INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

THE USE OF THE ANOREXICS dexfenfluramine or fenfluramine for >3 mo increases the risk ratio of developing primary pulmonary hypertension by a factor of 23 (1). The pathogenesis of primary pulmonary hypertension is poorly understood, and, consequently, the possibilities for treatment are limited (13, 15). It has previously been shown that aminorex, fenfluramine, and dexfenfluramine inhibit K⁺ currents in rat pulmonary artery (PA) smooth muscle cells (SMCs) and that dexfenfluramine causes reversible membrane depolarization in these cells (12, 16). Although an increase in vascular smooth muscle intracellular Ca²⁺ concentration ([Ca²⁺]_i) would be expected as a result of membrane depolarization and as a prerequisite for vasoconstriction, changes in [Ca²⁺]_i in response to dexfenfluramine have not been determined. In this study, we measured the changes in [Ca²⁺]_i elicited by dexfenfluramine in rat PASMCs and evaluated the relative contribution of extracellular Ca²⁺ influx versus release of Ca²⁺ from intracellular stores. K⁺ currents and membrane potentials were also measured with the amphotericin-perforated whole cell patch-clamp technique (11).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Isolation of rat PASMCs. All animal studies were conducted in accordance with institutional guidelines. Fresh rat PASMCs were dissociated in the same manner for both Ca²⁺ and patch-clamp studies. Male Sprague-Dawley rats (250-350 g) were anesthetized with pentobarbital sodium (50 mg/kg), and their heart and lungs were removed en bloc. Resistance PAs (third to fifth division) were dissected, cleaned of connective tissue, cut open, and placed in "Ca²⁺-free" Hanks' solution composed of (in mM) 145 NaCl, 4.2 KCl, 1.0 MgCl₂, 1.2 KH₂PO₄, 10 HEPES, and 0.1 EGTA (pH 7.4 with NaOH) for 10 min at 4°C. The arteries were then transferred to Hanks' solution without EGTA ("low Ca²⁺") containing 1 mg/ml of papain, 0.8 mg/ml of albumin, and 0.75 mg/ml of dithiothreitol and kept at 4°C for 20 min. Subsequently, the arteries were incubated at 36°C for 12 min. After digestion, the cells were washed in low-Ca²⁺ Hanks' solution to remove residual enzymes and maintained at 4°C. The arteries were triturated before the experiments to produce a suspension of single cells.

Measurement of intracellular Ca²⁺. [Ca²⁺]_i was measured by dual-excitation imaging with fura 2 (4). Freshly dispersed cells were transferred to imaging dishes (Molecular Probes, Eugene OR) and incubated in low-Ca²⁺ Hanks' solution with the cell-permeable acetoxymethyl ester form of fura 2 (0.1 µM) and Pluronic F-127 (0.8 µM) for 15 min at room temperature. The plates were then washed with HEPES buffer containing 1.5 mM Ca²⁺ (see solution composition in Electrophysiology) and incubated at room temperature for a further 20 min. The plates were then washed again and placed on a heated microscope stage (33°C). The drugs were added directly to the cells as a bolus by microinjection. All drugs were given in 10-µl volumes to remove potential volume-induced artifacts. Ten-microliter injections of saline had no effect on [Ca²⁺]_i. Changes in [Ca²⁺]_i were recorded in individual cells with a MetaFluor (Universal Imaging, West Chester, PA)-driven 340/380 filter imaging system and cooled charge-coupled device camera (Photometrics, Tucson, AZ). This system allows discrete areas within single cells to be imaged so that relaxed SMCs can be identified while other cell types (endothelial cells and fibroblasts) can be excluded. Background fluorescence was recorded from each dish of cells and subtracted before calculation of the 340- to 380-nm ratio. Measurements were made every 5 s. [Ca²⁺]_i was calculated according to the method of Grynkiewicz et al. (4). A dissociation constant of 325 nM was calculated from the in vitro calibration. Maximal and minimal ratio values were determined at the end of each experiment by first treating the cells with 1 µM ionomycin (maximal ratio) and then chelating all free Ca²⁺ with 2 mM EGTA (minimal ratio). For studies with Ca²⁺-free medium, the extracellular solution had no added Ca²⁺ and was supplemented with the Ca²⁺ chelator EGTA (10 mM) and 1.5 mM MgCl₂ to replace the Ca²⁺. Cells were exposed to the Ca²⁺-free medium for 1 min before the addition of dexfenfluramine.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Dexfenfluramine and intracellular Ca²⁺ in rat PASMCs. Dexfenfluramine caused a dose-dependent elevation in [Ca²⁺]_i in rat PASMCs, with an initial increase in Ca²⁺ at 1 µM (P < 0.05; n = 46 cells; Fig. 1). Thirty-seven of the forty-six cells tested in this way (80.4%) responded to 1 µM dexfenfluramine. The source of the rise in [Ca²⁺]_i was determined by multiple recordings of the dexfenfluramine-mediated responses after blockade of the influx of Ca²⁺ or release of the intracellular pools of Ca²⁺. At 100 µM (average rise in Ca²⁺ = 1,234 ± 122 nM; n = 25 cells), pretreatment of the cells for 1 min with 1 mM CoCl₂ to block the influx of extracellular Ca²⁺ reduced but did not abolish the elevation in [Ca²⁺]_i in response to 100 µM dexfenfluramine (Fig. 2A). In the absence of extracellular Ca²⁺, i.e., with the cells bathed in Ca²⁺-free perfusate (see METHODS), there was a similar reduction in the response to 100 µM dexfenfluramine (Fig. 2A). The component of the rise in [Ca²⁺]_i that was not prevented by removal or blockade of the influx of extracellular Ca²⁺ was investigated after pretreatment of the cells with thapsigargin (to deplete all intracellular Ca²⁺ stores) or caffeine (to deplete the ryanodine-sensitive stores). Thapsigargin (0.1 µM) caused a slow, sustained rise in [Ca²⁺]_i consistent with a leak of Ca²⁺ from the intracellular stores and then a block of reuptake (n = 11 cells) (14). The increase in [Ca²⁺]_i caused by 1 and 10 µM dexfenfluramine was completely inhibited by pretreatment of the cells with 0.1 µM thapsigargin (Fig. 3A), whereas the response to 100 µM dexfenfluramine was significantly reduced from the control value (n = 11 cells; Figs. 2B and 3A) but was not abolished. Caffeine (1 mM) caused a classic, transient increase in [Ca²⁺]_i, after which the response to 100 µM dexfenfluramine was also significantly reduced to 454 ± 57 nM (P < 0.05; Figs. 2B and 3B).

View larger version (10K): [in this window] [in a new window]   Fig. 1.   Changes in intracellular Ca2+ concentration ([Ca2+]i) after sequential applications of 1, 10, and 100 µM dexfenfluramine. [dexfenfluramine], Dexfenfluramine concentration. Values are means ± SE; n = 37 cells. * P < 0.05 compared with baseline.

View larger version (27K): [in this window] [in a new window]   Fig. 2.   [Ca2+]i to 100 µM dexfenfluramine (DEX) alone and after pretreatment of cells with Ca2+-free solution and CoCl2 (A) or caffeine, thapsigargin, and ketanserin (B). Values are means ± SE; nos. in parentheses, no. of cells. * P < 0.05 compared with DEX response.

View larger version (11K): [in this window] [in a new window]   Fig. 3.   A: representative example of actual changes in 340- to 380-nm (340/380) ratio to DEX after pretreatment with thapsigargin. B: representative example of actual changes in 340- to 380-nm ratio to DEX after pretreatment with caffeine. Arrows, drug application.

Dexfenfluramine, K⁺ currents, and membrane potential in rat PASMCs. With the use of the amphotericin- perforated patch clamp, outward currents were recorded from single PASMCs. The average current density was 265 ± 33 pA/pF (at +50 mV; n = 15 cells), and the outward currents could be almost completely inhibited by 1 mM 4-aminopyridine. Figure 4A shows a dose-response curve of the inhibitory effect of dexfenfluramine. In contrast to the rise in [Ca²⁺]_i observed with 1 µM dexfenfluramine, the currents were only inhibited at concentrations of 10 µM and above as previously described (12, 16). The average resting membrane potential recorded from the PASMCs was 46.1 ± 2 mV (n = 6 cells). The membrane potential was recorded for 1 min to establish stability, and then doses of 1, 10, and 100 µM dexfenfluramine were applied to the cell for 2 min each. Each dose was washed out before the next one was applied. There was no effect of dexfenfluramine on membrane potential until 100 µM (Fig. 4B).

View larger version (23K): [in this window] [in a new window]   Fig. 4.   A: percentage of K+ current (IK) inhibited at 10 mV by increasing doses of DEX in rat pulmonary artery smooth muscle (n = 3 cells/dose). * Significant change in IK inhibition, P < 0.05. B: membrane potential (Em) recorded during control (n = 6 cells) or after 2-min exposure to 1 (n = 3 cells), 10 (n = 5 cells), or 100 (n = 6 cells) µM DEX. Values are means ± SE. * Significant change in Em, P < 0.05.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The mechanism of pulmonary hypertension associated with dexfenfluramine intake remains to be elucidated. The present study shows that dexfenfluramine increases [Ca²⁺]_i in rat PASMCs. A similar dose-dependent increase was observed in cultured human SMCs (Soper and Archer, unpublished observations), but because the cells used were an immortalized cell line taken from conduit human PAs, the full study was undertaken with freshly dispersed rat PASMCs. Increased [Ca²⁺]_i is an important signal that activates the contractile apparatus but can also stimulate cellular proliferation (7, 8). Both vasoconstriction and mesenchymal cell proliferation are important in the pathogenesis of pulmonary hypertension (15).

The sustained plasma concentration of fenfluramine that "correlates with the best rate of weight loss" is said to be 1 µM (2). We show an increase in [Ca²⁺]_i at 1 µM dexfenfluramine, whereas there is little effect on K⁺ current and no effect on membrane potential at this concentration. This change at 1 µM dexfenfluramine could be completely prevented by prior release of intracellular stores by 0.1 µM thapsigargin (14, 18), suggesting that release occurs before influx of extracellular Ca²⁺ via sarcolemmal depolarization. Furthermore, blockade of the influx of extracellular Ca²⁺ by CoCl₂ or removal of extracellular Ca²⁺ reduced (by ~58%) but did not abolish the increase in [Ca²⁺]_i caused by 100 µM dexfenfluramine, suggesting that entry of extracellular Ca²⁺ is a source of dexfenfluramine-induced increase in [Ca²⁺]_i at a high concentration but that release of intracellular Ca²⁺ also contributes. It should be noted that thapsigargin has been reported to inhibit L-type Ca²⁺ channels at high concentrations (5 µM) (14). The concentration of thapsigargin used (0.1 µM) was chosen to avoid this potentially misleading effect. Caffeine depletes the sarcoplasmic reticulum of Ca²⁺ by binding to the ryanodine receptor (10). One millimolar caffeine also reduced the response to 100 µM dexfenfluramine, suggesting a role of the ryanodine-sensitive intracellular stores. Serotonin-induced rises in intracellular Ca²⁺ in cultured PASMCs have been suggested to be primarily due to release of inositol 1,4,5-trisphosphate stores (17). These data presented here suggest that dexfenfluramine, at least in isolated PASMCs, increases intracellular Ca²⁺ through a mechanism different from that of serotonin. A previous study (9) in canine PASMCs has suggested that the inhibition of K⁺ current that occurs in hypoxic pulmonary vasoconstriction may be secondary to a rise in [Ca²⁺]_i. An elegant study by the same group (5) indicated roles for caffeine and ryanodine-sensitive stores in hypoxic pulmonary vasoconstriction. Although it seems likely that dexfenfluramine releases intracellular Ca²⁺ before its inhibition of K⁺ current, it remains to be determined whether this rise in Ca²⁺ is directly responsible for the inhibition of K⁺-channel activity.

It has recently been reported that patients with both primary pulmonary hypertension (6) and anorexic-induced pulmonary hypertension (3) may have impaired endothelial function as indicated by reduced endogenous nitric oxide levels. In light of these data, it could be speculated that a small rise in [Ca²⁺]_i may occur at plasma levels of dexfenfluramine that occur clinically but that significant pulmonary constriction only occurs in the absence of normal, compensatory vasodilator mechanisms.