Abstract

The prefrontal cortex (PFC) dopamine system, which is critical for modulating PFC function, undergoes remodeling until at least young adulthood in primates. Catechol-o-methyltransferase (COMT) alters extracellular dopamine levels in PFC, and its gene contains a functional polymorphism (Val158Met) that has been associated with variation in PFC function. We examined COMT enzyme activity and protein immunoreactivity in the PFC during human postnatal development. Protein was extracted from PFC of normal individuals from 6 age groups: neonates (1–4 months), infants (5–11 months), teens (14–18 years), young adults (20–24 years), adults (31–43 years), and aged individuals (68–86 years; n = 5–8 per group). There was a significant 2-fold increase in COMT enzyme activity from neonate to adulthood, paralleled by increases in COMT protein immunoreactivity. Furthermore, COMT protein immunoreactivity was related to Val158Met genotype, as has been previously demonstrated. The significant increase in COMT activity from neonate to adulthood complements previous findings of protracted postnatal changes in the PFC dopamine system and may reflect an increasing importance of COMT for PFC dopamine regulation during maturation.

Introduction

The prefrontal cortex (PFC) shows a protracted postnatal developmental course, with refinement of synaptic innervation and connectivity occurring well into adulthood (Lewis 1997; Levitt 2003). This extended time course means that PFC function may, relative to other brain regions, be particularly sensitive to neurodevelopmental lesions (Levitt 2003) and, moreover, their functional consequences might not be manifest until adulthood (Weinberger 1987). Thus, the postadolescent age of onset of certain neurodevelopmental disorders that relate to dysfunction of the PFC, such as schizophrenia, may only become clinically apparent once this brain region has fully matured (Weinberger 1987; Weinberger and Marenco 2003). Therefore, elucidating the maturation of the PFC from birth until adulthood is critical both to our understanding of its normal function and also its dysfunction in neurodevelopmental disorders such as schizophrenia.

Dopamine plays a critical role in modulating normal PFC function (Goldman-Rakic and others 2000), and the PFC dopamine system undergoes extensive refinement during postnatal life (Lambe and others 2000). The primate PFC does not receive its full complement of synaptic dopamine innervation until around adolescence. Thus, the density of catecholamine-positive fibers and varicosities in the rhesus monkey PFC peaks in adolescence, before receding to adult levels (Rosenberg and Lewis 1995). Additionally, the density of catecholamine appositions onto pyramidal cells, but not onto interneurons, doubles from birth to adolescence in rhesus monkey PFC, suggesting that the increase in dopamine input might differentially affect excitatory versus inhibitory transmission in cortical circuits (Lambe and others 2000). Furthermore, the abundances of dopamine receptors and tyrosine hydroxylase, the rate-limiting enzyme for dopamine biosynthesis, are also dynamically regulated during primate postnatal development (Lidow and Rakic 1992; C.S. Weickert, M.J. Webster, P. Gondipalli, D. Rothmond, R.J. Fatula, M.M. Herman, J.E. Kleinman, and M. Akil, unpublished data). Mirroring these anatomical changes, working memory performance, which is known to be critically dependent on PFC dopamine function (Goldman-Rakic 2000), does not peak until after adolescence (reviewed in Lewis 1997; Luna and others 2004), consistent with the increase in dopaminergic signaling during this period. Interestingly, information on the expression of molecules responsible for dopamine inactivation during this period is sparse, despite the critical role that such molecules likely play in modulating PFC dopaminergic function.

The activity of the catechol-o-methyltransferase (COMT) enzyme modulates dopamine levels in the PFC (Karoum and others 1994; Gogos and others 1998; Tunbridge and others 2004). Elimination of dopamine via catabolism by COMT is thought to be particularly important in the PFC, relative to the striatum, because dopamine transporters in the former region are extrasynaptic, permitting greater neurotransmitter diffusion (Lewis and others 2001). The COMT gene contains a polymorphism (Val158Met) that affects the activity of the enzyme: Met158 homozygotes have approximately one-third less COMT enzyme activity in PFC than Val158 homozygotes (Chen and others 2004), and the reduced enzyme activity determined by the Met158 allele presumably results in increased PFC dopamine relative to Val158 carriers because COMT inhibition results in elevated dopamine efflux (Tunbridge and others 2004). Consistent with its role in modulating PFC dopamine levels, the Val158Met polymorphism is associated with performance on tests of working memory and executive function, which depend on PFC function. Thus, the high-activity Val158 allele is linked with relatively poorer performance on such tasks, relative to the Met158 allele (e.g., Egan and others 2001; Bilder and others 2002; Joober and others 2002; Malholtra and others 2002), presumably as a result of increased PFC dopamine catabolism. Paralleling these findings, administration of tolcapone, a COMT inhibitor, improves performance of extradimensional set shifting in rats (Tunbridge and others 2004), a task dependent on medial PFC in rats (Birrell and Brown 2000) and the dorsolateral PFC in primates (Dias and others 1996). Tolcapone also improves cognitive function in humans (Gaspirini and others 1997; Mattay and others 2004). COMT may also be genetically associated with schizophrenia (Egan and others 2001; Shifman and others 2002; discussed in Tunbridge, Harrison, and Weinberger 2006). Thus, knowledge of the developmental profile of COMT activity is relevant to understanding the development of the normal PFC dopamine system, as well as its dysfunction in schizophrenia and other neurodevelopmental disorders.

Studies of the postnatal expression profile of COMT are few (Agathopolous and others 1971; Stanton and others 1975; Brust and others 2004; see Discussion), and none have been conducted in human brain. Thus, we sought to determine COMT activity and expression during postnatal maturation in the PFC of normal humans. We demonstrate a dramatic increase in PFC COMT activity across the human postnatal lifespan, which correlates with increases in COMT protein immunoreactivity determined by immunoblotting.

Materials and Methods

Tissue Collection and Protein Extraction

Specimens were obtained through the Office of the Medical Examiners of the District of Colombia and were fresh frozen and processed as previously described (Kleinman and others 1995) in the Clinical Brain Disorders Branch of the National Institute of Mental Health. All subjects were free of psychiatric and neurologic disorders and of significant neuropathological findings. The cohort used in this study (see Table 1) consisted of neonates (1–4 months, n = 8), infants (5–11 months, n = 5), teens (14–18 years, n = 8), young adults (20–24 years, n = 6), adults (31–43 years, n = 7), and aged (68–86 years, n = 6). All individuals were of African American origin, except for one individual of Asian descent, included in the aged group. Pulverized dorsolateral PFC tissue (Brodmann's area 46) was thawed and homogenized over wet ice in extraction buffer (0.6% Tris and 50% glycerol containing protease inhibitors [0.024% AEBSF, 0.005% aprotinin, 0.001% pepstatin A, and 0.001% leupeptin]). Protein concentrations were determined using the Bradford method, and crude homogenates were stored in single-use aliquots (20 μL of a 1 μg/μL dilution) at −80 °C.

Table 1

Demographics of the developmental cohort

Group (nAge (years) Sex Val158Met genotypes Postmortem intervala Brain pH Cause of death (n
Neonates (8) 0.20 ± 0.03 3M/5F V/V: n = 2, V/M: n = 6, M/M: n = 0 48.8 ± 9.4 6.34 ± 0.11 SIDS (5); endocardial fibroelastosis (1); bronchopneumonia (1); undetermined (1). 
Infants (5) 0.60 ± 0.10 3M/2F V/V: n = 3, V/M: n = 2, M/M: n = 0 46.8 ± 5.8 6.53 ± 0.08 SIDS (2); bronchiolitis (1); diphenhydramine intoxication (1); cardiomegaly, atrial septal defect (1). 
Adolescents (8) 16.8 ± 0.62 8M V/V: n = 4, V/M: n = 4, M/M: n = 0 25.5 ± 5.0 6.48 ± 0.06 GSW to chest (3), back (2), abdomen (1); multiple GSW (1); stab wound to chest (1). 
Young adults (6) 22.2 ± 0.65 6M V/V: n = 3, V/M: n = 2, M/M: n = 1 37.2 ± 4.7 6.34 ± 0.12 GSW to torso (2), chest (1); stab wound to chest (1); fibrinous pericarditis (1); pulmonary embolism (1). 
Adults (7) 37.7 ± 1.71 7M V/V: n = 4, V/M: n = 3, M/M: n = 0 27.4 ± 6.2 6.43 ± 0.06 ASCVD (1); stab wound to chest (1); acute asthma attack (1); GSW to chest (1); pulmonary embolism (1); multiple GSW (1); GSW to torso (1). 
Aged (6) 76.7 ± 2.85 5M/1F V/V: n = 5, V/M: n = 1, M/M: n = 0 45.8 ± 6.2 6.33 ± 0.11 ASCVD (2); GSW wound to chest (1); sepsis (1); pulmonary embolism (1); undetermined (1). 
Group (nAge (years) Sex Val158Met genotypes Postmortem intervala Brain pH Cause of death (n
Neonates (8) 0.20 ± 0.03 3M/5F V/V: n = 2, V/M: n = 6, M/M: n = 0 48.8 ± 9.4 6.34 ± 0.11 SIDS (5); endocardial fibroelastosis (1); bronchopneumonia (1); undetermined (1). 
Infants (5) 0.60 ± 0.10 3M/2F V/V: n = 3, V/M: n = 2, M/M: n = 0 46.8 ± 5.8 6.53 ± 0.08 SIDS (2); bronchiolitis (1); diphenhydramine intoxication (1); cardiomegaly, atrial septal defect (1). 
Adolescents (8) 16.8 ± 0.62 8M V/V: n = 4, V/M: n = 4, M/M: n = 0 25.5 ± 5.0 6.48 ± 0.06 GSW to chest (3), back (2), abdomen (1); multiple GSW (1); stab wound to chest (1). 
Young adults (6) 22.2 ± 0.65 6M V/V: n = 3, V/M: n = 2, M/M: n = 1 37.2 ± 4.7 6.34 ± 0.12 GSW to torso (2), chest (1); stab wound to chest (1); fibrinous pericarditis (1); pulmonary embolism (1). 
Adults (7) 37.7 ± 1.71 7M V/V: n = 4, V/M: n = 3, M/M: n = 0 27.4 ± 6.2 6.43 ± 0.06 ASCVD (1); stab wound to chest (1); acute asthma attack (1); GSW to chest (1); pulmonary embolism (1); multiple GSW (1); GSW to torso (1). 
Aged (6) 76.7 ± 2.85 5M/1F V/V: n = 5, V/M: n = 1, M/M: n = 0 45.8 ± 6.2 6.33 ± 0.11 ASCVD (2); GSW wound to chest (1); sepsis (1); pulmonary embolism (1); undetermined (1). 

Note: SIDS, sudden infant death syndrome; GSW, gunshot wound; ASCVD, arteriosclerotic cardiovascular disease; M, male; F, female.

a

Differs between groups at trend significance level.

COMT Activity Assay

COMT enzyme activity was assayed by measuring the incorporation of a radioactive methyl group into a catechol substrate in crude tissue homogenates, as previously described (Chen and others 2004). Because this COMT assay is linear over a wide range of protein concentrations (0–150 μg tissue; Chen and others 2004), we used 20 μg of crude protein, which we found to yield a linear relationship between measured COMT activity and reaction time (r = 0.995; data not shown). Background levels of radioactivity were determined by adding 10 mM tolcapone to randomly selected samples (4 per assay), and this background was subtracted from measured levels of COMT activity. COMT enzyme activity was assayed in all samples at one time. The assay was repeated 4 times, and data presented are the mean values for these 4 experiments, with each sample normalized to mean adult COMT activity levels, considered to be 100%. The raw data are presented in the supplementary information.

Western Blotting

Crude tissue homogenates (2.5 μg total protein) were taken from the same protein samples used for the enzyme activity assay (determined during pilot studies to be within the linear range of protein quantification for both COMT and β-actin) and diluted to a volume of 10 μL in sample buffer (LDS; Invitrogen, Carlsbad, CA) and sterile water and boiled for 7 min. They were then run on 15% Tris glycine polyacrylamide gels (120 V for ∼2 h) and transferred to polyvinylidene fluoride membrane (16 h at 30 V). Membranes were labeled with anti-COMT antibody (Chemicon, Temecula, CA; Catalog number AB5873; 1:5000 dilution) as previously described (Tunbridge, Weinberger, and Harrison 2006) and were apposed to Kodak Biomax MR film. Membranes were then rinsed and incubated with anti-actin antibody (Chemicon; Catalog number MAB1501; 1:3000 dilution) to monitor variations in sample handling and western blotting. Films were digitized, and individual bands were quantified using NIH Image (v1.32j; profile plots function). COMT immunoreactivity was normalized to the corresponding β-actin measurement, although analysis of raw data produced similar results. Western blotting was performed 3 times, and data presented are the mean normalized values for these 3 experiments.

Genotyping

Genomic DNA was extracted from pulverized cerebellar hemisphere using Puregene reagents (Gentra, Minneapolis, MN). Samples were genotyped for the COMT Val158Met polymorphism using a Taqman 5′ exonuclease assay as previously described (Chen and others 2004). The distribution of COMT Val158Met genotypes in the different age groups is shown in Table 1. Allele frequencies are as expected for a cohort with this racial profile (Palmatier and others 2004).

Data Analysis

All experiments and image analyses were performed blind to the identities of the samples. One-way analysis of variance (ANOVA) was used to investigate whether age groups were matched for demographic variables (Table 1), and the effect of brain pH and postmortem interval on COMT enzyme activity and protein expression was investigated using Pearson's correlations. The effect of gender was examined by t-test in the infant and neonate groups combined (because older age groups consisted largely, or entirely, of males). The effect of age group on COMT enzyme activity was explored using one-way ANOVA, with pH and postmortem interval included as covariates (although inclusion or exclusion of these covariates had no effect on the findings), and correlation. The relationship between COMT enzyme activity and aspects of COMT protein expression was investigated using Pearson's correlations. The effect of Val158Met genotype on COMT enzyme activity and protein immunoreactivity was determined using t-tests or one-way ANOVA in order to investigate the effect of including age as a covariate (the single Met158 homozygote in the cohort was excluded from these analyses).

Results

COMT Enzyme Activity during the Postnatal Lifespan

COMT enzyme activity was successfully detected in all individuals, and there was no significant effect of brain pH or postmortem interval (F values < 1). There was a significant increase in COMT enzyme activity from the neonate period to adulthood (main effect of age group: F5,25 = 4.0, P < 0.01; Fig. 1; see Table 2 for post hoc comparisons), which continued into adulthood (see Table 2) and was also reflected as a significant correlation between COMT enzyme activity and increasing age (r = 0.64, P < 0.0001). There was no significant effect of gender on COMT activity in infants and neonates (t = 1.1, P > 0.1).

Figure 1.

COMT enzyme activity in dorsolateral PFC across human postnatal development. There is a significant increase in COMT enzyme activity (expressed as a percentage of mean adult values) in the PFC during human postnatal development. COMT enzyme activity does not peak until adulthood, at which point it is significantly greater than in all other groups, except aged individuals (see Table 2).

Figure 1.

COMT enzyme activity in dorsolateral PFC across human postnatal development. There is a significant increase in COMT enzyme activity (expressed as a percentage of mean adult values) in the PFC during human postnatal development. COMT enzyme activity does not peak until adulthood, at which point it is significantly greater than in all other groups, except aged individuals (see Table 2).

Table 2

P values from post hoc comparisons of COMT enzyme activity between age groups.

 Neonates Infants Teens Young adults Adults 
Infants NS     
Teens NS NS    
Young adults NS NS NS   
Adults 0.001 0.004 0.006 0.029  
Aged 0.015 0.024 0.079 NS NS 
 Neonates Infants Teens Young adults Adults 
Infants NS     
Teens NS NS    
Young adults NS NS NS   
Adults 0.001 0.004 0.006 0.029  
Aged 0.015 0.024 0.079 NS NS 

Note: NS, not significant.

Correlations between COMT Enzyme Activity and Protein Immunoreactivity

Western blotting for COMT revealed the presence of 4 major bands at approximately 39, 32, 26, and 24 kDa. Whereas the 39-, 32-, and 24-kDa bands were present in all individuals, the 26-kDa band was not visible in some samples (Fig. 2). Based on its size, the 32-kDa band is presumed to be the membrane-bound form of catechol-o-methyltransferase (MB-COMT), and we have recently identified the 39-kDa band as a novel variant of COMT (presumably an isoform of MB-COMT, based on its size and relative abundance; Tunbridge, Weinberger, and Harrison 2006). The finding of a second doublet at around 24 and 26 kDa is novel, and, given that these bands are the approximate size of soluble COMT (S-COMT), these may represent isoforms of S-COMT, although it is also possible that they represent breakdown products of the larger COMT bands. None of the western blotting immunoreactive bands showed significant correlations with either brain pH (−0.30 < r < 0.29) or postmortem interval (−0.26 < r < 0.11), except for expression of the 39-kDa band, which was significantly correlated with brain pH (r = 0.42, P < 0.01).

Figure 2.

Examples of COMT and actin immunoblots in 2 different subjects. Immunoblotting with an anti-COMT antibody produced 4 major bands at 39, 32, 26, and 24 kDa (see left-hand panel). The 26-kDa band was not visible in some individuals (see example in right-hand panel). Immunoblotting with an antiactin antibody produced a single band at 50 kDa, the predicted size of β-actin.

Figure 2.

Examples of COMT and actin immunoblots in 2 different subjects. Immunoblotting with an anti-COMT antibody produced 4 major bands at 39, 32, 26, and 24 kDa (see left-hand panel). The 26-kDa band was not visible in some individuals (see example in right-hand panel). Immunoblotting with an antiactin antibody produced a single band at 50 kDa, the predicted size of β-actin.

There were significant correlations between COMT enzyme activity and immunoreactivity of the 32-kDa (r = 0.51, P < 0.001; Fig. 3b) and 24-kDa (r = 0.61, P < 0.001; Fig. 3d) bands, normalized to β-actin, and a trend correlation between COMT enzyme activity and immunoreactivity of the 39-kDa band (r = 0.28, P < 0.1; Fig. 3a), normalized to β-actin. However, COMT enzyme activity significantly correlates neither with immunoreactivity of the 26-kDa band (r = 0.21; Fig. 3c) normalized to β-actin nor with β-actin immunoreactivity (r = 0.16; data not shown). The positive correlations between COMT enzyme activity and immunoreactivity were reflected as significant correlations between normalized immunoreactivity and age for the 32-kDa (r = 0.42, P < 0.01) and 24-kDa (r = 0.36, P < 0.05) bands, but not for the 39-kDa (r = 0.19, P > 0.1) or 26-kDa (r = 0.11, P > 0.1), or β-actin (r = −0.014, P > 0.1), bands.

Figure 3.

Correlations between COMT immunoreactive bands and COMT enzyme activity (expressed as % mean adult values). (a) There was a trend level positive correlation between 39-kDa band immunoreactivity and COMT enzyme activity (r = 0.28; P < 0.1). (b) There was a significant positive correlation between immunoreactivity of the 32-kDa band and COMT enzyme activity (r = 0.51; P < 0.001). (c) There was no significant correlation between 26-kDa band immunoreactivity and COMT enzyme activity (r = 0.21). (d) There was a significant positive correlation between 24-kDa band immunoreactivity and COMT enzyme activity (r = 0.61; P < 0.001).

Figure 3.

Correlations between COMT immunoreactive bands and COMT enzyme activity (expressed as % mean adult values). (a) There was a trend level positive correlation between 39-kDa band immunoreactivity and COMT enzyme activity (r = 0.28; P < 0.1). (b) There was a significant positive correlation between immunoreactivity of the 32-kDa band and COMT enzyme activity (r = 0.51; P < 0.001). (c) There was no significant correlation between 26-kDa band immunoreactivity and COMT enzyme activity (r = 0.21). (d) There was a significant positive correlation between 24-kDa band immunoreactivity and COMT enzyme activity (r = 0.61; P < 0.001).

We also examined correlations between the immunoreactivity of the 39-kDa band expressed relative to the total MB-COMT (32 + 39 kDa bands) because we have demonstrated this measure to be altered in schizophrenia and bipolar disorder (Tunbridge, Weinberger, and Harrison 2006). There was a significant negative correlation between this measure and COMT enzyme activity (r = −0.536, P > 0.001), which was also reflected as a significant negative correlation with age (r = −0.450, P > 0.005). These data are discussed in more detail in the supplementary information. Protein immunoreactivity of the different bands broken down by group is shown in the supplementary information table.

Effect of Val158Met Genotype on COMT Activity and Protein Immunoreactivity

There was no significant difference in COMT enzyme activity between Val158 homozygotes and Val158Met heterozygotes (t = 0.90, P > 0.1; Table 3), and this remained nonsignificant when age was included as a covariate (F1,31 = 0.02, P > 0.1; see Fig. 4 for COMT activity in different genotype groups). In contrast, there was a significant difference between genotype groups for immunoreactivity of the 32-kDa band (t = 2.6, P < 0.05; Table 3) and a trend difference for the immunoreactivity of the 39-kDa band (t = 1.8, P < 0.1; Table 3), with Val158 homozygotes showing significantly greater immunoreactivity than heterozygotes. Furthermore, there was a significant effect of genotype on the immunoreactivity of the 39-kDa band, expressed relative to the total MB-COMT (32 + 39 kDa bands; t = −2.9, P < 0.01; Table 3; of interest given that it replicates our recent findings in postmortem PFC tissue taken from psychiatric patients and controls [Tunbridge, Weinberger, and Harrison 2006]; see supplementary online information). However, there were no significant genotype group differences for immunoreactivity of either the 26-kDa (t = 0.40; Table 3) or the 24-kDa (t = 1.63; Table 3) COMT bands, or β-actin (t = 0.13; Table 3).

Table 3

Effect of Val158Met genotype on enzyme activity and western blot variables

 Val/Val (n = 22) Val/Met (n = 17) Met/Met (n = 1) 
COMT enzyme activitya 69.6 ± 5.398 67.34 ± 6.906 44.704 
39-kDa bandb 1.84 ± 0.089 1.64 ± 0.114† 2.288 
32-kDa bandb 1.164 ± 0.083 0.895 ± 0.107* 0.619 
26-kDa bandb 0.198 ± 0.049 0.171 ± 0.063 0.812 
24-kDa bandb 0.404 ± 0.034 0.297 ± 0.044 0.305 
Actinc 4867 ± 234 4775 ± 300 3524 
Relative 39-kDa band immunoreactivityd 0.624 ± 0.014 0.676 ± 0.018** 0.777 
 Val/Val (n = 22) Val/Met (n = 17) Met/Met (n = 1) 
COMT enzyme activitya 69.6 ± 5.398 67.34 ± 6.906 44.704 
39-kDa bandb 1.84 ± 0.089 1.64 ± 0.114† 2.288 
32-kDa bandb 1.164 ± 0.083 0.895 ± 0.107* 0.619 
26-kDa bandb 0.198 ± 0.049 0.171 ± 0.063 0.812 
24-kDa bandb 0.404 ± 0.034 0.297 ± 0.044 0.305 
Actinc 4867 ± 234 4775 ± 300 3524 
Relative 39-kDa band immunoreactivityd 0.624 ± 0.014 0.676 ± 0.018** 0.777 

Note: Values are means ± standard error of the mean.

a

Expressed as percent mean adult value.

b

Expressed relative to actin expression.

c

Expressed as National Institutes of Health Image profile plot units.

d

Immunoreactivity of 39-kDa band relative to the total of 39 + 32 kDa bands.

Differs from Val/Val genotype group (P < 0.1).

*

Differs from Val/Val genotype group (P < 0.05).

**

Differs from Val/Val genotype group (P < 0.01).

Figure 4.

COMT enzyme activity increases in dorsolateral PFC during human postnatal development in both Val158Met genotype groups. COMT enzyme activity (expressed as a percentage of the mean adult value) increased with age in both Val/Val (indicated with squares) and Val/Met (indicated with circles) genotype groups. The single Met/Met individual is indicated with a triangle (young adult group).

Figure 4.

COMT enzyme activity increases in dorsolateral PFC during human postnatal development in both Val158Met genotype groups. COMT enzyme activity (expressed as a percentage of the mean adult value) increased with age in both Val/Val (indicated with squares) and Val/Met (indicated with circles) genotype groups. The single Met/Met individual is indicated with a triangle (young adult group).

Discussion

The main finding of this study is that there is a significant increase in COMT enzyme activity in human dorsolateral PFC during postnatal development, which continues into adulthood. Although this study lacks the power to statistically investigate interactions between age group and Val158Met genotype, this increase in COMT enzyme activity with age is unlikely to result from a chance difference in Val158Met genotype frequencies between age groups because the genotypes are fairly evenly distributed among the groups and an increase in COMT enzyme activity with age is seen in both Val158/Val158 and Val158/Met158 genotype groups (Fig. 4). Additionally, we demonstrate significant correlations between COMT enzyme activity and immunoreactivity of COMT protein isoforms and replicate our previous finding of an additional 39-kDa COMT immunoreactive band (Tunbridge, Weinberger, and Harrison 2006; see supplementary information).

These data show that, like other PFC dopaminergic markers (e.g., Lidow and Rakic 1992; Rosenberg and Lewis 1995; Lewis and others 1998; Lambe and others 2000; Levitt 2003; Weickert, M.J. Webster, P. Gondipalli, D. Rothmond, R.J. Fatula, M.M. Herman, J.E. Kleinman, and M. Akil, unpublished data), PFC COMT is variable depending on an individual's age and increases over a protracted postnatal period. However, in contrast to data obtained for other PFC dopaminergic markers (see Introduction and discussed further below), COMT enzyme activity continues to increase through adolescence, peaking only in adulthood, as demonstrated by our finding of a significant difference in this measure between the adult group and teen and young adult groups.

Our data are consistent with the limited available data on COMT activity during postnatal maturation in other tissues and species. In human liver, COMT activity increases postnatally, peaking in adulthood (Agathopolous and others 1971). Similarly, porcine COMT activity has been shown to increase after birth in several tissue types, including brain (Stanton and others 1975). Brust and others (2004) calculated that COMT activity increased in frontal cortex, but not in striatum or mesencephalon, in young adult pigs compared with newborns. However, these findings were extrapolated from positron emission tomography of 18F-fluorodopa utilization, rather than direct measurement of COMT enzyme activity. Interestingly, our findings extend those of Chen and others (2004) who found a weak but positive correlation between age and COMT enzyme activity in dorsolateral PFC in an adult cohort (mean age ± standard deviation: 40.6 ± 14.9 years, r = 0.194).

The increase in COMT activity during postnatal maturation is perhaps related, at least in part, to the general expansion and refinement of the PFC dopamine system that occurs over this period (e.g., Lambe and others 2000). That is, the increased PFC COMT activity from neonates to adulthood might reflect an increased need for dopamine catabolism, accompanying the expansion in dopaminergic innervation that occurs during adolescence. In this respect, increasing COMT expression and activity may be part of the developmental homeostasis that regulates dopamine function in the PFC during this critical postnatal developmental period, in which prefrontal function is maturing, and reward and learning mechanisms subserved by dopamine are also being elaborated and refined. However, our data suggest that the maturational elevation of COMT activity is more prolonged than other aspects of the dopaminergic system. This is in clear contrast to the postnatal profiles of other markers of dopamine function within the primate PFC. Thus, dopaminergic innervation of the monkey PFC peaks during adolescence (Rosenberg and Lewis 1995; Lewis 1997; Lambe and others 2000), and adult levels of the dopamine receptors are obtained by the onset of adolescence, following a peak early in postnatal development (Lidow and Rakic 1992). Similarly, there is no change in monoamine oxidase-A, or monoamine oxidase-B, activity in the human frontal cortex during the period between childhood and adulthood (Kornhuber and others 1989). Finally, although postnatal changes of levels in dopamine and its metabolites have not been systematically studied into full adulthood in a single human cohort, the available data from studies of monkey and human brain suggest that adult levels are reached during adolescence (Adolfsson and others 1979; Goldman-Rakic and Brown 1992). Thus, the increase in PFC COMT activity from the young adult period to adulthood demonstrated in this study stands in contrast to data obtained for a range of other dopaminergic markers within the PFC, which reach adult levels during or before adolescence.

Evidence that the functional role of COMT in PFC changes during adolescence comes from recent data obtained in subjects with velocardiofacial syndrome (VCFS, also known as DiGeorge syndrome; Shprintzen and others 1978). VCFS is a developmental disorder resulting from hemizygous deletion of chromosome 22q11, with the COMT locus comprising part of the critically deleted region; thus, individuals with VCFS are hemizygous for the Val158Met polymorphism. Of interest to the present study is the fact that the relationship between the Val158Met allele and cognition in VCFS subjects is critically dependent on the age of the individual (Gothelf and others 2005). Thus, prior to adolescence (11–13 years), Met158 hemizygotes perform significantly better than Val158 hemizygotes on tests of cognition (Bearden and others 2004; Gothelf and others 2005), similar to findings in normal adults (e.g., Egan and others 2001), whereas in late adolescence (16–18 years) Val158 hemizygotes outperform Met158 hemizygotes (Baker and others 2005; Gothelf and others 2005), due to a significant deterioration in performance in the Met158 hemizygote group (Gothelf and others 2005). The better PFC performance of Val158 hemizygotes compared with Met158 hemizygotes in the older VCFS cohorts is consistent with the suggested inverted-U–shaped relationship between PFC dopamine levels and PFC function (Goldman-Rakic and others 2000, Mattay and others 2003) because the presence of the Met158 allele on a hemizygous background combined with age-related increases in dopamine signaling likely represents suboptimal COMT function and superoptimal dopamine levels, compared with the more optimal dopamine levels found in Val158 hemizygotes (for further discussion, see Tunbridge, Harrison, and Weinberger 2006). It might be envisaged that the changing relationship between the COMT Val158Met polymorphism and PFC function in VCFS is related to fluctuations in the global dopaminergic state of the PFC at different postnatal developmental time points, determined by COMT enzyme activity and other dopaminergic markers. Such fluctuations might be expected to result from alterations in the balance between dopamine innervation and catabolism, as these opposing actions concurrently increase during PFC maturation.

It is important to note that the increase in COMT activity may, in part, be related to its methylation of substrates other than the catecholamines, such as catecholestrogens (Männistö and Kaakkola 1998). Certain catecholestrogens are toxic, and, therefore, the upregulation of COMT activity during ontogenesis may represent an attempt to metabolize any accumulation of these compounds. Catecholestrogens are bioactive, for example, in the periphery they modulate angiogenesis (Zhu and Conney 1998). However, although they are known to be present in the cortex, their function in brain remains unknown (Zhu and Conney 1998).

Western blotting using a previously described antibody (Chen and others 2004; Tunbridge, Weinberger, and Harrison 2006) identified several COMT immunoreactive bands. The band at approximately 32 kDa is of the anticipated size for MB-COMT, and we recently described the 39-kDa isoform and demonstrated that it is a COMT allozyme (Tunbridge, Weinberger, and Harrison 2006). Thus, based on their size and abundance relative to the lower bands (e.g., Chen and others 2004), we believe the 32- and 39-kDa bands to be isoforms of MB-COMT. This interpretation is consistent with the data above, showing their abundance to be correlated with COMT enzyme activity. In our previous study (Tunbridge, Weinberger, and Harrison 2006), S-COMT was expressed at too low an abundance to be reliably quantified; thus, our current report of a doublet of COMT immunoreactive bands at 24 and 26 kDa, approximately the reported size of S-COMT (25 kDa), represents a novel finding. It is currently not clear whether both of these bands are isoforms of S-COMT, breakdown products of the larger COMT isoforms, or whether one or both represents nonspecific antibody binding. However, the significant correlation between the abundance of the 24-kDa band and COMT enzyme activity supports the interpretation that the 24-kDa band is a COMT allozyme. Thus, correlations between COMT enzyme activity and western blotting results suggest that both MB-COMT and S-COMT allozymes contribute to measured COMT enzyme activity.

Although COMT enzyme activity was higher in the Val/Val homozygote group than the Val/Met heterozygote group, we found no significant effect of the Val158Met polymorphism on this measure in the current samples of human brain tissue. This is likely due primarily to the absence of a Met/Met homozygote group in combination with the relatively small sample size, confounded by the increased variance within the genotype groups due to the large age-related increase in COMT activity and the fact that the Val158Met polymorphism is not the only locus within the COMT gene that influences enzyme activity (Chen and others 2004). However, the Val158Met polymorphism was related to the abundance of the 32- and 39-kDa immunoreactive bands, in agreement with the destabilizing effect of the Met158 allele on COMT protein shown in previous studies (Shield and others 2003; Chen and others 2004). Furthermore, we replicated our finding of an association between the expression of the 39-kDa band, relative to total presumed MB-COMT immunoreactivity, and the Val158Met polymorphism (see supplementary online material; Tunbridge, Weinberger, and Harrison 2006). However, we found no significant effect of Val158Met genotype on expression of the 24- or 26-kDa bands, possibly due to the small sample size and their relatively lower abundance. Thus, although Val158Met genotype affected abundance of COMT immunoreactive bands as previously described (Shield and others 2003; Chen and others 2004), this failed to manifest as a significant effect on enzyme activity in this study.

In conclusion, we demonstrate that COMT enzyme activity increases during human postnatal development in the PFC, likely as the result of increased COMT protein. These results complement previous data showing a protracted development of the PFC dopamine system in primates.

Supplementary Material

Supplementary material can be found at: http://www.cercor.oxfordjournals.org/.

EMT is supported by an Advanced Track Scholarship from the National Institutes of Health/Oxford Graduate Partnership Program, and a Pathfinder/Brain Sciences award from the UK Medical Research Council. CSW, JEK, MMH, JC, BSK, and DRW are supported by the National Institute of Mental Health Intramural Research Program. Work in PJH's laboratory is supported by the Medical Research Council and the Stanley Medical Research Institute. The authors are grateful to T. Hyde for preparation of brain tissue and to S. Beltaifa, H. Cassano, R. Fatula, S. Lauderdale, D. Rothmond, and Y. Snitkovsky for technical assistance. We are indebted to the families of the deceased. Conflicts of Interest: None declared.

References

Adolfsson
R
Gottfries
CG
Roos
BE
Winblad
B
Post-mortem distribution of dopamine and homovanillic acid in human brain, variations related to age, and a review of the literature
J Neural Transm
 , 
1979
, vol. 
45
 (pg. 
81
-
105
)
Agathopolous
A
Nicolopoulos
D
Matsaniotis
N
Papadatos
C
Biochemical changes of catechol-o-methyltransferase during development of human liver
Pediatrics
 , 
1971
, vol. 
47
 (pg. 
125
-
128
)
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
, vol. 
58
 (pg. 
23
-
31
)
Bearden
CE
Jawad
AF
Lynch
DR
Sokol
S
Kanes
SJ
McDonald-McGinn
DM
Saitta
SC
Harris
SE
Moss
E
Wang
PP
Zackai
E
Emanuel
BS
Simon
TJ
Effects of a functional COMT polymorphism on prefrontal cognitive function in patients with 22q11.2 deletion syndrome
Am J Psychiatry
 , 
2004
, vol. 
161
 (pg. 
1700
-
1702
)
Bilder
RM
Volavka
J
Czobor
P
Malhotra
AK
Kennedy
JL
Ni
X
Goldman
RS
Hoptman
MJ
Sheitman
B
Lindenmayer
JP
Citrome
L
McEvoy
JP
Kunz
M
Chakos
M
Cooper
TB
Lieberman
JA
Neurocognitive correlates of the COMT Val(158)Met polymorphism in chronic schizophrenia
Biol Psychiatry
 , 
2002
, vol. 
52
 (pg. 
701
-
707
)
Birrell
JM
Brown
VJ
Medial frontal cortex mediates perceptual attentional set shifting in the rat
J Neurosci
 , 
2000
, vol. 
20
 (pg. 
4320
-
4324
)
Brust
P
Walter
B
Hinz
R
Fuchtner
F
Muller
M
Steinbach
J
Bauer
R
Developmental changes in the activities of aromatic amino acid decarboxylase and catechol-O-methyl transferase in the porcine brain: a positron emission tomography study
Neurosci Lett
 , 
2004
, vol. 
364
 (pg. 
159
-
163
)
Chen
J
Lipska
BK
Halim
N
Ma
QD
Matsumoto
M
Melhem
S
Kolachana
BS
Hyde
TM
Herman
MM
Apud
J
Egan
MF
Kleinman
JE
Weinberger
DR
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
, vol. 
75
 (pg. 
807
-
821
)
Dias
R
Robbins
TW
Roberts
AC
Dissociation in prefrontal cortex of affective and attentional shifts
Nature
 , 
1996
, vol. 
380
 (pg. 
69
-
72
)
Egan
MF
Goldberg
TE
Kolachana
BS
Callicott
JH
Mazzanti
CM
Straub
RE
Goldman
D
Weinberger
DR
Effect of COMT Val108/158 Met genotype on frontal lobe function and risk for schizophrenia
Proc Natl Acad Sci USA
 , 
2001
, vol. 
98
 (pg. 
6917
-
6922
)
Gaspirini
M
Fabrizio
E
Bonifati
V
Meco
G
Cognitive improvement during Tolcapone treatment in Parkinson's disease
J Neural Transm
 , 
1997
, vol. 
104
 (pg. 
887
-
894
)
Gogos
JA
Morgan
M
Luine
V
Ogawa
S
Pfaff
D
Karayiourgou
M
Catechol-o-methyltransferase-deficient mice exhibit sexually dimorphic changes in catecholamine levels and behaviour
Proc Natl Acad Sci USA
 , 
1998
, vol. 
95
 (pg. 
9991
-
9996
)
Goldman-Rakic
P
Brown
R
Postnatal development of monoamine content and synthesis in the cerebral cortex of rhesus monkeys
Dev Brain Res
 , 
1992
, vol. 
4
 (pg. 
339
-
349
)
Goldman-Rakic
PS
Muly
EC
Williams
GV
D1 receptors in prefrontal cells and circuits
Brain Res Rev
 , 
2000
, vol. 
31
 (pg. 
295
-
301
)
Gothelf
D
Eliez
S
Thompson
T
Hinard
C
Penniman
L
Feinstein
C
Kwon
H
Jin
S
Jo
B
Antonarakis
SE
Morris
MA
Reiss
AL
COMT genotype predicts longitudinal cognitive decline and psychosis in 22q11.2 deletion syndrome
Nat Neurosci
 , 
2005
, vol. 
8
 (pg. 
1500
-
1502
)
Joober
R
Gauthier
J
Lal
S
Bloom
D
Lalonde
P
Rouleau
G
Benkelfat
C
Labelle
A
Catechol-o-methyltransferase Val-108/158-Met gene variants associated with performance on the Wisconsin Card Sorting Test
Arch Gen Psychiatry
 , 
2002
, vol. 
59
 (pg. 
662
-
663
)
Karoum
F
Chrapusta
SJ
Egan
MJ
3-methoxytyramine is the major metabolite of released dopamine in the rat frontal cortex: reassessment of the effects of antipsychotics on the dynamics of dopamine release and metabolism in the frontal cortex, nucleus accumbens and striatum by a simple two-pool model
J Neurochem
 , 
1994
, vol. 
63
 (pg. 
972
-
979
)
Kleinman
JE
Hyde
TM
Herman
MM
Bloom
FE
Kupfer
DJ
Methodological issues in the neuropathology of mental illness
Psychopharmcology: the fourth generation of progress
 , 
1995
New York
Raven Press
(pg. 
859
-
864
)
Kornhuber
J
Konradi
C
Mack-Burkhardt
F
Riederer
P
Heinsen
H
Beckmann
H
Ontogenesis of monoamine oxidase-A and -B in the human brain frontal cortex
Brain Res
 , 
1989
, vol. 
499
 (pg. 
81
-
86
)
Lambe
EK
Krimer
LS
Goldman-Rakic
PS
Differential postnatal development of catecholamine and serotonin inputs to identified neurons in prefrontal cortex of rhesus monkey
J Neurosci
 , 
2000
, vol. 
20
 (pg. 
8780
-
8787
)
Levitt
P
Structural and functional maturation of the developing primate brain
J Pediatr
 , 
2003
, vol. 
143
 
4 Suppl
(pg. 
S35
-
S45
)
Lewis
DA
Development of the prefrontal cortex during adolescence: insights into vulnerable neural circuits in schizophrenia
Neuropsychopharmacology
 , 
1997
, vol. 
16
 (pg. 
385
-
398
)
Lewis
DA
Melchitzky
DS
Sesack
SR
Whitehead
RE
Auh
S
Sampson
A
Dopamine transporter immunoreactivity in monkey cerebral cortex: regional, laminar, and ultrastructural localization
J Comp Neurol
 , 
2001
, vol. 
432
 (pg. 
119
-
136
)
Lewis
DA
Sesack
SR
Levey
AI
Rosenberg
DR
Dopamine axons in primate prefrontal cortex: specificity of distribution, synaptic targets, and development
Adv Pharmacol
 , 
1998
, vol. 
42
 (pg. 
703
-
706
)
Lidow
MS
Rakic
P
Scheduling of monoaminergic neurotransmitter receptor expression in the primate neocortex during postnatal development
Cereb Cortex
 , 
1992
, vol. 
2
 (pg. 
401
-
416
)
Luna
B
Garver
KE
Urban
TA
Lazar
NA
Sweeney
JA
Maturation of cognitive processes from late childhood to adulthood
Child Dev
 , 
2004
, vol. 
75
 (pg. 
1357
-
1372
)
Malholtra
AK
Kestler
LJ
Mazzanti
C
Bates
JA
Goldberg
T
Goldman
D
A functional polymorphism in the COMT gene and performance on a test of prefrontal cognition
Am J Psychiatry
 , 
2002
, vol. 
159
 (pg. 
652
-
654
)
Männistö
PT
Kaakkola
S
Catechol-o-methyltransferase (COMT): biochemistry, molecular biology, pharmacology and clinical efficacy of the new selective COMT inhibitors
Pharmacol Rev
 , 
1999
, vol. 
51
 (pg. 
593
-
628
)
Mattay
VS
Apud
JA
Rasetti
R
Cayenne
C
Das
S
Alce
G
Callicott
J
Goldberg
T
Egan
M
Weinberger
DR
Tolcapone enhances prefrontal function: an fMRI study in normal healthy volunteers. Program No. 81.15, 2004 Abstract Viewer/Itinerary Planner
 , 
2004
Washington, DC
Society for Neuroscience
 
Available from: http://www.sfn.org. 15 October 2004.
Palmatier
MA
Pakstis
AJ
Speed
W
Paschou
P
Goldman
D
Odunsi
A
Okonofua
F
Kajuna
S
Karoma
N
Kungulilo
S
Grigorenko
E
Zhukova
OV
Bonne-Tamir
B
Lu
RB
Parnas
J
Kidd
JR
DeMille
MM
Kidd
KK
COMT haplotypes suggest P2 promoter region relevance for schizophrenia
Mol Psychiatry
 , 
2004
, vol. 
9
 (pg. 
859
-
870
)
Rosenberg
DR
Lewis
DA
Postnatal maturation of the dopaminergic innervation of monkey prefrontal and motor cortices: a tyrosine hydroxylase immunohistochemical analysis
J Comp Neurol
 , 
1995
, vol. 
358
 (pg. 
383
-
400
)
Shifman
S
Bronstein
M
Sternfeld
M
Pisante-Shalom
A
Lev-Lehman
E
Weizman
A
Reznik
I
Spivak
B
Grisaru
N
Karp
L
Schiffer
R
Kotler
M
Strous
RD
Swartz-Vanetik
M
Knobler
HY
Shinar
E
Beckmann
JS
Yakir
B
Risch
N
Zak
NB
Darvasi
A
A highly significant association between a COMT haplotype and schizophrenia
Am J Hum Genet
 , 
2002
, vol. 
71
 (pg. 
1296
-
1302
)
Shield
AJ
Thomae
BA
Eckloff
BW
Wieben
ED
Weinshilboum
RM
Human catechol O-methyltransferase genetic variation: gene resequencing and functional characterization of variant allozymes
Mol Psychiatry
 , 
2003
, vol. 
9
 (pg. 
151
-
160
)
Shprintzen
RJ
Goldberg
RB
Lewin
ML
Sidoti
EJ
Berkman
MD
Argamaso
RV
Young
D
A new syndrome involving cleft palate, cardiac anomalies, typical facies, and learning disabilities: velo-cardio-facial syndrome
Cleft Palate J
 , 
1978
, vol. 
5
 (pg. 
56
-
62
)
Stanton
HC
Cornejeo
RA
Mersmann
HJ
Brown
LJ
Mueller
RL
Ontogenesis of monoamine oxidase and catechol-O-methyl transferase in various tissues of domestic swine
Arch Int Pharmacodyn Ther
 , 
1975
, vol. 
213
 (pg. 
128
-
144
)
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
, vol. 
24
 (pg. 
5331
-
5335
)
Tunbridge
EM
Harrison
PJ
Weinberger
DR
Catechol-o-methyltransferase, cognition and psychosis: Val158Met and beyond
Biol Psychiatry
 , 
2006
 
(doi:10.1016/j.biopsych.2005.10.024).
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
, vol. 
11
 (pg. 
116
-
117
)
Weinberger
DR
Implications of normal brain development for the pathogenesis of schizophrenia
Arch Gen Psychiatry
 , 
1987
, vol. 
44
 (pg. 
660
-
999
)
Weinberger
DR
Marenco
S
Hirsch
SR
Weinberger
DR
Schizophrenia as a neurodevelopmental disorder
Schizophrenia
 , 
2003
2nd ed
MA
Blackwell Publishing, Malden, Massachusetts
(pg. 
326
-
348
)
Zhu
BT
Conney
AH
Functional role of estrogen metabolism in target cells: review and perspectives
Carcinogenesis
 , 
1998
, vol. 
19
 (pg. 
1
-
27
)