Comparative Study on the Vasorelaxant Effects of Three Harmala Alkaloids in Vitro

Jpn. J. Pharmacol. 85, 299 – 305 (2001)

Comparative Study on the Vasorelaxant Effects of Three Harmala Alkaloids In Vitro

Chuen-Chao Shi1, Jyh-Fei Liao1,* and Chieh-Fu Chen1,2

1Department and Institute of Pharmacology, National Yang-Ming University, Taipei 112, Taiwan 2National Research Institute of Chinese Medicine, Taipei 112, Taiwan

Received November 6, 2000 Accepted December 19, 2000

ABSTRACT—Three psychological active principles from the seeds of Peganum harmala L., harmine, har- maline and harmalol, showed vasorelaxant activities in isolated rat thoracic aorta preparations precontracted by phenylephrine or KCl with rank order of relaxation potency of harmine > harmaline > harmalol. The vasorelaxant effects of harmine and harmaline (but not harmalol) were attenuated by endothelium removal or pretreatment with a nitric oxide (NO) synthase NM-nitro-L-arginine methyl ester. In cultured rat aortic endothelial cells, harmine and harmaline (but not harmalol) increased NO release, which was dependent on the presence of external Ca2+. In endothelium-denuded preparations, pretreatment of harmine, harmaline or harmalol (3 – 30 mM) inhibited phenylephrine-induced contractions in a non-competitive manner. Receptor binding assays indicated that all 3 compounds interacted with cardiac a 1-adrenoceptors with comparable affinities (Ki value around 31 – 36 mM), but only harmine weakly interacted with the cardiac 1,4-dihydro- 2+ pyridine binding site of L-type Ca channels (Ki value of 408 mM). Therefore, the present results suggested that the vasorelaxant effects of harmine and harmaline are attributed to their actions on the endothelial cells to release NO and on the vascular smooth muscles to inhibit the contractions induced by the activation of receptor-linked and voltage-dependent Ca2+ channels. The vasorelaxant effect of harmalol was not endothe- lium-dependent.

Keywords: Harmine, Harmaline, Harmalol, Vasorelaxant, Nitric oxide

Harmala alkaloids harmine, harmaline and harmalol frequently followed by a secondary increase; and the effects (Fig. 1) are the psychological active principles from the of harmalol on these two parameters are inconsistent (18). seeds of Peganum harmala L., which are also distributed Recently, we reported the in vivo cardiovascular effect and widely in other medicinal plants and found endogenously in in vitro vasorelaxant effect of harman (Fig. 1), another har- mammalian tissues (1, 2). These harmala alkaloids have a mala alkaloid (19). Our results suggested that the vasore- wide spectrum of pharmacological actions in the central laxant effect of harman may be involved in its hypotensive nervous system such as tremorogenesis (3, 4), hypothermia effect, and the vasorelaxant effect is attributed to its actions (5), hallucinogenesis (6, 7), central monoamine oxidase in- on the endothelial cells to release nitric oxide (NO) and on hibition (8 – 10), convulsive or anticonvulsive actions (11) the vascular smooth muscles to inhibit the contractions and binding to various receptors including 5-HT receptors induced by the activation of receptor-linked and voltage- 2+ and the benzodiazepine binding site of GABAA receptors dependent Ca channels. Although the cardiovascular effects (12 – 14). In addition, these compounds also have antioxi- of these harmala alkaloids may not be of practical impor- dative (15), platelet aggregation inhibitory (16) and immu- tance (or for clinical use), the related information for these nomodulatory effects (17). There are also some reports potential hallucinogens should be important for their toxi- concerning the cardiovascular actions of these harmala cology. Because the chemical structures of these 4 com- alkaloids. For example, it has been reported that harmine pounds are very similar, the present study was carried out reduces systemic arterial blood pressure and total periph- to compare harmine, harmaline and harmalol in terms of eral vascular resistance; harmaline-evoked decreases are vasorelaxant effects on the isolated rat thoracic aorta preparations and stimulating effects on NO release from *Corresponding author. FAX: +886-2-28264372 cultured rat aortic endothelial cells. To compare with har- E-mail: [email protected] man (19), their affinities for cardiac a 1-adrenoceptors and

299 300 C.-C. Shi, J.-F. Liao & C.-F. Chen

37°C. The composition of Krebs’ solution was as follows: 118 mM NaCl, 4.7 mM KCl, 25 mM NaHCO3, 1.2 mM KH2PO4, 2.5 mM MgSO4, 2.5 mM CaCl2 and 11.1 mM glucose. Isometric tension change was measured with a Grass FT03 force transducer and recorded on a 4-channel polygraph (Gould RS3400 polygraph; Oxnard, CA, USA). Before starting the experiment, all preparations were al- lowed to equilibrate for 60 min, during which time Krebs’ solution was replaced at least twice.

Vasorelaxant effects To evaluate the vasorelaxant and endothelium-dependent effects of test compounds, endothelium-intact and -denuded preparations were pretreated with phenylephrine (PE, 0.3 mM) or KCl (60 mM) to produce sustained contractions (19). Lack of functional vascular endothelium was confirmed by the loss of relaxant response to 3 mM acetylcholine before the experiment began. After the contraction had reached a stable plateau, cumulative concentrations of test compound were added. The vasorelaxant effect of test compound was expressed as a percentage of relaxation, and the IC50 (the concentration to produce a 50% maximal relaxation) value was determined from the concentration-response curve by data fitting with computer software GraFit (Erithacus Soft- ware, Staines, Middlessex, UK). The involvement of the mediator for endothelium-related vasorelaxation induced by test compound was examined by pretreatment of preparations with NM-nitro-L-arginine methyl ester (a NO synthase inhibitor), tetraethylammonium (a K+ channel blocker) or indomethacin (a cyclooxygenase inhibitor). A series of experiments was designed to assess the in- volvement of a 1-adrenoceptors in the vasorelaxant effect of test compound in endothelium-denuded aortic preparations. m Fig. 1. Chemical structures of harmala alkaloids. Various concentrations of test compound (3, 10, 30 M) were added 10 min before the construction of cumulative concentration-response curves with PE. The results were expressed as the percentage of the maximum contractile the 1,4-dihydropyridine (DHP) binding site of L-type Ca2+ tension to PE before and after pretreatment with test com- channels were also studied with receptor binding assays. pound.

MATERIALS AND METHODS NO measurement The culture of rat aortic endothelial cells and the mea- Isolated thoracic aorta preparations surement of NO in the medium were according to the meth- Sprague-Dawley rats weighing 180 – 280 g were sacri- ods of Wang et al. (20, 21). In brief, endothelial cells were ficed by decapitation. The thoracic aorta was excised and grown in 35-mm2 dishes in 1 ml of Dulbecco’s modified adhesive connective tissues were carefully cleared away. Eagle’s medium supplemented with 10% fetal calf serum The aorta was cut into approximately 3 – 4-mm-long ring and antibiotics. Upon reaching confluence in about 4 days, segments. Denuded aorta ring segments were obtained by the medium was changed to 1 ml of Hanks’ balanced salt gently rubbing with the finger tip. The isolated aorta ring solution (HBSS) with L-arginine (100 mM) and added was suspended under a basal tension of 1 g in a 10 ml organ CaCl2 (to 2.5 mM). The changing over to HBSS was neces- bath containing Krebs’ solution continuously aerated with a sary because it provided the least interference in the assay. 2+ gas mixture of 95% O2 and 5% CO2 and maintained at However, additional Ca was required to make the final Vasorelaxation of Harmala Alkaloids 301

concentration comparable to that in normal Krebs’ solution. (CaCl2), potassium chloride (KCl), EGTA, sodium chloride The cells were then equilibrated for 60 min at 37°C. Ali- (NaCl), magnesium sulfate heptahydrate (MgSO4 · 7H2O), quots (250 ml) of the supernatant were collected for analy- monopotassium phosphate (KH2PO4), sodium bicarbonate sis of nitrite by chemiluminescence, and the total content (NaHCO3), and D-glucose were purchased from Sigma of NO in the medium before test compound treatment was (St. Louis, MO, USA). Dulbecco’s modified Eagle’s medi- calculated and taken as basal 100%. Vehicle (HBSS with- um, fetal calf serum, and HBSS were purchased from out Ca2+) or test compound (1, 10, 100 mM) was then added Gibco Life Technologies (Grand Island, NY, USA). Vana- for 30 min, and then the cell supernatants (100 ml) were dium chloride was purchased from Aldrich (Milwaukee, collected for analysis of nitrite to examine the change of WI, USA). Radioligands [7-methoxy-3H]-prazosin (76.2 Ci NO content. Similar experiments also were carried out in /mmol) and [isopropyl-1,3-3H]-nimodipine (121.1 Ci/mmol) Ca2+-free HBSS to examine the role of external Ca2+. were purchased from NEN, Du Pont (Boston, MA, USA). Samples (100 ml) containing nitrite were measured by add- ing a reducing agent (0.8% VCl3 in HCl) to the purge vessel RESULTS to convert nitrite to NO, which was then carried by a flow of helium to the NO analyzer (Model 280; Sievers Re- As shown in Fig. 2, harmine, harmaline and harmalol search, Boulder, CO, USA). Nitrite concentrations were cal- (0.3 – 100 mM) concentration-dependently relaxed endo- culated by comparison with a standard solution of sodium thelium-intact rings precontracted with PE (0.3 mM), and nitrite. the vasorelaxant responses to harmine and harmaline, but not harmalol, were significantly depressed in endothelium- Receptor binding assays denuded preparations. The IC50 (the concentration to pro- According to previous reports (22, 23), the interaction of duce a 50% maximal relaxation) values of harmine and test compound with a 1-adrenoceptors or the DHP binding harmaline for PE-induced contractions were significantly site of L-type Ca2+ channels was assessed in rat heart mem- increased to 2.1-fold and 1.3-fold, respectively, in endo- brane preparations. In brief, binding assays were initiated thelium-denuded preparations as compared with that in en- by the addition of a receptor membrane preparation in an dothelium-intact preparations (Table 1). The IC50 values for appropriate buffer containing the specific radioligand PE-induced contractions showed that the rank order of 3 3 ([ H]prazosin for a 1-adrenoceptors or [ H]nimodipine for relaxation potency was harmine > harmaline > harmalol in DHP binding sites). After incubation, bound ligands were both endothelium-intact and -denuded preparations (Table 1). separated from free ligands by vacuum filtration through a Except for harmalol, harmine and harmaline also concen- 24-mm glass fiber filter (Whatman GF/C; Whatman, tration-dependently relaxed KCl (60 mM)-induced contrac- Maidstone, UK). The radioactivity of bound radioligand tions, having a similar rank order of relaxation potency of was then counted in 10 ml of scintillation counting fluid harmine > harmaline in both endothelium-intact and -de- (FluoranSafe 2; BDH Laboratory Supplies, Poole, UK) nuded preparations (Table 1). with a b-counter (LS 3801; Beckman, San Ramon, CA, Similar to endothelium removal, the vasorelaxant effects USA). Test compound was tested in various concentrations of harmine and harmaline on PE-induced contractions in from 0.1 to 1000 mM. The competition binding curve was endothelium-intact aortic preparations were attenuated by M analyzed to determine the IC50 (concentration of the test pretreatment with N -nitro-L-arginine methyl ester (500 mM, compound to competitively displace 50% of specifically 10 min), but not by tetraethylammonium (10 mM, 60 min) bound radioligand) using the computer software GraFit or indomethacin (30 mM, 45 min) (Fig. 3). These pretreat- (Erithacus Software, Staines, Middlessex, UK). The Ki ments had no effects on the vasorelaxant effect of harmalol value was calculated from the IC50 value using the Cheng- (Fig. 3). Prusoff equation (24). The effects of these 3 compounds on the NO release from cultured aortic endothelial cells are shown in Fig. 4. Statistical analyses Harmine and harmaline, but not harmalol (1 – 100 mM), Data are expressed as the mean ± S.E.M. and were ana- concentration-dependently stimulated NO production from lyzed by Student’s t-test or one way analysis of variance endothelial cells in medium containing 2.5 mM Ca2+ but not followed by the Newman-Keuls test with a significance in Ca2+-free medium, indicating that the external Ca2+ was level of P < 0.05. necessary for their actions. To determine whether these 3 compounds act on the a 1- Drugs adrenoceptors of vascular smooth muscles, cumulative Harmine, harmaline, harmalol, acetylcholine chloride, concentration-response curves of PE (1 nM to 100 mM) phenylephrine HCl (PE), NM-nitro-L-arginine methyl ester, was established in endothelium-denuded aorta preparations tetraethylammonium chloride, calcium chloride dihydrate pretreated with various concentrations of harmine, harma- 302 C.-C. Shi, J.-F. Liao & C.-F. Chen

Table 1. Vasorelaxant potencies of harmine, harmaline and harmalol on phenylephrine (0.3 mM)- and KCl (60 mM)-induced contractions of endothelium-intact (+EC) and endothelium-denuded (-EC) aortic preparations

IC50 value (mM) Compounds Phenylephrine KCl

+EC Harmine 8 ± 110± 1 Harmaline 41 ± 433± 7 Harmalol 109 ± 5 >1000 -EC Harmine 17 ± 217± 2 Harmaline 53 ± 270± 2 Harmalol 188 ± 40 >1000

Data are the mean ± S.E.M (n = 3 – 20) of IC50 (the concentration to produce a 50% maximal relaxation) values.

comparable Ki values around 31 – 36 mM, and only harmine had weak interaction with the DHP binding site of 2+ cardiac L-type Ca channels with a Ki value of 408 mM (Table 3).

DISCUSSION

Similar to the previous study on harman (19), the present study demonstrated that the vasorelaxant effects of harmine and harmaline were associated with the actions on both endothelial cells and vascular smooth muscles. However, the vasorelaxant effect of harmalol was endothelium- independent. Endothelium removal had the greatest influence on the vasorelaxant effect of harmine among these 3 compounds in PE-induced contractions, but the twofold increase in IC50 values for harmine (changed from 8 to 17 mM) was less than that for harman (fivefold increase, changed from 9 to m Fig. 2. Effects of harmine (A: 3 – 100 mM), harmaline (B: 3 – 48 M) (19). To examine the involvement of a mediator in 100 mM) and harmalol (C: 3 – 100 mM) on phenylephrine-induced this endothelium-related vasorelaxation, the results obtained contractions in isolated rat endothelium-intact (+EC) and -denuded with the use of different blockers (NM-nitro-L-arginine me- (-EC) thoracic aortic preparations. When phenylephrine (0.3 mM)- thyl ester, tetraethylammoniumm and indomethacin) indi- induced contraction reached the plateau, harmine, harmaline or har- malol was cumulatively added. Data are the mean ± S.E.M. (n = 3 – cated that NO, but not endothelium-derived hyperpolariza- 20 in each group) and are expressed as % of control tension. ing factor or prostacyclin, is involved in the vasorelaxant *P<0.05, **P<0.01, as compared with the +EC. effect of harmine or harmaline. The action of harmine or harmaline on endothelial cells to release NO was confirmed in experiments directly with cultured aortic endothelial line or harmalol (3, 10 and 30 mM). All 3 compounds shift- cells. Although the underlying mechanism of action was ed the concentration-response curves of PE rightward, as unclear, external Ca2+ was necessary for the action of revealed by the decrease in the ED50 (-logM) value of harmine or harmaline to increase NO release from endo- PE, but suppressed the maximal response (Fig. 5, Table 2), thelial cells. Consistent with its endothelium-independent indicating that these antagonistic effects on vascular a 1- vasorelaxation, harmalol had no significant effect on NO adrenoceptors are not competitive. release from endothelial cells. Receptor binding assays indicated that all 3 compounds At concentrations higher than 3 mM, the vasorelaxant were able to interact with cardiac a 1-adrenoceptors with effects of these 3 compounds were only partially or not at- Vasorelaxation of Harmala Alkaloids 303

Fig. 4. Effects of harmine (A: 1, 10 or 100 mM), harmaline (B: 1, 10 or 100 mM), and harmaolol (C: 1, 10 or 100 mM) on NO produc- tion from cultured rat aortic endothelial cells. NO content in the medium was quantified before (as basal) and after cells were treated

M with vehicle control, harmine, harmaline or harmalol for 30 min in Fig. 3. Effects of N -nitro-L-arginine methyl ester, indomethacin 2+ 2+ 2.5 mM Ca medium or Ca -free medium. Data are the mean and tetraethylammonium on the vasorelaxant action of harmine ± S.E.M. (n = 3 – 10 in each group). *P<0.05, **P<0.01, as com- (A: 3 – 100 mM), harmaline (B: 3 – 100 mM) and harmalol (C: 3 – pared with the corresponding control. 100 mM) in endothelium-intact aortic preparations precontracted with phenylephrine. Preparations were pretreated with NM-nitro-L-arginine methyl ester (L-NAME, 500 mM for 10 min), indomethacin (30 mM for 45 min) or tetraethylammonium (TEA, 10 mM for 60 min) prior to harmine, harmaline or harmalol addition. Data are the mean supporting a direct action on the vascular smooth muscle ± S.E.M. (n = 3 – 20 in each group) and are expressed as % of control cells. The cumulative concentration-response curves to PE < < tension. *P 0.05, **P 0.01, as compared with the control. were shifted to the right with the maximum response at- tenuated by these 3 compounds, indicating that the antago- nistic effects of these 3 compounds on a 1-adrenoceptors of tenuated by endothelium removal, indicating a direct action vascular smooth muscle cells were not competitive. How- on the vascular smooth muscle cells. Indeed, pretreatment ever, in a similar concentration range, all 3 compounds are with any of these 3 compounds (3, 10, 30 mM) inhibited able to interact with cardiac a 1-adrenoceptors with compa- PE-induced contractions in endothelium-denuded aortic rings, rable Ki values around 31 – 36 mM. Therefore, it is possible 304 C.-C. Shi, J.-F. Liao & C.-F. Chen

Table 2. Effects of harmine, harmaline and harmalol on Emax and EC50 values of cumulative phenylephrine-induced contractions in endothelium-denuded aortic preparations

Treatment Emax (%) EC50 (-logM) Control 100 ± 0 6.48 ± 0.12 Harmine 3 mM80± 7** 5.75 ± 0.13 Harmine 10 mM83± 5* 5.66 + 0.14 Harmine 30 mM50± 9** 4.69 ± 0.73* Control 100 ± 0 6.77 ± 0.07 Harmaline 3 mM91± 7 6.15 ± 0.14** Harmaline 10 mM83± 6* 5.75 ± 0.02** Harmaline 30 mM77± 6** 5.43 ± 0.02** Control 100 ± 0 6.82 ± 0.05 Harmalol 3 mM96± 2 6.69 ± 0.21 Harmalol 10 mM74± 3** 5.80 ± 0.08** Harmalol 30 mM56± 7** 5.06 ± 0.33**

Data are the mean ± S.E.M. (n = 3 – 9) of Emax (% of control maxi- mum contraction) and EC50 (the concentration to produce a 50% of maximum contraction) values. *P<0.05, **P<0.01, as compared with the corresponding control.

Table 3. The affinities of harmine, harmaline and harmalol for cardiac a 1-adrenoceptors and 1,4-dihydropyridine (DHP) binding site of L-type Ca2+ channels

Ki value (mM)

Compound a 1 DHP binding site

Harmine 31 ± 5408± 70 Harmaline 36 ± 4 >1000 Harmalol 33 ± 1 >1000

Data are the mean ± S.E.M. of Ki values obtained from 3 – 5 experi- ments with duplicate determinations.

increase of 45Ca2+ uptake induced by high K+, norepineph- Fig. 5. Effects of harmine (A: 3, 10 or 30 mM), harmaline (B: 3, 10 rine and carbachol in rabbit aorta and guinea pig taenia, or 30 mM) and harmalol (C: 3, 10 or 30 mM) on the concentration indicating an inhibition on Ca2+ channels of vascular and response curves of phenylephrine in endothelium-denuded aortic intestinal smooth muscles (26). Furthermore, the inhibitory preparations. Harmine, harmaline or harmalol was pretreated for 10 min prior to construction of the concentration response curve of effects of harmaline on these smooth muscles were not phenylephrine. Data are the mean ± S.E.M. (n = 3 – 9 in each group) reversed by Bay K 8644 (26), a dihydropyridine that stimu- and are expressed as % of maximum contraction. lates Ca2+ through L-type Ca2+ channels (27), indicating that the DHP binding site of L-type Ca2+ channels is not involved. Consistent with this report, the present results reveal a that this noncompetitive antagonism is caused by another vasorelaxation effect of harmine or harmaline on KCl- vasorelaxant action combined with the a 1-adrenoceptor induced contractions in endothelium-denuded preparations. blockade. However, due to the difference of receptor sub- The receptor binding assay on the DHP binding site of L- 2+ types (a 1B-adrenoceptors in rat heart and a 1D-adrenocep- type Ca channels indicated that only harmine had a weak tors in rat aorta) (25), it should be determined whether interaction, suggesting that other binding sites are involved. these compounds have similar interactions with vascular The precise site of action of these compounds on the Ca2+ a 1D-adrenoceptors. channels of smooth muscle cells needs to be further studied. Harmaline, at the concentrations needed to inhibit the In terms of the structure-activity relationship, these results smooth muscle contraction, has been reported to inhibit the indicated that: a) change of dihydro-b-carboline to b-car- Vasorelaxation of Harmala Alkaloids 305 boline (i.e., harmaline vs harmine) increases the potency Theoretical structure-activity studies of b-carboline analogs. of vasorelaxation and the affinity for the DHP binding site; Requirements for benzodiazepine receptor affinity and antago- nist activity. Mol Pharmacol 28, 17 – 31 (1985) b) change of OH to OCH3 group in the C7 position (i.e., 12 Lamarre Y and Puil E: Induction of rhythmic activity by harma- harmalol vs harmaline) increases the action on the endo- line. Can J Physiol Pharmacol 52, 905 – 908 (1974) thelium to release NO (but this action also appeared in 13 Rommelspacher H, Nanz C, Borbe HO, Fehske KJ, Muller WE harman (19), which has no functional group in this posi- and Wollert U: 1-Methyl-beta-carboline (harmane), a potent tion); and c) the functional group in the C7 position plays endogenous inhibitor of benzodiazepine receptor binding. no important role in the affinity for the cardiac a 1-adreno- Naunyn Schmiedebergs Arch Pharmacol 314, 97 – 100 (1980) ceptors. 14 Rommelspacher H, Bruning G, Susilo R, Nick M and Hill R: Pharmacology of harmalan (1-methyl-3,4-dihydro-b-carboline). Acknowledgments Eur J Pharmacol 109, 363 – 371 (1985) This work was supported by grants from the National Science 15 Tse SYH, Mak IT and Dickens BF: Antioxidative properties of Council, Executive Yuan (NSC-86-2314-B-077-004-M13) and the harmane and b-carboline alkaloids. Biochem Pharmacol 42, National Research Institute of Chinese Medicine (NRICM-86201), 459 – 464 (1991) Taipei, Taiwan. Mr. Liao received an award from the Medical Re- 16 Saeed SA, Farnaz S, Simjee RU and Malik A: Triterpenes and search and Advancement Foundation in Memory of Dr. Chi-Shun B-sitosterol from piper betel: isolation, antiplatelet and anti- Tsou. inflammatory effects. Biochem Soc Trans 21, 462S (1993) 17 Li WK: Extraction of alkaloids from Peganum harmala L and REFERENCES study on their antihydatid chemical composition. J Lanzhou Med Coll 22, 16 – 18 (1996) 1 Airaksinen MM and Kari I: >-Carbolines, psychoactive compounds 18 Aarons DH, Rossi GV and Orzechowski RF: Cardiovascular in the mammalian body. Part I: occurrence, origin and metabo- actions of three harmala alkaloids: harmine, harmaline, and lism. Med Biol 59, 21 – 34 (1981) harmalol. J Pharm Sci 66, 1244 – 1248 (1977) 2 Beck O and Faull KF: Concentrations of the enantiomers of 19 Shi CC, Chen SY, Wang GJ, Liao JF and Chen CF: Vasorelax- 5-hydroxymethtryptoline in mammalian urine: implications for ant effect of harman. Eur J Pharmacol 390, 319 – 325 (2000) in vivo biosynthesis. Biochem Pharmacol 35, 2636 – 2639 20 Wang GJ, Shan J, Pang PKT, Yang MC, Chou CJ and Chen CF: (1986) The vasorelaxing action of rutaecarpine: direct paradoxical 3 Poirier LJ, Sourkes TL, Bouvier G, Boucher R and Carabin S: effects on intracellular calcium concentration of vascular smooth Striatal amines, experimental tremor and the effect of harmaline muscle and endothelial cells. J Pharmacol Exp Ther 276, 1016 – in the monkey. Brain 89, 37 – 52 (1966) 1021 (1996) 4 Lutes J, Lorden JF, Beales M and Lotmans GA: Tolerance to 21 Wang GJ, Wu XC, Chen CF, Lin LC, Huang YT, Shan J and the tremorogenic effects of harmaline: Evidence for altered Pang PKT: The vasorelaxing action of rutaecarpine: effects of olivo-cerebellar function. Neuropharmacology 27, 849 – 855 rutaecarpine on calcium channel activities in vascular endothe- (1988) lial and smooth muscle cells. J Pharmacol Exp Ther 289, 1237 – 5 Bruinvels J and Sourkes TL: Influence of drugs on the tempera- 1244 (1999) ture-lowering effect of harmaline. Eur J Pharmacol 4, 31 – 39 22 Yu LL, Liao JF, Chou CJ and Chen CF: Evaluation of the (1968) water extract and some pure principles of Evodiae Fructus on 6 Zetler G, Back G and Iven H: Pharmacokinetics in the rat of 1,4-dihydropyridine binding site of L-type calcium channels. the hallucinogenic alkaloids harmine and harmaline. Naunyn J Chin Med 6, 149 – 157 (1995) Schmiedebergs Arch Pharmacol 285, 273 – 292 (1974) 23 Chiou WF, Liao JF, Shum YC and Chen CF: Mechanisms of 7 Grella B, Dukat M, Young R, Teitler M, Herrick-Davis K, vasorelaxant effect of dehydroevodiamine: a bioactive iso- Gauthier CB and Glennon RA: Investigation of hallucinogenic quinazolinocarboline alkaloid of plant origin. J Cardiovasc Phar- and related beta-carbolines. Drug Alcohol Depend 50, 99 – 107 macol 27, 845 – 853 (1996) (1998) 24 Cheng YC and Prusoff WH: Relationship between the inhibition 8 Nelson DL, Herbet A, Petillot Y, Pichat L, Glowinski J and constant (Ki) and the concentration of inhibitor which causes 3 Hamon M: [ H]Harmaline as a specific ligand of MAOA – I. 50 percent inhibition (IC50) of an enzymatic reaction. Biochem Properties of the active site of MAO A from rat and bovine Pharmacol 22, 3099 – 3108 (1973) brains. J Neurochem 32, 1817 – 1827 (1979) 25 Graham RM, Perez DM, Hwa J and Piascik MT: a 1-Adrenergic 9 Fuller RW, Wong CJ and Hemrick-Luecke SK: MD 240928 and receptor subtypes. Molecular structure, function, and signaling. harmaline: Opposite selectivity in antagonism of the inactivation Circ Res 78, 737 – 749 (1996) of types A and B monoamine oxidase by pargyline in mice. Life 26 Karaki H, Kishimoto T, Ozaki H, Sakata K, Umeno H and Sci 38, 409 – 412 (1986) Urakawa N: Inhibition of calcium channels by harmaline and 10 May T, Rommelspacher H and Pawlik M: [3H]Harman binding other harmala alkaloids in vascular and intestinal smooth experiments. I: A reversible and selective radioligand for muscles. Br J Pharmacol 89, 367 – 375 (1986) monoamine oxidase subtype A in the CNS of the rat. J Neuro- 27 Schramm M, Thomas G, Towart R and Franckowiak G: Novel chem 56, 490 – 499 (1991) dihydropyridines with positive inotropic action through activa- 11 Loew GH, Nienow J, Lawson JA, Toll L and Uyeno ET: tion of Ca2+ channels. Nature 303, 535 – 537 (1983)