Chloride increases adrenergic receptor–mediated platelet and vascular responses^*

Abstract

Background:

We postulated that increasing intracellular chloride concentration ([Cl⁻]_i) in human platelets would potentiate α₂ adrenergic receptor (A2AR)—mediated platelet aggregation, and that vascular reactivity would also be increased by raising [Cl⁻]_i in blood vessels. We further hypothesized that ligands binding to the A2AR would increase [Cl⁻]_i by stimulating carbonic anhydrase-dependent chloride/bicarbonate exchange. Because diuretics are potent inhibitors of carbonic anhydrase, we speculated that these agents inhibit platelet aggregation and vascular contractility through inhibition of chloride influx by decreasing carbonic anhydrase activity, and subsequently, chloride/bicarbonate exchange. The aim of this study was to test these hypotheses.

Methods:

Platelet aggregation was measured by determining changes in optical density of platelet-rich plasma. Contractile responses to A2AR agonists were recorded in isolated vascular smooth muscle. The substances [Cl⁻]_i and intracellular pH (pH_i) were measured using microfluorometric methods. Carbonic anhydrase activity and chloride/bicarbonate exchange were determined by an in vitro assay based on the Stewart cycle.

Results:

Increasing [Cl⁻]_i potentiated platelet aggregation and vascular contractility, and epinephrine raised [Cl⁻]_i by stimulating carbonic anhydrase-dependent chloride/bicarbonate exchange. Furthermore, diuretic-dependent inhibition of carbonic anhydrase activity decreased chloride/bicarbonate exchange.

Conclusions:

Our data support the concept that diuretics inhibit carbonic anhydrase activity and chloride/bicarbonate exchange in platelets and vascular smooth muscle. The ensuing reduction in [Cl⁻]_i that is induced by diuretics in these tissues could play a role in reducing the effect of catecholamines on precipitating thrombotic stroke or myocardial infarction. Am J Hypertens 2002;15:492–498 © 2002 American Journal of Hypertension, Ltd.

The α₂ adrenergic receptor (A2AR) binds norepinephrine and epinephrine; the subsequent steps involved in postreceptor signal transduction induced by these ligands are poorly understood. We previously demonstrated that inhibition of chloride transport decreases epinephrine-mediated platelet aggregation.¹ It is possible that changes in the intracellular chloride concentrations ([Cl⁻]_i) play a significant role in A2AR-mediated signal transduction in other tissues such as blood vessels. [Cl⁻]_i is maintained primarily by coupling the transport of chloride to the sodium gradient and via the Cl⁻/HCO₃ exchanger. There are also a number of passive chloride channels that facilitate chloride transport across the plasma membrane according to the electrochemical gradient. These chloride transporters and channels are classified according to their mechanism of activation and amino acid sequence data.^2–4 In this study, we further examine the roles of [Cl⁻]_i and the chloride/bicarbonate exchanger on mediating A2AR-dependent platelet aggregation and vascular contraction.

Methods

All reagents were from Sigma Chemical Co. (St. Louis, MO). To study the role of [Cl⁻]_i on vascular contractility, we used adult male Sprague-Dawley rats weighing 275 g (Harlan Industries, Indianapolis, IN or Charles River Laboratories, Wilmington, MA). The rats were anesthetized, and the tail arteries or aortae were excised, dissected free of loose connective tissue, and cut into helical strips; these were then suspended in a muscle bath and attached to force transducers (model FT03, Grass Instruments, Quincy, MA) for measurement of isometric tension generated. The tissues were incubated in a standard physiologic salt solution (PSS) of the following composition (in mmol/L): NaCl, 130; KCl 4.7; NaH₂PO₄, 1.18; MgSO₄-7H₂O, 1.7; CaCl₂-2H₂0, 1.6; NaHCO₃, 14.9; dextrose, 5.5; and the PSS was gased with 95%-5% O₂-CO₂ at 37° C. In some experiments, the vascular smooth muscle was preincubated for ≥20 min in a PSS in which the NaCl was iso-osmotically replaced with sodium gluconate.

Differences in epinephrine-mediated platelet aggregation were determined by measuring changes in optical density using standard aggregometry techniques (Chronolog Instruments, Havertown, PA). We used a chloride sensitive fluorescent dye N-ethoxycarbonylmethyl-6-methoxy-quinolium bromide to determine [Cl⁻]_i in platelets as described.^1,5 These protocols were approved by our institutional committee for the protection of human subjects in research, and each volunteer gave informed consent. All subjects were healthy and none were taking any medication. Blood was drawn after the subjects rested for 1 h. Fluorometric data were obtained with a spectrofluorometer (Spex Industries, Edison, NJ) at 37°C. In a few experiments, intracellular pH (pH_i) was also measured using a pH-sensitive dye, 2′-7′bis(2-carboxyethyl)-5 (6)carboxy-fluorescein, as described elsewhere.⁶

Carbonic anhydrase hydratase activity was monitored directly by substrate-dependent changes in pH, or esterase activity was measured in vitro by determining the rate of hydrolysis of phenylacetate by erythrocyte lysates isolated from whole blood from healthy volunteers.^7–9 Carbonic anhydrase II is the subtype of enzyme that is responsible for the majority of the dehydratase activity in erythrocytes, and it is the major subtype found in platelets.¹⁰ To determine the effect of epinephrine on carbonic anhydrase-dependent anion exchange in platelets, an in situ assay was used. Briefly, we measured the rate of swelling of human platelets placed in a PSS with 15 mmol/L HCO₃⁻, pH 7.4, and in which NaCl was replaced with an equimolar, iso-osmotic concentration of NH₄Cl.^11,12 In aqueous solutions of NH₄Cl, free NH₃ easily diffuses across cell membranes. At physiologic pH, intracellular NH₃ is protonated by the H⁺ that is generated from the hydration of CO₂ by intracellular carbonic anhydrase, and NH₄⁺ is formed. This impermeant molecule is osmotically active and, as a result of NH₄⁺ accumulation, the platelet swells. This increase in platelet size correlates with the H⁺ made available by the intracellular carbonic anhydrase. Changes in median platelet size were followed over varying time periods in a model ZM Coulter Counter with a 70-μm aperture and Channelyzer 256 cell sizing instrument (Coulter Industries, Hialeah, FL).

Data were analyzed by the Prism GraphPad (San Diego, CA) statistical analysis program. Group comparisons were made with a paired t test, and when multiple comparisons were made, a Bonferroni adjustment was made. EC50 values were calculated from log transformation of dose response curves, and the EC50s for the various treatment groups were similarly compared using a paired t test. Values were considered to be significant at P < .05.

Results

Iso-osmotic replacement of NaCl with sodium gluconate significantly decreased the magnitude of the contractile responses of isolated rat aortae to the A2AR agonist clonidine. Maximal contractile responses to clonidine were reduced (in milligrams of tension ± SEM) from 289 ± 18 to 120 ± 11, P < .05 (Fig. 1). Recognizing that gluconate can bind calcium ions, we measured the free calcium concentration in chloride-free PSS with an ion-specific electrode; the ionized calcium concentration fell from 1.6 mmol/L in PSS with chloride to approximately 0.9 mmol/L in the PSS in which the NaCl had been iso-osmotically replaced with sodium gluconate. However, in a separate set of experiments, this degree of reduction in the total extracellular calcium content did not significantly diminish contractile responses to clonidine (data not shown).

Effect of sodium gluconate on contractile response to clonidine. Iso-osmotic substitution of gluconate for chloride significantly inhibited the contractile responses mediated by the postsynaptic α₂ adrenergic receptor in isolated aortae. Although maximum contractile responses to clonidine were diminished in the chloride-free physiologic salt solution, the EC50 values were significantly diminished from (in μmol/L ± SEM) 0.55 ± 0.10 to 0.49 ± 0.1, P < .005.

We previously reported that acetazolamide, a potent inhibitor of carbonic anhydrase, decreased basal [Cl⁻]_i and epinephrine-stimulated accumulation of [Cl⁻]_i in platelets.¹ Assuming that similar effects would be seen in vascular smooth muscle, we determined that clonidine-mediated contractions of isolated rat aortae were indeed sensitive to inhibition with acetazolamide. Acetazolamide significantly decreased the maximum contractile responses to clonidine from (in milligrams of force ± SEM) 311 ± 26 to 127 ± 12, P < .05 (Fig. 2).

The effect of acetazolamide on α₂ adrenergic receptor-mediated contractile responses. Inhibition of carbonic anhydrase with acetazolamide also inhibited clonidine-induced contractile responses in isolated aortae. Acetazolamide not only decreased maximal responses to clonidine, but inhibition of carbonic anhydrase significantly shifted the EC50 values from 0.26 ± 0.08 to 0.51 ± 0.09, P < .0001.

We next demonstrated that increases in [Cl⁻]_i would augment clonidine-induced contractile responses. Agents that attack membrane thiol groups increase chloride transport in various cell lines.^13,–15 We found that micromolar concentrations of HgCl₂ significantly increased [Cl⁻]_i above baseline values in isolated platelets (Fig. 3). The addition of 50 μmol/L HgCl₂ increased intraplatelet chloride concentrations by nearly 25% over 5 min. Dose-dependent increases in platelet [Cl⁻]_i were also noted with other mercurials such as phenylmercuric acetate and thimerosal (data not shown). When isolated rat tail arteries were incubated with 100 μmol/L phenylmercuric acetate, contractile responses to clonidine were significantly increased (Fig. 4). In the presence of this mercurial, maximal contractile responses to this A2AR agonist were increased from (in milligrams of tension ± SEM) 1223 ± 116 to 1548 ± 113, P < .05. Mercurial-induced augmentation of [Cl⁻]_i was accompanied by significantly enhanced epinephrine-mediated platelet aggregation (Fig. 5). The effect of the mercurial on A2AR-mediated responses in platelets was dose-dependent upon the concentration of thimerosal used (data not shown).

The effect of mercury on intraplatelet [Cl⁻]_i. Micromolar concentrations of HgCl₂ significantly increased the accumulation of intraplatelet [Cl⁻]_i. HgCl₂ increased intraplatelet [Cl⁻]_i from (in mmol/L ± SEM) from 97.8 ± 0.36 to 126.2 ± 0.30. Similar results were obtained with incubation of platelets in other mercurials such as thimerosal and phenylmercuric acetate (not shown).

The effect of phenylmercuric acetate (PMA) on α₂ adrenergic receptor (A2AR)-mediated contractile responses. Preincubation of isolated rat tail arteries with PMA resulted in significant increases in vascular contractility as reflected in a significant decrease in the mean EC50 for clonidine in the presence of PMA from 0.13 to 0.04 μmol/L, P < .05. Whereas diminished [Cl⁻]_i attenuated clonidine-induced contractions in isolated blood vessels, an increase in [Cl⁻]_i enhanced A2AR-dependent vascular reactivity. PSS = physiologic salt solution.

The effect of thimerosal in epinephrine-mediated platelet aggregation. Increasing intraplatelet [Cl⁻]_i with thimerosal enhanced α₂ adrenergic receptor-mediated platelet aggregation. Preincubation of platelets in physiologic salt solution with 50 μmol/L thimerosal decreased the EC50 for epinephrine from (in μmol/L) 0.86 to 0.10, n = 6, P < .01.

We next used an in situ assay to measure the effect of mercurials and epinephrine on carbonic anhydrase-dependent anion exchange activity. The swelling that occurs when some hematopoietic cells are placed in an isotonic solution of ammonium chloride is markedly enhanced by the addition of bicarbonate. This phenomenon was first explained by Jacobs and Stewart.¹¹ As shown in a schematic representation in Fig. 6, the relatively lipid soluble moiety NH₃ is generated in aqueous solutions of NH₄Cl and, with increasing concentrations of NH₄Cl, more NH₃ is available to diffuse across the cell membrane, where it serves as a conjugate base and accepts the H⁺ generated by the carbonic anhydrase catalyzed hydration of CO₂. Because the newly formed NH₄⁺ is osmotically active, and H⁺ is not, the formation of NH₄⁺ from H⁺ and NH₃ results in an increase in cell size. The bicarbonate generated from the hydration of CO₂ is free to exchange with extracellular chloride. Inhibition of carbonic anhydrase with acetazolamide results in a decrease in the hydration of CO₂ into H⁺ and bicarbonate; however, because H⁺ is a stronger acid than bicarbonate is a base, intracellular pH should be expected to rise, and it does (Fig. 6, inset). Because of the intracellular alkalinization that occurs in the presence of acetazolamide, less NH₄⁺ is formed from intracellular NH₃, and platelet swelling is diminished (Fig. 7). Furthermore, inhibition of chloride/bicarbonate exchange should decrease the export of bicarbonate and shift the equilibrium for the carbonic anhydrase reaction. The net result of decreased formation of NH₄⁺ and diminished anion exchange is an attenuation of platelet swelling. Accordingly, changes in platelet size can be used to measure carbonic anhydrase-dependent chloride/bicarbonate exchange activity.

Effect of carbonic anhydrase activity on chloride/bicarbonate exchange and [Cl⁻]_i. When platelets are placed in an iso-osmotic solution of NH₄Cl, intraplatelet NH₃ complexes with the H+ produced by carbonic anhydrase and the relatively impermeant, but osmotically active, NH₄⁺ accumulates intracellularly, and platelets swell. The conjugate anion formed by this reaction, HCO₃⁻, is exchanged with extracellular [Cl⁻]_i. As a result, an intracellular chloride concentration, which is higher than predicted from its equilibrium potential can be achieved. The degree of platelet swelling can be used as an indicator of carbonic anhydrase-dependent chloride/bicarbonate exchange. As shown in the inset, acetazolamide should increase pH_i; bicarbonate is a weak base to the conjugate H⁺ acid; inhibition of carbonic anhydrase did indeed alkalinize the platelets.

Left) Platelets placed in an iso-osmotic physiologic salt solution containing 65 mmol/L NH₄Cl and 65 mmol/L NaCl swelled considerably over 10 min. This change in platelet size was significantly attenuated when the platelets were incubated with an inhibitor of the chloride/bicarbonate exchanger, DIDS, or when carbonic anhydrase was blocked with acetazolamide. Right) A quantity of 50 μmol/L HgCl₂ increased NH₄Cl⁻induced platelet swelling (values expressed as percent change from baseline size ± SEM) from 63 ± 4 to 229 ± 24, P < .01). DIDS = 4",4-diisothiocyano-2-2",2-stilbenedisulphonic acid.

Platelets placed in a PSS in which 65 mmol/L of NaCl was replaced with NH₄Cl resulted in nearly a 20% increase in median platelet volume (Fig. 7). This NH₄Cl⁻mediated platelet swelling was significantly inhibited with either inhibition of carbonic anhydrase by acetazolamide or the chloride/bicarbonate exchange inhibitor, 4′,4-diisothiocyano-2-2′,2-stilbenedisulphonic acid (DIDS). Furthermore, increasing the concentration of NH₄Cl in the PSS increased the NH₃ available to complex with H⁺ and, accordingly, resulted in a greater rate of platelet swelling.

We determined whether HgCl₂ could increase [Cl⁻]_i by increasing anion exchange. When platelets were incubated with 50 μmol/L of HgCl₂, the rate of platelet swelling was drastically increased. It is possible that mercurials increase intraplatelet chloride either by increasing the carbonic anhydrase activity–dependent concentration of intracellular bicarbonate available for exchange with extracellular chloride, or by stimulating the chloride/bicarbonate anion exchanger directly (Fig. 7).

Subsequently, we used this in situ assay to determine whether A2AR ligands increase platelet [Cl⁻]_i by increasing carbonic anhydrase-dependent chloride/bicarbonate exchanger activity. When incubated with 10 μmol/L epinephrine, the rate of platelet swelling induced by incubation in NH₄Cl significantly increased. In these experiments, incubation of platelets in PSS in which 130 mmol/L NaCl had been replaced with an equimolar concentration of NH₄Cl resulted in an 18.4 ± 0.5% increase in platelet size. Platelets that were incubated with 10 μmol/L epinephrine in NH₄Cl⁻PSS increased in size by over 25.6 ± 0.4% over the same 5-min period, P < .005 (Fig. 8). Epinephrine clearly augmented carbonic anhydrase-dependent increases in [Cl⁻]_i.

Effect of epinephrine (EPI) on NH₄Cl⁻ induced platelet swelling. Epinephrine increased platelet swelling in response to NH₄Cl by 33% over control values.

The specific binding of radiolabeled thiazides to sodium chloride cotransporters has not been convincingly demonstrated in many tissues, but thiazides inhibit carbonic anhydrase activity.¹⁶ We postulated that the salutary actions of the diuretics result not from their action on renal thiazide-sensitive cotransporters but, rather, through their ubiquitous ability to inhibit carbonic anhydrase, and thus decrease chloride/bicarbonate exchange. Using a standard in vitro assay of carbonic anhydrase, we found that both acetazolamide and hydrochlorothiazide significantly diminished the activity of this enzyme. Carbonic anhydrase activity has been reported to be sensitive to chloride concentration; we, too, found such was the case. When carbonic anhydrase activity was measured in the presence of chloride at concentrations that were close to those values of [Cl⁻]_i we observed in platelets, the sensitivity of carbonic anhydrase to inhibition by thiazides shifted significantly further, ie, micromolar concentrations of thiazide further inhibited the residual carbonic anhydrase activity present with physiologic [Cl⁻]_i (Fig. 9).

Left) The thiazide diuretics have been demonstrated to significantly inhibit the various isozymes of carbonic anhydrase. We believed that this inhibition was not significant, given the high concentration of carbonic anhydrase found in most cells; however, we also found that chloride, at a concentration similar to that which is found in platelets and vascular smooth muscle cells, itself inhibits carbonic anhydrase activity. Middle) Because of the high [Cl⁻]_i of platelets and vascular smooth muscle cells, it is likely that thiazide diuretics do indeed have a significant effect on carbonic anhydrase activity and mediating physiologic responses that are dependent upon modulating chloride/bicarbonate exchange and [Cl⁻]_I. Right) Effect of intracellular chloride and HCTZ together on carbonic anhydrase activity in platelets.

Discussion

We have presented further evidence that A2AR-dependent platelet aggregation and vascular contraction are dependent upon increases in [Cl⁻]_i. The contributory roles of the various anion transport mechanisms in mediating agonist-induced platelet aggregation or vascular reactivity have not been well studied. Our new finding that Hg-induced increases in [Cl⁻]_i potentiate epinephrine-mediated platelet aggregation supports our contention that shifts in chloride are a necessary step to initiate A2AR-dependent thrombus formation. The mechanism(s) by which mercurials increase [Cl⁻]_i above a concentration predicted for its equilibrium potential remains unknown. The Hg-induced rises in [Cl⁻]_i were not inhibitable by acetazolamide (data not shown); it is not likely that these compounds directly stimulate carbonic anhydrase-dependent anion exchange. It is reasonable that the mercurials stimulate the activity of thiazide-sensitive Na/Cl cotransporter or the bumetanide sensitive Na/K/2Cl cotransporter; conversely, mercurials may decrease K/Cl cotransport that is primarily a chloride efflux pathway.^13,,15,17,18 It is unclear that the first two of these functional transport proteins are found in both platelets and vascular smooth muscle.¹⁹ Mercurials bind to active sulfhydryl moieties, and oxidation of sulfhydryl groups has been demonstrated to increase K/Cl cotransport.^14,15 As a result, it could be argued that mercurial-dependent loss of active sulfhydryl groups on transport proteins decreases potassium coupled-chloride efflux. However, in those studies, chloride-dependent rubidium efflux was measured; it is possible that augmented K/Cl efflux is secondary to a net increase in [Cl⁻]_i accumulation. Also, we do not argue that chloride/bicarbonate exchange is the sole mechanism responsible for mediating chloride-dependent platelet aggregation and vascular contractions. The probability that the Na/H exchanger is involved in platelets is also very high, as this antiporter can be stimulated by catecholamines and induce a rise in intracellular pH. Such a rise in pH_i would increase CO₂ diffusion into the cell which would then increase bicarbonate formation and isoelectrically raise [Cl⁻]_i. Despite these unanswered questions, heightened [Cl⁻]_i in platelets is clearly associated with enhanced epinephrine-mediated platelet aggregation.

Many of the cellular transport mechanisms observed in platelets are similarly operative in vascular smooth muscle. However, as it is an excitable tissue, discerning the role of [Cl⁻]_i in vascular responsiveness to A2AR agonists is predictably more complicated than delineating the role of [Cl⁻]_i in A2AR-mediated platelet aggregation. Again, [Cl⁻]_i is accumulated by transport against its electrochemical gradient. When A2AR agonists are added in the presence of Hg to vascular smooth muscle, any increase in chloride conductance would result in a greater chloride efflux, augmented loss of negative charges from the cell, depolarization, and greater contractile responses to clonidine. Indeed, we observed greater contractile responses to clonidine in the presence of Hg. Conversely, when blood vessels were preincubated in either low chloride PSS or with acetazolamide, contractile responses to clonidine were greatly diminished.

Finally, we postulated that epinephrine-induced increases in [Cl⁻]_i result from enhanced carbonic anhydrase-dependent formation of bicarbonate. Accordingly, we sought to demonstrate this phenomenon using an assay of carbonic anhydrase-dependent anion exchange in intact cells. We determined whether epinephrine could increase carbonic anhydrase-dependent cell swelling induced by NH₄Cl. Because A2AR-binding of agonist is reported to be sodium dependent,²⁰ we first determined whether NH₄⁺ could substitute for this cation in a binding assay. In the presence of NH₄Cl or NaCl, we found similar specific binding for tritiated yohimbine, although the affinity of the receptor for this ligand was substantially reduced (data not shown). Despite a possible reduction in epinephrine binding to platelet A2ARs in this assay, we still observed a remarkable acetazolamide-sensitive increase in NH₄Cl⁻dependent cell swelling in the presence of the catecholamine. The A2AR agonists clearly have the potential of activating carbonic anhydrase-dependent anion exchange.

Our observations offer a possible rationale for the salutary effect of thiazides on cardiovascular morbidity.^21–23 We argue that thiazides, by virtue of their known ability to attenuate carbonic anhydrase directly, may decrease A2AR-mediated platelet aggregation and vascular reactivity by decreasing [Cl⁻]_i. Furthermore, this salutary action of the thiazides is not dependent upon the Na/Cl cotransporter that may be absent in platelets and in vascular smooth muscle. It could be argued that the concentration of thiazides required to significantly inhibit carbonic anhydrase is greater than the concentration normally found in patients using these drugs therapeutically. However, this opposition to a potential salutary benefit of the thiazides being mediated by carbonic anhydrase ignores the well established inhibitory effect of chloride on carbonic anhydrase activity that we, too, have found. Epinephrine-stimulated platelets have very high [Cl⁻]_i; this blood element is incapable of new protein synthesis; and these inhibitors bind very tightly to carbonic anhydrase.²⁴ Irreversible inhibition of carbonic anhydrase by low concentrations of thiazide can be expected to result in significant inhibition of platelet aggregation in vivo.

Our findings suggest a mechanism by which thiazides may exert some of their healthful effects; in any case, the mechanisms of thiazide action should be revisited. Our observations are also supported by recent work that shows hydrochlorothiazide can exert direct vasodilator effects by activation of pH-sensitive calcium-activated K⁺(K_ca) channels in human resistance arteries.^25,26 Also, it has been reported that bendroflumethiazide, a thiazide diuretic with minimal inhibitory effects on carbonic anhydrase, had little effect on norepinephrine-induced tone when compared with hydrochlorothiazide. Others have supported our contention that the vasodilatory effect of diuretics is dependent on the inhibition of vascular smooth muscle carbonic anhydrase.²⁷ Finally, it should be remembered that acetazolamide is one of the most dramatic direct-acting vasodilators in some blood vessels such as those in the cerebral vasculature, where thrombotic events are most problematic.²⁸ It has also been reported that the activity of the anion exchanger is increased in a subset of hypertensive patients, and thiazides may specifically benefit this particular subset of patients.²⁹ In any event, our results on the significance of chloride in mediating vascular contractility reinforce the role of this anion, rather than sodium, in determining blood pressure in salt-sensitive individuals.³⁰

Acknowledgment

The authors appreciate the mentoring of Horace W. Davenport, PhD, DSc, Professor and Chairman Emeritus of the Department of Physiology at the University of Michigan, who recognized the importance of carbonic anhydrase on June 18, 1937.

References

Author notes