sign in sign up
<<
Home , Chlorphentermine, Substituted amphetamine

0022-3565/06/3182-604–610 THE JOURNAL OF PHARMACOLOGY AND EXPERIMENTAL THERAPEUTICS Vol. 318, No. 2 U.S. Government work not protected by U.S. copyright 101618/3124172 JPET 318:604–610, 2006 Printed in U.S.A.

Amphetamine Analogs Increase Plasma : Implications for Cardiac and Pulmonary Disease

Dorota Zolkowska, Richard B. Rothman, and Michael H. Baumann Clinical Section, Intramural Research Program, National Institute on Drug Abuse, National Institutes of Health, Baltimore, Maryland Received January 18, 2006; accepted April 25, 2006

ABSTRACT Downloaded from Elevations in plasma serotonin (5-HT) have been implicated in measured using a novel microdialysis method in whole blood. the pathogenesis of cardiac and pulmonary disease. Normally, We found that baseline dialysate levels of 5-HT are ϳ0.22 nM, plasma 5-HT concentrations are kept low by transporter-medi- and analogs evoke large dose-dependent in- ated uptake of 5-HT into platelets and by metabolism to 5-hy- creases in plasma 5-HT ranging from 4 to 20 nM. The ability of droxyindoleacetic acid (5-HIAA). Many abused drugs (e.g., sub- drugs to elevate plasma 5-HT is positively correlated with their stituted ) and prescribed medications (e.g., potency as 5-HT transporter substrates. produced fluoxetine) target 5-HT transporters and could thereby influence small, but significant, increases in plasma 5-HT. Although the jpet.aspetjournals.org circulating 5-HT. We evaluated the effects of amphetamines drug-evoked 5-HT concentrations are below the micromolar analogs [(Ϯ)-, (Ϯ)-3,4-methylenedioxymetham- levels required for contraction of pulmonary arteries, they ap- phetamine, (ϩ)-, (ϩ)-amphetamine, phenter- proach concentrations reported to stimulate mitogenesis in mine] on extracellular levels (i.e., plasma levels) of 5-HT and pulmonary artery smooth muscle cells. Additional studies are 5-HIAA in blood from catheterized rats. Effects of the 5-HT needed to determine the effects of chronic administration of uptake blocker fluoxetine were examined for comparison. amphetamines on circulating 5-HT. Drugs were tested in vivo and in vitro; plasma indoles were at ASPET Journals on May 26, 2015

Serotonin (5-hydroxytryptamine, 5-HT) is an endogenous via reversal of SERT (Rothman et al., 1999). One hypothesis bioactive compound that is widely distributed in neurons, to explain the ability of these agents to increase the risk of mast cells, enterochromaffin cells, and blood platelets (Coo- developing PPH is that they increase plasma 5-HT by stim- per et al., 2003; Gershon, 2004). Under normal physiological ulating 5-HT release from platelets (i.e., “the 5-HT hypothe- conditions, plasma 5-HT levels are kept exquisitely low (i.e., sis” of PPH) (MacLean et al., 2000). Others have invoked the Ͻ1 nM) due to transporter-mediated uptake of 5-HT into same mechanism to explain fenfluramine-induced VHD blood platelets and via metabolism of 5-HT to 5-hydroxyin- (Fishman, 1999), although a more likely mechanism in this doleacetic acid (5-HIAA) by (MAO). In- case involves activation of 5-HT2B receptors by the N-deethy- cluded among the many physiological effects of 5-HT are lated metabolite of fenfluramine, (Rothman mitogenesis and vasoconstriction. Accordingly, altered regu- et al., 2000a). lation of 5-HT levels in blood has been implicated in the Despite the importance of the 5-HT hypothesis to current pathogenesis of cardiac valvular heart disease (VHD) (Robio- dogma regarding the etiology of drug-induced PPH and VHD, lio et al., 1995) and primary pulmonary hypertension (PPH) the effects of agents on plasma 5-HT have received (MacLean et al., 2000). Many an- little attention. The paucity of data in this regard could be alogs (e.g., fenfluramine and ) are substrates for related to the fact that measuring plasma 5-HT is technically 5-HT transporters (SERTs) and release 5-HT from neurons challenging. Given that basal 5-HT levels are quite low and that over 99% of blood 5-HT is stored in platelets, even minor disturbance of platelets during sample handling will cause This work was supported by the National Institute on Drug Abuse IRP. large artificial increases in plasma 5-HT concentrations. Article, publication date, and citation information can be found at http://jpet.aspetjournals.org. Studies conducted in the 1990s indicate that acute admin- doi:10.1124/jpet.106.101618. istration of (ϩ)-fenfluramine does not increase plasma 5-HT

ABBREVIATIONS: 5-HT, 5-hydroxytryptamine or serotonin; 5-HIAA, 5-hydroxyindoleacetic acid; MAO, monoamine oxidase; VHD, valvular heart disease; PPH, primary pulmonary hypertension; SERT, ; IRP, Intramural Research Program; MDMA, (Ϯ)-3,4-methyl- enedioxymethamphetamine; HPLC, high-performance liquid chromatography; ECD, electrochemical detection; ANOVA, analysis of variance; SSRI, serotonin-selective reuptake inhibitor. 604 Amphetamines Increase Plasma Serotonin 605

in rats (Martin and Artigas, 1992), and chronic administra- In Vitro Drug Administration. Between 8:00 and 9:00 AM, rats tion of fenfluramine lowers blood 5-HT in humans (see Roth- were moved into the testing room and allowed to acclimate to the man et al., 2000b and references therein). In the present surroundings for 1 h. Extension tubes were attached to catheters, study, we developed a novel microdialysis method to assess and 0.5 ml of heparin flush (48 IU/ml in saline) was injected as noted the effects of amphetamine analogs on plasma levels of 5-HT above. In these experiments, serial blood samples were withdrawn from untreated donor rats and transferred into 300-␮l polypropylene in whole blood samples obtained from conscious catheterized tubes that were kept at 25°C. These tubes were prefilled with 20 ␮l rats. The results show that amphetamine analogs produce of heparin (1000 IU/ml). Test drugs or vehicle were added directly to significant dose-dependent increases in plasma 5-HT, and blood samples in 10-␮l volumes to yield final concentrations of 0.3, 1, this effect is proportional to drug potency as SERT sub- 3, or 33 ␮M. Microdialysis probes were placed into the blood samples, strates. 5-HT uptake inhibitors, such as fluoxetine, also can and dialysate efflux was collected for 15 min and assayed for 5-HT increase plasma 5-HT, but to a lesser extent. The physiolog- and 5-HIAA using HPLC-ECD. Each blood sample was dialyzed for ical significance of these findings is discussed. 15 min to generate a single dialysate sample. Two baseline samples were collected before addition of test drugs to subsequent samples. Materials and Methods Probe recoveries were performed before and after blood sampling using a 10-pg 5-HT standard. Animals. Male Sprague-Dawley rats (Charles River Laboratories, HPLC-ECD Analysis of 5-HT and 5-HIAA. Aliquots of the Wilmington, MA) weighing 350 to 450 g were singly housed with food dialysate (5 ␮l) were injected onto a microbore HPLC column (Unijet, and water freely available. Rats were maintained in facilities accred- 100 ϫ 1 mm, 5 ␮M octadecylsilane; Bioanalytical Systems, Inc., West

ited by the American Association of the Accreditation of Laboratory Lafayette, IN) that was coupled to an amperometric detector (Model Downloaded from Animal Care, and procedures were carried out in accordance with the LC-4C; Bioanalytical Systems, Inc.). A glassy carbon electrode was Animal Care and Use Committee of the National Institute on Drug set at a potential of ϩ650 mV relative to Ag/AgCl reference. Mobile ␮ Abuse Intramural Research Program (IRP). phase consisted of 180 MNa2EDTA, 150 mM monochloroacetic Drugs and Reagents. (Ϯ)-Fenfluramine HCl (fenfluramine, FW acid, 125 mM NaOH, and 690 ␮M sodium octanesulfonic acid, with

267.7), (Ϯ)-3,4-methylenedioxymethamphetamine (MDMA, FW 7.5% MeOH and 7.5% CH3CN/l water (final pH 3.15). Mobile phase 229.7), (ϩ)-methamphetamine HCl (methamphetamine, FW 185.7), was pumped through the column at 60 ␮l/min (260D, syringe pump;

(ϩ)-amphetamine sulfate (amphetamine, FW 368.5), Teledyne ISCO, Lincoln, NE). Chromatographic data were acquired jpet.aspetjournals.org HCl (phentermine, FW 185.7), and pentobarbital sodium were ob- on-line and exported to a Millennium software system (Waters As- tained from the National Institute on Drug Abuse, IRP Pharmacy. sociates, Milford, MA) for peak amplification, integration, and anal- Fluoxetine HCl (fluoxetine, FW 345.8) was purchased from Spectrum ysis. The concentration of 5-HT and 5-HIAA in dialysate samples Chemical Manufacturing Company (New Brunswick, NJ). Monochlo- was compared with known standards, and the lower limit of detec- roacetic acid was obtained from Mallinckrodt Baker Inc. (Phillips- tion was ϳ0.05 pg/5 ␮l (0.047 nM) for both indoles. burg, NJ), and all other reagents were obtained from Sigma Chem- Statistical Analyses. In all studies, the first two dialysate sam- ical Co. (St. Louis, MO). Drug solutions for the in vivo and in vitro ples collected before any treatment were considered baseline sam- studies were prepared in saline immediately before use, and doses ples. 5-HT and 5-HIAA measures are mean Ϯ S.E.M. expressed as at ASPET Journals on May 26, 2015 are expressed as the salt. picograms per 5-␮l sample. For in vivo experiments, data were eval- Surgical Procedures. Rats received sodium pentobarbital (60 uated by two-way ANOVA (drug treatment ϫ time) and one-way mg/kg i.p.) for surgical anesthesia. Indwelling jugular catheters, ANOVA (at each time point). For in vitro experiments, data were made of Silastic Medical Grade tubing (Dow Corning, Midland, MI), evaluated by one-way ANOVA (drug dose). When significant F val- were implanted into the right jugular vein and advanced to the ues were obtained, Newman-Keuls post hoc tests were performed to atrium as described previously (Baumann et al., 2001). Rats were compare group means. For data correlations, peak drug effects mea- allowed to recover for 7 to 10 days postoperatively. sured in vivo after 1.0 mg/kg were compared with peak drug effects ␮ 3 In Vivo Drug Administration. Between 8:00 and 9:00 AM, rats measured in vitro after 3 M and with drug EC50 values for [ H]5- were moved into the testing room and allowed to acclimate to the HT release from synaptosomes (Rothman et al., 2001). A value of p Ͻ surroundings for 1 h. Extension tubes were attached to catheters, 0.05 was considered as the minimum criterion for statistical signif- and 0.5 ml of heparin flush (48 IU/ml in saline) was injected. Blood icance. samples (0.3 ml) were withdrawn into 1-ml syringes and gently transferred into 300-␮l polypropylene vials that were chilled on crushed ice. Vials contained 20 ␮l of heparin (1000 IU/ml in saline) Results as an anticoagulant. Heparin flush was injected into the rats after Baseline 5-HT Levels in Plasma. For all animals used in each sample to maintain volume homeostasis. A 4 ϫ 0.6-mm dialysis this study (n ϭ 172 rats), the mean basal concentration of probe (MAB 6; SciPro, Inc., Sanborn, NY) was immediately im- Ϯ ␮ Ϯ mersed in the chilled blood sample. Ringer’s solution containing dialysate 5-HT in blood was 0.23 0.01 pg/5 l (i.e., 0.22 150.0 mM Naϩ, 3.0 mM Kϩ, 1.4 mM Ca2ϩ, 0.8 mM Mg2ϩ, 1.0 nM P, 0.01 nM). Baseline dialysate 5-HT levels differed slightly and 155 mM ClϪ was pumped through the probe at 1.7 ␮l/min, and depending upon experimental conditions. In the in vivo drug each blood sample was dialyzed for 15 min to generate a single administration experiments (n ϭ 120 rats), blood samples dialysate sample. Serial blood samples were collected and dialyzed were maintained on ice to reduce SERT-mediated “leak” of every 15 min. Dialysates were assayed for 5-HT and 5-HIAA using 5-HT from platelets, and basal plasma 5-HT was 0.20 Ϯ 0.01 high-performance liquid chromatography (HPLC) with electrochem- pg/5 ␮l. In the in vitro drug experiments (n ϭ 52 rats), blood ical detection (ECD) as described below. After three to four baseline samples were kept at room temperature to optimize SERT samples were obtained and dialyzed, drug treatments were admin- function, and this resulted in a slightly higher basal level of istered through i.v. catheters. Fenfluramine, MDMA, methamphet- plasma 5-HT, 0.29 Ϯ 0.02 pg/5 ␮l. Baseline 5-HT did not , amphetamine, and fluoxetine were dissolved in saline and administered as i.v. bolus injections of 0.3 and 1.0 mg/kg. Phenter- differ significantly between the various treatment groups mine was dissolved in saline and injected i.v. at doses of 1.0 and 3.0 within the in vivo and in vitro conditions. It should be noted mg/kg. Blood samples were collected at 15-min intervals for 90 min that microdialysis probes had in vitro recovery rates of ap- postinjection. Probe recoveries were performed before and after blood proximately 25% when tested in a physiological salt solution, sampling using a 10-pg 5-HT standard prepared in Ringer’s solution. and this value did not vary significantly before, during or 606 Zolkowska et al. after experiments where probes were immersed in sequential inhibitor, fluoxetine, produced a modest increase in plasma blood samples. Preliminary studies have shown that in vitro 5-HT [F(2,120) ϭ 49.16; p Ͻ 0.0001] with 0.3 and 1.0 mg/kg probe recoveries determined in artificial salt solutions do not doses increasing 5-HT levels 4- and 7-fold. Interestingly, necessarily reflect probe recovery characteristics in complex none of the in vivo treatments affected plasma 5-HIAA levels biological matrices. Thus, we did not “correct” 5-HT or (data not shown). Fenfluramine, MDMA, and methamphet- 5-HIAA values for probe recovery. In all cases, separate sa- amine produced transient increases in plasma 5-HT that had line control groups were tested in parallel with each drug mostly returned to baseline values after 90 min. treatment condition. Some investigators (Ulus et al., 2000) have proposed that In Vivo Drug Administration Experiments. Figure 1 pharmacological doses of phentermine and other amphet- depicts the effects of fenfluramine, MDMA, methamphet- amine analogs will block MAO activity in vivo, an effect that amine, amphetamine, phentermine, and fluoxetine on dialy- can be detected as decreased 5-HIAA levels. Others have sate 5-HT levels in whole blood, when drugs were adminis- suggested that coadministration of phentermine with fenflu- tered via i.v. catheters in vivo. Fenfluramine significantly ramine would enhance fenfluramine-induced increases in increased dialysate 5-HT [F(2,120) ϭ 56.55; p Ͻ 0.0001], and plasma 5-HT (Fishman, 1999; Ulus et al., 2000). To address this stimulatory action was dose-dependent. Fenfluramine these hypotheses, we compared the effects of phentermine elevated extracellular 5-HT to 15- and 19-fold after 0.3 and and fenfluramine, alone and in combination, at 1.0 mg/kg 1.0 mg/kg, respectively. MDMA also significantly increased doses. As shown in Fig. 2, no treatment altered plasma 5-HT levels [F(2,120) ϭ 64.22; p Ͻ 0.0001], with 12- and 5-HIAA, and coadministration of phentermine plus fenflura- 21-fold elevations at the 0.3 and 1.0 mg/kg doses. Metham- mine did not enhance plasma 5-HT higher than fenfluramine Downloaded from phetamine significantly increased plasma 5-HT [F(2,120) ϭ alone. 29.75; p Ͻ 0.0001] but to a lesser extent than fenfluramine, In Vitro Drug Administration Experiments. Figure 3 with 7- and 12-fold elevations at 0.3 and 1.0 mg/kg doses. shows the effects of fenfluramine, MDMA, methamphet- Amphetamine had weak effects on 5-HT levels [F(2,120) ϭ amine, amphetamine, phentermine, and fluoxetine on dialy- 3.64; p ϭ 0.03], and only the 1 mg/kg dose caused a signifi- sate 5-HT levels in blood, when drugs were administered cant 5-fold rise. Phentermine had a very weak effect on 5-HT, directly into blood samples in vitro. Fenfluramine signifi- jpet.aspetjournals.org and post hoc tests failed to demonstrate any significant ef- cantly increased plasma 5-HT in a dose-dependent manner fects of the 1.0 and 3.0 mg/kg doses, possibly due to variabil- [F(3,23) ϭ 64.54; p Ͻ 0.0001], producing 16- and 91-fold ity in the saline-injected control group. The 5-HT reuptake elevations when administered at concentrations of 3 and 33 ␮M, respectively. MDMA had similar effects [F(3,23) ϭ 16.15; p Ͻ 0.0001] and increased dialysate 5-HT to 8- and 42-fold above baseline at 3 and 33 ␮M doses. Methamphet- amine [F(3,23) ϭ 96.19; p Ͻ 0.0001] and amphetamine at ASPET Journals on May 26, 2015 [F(3,23) ϭ 50.03; p Ͻ 0.0001] were somewhat less efficacious than MDMA, significantly increasing 5-HT by 4- and 44-fold at the 3 and 33 ␮M doses, respectively. Phentermine signif- icantly increased 5-HT levels only at the 33 ␮M dose [F(3,23) ϭ 9.71; p Ͻ 0.0004]. Fluoxetine dose dependently increased plasma 5-HT [F(3,23) ϭ 17.76; p Ͻ 0.0001] to 3-, 14,- and 22-fold above baseline at 0.3, 3, and 33 ␮M, respec- tively. The data in the top panel of Fig. 4 show that the ability of amphetamine analogs to increase plasma 5-HT in vitro at a 3 ␮M dose is significantly correlated with the increase in 5-HT levels produced after in vivo administration of 1.0 mg/kg (p Ͻ

Fig. 2. In vivo effects of fenfluramine (fen) and phentermine (phen) on dialysate 5-HT and 5-HIAA levels measured in blood from conscious rats. Fig. 1. In vivo effects of amphetamine analogs and fluoxetine on dialysate Drugs were dissolved in sterile saline and administered by the i.v. route. 5-HT levels measured in blood from conscious rats. Drugs were dissolved Fen or phen was given at a dose of 1.0 mg/kg, whereas fen ϩ phen was in sterile saline and administered i.v. at 0 min. Serial blood samples were administered as a mixture to yield a final concentration of 1.0 mg/kg for withdrawn at 15-min intervals and immediately dialyzed as described each drug. Serial blood samples were withdrawn at 15-min intervals and under Materials and Methods. Data are mean Ϯ S.E.M. for n ϭ 6 rats/ were immediately dialyzed. Data show peak effects determined 15 min p Ͻ 0.05 with ,ء .p Ͻ 0.05 compared with saline controls at corresponding time postinjection and are mean Ϯ S.E.M. for n ϭ 6 rats/group ,ء .group points. respect to saline-treated control. Amphetamines Increase Plasma Serotonin 607 Downloaded from Fig. 4. Drug-induced elevations of plasma 5-HT in vivo are correlated

with increases in plasma 5-HT measured in vitro (top) and EC50 values for [3H]5-HT release measured in synaptosomes (bottom). Top, percent increase in plasma 5-HT induced by 1 mg/kg in vivo was plotted against percent increase induced by 3 ␮M in vitro for each amphetamine tested. Bottom, percentage increase in plasma 5-HT induced by 1 mg/kg in vivo 3 was plotted against the EC50 for [ H]5-HT release from synaptosomes

previously reported for each amphetamine tested. The in vivo and in vitro jpet.aspetjournals.org ϭ plasma 5-HT data are means for n 6 rats/group, whereas the EC50 data are means for n ϭ 3 separate experiments.

Fig. 3. In vitro effects of amphetamine analogs and fluoxetine on dialy- sate 5-HT levels measured in blood from conscious rats. Drugs were Discussion added directly to blood samples obtained from untreated donor rats to Fenfluramine, MDMA, methamphetamine, and amphet- yield final concentrations shown. Blood samples were immediately dia- lyzed as described under Materials and Methods. Data show peak effects amine are known to be substrates for SERT proteins and to measured 15 min after addition of drugs and are mean Ϯ S.E.M. for n ϭ release 5-HT from neurons in the brain (Berger et al., 1992; at ASPET Journals on May 26, 2015 p Ͻ 0.05 compared with saline control. Crespi et al., 1997; Rothman et al., 2001). The anorectic ,ء .rats/group 6 0.04). This finding suggests that these drugs are increasing extracellular 5-HT in blood by a similar mechanism under in vitro and in vivo conditions. We previously reported the EC50 values for drugs releasing [3H]5-HT from synaptosomes, which is a measure of the potency of these compounds as substrates for SERT (Rothman et al., 2001). The bottom panel of Fig. 4 demonstrates that the calculated EC50 values for test drugs to release [3H]5-HT from synaptosomes is highly correlated with the ability of the same agents to in- crease plasma 5-HT in vivo (p Ͻ 0.01). Taken together, these correlative relationships suggest that amphetamine analogs increase plasma 5-HT in whole blood by a process involving SERT proteins, possibly those found on platelets. Studies in nervous tissue have shown that uptake inhibi- tors can attenuate the ability of transporter substrates to release 5-HT via carrier-mediated exchange (Baumann et al., 2001; Rothman et al., 2001). To test whether uptake blockers could affect drug-induced 5-HT efflux in blood samples, we coadministered fluoxetine and MDMA in blood samples in vitro and performed microdialysis. The top panel of Fig. 5 demonstrates that low-dose fluoxetine (i.e., 0.3 ␮M) did not block the rise in plasma 5-HT produced by low-dose MDMA (i.e., 1.0 ␮M). This type of experiment is complicated by the Fig. 5. In vitro effects of fluoxetine on MDMA-induced increases in fact that fluoxetine pretreatment alone causes increases in dialysate 5-HT levels measured in blood from conscious rats. Drugs were plasma 5-HT that are comparable with the effects of low-dose added directly to blood samples obtained from untreated donor rats to MDMA. However, as reported in the bottom panel of Fig. 5, a yield final concentrations shown as micromolar equivalents. Blood sam- ␮ ples were immediately dialyzed. Data show peak effects measured 15 min higher dose of fluoxetine (i.e., 3 M) significantly reduced the Ϯ ϭ ␮ after addition of drugs and are mean S.E.M. for n 6–10 rats/group. p Ͻ 0.05 compared with saline control; #, p Ͻ 0.05 with respect to all ,ء large increase in plasma 5-HT produced by 33 M MDMA by 50% [F(3,39) ϭ 70.61; p Ͻ 0.0001]. other groups. 608 Zolkowska et al. medication, fenfluramine, is associated with the occurrence rat brain synaptosomes (79 nM) (Rothman et al., 2001). of PPH and VHD and was withdrawn from the marketplace These observations support the notion that amphetamine in 1997. Determining the mechanism(s) whereby fenflura- analogs act as substrates for platelet SERT proteins, thereby mine increases the risk of developing VHD and PPH is im- explaining their ability to increase plasma 5-HT. This inter- portant to understand, so that newly discovered pretation is further supported by the finding that fluoxetine medications will not produce these serious adverse effects could partially reverse the ability of high-dose MDMA to (Rothman and Baumann, 2003). Many investigators have release 5-HT in vitro (see Fig. 5). In this experiment, we theorized that fenfluramine increases the risk of PPH and believe that fluoxetine prevents MDMA from entering plate- VHD by elevating plasma 5-HT levels in blood, consistent lets by blocking SERT sites, thus preventing massive release with its known 5-HT-releasing properties in nervous tissue of 5-HT via disruption of storage vesicles (Schuldiner et al., (Connolly et al., 1997; Fishman, 1999; MacLean et al., 2000). 1993). Both methamphetamine and amphetamine release Studies conducted in the 1990s provide little support for this 5-HT from synaptosomes with low potency (ϳ1 ␮M), but hypothesis. For example, Martin and Artigas (1992) reported methamphetamine is much more potent at increasing extra- that acute administration of (ϩ)-fenfluramine does not in- cellular 5-HT in brain via a SERT-related mechanism (Roth- crease plasma 5-HT in rats. Moreover, chronic administra- man and Baumann, 2003; Rothman et al., 2005). Although tion of fenfluramine lowers blood 5-HT in humans (see Roth- we have no explanation for this peculiar phenomenon, it also man et al., 2000b and references therein). Given the occurs in blood since methamphetamine, but not amphet- importance of the 5-HT hypothesis to current dogma regard- amine, increases plasma 5-HT at the 0.3 mg/kg dose. As ing fenfluramine-associated PPH and VHD, we decided to noted above, the 40-fold increase in plasma 5-HT produced by Downloaded from directly investigate this issue by using a novel microdialysis 33 ␮M amphetamine in vitro likely results from disruption of method to assess the effects of amphetamine analogs on 5-HT storage vesicles in the platelets. plasma levels of 5-HT in conscious rats. In addition to fen- Steady-state plasma 5-HT levels are maintained in the fluramine, we assessed the actions of other substituted am- subnanomolar range in part by SERT-mediated uptake of phetamines on plasma 5-HT, both therapeutic agents such as 5-HT into platelets. Thus, it is not surprising that the 5-HT amphetamine and drugs of abuse such as MDMA and meth- uptake inhibitor fluoxetine also increases plasma 5-HT in jpet.aspetjournals.org amphetamine. vivo and in vitro. The stimulatory effect of fluoxetine is much Measuring plasma 5-HT is technically challenging. Given smaller in magnitude than the effects of amphetamines, sug- that over 99% of blood 5-HT is stored in platelets, and plate- gesting that uptake blockers produce minor changes in lets are very fragile, even minor disturbance or damage to plasma 5-HT compared with SERT substrates. It should be platelets during sample handling will cause large artificial noted that fluoxetine, unlike SERT substrates, cannot be increases in plasma 5-HT. The method we developed to mea- removed from the plasma by translocation into cells that sure dialysate 5-HT levels ex vivo minimizes trauma to plate- express SERT, and this could prolong the actions of fluox- at ASPET Journals on May 26, 2015 lets and permits an accurate determination of plasma 5-HT. etine and its bioactive metabolite, norfluoxetine. Because the The mean baseline level of dialysate 5-HT in blood collected ability of fluoxetine to increase plasma 5-HT arises from its from rats in these studies (n ϭ 172) was 0.22 Ϯ 0.01 nM. inhibition of SERT, other serotonin-selective reuptake inhib- Because in vitro probe recoveries averaged approximately itors (SSRIs) might cause small and transient increases in 25%, the “actual” corrected baseline level of plasma 5-HT in plasma 5-HT. It is unlikely that SSRI-induced increases in our experiments is likely in the range of 0.88 nM. This level plasma 5-HT contribute to the effects of this is comparable with plasma 5-HT levels in humans, which are class of medication. However, it is tempting to speculate that reportedly in the subnanomolar range (Herve et al., 1995). SSRI-induced increases in plasma 5-HT contribute to the With the exception of phentermine, all of the test drugs ability of SSRIs to increase ejaculatory threshold (de Jong et produced dose-dependent increases in plasma 5-HT when al., 2006). administered in vivo (see Fig. 1). These findings differ from Our findings provide at least partial confirmation of the those of Martin and Artigas (1992), who did not observe 5-HT hypothesis that fenfluramine can increase plasma elevations in plasma 5-HT after administration of (ϩ)-fenflu- 5-HT. However, it is possible that the ability of fenfluramine ramine (2.5 mg/kg) to rats. The reason for the discrepancy and other SERT substrates to increase plasma 5-HT might be between our results and those of others is not known but different after chronic administration. In this instance, plate- could be related to differences in blood sampling procedures. let 5-HT is markedly reduced (Raleigh et al., 1986; Celada et Specifically, we obtained blood samples within 15 min of al., 1994), and it is not yet known if plasma 5-HT is reduced fenfluramine administration, whereas Martin and Artigas under these circumstances or if the ability of fenfluramine to sampled 60 min after injection, possibly missing the peak increase plasma 5-HT will be attenuated. In the case of effect of the drug on circulating 5-HT. chronic treatment with SERT inhibitors, both platelet and The ability of amphetamine analogs to increase plasma plasma 5-HT are markedly reduced (Celada et al., 1992a). 5-HT in vivo is correlated with their capacity to increase Regarding PPH, another factor to consider is whether the plasma 5-HT in vitro and with their potency at releasing maximal drug-induced plasma 5-HT concentrations ap- [3H]5-HT from synaptosomes (see Fig. 4). Because platelets proach those needed to contract pulmonary arteries or to and neurons express the same SERT protein (Lesch et al., stimulate mitogenesis in pulmonary artery endothelium or 1993), the potency of agents at releasing [3H]5-HT from smooth muscles. The maximal concentration of drug-induced platelets and from synaptosomes should be similar. Indeed, dialysate 5-HT was ϳ5 nM when high i.v. drug doses were the EC50 value reported by Schuldiner et al. (1993) for fen- administered; correcting for probe recovery estimates, this fluramine-induced release of [3H]5-HT from platelets (170 value is increased to ϳ20 nM. Serotonin contracts human nM) is similar to the value we reported for fenfluramine in pulmonary arteries with EC50 values that range from ap- Amphetamines Increase Plasma Serotonin 609 proximately 100 nM (Morecroft et al., 1999) to the micromo- amine other hypotheses concerning the pathogenesis of fen- lar range (Cortijo et al., 1997). Thus, it would seem unlikely fluramine-associated VHD. Ulus et al. (2000) suggested that that fenfluramine and other SERT substrates could increase phentermine and other amphetamines block MAO activity in plasma 5-HT to a concentration that would directly contract vivo, and this action could contribute to the development of pulmonary arteries. VHD when phentermine and fenfluramine are administered Recent studies indicate that SERT plays a key role in together. MAO is an enzyme that metabolizes 5-HT to mitogenic effect of 5-HT on pulmonary artery smooth muscle 5-HIAA. Inhibition of MAO produces measurable decreases cells, and this effect can be prevented by SERT inhibitors in plasma 5-HIAA. Several types of evidence obtained in rat fluoxetine and but not by the 5-HT2A receptor nervous tissue suggest that phentermine and fenfluramine antagonist (Eddahibi et al., 2002). The threshold do not affect MAO (Baumann et al., 2000; Kilpatrick et al., concentration for 5-HT-stimulated mitogenic responses in 2001), and the present data confirm that high i.v. doses of cultured human pulmonary artery smooth muscle cells is these drugs do not affect plasma 5-HIAA (see Fig. 2), consis- approximately 10 nM (Eddahibi et al., 2001; Marcos et al., tent with data from nonhuman primates (Alexander et al., 2003), although higher 5-HT (ϳ100 nM) concentrations are 2005). Other investigators proposed that phentermine would needed to stimulate mitogenic responses in rat pulmonary enhance fenfluramine-induced increases in plasma 5-HT artery smooth muscle cells (Pitt et al., 1994; Eddahibi et al., (Fishman, 1999; Ulus et al., 2000). Our data clearly show 1999). It seems possible that high doses of fenfluramine, that phentermine does not enhance the rise in plasma 5-HT MDMA, and methamphetamine could transiently produce produced by fenfluramine. plasma 5-HT concentrations sufficient to stimulate mitogenic In summary, contrary to previously published data (Mar- Downloaded from responses in pulmonary smooth muscle cells. Consequently, tin and Artigas, 1992), we demonstrate that fenfluramine these drugs could increase the risk of developing PPH in and other SERT substrates produce significant dose-depen- susceptible individuals, should the exposure to higher than dent increases in plasma 5-HT when administered under in normal plasma 5-HT continue long enough. vivo and in vitro conditions. This effect most likely involves Regarding VHD, it is well established that 5-HT can in- SERT-mediated exchange of drug molecules for platelet duce mitogenic responses in human and animal heart inter- 5-HT. Agents with less potent SERT substrate activity, such jpet.aspetjournals.org stitial cells in vitro (Rajamannan et al., 2001; Jian et al., as amphetamine and phentermine, are considerably less ef- 2002; Setola et al., 2003). Although mitogenic responses have fective in elevating plasma 5-HT compared with fenflura- been observed with 5-HT concentrations as low as 10 nM mine, MDMA, and methamphetamine. The fact that MDMA (Hafizi et al., 2000; Rajamannan et al., 2001), other studies and methamphetamine increase plasma 5-HT to concentra- report that 5-HT concentrations in the micromolar range are tions that stimulate mitogenic responses in pulmonary ar- required to stimulate mitogenic responses (Jian et al., 2002; tery smooth muscle cells suggests that this effect may con- Setola et al., 2003). Importantly, cells must be exposed to tribute to the cardiovascular toxicities associated with illicit at ASPET Journals on May 26, 2015 5-HT for 24 to 48 h to produce the mitogenic response, a use of these drugs. Further research will be needed to eluci- period of time far greater than the 60 to 90 min of elevated date the physiological significance of our findings. plasma 5-HT produced by administration of fenfluramine and the other substituted amphetamines. Thus, it seems Acknowledgments unlikely that a single i.v. dose of fenfluramine, MDMA, or We thank Michelle Nunez and Robert Clark for expert technical methamphetamine could transiently produce high enough assistance during the initial phases of this study. plasma 5-HT concentrations to stimulate mitogenic re- sponses in cardiac valvular cells. This interpretation of our References data is supported by the fact that high 5-HT concentrations Alexander M, Rothman RB, Baumann MH, Endres CJ, Brasic JR, and Wong DF Ͼ (2005) Noradrenergic and effects of (ϩ)-amphetamine-like stimu- ( 500 nM) are required to produce valvulopathy in carcinoid lants in the baboon Papio anubis. Synapse 56:94–99. syndrome (Robiolio et al., 1995). Modest 2- to 3-fold eleva- Artigas F, Sarrias MJ, Martinez E, Gelpi E, Alvarez E, and Udina C (1989) Increased plasma free serotonin but unchanged platelet serotonin in bipolar patients treated tions of plasma 5-HT, such as those occurring with the pre- chronically with lithium. Psychopharmacology (Berl) 99:328–332. scribed use of lithium (Artigas et al., 1989) and MAO inhib- Baumann MH, Ayestas MA, Dersch CM, Brockington A, Rice KC, and Rothman RB (2000) Effects of phentermine and fenfluramine on extracellular and itors (Celada et al., 1992b) are not associated with VHD. Our serotonin in rat : therapeutic implications. Synapse 36:102– previous findings indicated that medications known to cause 113. VHD (fenfluramine, , ) produce me- Baumann MH, Ayestas MA, Dersch CM, and Rothman RB (2001) 1-(m-Chlorophe- nyl) (mCPP) dissociates in vivo serotonin release from long-term sero- tabolites that are 5-HT2B receptor (Rothman et al., tonin depletion in rat brain. Neuropsychopharmacology 24:492–501. 2000a; Setola et al., 2003). More recently, two medications Berger UV, Gu XF, and Azmitia EC (1992) The substituted amphetamines 3,4- methylenedioxymethamphetamine, methamphetamine, p-chloroamphetamine used to treat Parkinson’s disease, and , and fenfluramine induce 5-hydroxytryptamine release via a common mechanism were identified as 5-HT agonists that also cause VHD (for blocked by fluoxetine and . Eur J Pharmacol 215:153–160. 2B Celada P, Dolera M, Alvarez E, and Artigas F (1992a) Effects of acute and chronic review, see Setola and Roth, 2005). Collectively the data treatment with on extracellular and platelet serotonin in the blood of indicate that fenfluramine produces VHD through the acti- major depressive patients: relationship to clinical improvement. J Affect Disord 25:243–249. vation of cardiac valvular 5-HT2B receptors by its metabolite, Celada P, Martin F, and Artigas F (1994) Effects of chronic treatment with dexfen- norfenfluramine, and not from elevation of plasma 5-HT. On fluramine on serotonin in rat blood, brain and lung tissue. Life Sci 55:1237–1243. Celada P, Perez J, Alvarez E, and Artigas F (1992b) Monoamine oxidase inhibitors the other hand, the translocation of 5-HT transporter sub- and brofaromine increase plasma serotonin and decrease 5-hydroxyin- strates into cells in exchange for endogenous 5-HT and the doleacetic acid in patients with major : relationship to clinical improve- ment. J Clin Psychopharmacol 12:309–315. subsequent “trapping” of these agents inside of cells could Connolly HM, Crary JL, McGoon MD, Hensrud DD, Edwards BS, Edwards WD, and contribute to the mechanism of toxicity (Rothman et al., Schaff HV (1997) Valvular heart disease associated with fenfluramine- phentermine. N Engl J Med 337:581–588. 1999). Cooper JR, Bloom F, and Roth RH (2003) The Biochemical Basis of Neuropharma- Our experiments provided the unique opportunity to ex- cology, 8th ed, Oxford University Press, New York. 610 Zolkowska et al.

Cortijo J, Marti-Cabrera M, Bernabeu E, Domenech T, Bou J, Fernandez AG, Beleta receptors mediating contraction in human small muscular pulmonary J, Palacios JM, and Morcillo EJ (1997) Characterization of 5-HT receptors on arteries: importance of the 5-HT1B receptor. Br J Pharmacol 128:730–734. human pulmonary artery and vein: functional and binding studies. Br J Pharma- Pitt BR, Weng W, Steve AR, Blakely RD, Reynolds I, and Davies P (1994) Serotonin col 122:1455–1463. increases DNA synthesis in rat proximal and distal pulmonary vascular smooth Crespi D, Mennini T, and Gobbi M (1997) Carrier-dependent and Ca2ϩ-dependent muscle cells in culture. Am J Physiol 266:L178–L186. 5-HT and dopamine release induced by (ϩ)-amphetamine, 3,4-methyl- Rajamannan NM, Caplice N, Anthikad F, Sebo TJ, Orszulak TA, Edwards WD, Tajik enedioxymethamphetamine, p-chloroamphetamine and (ϩ)-fenfluramine. Br J J, and Schwartz RS (2001) Cell proliferation in carcinoid valve disease: a mecha- Pharmacol 121:1735–1743. nism for serotonin effects. J Heart Valve Dis 10:827–831. de Jong TR, Veening JG, Waldinger MD, Cools AR, and Olivier B (2006) Serotonin Raleigh MJ, Brammer GL, Ritvo ER, Geller E, McGuire MT, and Yuwiler A (1986) and the neurobiology of the ejaculatory threshold. Neurosci Biobehav Rev, in press. Effects of chronic fenfluramine on blood serotonin, cerebrospinal fluid metabolites Eddahibi S, Fabre V, Boni C, Martres MP, Raffestin B, Hamon M, and Adnot S and behavior in monkeys. Psychopharmacology (Berl) 90:503–508. (1999) Induction of serotonin transporter by hypoxia in pulmonary vascular Robiolio PA, Rigolin VH, Wilson JS, Harrison JK, Sanders LL, Bashore TM, and smooth muscle cells: relationship with the mitogenic action of serotonin. Circ Res Feldman JM (1995) Carcinoid heart disease: correlation of high serotonin levels 84:329–336. with valvular abnormalities detected by cardiac catheterization and echocardiog- Eddahibi S, Humbert M, Fadel E, Raffestin B, Darmon M, Capron F, Simonneau G, raphy. Circulation 92:790–795. Dartevelle P, Hamon M, and Adnot S (2001) Serotonin transporter overexpression Rothman RB, Ayestas MA, Dersch CM, and Baumann MH (1999) Aminorex, fenflu- is responsible for pulmonary artery smooth muscle hyperplasia in primary pul- ramine and are serotonin transporter substrates: implications monary hypertension. J Clin Investig 108:1141–1150. for primary pulmonary hypertension. Circulation 100:869–875. Rothman RB and Baumann MH (2003) Monoamine transporters and psychostimu- Eddahibi S, Raffestin B, Hamon M, and Adnot S (2002) Is the serotonin transporter lant drugs. Eur J Pharmacol 479:23–40. involved in the pathogenesis of pulmonary hypertension? J Lab Clin Med 139: Rothman RB, Baumann MH, Dersch CM, Romero DV, Rice KC, Carroll FI, and 194–201. Partilla JS (2001) Amphetamine-type central nervous system release Fishman AP (1999) Aminorex to fen/phen: an epidemic foretold. Circulation 99:156– more potently than they release dopamine and serotonin. Synapse 161. 39:32–41. Gershon MD (2004) Review article: serotonin receptors and transporters: roles in Rothman RB, Baumann MH, Savage JE, Rauser L, McBride A, Hufeisen SJ, and normal and abnormal gastrointestinal motility. Aliment Pharmacol Ther 20 Roth BL (2000a) Evidence for possible involvement of 5-HT2B receptors in the (Suppl 7):3–14. cardiac valvulopathy associated with fenfluramine and other serotonergic medi- Downloaded from Hafizi S, Taylor PM, Chester AH, Allen SP, and Yacoub MH (2000) Mitogenic and cations. Circulation 102:2836–2841. secretory responses of human valve interstitial cells to vasoactive agents. J Heart Rothman RB, Blough BE, Woolverton WL, Anderson KG, Negus SS, Mello NK, Roth Valve Dis 9:454–458. BL, and Baumann MH (2005) Development of a rationally designed, low abuse Herve P, Launay JM, Scrobohaci ML, Brenot F, Simonneau G, Petitpretz P, Poubeau potential, biogenic amine releaser that suppresses cocaine self-administration. P, Cerrina J, Duroux P, and Drouet L (1995) Increased plasma serotonin in J Pharmacol Exp Ther 313:1361–1369. primary pulmonary hypertension. Am J Med 99:249–254. Rothman RB, Redmon JB, Raatz SK, Kwong CA, Swanson JE, and Bantle JP (2000b) Jian B, Xu J, Connolly J, Savani RC, Narula N, Liang B, and Levy RJ (2002) Chronic treatment with phentermine combined with fenfluramine lowers plasma Serotonin mechanisms in heart valve disease: I. Serotonin-induced up-regulation serotonin. Am J Cardiol 85:913–915, A910.

of transforming growth factor-beta1 via G-protein signal transduction in aortic Schuldiner S, Steiner-Mordoch S, Yelin R, Wall SC, and Rudnick G (1993) Amphet- jpet.aspetjournals.org valve interstitial cells. Am J Pathol 161:2111–2121. amine derivatives interact with both plasma membrane and secretory vesicle Kilpatrick IC, Traut M, and Heal DJ (2001) Monoamine oxidase inhibition is un- biogenic amine transporters. Mol Pharmacol 44:1227–1231. likely to be relevant to the risks associated with phentermine and fenfluramine: a Setola V, Hufeisen SJ, Grande-Allen KJ, Vesely I, Glennon RA, Blough B, Rothman comparison with their abilities to evoke monoamine release. Int J Obes Relat RB, and Roth BL (2003) 3,4-Methylenedioxymethamphetamine (MDMA, “Ec- Metab Disord 25:1454–1458. stasy”) induces fenfluramine-like proliferative actions on human cardiac valvular Lesch KP, Wolozin BL, Murphy DL, and Reiderer P (1993) Primary structure of the interstitial cells in vitro. Mol Pharmacol 63:1223–1229. human platelet serotonin uptake site: identity with the brain serotonin trans- Setola V and Roth BL (2005) Screening the receptorome reveals molecular targets porter. J Neurochem 60:2319–2322. responsible for drug-induced side effects: focus on “fen-phen.” Expert Opin Drug MacLean MR, Herve P, Eddahibi S, and Adnot S (2000) 5-Hydroxytryptamine and Metab Toxicol 1:377–387.

the pulmonary circulation: receptors, transporters and relevance to pulmonary Ulus IH, Maher TJ, and Wurtman RJ (2000) Characterization of phentermine and at ASPET Journals on May 26, 2015 arterial hypertension. Br J Pharmacol 131:161–168. related compounds as monoamine oxidase (MAO) inhibitors. Biochem Pharmacol Marcos E, Adnot S, Pham MH, Nosjean A, Raffestin B, Hamon M, and Eddahibi S 59:1611–1621. (2003) Serotonin transporter inhibitors protect against hypoxic pulmonary hyper- tension. Am J Respir Crit Care Med 168:487–493. Address correspondence to: Dr. Michael H. Baumann, Clinical Psycho- Martin F and Artigas F (1992) Simultaneous effects of p-chloroamphetamine, D- pharmacology Section, IRP, NIDA, National Institutes of Health, Department fenfluramine and on free and stored 5-hydroxytryptamine in brain and of Health and Human Services, 5500 Nathan Shock Drive, Baltimore, MD blood. J Neurochem 59:1138–1144. 21224. E-mail: [email protected] Morecroft I, Heeley RP, Prentice HM, Kirk A, and MacLean MR (1999) 5-Hydroxy-