1999a). Vandenbergh et al. (1997) found that the COMT H allele was significantly more common among 200 white polysubstance abusers than 100 controls. The association was explained through impulsivity or dyscontrol, not by liability to excessive anxiety.

Altogether, association studies on the COMT genotype and alcoholism were rare in the 1990s.