manuscript, Dr. John Baxter for critical comments of the text, and K. Nakamura, K. Nguyen, and R. Brown for excellent technical assistance. This work was ...
Arachidonic acid metabolites regulate the secretion of atrial natriuretic peptide in cultured rat atrial cardiocytesl BRANKA KOVACIC-MILIVBJEVIC Metabolic Research Unit, University of California at Sun Francisco, Sun Francisco, CA 94143, U.S.A.
HAROLD D. SCHULTZ Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by Depository Services Program on 05/28/13 For personal use only.
University sf California at San Francisco, Sun Francisco, CA 94143, U.S.A.
DAVID G. GARDNER~ Department of Medicine and Metabolic Research Unit, University of California at San Francisco, Sun Francisco, CA 94143, U.S.A. Received August 10, 1990 K O ~ A ~ I ~ - M I L ~B., ~ OSCHULTZ, J B ~ I ~ ,PI. D., and GARDNER, D. G. 1991. Arachidonic acid metabolites regulate the secretion of atrial natriuretic peptide in cultured rat atrial cardiocytes. Can. J. Physiol. Pharmacol. 69: 1493- 1499. Prostaglandins F,, and E, increase release of immunoreactive (irANP) in primary cultures of rat atrial cardiocytes. This effect is independent of cell density in the cultures and does not appear to operate through a CAMP-dependent mechanism. Studies that probed the PCF,, effect with a number of different pharmacological antagonists suggest that it is tied to a calmodulin-dependent step. This latter effect does not appear to be related to increased calcium entry through voltage-gated channels in the plasma membrane nor to mobilization of ryanodine-sensitive intracellular calcium p o l s . Inhibitors of the Iipxygenase pathway, a second avenue of arachidonate metabolism, resulted in a decrease in irANP release from cultured atrial or ventricular cardiscytes. Leukotriene C,, a lipoxygenase product, had a modest effect to promote irANP release over a 24-h p e r i d . However, pretreatment of anesthetized rats with mordihydroguarietic acid, a lipoxygenase inhibitor, had no effect on stretch-dependent release of irANP from the heart in vivo. 'These findings suggest that the prostaglandins represent the more important group of arachidonate metabolites in regulating irANP release physiologically. Key words: ANP secretion, prostaglandins, leukotrienes , cardiovascular homeostasis. KOVACIC-MIL~VOJEV~C, B., SCHULTZ,Pi. D., et GARDNBR, D. G. 1991. Arachidonic acid metabolites regulate the secretion of atrial natriuretic peptide in cultured rat atrial cardiocytes. Can. J. Bhysiol. Pharmacol. 69 : 1493 - 1499. Has prostaglandines I;, et E, augmentent la liberation d'ANP immunor6actif (ANPir) dans les cultures primaires de cardiocytes auriculaires de rat. Cet effet est indkpendant de la densite des cellules cultivCes et ne semble pas lie 2t un micanisme dependant de 1'AMPc. Des etudes, qui ont examink l'effet de 19utilisation de la PGF,, avec diffkrents antagonist= pharmacologiques, suggkrent qu'il est lie % une phase dkpendante de la calmduline. Cet effet ne semble associC ni B l'ouvebture des canaux de calcium dans la membrane plasmatique ni a une mobilisation de pools de calcium intracellulaire sensible % la ryandine. Les inhibiteurs de la voie de la lipoxygCnase, une des deux voies du mCtabolisme de l'arachidonate, ont p r o v q u t une diminution de la liberation d'ANPir des cardiscytes ventriculaires ou auriculaires cultivts. Le leucotrihne C,, urn prduit de la lipxyg4nase, n'a que faiblement influence la libkration d'ANPir sur une pCriode de 24 h. Toutefois, un pretraitement de rats anesthksies avec l'acide nordihydroguariktique, un inhibiteur de la lipoxygknase, n'a pas eu d9effet sur la libkration d'ANPir dependante de l'ktirement cardiaque in vivs. Ces rksultats suggkrent que les prostaglandines representent le plus important groupe de mCtaboIites de 19arachidonate pour la regulation physiologique de la libkmtion d' ANPir. Mots clds : skcrktion d'ANP, prostaglandines, leucotri&nes, homeostase cardiovasculaire. [Traduit par la rCdaction]
Introduction The atrial natriuretic peptide (ANP) is a cardiac hormone that is regulated by a variety of neurotransmitters (Sonnenberg and Veress 1984; Schiebinger et al. 1989; Shields and Glembotski 1989), hormones (Gardner et d.1989; Ladenson et al. 1988; Gardner et al. 1988; Gardner and Schultz 19W), and physical factors that increase myocardial wall tension (Dietz 1984; Eang et d. 1984; Burnett et d. 1986). Recently we reported that prostaglandins (PGs), specifically PGF2, and PG&, stimulate release of immunoreactive atrial natriuretic peptide (irANP) and accumulation of ANP mRNA transcripts in cultured neonatal rat atrial and ventricular cells 'This paper was presented at the Satellite Symposium of the 13th Scientific Meeting of the International Society of Hypertension, called A Decade of ANF Research, held June 21 -23, 1990, Ottawa, Ont., Canada, and has undergone the Journal's usual peer review. 2Author for correspondence. Prtnted In Canda / Imprim6 iau Canada
(Gardner and Schultz 1990). The effect of the PGs on secretion seemed to be reduced rather than accentuated in the presence of isobutylmethylxanthine, a phosphodiesterase inhibitor, was not shared by prostacyclin (PG12), a potent activator of adenylate cyclase, and was inhibitable with high concentrations of forskolin. In the present study we have hrther explored the relationship of PG stimulation to irANP release and have extended our investigation to arachidonate metabolites generated in the parallel lipxygenase pathway.
Materids and methods Materials [y-32P]ATP was purchased from Amersham Corp., Arlington Heights, IL. Prostaglandins, calmidazolium, diethylcarbazine, and nordihydroguarietic acid (NDGA) were obtained from Sigma Chemical Co., St. Louis, MO. Baclein, caffeic acid, ryansdine, TMB 8, and leukotriene C, were purchased from Biomol Research Laboratories, Plymouth Meeting, PA. Qther reagents were obtained from standard comercia1 suppliers and represent the best grades available.
f 494
CAN. J. PHYSIOL. PHAWMACOL. VOL. 69, 1991
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by Depository Services Program on 05/28/13 For personal use only.
Methods
Cultures of neonatal rat cardiocytes, enriched for either atrial or ventricular cells, were prepared as described previously (Gardner et al. 1988). Cells were cultured for 5 days in DME W21 medium containing lO % fetal calf serum (FCS), 2 mM glutamine, 100 U/mL penicillin, and 100 pg/mL streptomycin. At that point cells were changed to serum-free media containing 18% serum substitute (Bauer et al. l 976), glutamine, penicillin, streptomycin, and the additive(s) indicated. Media was collected at 24 k and assayed for irANP as described previously (Gardner et al. 1988). Bw selected experiments a media aliquot was taken at 5 and 30 rnin following addition sf agonist and subjected to similar analysis for irAP.JP. For measurement sf cellular CAMPand (or) cGMP levels, cells were lysed with 6% cold trichloroacetic acid (TCA) after 5 min of incubation with the agonist. Cellular material was extracted for 30 min at 4°C. The cellular debris was then pelleted; the TCA extract was removed and extracted (three times) with two volumes of watersaturated ether. The aqueous phase was then neutralized with 1 ha NaHCO, (5% by volume) and assayed for CAMP or cGMP using commerciaIly available kits (Arnerskam). Protein in the neutralized TCA extract was determined using the BCA protein assay reagent (Pierce Chemicals, Rockford, Ik) . Measurements of irANP, CAMP, and cGMP were subjected to one-way analysis s f variance with a post-hoe Neuman -Keuls test for significance. For the in vivo experiment, female Sprague- Dawley rats (250 350 g) were anesthetized with 2 % chloralose and 25 % urethane (3 mL/ kg i.p.). An intm-arterial catheter (PE-50) was introduced into the left carotid artery for measurement of mean arterial blood pressure (MAP). Catheters were also introduced into the right internal jugular vein for the measurement of central venous pressure ((CVP) and into the lefi femoral vein for administration of volume and pharmacological agents. Catheter patency was maintained with periodic administration of he~arinizedsaline. TOkvaluate the effect of NDGA on stretch-dependent ANP release, the following protocol was employed. A blood sample (1 mL) was collected in 10 mM ethylenediaminetetraacetic acid (EmA) at time zero and placed on ice. Aker H rnin an infusion of 5 rnL of saline was delivered into the femoral vein over a 60-s period. Additional blood samples were taken 1 and 4 min later. After a 30- to 45-min re-equilibration period, a fourth blood sample was taken. One minute later an 8-mL saline infusion was delivered over 40 s. Blood samples were taken H and 4 min later. For the NNA-treated group the drug (1 mg/ kg) was administered 30 min prior to the 5-mL infusion and again 30 min prior to the 8-mL infusion. Blood samples were centrifuged and plasma was decanted. Plasma (0.5 d)was extracted over a Sep-Pak C,, column, as described by Lang et al. (1985), and dried under a nitrogen stream. The dessicated material was resuspended in 250 pL of buffer and duplicate 100-yL aliquots were assayed for irAMP as described previously (Gardner and Sckultz 1990).
As noted previously (Gardner and Schultz 1990), prostaglandins E2 and F2a routinely foster a significant increment (approx. two- to three-fold) in irANP release from neonatd rat cardiocytes. As a function of time in culture, these cardiocytes organize into clusters and often beat in synchrony. Since these are phenomena that might be promoted by increased cell density, we asked whether the PG effects were dependent on the number of myocardial cells present in our cultures. As shown in Table I, the answer appears to be negative. Although cell densities varied over a fivefold range, the PG stimulation was preserved in all but the lowest density PGF2,-treated cultures. In each case the relative magnitude of the response remained at the same level (approx. two- to three-fold increment vs. the control) suggesting that cell density per se neither amplifies
TABLE1. Effect of cell density on irANP release from cultured atrial cardiocytes Cell density (cells/well) -
10 x 103
2% x lo3
lrANP Treatment (ng * mL-' media . 24 h-' ) -
-
Control PGF2, PGE, Control PGF2, PGE,
50 x 103
Control PGF2, PGE2
NOTE:Cells were cultured at densities o f 10, 25, or 50 x 10"well in a 96-well dish. After 5 days, serum-containing media were replaced with DME H21 containin serum substitute in the presence or absence of the prostaglandin (loPf M) indicated. After 24 h media were collected for assay of irANP; values represent the means & SD; n = 4 for each group. *p < 0.05.
nor abrogates the response to the eicssanoids. As mentioned above, earlier evidence suggested that PG-mediated increments in irANP release were not dependent on the ability of these agents to foster accumulation of intracelltalar CAMP. To approach this question more rigorously we M PGF2, in the presence o r absence treated cells with s f isobutylmethylxanghine (3 x 1W4 M, IBMX) for 5 min, collected media for measurement of irANP, and made cytosolic extracts for assessment of intracellular cAMP and cGMP levels. As shown in Fig. 1 , PGF2, effected a significant increment in irANP release. This was accompanied by a reduction in cellular CAMP levels. Inclusion of IBMX done in the culture medium resulted in no change in irANP release or cellular cAMP levels. When administered together with BGF2,, however, IBMX promoted an increase in cAMP levels while, at the same time, reducing BGF2,-stimulated irANB release. Taken together these data suggest a dissociation between CAMP accumulation and the secretory event. PGF2, had no effect on cGMP accumulation in the presence o r absence of IBMX. IBMX alone increased cGMP levels approximately twofold. We have shown previously that PGF2, stimulation of irANP release is independent of calcium ion concentration in the culture medium (Gardner and Schultz 1980)). To further explore the possible role of calcium ion in mediating the PGF2, stimulation of ANP release, we examined the effect of several pharmacological agents known to be capable of perturbing intracellular calcium homeostasis in myocardial cells. As shown in Table 2, calcium channel blockade with veraparnil had little effect on EF2,-stimulated irANP release. Wyanodine, an agent that is thought to reduce available pools of releasable calcium in the sarcoplasmic reticulum (Fabiats B985), was without effect. TMB-8, an agent that similarly reduces intracellular calcium release (Chiou and Malagodi 19751, appeared to diminish the relative response to PGF2,, largely by virtue of its tendency to increase basal secretion of irANB. Of interest, however, treatment with the calmodulin antagonist calmidazolium reduced basal secretion of irANB by approximately 55 % and totally eliminated the PGF2, stimulation of peptide release, implying that ANP release involves a calmodulin-dependent step. Additional experiments dernonstrated that the inhibition of irANP release could be demon-
'IC-MILIVOJEVIC BT AL.
43
Can. J. Physiol. Pharmacol. Downloaded from www.nrcresearchpress.com by Depository Services Program on 05/28/13 For personal use only.
p
