Abstract
Cardiac-valve regurgitation observed in Parkinson patients treated with the ergoline dopamine receptor agonist 8β-methylthiomethyl-6-propylergoline (pergolide) has been associated with the agonist efficacy of the drug at 5-hydroxytryptamine2B (5-HT2B) receptors. 5-HT2A receptors may also play a role in pergolide-induced cardiac-valve regurgitation. We studied the pharmacological profile of pergolide and eight derivatives in porcine vascular rings endowed with 5-HT2B and 5-HT2A receptors to detect the molecular fragment of the pergolide molecule that may be responsible for agonism at these receptors. Pergolide derivatives showed a different substitution pattern at N(6), and the side chain at C(8) was modified by replacement of the sulfur against an oxygen atom. We demonstrate that the potent agonism of pergolide both at 5-HT2B and 5-HT2A receptors is retained when the N(6) propyl substituent is replaced by ethyl. However, agonism can be converted into antagonism if N(6) propyl is replaced by methyl. The N(6)-unsubstituted derivative was a low efficacy 5-HT2B receptor partial agonist and a 5-HT2A receptor antagonist. This pharmacological pattern was also applicable for pergolide congeners with an oxygen in the side chain at C(8). 6-Methylpergolide retained agonist efficacy and potency compared with pergolide at human (h) D2LONG(L) and hD2SHORT(S) receptors as determined by guanosine 5′-O-(3-[35S]thio)triphosphate binding. Based on the ability of pergolide to produce potent agonism at 5-HT2B receptors and the failure of 6-methylpergolide to act as an agonist but as a potent antagonist, we conclude that the N(6) propyl substituent of pergolide is crucial for 5-HT2B receptor agonism and thus a determinant of valvular regurgitation.
8β-Methylthiomethyl-6-propylergoline (pergolide) is a potent agonist of the D1 and D2 class of dopamine receptors (Fuller and Clemens, 1991) and has been used in the treatment of Parkinson's disease (PD) (Standaert and Young, 2005). Pergolide was approved for use in 1989. Two studies recently showed that patients with PD who were treated with pergolide had an increased chance of serious damage to their heart valves when compared with patients who did not receive the drug. The risk was high among patients who had taken daily doses of pergolide that exceeded 3 mg for 6 or more months (Schade et al., 2007; Zanettini et al., 2007). As a consequence, pergolide was retracted from the U.S. market in 2007.
The incidence of the adverse effect (cardiac-valve regurgitation) caused by pergolide is predominantly associated with an involvement of serotonin 5-HT2B receptor activation in cardiac valves (Roth, 2007). Stimulation of this type of receptors initiates myofibroblast proliferation within the tissue and in turn thickening, retraction, and stiffening of the valves, resulting in incomplete leaflet coaptation and valve regurgitation (Setola et al., 2003; Baseman et al., 2004; Horvath et al., 2004; Van Camp et al., 2004; Schade et al., 2007; Zanettini et al., 2007). Although activation of 5-HT2B receptors has been shown to be necessary to produce cardiac valvulopathy (Rothman et al., 2000), a participation of 5-HT2A receptors in the development of pergolide-induced cardiacvalve regurgitation cannot be completely ruled out (Fitzgerald et al., 2000). Several lines of evidence suggest that: 1) 5-HT2A receptor activation induces mitogenic responses (Xu et al., 2002); 2) human heart valves have high mRNA levels of 5-HT2A receptors (Fitzgerald et al., 2000); 3) stimulation of [3H]thymidine deoxyribose incorporation into newly synthesized DNA was not completely blocked by the 5-HT2B/2C receptor antagonist SB206553 (Setola et al., 2003); and 4) pergolide behaved as a potent full agonist at recombinant human 5-HT2A receptors (Newman-Tancredi et al., 2002).
Pergolide is an ergot alkaloid derivative that has recently been reported to act as a full 5-HT2B receptor agonist showing similar potency at recombinant human and native porcine 5-HT2B receptors (Newman-Tancredi et al., 2002; Jähnichen et al., 2005). In contrast to the agonist properties of pergolide, another ergot alkaloid derivative, lisuride, which is also used to treat PD, was devoid of agonist activity both at human and porcine 5-HT2B receptors (Newman-Tancredi et al., 2002; Jähnichen et al., 2005). No link has been found in the literature between lisuride use and fibrotic cardiac valvulopathy, in agreement with the 5-HT2B receptor antagonist effect of this drug (Hofmann et al., 2006). Unlike the antagonist properties of lisuride at 5-HT2B receptors this ergot compound acted as a partial agonist at recombinant human 5-HT2A receptors (Jerman et al., 2001; Newman-Tancredi et al., 2002). This finding would seem to rule out a prominent role for the 5-HT2A receptor in human heart valve disease.
The observation that pergolide behaved as a potent full 5-HT2B receptor agonist prompted us to focus on the structural fragment of the pergolide molecule that might be responsible for agonism. We modified the substituents at N(6) and C(8). To study 5-HT2B receptor-mediated effects of pergolide and its derivatives, we used the porcine isolated pulmonary artery in which 5-HT has been shown to elicit endothelium-dependent relaxation (Glusa and Pertz, 2000). In addition, we wanted to know the influence of structural modification on the pharmacological properties of pergolide at 5-HT2A receptors that might be involved, at least in part, in heart valve damage induced by the drug (see above). To study 5-HT2A receptor-mediated effects, we used the porcine isolated coronary artery in which 5-HT has been shown to elicit contractile responses (Cushing and Cohen, 1993). To show whether structural modification of the pergolide molecule would affect the efficacy at dopamine D2 receptors, which are the major therapeutic target of pergolide, we also studied the effects of selected compounds at recombinant hD2S and hD2L receptors, stably expressed in Chinese hamster ovary (CHO) cells, by measuring G-protein activation using a [35S]GTPγS binding assay.
We examined the effects of pergolide and eight derivatives of the drug that showed the following structural features: the propyl group at N(6) of pergolide was replaced by ethyl, methyl, or hydrogen. Furthermore, the sulfur located in the side chain attached to position 8 of pergolide was substituted by an oxygen, and this drug was also modified by replacing the N(6) propyl group by ethyl, methyl, or hydrogen. In this series, the N(6) propyl group was additionally replaced by a bulky cyclopropylmethyl substituent (ninth compound) (Fig. 1).
Materials and Methods
Tissue Preparation. All experimental procedures carried out in the present study where within the Guidelines of the Animals (Scientific Procedures) Act 1986. Lungs and hearts from pigs were obtained from the local slaughterhouse (Lehrund Versuchsanstalt für Tierzucht und Tierhaltung, Teltow Ruhlsdorf, Germany). Small branches of pulmonary arteries were dissected from the lungs. The left anterior descending artery (Ramus interventricularis anterior) and the right coronary artery (arteria coronaria dextra) were dissected from hearts of pigs used for studies on pulmonary arteries. Vessels were placed in Krebs-Henseleit solution of the following composition: 118 mM NaCl, 4.7 mM KCl, 2.5 mM CaCl2 (1.6 mM for coronary arteries), 1.2 mM MgSO4, 1.2 mM KH2PO4, 25 mM NaHCO3, and 10 mM d-glucose, pH 7.4. The solution was aerated with 95% O2/5% CO2. After removal of fat and connective tissue, the vessels were cut into rings (coronary artery, 3 to 4-mm long and 2 to 3-mm wide; pulmonary arteries, 2 to 3-mm long and 2-mm wide). In experiments with coronary artery, the intimal surface of the rings was gently rolled with a pair of tweezers to destroy the endothelium. Rings of both tissues were horizontally suspended between two L-shaped stainless steel hooks (300-μm diameter) and mounted in a water-jacketed 20-ml organ chamber filled with Krebs-Henseleit solution for measurement of isometric changes in tension. The solution was kept at 37°C and continuously aerated with 95% O2/5% CO2, pH 7.4.
Porcine Pulmonary Artery (Functional 5-HT2B Receptor Assay). Resting tension was adjusted to 20 mN at the beginning of the experiment. During an initial stabilization period of 60 min, the bathing medium was replaced once after 30 min. The tissue rings were then stimulated at intervals of 45 min once with KCl (30 mM) and three times with U46619 (0.01 μM) until the contractile response had become constant. The presence of endothelium was verified by the ability of bradykinin (0.01 μM) to cause relaxation after the second contraction with U46619. Resting tension was readjusted after the contraction with KCl (30 mM) and after the first contraction with U46619 (0.01 μM). The relaxant response to 5-HT (or ergot alkaloid derivative) was studied after the third 0.01 μM U46619-induced contraction had stabilized. In agonist experiments with ergot alkaloid derivatives, a noncumulative concentration-response curve to the agonist (0.3–1000 nM) was established by adding only one concentration of agonist to each tissue. This method was used because it is known that many tissues respond only to the first concentration of ergots, and the cumulative concentration-response technique cannot be applied (Müller-Schweinitzer, 1990). Agonist experiments were performed in the absence or presence of the selective serotonin 5-HT2B receptor antagonist SB204741 (3 μM) added 30 min before the construction of the concentration-response curve. In experiments where the antagonist or partial agonist properties of ergot alkaloid derivatives were studied, a cumulative concentration-response curve to 5-HT was constructed on each tissue 60 min after the addition of the ergot alkaloid derivative. When the maximal relaxant response to 5-HT or the test agonist had been attained, relaxation was accomplished by the addition of bradykinin (0.01 μM). Relaxant effects were expressed as a percentage of the relaxation induced by the agonist plus bradykinin. All experiments were performed in the continuous presence of ketanserin (0.1 μM) to block 5-HT2A receptors.
Porcine Coronary Artery (Functional 5-HT2A Receptor Assay). Resting tension was adjusted to 20 mN at the beginning of the experiment. The tissues were stabilized for 60 min with replacement of the bathing medium after 30 min. During the following equilibration period (115 min), the vessels were stimulated twice with KCl (50 mM) for 30 min with a washing period of 10 min between each KCl challenge. The absence of endothelium was verified by the failure of bradykinin (0.1 μM) to cause relaxation after the second contraction with KCl. Resting tension was readjusted to 20 mN after the first contraction with KCl (50 mM). Cumulative concentration-response curves to 5-HT or ergot alkaloid derivative were constructed in the absence or presence of ketanserin (0.01 μM) until a maximal response was observed. The full pA2 value for ketanserin against 5-HT was estimated from the Schild plot. pA2 of 8.88 ± 0.03 (slope, 1.05 ± 0.05; not significantly different from unity, n = 7) was in agreement with the pA2 of 8.7 ± 0.04 estimated by Cushing and Cohen (1993) in this tissue. In separate experiments, a cumulative concentration response curve to 5-HT was established in the absence or presence of a single concentration of an ergot alkaloid derivative. Antagonists were added to the bathing medium 60 min before the construction of a concentration-response curve. Contractile effects were expressed as a percentage of the second KCl-induced contraction. All experiments were performed in the continuous presence of prazosin (0.1 μM), cocaine (6 μM), and indomethacin (5 μM) to block α1-adrenoceptors and to inhibit neuronal uptake of 5-HT and vascular eicosanoid production by cyclooxygenase, respectively.
[35S]GTPγS Binding (Functional Dopamine D2 Receptor Assay). Efficacy of selected compounds at the dopamine receptor subtypes D2L and D2S was investigated in a [35S]GTPγS assay as previously described (Schlotter et al., 2005) using membranes of CHO cells stably expressing hD2L and hD2S receptors (Hayes et al., 1992). Homogenates of membranes were prepared according to the literature (Hübner et al., 2000) with receptor densities of Bmax = 1.11 and 1.44 pmol/μg for D2L and D2S, respectively, diluted in HEPES buffer (20 mM HEPES, 10 mM MgCl2, 100 mM NaCl, pH 7.4), and incubated at 37°C with 1 μM GDP (HEPES buffer) and the test compound (in HEPES buffer supplemented with 0.1 mM dithiothreitol) applying eight different concentrations (0.01–100,000 nM) as hexaduplicates at a final volume of 200 μl in 96-well microplates. After 30 min, 0.1 nM [35S]GTPγS (specific activity, 1250 Ci/mmol; PerkinElmer Life and Analytical Sciences, Boston, MA) was added, and the incubation was continued for a further 30 min. The experiment was terminated by rapid filtration through GF/B filters using an automated cell harvester, the filters were washed five times with ice-cold washing buffer (140 mM NaCl, 10 mM KCl, 1.5 mM KH2PO4, 8 mM Na2HPO4, pH 7.4), dried at 60°C for 3 h, and the trapped radioactivity was counted in a microplate scintillation counter. For all experiments, the maximal stimulation induced with 10,000 nM quinpirole was approximately 35% over basal activity. [35S]GTPγS binding data were normalized from each individual experiment related to the full effect of the reference agonist quinpirole.
Drugs. U46619 was obtained as a gift from Upjohn (Kalamazoo, MI). The following drugs were purchased: bradykinin triacetate, indomethacin, prazosin hydrochloride, and quinpirole hydrochloride (Sigma-Aldrich, Taufkirchen, Germany). Cocaine hydrochloride was from Merck (Darmstadt, Germany). 5-HT and ketanserin tartrate were from Janssen (Beerse, Belgium). SB204741 was from Tocris (Bristol, UK). Pergolide mesylate, 6-norpergolide, 6-nor-6-methylpergolide (6-methylpergolide), 6-nor-6-ethylpergolide (6-ethylpergolide), 8β-(methyloxymethyl)-6-propylergoline (O-pergolide), 8β-(methyloxymethyl)ergoline (6-nor-O-pergolide), 8β-(methyloxymethyl)-6-methylergoline (6-methyl-O-pergolide), 8β-(methyloxymethyl)-6-ethylergoline (6-ethyl-O-pergolide), and 8β-(methyloxymethyl)-6-cyclopropylmethylergoline (6-cyclopropylmethyl-O-pergolide) were synthesized by Alfarma s.r.o. (Cernosice, Czech Republic) according to previously published procedures with minor modifications (Misner et al., 1997; Anastasia et al., 2001; Cabri et al., 2006).
Drugs were dissolved in distilled water, dimethylsulfoxide (SB204741; ergot alkaloid derivatives in the [35S]GTPγS assay), 50% (v/v) ethanol (indomethacin and prazosin), or ethanol (U46619) to a 1 to 30 mM stock solution. In 5-HT receptor assays, ergot alkaloid derivatives were made soluble in a mixture of 50% (v/v) ethanol and an equimolar amount of 1N HCl. Stock solutions were stored at –18°C and freshly diluted in distilled water, ethanol (SB204741), or in HEPES buffer before the beginning of the experiment. Final concentrations of ethanol and dimethylsulfoxide present in the organ bath did not exceed 0.1 and 0.01%, respectively.
Data Analysis and Presentation. Data are presented as mean values ± S.E.M. for n animals or n individual experiments. Concentration-response curves were fitted to the Hill equation by an iterative least-squares method (GraphPad Prism 4.0; GraphPad Software Inc., San Diego, CA) to provide estimates of the maximum response Emax (percentage of the maximal response to a reference compound) and the half-maximal effective concentration pEC50 (the negative logarithm of the agonist concentration producing 50% of the maximal response). Affinities of partial agonists (–log KP = pKP values) were estimated in agonist experiments according to the method of Kenakin (1993) and in antagonist experiments according to the method of Marano and Kaumann (1976). Using the method of Kenakin (1993), we estimated the equilibrium dissociation constant KP for the partial agonist/receptor complex by comparing equiactive molar concentrations of the full agonist A (5-HT) and the partial agonist P (ergot derivative) according to the equation: c(A) = m · c(A)/c(P) + b with m = KP/(ϵP/ϵA – 1), where c(A) is the molar concentration of A, c(P) the molar concentration of P, m the slope, and b the ordinate intercept of the regression line of c(A) versus c(A)/c(P). ϵA and ϵP represent the intrinsic efficacies of A and P, respectively. If ϵP << ϵA, pKP = –log KP can be calculated from: –log KP = log m. Using the method of Marano and Kaumann (1976), we estimated the equilibrium dissociation constant KP for the partial agonist/receptor complex by comparing equiactive molar concentrations of the full agonist A (5-HT) in the absence and presence of the partial agonist P (ergot derivative) according to the equation: c(A) = m · c(A)* + b, with m = 1/[1 + (1 – ϵP/ϵA) · c(P)/KP], where c(A) is the molar concentration of A in the absence of P, c(A)* the molar concentration of A in the presence of P, m is the slope of a weighted regression line of c(A) versus c(A)*, b is the ordinate intercept, and c(P) is the molar concentration of P. If ϵP << ϵA, pKP = –log KP can be calculated from: log [(1/m) – 1] = log c(P) – log KP. Antagonist affinities of silent antagonists were expressed as an apparent pA2 value. pA2 was calculated from a single concentration of antagonist using the following equation: pA2 = –log c(B) + log (r – 1), where c(B) is the molar concentration of the antagonist, and r is the ratio of agonist EC50 determined in the presence and absence of the antagonist (Furchgott, 1972). The full pA2 value for ketanserin against 5-HT was determined according to the method of Arunlakshana and Schild (1959). Student's t test (unpaired, two-tailed) was used to assess differences between two mean values, with P < 0.05 being considered as significant. When more than two groups of treatments were compared, a one-way analysis of variance was performed.
Results
Effects of Pergolide and Its Derivatives at Endothelial 5-HT2B Receptors in Porcine Pulmonary Arteries. Derivatives of pergolide and O-pergolide with an ethyl or cyclopropylmethyl group at N(6) caused relaxant responses similar to 5-HT, pergolide, and O-pergolide in porcine pulmonary arteries. Pergolide, 6-ethylpergolide, O-pergolide, 6-ethyl-O-pergolide, and 6-cyclopropylmethyl-O-pergolide behaved as full agonists in this tissue because their Emax values did not differ from the Emax value of the respective 5-HT control curve. The rank order of agonist potency was: 5-HT ≈ pergolide > O-pergolide ≈ 6-cyclopropylmethyl-O-pergolide ≈ 6-ethylpergolide ≈ 6-ethyl-O-pergolide. The relaxant response to pergolide has recently been shown to be inhibited by the 5-HT2B/2C receptor antagonist SB206553 (1 μM; Jähnichen et al., 2005). Relaxations to 6-ethylpergolide, O-pergolide, 6-ethyl-O-pergolide, and 6-cyclopropylmethyl-O-pergolide were inhibited by the selective 5-HT2B receptor antagonist SB204741 (3 μM) (Fig. 2). The estimated pA2 values for SB204741 were in the same range as the pA2 for SB204741 against 5-HT and argue for an involvement of 5-HT2B receptors in the relaxant response to the drugs (Table 1). 6-Norpergolide (0.3 μM) and 6-nor-O-pergolide (1 μM) elicited slight relaxations (Emax, 8 ± 1 and 12 ± 2%, respectively) that were totally abolished by SB204741 (3 μM; data not shown). However, 6-methylpergolide (0.01 μM) and 6-methyl-O-pergolide (0.03 μM) failed to show relaxation but antagonized the relaxant response to 5-HT (Fig. 3). High concentrations (1 μM) of 6-methylpergolide and 6-methyl-O-pergolide induced slight relaxations. Relaxation to 6-methylpergolide (1 μM) of 15 ± 4% (n = 6) remained unaffected in the presence of SB204741 (P > 0.05, n = 6), whereas relaxation to 6-methyl-O-pergolide (1 μM) of 19 ± 4% (n = 6) was reduced to 7 ± 2% by SB204741 (P = 0.04, n = 6). Agonist and antagonist effects of the drugs are summarized in Table 1.
Effects of Pergolide and Its Derivatives at Smooth Muscle 5-HT2A Receptors in Porcine Coronary Arteries. Pergolide, 6-ethylpergolide, O-pergolide, 6-ethyl-O-pergolide, and 6-cyclopropylmethyl-O-pergolide produced concentration-dependent contractile effects in porcine coronary arteries with pEC50 values higher than that of 5-HT. The rank order of agonist potency was O-pergolide > pergolide > 6-cyclopropylmethyl-O-pergolide ≈ 6-ethylpergolide ≈ 6-ethyl-O-pergolide > 5-HT. The drugs were partial agonists relative to 5-HT (Fig. 4). The pEC50 value for the respective partial agonist was in good agreement with the respective –log KP = pKP value calculated from equiactive concentrations of the partial agonist and 5-HT (Table 2). The Emax values of the partial agonists did not significantly differ among themselves (P > 0.05, one-way analysis of variance, n = 4–5). The 5-HT2A receptor antagonist ketanserin (0.01 μM) inhibited the contractile response to the agonists. The estimated pA2 values for ketanserin against the compounds were in the same range as the pA2 value of the antagonist against 5-HT and argue for an involvement of 5-HT2A receptors in the contractile response to the compounds (Table 2). As expected from classic partial agonists, pergolide, 6-ethylpergolide, O-pergolide, 6-ethyl-O-pergolide, and 6-cyclopropylmethyl-O-pergolide caused antagonism of the contractile response to 5-HT (Fig. 4, insets). The estimated –log KP = pKP value for the respective partial agonist-5-HT2A receptor complex matched the pEC50 value for the respective partial agonist-induced contraction (Table 2). In contrast, 6-methylpergolide and 6-methyl-O-pergolide did not induce contractions up to a concentration of 1 μM. 6-Methylpergolide (0.02 μM) and 6-methyl-O-pergolide (0.02 μM) behaved as surmountable antagonists of the 5-HT response in porcine coronary arteries (Fig. 5). The two drugs with a hydrogen at N(6) of the ergoline molecule, 6-norpergolide (1 μM) and 6-nor-O-pergolide (1 μM), acted as antagonists as well. Both compounds shifted the concentration-response curve to 5-HT to the right and depressed the maximal 5-HT response (Fig. 5). Agonist and antagonist effects of these drugs are summarized in Table 2.
Effects of Pergolide, 6-Methylpergolide, and 6-Norpergolide at hD2S and hD2L Receptors, Stably Expressed in CHO Cells. Pergolide and 6-methylpergolide behaved as full agonists compared with the selective dopamine D2 receptor agonist quinpirole at hD2S and hD2L receptors (Fig. 6). Substitution of the propyl group at N(6) (pergolide) against a methyl group (6-methylpergolide) did not affect agonist potency either at hD2S or at hD2L receptors (Table 3). The pergolide metabolite 6-norpergolide was a high-efficacy partial agonist at both sites exhibiting a 10-fold lower agonist potency compared with the parent drug pergolide (Table 3; Fig. 6).
Discussion
Among dopamine receptor agonists, the ergot alkaloid derivatives bromocriptine, pergolide, and cabergoline are used in the treatment of PD and hyperprolactinemia. From a structural viewpoint, pergolide and cabergoline do not possess a N(6) methyl substituent, which is characteristic of bromocriptine and other therapeutically used ergot alkaloid derivatives. However, pergolide bears a N(6) propyl group and cabergoline a similarly sized N(6) allyl group. In the treatment of PD with pergolide, cabergoline, and bromocriptine, the antiparkinsonian effect of pergolide was the strongest, whereas cabergoline showed the longest duration of action (Arai et al., 1995; Deleu et al., 2002). These findings may be associated with the enhanced hydrophobicity of pergolide and cabergoline induced by the introduction of bulkier N(6) substituents such as propyl or allyl (Seeman et al., 1985). Numerous reports on ergot alkaloid derivatives used in the treatment of PD have shown that restrictive cardiac valvulopathies predominantly occurred in the therapy with pergolide and cabergoline, and 5-HT2B receptor stimulation by these drugs seems to be consistent with their potential valve-damaging effect (Pritchett et al., 2002; Horvath et al., 2004; Van Camp et al., 2004). In contrast, only one case of severe tricuspid-valve regurgitation was reported after 5 years of therapy with bromocriptine (Serratrice et al., 2002). Using a population-based cohort comprising 11,417 subjects, no patients with newly diagnosed valve regurgitation were identified among those treated with bromocriptine (Schade et al., 2007). Based on these observations, we hypothesize that the N(6) propyl or allyl substituent plays a role as a determinant of 5-HT2B receptor activation and in turn of valvular regurgitation. Further evidence that the N(6) substituent may play a crucial role for 5-HT2B receptor activation has recently been provided using the ergot alkaloid derivative proterguride. This compound, which is the N(6)-propyl derivative of terguride, acted as a high-efficacy partial agonist at 5-HT2B receptors in porcine pulmonary arteries, whereas the parent compound, terguride, with a methyl group at N(6), was devoid of intrinsic efficacy (Jähnichen et al., 2005; Schurad et al., 2006).
We have recently shown that pergolide and cabergoline both behaved as potent full agonists at native porcine 5-HT2B receptors (Jähnichen et al., 2005). Agonist potencies and efficacies of pergolide and cabergoline were in line with those at recombinant human 5-HT2B receptors (Newman-Tancredi et al., 2002). The main aim of the present study was to clarify whether the effect of different N(6) substituents on the efficiency of pergolide derivatives to act as agonists or antagonists follows a predictable pattern. Therefore, we replaced the propyl substituent of pergolide with ethyl, methyl, and hydrogen, respectively. The pharmacological profile of the N(6)-unsubstituted pergolide derivative is of special interest in the therapy with pergolide because the drug has been shown to be transformed into the N-despropyl metabolite, 6-norpergolide, in humans (Rubin et al., 1981). We further replaced the sulfur of pergolide, which is rare within the family of ergot compounds, by oxygen to find out whether such a modification of the side chain at C(8) might affect the pharmacological profile of differentially N(6)-substituted pergolide derivatives.
The present study shows that the effect of N(6) substituents on intrinsic efficacy for pergolide and O-pergolide derivatives at 5-HT2B receptors in porcine pulmonary arteries followed a certain pattern; compounds with propyl, ethyl, or cyclopropylmethyl as N(6) substituent were full 5-HT2B receptor agonists, those with methyl at N(6) were silent antagonists, and 6-nor derivatives were partial agonists of minimal efficacy. Thus, the replacement of a propyl against a methyl group at N(6) of the ergoline skeleton indeed is able to convert full agonism into silent antagonism at 5-HT2B receptors in porcine pulmonary arteries. It is not clear whether the partial agonist activity of the pergolide metabolite, 6-norpergolide, is of clinical relevance. However, it should be emphasized that intrinsic activity may vary depending on the tissue used, the species, and the degree of receptor/G-protein reserve (Hoyer and Boddeke, 1993). Derivatives of O-pergolide showed similar pharmacological effects compared with their analogs in the pergolide series. Therefore, we hypothesize that the sulfur atom is not a determinant for 5-HT2B receptor agonism.
Pergolide and its derivatives showed a similar pharmacological profile at 5-HT2A receptors in porcine coronary artery compared with their effects at porcine 5-HT2B receptors. This is not unexpected because ergot alkaloids are drugs that possess low discriminatory power among the subtypes of the 5-HT2 receptor family, 5-HT2A, 5-HT2B, and 5-HT2C (Pertz and Eich, 1999). Compounds with propyl, ethyl, and cyclopropylmethyl substituents at N(6) were partial agonists compared with 5-HT at 5-HT2A receptors, whereas compounds with a methyl group at N(6) and 6-nor derivatives behaved as silent antagonists of the 5-HT contractile response. This was true irrespective of whether the side chain attached at C(8) of the compounds contained a sulfur or an oxygen atom. Thus, the sulfur in the side chain is not crucial for 5-HT2A receptor agonism of pergolide. The pharmacological profile of 6-methylpergolide was in good agreement with that of 6-methyl-O-pergolide at 5-HT2A receptors; these compounds behaved as silent and surmountable antagonists of the contractile 5-HT response, showing nanomolar affinity for the 5-HT2A receptor. Pergolide has previously been described to act as a full agonist at recombinant human 5-HT2A receptors expressed in CHO cells showing an agonist potency that was 8-fold higher than the agonist potency estimated in the present study (Newman-Tancredi et al., 2002). It is not unusual that differences in efficacy and agonist potency exist between studies using cells transfected with recombinant receptors and those using intact native receptors. Overexpression of receptors in cells that do not normally synthesize those receptors may produce aberrant G-protein coupling or receptor cross-talk (Sanders-Bush and Canton, 1995) with the consequence that efficacy and/or agonist potency may be enhanced. It should be mentioned that ergolines have previously been shown to possess nearly identical affinities for pig and human 5-HT2A receptors (Nelson et al., 1993). Moreover, pig and human 5-HT2A receptors have been reported to possess 97% sequence homology (Ullmer et al., 1995).
We have shown that the existence of a sulfur atom in the side chain at C(8) of pergolide is not critical for agonism of the drug both at 5-HT2B and 5-HT2A receptors (see above), but what is the role of the stereochemistry at C(8)? In this connection, a comparison between the 8α-substituted ergot alkaloid derivative terguride and the 8β-substituted compound 6-methylpergolide justifies special mention. Both compounds with different stereochemistry at C(8) but the same N(6) substituent (methyl) were silent antagonists at porcine 5-HT2B and 5-HT2A receptors (T. Görnemann and H. H. Pertz, unpublished data; Jähnichen et al., 2005). Irrespective of the stereochemistry at C(8), antagonism was converted into agonism when the N(6) substituent was propyl instead of methyl (Jähnichen et al., 2005; Schurad et al., 2006). Thus, the N(6) substituent plays the major role as a determinant for agonism or antagonism at 5-HT2B and 5-HT2A receptors and not the stereochemistry at C(8).
The observation of the present study that the agonist efficacy of 6-methylpergolide at hD2S and hD2L receptors is retained compared with pergolide is of particular importance. It is interesting to note that the active metabolite of pergolide, 6-norpergolide, which showed minimal efficacy at 5-HT2B receptors and lacked efficacy at 5-HT2A receptors (see above), behaved as a high-efficacy partial agonist both at hD2S and hD2L receptors. The finding that the agonist character of 6-methylpergolide at D2 receptors is not significantly perturbed could legitimize the compound as a potential experimental therapeutic.
In summary, the present study shows that the potent agonism of pergolide both at 5-HT2B and 5-HT2A receptors can be converted into antagonism when the N(6) propyl substituent is replaced by a methyl substituent. Substitution of the N(6) propyl group by a methyl group neither affects efficacy nor potency at D2 receptors, which are the major therapeutic target of pergolide in the treatment of PD. It has been demonstrated in numerous studies that 5-HT2B receptor activation participates in appearance and progression of pergolide-induced cardiac-valve regurgitation, but 5-HT2A receptor activation may also be involved in this severe side effect of pergolide. Based on the ability of pergolide to produce potent agonism and the failure of 6-methylpergolide to act as an agonist but as a potent antagonist, we conclude that the N(6) propyl substituent of pergolide is the molecular fragment that is especially responsible for 5-HT2B receptor agonist effects and thus a determinant of valvular regurgitation. However, it should be emphasized that other structural features of the ergoline molecule may also contribute to the ability of these drugs to induce valvular heart disease. Admittedly, a case of mitral valve disease attributable to chronic administration of the ergopeptide derivative ergotamine, a potent agonist at 5-HT2B receptors (Glusa and Roos, 1996), has been reported (Flaherty and Bates, 1996). Therefore, further studies are needed that substantiate that 6-methylpergolide, in contrast to pergolide, fails to induce restrictive cardiac valvulopathies.
Acknowledgments
We thank Th. Paulke and M. Uwarow of the Lehrund Versuchsanstalt für Tierzucht und Tierhaltung (Teltow-Ruhlsdorf, Germany) for providing pig lungs and hearts for the studies.
Footnotes
-
This study was supported in part by grants from Europäischer Fonds für Regionale Entwicklung (Project 10021471) and EUREKA (Project OE159) of the Ministry of Education, Youth, and Sports of the Czech Republic.
-
Article, publication date, and citation information can be found at http://jpet.aspetjournals.org.
-
doi:10.1124/jpet.107.133165.
-
ABBREVIATIONS: Pergolide, 8β-methylthiomethyl-6-propylergoline; PD, Parkinson's disease; 5-HT, 5-hydroxytryptamine creatinine sulfate; SB206553, 5-methyl-1-(3-pyridylcarbamoyl)-1,2,3,5-tetrahydropyrrolo[2,3-f]indole; h, human; CHO, Chinese hamster ovary; [35S]GTPγS, guanosine 5′-O-(3-[35S]thio)triphosphate; U46619, 9,11-dideoxy-11α,9α-epoxymethanoprostagnadin F2α; SB204741, N-(1-methyl-1H-5-indolyl)-N′-(3-methyl-5-isothiazolyl)urea.
- Received October 17, 2007.
- Accepted December 19, 2007.
- The American Society for Pharmacology and Experimental Therapeutics