RP-HPLC is a non-radioactive method which does not involve many steps in processing the samples, making it easy to perform and automatize. The coulometric detection system proved to be a reliable method to analyze the COMT reaction products (I). This detection method means that the present set-up is the most sensitive of its kind used in the COMT activity analysis. Thus, it is suitable for the enzyme kinetic analyses and can detect lower COMT activites than previous methods. It allows the analysis of vanillic acid and isovanillic acid, and thus one can calculate the meta/para ratios, from several tissue sources when DHBAc is used as a substrate. Another COMT assay utilizing the described detection system has been developed to be used with dihydroxybenzylamine as a substrate (Ellingson et al. 1999). Since coulometric detection has been used for the analysis of catecholamines and their metabolites (Törnwall et al. 1994), the endogenous cateacholamines should also be applicable as substrates for COMT activity analysis.
The highest variation in COMT activity was found in repeated analysis of apparently similar pieces of tissues obtained from different individuals. The interindividual variation may be due to at least two sources, i.e. COMT assay and genetic differences. COMT assay includes the variation from analytical and sample preparation steps. Previously, we noticed that the variation in specific COMT activity in erythrocytes was mainly affected by the variation in HPLC and protein analysis while the handling of the samples produced less variation (Tuomainen et al. 1996). In the present assay the protein analysis introduces more variation than the enzyme reaction and the HPLC analysis. The genetical variation of COMT activity in humans could be observed by the thermolability of COMT enzyme (Weinshilboum et al. 1999). Although a different level of COMT activity between different inbred rat strains has been described, no thermolability of COMT enzyme has been noticed within a single rat strain (Goldstein et al. 1980; Lotta et al. 1995).
Kinetics. The Km value for recombinant MB-COMT (27.2 mM) obtained (II) was comparable to that observed earlier (22.2 mM, Lotta et al. 1995). Also the higher affinity and lower methylation capacity of recombinant MB-COMT compared to that of recombinant S-COMT is in agreement with the observations with partially purified enzyme preparations. Meta/para ratios for DHBAc with recombinant COMT proteins were similar to those reported earlier (23.7 and 5.1 for MB-COMT and S-COMT, respectively, Nissinen 1984b). The increase of meta/para ratio of recombinant MB-COMT, but not recombinant S-COMT, was also found with declining substrate concentrations as described previously (Nissinen 1984b). The high meta/para ratios with low substrate concentrations resemble the values encountered in vivo.
DHBAc offers some advantages over most endogenic substrates. It is not metabolized further and it is not easily oxidized. The high sensitivity of the detection of reaction products means that one can use a non-saturating concentration of DHBAc. This was achieved in 900 x g supernatant which contains both COMT isoforms. Previously, the activity of the two enzyme forms, MB-COMT and S-COMT, has been analyzed in the same homogenate sample of human brain with a low concentration of dopamine for MB-COMT and a high concentration for S-COMT (Rivett et al. 1983a). Dopamine differs in its affinity and reaction velocity between MB-COMT and S-COMT. Using DHBAc in rat tissues, however, the difference is not so great.
The sum of metabolism of DHBAc through MB-COMT and S-COMT could also be calculated by adding the reaction velocities of both isoenzymes (Rivett et al. 1983a). Kinetic constants determined for recombinant human COMT enzyme forms (Lotta et al. 1995) could be assumed to represent the values of pure natural enzymes. Since the Vmax values were given as catalytic numbers (kcat in 1/min) they could be converted to Vmax values (expressed as mmol/min of product formed) when total enzyme concentrations (Etot in nM) are known. These calculations are shown in Table 5. These Vmax values could be interpreted so that 64 nmol of recombinant MB-COMT produces maximally 1.41 mmol/min of the reaction product and 32 nM of recombinant S-COMT produces maximally 1.39 mmol/min of the reaction product. The molar ratio of 64 nM : 32 nM of recombinant COMT isoforms is close to the relative ratio of 70 % : 30 % (MB-COMT : S-COMT) obtained by COMT protein blotting in human brain (Tenhunen et al. 1994). By using the analyzed Km values and the corresponding calculated Vmax values, the approximation of total metabolism of DHBAc in a hypothetical human brain homogenate containing these amounts of COMT enzymes were computed. It is assumed that SAM is present at saturating concentrations and does not affect the kinetic values. At 10-400 mM DHBAc concentration range, 51-55 % of the metabolism is account for MB-COMT by our hypothetical human brain homogenate (Fig. 10A). In the rat brain, the ratio of COMT protein isoforms is 30 % : 70 % (MB-COMT : S-COMT) (Tenhunen et al. 1994). Assuming that the molar amount of MB-COMT is half of the corresponding amount of human brain (32 nM) and the amount of S-COMT is twice as high as in human brain (64 nM), the Vmax values were calculated to be half and twice for MB-COMT and S-COMT, respectively. By substituting these Vmax values to the corresponding equations for reaction velocities, the estimation of DHBAc metabolism in rat brain homogenate was computed (Fig. 10B). It revealed that only 21-24 % of 10-400 mM DHBAc is metabolized via MB-COMT. Since the reported Km value for S-COMT is lower than that of rat brain COMT (Nissinen 1985), these calculations slightly underestimate the metabolism through S-COMT. Nonetheless, with DHBAc as a substrate, it seems that half of the metabolism is via MB-COMT in human brain tissue homogenate independently of substrate concentration. Also, the present analysis (at 240 mM concentration of DHBAc, I) of rat brain tissue homogenates appears to assay preferentially the S-COMT activity.
TABLE 5. Calculated kinetic values for DHBAc.
Vmax = kcat x Etot, kcat = catalytic number, Etot = concentration of the enzyme in the assay.
a
determined with recombinant human MB-COMT and recombinant human S-COMT (Lotta et al. 1995).
Figure 10. Calculated kinetics for brain COMT using DHBAc as a substrate. The Km and Vmax values from Table 5. were substituted to enzyme kinetic equation (v = Vmax,MB x substrate concentration/(Km,MB + substrate concentration) + Vmax,S x substrate concentration/(Km,S + substrate concentration) (Rivett et al., 1983a). Based on the relative amount of COMT isoforms present in the brain (Tenhunen et al., 1994), the Vmax values corresponding to 63 nM and 32 nM for MB-COMT and S-COMT, respectively, were used for the human brain (A) while Vmax values corresponding to 32 nM and 64 nM for MB-COMT and S-COMT, respectively, were used for the rat brain (B).
Based on analogous calculations for endogenous substrates (within 10-400 mM concentrations), L-DOPA behaves similarly as DHBAc while dopamine and noradrenaline are metabolized primarily (50-90 %) via MB-COMT in hypothetical human brain homogenate (data not shown). In a hypothetical rat brain homogenate, L-DOPA again acts like DHBAc as a substrate while dopamine and noradrenaline are metabolized primarily via MB-COMT only at lower substrate concentrations (less than 100 mM and 50 mM for dopamine and noradrenaline, respectively, data not shown). These approximations represent a situation when other metabolism is blocked and the reaction is made at saturating concentrations of SAM. Also, in rat brain, the affinities of catecholamines for S-COMT are higher than with recombinant S-COMT (Lotta et al. 1995; Nissinen 1985). Altogether, these approximations are in line with earlier calculations (Rivett et al. 1982) supporting the importance of MB-COMT in the metabolism of endogenous catecholamines in vivo.
Amount of COMT. Although the actual amount of COMT proteins with respect to total protein content in the brain is not known, a rough estimation of the amount of COMT enzyme in the present COMT assay (I) could be calculated. Recombinant COMT, derived from rat liver S-COMT sequence, has been purified to near homogeneity (Lundström et al. 1992). Up to 98 % of purity (Vidgren et al. 1991) has been reported for the enzyme (denoted as fraction b by Lundström et al. 1992) in crystallization studies. This preparation has a specific activity of 500 nmol/min/mg protein (Lundström et al. 1992) with 400 mM concentration of DHBAc. If it is assumed that this preparation is 100 % active and pure soluble COMT, then 1 mg of COMT protein produces 500 nmol/min of the reaction product as a maximal capacity of methylation. Since 1 mg protein of striatal homogenate produces 45.8 pmol/min of the reaction product (note: with 240 mM concentration of DHBAc, I), only 91.6 ng of pure S-COMT protein is needed to produce that activity. This amount of COMT represents 0.09 % of the total protein in the rat brain homogenate. Since the molecular weight of recombinant S-COMT is 25 kD (Lundström et al. 1992), 9.1 ng of S-COMT protein corresponds to about 3.7 pmol of S-COMT in 1 mg of total protein in rat brain homogenate. In the present assay, approximately 400 mg of total protein has been normally used in the COMT reaction. This corresponds to 1.5 pmol of COMT protein, which is close to the calculated molar range of recombinant COMT proteins (16 pmol and 8 pmol for recombinant MB-COMT and S-COMT, respectively) used in kinetic analyses (Lotta et al. 1995). In another study, 400 mM concentration of DHBAc produced about 0.15 nmol/min of the reaction product in 1 mg of total rat brain homogenate protein (Tilgmann et al. 1992). By analogous calculations, this homogenate contains 30 ng or 1.2 pmol (0.03 %) of S-COMT protein in 1 mg of total protein. These approximations, however, are likely to overestimate the amount of S-COMT since the specific activities obtained in these studies also contain the activity derived by MB-COMT (about 20 %, Fig. 10). On the other hand, since the present assay utilizes the 900 x g supernatant, the loss of COMT activity (approximately 10 %) in the pellet is clear due mostly to loss of S-COMT (Ulmanen et al. 1997). Assuming that COMT activity in 100 000 x g supernatant is derived only from S-COMT, which has a Vmax value of 186 pmol/min/total protein (Nissinen 1985), the amount of COMT in this preparation is 1.57 ng or 63 pmol (about 0.16 %) in 1 mg of the total protein in this fraction. Altogether, these calculations suggest that the amount of COMT protein is less than 1 % of the total protein of the rat brain homogenate. Also, this molar amount of COMT enzyme in the assay could be physiologically relevant e.g. in kinetic analyses.
Cell cultures. Earlier, the COMT activity in brain cell cultures has been analyzed after homogenization of the collected cells (Hansson 1984). In our studies (IV), COMT activity was analyzed by adding the substrate directly to the viable cell cultures without addition of SAM, which penetrates poorly through cell membranes (Baldessarini 1987). Thus, the product formation must have been occurred inside the cells confirming the intracellular localization of COMT (Trendelenburg 1990; Männistö et al. 1992b; Kaakkola et al. 1994; Ulmanen et al. 1997). Although preliminary kinetic data suggest non-saturating conditions with the current concentration of the substrate, it cannot be excluded that the saturation occurred in certain cultures (data not shown). Also, the penetration of DHBAc inside the cell and the low micromolar concentration of intracellular SAM (Baldessarini 1988) affect the overall pace of the reaction. In peripheral tissues, addition of substrate in low concentrations seems to saturate COMT, below the level of uptake saturation (Trendelenburg 1986, 1990; Wilson et al. 1988). It can be speculated that MB-COMT, which has a higher affinity than S-COMT, is responsible for the methylation at low micromolar concentrations of substrates. Also, apparently high meta/para ratios in culture studies (IV) support the primary metabolism through MB-COMT over S-COMT.
Brain regions. The present results with cultured cells from discrete regions of the brain (IV) agree well with concept that COMT is widely distributed throughout the rat brain (Kaplan et al. 1979; Hansson 1984; Roth 1992; Karhunen et al. 1994). Also, the highest COMT activity was found in cerebellar cultures (IV), which corresponds to COMT staining of Bergmann glia (Kaplan et al. 1979; Karhunen et al. 1994). The basal COMT activities found here in brain cell cultures were comparable with earlier reports of primary glial cultures (Hansson 1984). In addition, in the present cultures (IV), the COMT activities are at the same level as in striatal (I,II), hypothalamic and hippocampal homogenates from rat brain.
Brain cells. Cell cultures studies (IV) confirm the presence of COMT in neurons and glia (Karhunen et al. 1994, 1995b). For the first time, the COMT activity was demonstrated in primary cultures of brain neurons. Cultured fetal basal forebrain and midbrain neurons have commonly been used as a model of striatal and nigral neurons, respectively (McMillian et al. 1995, 1997). For basal forebrain neurons, neuronal COMT is probably located in site postsynaptical to nigrostriatal dopaminergic neurons. Midbrain neuronal cultures confirm the observations of nigral COMT activity (Guldberg and Marsden 1975; Rivett et al. 1983a). Only a weak COMT immunoreactivity has been observed in dopaminergic neurons in human substantia nigra (Kastner et al. 1994). Also, a nigral lesion does not change the COMT activity in striatum (Kaakkola et al. 1987). Thus, COMT must be located in neurons other than the dopaminergic neurons in the substantia nigra.
Some of the COMT activity data supported the predominant role of COMT in glia, especially in basal forebrain neurons. First, the lowest COMT activity was found in the most pure neuronal cultures. Second, when neurons were grown on top of striatal and hypothalamic glial cells, the specific COMT activity was not changed. Indeed, when increasing amounts of neurons were plated on top of glia (12 500 - 200 000 neurons/plate) the total COMT activity increased while specific activity did not (data not shown). Third, in basal forebrain neuron enriched cultures, the increase of COMT activity corresponded to glial proliferation, which was indicated by increased GFAP staining.
Microglia. A novel finding was the presence of COMT in the microglial cells in the striatum (III). This was demonstrated three days after fluorocitrate infusion by the increased COMT activity and colocalization with double stained activated microglial cells. The astrogliosis, which takes place about one week after the toxin lesion (Rivett et al. 1983a; Kaakkola et al. 1987) can be excluded since MAO B activity, which is present in astroglia and absent in microglia (Ekblom et al. 1994), was decreased and GFAP staining was low at the injection site throughout the study. The appearance of activated microglia at the lesion site was demonstrated by increased alk-PDE activity, a marker for activated peripheral macrophages (Morahan et al. 1980), and high OX-42 staining, a marker for microglia (Graeber et al. 1989). The localization of COMT in macrophages has earlier been speculated as a cause of the transient increase of COMT activity in virus-infected brain (Guchhait and Monjan 1980). In addition, COMT immunoreactivity has been detected in peripheral macrophages (Inoue and Creveling 1986; Inoue et al. 1991). Since the meta/para ratios were not greatly changed, the dominance of either COMT isoform in microglia could not be determined. One function of increased COMT activity may be to metabolize the catecholamines which are leaking out of damaged cells. Due to reduced COMT and MAO B activity at 1-2 days after glial damage, the amount of extracellular dopamine could increase and represent as a possible factor for microglial activation. Depending on whether the effect of increased COMT activity in microglia/macrophages is regenerative (or degenerative) on brain tissue, the brain penetrating COMT inhibitors could be either clinically harmful (or useful) in pathological situations.
Although fluorocitrate is taken up by astrocytes (Clarke et al. 1970), the selectivity of this toxin is reduced at higher doses. This has been shown by decreased activity of a cholinergic neuronal marker after intrastriatal infusion of fluorocitrate (Paulsen et al. 1987). However, the major effect of fluorocitrate on glial cells was seen as a decrease of MAO B activity and by a decrease of COMT activity for two days after fluorocitrate infusion evidence of the presence of COMT in glial cells. A trend towards increased meta/para ratio 24 h after the 2 nmol dose of fluorocitrate suggested that the reduction of total COMT activity is due to decreased glial S-COMT activity. The higher dose seems to reduce almost equally the activity of both COMT isoforms. Apparently fluorocitrate affected also dopaminergic neurons, since staining of TH at the site of lesion was decreased and increased outside the lesion suggesting induction of compensatory dopaminergic tone. Differential staining of TH also explains the inconsistent results in TH activity. No studies are available concerning the long term effects of fluorocitrate.
Kidney. In kidney COMT activity is concentrated in cortex but the deeper parts also have a considerable COMT activity (V). Little is known about the distribution of COMT activity in the different regions of kidney tissue. COMT activity in kidney cortex has been reported to be higher than in medulla (Goldstein et al. 1980). The COMT mRNA (Meister et al. 1993) and protein (Kaplan et al. 1979; Karhunen et al. 1994) have been demonstrated to reside in proximal and distal tubules in addition to the collecting duct and ureter in rat kidney. The present results on the distribution of COMT activity suggest that metabolism via COMT could take place throughout the kidney but the primary site is in cortical areas, e.g. proximal tubular cells.
Recombinant COMT. The direct effect of ethanol on COMT activity was studied in vitro with recombinant COMT enzymes (II). Ethanol inhibited recombinant MB-COMT activity but this was significant only at 1000 mM concentration. At this ethanol concentration, recombinant MB-COMT activity was inhibited with a mixed type of inhibition. With recombinant S-COMT only the formation of vanillic acid was affected by 1000 mM ethanol concentration, which increased the Vmax value of recombinant S-COMT. These opposite effects of ethanol on MB-COMT and S-COMT seemed to cancel each other out, producing mainly a decrease of the formation of isovanillic acid in striatal tissue homogenates. As expected, all these changes were reflected in the meta/para ratios.
There was a trend towards reduction of MB-COMT activity already at 100 mM ethanol concentration (II). Previously, no evidence for the inhibition of (apparently S-) COMT by ethanol at concentrations up to 90 mM have been found in vitro (Lahti and Majchrowicz 1974; Giovine et al. 1977; Hoffman et al. 1981). Due to the opposite effect of ethanol on COMT isoforms these changes have not been detected earlier. However, an acute in vivo administration of ethanol (1 g/kg, i.p., which produced 25 mM concentration of ethanol in the blood) on rats has decreased the 3-MT concentration in the nucleus accumbens of non-alcohol prefering rats and (pargyline treated) Wistar rats (Honkanen et al. 1994). Also, in the mouse striatum, a decrease of 3-MT by ethanol (3.5 g/kg, i.p.) has been noticed (Milio and Hadfield 1992). In contrast, ethanol and other reinforcing drugs, rather increase the release of dopamine in the brain, especially in nucleus accumbens. Thus, it is tempting to speculate that decreased 3-MT formation is caused by selective inhibition of postsynaptic MB-COMT, possibly leading to elevated dopaminergic tone. The present experiments give only faint support for the inhibition of MB-COMT by 100 mM concentration of ethanol, which is a clinically intoxicating blood concentration (4.6 g/l). Thus, a slight potentiation of COMT inhibitor toxicity could only be possible at very high concentrations of ethanol.
Cell cultures. In primary cultures of rat brain cells, COMT activity was effectively decreased by the nitrocatechol-type inhibitors with tolcapone being slightly more potent than entacapone in cultures containing glial cells (IV). The approximated 50 % inhibition of COMT activity was achieved with 10-150 nM concentration of entacapone and tolcapone. This is in agreement with the reported IC50 values of entacapone and tolcapone (2.2-160 nM for brain and liver COMT) (Zürcher et al. 1990; Nissinen et al. 1992). Based on Ki values determined with recombinant human COMT enzymes, tolcapone is slightly more potent against MB-COMT (2.0 nM and 0.3 nM for entacapone and tolcapone, respectively), while the Ki values of both drugs do not differ for S-COMT (Lotta et al. 1995). At the catalytic site, the binding should be similar suggesting equal affinity for both drugs (Lotta et al. 1995; Vidgren and Ovaska 1997). The apparent difference in potency between entacapone and tolcapone is probably due to their membrane penetrating ability. Since tolcapone is a brain penetrating drug (Männistö et al. 1992a) it could reach the enzyme slightly more effectively.
Examination of the approximated 50 % inhibitory concentrations of the nitrocatechol-type drugs revealed a trend for more sensitive inhibition of neuronal COMT than glial COMT. This was supported by the reduction of COMT inhibitor efficacy after the proliferation of glial cells in neuron-enriched cultures. Both entacapone and tolcapone were equipotent in these neuronal cultures. Since tolcapone has been claimed to be more effective against MB-COMT than S-COMT (Borges et al. 1997; Vieira-Coelho and Soares-da-Silva 1999) and primary neuronal cultures apparently contain about the same amount of MB-COMT and S-COMT protein (Karhunen et al. 1995b), a slight sensitivity of neuronal COMT to inhibition by both entacapone and tolcapone could possibly explain these results.
CGP 28014 did not affect COMT activity in any of the cultures (IV). Also previous experiments with tissue homogenates (Waldmeier et al. 1990) and with aggregate cultures (Wiese et al. 1993) demonstrated that CGP 28014 does not inhibit COMT in vitro. Thus, CGP 28014 is not metabolized to an active COMT inhibitor by brain cells. Since the main metabolite of CGP 28014 was also ineffective in vitro, inhibition of glial uptake2 was suggested (Waldmeier et al. 1990). However, in our laboratory CGP 28014 did not reduce the uptake of [3H]-dopamine into glial cells (M. Törnwall 1994, unpublished results). Thus, the mechanism of the inhibition of O-methylation CGP 28014 remains unknown.
Kidney. In kidney (V), increased natriuresis by entacapone was demonstrated to be caused by inhibition of COMT activity. Furthermore, the inhibition of COMT activity by entacapone was seen to be equal in all regions of the kidney. Previously, the inhibition of COMT activity with nitecapone has been noted, but this was not fully demonstrated (Eklöf et al. 1997). The natriuretic effect was produced by local COMT inhibition in the kidney since only a transient effect of entacapone on brain COMT was observed. At this dose entacapone affects briefly also brain COMT activity (Kaakkola and Wurtman 1992). However, due to the shorter duration of COMT inhibition in brain than in kidney, a local effect on kidney function predominates.
Both entacapone and L-DOPA induced natriuresis, which was mediated by D1 receptors, as has been reported before (Eklöf et al. 1997; Hansell et al. 1998). Nitecapone and gamma-L-glutamyl-DOPA, a kidney specific dopamine precursor, have additive effect on natriuresis (Eklöf et al. 1997). This also emphasizes the important role of kidney dopamine in the regulation of salt balance.
In addition to the natriuretic effect, the inhibition of COMT was accompanied by a slight increase of dopamine levels and over three-fold increase of DOPAC levels in the urine whereas L-DOPA produced a smaller natriuretic effect but greater efflux of dopamine and less increase in DOPAC excretion. This suggests that L-DOPA is metabolized first to dopamine outside of the kidney while entacapone acts more locally to elevate kidney dopamine concentrations providing more dopamine to stimulate its receptors within the kidney.