Effect of anti-calmodulin agents in vasopressin release in vitro to depolarization and calcium ionophore

Effect of anti-calmodulin agents in vasopressin release in vitro to depolarization and calcium ionophore

Life Sciences, Vol. 46, pp. Printed in the U.S.A. Pergamon Press 1091-1098 EFFECT OF ANTI-CALMOOULIN AGENTS ON VASOPRESSIN RELEASE IN VITRO TO DEPO...

746KB Sizes 0 Downloads 32 Views

Life Sciences, Vol. 46, pp. Printed in the U.S.A.

Pergamon Press

1091-1098

EFFECT OF ANTI-CALMOOULIN AGENTS ON VASOPRESSIN RELEASE IN VITRO TO DEPOLARIZATION AND CALCIUM IONOPHORE Noreen

F.

Rossi

Departments of Internal Medicine and Physiology Wayne State University School of Medicine 540 East Canfield Avenue Detroit, Michigan 48201

(Received in final form February 7, 1990)

Summary 2+ Calmodulin has been implicated in transducing the effects of Ca on synaptic transmission and hormone release, including osmoticallystimulated vasopressin (AVP) release. If the an$i-calmodulin agents block AVP release secondary to inhibition of Ca -calmodulin interthese drugs should inhibit AVP release to stimuli increasaction f+ influx via different mechanisms. Hypothalamo-neurohypophying Ca sial complexes (HNc) were exposed to ionomycin, Bay K 8644, or veratridine either alone, with any one of thr?? distinct chemical classes of anti-calmodulin agent, or with a Ca channel antagonist. All the f?ti-calmodulin agents impaired AVP release to 2/onomycin, while Ca channel blockade did not. Converse1 y, Ca channe I antagonism completely blocked AVP release in response to Bay K 8644, but the anti-calmodulin agents had no effect. None of the inhibitors prevented veratridine-induced AVP release. These results are consistent with the hypothesis that the anti-calmodulin agents tested inhibit AVP release by2$heir membrane stabilizing properties rather than by antagonizing Ca -calmodulin $9 HNC. Depolarization initiated by Na influx may2gtimulate Na -Ca exchange by a mechanism independent of slow Ca channels as well. Ca

2+

ions play a csycial role in the secretion of argipine vasopressin (AVP). Extracellular Ca ions are are necessary for both K -depolarizationinduced and electrically-stimulate#5ho~Tone release from the neurohypophysis Demonstration of uptake of Ca (1,2). into neural lypes after electrical stimulation lends further support for involvement of Ca influx in initiating the intracellular mechanisms necessary for AVP release (3). Calmodulin, which binds Ca 2+ with high affinity (41, has been implicatF$ as the cytop I asmi c “receptor” responsible for transducing the effects of Ca on synaptic transmission (5) and hormone release (6). A substantial concentration of calmodulin has been identified in both mammalian hypothalamus (7) and posterior pituitary (8). Morepyer,2+caImoduIin stimulates the activity of several brain enzymes such as Ca -Mg protein kinases (6), and ATPase (9), phosphodiesterase (4) as well as neurohypophysial adeny I ate cyclase (10) which may be involved in signal transduction or exocytotic release of AVP. Recently, we reported that anti-calmodulin agents of three distinct chemical classes inhibit osmotically-mediated AVP release; however, these compounds did

Copyright

0024-3205190 (c) 1990

$3.00 +.oo Pergamon Press

plc

Calmodulin and AVP Release

1092

Vol. 46, No. 15, 1990

not have uniform effects on AVP secretion induced by angiotensin II (11). It is now known that the pharmacologic probes used to test calmodulin involvement in cellular processes exert actions other than calmodulin antagonism (12-15). The present experiments were designed to test the hypothesis that, $1 the anticalmodulin agents block AVP release secondary to inhibition of Ca -calmodulin in@-actions, then these drugs should inhibit AVP release to stimuli increasing Ca influx via different mechanisms.

Methods The HNC explants are dissected and cultured Pupaution ob f/NC up.&ant~. as previously described (16). Briefly, unanesthetized male Sprague-Dawley rats weighing 150-200 g were decapitated by guillotine. The cranium was removed, and the neural and vascular attachments to the base of the skull were cut using a caudal approach. The dura mater attaching to the sella turcica was severed to allow the removal of the pituitary with its attachments to the stalk intact. Under a dissecting microscope, the adenohypophysis was removed. The dura mater and the arachnoid were stripped from the ventral surface of the hypothalamus. The explant was obtained by cutting 1 mm rostra1 to the optic chiasm and 2-3 mm lateral to either side of the median eminence. The lateral cuts tapered back to a point 0.5 mm caudal to the infundibular stalk. The explant was undercut with a razor blade at a depth of 1 mm. This block of tissue contains the supraoptic nuclei lobe. with intact axonal projections to the neural The arcuate, suprachiasmatic. preoptic, ventromedial nuclei, and the organum vasculosum of the lamina terminalis are present within the explant as is the intermediate lobe, which is adherent to the neural lobe. The paraventricular nuclei and the subfornical organ are absent. lncubtion condtibnh. Individual HNCs were placed into separate incubation wells (Falcon. Oxnard, CA) and were supported ventral side down by Teflon CA) on stainless steel grids. Each incubation mesh (Spectrum, Los Angeles, well contained 0.75 ml of Ham’s F-12 nutrient mixture (Grand Island Biological, NY) fortified with 20% fetal bovine serum (Hyclone Laboratories, Grand Island, Logan, UT), 1.0 mg/ml glucose, 100 U/ml penicillin, and lOO,ug/mI streptomycin. The HNCs were The medium had a final osmolality of 298+3 mosmol/kg water. maintained in a humidified atmosphere of 9% 0,-r% CO at 37oC. The medium was changed at 24 h intervals or after the completion of 3 he experimental protocol After all experimental protocols were completed on a given set for that day. they were fixed in neutral buffered Formalin for microscopic of explants, examination to detect any tissue damage. Expehiment& p~~otocot4. In general, the experimental protocols followed a previously specified technique (16). All experiments were performed at 48 h after dissection. Earlier work has demonstrated that basal release rates and the responsiveness of cultured HNCs to stimulation are stable for up to 5 days At the beginning of a protocol, bacitracin (Sigma, St. Louis, MC) was (16). added to each well, giving a final concentration of 0.5 mg/ml and preventing Basal rate of AVP release was initially degradation of the AVP released (17). followed by determination of AVP release in assessed during a control hour, Samples for determinaresponse to the test substance in the subsequent hour. Al I tion of degradation were taken to verify the effectiveness of bacitracin. were immediately frozen and stored at -7O’C until radioimmunoassay was samples The remaining medium was removed for determination of osmolality, performed.2+ ionized Ca , and lactate dehydrogenase (LDH) activity. Three trifluoperazine Philadelphia,

classes

of anti-calmodulin agents were tested: a phenothiazine, 1 JIM gift of Smith Kline Beckman and French Pharmaceuticals, (TFP; 1 JJM W 5 or W 7 (Seikagaku the naphthalenesulfonamides, PA);

1093

Calmodulin and AVP Release

Vol. 46, No. 15, 1990

Kogyo, Tokyo, Japan); and a derivative of an2imidazolium compound, 2pM R 24571 (Research Biochemical, Natick, MA). The Ca channel antagonist, methoxyverapamil (D 600; Knoll Pharmaceutical, Ludwigshafen, Germany) was used at 0.5 PM. In separate explants, the effect of each of the anti-calmodulin agents or D 60$+ was examined during the test period either alone or in the presence of th? Ca channel agonist, 1 uM Bay K 8644 (Miles, West Haven, CT); or the Nf+ channel activator, 75 JJM veratridine (Aldrich, Milwaukee, WI); or the Ca ionophore, 1 )JM ionomycin (Calbiochem, La Jolla, CA). Similarly, the effects of identical concentrations of Bay K 8644, veratridine, and ionomycin alone were examined. AVP mcadwrementd. The AVP released into the medium was measured by radioimmunoassay (18). The medium was diluted 100 times with assay buffer that contained 0.15 M sodium phosphate, 0.01 M sodium EDTA, 0.1 g/100 ml sodium azide, and 0.1 g/100 ml bovine serum albumin (ICN-Miles, Irvine, CA) at pH 7.4. Radioimmunoassay was performed in duplicate directly on the diluted samples. Standard curves were prepared with purified AVP with a designated activitr250f 400 U/mg (Ferring, Malmo, Sweden); standards ranged from 0.5-50 pg/tube. IAVP was synthesized via chloramine-T iodination of AVP (19). The anti-AVP antiserum 2849 (gift of Drs. M.D. Lindheimer and J. Durr, Universities of Chjcago and Colorado, respectively) was used at a final dilution of 1:3.6 x 10 . Separation of bound from free AVP was accomplished by precipitation of the bound fraction with 2% bovine gamma globulin (Sigma, St. Louis, MD) and 25% polyethylene glycol (Fisher Scientific, Fair Lawn, NJ). and 50% displacement was 4.1 pg/tube. Assay sensitivity was 0.1 pg/tube, variabi I ity at Intra-assay variability in the middle and high range was 9%; Interassay variability of seven separate less than 0.1 pg/tube was 13.4%. All samples from a given protocol were assayed simultaneously. assays was 15%. Cross-reactivities with related neuropeptides have been previously published Additives and test agents at lOO-fold dilutions did not affect the (11). Incubation medium from a culture well not parallelism of the standard curve. the AVP measured reflected the containing an explant was diluted 10-fold; the AVP concentration of content of the fetal bovine serum. If detectable, The AVP concentrathis medium was then subtracted from that in each sample. tion of the incubation medium was usually below the limits of sensitivity of the radioimmunoassay and never exceeded 1.0 pg/ml. [email protected] method4 . Osmolality was determined by freezing point depression (Precision Systems 5004, Sudbury,2+MA) on medium from control and test ion-selective was measured with an periods for each explant. Ionized Ca LDH activity was determined using a electrode (Nova Biochemical, Newton, MA). standard spectrophotometric procedure (20). SXatioti& anaIydti. Control and test hours were used in all experiments as specified above. Release of AVP during the test hour was normalized as the Each explant acted percent of AVP released during the preceding control hour. as its own control. The data are expressed as the mean + standard error of the Differences between control mean, with “nw taken as the number of explants. Analysis and test hour release of AVP were compared using the paired t- test. of variance with Scheffe’s modification was used when multiple comparisons were made among test hour AVP release rates. A P-value less than 0.05 was considered statistically significant.

Resu I ts

release

Explants exposed compared with

to 1pM the control

ionomycin showed an 593+97% 100~22% (P < O.bol, hour

increase n = 14).

in TFP,

AVP W 7,

1094

Calmodulin and AVP Release

Vol. 46. No. 15, 1990

2+ and R 24571 significantly decreased the response to ionomycin, while the Ca channel antagonist D 600 failed to inhibit ionophore-stimulated AVP relea?? (Fig. 1). Medium osmolality was 30521 mosm/kg water (n=41). Ionized Ca concentration was 0.831+0.004 mM fn = 11) with ionomycin alone and did not significantly vary among The groups (P > 0.05). Likewise, the LDH activity did not significantly differ (Table 1).

TABLE Medium LDH Activity in ____-_-____--_____--___--__

Calcium

Channel

1

Activator-

and

lonophore-stimulated ---------__-_-

Bay

K 8644

n

LDH mU/mg wet wt/h

lonomycin

n

Bay Bay Bay Bay Bay

K K K K K

8 8 7 8 6

0.177 0.186 0.193 0.209 0.195

I I I I I

11 7 6 8 6

+ + + +

TFP W 7 R 24571 D 600

+ 5 5 -; E

__________--____----____ Values are comparisons.

0.006 0.013 0.006 0.007 0.004

+ TFP +w7 + R 24571 + D 600

HNCs

LDH mU/mg wet wt/h

0.180 0.154 0.143 0.194 0.194

+ T T 5 z

0.011 0.020 0.007 0.008 0.013

-___I__________-_---~~~~~-------______

mean + SE;

I,

by

ionomycin;

K,

Bay

K 8644.

P > 0.05

for

al I

lonomycin k

800

.

Y .

800’

a b

800’

8

1

l

700’

400

.

(

‘0 I*

500’

l

*

. . . .

t 0 L

100’

TFP

w7

FIG.

Ml671

Dsoo

1

Effect of anti-calmodulin agents on AVP release in HNC exposed to ioq?mycin; to ionomycin channel or to ionomycin and a Ca and a calmodul in “inhibitor”; blocker. For each treatment, AVP release is given as the percent of AVP Control hour AVP release released in the preceding control hour (not shown). n = 41. Data are mean + SE; n = 14, 7, 6, 8, was 130 + 37 pg/HNC/h overall, ** P < 0.05 and 6 for each group, respectively. * P < 0.05 versus-control; versus ionomycin alone.

Vol.

46, No.

Calmodulin

15. 1990

and AVP

Release

1095

~JJM Bay K 8644 stimulated AVP release to 499274% of control In contrast, but nyqe of the anti-calmodulin agents significantly (P < 0.005, n = 8); channel blockade completely prevented the reinhibited this response. Ca Medium osmolality was 30421 mosm/kg water lease of AVP to Bay K 8944 (Fig. 2). concentrations and LDH activities were not different Ionized Ca (n = 37). among the groups (Table 1).

Bay L

700

3

600

-

600

K

8644

. E f

400

6

300

m

:

200

4 f

100

h 0

*

0 TFP

WI

R24671

FIG.

Bay

K

6644

Iay

K

6644

l hwltw

WOO

2

Effect of anti-calmodulin agents on AVP release in HNC exposed to B?u K 8644; channel Bay K 8644 and calmodulin "inhibitor"; or to Bay K 8644 and a Ca AVP release is given as the percent of AVP For each treatment, blocker. Control hour AVP release in the preceding control hour (not shown). released Data are mean + SE; n = 8, 8, 7, 8, and was 107 t 13 pg/HNC/h overall, n = 37. 6 for each group, respectively. * P< 0.05 versus control; ** P< 0.05 versus Bay K 8644 alone.

TABLE Effect of Anti-Calmodulin ____------_--______----

Agents

2

on Veratridine-induced AVP Release ___--_-----_----__--

AVP X of control/HNC/h

n

LDH mU/mg wet wt/h

in HNCs

Ionized Ca mM

2+

Test Basal _-------~_--_-------__________I_____-_--~--------~~~~---~~~~ V V + TFP v+w7 V + R 24571 V + D 600

Values versus

14 6 6 6 14

100 100 100 100 100

t -; -, + +

26 45 38 29 28

419 + 59+ 313 7 82 + ++ 17437461 ' 515 7 189* 1146 i 268+

are mean + SE; V, veratridine; control; -** P < 0.02 versus V.

0.165 0.146 0.225 0.212 0.185

* P < 0.05

+ 7 5 + 7 -

0.014 0.009 0.011 0.004 0.011

versus

control;

0.93 0.93 0.84 0.84 0.90

+ 5 7 5 E

0.01 0.01 0.01 0.01 0.01

+ P < 0.001

1096

Calmodulin and AVP Release

Vol. 46. No. 15, 1990

Exposure to 75 #4 veratridine produced a comparable rise in AVP release (fable 2). While none of the anti-calmodulin agents inhibited AVP release in response to veratridine, the presence of W 7 resulted in a significantly higher release of AVP compared with the response to veratridlne alone. D 600 also did not impair hormone release to Na* channel activation, but did not result in a statistically significant increase in hormone release compared with veratridine alone (P = 0.17, n = 14). None of the inhibitory agents control release rates (Figure 3). antagonist of calmodulin, produced

2

alone resulted in AVP release different from W 5, a less potent naphthalenesulfonamide results similar to W 7 in all cases.

200

i b $ O 6

100

0

uunhd

0

inNbitor

z

1 % *

c TFP

W7

R24671

FIG.

0600

3

2+ Effect of anti-calmodulin agents and a Ca channel antagonist on AVP release For each in HNC exposed to the inhibitory agents alone during the test hour. treatment AVP is given as the percent of AVP released in the preceding hour as shown. Data are mean 2 SE; n = 7, 6, 6, and 17. P > 0.05 ~$~~! comparisons.

Discussion AVP secretion can be stjOulated in vitro by depolarization of the neurohy_Our observations in HNC are in pophysis (1,2,21,22) and Ca ionophore (23). the present results extend previous agreement with these reports. Moreoves$ findings channel activator, by demonstrating thff the Ca Bay K 8644 (241, and the mOre ion-selective Ca ionophore, ioncmycin (251, also stimulate AVP release. While Nat-dependent action2,potentials are not an absolute requirement (26), a rise in intracellular Ca (27). is an essential step for AVP release Acsgrding to the calmodulin hypothesis of neurotransmission, intracellular free Ca intracellular binds to calmodulin and this complex initiates several Wakerly (28) has mechanisms ultimately leading to secretory activity (5). Consistent proposed that calmodulin may be involved in the secretion of AVP. with this hypothesis, osmotically-stimulated AVP release is blocked by three different classes of anti-calmodulin agents (11).

Vol.

46, No. 15, 1990

Calmodulin and AVP Release

1097

The anti-calmodulin drugs, however, are known to exert actions independent If TFP, W 7, and R 24571 were all acting of calmodulin antagonism (12-121. -caImoduIin interactions in the magnocellular to antagonize intracellular Ca then these compounds should have inhibltg$ neurons responsible for AVP release, AVP release both to depolarization with veratridine and Bay K 8644 a@ to Ca channel D 600 completely prevented AVP release to Ca via ionophore. entry but none of the anti-calmodulin compounds blocked hormone release activation, to Bay K 8644. These data might suge$st that calmodulin was not involved in AVP were involved In AVP release Alternatively, if Ca -caImoduIin and R 2457$+were not effective in influx occurred, then TFP, W 7, inhibitors channe I blocking this interaction nor were they acting as Ca (13,23,29). none of the anti-calmodulin agents prevepted veratridineSimilarly, Unlike AVP release in response to Na channel activainduced AVP secretion. D 600 failed to antagonize AVP in acutely dissected neural lobes (211, tion Some disparity in observations may have resulted secretion by cultured HNC. I obes from differences in the preparations (acute versus cultured; neural the concentrations of D 600 differed lDDC-fold. Additionally, versus HNC). The submicromolar dose of D 600 in t5$ present experiments was sufficient to inhibit AVP release stimulated by Ca influx through potential operated channels activated by Bay K 8644. The ability of D 600 to block veratridine-induced AVP rejgase from neurohypophysis may have resulted from a nonspecific action of the Ca 2+channel antagonist at the millimolar concentration used (21). AlthougQ channels are often involved in+neural depolarization initiated by Na slow Ca an increase in intracellular Na ion concentration can also2+stimulate influx 2+ channels Na -Ca exchange and cell depolarization independent of the slow Ca If such a mechanism were involved in veratridine-induced AVP release, (30). in accord with the current findings, then, D 600 would not effectively prevent this response. While the apparent augmentation of AVP release with W 7 is less easy to interpret, the failure of any of the anti-calmodulin compounds to block AVF+release stimulated by veratridine , as with Bay K 8644, suggests that either Ca -calmodulin is not involved or these agents are not specifically blocking Ca2+-caImoduIin interactions in HNC. Ionomycin-induced AVP release is not voltage deefndent, and it is known that the ionophore complexes with and transports Ca ions across the lipophilic plasma membrane (31). Predictably, D 600 did not block effect this effect; however, all the other agents tested did. The membrane stabilizing properties of some of these drugs correlate closely with their ability to inhibit calmodulin (32). The drugs can impair ion fluxes across the plasma membrane through either direct membrane stabilization by virtue of their hydrophobicity or secondary membrane effects by inhibition of calmodulin-stimulated phospholipases (12,15,32). Blockade of AVP secretion in response to ionophore but not to depolarization may have resulted from just such membrane stabilization. Alternatively, the ionophore renders the intracellular compartment more accessible to the inhibitors despite the lack of a sufficient effect on the integrity of the plasma membrane to increase LDH activity in the culture medium. Either $/rect, independent inhibition of calmodulin-dependent enzymes or blockade of Ca -calmodulin interactions could thereby be facilitated. In summary, exposure to depolarizing stimuli and iqnophore led to Ay+P secretion by HNC, presumably secondary to increased Ca influx. The Ca channel antagonist successfully inhibited AV+p release in response to the channel activptor, while release initiated by Na influx appeared to be independent of the Ca channel. Finally, the anti-calmodulin agents failed to block AVP release to all the stimuli except the ionophore, consistent with the hypothesis that 2fhese drugs are acting to impair secretion by some mechanism independent of Ca -calmodulin antagonism, such as direct membrane stabilization.

1098

Calmodulin and AVP Release

Vol. 46,

No.

15,

1990

Acknowledgements

I thank Professor Dr. Oberdorf of Knoll Pharmaceutical for the generous supply of methoxyverapamil, Dr. Alexander Scriabine of Miles Laboratories for the gift of Bay K 8644, and Drs. Murray Lindheimer and Jacques Durr for the superb AVP antibody used in these experiments. This research was supported by a grant from the National Kidney Foundation of Michigan. References 1. 2. 3. 4. 5. 6. 7. 8. 9. IO. 11. 12. 13. 14.

15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32.

w. w.

DOUGLAS and A. M. POISNER, J. Physiol. Lond. 172:1-18 (1964). J. T. RUSSELL and N. A. THORN, Acta Endocrinol. 76:m-470 (1974). .I. J. NORDMANN, G. DAYANITHI, and M. CAZALIS, IT Vasopressin, edited by R. W. Schrier, pp. 375-383, Raven, New York (1985). w. Y. CHEUNG, Biochem. Biophys. Res. Commun. 38:533-538 (1970). R. J. DeLORENZO, Cell Calcium 2:365-385 (1981r B. L. BROWN, S. W. WALKER, and S. TOMLINSON, Clin. Endocrinol. -23:201-218 (1985,. \ J. C. EGRIE, J.A. CAMPBELL, A. L. FLANGAS, and F. L. SIEGEL, J. Neurochem. 28:1207-1213 (1977). r M. SHEAVES, D. BOWMAN, and D. B. HOPE, J. Neurochem. 2:909-914 (1982). SOBUI H. ICHIDA, H. YOSHIDA, R. ZAMAZAKI, and L. KAKIUCHI, FEBS Lett. :93199-20; (1979). r A. DARTT, C. TORP-PEDERSEN, and N. THORN, Brain Res. 204:121-128 (1981) N. F. ROSSI and R. W. SCHRIER, Am. J. Physiol. 256:E516-m3 (1989). P. SEEMAN, Pharmacol. Rev. 24:583-655 (1972). R. J. GOULD, K. M. M. MURKY, I. J. REYNOLDS, and S. H. SNYDER, Proc. Natl. Acad. Sci. USA 80:5122-5125 (1983). R. J. BALDESSARINI, InThe Pharmacological of Therapeutics (7th ed.) --Basis edited by A. G. GoodmanTd L. S. Gilman, pp. 387-445, Macmillan, New York (1985). J. STOCLET, D. GERARD, M. KILHOFFER, C. LUGN IER, R. MILLER, and P. SCHAFFER, Prog. Neurobiol. 29:321-364 (1987). C. D. SLADEK and K. M. KNIGE, Endocrinol. 101:411-420 (1977). C. M. GREGG, Neuroendocrinol. 44:378-383 (1986). J. M. DAVISON, E. A. GILMORE, r DURR, G. ROBERTSON, and M. D. LINDHEIMER, Am. J. Physiol. 246:F105-F109 (1984). R. J. ANDERSON, P.G. PLUSS, A. S. BERNS, J. T. JACKSON, P. E. ARNOLD, R. W. SCHRIER, and K. M. McDONALD, J. Clin. Invest. 62:769-777 (1978). H. U. BERGMEYER, In Methods of Enzymatic Analyss, vol. 2, pp. 574-578, Academic, New York (1974). J. J. NORDMANN and R. E.-J. DYBALL, J. Gen. Physiol. 72:297-304 (1978). C. A BONDY, H. GAINER, and J. T. RUSSELL, Neuroendocrinol. -46:258-267 (1987). S. ISHIKAWA and R. W. SCHRIER, Am. J. Physiol. 244:R703-R708 (1983). M. T. SCHRAMM, G. TOWARD, and R. FRANCKOWIAK, Nature 303:535-537, 1983. C. LIU and T. E. HERMANN, J. Biol. Chem. 253:5892-589rl978. J. J. NORDMANN and J. J. DREIFUSS, Brain E. 45:604-607 (1972). J. J. NORDMANN, Prog. Brain Res. 60:281-304 (1?%3). J. B. WAKERLY In Vasopressin: Principles and Properties, edited by D. M. Gash and G. J. Boer, pp. 211-256, Plenum, NewYork (1987). R. A. JANIS and D. J. TRIGGLE, Drug Develop. Res. 4:257-274 (1984). S. LANG and M. P. BLAUSTEIN, Circ. Res. 46:463-470 (1980). M. H. CARVALHO, J . C. PRAT, A. G. GRCIA, and S. M. KIRPEKAR, Am. J. Physiol. 242:El37-El45 (1982). Y. LANDRYX. AMELLAL, and M. RUCKSTUHL, Biochem. Pharmacol. -30:2031-2032 (1981).