Phenylethanolamine N-methyltransferase has β-carboline 2N-methyltransferase activity: hypothetical relevance to Parkinson’s disease

Abstract

Mammalian brain has a β-carboline 2N-methyltransferase activity that converts β-carbolines, such as norharman and harman, into 2N-methylated β-carbolinium cations, which are structural and functional analogs of the Parkinsonian-inducing toxin 1-methyl-4-phenylpyridinium cation (MPP⁺). The identity and physiological function of this β-carboline 2N-methylation activity was previously unknown. We report pharmacological and biochemical evidence that phenylethanolamine N-methyltransferase (EC 2.1.1.28) has β-carboline 2N-methyltransferase activity. Specifically, purified phenylethanolamine N-methyltransferase (PNMT) catalyzes the 2N-methylation (21.1 pmol/h per unit PNMT) of 9-methylnorharman, but not the 9N-methylation of 2-methylnorharmanium cation. LY134046, a selective inhibitor of phenylethanolamine N-methyltransferase, inhibits (IC₅₀ 1.9 μM) the 2N-methylation of 9-methylnorharman, a substrate for β-carboline 2N-methyltransferase. Substrates of phenylethanolamine N-methyltransferase also inhibit β-carboline 2N-methyltransferase activity in a concentration-dependent manner. β-Carboline 2N-methyltransferase activity (43.7 pmol/h/mg protein) is present in human adrenal medulla, a tissue with high phenylethanolamine N-methyltransferase activity.

We are investigating the potential role of N-methylated β-carbolinium cations in the pathogenesis of idiopathic Parkinson’s disease. Presuming that phenylethanolamine N-methyltransferase activity forms toxic 2N-methylated β-carbolinium cations, we propose a novel hypothesis regarding Parkinson’s disease—a hypothesis that includes a role for phenylethanolamine N-methyltransferase-catalyzed formation of MPP⁺-like 2N-methylated β-carbolinium cations.

Introduction

We hypothesize that β-carboline N-methyltransferases and the products of their catalytic activities, N-methylated β-carbolinium cations, contribute to the development of idiopathic Parkinson’s disease (Collins and Neafsey, 1985, Collins, 1994). The original basis for this hypothesis was the remarkable structural similarity between N-methylated β-carbolinium cations and neurotoxic 1-methyl-4-phenylpyridinium cation (MPP⁺). In the central nervous system of human and nonhuman primates, MPP⁺ selectively destroys the dopaminergic nigrostriatal system, resulting in a Parkinsonian condition (Burns et al., 1984, Chiueh et al., 1984, Langston et al., 1983, Langston, 1995). Previous work indicates that N-methylated β-carbolinium cations, including 2-methylnorharmanium cation and 2,9-dimethylnorharmanium cation, share several functional and toxic properties with MPP⁺ (Cobuzzi et al., 1994, Albores et al., 1990, Collins et al., 1995, Collins et al., 1996, Collins et al., 1992b, Collins, 1994; Fields et al., 1992a, Fields et al., 1992b; Drucker et al., 1990, Neafsey et al., 1989, Neafsey et al., 1995, Matsubara et al., 1992, Matsubara et al., 1998a, Matsubara et al., 1998b, Gearhart et al., 2000a).

N-Methylated β-carbolinium cations are present in postmortem brain from neurologically normal individuals, but no one has investigated the levels of these toxins in postmortem brain from Parkinson’s cases (Matsubara et al., 1993, Matsubara, 1996). Consistent with the postulate that N-methylated β-carbolinium cations may play a role in Parkinson’s disease, Matsubara et al. reported that the levels of these toxins are significantly elevated in the lumbar cerebrospinal fluid from Parkinson’s patients (Matsubara et al., 1995).

Two distinct S-adenosyl-l-methionine (SAM)-dependent N-methyltransferase activities result in the formation of toxic N-methylated β-carbolinium cations (Matsubara et al., 1992, Collins et al., 1992b). We named these enzymatic activities β-carboline 2N-methyltransferase and β-carboline 9N-methyltransferase, for N-methylation occurring at the pyridyl-nitrogen (2N-) and indole nitrogen (9N-), respectively. β-Carboline 2N-methyltransferase and β-carboline 9N-methyltransferase activities are present in mammalian brain homogenates (Collins et al., 1992b, Matsubara et al., 1992, Matsubara et al., 1993, Matsubara, 1996, Gearhart et al., 1997, Gearhart et al., 2000b), including postmortem human brain from control (Matsubara et al., 1993, Gearhart et al., 2000b) and Parkinson’s disease cases (Gearhart et al., 2000b). Consistent with the postulate that β-carboline N-methylation may play a role in Parkinson’s disease, brain β-carboline 9N-methyltransferase activity is increased to 400% of control activity in the supernatant fraction prepared from postmortem frontal cortex from Parkinson’s disease cases (Gearhart et al., 2000b). Unfortunately, no one has purified β-carboline 2N-methyltransferase or β-carboline 9N-methyltransferase.

The focus of this report is the β-carboline 2N-methyltransferase activity, which catalyzes the 2N-methylation of simple β-carbolines, such as norharman and harman. The reaction products are 2N-methylated β-carbolinium cations, which are structural and functional analogs of the Parkinsonian-inducing MPP⁺. Characterization of β-carboline 2N-methyltransferase is important because its activity forms toxins, and β-carboline 2N-methyltransferase substrates (norharman and harman) are elevated in the plasma (Kuhn et al., 1995) and CSF (Kuhn et al., 1996) of individuals diagnosed with Parkinson’s disease. Michaelis–Menten enzyme kinetics predicts that reaction velocity (i.e. product formation) increases with increased substrate concentration (Lehninger, 1975, Cornish-Bowden, 1979); therefore, we predict that individuals with elevated β-carbolines will also have increased levels of toxic N-methylated β-carbolinium cations. We consider norharman and harman—substrates for the β-carboline 2N-methyltransferase activity discussed here—as multisource protoxicants, because these simple β-carbolines are widespread in the environment (Adachi et al., 1991, Collins, 1983, Bosin et al., 1988, Felton and Knize, 1990, Gross et al., 1993, Poindester and Carpenter, 1962, Rommelspacher and Schmidt, 1985). β-Carboline 2N-methyltransferase substrates (norharman and harman) are also present at picomolar levels in mammalian tissues and body fluids, including human brain and cerebrospinal fluid (Airaksinen and Kari, 1981, Bosin et al., 1989, Collins, 1983, Fekkes et al., 1992, Fekkes and Bode, 1993, Matsubara, 1996, Peura et al., 1989, Rommelspacher et al., 1991, Rommelspacher and Schmidt, 1985, Kuhn et al., 1995, Kuhn et al., 1996).

We were unable to purify β-carboline 2N-methyltransferase from mammalian brain; therefore, we used pharmacological and biochemical methods in an effort to identify the physiological role of β-carboline 2N-methyltransferase. We speculated that β-carboline 2N-methyltransferase activity may be due to any one of a number of previously characterized CNS-residing N-methyltransferases (Nagatsu et al., 1977, Naoi et al., 1989, Naoi et al., 1997, Saavedra et al., 1973, Ansher and Jakoby, 1987, Ansher et al., 1986), including SAM: phenylethanolamine N-methyltransferase (EC 2.1.1.28) (Eagles and Iqbal, 1974, Nagatsu et al., 1977, Vogel et al., 1976). Phenylethanolamine N-methyltransferase (PNMT) is a cytosolic, SAM-dependent N-methyltransferase, which catalyzes the formation of epinephrine from norepinephrine. PNMT activity is present in the adrenal medulla (Kitabchi and Williams, 1969, Axelrod, 1971), brain (Fuller, 1982, Kopp et al., 1979, Lew et al., 1977, Mefford, 1988, Vogel et al., 1976) and other tissues (Kennedy et al., 1995).

We report evidence that PNMT has a β-carboline 2N-methyltransferase activity: purified PNMT 2N-methylates a β-carboline; a selective inhibitor of PNMT inhibits β-carboline 2N-methyltransferase activity; substrates for PNMT inhibit β-carboline 2N-methyltransferase activity; and adrenal medulla exhibits β-carboline 2N-methyltransferase activity.