J. Neurochem. (2010) 114, 1745–1755.
Catechol-O-methyltransferase (COMT) plays an active role in the metabolism of dopamine (DA) in the prefrontal cortex (PFC). Because of low levels of dopamine transporter (DAT), it is proposed that the majority of released DA is taken up by either norepinephrine transporter (NET) and subsequently metabolized by monoamine oxidize (MAO) or by uptake2 (to glial cells and post-synaptic neurons) and metabolized by COMT. However, a comprehensive in vivo study of rating the mechanisms involved in DA clearance in the PFC has not been done. Here, we employ two types of microdialysis to study these pathways using DAT, NET and MAO blockers in conscious mice, with or without Comt gene disruption. In quantitative no-net-flux microdialysis, DA levels were increased by 60% in the PFC of COMT-knockout (ko) mice, but not in the striatum and nucleus accumbens. In conventional microdialysis studies, we showed that selective NET and MAO inhibition increased DA levels in the PFC of wild-type mice by two- to fourfold, an effect that was still doubled in COMT-ko mice. Inhibition of DAT had no effect on DA levels in either genotype. Therefore, we conclude that in the mouse, PFC COMT contributes about one half of the total DA clearance.
area under the curve
Catechol-O-methyltransferase (COMT) catalyses the transfer of a methyl group from S-adenosyl-l-methionine to one of the two hydroxyl groups of catecholic compounds, including l-dopa, catecholestrogens, endogenous and exogenous catecholamines and their hydroxylated metabolites (Guldberg and Marsden 1975; Männistö and Kaakkola 1999). COMT is strictly an intracellular enzyme that appears as soluble-COMT and membrane-bound-COMT isoforms, both of which are products of the same gene (Salminen et al. 1990; Lundström et al. 1991). In the brain, COMT has been localized to post-synaptic neurons and glial cells (Rivett et al. 1983; Kaakkola et al. 1987; Karhunen et al. 1995a,b). Notably, uptake2 is needed to transport dopamine (DA) into these cells (Wilson et al. 1988; Trendelenburg 1990; Männistöet al. 1992).
In the striatum, uptake1 mediated by a dopamine transporter (DAT) is the primary way of terminating the DA signal (Cass et al. 1993; Giros et al. 1996; Eisenhofer et al. 2004). DAT rapidly uptakes most of the released DA into the pre-synaptic neuron after which it is either packed to storage vesicles by vesicular monoamine transporter 2 or metabolized to 3,4-dihydroxyphenylacetic acid (DOPAC) by monoamine oxidase (MAO). In contrast, in brain areas with low DAT density (Sesack et al. 1998), e.g. in the prefrontal cortex (PFC), metabolism by COMT (Matsumoto et al. 2003) and uptake by the norepinephrine transporter (NET) (Carboni et al. 1990; Di Chiara et al. 1992; Tanda et al. 1994, 1997; Yamamoto and Novotney 1998; Mundorf et al. 2001; Mazei et al. 2002; Morón et al. 2002) appear to have a more important role in the control of dopaminergic transmission. This is supported also by the fact that NET has a stronger affinity for DA than DAT (Giros et al. 1994; Gu et al. 1994; Eshleman et al. 1999). However, very little is known of the relative magnitude of these effects. Previous studies in DAT-knockout (ko) mice have confirmed the negligible effect of DAT deficiency on DA levels in the PFC, whereas in the striatum the levels were indeed increased (Shen et al. 2004). With regards to COMT, it has been confirmed that 3-methoxytyramine (3-MT), the DA metabolite produced by COMT in the CNS, comprises more than 60% of the total basal DA turnover in the rat PFC, whereas in the striatum and nucleus accumbens 3-MT accounts for < 15% of basal DA turnover (Karoum et al. 1994). Also, it was previously shown in vivo that COMT-ko mice have a twofold longer DA elimination time in the PFC than wild-type (wt) animals, whereas the rapid DA clearance observed in the striatum is not affected by COMT deficiency (Yavich et al. 2007). Under baseline conditions, Comt gene disruption does not affect the DA levels in striatal, cortical or hypothalamic tissues, despite considerably elevated DOPAC levels (Huotari et al. 2002a).
A number of studies on the effect of COMT on DA metabolism have been conducted before (Huotari et al. 2002a; Tunbridge et al. 2004; Yavich et al. 2007; Lapish et al. 2009), but none of them have attempted a comprehensive in vivo survey of the importance of COMT activity on DA clearance in the PFC. We have now employed two types of microdialysis techniques to clarify the pathways of DA clearance in conscious mice carrying a mutation in the Comt gene as well as their wt littermates. Our results provide unequivocal converging evidence for the central role of COMT in DA clearance in the mouse PFC.
Materials and methods
Homozygous male and female COMT-ko mice originally generated by Gogos et al. (1998) and their wt littermates, aged 2–6 months were used. The mice were bred in the Laboratory Animal Centre in University of Helsinki, Finland. Genotyping of the COMT-ko mice was performed as described elsewhere (Tammimäki et al. 2008). The mice were housed individually at an ambient temperature of 21–23°C under 12 : 12 h light cycle with free and continuous access to food pellets and drinking fluid. The oestrus phase was not determined. All procedures with animals were performed according to European Community Guidelines for the use of experimental animals (European Communities Council Directive 86/609/EEC) and the experimental setup was reviewed by State Provincial Office of Southern Finland and approved by the Animal Experiment Board in conformance with the current legislation.
Tissue preparation, immunofluorescence and immunohistochemistry
The mice were implanted with guide cannulas (MAB-4; Agn Tho’s AB, Lidingö, Sweden) under isoflurane anaesthesia and were given a buprenorphine injection (0.05 mg/kg subcutaneously) for pain relief 30 min before and 12 h after the operation. The coordinates for guide cannulas were calculated relative to bregma. They were aimed at the point above the nucleus accumbens (A/P = + 1.4, L/M = + 0.9, D/V = −3.8), the dorsal striatum (A/P = + 0.6, L/M = + 1.8, D/V = −2.2), or the medial PFC (A/P = + 2.0, L/M = + 0.5, D/V = −1.0; Fig. 1) according to the mouse brain atlas by Franklin and Paxinos (1997). Because of the small size of the mouse nucleus accumbens, the probes were collecting the fluid from both core and shell compartments. The cannula was attached to the scull with dental cement (Aqualox, Voco, Germany) and two stainless steel screws. After the surgery, the animals were individually housed into test cages (30 × 30 × 40 cm) and allowed to recover for 5–7 days before the experiment.
In the morning of the experiment day, a microdialysis probe (MAB-4; Agn Tho’s AB; membrane length 1 mm for nucleus accumbens and dorsal striatum and 2 mm for the PFC, outer diameter 0.2 mm) was inserted into the guide cannula, and the probe was infused with a modified Ringer solution (147 mM NaCl, 1.2 mM CaCl2, 2.7 mM KCl, 1.0 mM MgCl2) at a flow rate of 0.6 μL/min. Three baseline samples (every 30 min) were collected after the 3-h stabilization period. Thereafter, four concentrations of DA in Ringer solution (Cin; 0, 2, 10 and 20 nM) were perfused through the probes in a random order. Following a 30-min equilibration period, two 30-min samples were collected at each Cin for HPLC analysis. At the end of the experiment, the animals were decapitated, their brains were removed from the skull and frozen rapidly on dry ice. The positions of the microdialysis probes were verified histologically from brain slices prepared post-mortem.
In the morning of the experiment day a microdialysis probe (MAB-4; Agn Tho’s AB; membrane length 1 mm for nucleus accumbens and dorsal striatum and 2 mm for the PFC, outer diameter 0.2 mm) was inserted into the guide cannula, and the probe was infused with a modified Ringer solution (147 mM NaCl, 1.2 mM CaCl2, 2.7 mM KCl, 1.0 mM MgCl2) at a flow rate of 2 μL/min. Baseline samples (every 20 min) were collected after the 2-h stabilization period. Following collection of 3–4 stable baseline samples, the mice that were scheduled for DAT, NET or MAO inhibition received either: (i) GBR 12909 (20 mg/kg; Tocris bioscience, Bristol, UK) for DAT inhibition, (ii) reboxetine (20 mg/kg; Toronto Research Chemicals, North York, Ontario, Canada) for NET inhibition or (iii) pargyline (100 mg/kg; Sigma Aldrich, St. Louis, MO, USA) for MAO A/B inhibition. The doses were selected according to previous experiments where clear effects on DA levels or behaviour in mice or rats was demonstrated (pargyline: Sharp et al. 1986; GBR 12909: Ichihara et al. 1993; Huotari et al. 2002b; reboxetine: Page and Lucki 2002; Dziedzicka-Wasylewska et al. 2006). At the end of the experiment, the animals were decapitated and the probe placement was verified as described above. The results were not corrected for probe recovery.
Quantification of extracellular concentrations DA and DA metabolites
The samples acquired by microdialysis were analysed by an HPLC system coupled with an electrochemical detector. The detailed information is given in the Appendix S1.
Only data from those mice with accurate probe placements were included in the data analysis. In the no-net-flux studies, a linear equation was constructed for each animal by plotting the net flux of DA through the probe (DAin − DAout) against DAin, where DAout is the dialysate DA concentration acquired during the perfusion and DAin is the DA concentration of the perfusion fluid. Based on this equation, the extracellular DA level (DAext) and the in vivo extraction fraction (Ed) were calculated as described by Parsons and Justice (1992). The DAext value stands for the perfusion fluid DA concentration at which there is no net flux of DA through the probe (DAin − DAout = 0). Ed, on the other hand, has been shown to describe combined DAT, NET and uptake2 functions (Justice 1993; Smith and Justice 1994; He and Shippenberg 2000; Chefer et al. 2006).
Two-way anova was used to test the no-net-flux data and two-way repeated measures anova was used to test the conventional microdialysis data with sex and genotype as independent variables. Bonferroni comparisons or contrast analyses were used as post hoc tests where appropriate. Results are given as means ± SEM. Area under the concentration–time curve (AUC) was calculated from our conventional microdialysis results according to the trapezoidal rule for 60–360 min. Analyses were conducted with SPSS Statistics 17.0 statistical analysis software (SPSS Inc., Chicago, IL, USA). Results were considered significant at p < 0.05.
Site-specific presence of DAT and NET in the mouse brain
It has been previously shown that DAT is the dominating transporter in the striatum, whereas NET predominates over DAT in the PFC (Javitch et al. 1984; Scatton et al. 1985; Sesack et al. 1998). We confirmed this pattern of expression in our COMT-ko mice and their wt littermates using immunohistochemistry (Fig. 2). Moreover, we showed that in the COMT-ko mice, there was no up-regulation of DAT or NET in either brain region.
Baseline striatal, accumbal, and prefrontal cortical DA and DA metabolite levels in conventional microdialysis
The basal extracellular DA and DA metabolite levels in the striatum, nucleus accumbens and PFC of COMT-ko and wt mice were assessed by calculating the mean values of 3–4 post-stabilization samples acquired by conventional microdialysis. No sex- or genotype-dependent alterations in DA levels were found (Fig. 3). Accumbal DA levels were 38–56% and the prefrontal cortical levels 3–7% of the concentrations found in the striatum. A lack of COMT induced an increase in the extracellular DOPAC in all brain areas studied (striatum: genotype effect F1, 53 = 17.681, p < 0.001; nucleus accumbens: genotype effect F1,40 = 31.911, p <0.001; and the PFC: genotype effect F1,34 = 49.126, p < 0.001; Fig. 3). DOPAC elevation (five- to sixfold) by COMT deletion was most striking in the PFC. Homovanillic acid (HVA) was not detected in COMT-ko animals in any of the brain areas tested.
Conventional microdialysis in PFC using uptake and MAO blockers
As a result of low levels of DAT in the PFC, it has been proposed that COMT plays a key role in the metabolism of DA in this brain area. However, this prediction has not been tested previously in vivo in conscious, freely moving mice. To address this, we examined how blocking-known mechanisms of DA clearance one-by-one (DAT, NET and MAO vs. COMT) affects levels of DA and its metabolites in the PFC.
In the wt mice, an inhibition of DAT caused no changes in DA levels in the PFC (Fig. 4a and b). Nor did GBR 12909 change DOPAC or HVA levels in either sex (Fig. 4c–f). Instead, both NET (Fig. 5a and b) and MAO (Fig. 6a and b) inhibition elevated the DA levels by two- to fourfold in the PFC. No sex effects were seen. Unexpectedly, NET inhibition did not reduce DOPAC levels at all and also the HVA levels remained unchanged (Fig. 5c–f). Pargyline, on the other hand, effectively decreased DOPAC and HVA to very low levels in both sexes (Fig. 6c–f).
Inhibition of DAT did not affect DA levels in the PFC of COMT-ko mice (Fig 4a and b). However, a lack of COMT had a marginal (up to 15% in AUC60–360min values) DA elevating effect above the values of wt mice. Also the very high DOPAC levels in COMT-ko mice (genotype effect F1,22 = 56.945, p < 0.001) were unaffected by GBR 12909 (Fig. 4c and d). Following NET and MAO inhibition, the elevation in DA levels were nearly doubled by Comt gene disruption [reboxetine: genotype effect F1,25 = 5.998, p < 0.05 (Fig. 5a and b); pargyline: genotype effect F1,20 = 11.084, p < 0.01 (Fig. 6a and b)]. NET inhibition did not further modify DOPAC levels in either genotype and the levels remained higher in COMT-ko mice than in their wt littermates (genotype effect F1,25 = 90.050, p < 0.001; Fig. 5c and d). Although an effective reduction of extracellular DOPAC because of MAO inhibition was seen, also in pargyline treated COMT-ko mice the DOPAC levels were significantly higher than in their wt littermates (genotype effect F1,20 = 63.898, p < 0.001; Fig. 6c and d). Again, no HVA was detected in COMT-ko animals (Figs 4e and f, 5e and f, and 6e and f).
Baseline striatal, accumbal and prefrontal cortical DA levels in no-net-flux microdialysis
Finally, to supplement our conventional microdialysis studies, we wanted to determine the basal extracellular DA levels by a more sophisticated quantitative no-net-flux microdialysis (Parsons and Justice 1992; He and Shippenberg 2000). We found a significant genotype effect with higher DA levels in the medial PFC of COMT-ko mice (F1,27 = 5.898, p < 0.05; Fig. 7c). A lack of COMT did not alter the extracellular DA levels in the two other brain areas studied (Fig. 7a and b). The in vivo extraction fractions remained unaltered in all brain areas (Table 1).
|In vivo extraction fraction (Ed)||Males||Females|
|Wild type||COMT-ko||Wild type||COMT-ko|
|Dorsal striatum||0.31 ± 0.04||0.28 ± 0.02||0.24 ± 0.05||0.36 ± 0.07|
|Nucleus accumbens||0.31 ± 0.05||0.28 ± 0.03||0.33 ± 0.06||0.30 ± 0.04|
|PFC||0.63 ± 0.05||0.69 ± 0.05||0.62 ± 0.08||0.75 ± 0.05|
We made several important findings. First, the NET and MAO inhibition, in contrast to DAT inhibition, increased the prefrontal DA levels several-fold, an effect that was doubled in COMT-ko mice. This provides additional evidence for the pivotal role of the O-methylation route in the PFC when other elements are pharmacologically inhibited. Quite surprisingly, when excluding the function of the pre-synaptic MAO by NET inhibition, the extracellular DOPAC levels did not decrease, predicting an effective compensation by MAO located in post-synaptic neurons and glial cells. Second, although a lack of COMT itself only marginally elevated the prefrontal DA acquired by low recovery conventional microdialysis, our no-net-flux study unequivocally demonstrated that the absolute extracellular DA levels were exclusively elevated in the PFC, but not in the striatum or nucleus accumbens. Elevation of extracellular DA levels in COMT-ko mice has not been shown before and represents a novel finding.
Quantification of various DA clearing mechanisms in the mouse or rat PFC has been done very seldom. Karoum et al. (1994) found that 3-MT, a putative COMT product represented about 60% of the rat PFC DA metabolism, in contrast to < 15% in the striatum. It has been shown that extracellular DA is cleared from the mouse PFC primarily by NET (Carboni et al. 1990; Di Chiara et al. 1992; Tanda et al. 1994, 1997; Yamamoto and Novotney 1998; Mundorf et al. 2001; Mazei et al. 2002). For example, in PFC synaptosomes from NET-ko mice, the DA uptake was 55% lower than that from wt mice (Morón et al. 2002). No such effect was seen in synaptosomes from DAT-ko mice. A low rate of DA uptake allows its diffusion to the surrounding regions, including sites where NET can be found. These finding are not far from our estimates where NET (followed by MAO) is about as important as COMT (preceded by uptake2) in the clearance of DA in the PFC.
Earlier studies in COMT-ko mice have failed to detect significant genotypic differences in the brain tissue DA concentrations under normal conditions (Gogos et al. 1998; Huotari et al. 2002a). Our findings that COMT deficiency does not induce significant changes in striatal and accumbal extracellular DA levels, under baseline conditions in freely moving mice, is in agreement with these and previous findings from microdialysis studies performed under chloral hydrate anaesthesia (Huotari et al. 2004). Furthermore, treatment with the selective COMT inhibitors, tolcapone or entacapone, has not affected the dorsal striatal or accumbal extracellular and tissue DA levels or DA outflow (Acquas et al. 1992; Kaakkola and Wurtman 1992; Li et al. 1998; Budygin et al. 1999; Huotari et al. 1999; Napolitano et al. 2003). Nevertheless, it is noteworthy that although treatment with tolcapone alone was not sufficient to affect extracellular DA in the PFC of rats, it potentiated the clozapine-induced DA efflux (Tunbridge et al. 2004).
Although Comt gene disruption could, in principle, result in the emergence of compensatory mechanisms, MAO activity and MAO protein levels were unaltered in the brain, kidneys and liver of COMT-ko animals (Huotari et al. 2002a; Odlind et al. 2002; Haasio et al. 2003). Similarly, DAT levels and function (Huotari et al. 2002b; Yavich et al. 2007), and also the conjugation of L-dopa (Forsberg et al. 2004), appear to be unaltered in these mice. To further rule out the possibility of compensatory mechanisms, in this report we provide evidence that Comt gene disruption does not change the general pattern of NET and DAT immunoreactivities in the mouse brain. Nevertheless, the possibility of enhanced DA uptake either into non-neuronal cells or post-synaptic neurons by uptake2 or to noradrenergic neurons by NET could not be excluded.
Our conventional microdialysis study revealed the negligible effect of DAT inhibition on DA levels in the PFC of both COMT-ko and wt mice. Although no significant differences were found, the AUC60–360min values of DA levels in COMT-ko animals were approximately 15% higher than in wt mice. In this case, it seems that the action of NET, followed by MAO, is sufficient enough to eliminate synaptic DA in the PFC, and that the roles of DAT and uptake2 are relatively small. However, it has to be pointed out that when assayed under no-net-flux conditions, DA levels in the PFC in contrast to other brain areas were indeed significantly increased by 60% in the COMT-ko mice. In the conventional microdialysis, in vitro recovery estimates (about 10%) of the probes in buffer or saline solution reflect poorly the complex processes of neurotransmitter release, uptake and metabolism associated to the brain tissue. It could well be that relative differences are faded because of a poor and inconsistent probe recovery of already low basal prefrontal DA levels. In the no-net-flux microdialysis, by perfusing known concentrations of the analyte (DA) in vivo through the probe, we are able to generate a series of points that can be interpolated to determine the exact point where the concentration of the analyte is the same in the perfusate as it is in the extracellular fluid (Lönnroth et al. 1987; Parsons and Justice 1992). The latter method can be kept more sensitive and reliable.
The fact that DAT inhibition did not increase DA levels in the PFC is in agreement with similar studies done in rats (Di Chiara et al. 1992; Mazei et al. 2002) and DAT-ko mice (Shen et al. 2004). Following NET and MAO inhibition, on the other hand, prefrontal extracellular DA levels were greatly elevated, an effect that was further increased almost twofold by a lack of COMT. In a sharp contrast, an increase in DA efflux in the striatum caused by MAO inhibition is not enhanced by supplemental COMT inhibition (Tuomainen et al. 1996). Collectively, these results provide a strong experimental evidence for a specific role of COMT in the PFC. Under normal conditions, COMT may account for up to one half of DA metabolism in the PFC, an effect of even a higher magnitude occurring during NET and MAO inhibition.
The accumulation of DOPAC caused by interruption in the O-methylation pathway was clearly seen as elevated levels in our COMT-ko mice. The most accentuated elevation of basal DOPAC levels was seen in the PFC also reflecting the importance of COMT in this area. Interestingly, the inhibition of NET did not affect DOPAC levels in either COMT-ko or wt mice, suggesting that post-synaptic MAO, acting either in glial cells or post-synaptic neurons (after action of uptake2), can effectively take over when pre-synaptic MAO has been excluded. Unfortunately the 3-MT levels in these mice could not be measured with our method. Also conjugation, which may play some role in catecholamine metabolism (Kopin 1984; Buu et al. 1985), could not be quantified in our experimental set-up. Considering the fact that extracellular DA levels (picomolar) are only 1/1000 of the levels of DA metabolites (nanomolar), any alternative pathway would easily prevent DA levels to increase too much. The apparent lack of compensatory mechanisms in COMT-ko mice indicates nonetheless, that DA elimination through any other mechanism cannot be quantitatively very important.
Our findings may have clinical implications especially within psychiatry. COMT activity is genetically polymorphic (Weinshilboum and Raymond 1977; Boudíkováet al. 1990) determined primarily by a common Val108/158Met polymorphism and the thermolability of the 158Met variant (Lotta et al. 1995; Lachman et al. 1998). Although the frequencies of the Val and Met alleles vary fourfold among populations (DeMille et al. 2002), the COMT activity fluctuates much less (ca. 40% according to Chen et al. 2004). The impact of this polymorphic activity on cognition and on susceptibility to psychiatric disorders has recently gained considerable attention (Karayiorgou et al. 1997, 1999; Egan et al. 2001; Gothelf et al. 2005; Mata et al. 2006; Tunbridge et al. 2006; Pooley et al. 2007; Raux et al. 2007). Despite numerous reports, the role of Val108/158Met COMT variant in schizophrenia remains unclear (Munafòet al. 2005) although a recent meta-analysis appears to support a role of Val108/158Met in obsessive-compulsive disorder (Pooley et al. 2007). Another large meta-analysis described a modest association between the Val108/158Met genotype and executive performance in healthy individuals, but not in individuals with schizophrenia (Barnett et al. 2007). It is also noteworthy that a low COMT activity may augment of the effect of olanzapine on working memory in patients with schizophrenia (Bertolino et al. 2004). COMT Val108/158Met polymorphism may also be a risk factor of addictive diseases but its influence may vary depending on the substance, sex and population (Enoch et al. 2006). Because abnormal prefrontal DA metabolism may underlie the risk or expression of a number of human behavioural, cognitive and psychiatric diseases, our results provide a framework for understanding how variation in COMT activity may modulate this risk, expression or both. COMT activity differences based on Val/Met polymorphism may be assumed to cause even more modest changes in DA metabolism than a full lack of COMT. Therefore, the actual importance of Comt gene polymorphisms should be much smaller than anticipated. In fact, the best population studies have reached the same conclusion (see above).
In summary, our results deepen the understanding of DA clearance in the PFC of the mouse. These findings – based on blocking major DA elimination routes one-by-one, but neglecting few routes (conjugations) – suggest that COMT contributes up to one half of the prefrontal DA clearance under basal conditions and even more in absence of NET or MAO. Overall, our results highlight the role of uptake1 by NET (instead of DAT) as a significant route for prefrontal DA clearance but uptake2-dependent mechanisms have a role of a similar magnitude. Our findings also explain why COMT-ko mice have a twofold longer DA elimination time in the PFC than the wt animals, whereas DA clearance in the striatum remains unaltered (Yavich et al. 2007).
The authors want to thank Ms Anna Niemi and Mrs Kati Rautio for excellent technical assistance. These studies were supported by the Academy of Finland (No. 117881/2006 and 1131915/2009), Sigrid Juselius Foundation, and Helsinki University Research Grants to PTM, and The Finnish Parkinson Foundation to MK.