The role of adenosine receptors and endogenous adenosine in citalopram-induced cardiovascular toxicity.

AIM: We investigated the role of adenosine in citalopram-induced cardiotoxicity.
MATERIALS AND METHODS: Protocol 1: Rats were randomized into four groups. Sodium cromoglycate was administered to rats. Citalopram was infused after the 5% dextrose, 8-Cyclopentyl-1,3-dipropylxanthine (DPCPX; A1 receptor antagonist), 8-(-3-chlorostyryl)-caffeine (CSC; A2a receptor antagonist), or dimethyl sulfoxide (DMSO) administrations. Protocol 2: First group received 5% dextrose intraperitoneally 1 hour prior to citalopram. Other rats were pretreated with erythro-9-(2-hydroxy-3-nonyl) adenine (EHNA; inhibitor of adenosine deaminase) and S-(4-Nitrobenzyl)-6-thioinosine (NBTI; inhibitor of facilitated adenosine transport). After pretreatment, group 2 received 5% dextrose and group 3 received citalopram. Adenosine concentrations, mean arterial pressure (MAP), heart rate (HR), QRS duration and QT interval were evaluated.
RESULTS: In the dextrose group, citalopram infusion caused a significant decrease in MAP and HR and caused a significant prolongation in QRS and QT. DPCPX infusion significantly prevented the prolongation of the QT interval when compared to control. In the second protocol, citalopram infusion did not cause a significant change in plasma adenosine concentrations, but a significant increase observed in EHNA/NBTI groups. In EHNA/NBTI groups, citalopram-induced MAP and HR reductions, QRS and QT prolongations were more significant than the dextrose group.
CONCLUSIONS: Citalopram may lead to QT prolongation by stimulating adenosine A1 receptors without affecting the release of adenosine.

Introduction

Citalopram is a selective serotonin reuptake inhibitor (SSRI) that is widely prescribed for use a depressive illness and panic disorder. SSRIs are the second most common cause of antidepressant poisonings.[1] Citalopram is reported to be the only SSRI which requires routine cardiac monitoring in overdose.[2] In acute citalopram overdose, electrocardiagraphic abnormalities has been described in a number of large case series, including QT prolongation and torsades de pointes.[3]

Adenosine is an endogenous nucleoside that shows its well-known cardiovascular effects by A1, A2a, and A2b receptors.[4] Activation of A1 receptors depresses heart by negative inotropic, chronotropic, and dromotropic effects. Activation of A2 receptors cause a reduction in mean arterial pressure by causing a relaxation in vascular smooth muscle cells.[5]

Some studies have shown that adenosine A1 receptor stimulation and/or endogenous adenosine may have a role in amitriptyline—a tricyclic antidepressant (TCA)—induced cardiovascular toxicity such as hypotension, QRS, and QT prolongation.[678910] There is only one experimental study about the role of adenosine in cardiovascular effects of the citalopram.[11] It was observed that the negative inotropic and chronotropic effects induced by citalopram can be explained by the inhibition of re-uptake of adenosine or the activation of adenosine A1 receptors. Therefore, aim of this study is to clarify the role of adenosine receptors and/or endogenous adenosine in the mechanism of the cardiovascular toxic effects induced by citalopram overdose in rats.

Materials and Methods

This study was supported by a grant from the Scientific and Technological Research Council of Turkey (TUBITAK, Project Number: 107S251). This experimental study was performed with adult male Wistar rats (n = 77), weighing 250 − 280 g. The animal experiments approved by the Committee of Animal Care and Use.

All rats were fasted overnight with free access to water. Rats were anesthetized with urethane/chloralose (500 mg/kg/50 mg/kg intraperitoneally). The trachea was cannulated for spontaneous breathing. The right common carotid artery was cannulated (PE 50 OD mm [in.].97 [.038] ID mm [in.].58 [.023]) for blood pressure measurements. The left external jugular vein and left femoral vein were cannulated for drug administration (0.05 mI/kg/min, Braun, Perfusor Compact S, Germany). After the cannulation procedure, animals were allowed to become stabilized for 15 minutes. Rats were excluded from the study which had a mean arterial pressure under 100 mmHg. The body temperature was kept at 37°C.[6]

The mean arterial pressure (MAP), heart rate (HR), electrocardiogram (ECG), and survival time were recorded for each rat during 60 minutes (MLT844 Physiological Pressure Transducer, Interlab LTD, Istanbul, Turkey; Powerlab/8SP Data Acquisition System, AD Instruments, United Kingdom).

Experimental protocol 1: Evaluation of adenosine receptors in citalopram-induced cardiovascular toxicity

We tested 0.5 mg/kg/min, 1 mg/kg/min, 2 mg/kg/min; 4 mg/kg/min, and 8 mg/kg/min infusion doses of citalopram to determine a toxic dose of citalopram (n = 18). Citalopram infusion of 4 mg/kg/min caused a significant reduction in MAP and HR and a significant prolongation in QT interval and QRS durations after 10th minute (P < 0.001, for all). Citalopram infusion of 8 mg/kg/min caused death at 15th minute. Hence, 4 mg/kg/min citalopram infusion was used in the experimental protocol.

The cardiovascular effects of the mediators released from mast cells were prevented by the stimulation of adenosine A3 receptors by using sodium cromoglycate—a mast cell stabilizator. The safe dose of sodium cromoglycate that did not significantly alter MAP, HR, QRS duration, and QT intervals was found to be 20 mg/kg bolus (n = 8). After the stabilization period, sodium cromoglycate was administered to all animals intravenously (i.v). After 10 minutes, rats were randomized into four groups [Table 1] as follows:

Table 1

Experimental design of Protocol 1 of the study to evaluate the role of adenosine receptors in citalopram-induced cardiovascular toxicity in rats

Group 1 (control, 5% dextrose, n = 7): Following the 20 minute infusion of 5% dextrose, citalopram (4 mg/kg/min) was infused for 60 minutes

Group 2 [8-Cyclopentyl-1,3-Dipropylxanthine, DPCPX, n = 7]: Following the 20 minute infusion of 20 μg/kg/min of DPCPX (selective adenosine A1 receptor antagonist)[6], citalopram (4 mg/kg/min) was infused for 60 minutes

Group 3 [8-(3-chlorostyryl) caffeine, CSC, n = 7]: Following the 20 minute infusion of 24 μg/kg/min of CSC (selective adenosine A2a receptor antagonist)[6], citalopram (4 mg/kg/min) was infused for 60 minutes

Group 4 (dimethyl sulfoxide, DMSO, n = 3): Following the 20 minute infusion of 10% DMSO (solvent of DPCPX and CSC), citalopram (4 mg/kg/min) was administered for 60 minutes.

Protocol 2: Evaluation of endogenous adenosine in citalopram-induced cardiovascular toxicity

Ten percent DMSO [a solvent of erythro-9-(2-hydroxy-3-nonyl) adenine (EHNA) and S-(4-nitrobenzyl)-6-thioinosine (NBTI) in a preliminary study (n = 3)] did not increase the endogenous adenosine concentrations (0.7293 ± 0.06742, 0.6823 ± 0.1916; P > 0.05, the plasma adenosine concentrations at the beginning and end of the experiment, respectively) and it did not cause any change in the cardiovascular parameters.

Blood samples of 2 mL were collected from the rats under anesthesia from the tail vein 15 days before the experiment to measure basal plasma adenosine levels. After 1 week, the rats were randomized into three groups as follows [Table 2]:

Table 2

Experimental design of Protocol 2 of the study to evaluate the role of endogenous adenosine in citalopram-induced cardiovascular toxicity in rats

Group 1 (control, n = 8): The control group received 0.5 mL of 5% dextrose solution i.p. 1 hour before the cannulation (equal volume as the other groups). After the stabilization period, citalopram (4 mg/kg/min) was infused during 60 minutes

Group 2 (n = 8): EHNA, 10 mg/kg i.p—an inhibitor of adenosine deaminase (ADA) and NBTI, 1 mg/kg i.p—an inhibitor of facilitated adenosine transport were administered to the rats 1 hour before cannulation with the total volume of 0.5 mL.[12] After pretreatment with EHNA and NBTI, 5% dextrose (0.05 mL/kg/min) was infused for 60 minutes. This group was designed to determine the increase in adenosine availability

Group 3 (n = 8): EHNA (10 mg/kg i.p) and NBTI (1 mg/kg i.p) were administered to the rats 1 hour before the cannulation with the total volume of 0.5 mL.[12] After the pretreatment with EHNA and NBTI, citalopram (4 mg/kg/min) was infused during 60 minutes.

At the end of experiment, blood samples were collected from the carotid artery. The HPLC-fluorometry system (Shimadzu, Osaka, Japan) was used for measurement of plasma adenosine concentrations according to the method of Zhang and Saito.[1314]

Drugs

Citalopram was obtained from Fako-Actavis Company (Istanbul, Turkey) and was prepared in distilled water (80 mg/mL). Urethane, α-chloralose, DPCPX, CSC, adenosine, dilazep dihydrochloride, indomethacin, EHNA, NBTI, EDTA, G- EDTA, trichloroacetic acid, potassium hydroxide, zinc sulfate, and barium hydroxide were obtained from Sigma Chemical (St. Louis, MO, USA). DMSO and chloroacetaldehyde were obtained from Aldrich Chemical. Urethane and α-chloralose were prepared as 300 mg/mL and 40 mg/mL stock solutions in distilled water, respectively. Sodium cromoglycate was prepared at a concentration of 12 mg/mL in distilled water. DPCPX was prepared as 4 mg/mL stock solution in DMSO. CSC was prepared as 6 mg/mL stock solution in DMSO. EHNA was prepared at a concentration of 10 mg/mL in DMSO. NBTI was prepared as 1 mg/mL stock solution in DMSO.

Statistical Analysis

Statistical analysis was carried out by calculating the percentage change in cardiovascular parameters. Statistical analysis of data within groups was evaluated by repeated measures of analysis of variance (ANOVA) followed by Tukey's multiple comparison tests. To analyze the differences among groups, ANOVA and Tukey's multiple comparison tests were performed. Duration of survival was compared using survival analysis based on the Kaplan-Meier procedure (GraphPad Instat™, 1990 − 1994, GraphPad Software V2.05a 9342, USA). P values of < 0.05 were considered to be statistically significant.

Results

Protocol 1: The Adenosine Receptors in Citalopram-induced Cardiovascular Toxicity

Citalopram infusion followed by the dextrose infusion caused a significant decrease in MAP and HR (P < 0.001 and P < 0.01). Citalopram prolonged QT interval and QRS duration after 30th minute (P < 0.01 and P < 0.05). Citalopram infusion followed by the DPCPX infusion caused a significant decrease in MAP and HR (P < 0.001 and P < 0.01). Citalopram infusion prolonged QRS duration at 60th minute significantly (P < 0.05). There was no significant change in the QT interval.

Citalopram infusion followed by the CSC infusion caused a significant decrease in the MAP and HR after 20th minute (P < 0.001 and P < 0.01). Citalopram prolonged QT interval and QRS duration (P < 0.05 and P < 0.01). Citalopram infusion followed by DMSO infusion caused a significant decrease in the MAP and HR, (P < 0.05 and P < 0.00l). Citalopram prolonged QRS duration after 10th minute significantly (P < 0.05). There was a prolongation in the QT interval after 20th minute but this prolongation was not significant [Table 3].

Table 3

The effects of citalopram on cardiovascular parameters following administration of adenosine receptor antagonists

There was no statistically significant difference between MAP and HR reduction among the groups (P > 0.05), [Figure 1a and b]. DPCPX infusion prevented the prolongation of QT interval induced by citalopram when compared to control group significantly (P < 0.05 at 20th min and 30th min; P < 0.01 at 40th min, 50th min and 60th min) [Figure 1c]. DPCPX or CSC infusions did not prevent QRS prolongation induced by citalopram at any time, significantly. DPCPX infusion prevented QRS prolongation when compared to DMSO group, significantly (P < 0.05) [Figure 1d].

Figure 1

The effects of citalopram on cardiovascular parameters after the pretreatments with dextrose, DPCPX, CSC, and DMSO infusions. (c) †, P < 0.05; DPCPX versus control group, ††, P < 0.01; DPCPX versus control group, (d) †, P < 0.05; DPCPX versus DMSO group. (ANOVA and Tukey's multiple comparison tests were performed.). End of the infusion = End of the pretreatment of DPCPX, CSC and DMSO, MAP = Mean arterial pressure, HR = Heart rate, DPCPX = 8-Cyclopentyl-1,3-Dipropylxanthine, CSC = 8-(3-chlorostyryl) caffeine, DMSO = dimethyl sulfoxide, Cit = Citalopram

Survival rate was 100% (7/7) for all groups. There was not any significant difference for survival times among groups (P > 0.05).

Plasma adenosine concentrations

In the control group, citalopram infusion did not change plasma adenosine concentrations significantly (P > 0.05). In the other two groups, plasma adenosine concentrations increased (P < 0.01, P < 0.05 at group 2 and 3, respectively). When compared with the control group, a significant increase in the groups pretreated with EHNA/NBTI was observed (P < 0.05, P < 0.05 at group 2 and 3, respectively) [Table 4].

Table 4

Baseline and end-experiment plasma adenosine levels for protocol 2

Cardiovascular parameters

Citalopram infusion followed by dextrose injection caused a significant decrease in MAP and HR (P < 0.001 and P < 0.001). Citalopram prolonged QT interval and QRS duration (P < 0.001 and P < 0.001). Dextrose infusion followed by EHNA and NBTI injections caused a significant decrease in MAP and HR (P < 0.01 and P < 0.05). Dextrose infusion followed by EHNA and NBTI injections prolonged QRS duration at 50th minute (P < 0.05). There was no significant change in QT interval. Citalopram infusion followed by EHNA and NBTI injections caused a significant decrease in MAP and HR (P < 0.001 and P < 0.001). Citalopram infusion prolonged QT interval and QRS duration (P < 0.001 and P < 0.001) [Table 5].

Table 5

The changes in cardiovascular parameters of the comparator groups for protocol 2

The comparison of group 1 and group 3

Decrease in MAP in EHNA/NBTI pretreated and citalopram infused rats was significantly higher than that of dextrose and citalopram infused rats at 10th minute (P < 0.05) [Figure 2a]. When we compared two groups for decrease in HR, citalopram followed by EHNA/NBTI-induced decrease in HR was significantly higher than that of citalopram followed by dextrose infusion-induced decrease at 40th minute (P < 0.01) [Figure 2b]. There was no statistically significant difference in QT and QRS prolongations between two groups (P > 0.05) [Figure 2c and d].

Figure 2

The effects of the groups on the cardiovascular parameters. Control group (citalopram infusion), Group 2 (EHNA/NBTI administration), Group 3 (citalopram infusion following EHNA/NBTI administration) (a) *P < 0.05; group 3 versus control group, †, P < 0.05, ††, P < 0.01; group 3 versus group 2, (b) **P < 0.01; group 3 versus control group, ††, P < 0.01; group 3 versus group 2, †††, P < 0.001; group 3 versus group 2, (c) †, P < 0.05; group 3 versus group 2, ††, P < 0.01; group 3 versus group 2, (d) †, P < 0.05; group 3 versus group 2. (ANOVA and Tukey's multiple comparison tests were performed.) MAP = Mean arterial pressure, HR = Heart rate, EHNA = Erythro-9-(2-hydroxy-3-nonyl) adenine, NBTI = S-(4-nitrobenzyl)-6-thioinosine

The comparison of group 2 and group 3

Decrease in the MAP in EHNA/NBTI pretreated and citalopram infused rats was significantly greater than that of EHNA/NBTI pretreated and dextrose infused rats at 10th, 20th, and 30th minutes (P < 0.05, P < 0.05 and P < 0.01, respectively) [Figure 2a].

Citalopram followed by EHNA/NBTI-induced decrease in HR was significantly higher than that of dextrose followed by EHNA/NBTI infusion-induced decrease after 20th minute (P < 0.001, P < 0.01, P < 0.001 and P < 0.001, respectively) [Figure 2b].

Prolongation in QT interval in EHNA/NBTI pretreated and citalopram infused rats was greater than that of EHNA/NBTI pretreated and dextrose infused rats at 10th and 20th minutes (P < 0.01 and P < 0.05, respectively) [Figure 2c].

Prolongation in QRS interval in EHNA/NBTI pretreated and citalopram-infused rats was statistically more significant than that of EHNA/NBTI pretreated and dextrose-infused rats after 20th minute (P < 0.05) [Figure 2d].

Survival Analysis

The survival rates for 60 minutes were 50% (4/8) for group 1, 100% (8/8) for group 2, and 62.5% (5/8) for group 3. While four rats died at 25th, 49th, 50th, and 52th minutes in the control group, three rats died at 31th, 31th, and 34th minutes in group 3. There was no significant difference in survival rates (P > 0.05) [Figure 3].

Figure 3

Comparison of survival times among the groups. (Kaplan- Meier analysis was used.). EHNA = Erythro-9-(2-hydroxy-3-nonyl) adenine, NBTI = S-(4-nitrobenzyl)-6-thioinosine

Discussion

Our study consisted of two parts in which we investigated the role of adenosine receptors and endogenous adenosine in citalopram-induced cardiovascular toxicity.

In the first part of our study, citalopram infusion led to a significant reduction in mean arterial pressure (MAP) and heart rate (HR), and significantly prolonged QRS duration and the QT interval. Administration of DPCPX prior to citalopram infusion prevented only the prolongation of the QT interval as compared with the control group. However, administration of CSC prior to citalopram infusion did not ameliorate citalopram-induced deterioration of cardiovascular parameters. No statistically significant difference in survival rates was found among the groups.

The citalopram-induced cardiotoxic effects observed in our rat model are compatible with clinical evidence in the literature. Although exposure to high doses of selective serotonin reuptake inhibitors (SSRIs) is known to be safer than that of tricyclic antidepressants (TCAs), high-dose exposure to citalopram is not safe and can lead to critical cardiotoxic effects such as hypotension, tachycardia, bradycardia, bundle-branch block, and ECG abnormalities.[3151617] In a retrospective study that compared the cardiotoxic effects of SSRIs, citalopram was determined to prolong the QT and QTc intervals more significantly than fluoxetine, fluvoxamine, paroxetine, and sertraline.[2]

Adenosine causes negative chronotropic, dromotropic, and inotropic effects, as well as cardiac depression via A1 receptors. Adenosine produces its known effects through cyclic adenosine monophosphate (cAMP)-dependent (indirect or antiadrenergic effect) and cAMP-independent pathways via A1 receptors. In the cAMP-dependent pathway, adenosine antagonizes electrophysiological and biochemical effects of β-adrenergic agonists. In cases where a β-adrenergic agonist is not present, adenosine does not affect the ventricular action potential or calcium flow. In the cAMP-independent (direct) pathway, stimulation of A1 receptors causes K+ loss by Gi protein-gated inwardly rectifying K+ channels. This causes shortening of the action potential of atrial cells, hyperpolarization of sinoatrial (SA) node cells, and depression of the action potential of atrioventricular nodal cells.[18]

In our study, blocking adenosine receptors by the selective adenosine A1 and A2a antagonists DPCPX and CSC did not prevent citalopram-induced reductions in MAP and HR. In an isolated guinea pig atrium study, negative inotropic and chronotropic effects caused by citalopram could not be prevented by an adenosine A2 receptor antagonist (3,7 dimethyl-1- dipropargylxanthine, DMPX); however, these effects were significantly blocked by a selective adenosine A1 receptor antagonist (DPCPX) and a non-selective adenosine A1/A2 receptor antagonist (theophylline).[11] Negative inotropic and chronotropic effects of citalopram were explained by adenosine re-uptake inhibition or by activation of A1 receptors. Conversely, a clear conclusion could not be reached. Further studies are needed to investigate whether adenosine receptor antagonists can prevent citalopram-induced reductions in MAP and HR.

The observed inhibition of citalopram-induced QT prolongation by an A1 receptor antagonist (DPCPX) suggests that endogenous adenosine and/or stimulation of adenosine A1 receptors plays a role in the mechanism underlying QT interval prolongation. Adenosine has minimal and/or non-existent effects on ventricular myocardial cells in the absence of catecholamines. This effect is associated with the absence of potassium acetylcholine channels (KAch) in ventricular myocytes under basal conditions.[19] If an adenosine-mediated mechanism plays a role in citalopram toxicity, this information will help us ignore the impact of a cAMP-independent (direct) pathway in the mechanism of QT interval prolongation. QT prolongation is related to the ventricular action potential. Thus, prolongation of the ventricular action potential leads to prolongation of the QT interval. Theoretically, prolongation of the action potential is possible through either an increase of inward depolarizing currents or a reduction of outward repolarization currents carried by potassium ions.[20] Adenosine shows antiadrenergic activities over delayed rectifier potassium currents (IK). Action potential prolongation occurs through the inhibition of catecholamine-related IK currents.[21]

In a study performed by Witchel et al., in isolated guinea pig cardiomyocytes, citalopram was shown to inhibit human ether-à-go-go related gene (hERG)-related K+ channel currents.[22] hERG ion channel is responsible for the repolarizing rapid delayed rectifier potassium current (IKr) and electrical conduction in the myometrium. Inhibition of this channel results in prolongation of both the QT interval and ventricular action potential.[23] Based on the results of some studies, both citalopram and adenosine utilize IK currents while generating their effects. Considering these data, adenosine-mediated rapid delayed rectifier K+ current (IKr) inhibition most likely plays a role in citalopram-induced QT prolongation. Further studies are needed to investigate the role of potassium channels in adenosine-mediated mechanism on citalopram toxicity.

Based on the results of this part of our study, it is difficult to assert whether the increase in endogenous adenosine or direct adenosine receptor stimulation plays a role in the cardiotoxic effects induced by citalopram.

In the second part of our study, we evaluated the role of endogenous adenosine in the cardiotoxic effects of citalopram poisoning. We found no significant difference in rat plasma adenosine concentrations with high-dose citalopram infusion. A significant increase in plasma adenosine concentrations was observed in groups pre-treated with ENHA/NBTI. In the control group, citalopram infusion caused a significant prolongation of QRS duration and the QT interval, and a significant reduction in MAP and HR. In addition to an increase in plasma adenosine concentrations after ENHA/NBTI administration, a significant decrease was observed in MAP and HR. Further, QRS duration was significantly prolonged in the 50th minute. These alterations can be explained by the known cardiovascular effects of adenosine through stimulation of its receptors.[518] Citalopram, administered after ENHA/NBTI infusion, potentiated the effects on MAP, HR, QRS duration, and the QT interval of the group that was only administered ENHA/NBTI. No statistically significant difference in survival rates was found among the groups.

In our study, high-dose citalopram infusion caused a reduction in MAP and HR, and also prolonged QRS duration and QT intervals without increasing plasma adenosine concentrations. Citalopram can increase susceptibility to endogenous adenosine without changing plasma adenosine levels.

Our results show that citalopram may lead to QT prolongation by stimulating adenosine A1 receptors without affecting the release of adenosine. Further studies are needed to clarify the role of potassium channels via adenosine A1 receptors stimulation on mechanism of citalopram induced-QT prolongation.