Comparative Effects of α-, β-, and γ-Carbolines on Platelet Aggregation and Lipid Membranes

Cigarette smoking and alcohol consumption possibly affect platelet functions. To verify the hypothesis that some α-, β-, and γ-carboline components in cigarette smoke and alcoholic beverages may change platelet aggregability, their effects on human platelets were determined by aggregometry together with investigating their membrane effects by turbidimetry. Carbolines inhibited platelet aggregation induced by five agents with the potency being 3-amino-1,4-dimethyl-5H-pyrido[4,3-b]indole > 3-amino-1-methyl-5H-pyrido[4,3-b]indole > 1-methyl-9H-pyrido[3,4-b]indole. The most potent 3-amino-1,4-dimethyl-5H-pyrido[4,3-b]indole showed 50% aggregation-inhibitory concentrations of 6–172 μM. Both γ-carbolines interacted with phosphatidylcholine membranes to lower the lipid phase transition temperature with the potency correlating to the antiplatelet activity, suggesting that the interaction with platelet membranes to increase their fluidity underlies antiplatelet effects. Given their possible concentration and accumulation in platelets, γ- and β-carbolines would provide cigarette smokers and alcohol drinkers with reduced platelet aggregability, and they may be responsible for the occurrence of hemorrhagic diseases associated with heavy smoking and alcoholics.


Introduction
Cigarette smoking has the possibility to affect the functions of platelets. Platelet aggregation stimulated by collagen is diminished in smokers compared with nonsmokers, suggesting that habitual smoking reduces platelet aggregability [1].
Smokers are more susceptible to aspirin in collagen-and adenosine 5 -diphosphate-(ADP-) induced platelet aggregation than nonsmokers [2]. Smoking also promotes the platelet inhibition mediated by an antiplatelet prodrug clopidogrel [3]. Cigarette smoking and alcohol drinking are often characterized by concurrent use, and alcohol drinking is generally accompanied by an increase in smoking. Platelet defects are also noted in alcoholics [4]. Alcohol consumption is inversely related to platelet aggregation in response to collagen, epinephrine, and ADP [5], and it reduces platelet aggregability [6,7]. However, the influences of cigarette smoking and alcohol consumption on platelet aggregability appear to be conflicting, with some studies indicating a decrease and others an increase of aggregation. Platelet aggregation induced by epinephrine and ADP is increased in habitual smokers [8,9]. The platelets of alcoholics are initially hypoaggregable but become hyperaggragable after the cessation of drinking [10]. ADP-induced platelet aggregation is increased in alcoholics [11].
Cigarette smoke contains a variety of bioactive compounds [9]. Alcoholic beverages are not consumed as pure ethanol. Therefore, some components in cigarette smoke and alcoholic beverage are assumed to influence the properties of platelets. The objective of this study was to determine antiplatelet effects of the assumed compounds and to relate cigarette smoking and alcohol consumption to hypoaggregability. In addition, their effects on lipid membranes were investigated to address one possible mode of antiplatelet action. From the viewpoint of this mechanistic membrane interaction, an opposite phenomenon, hyperaggregability, in smokers and alcoholics was also discussed.
Cigarette smoking and alcohol drinking have been referred to as risk factors for intracerebral hemorrhage, subarachnoid hemorrhage, and hemorrhagic stroke [12][13][14][15][16]. These factors correlate with each other. Reduced platelet aggregability is involved in the development of hemorrhagic 2 Journal of Toxicology complications [17,18]. The occurrence of hemorrhage is increased by heavy cigarette smoking and binge alcohol drinking [12,19]. A graded increase in risk of intracerebral hemorrhage and hemorrhagic stroke depends on how many cigarettes are smoked [13], and several chemicals in smoke are likely to relate to the increased risk of intracranial hemorrhage [20]. Heavy alcohol drinking, but not light-to-moderate alcohol intake, increases the risk of intracerebral and subarachnoid hemorrhage [12,14]. These suggest that certain antiplatelet components in cigarette smoke and alcoholic beverages are possibly associated with serious hemorrhagic diseases. This study aimed at discussing such a possibility by determining their effects on human platelets.
α-, β-, and γ-carbolines exhibit a wide range of bioactivities such as neuropsychiatric, hallucinogenic, and carcinogenic, which have been exclusively studied [33][34][35]. In addition to these known bioactivities, α-, β-, and γ-carbolines may exert some effects at the periphery, because they are found in human blood and platelets. Certain carbolines such as harmane, harmine, and harmol were recently reported to show the inhibitory effects on collagen-induced platelet aggregation [36]. The structure and antiplatelet activity relationship suggested that the presence of a methyl group in the pyrido moiety is important for carboline compounds to inhibit platelet aggregation [37].

Platelet Aggregation Assay.
The experiments were designed and performed according to the guidelines of the Japanese Pharmacological Society. Human platelet-rich plasma (PRP) and platelet-poor plasma (PPP) were prepared with citrated blood which was obtained from healthy male donors aged 38-48 years (n = 6) of nonsmokers and nonalcoholics without taking any drugs at least for one month according to the previous method [42]. Only male subjects were employed as blood donors because platelet aggregation induced by collagen, PAF and ADP was different between males and females in the study which evaluated the comparative effects of the smoking habit and the act of smoking on platelet aggregability [43]. Sharp increases of the carboline concentrations in plasma and platelets were found in nonsmokers after 5 minutes following smoking one cigarette [25][26][27] and in nonalcoholics 30 minutes after drinking alcoholic beverages [32]. Therefore, PRP and PPP were prepared from nonsmokers and nonalcoholics in order to determine the inherent effects of α-, β-, and γ-carbolines on platelet aggregation without exposing platelets to carboline compounds. The platelet count of PRP was adjusted to be 3 × 10 5 platelets/μL by diluting PRP with PPP. Both PRP and PPP were used within 3 hours after preparation.
Platelet aggregation was assayed by the same method as the aggregometry using an antiplatelet reference aspirin which was recently reported from our laboratory [37]. Since its methodological validity had been confirmed [44,45], this study was performed without using references and the obtained results were discussed with referring to our recent report [37]. The aggregation response of PRP was monitored using a HEMA Tracer 601 aggregometer (Niko Bioscience, Tokyo, Japan) by an increase of percent light transmission (% T) at 660 nm as a function of time. PPP and unstimulated PRP were defined as 100% T and 0% T, respectively. The tested carbolines were dissolved in and diluted with DMSO and/or water. After adjustment of % T to 0, 20 μL of the sample solutions were added to 170 μL of PRP. Vehicle alone was added for controls. The volume of DMSO was kept less than 0.1%-1.0% (v/v) of the total volume so as not to affect platelet aggregability and aggregation inducers. After treating PRP with each carboline for 1 minute, platelet aggregation was induced by adding 10 μL of the aqueous solutions of collagen (50 μg/mL), PAF (5 μM), epinephrine (40 μg/mL), ADP (60 μM) or thrombin (5 units/mL). The time of their addition was defined as 0 minute. A maximal % T of aggregation response (T max ), an area under curve of aggregation response (AUC, from 0 to 5 minutes) and a single slope of aggregation response were measured in platelet aggregation induced by collagen, PAF, ADP, and thrombin. Since epinephrine showed the biphasic aggregation, first T max (at 1 minute) and second T max (at 5 minutes), first AUC (from 0 to 1.5 minutes) and second AUC (from 1.5 to 5 minutes), and slopes of the first (% T at 45 seconds) and second (% T at 4 minutes) phases were measured. The tested compounds were assayed at various concentrations (1-1000 μM). Their % inhibitions were plotted against their concentrations to prepare concentration-inhibition curves, from which 50% inhibitory concentrations were calculated [46].

Membrane Interaction Assay.
The changes in turbidity of liposomal membrane suspensions reflect the thermotropic phase transition of membrane-constituting phospholipids from the gel to the liquid crystalline phase and the phase transition temperatures are altered by membrane-interacting compounds [47]. The membrane interactions of antiplatelet ADγC and AMγC were assayed as reported previously [48]. Briefly, DPPC liposomes were prepared by hydrating the dry film of DPPC (final concentration of 0.4 mM) with water, and then treated with ADγC and AMγC (0, 0.5, 1 and 1.5 mM for each). The absorption of DPPC liposome suspensions was measured at 450 nm by a UV-260 spectrophotometer equipped with a CPS thermocontroller (Shimadzu, Kyoto, Japan) as the temperature was increased. The transition between gel and liquid crystalline phases in phospholipid liposomes dispersed in the aqueous medium is accompanied by a sudden change in turbidity [49]. The phase transition temperatures defined as the mid-point of the abrupt step in absorbance were calculated.

Statistical Analysis.
All results are expressed as mean ± SE (n = 7 for antiplatelet experiments and n = 6 for membrane experiments). Data were analyzed by ANOVA, followed by post hoc Fisher's PLSD test using StatView version 5.0 (SAS Institute, Cary, NC, USA). Values of P < .05 were considered statistically significant.

Antiaggregatory Effect.
Carboline compounds affected human platelets with the potency different between subclasses. Their inhibitory effects on platelet aggregation induced by collagen, PAF, epinephrine, ADP and thrombin are shown in Figures 2, 3, 4, 5, 6, and 7. The comparisons showed that γ-carbolines were the most effective in inhibiting platelet aggregation, followed by less effective βcarboline, with α-carboline being almost ineffective at the tested concentrations (250 or 500 μM). AMαC showed a 50% inhibitory concentration of 499.6 ± 20.3 μM only for AUC of aggregation response to ADP. In contrast, ADγC and AMγC decreased T max and AUC of aggregation responses to all of collagen, PAF, epinephrine, ADP, and thrombin. Both γ-carbolines also decreased the slopes of collagen-and epinephrine-induced platelet aggregation. When comparing 50% inhibitory concentrations, ADγC was more potent than AMγC. Even at 10 μM or less, ADγC inhibitorily influenced the aggregability of platelets stimulated by PAF and epinephrine.

Discussion
The comparative studies have revealed that βand γ-carbolines, but not α-carboline, inhibit platelet aggregation induced by five different agents with the potency being ADγC > AMγC > MβC. The relation between structure and  antiplatelet activity indicates that the basic structure of 5Hpyrido [4,3-b]indole (γ-carboline) is important for inhibiting platelet aggregation, followed by the structure of 9Hpyrido [3,4-b]indole (β-carboline). An additional methyl group in the pyrido moiety provides γ-carbolines with higher activity. Since ADγC and AMγC themselves do not aggregate platelets despite partly resembling serotonin in structure [50], these γ-carbolines are referred to as potent antiplatelet compounds.
A question arises as to whether carboline components actually affect the aggregability of platelets of cigarette smokers and alcohol drinkers. α-, β-, and γ-carbolines are contained in mainstream cigarette smoke (≥22 ng/cigarette for AMαC, ≥2.2 μg/cigarette for MβC, ≥0.5 ng/cigarette for ADγC and ≥1.1 ng/cigarette for AMγC), and higher levels are found in sidestream cigarette smoke (≥3.0 μg/cigarette for MβC) [21,22,24,25]. β-carbolines are also present in various alcoholic beverages at the concentrations of ≥3.2 μM for MβC [25,51]. Compared with nonsmokers (≥23 pM), the plasma levels of MβC are higher in smokers (≥48 pM) and increase to ≥165 pM by smoking a cigarette [25]. The platelet concentrations of MβC are ≥0.015 pmol/10 9 platelets for nonsmokers but ≥0.060 pmol/10 9 platelets for smokers, and MβC shows the increasing platelet concentrations to be ≥0.142 pmol/10 9 platelets 13 minutes after smoking [26]. β-carbolines increase in blood rapidly following cigarette smoking and alcohol drinking, and their concentrations in platelets are much higher than in plasma, indicating their significant concentration and accumulation in platelets. The accurate γ-carboline concentrations in platelets of smokers and alcoholics have been unknown. In the dosing experiment using rabbits, however, ADγC was present in blood, especially in red blood cells, for a long time after oral dosing [52]. Although aspirin is less potent in aggregation inhibition than ADγC, it has been frequently used as an antiplatelet drug. Considering that cigarette smoke-and alcoholic beverage-derived βand γ-carbolines are concentrated and accumulated in platelets, the possibility for them to affect platelet aggregability is not necessarily excluded.
Although the platelet aggregation mechanisms for collagen, PAF, epinephrine, ADP, and thrombin differ, ADγC and AMγC were inhibitory on platelet aggregation induced by all of these aggregants, and MβC by two aggregants. These results indicate that γand β-carbolines influence the step common to different aggregation agonists in addition to their specific action at the receptor and enzyme levels. The physicochemical property of biomembranes, such as fluidity, plays a crucial role in signal transduction and affects the activity of platelets. Membrane fluidity modulates platelet aggregability and membrane-fluidizing compounds attenuate collagenand thrombin-induced platelet aggregation [53]. A change in platelet membrane fluidity is mechanistically related to various antiplatelet compounds [38,54].
The membrane effects of toxins and drugs have been most frequently studied by measuring fluorescence polarization of liposomal and cellular membranes labeled with fluorescent probes [37][38][39]54]. However, such a method was not applicable to antiplatelet γ-carbolines, because they are naturally fluorescent with the maximal excitation and emission wavelengths almost similar to those of typical fluores-cent probes. Therefore, turbidimetry was used for them. The changes in turbidity of liposome suspensions reflect the promotion of a gel to liquid crystalline transition of membrane phospholipids, indicating an increase of membrane fluidity [47,48]. Consequently, ADγC and AMγC were found to lower the phase transition temperature of membrane DPPC at 500 μM-1.5 mM with the potency correlating to their relative antiplatelet effects. Antiplatelet γ-carbolines appear to interact with lipid membranes and increase their fluidity at platelet aggregation-inhibitory concentrations.
The antiplatelet mechanisms previously reported for βand γ-carbolines and their relating structures include the inhibition of aggregation-relevant enzymes and receptors, and the suppression of cytosolic calcium mobilization and arachidonic acid liberation. βand γ-carbolines influence cyclooxygenase activity and arachidonic acid metabolism to reduce the production of prostaglandins and thromboxane [55]. Antiplatelet γ-carbolines (ADγC and AMγC) and β-carbolines (MβC) are the potent inhibitors of monoamine oxidase and serotonin uptake of platelets [33,50]. β-carbolines like MβC inhibit phospholipase Cγ2 and protein tyrosine phosphorylation [36]. Biological membranes require the lipid bilayer environments optimal for membrane-embedded enzymes, receptors, and transporter systems. The fluidity changes of platelet membranes modify the activities of phospholipase, cyclooxygenase, and aggregation agonists' receptors with the subsequent inhibition of phosphoinositide breakdown, and of prostaglandin and thromboxane formation [38,39]. The increased membrane fluidity also implies that aggregation-relevant receptors on platelet membranes are less exposed to the external environment.
The present results suggest that platelet hypoaggregability might be induced in smokers and alcoholics by γand β-carbolines. However, the influences of cigarette smoking and alcohol consumption on platelet aggregation have been conflicting, with some studies reporting a reduction of aggregability [1][2][3][4][5][6][7] but others an enhancement of aggregability [8][9][10][11]. In contrast to fluid membranes (with increased fluidity) induced by ADγC and AMγC, rigid membranes (with decreased fluidity) show the enhanced platelet aggregability in response to epinephrine and thrombin [40,41]. The decreased membrane fluidity renders platelet receptors more exposed to the external environment and makes the binding of agonists to the receptors more efficiently, resulting in an increase of platelet sensitivity to aggregants (hyperaggregability) [56,57]. Varying lipid compositions modify the fluidity of biomembranes and cholesterol is one of determinants for decreasing membrane fluidity. While the enhanced platelet aggregability could be found in chronic smokers, subjects who had smoked 10 ± 2 cigarettes per day for 7-10 years showed the decreased fluidity of platelets, which was due to an increase of cholesterol in platelet membranes [58]. Platelet aggregability in response to ADP and collagen was enhanced in alcoholics with increasing cholesterol in platelet membranes [11]. The conflicting phenomena, hypoaggregability and hyperaggregability, associated with cigarette smoking and alcohol consumption appear to be explained by the biphasic effects of membrane fluidity changes.
Reduced platelet aggregability is important in the development of intracranial hemorrhage [17,28]. Antiplatelet medication is known to increase the incidence of intracerebral hemorrhage, hemorrhagic stroke and other hemorrhagic complication [59,60] and also increase the recurrence risk of intracerebral hemorrhage [61]. Intracerebral hemorrhage, subarachnoid hemorrhage, and hemorrhagic stroke are closely associated with smoking and drinking habits [12][13][14][15][16]. The present results suggest the possibility that γand β-carboline components in cigarette smoke and/or alcoholic beverages are pharmacotoxicologically relevant to these serious hemorrhagic diseases as well as antiplatelet agents by decreasing platelet aggregation responses to different aggregants.
Another etiological role of cigarette smoking and alcohol consumption has been indicated in cardiovascular events [62,63] although it is conflicting similarly to their influence on platelet aggregability. Platelet aggregation is pathologically related to coronary artery disease and coronary thrombosis leads to myocardial infarction. The antiplatelet effects of γand β-carbolines are in line with several studies that cigarette smoking and alcohol drinking protect against coronary heart disease and also lower the risk of coronary artery disease and myocardial infarction [63][64][65]. However, such effects are inconsistent with other studies that cigarette smoking is a risk factor for coronary artery disease and binge alcohol drinking precipitates fatal myocardial infarction [62,64]. Chronic smokers and heavy drinkers show an increase of cholesterol in platelet membranes [6,58], by which the membrane fluidity of platelets is decreased. The aggregability should be enhanced in such rigid platelet membranes, increasing cardiovascular morbidity.
In conclusion, the reduced platelet aggregability found in cigarette smokers and alcoholic drinkers would be attributable to antiplatelet carboline components in cigarette smoke and alcoholic beverages. γand β-carbolines may be responsible for the occurrence of hemorrhagic diseases associated with heavy smoking and alcoholics by inhibiting platelet aggregation.