Relationship between decreased function and O2 consumption caused by cyclic GMP in cardiac myocytes and L-type calcium channels

Lin Yan 1, Gary X. Gong 1, James Tse 2, Peter M. Scholz 3, Harvey R. Weiss 1

1 Heart and Brain Circulation Laboratory, Department of Physiology and Biophysics, Uni-versity of Medicine and Dentistry of New Jersey, Robert Wood Johnson Medical School,Piscataway, NJ 08854-5635, USA2 Department of Anesthesia, University of Medicine and Dentistry of New Jersey, RobertWood Johnson Medical School, Piscataway, USA3 Department of Surgery, University of Medicine and Dentistry of New Jersey, Robert WoodJohnson Medical School, Piscataway, USA

Received: 24 April 1998 / Accepted: 15 July 1998

Res Exp Med (1998) 198: 109–121

Abstract. We tested the hypothesis that part of the decreased function and me-tabolism caused by cyclic guanosine monophosphate (GMP) in beating car-diac myocytes is related to inhibition of L-type calcium channels. The steadystate oxygen consumption (VO2) of a suspension of ventricular myocytes iso-lated from hearts of New Zealand white rabbits was measured using oxygenelectrodes. Cellular cyclic GMP levels were determined by radioimmunoas-say. Cell shortening was measured with a video edge detector. The VO2 wasobtained after: (1) adding sodium nitroprusside (NP 10–8,–6,–4 M), (2) pretreat-ment by BAY K8644 10–5 M (BAY, L-type calcium channel activator), nifed-ipine 10–4 M (NF, L-type calcium channel blocker) or forskolin 10–7 M (FK,adenylate cyclase activator), then adding NP10–8,–6,–4 M, (3) pretreatment withboth FK 10–7 M and NF 10–4 M and subsequently adding NP 10–8,–6,–4 M. NP10–4 M decreased VO2 from 707±34 to 410±13 (nl O2/min per 105 myocy-tes), decreased the percentage of shortening (Pcs) from 5.7±0.6 to 3.7±0.5and the rate of shortening (Rs) from 65.5±4.5 (µm/s) to 46.2±5.5. NP 10–4 Malso increased cyclic GMP from 264±70 (fmol/105 myocytes) to 760±283.Both BAY and FK increased VO2, Pcs and Rs without changing cyclic GMP.NF decreased Pcs, Rs and VO2. Similar metabolic and functional effects ofNP were observed with pretreatment with these agents separately, comparedto NP alone, and the elevation of cyclic GMP level was not different from thecontrol group. With FK alone, NP 10–4 M decreased VO2 by 51%, Pcs by 44%and Rs by 39%. In the presence of both FK and NF, the negative effects of NPwere diminished significantly. NP 10–4 M decreased VO2 by 37%, Pcs by 25%and Rs 20%. Thus, in beating cardiac myocytes, the negative metabolic and

Springer-Verlag 1996© Springer-Verlag 1997

Correspondence to: H. R. Weiss, e-mail: hweiss@umdnj.edu, Tel.: +1-732-235-4552, Fax: +1-732-235-5038

© Springer-Verlag 1998

functional effects of cyclic GMP were related to inhibition on L-type calciumchannels only when adenylate cyclase was stimulated.

Key words: Myocyte contraction – Myocyte oxygen consumption – Cyclicguanosine monophosphate – Nitric oxide – Calcium channels

Introduction

Cyclic guanosine monophosphate (GMP) can reduce inotropy, shortening,force development and metabolism in the heart [12, 19, 22]. Cyclic GMP alsoantagonizes the positive metabolic and inotropic effects of cyclic AMP in car-diac myocytes [2, 12, 14, 28]. From previous work done in our laboratory, ithas been shown that in rabbit hearts and cardiac myocytes increases in thelevel of cyclic GMP decreased oxygen consumption both in vivo and in vitro[6, 28]. The negative functional and metabolic effects of cyclic GMP havebeen reported to act through three mechanisms: (1) inhibition of L-type cal-cium channel current, (2) activation of cyclic GMP-dependent protein kinase,(3) activation of cyclic GMP-stimulated or inhibited cyclic adenosine mono-phosphate (AMP) phosphodiesterases [12, 13, 15, 18, 19, 21, 22, 25]. Thereis evidence to indicate that cyclic GMP reduces calcium influx [16], shortensaction potential duration [26] and inhibits calcium currents [1].

It has been suggested that the major factor controlling the action of cyclicGMP in cardiac myocytes is its effects on calcium current, primarily throughL-type calcium channels [10, 13–15, 18, 22, 25]. In cardiac myocytes, mainte-nance of ionic gradients is an important energy requirement of the cell. In ratcardiac myocytes, there is a correlation between increase in cytosolic free cal-cium ions and stimulation of oxygen uptake [17]. L-type calcium channels area major entry for free calcium ions into cells during the action potential. Thereare two kinds of calcium channels on ventricular myocytes, T-type and L-type.In guinea pig ventricular myocytes, the density of T-type calcium channels isroughly 10–20 times smaller than the density of L-type calcium channels [3].Much of the intracellular calcium during an action potential comes from thesarcoplasmic reticulum [20]. The link between the effects of cyclic GMP onfunction and metabolism and L-type calcium channels has been questioned. Inthe rabbit heart, no link was found between the negative effect of cyclic GMPon oxygen consumption and L-type calcium channel activity [11]. In addition,we reported that under basal quiescent conditions in myocytes, the negative ef-fects of cyclic GMP were not related to L-type calcium channel activity [30].In this study, we tried to elucidate whether the inhibitory effects of cyclic GMPon L-type calcium channels were the cause of negative metabolic and func-tional effects of cyclic GMP in contracting myocytes and whether increases inadenylate cyclase activity would enhance this relationship.

We tested the hypothesis that the negative functional and metabolic effectsof cyclic GMP on beating cardiac myocytes act through inhibition on L-typecalcium channels. We also determined whether stimulation of adenylate cy-clase would alter this relationship. This hypothesis was tested on beating car-diac myocytes isolated from New Zealand white rabbits. We used sodium ni-troprusside to increase the level of cyclic GMP, nifedipine and BAY K8644 to

110

change the activity of L-type calcium channels and forskolin to increase aden-ylate cyclase activity. The effects of generation of nitric oxide (NO) by so-dium nitroprusside were observed at different levels of activities of L-type cal-cium channels. We only observed an alteration in the metabolic and functionaleffects of cyclic GMP by L-type calcium channels with high adenylate cyclaseactivity.

Materials and methods

New Zealand white rabbits (n = 8), weighing 2–3 kg, were used in this study. Experimentswere performed on ventricular myocytes isolated from the hearts of these rabbits. All ex-periments were conducted in accordance with Guide for the Care of Laboratory Animals (DHHS Publication No. 85-23, revised 1985) and were approved by our Institutional Ani-mal Care and Use Committee.

Cell dissociation

Freshly isolated ventricular myocytes were prepared by a standard method as described pre-viously [6, 23] with the following modifications. The rabbits were anesthetized (35 mg/kgsodium pentobarbital) and then heparinized (10 units/gram body weight) using the circum-flex ear vein. The heart was rapidly removed after an overdose of pentobarbital (60 mg/kg).Retrograde aortic perfusion of the heart was immediately begun at 70 mm Hg constant pres-sure with HEPES (N-2-hydroxyethylpiperazine-N′-2-ethane-sulfonic acid; pH = 7.5) buf-fered minimal essential medium (MEM). This solution contains (mM): 117 NaCl, 5.7 KCl,11 NaHCO3, 1.5 NaH2PO4, 1.7 MgCl2, 21.1 HEPES, 11.7 glucose, amino acids and vita-mins. We added 2 mM L-glutamine, 10 mM taurine and the pH was adjust to 7.2 with NaOH.This low-Ca2+ MEM solution had an osmolality of 296 mOsm and a free Ca2+ activity of2–5 µM. After 5 min of perfusion with low-Ca2+ MEM, the heart was perfused at 50 mmHg with a 60 ml volume of low-Ca2+ MEM supplemented with 0.1% collagenase (Worth-ington type II). All perfusion solutions were maintained at 37°C and equilibrated with oxy-gen. After 25 min of collagenase perfusion with recirculation, the heart was removed fromthe perfusion apparatus and cut into 8–10 pieces in MEM containing 1.0 mM CaCl2 and0.5% bovine serum albumin (BSA, Fraction V, Sigma, St. Louis, Mo.). This Ca2+-MEMwas supplemented with 0.1% collagenase. The tissue suspension was gently swirled in 50-ml centrifuge tubes at 37°C by a wrist action shaker (Multi-Mixer, Lab-Line Inst. Mel-rose Park, Ill., 2 cycles/s) for 5 min. A slurry containing isolated heart cells was decantedfrom the tissue suspension. The isolated cells were washed three times with the aid of lowspeed centrifuge (34 g) to completely remove the collagenase and some subcellular debris,then resuspended in Ca2+-MEM solution. Incubation of the remaining tissue with collage-nase was repeated at least two more times. The combined, washed cells were then main-tained at room temperature. The viability of the myocytes in MEM suspension was about70–80%. Yields were typically 10–14×106 rod-shaped cells per heart. Cells previously iso-lated in this fashion have been shown to have an intact cell coat as well as a functional sar-colemma and normal permeability barriers to extracellular ions, adenosine diphosphate(ADP) and succinate [29].

Myocyte oxygen consumption measurement

Steady state oxygen consumption (VO2) was recorded continuously with a two-channel Ox-imeter (University of Pennsylvania, Philadelphia, Pa.) fitted in a customized glass record-ing chamber. All experiments described below were performed under normoxic conditions.The recording started with a PO2 at 115 mm Hg and ended at no less than 25 mm Hg. An-

111

aerobic metabolism occurs only at PO2 levels below 5–10 mm Hg [17]. Gradients in PO2were not likely in the chamber, since the cell suspension was well stirred.

The chamber was made of glass and contained a small Teflon-coated stirring bar to main-tain the cells in suspension by slow rotation. The recording chamber was bathed with 37°Ccirculating water. The cuvette was mounted on a magnetic stirrer. A ground glass stopperwas used to eliminate the gas phase. This stopper also provided access to the assay medi-um via a central hole (1.3 mm internal diameter) for addition of agents during the experi-ment. The volume of the recording chamber is 1.5 ml. Myocytes were added to the cham-ber and the average number added was 43170±3180. The total volume of all drugs addedto the chamber were less than 100 µml so no significant dilution of the myocyte suspensionoccurred.

The VO2 was measured polarographically with a Clark-type electrode tightly fitted toone end of the recording chamber in direct contact with the myocyte suspension. The elec-trode was calibrated by putting the electrode into a solution saturated with two known con-centrations of oxygen. When calibrated, a 95% response could be obtained within 3 to 4 s.Myocytes were paced with electric field stimulation (1 Hz, 5 ms duration, voltage at 10%above threshold, and the polarity was alternated after each pulse) by two platinum wires in-serted into the center of the myocyte suspension. The rate of fall in oxygen tension withinthe chamber was used to determine VO2 of the myocytes. Data were collected with a desk-top computer and analyzed off-line. VO2 was expressed as nl O2/min per 100,000 myocy-tes; VO2 determinations were obtained in the same MEM solution used to resuspend thecells. The sample was stirred at a rate sufficient to keep the cells suspended and yet not sorapidly as to compromise the viability of the cells. Inspection of cellular morphometry, madeat random, indicated that at the completion of the experiment there was little loss of viabil-ity. We found at least 70–80% of the cells were quiescent rod-shaped cells at the comple-tion of the experiment.

Experiment protocol

The following protocol was used for the VO2 recordings and measurements of cell short-ening. Myocytes were suspended in the chamber with 2.0 mM Ca2+-MEM at an appropri-ate myocyte concentration and the cells were stabilized for 10–15 min. The average num-ber of myocytes in the chamber was 43170±3180. The myocytes were paced with electri-cal field stimulation. A 5-min recording was obtained as baseline. The VO2 was then ob-tained after: (1) adding sodium nitroprusside (NP 10–8,–6,–4 M), (2) pretreatment with BAYK8644 10–5 M (BAY, L-type calcium channel activator), nifedipine 10–4 M (NF, L-type cal-cium channel blocker) or forskolin 10–7 M (FK, adenylate cyclase activator), then addingNP10–8,–6,–4 M or (3) pretreatment with both FK 10–7 M and NF 10–4 M sequentially, thenNP 10–8,–6,–4 M. For cell shortening data and cyclic GMP data, the same protocol was used,except that we only used NP 10–4 M and NP 10–8,–4 M, respectively. At the end of experi-ment, the percentage of live cells was checked and found to be about the same. For the cy-clic GMP determination, the same experimental manipulations were performed on a por-tion of myocytes incubated in a flask following the same treatment procedure (timing, dos-age temperature, stirring, etc.). These myocytes were frozen in liquid nitrogen upon com-pletion of each respective drug treatment for later cyclic GMP determination.

Cell shortening measurement

Isolated cardiac myocytes were put into a chamber (37°C) on the stage of an inverted mi-croscope (Zeiss Axiovert 125) in 2.0 mM Ca2+ MEM solution. The volume of the chamberwas 3.5 ml. Myocytes were paced with electrical field stimulation (1 Hz, 5 ms duration,voltage at 10% above threshold and alternated polarity) by two platinum wires inserted intothe center of the myocyte chamber. Unloaded cell shortening was measured on-line usinga Myotrack system (Crystal Biotech) including a camera (Pulnix, TM-640) and a video-edge detector (Crystal Biotech, Model VED-114) that detected the change of the position

112

of both edges of the cell. The data were collected continuously. The output of video-edgedetector was fed into both a television monitor and a desktop computer, which was used toanalyze the data.

A 5-min stabilization was allowed after which contraction data for each myocyte wererecorded for a minimum of ten consecutive contractions. We used five myocytes for eachgroup per experiment. The experiments were divided into five groups: adding NP 10–4 M,adding BAY K8644 10–5 M, NF 10–4 M or FK 10–7 M separately, then adding NP 10–4 Mand adding forskolin 10–7 M and nifedipine 10–4 M sequentially, then sodium nitroprusside10–4 M. We then determined percentage shortening (%) and velocity of shortening (um · s–1).

Cyclic GMP measurement

In order to determine cyclic GMP levels, frozen samples were warmed to 0°C and homog-enized in ethanol using a Brinkmann Polytron in an ice bath. The homogenate was centri-fuged at 30,000 g for 15 min in a Sorvall RC-5B centrifuge. The supernate was recovered.The pellet was resuspended in 1 ml of 2:1 ethanol-water and centrifuged as before. Thecombined supernatants were evaporated to dryness in a 60°C bath under a stream of nitro-gen gas. The final residue was dissolved in 1.5 ml of assay buffer (0.05 M sodium acetate,pH 5.8, containing sodium azide). Cyclic GMP levels were determined by radioimmunoas-say using a scintillation proximity assay (Amersham). This assay measures the competitivebinding of 125I-labeled cyclic GMP to a cyclic GMP specific antibody. After constructionof a standard curve, cyclic GMP levels were determined directly from the counts in pico-moles divided by the number of myocytes per tube times 100,000.

Statistics

Results are expressed as means±SEM. A repeated measure analysis of variance (ANOVA)was used to compare variables measured in the experimental and control conditions.Duncan’s multiple range test was used to compare differences post hoc. This analysis wasused to determine differences between groups and treatments for myocyte function, VO2and cyclic GMP levels. We also used the paired-T test to test the effects of BAY, NF andFK on function and VO2. In all cases, a P<0.05 was accepted as significant.

Results

Control percentage shortening (Pcs) averaged 5.7±0.6% and the rate of short-ening (Rs) was 65.5±4.5 (µm/s). The effects of NP, BAY K8644, NF and FKon cell shortening data are showed in Figs. 1 and 2. The basal VO2 of theelectrically stimulated cardiac myocytes was 706±34 nl O2/min per 105 myo-cytes, Fig. 3. All cyclic GMP data are summarized in Table 1 and the controlbaseline value was 264±70 pmol per 105 myocytes. There were no statisti-cally significant differences in the initial basal conditions for Pcs, Rs or VO2between any groups.

As showed in Table 1, cyclic GMP was increased by NP in a dose-depen-dent pattern. This difference from baseline was significant at a nitroprussideconcentration of 10–4 M and represented an increase of 288% over the controlvalue. BAY K8644, NF and FK did not change the level of cyclic GMP frombaseline. In the presence of these agents, the increases in cyclic GMP causedby NP were not different from those occurring without these agents.

113

114

Fig. 1. The effects of sodium nitroprusside (NP) under control conditions or with BAYK8644 (BAY) or nifedipine (NF) (top) and with forskolin (FK) with and without NF (bottom) on the percentage shortening of beating myocytes isolated from rabbits hearts (n = 8, cell number = 4 for each treatment). Note that percentage decline caused by NP wasnot altered by BAY or NF. However, NF reduced the percentage decline after FK. * Signif-icantly different from baseline; ** significantly different from baseline (paired-T test); ‡ significantly different from BAY 10–5 M; ¤ significantly different from FK 10–7 M

115

Fig. 2. The effects of NP under control conditions or with BAY or NF (top) and with FKwith and without NF (bottom) on the rate of shortening of beating myocytes isolated fromrabbits hearts (n = 8, cell number = 4 for each treatment). Note that percentage declinecaused by NP was not altered by BAY or NF. However, NF reduced the percentage declineafter FK. * Significantly different from baseline; ** significantly different from baseline (paired-T test); ‡ significantly different from BAY 10–5 M; ¤ significantly different fromFK 10–7 M

Sodium nitroprusside 10–4 M decreased Pcs of the myocytes from 5.7±0.6to 3.7±0.5. BAY K8644 10–5 M increased Pcs from 5.3±0.8 to 7.2±0.6, whileNF decreased Pcs from 6.5±0.6 to 4.3±0.6 (Fig. 1). In the presence of eitherBAY K8644 10–5 M or NF 10–4 M, the effects of NP on Pcs were not changed.FK 10–7 M elevated Pcs (4.3±0.5 vs. 5.7±0.4). In the presence of FK 10–7 M,NP diminished Pcs from 5.7±0.4 to 3.2±0.3; the percentage decrease was44%. In the group pretreated with FK and NF sequentially, this value was 25%,which was significantly different from the other groups.

116

Fig. 3. The effects of NP under control conditions or with BAY or NF (top) and with FKwith and without NF (bottom) on the oxygen consumption of beating myocytes isolatedfrom rabbits hearts (n = 7). Note that percentage decline caused by NP was not altered byBAY or NF. However, NF reduced the percentage decline after FK. * Significantly differ-ent from baseline; ** significantly different from baseline (paired-T test); † significantlydifferent from NP 10–8 M; ‡ significantly different from BAY 10–5 M; ° significantly dif-ferent from NF 10–4 M; ¤ significantly different from FK 10–7 M

The rate of shortening was decreased by nitroprusside 10–4 M from65.5±4.5 to 46.2±5.5 µm/s. BAY K8644 10–5 M increased Rs from 59.2±12.6to 80.2±12.4, while NF significantly decreased Rs (Fig. 2). In the presence ofeither BAY K8644 10–5 M or NF 10–4 M, the effects on Rs of NP were notchanged. FK 10–7 M elevated Rs (51.2±3.6 vs. 67.0±5.0). In the presence ofFK 10–7 M, NP diminished Rs from 67.0±5.0 to 40.6±4.3 and this percent-age decrease was 39%. In the group pretreated with FK and NF sequentially,this value was 20%, which was different from the other groups.

Adding NP to the myocytes suspension resulted in a dose-dependent de-crease in myocyte VO2. With the highest dose of NP, 10–4 M, VO2 was de-creased from 706±34 (nl O2/min per 100,000 myocytes) to 410±13. NF alsoinduced a drop of VO2. Both BAY K8644 and FK increased VO2. These ef-fects were significant using a paired-T test. The effects of NP on VO2 werenot changed by pretreatment with BAY K8644, NF, or FK. In the group treatedwith FK alone, VO2 was decreased by NP 10–4 M from 756±35 to 375±16and the percentage decrease was 50%. In the group treated with both FK 10–7 Mand NF 10–4 M, FK increased VO2 and NF decreased it. In this group, NP10–4 M decreased VO2 from 528±5 to 333±15, which was a 37% decrease.The percentage decrease was significantly different from the group treatedwith FK alone.

Discussion

The major finding of this study was that in contracting cardiac myocytes, thenegative metabolic and functional effects of cyclic GMP were not affected byeither stimulation or inhibition of L-type calcium channels. However, whenadenylate cyclase was stimulated with FK, part of the negative metabolic andfunctional effects of cyclic GMP appeared to act through inhibition of L-typecalcium channels. Adding NP led to an increase in the cyclic GMP level anda decrease on both VO2 and cell shortening in a dose-dependent manner. Others have reported that increases in myocyte cyclic GMP inhibit the L-type

117

Table 1. Effects of sodium nitroprusside on the level of cyclic guanosine monophosphate(GMP) (fmol/105 myocytes) with different degrees of L-type calcium channel activity

Control BAY K8644 Nifedipine Forskolin Forskolin +nifedipine

Control BAY 10–5 M NF 10–4 M FK 10–7 M FK/NF264 ± 70 304 ± 82 308 ± 61 258 ± 51 287 ± 64

NP 10–8 M NP 10–8 M NP 10–8 M NP 10–8 M NP 10–8 M462 ± 127 541 ± 196 427 ± 88 308 ± 45 376 ± 95

NP 10–4 M NP 10–4 M NP 10–4 M NP 10–4 M NP 10–4 M760 ± 283* 723 ± 297*†‡ 806 ± 220° 931 ± 332 ¤ 767 ± 159 ¤°†

* Significantly different from baseline, † significantly different from NP 10–8 M, ‡ signif-icantly different from BAY 10–5 M, ° significantly different from NF 10 –4 M, ¤ significant-ly different from FK 10–7 M

calcium channel [9, 10, 21, 24, 27]. We found that with electrical stimulation,when we added either an L-type calcium channel activator or blocker, the ef-fects of NP were not changed. When we added an adenylate cyclase activatorand an L-type calcium channel blocker sequentially, the effects of NP weresignificantly reduced. This indicated that the L-type calcium channel was im-portant in the negative functional and metabolic effects of cyclic GMP onlywhen it was activated by adenylate cyclase.

The use of isolated ventricular myocytes in this study reduced concernsarising from the use of heart tissues containing heterogeneous cell types, whichcould act as confounding sources of both cyclic GMP and oxygen consump-tion measured. The yields were high with 70–80% of rod-shaped healthy cells.The viability of the cells was confirmed by re-checking the percentage of rod-shaped cells at the end of each experiment. The cells were still capable of in-creasing VO2 at the end of the experiment. Dinitrophenol at least doubled thebasal oxygen consumption. Measuring errors with regard to VO2 and cyclicGMP due to damaged cells should be small, with approximately 20% roundcells. These cells still metabolized to an unknown extent; however, this wouldonly lead to a shift of absolute values without altering the conclusions drawnfrom experiments. With regard to the functional measurements, we only chosecells that shortened at least 4% and would contract in the steady state for20 min. We performed functional measurements on four cells per heart.

Cyclic GMP exerts negative metabolic and functional effects on cardiacmyocytes. The negative effects of cyclic GMP were shown to be greater underconditions of higher cellular metabolism [6], which could occur with high cal-cium or electrical stimulation. The effects of cyclic GMP also appeared to beenhanced during sympathetic activation or increased levels of cyclic AMP [2,12, 14, 28]. These negative effects may act through inhibition of L-type cal-cium channel, protein phosphorylation by cyclic GMP-dependent protein ki-nases or activation of cyclic GMP-stimulated or inhibited cyclic AMP phos-phodiesterases. In frog cardiac myocytes, the work of Fischmeister and Hart-zell showed that an inhibitor of cyclic GMP-stimulated cyclic AMP phospho-diesterases could reversed the negative inotropic effect of cyclic GMP [5, 7].One study reported that cyclic GMP might exert a direct stimulatory effect onL-type calcium channels [14]. There are were several studies using patch clampthat showed that there was an inhibition of L-type calcium channels with anincrease of cyclic GMP level in various species [9, 10, 21, 24, 27]. It was alsoshown that cyclic GMP reduces calcium influx [16]. One study in rat myocy-tes suggested that cyclic GMP inhibited L-type calcium channel activity byactivating cyclic GMP-dependent protein kinase [13]. It was also reported thatcyclic GMP had some cross-activation of cyclic AMP protein kinase [2]. Sincethere was evidence that cyclic GMP affects calcium entry through L-type cal-cium channels [10, 13–15, 18, 21, 22, 25], it was logical to assume that thiswould have functional and metabolic consequences.

Intracellular calcium concentration has been put forward as a key playerin determining the metabolic rate of the heart. It has been reported that rateof oxygen uptake of isolated cardiac myocytes was determined as a functionof the cytosolic free calcium level [17]. The L-type calcium channel is themajor point of entry for calcium ions into the cell during action potentials. Inthe present study, all of the experiments were carried out in 2.0 mM Ca2+ me-

118

dium. We used electrical field stimulation to pace the cells. With electricalstimulation, cells were in a contracting, high metabolic state and L-type cal-cium channel activity was elevated. Under these conditions, the activity ofL-type calcium channels could play a role in the effects of cyclic GMP inmyocytes.

From a previous study performed by our laboratory, we know that in quies-cent cardiac myocytes neither L-type calcium channel activation nor block-ade affected the negative metabolic effects of cyclic GMP [30]. This indicateda lack of importance of L-type calcium channels in the effects of cyclic GMPat rest. In our current study, we used electric field stimulation, which shouldhave increased the relative importance of L-type calcium channels in the neg-ative effects caused by cyclic GMP. Increasing doses of NP led to an eleva-tion of cyclic GMP level and a decrease in VO2, Pcs and Rs. If the inhibitionof cyclic GMP on L-type calcium channels played a crucial role in these ef-fects, then activators or blockers of L-type calcium channels should alter therelationship between cyclic GMP level and VO2 and functional parameters.We did not find an effect of these pharmacological agents. This was in agree-ment with the work of Finkel et al. [4]. They found that calcium channel ac-tivation did not affect myocyte responses to NO. A recent study also demon-strated that in the in vivo rabbit heart, blockade of L-type calcium channelsdid not affect the positive inotropic and metabolic effects of lowering the levelof cyclic GMP [11].

In the two experimental groups in which we added FK, there was an in-crease in both VO2 and Pcs. The negative effects of increasing cyclic GMPwere not changed after adenylate cyclase activation. However, when we addedNF to FK-pretreated cells, the negative metabolic and functional effects of in-creasing doses of NP were significantly less. The functional and metabolicvalues after both FK and NF were not significantly different from other groupsprior to the addition of nitroprusside. This data indicated that the negative func-tional and metabolic effects of cyclic GMP were partly caused by blockade ofL-type calcium channels under these conditions. There are interactionsbetween the second messengers cyclic AMP and cyclic GMP at the L-type cal-cium channel [7, 13, 14, 21, 22].

Previous studies have suggested that cyclic GMP inhibited cyclic AMP-stimulated L-type calcium channel activity [7, 12, 21]. When cyclic AMP lev-els were high, the cyclic AMP-protein kinases were activated. This may phos-phorylate the myocyte L-type calcium channel [8]. Under these conditions,cyclic GMP may have some cross effects with cyclic AMP-dependent proteinkinase. There is evidence that some of the effects of cyclic GMP at the L-typecalcium channel are related to a cyclic GMP dependent protein kinase [9, 13,21, 22, 24, 27]. It may be that this effect is important only when the channelhas been previously phosphorylated by cyclic AMP protein kinase. Increas-ing the level of cyclic GMP also leads to the activation of cyclic GMP-stim-ulated or inhibited cyclic AMP phosphodiesterases, which also indicates inter-action between cyclic GMP and cyclic AMP pathways. These interactions havealso been suggested to be an important mechanism of action of cyclic GMP inmyocytes [12, 14]. The negative functional effects of raising cyclic GMP mayonly be important in cardiac myocytes when they have been prestimulated byadenylate cyclase activation.

119

It was also thought possible that the use of activators and blockers of the L-type calcium channel could affect the level of cyclic GMP, which could inducea change in the relationship between different doses of NP and cyclic GMP. Asshown in Table 1, there were no differences in the level of cyclic GMP at base-line or after pretreated with BAY K8644, NF, FK or FK and NF together. Thus,the effects of NP in terms of alterations in cyclic GMP were similar, regardlessof our activation or blockade of the L-type calcium channels.

In summary, we found that increasing doses of NP induced an elevation incyclic GMP levels and a decrease in VO2 and functional parameters in con-tracting cardiac myocytes isolated from rabbits. The relationship between thelevel of cyclic GMP and function and metabolism was not changed when L-type calcium channels were either pharmacologically activated or inhibited.When adenylate cyclase was activated with FK, metabolic and functional pa-rameters increased. With FK, blockade of L-type calcium channels signifi-cantly diminished the negative metabolic and functional effects of increasesin cyclic GMP. We conclude that in beating cardiac myocytes, the negativemetabolic and functional effects of cyclic GMP were related to L-type calciumchannels only during activation of adenylate cyclase.

Acknowledgements. This study was supported, in part, by U. S. P. H. S grant HL 40320 anda grant-in-aid from the American Heart Association, New Jersey Affiliate.

References

1. Bkaily G, Sperelakis N (1985) Injection of guanosine-3′,5′-cyclic monophosphate intoheart cells blocks calcium slow channels. Am J Physiol 248:H745–H749

2. Cornell TL, Arnold E, Boerth NJ, Lincoln TM (1994) Inhibition of smooth musclegrowth by nitric oxide and activation of cAMP-dependent protein kinase by cGMP. AmJ Physiol 267:C1405–C1413

3. Droogmans G, Nilius B (1989) Kinetic properties of the cardiac T-type calcium chan-nel in the guinea-pig. J Physiol (Lond) 419:627–650

4. Finkel MS, Oddis CV, Mayer OH, Hattler BG, Simmons RL (1995) Nitric oxide syn-thase inhibitor alters papillary muscle force-frequency relationship. J Pharmacol ExpTher 272:945–952

5. Fischmeister R, Hartzell HC (1987) Cyclic guanosine-3′,5′-monophosphate regulatesthe calcium current in single cells from frog ventricle. J Physiol (Lond) 387:453–472

6. Gong GX, Weiss HR, Tse J, Scholz PM (1997) Cyclic GMP decreases cardiac myocyteoxygen consumption to a greater extent under conditions of increased metabolism. J Cardiovasc Pharmacol 30:537–543

7. Hartzell HC, Fischmeister R (1986) Opposite effects of cyclic GMP and cyclic AMPon Ca2+ current in single heart cells. Nature 323:273–275

8. Hove-Madsen L, Mery PF, Jurevicius J, Skeberdis AV, Fischmeister R (1996) Regula-tion of myocardial calcium channels by cyclic AMP metabolism. Basic Res Cardiol 91[Suppl 2]:1–8

9. Kristein M, Rivet-Bastide M, Hatem S, Benardeau A, Mercadier JJ (1995) Nitric ox-ide regulates the calcium current in isolated human atrial myocytes. J Clin Invest95:794–802

10. Kumar R, Namiki T, Joyner RW (1997) Effects of cGMP on L-type calcium current ofadult and newborn rabbit ventricular cells. Cardiovasc Res 33:573–582

11. Leone R, Naim KL, Scholz PM, Weiss HR (1998) Increased O2 consumption and pos-itive inotropy caused by cyclic GMP reduction are not altered by L-type calcium chan-nel blockade. Pharmacol 56:37–45

120

12. Lohmann SM., Fischmeister R, Walter U (1991) Signal transduction by cGMP in heart.Basic Res Cardiol 86:503–514

13. Mery P-F, Lohmann SM, Walter U, Fischmeister R (1991) Ca2+ current is regulated bycyclic GMP-dependent protein kinase in mammalian cardiac myocytes. Proc Natl Acad Sci 88:1197–1201

14. Mery P-F, Pavoine C, Belhassen L, Pecker F, Fischmeister R (1993) Nitric oxide reg-ulates cardiac Ca2+ current. Involvement of cGMP-inhibited and cGMP-stimulatedphosphodiesterases through guanylyl cyclase activation. J Biol Chem 268:26286–26295

15. Murad F (1994) The nitric oxide-cGMP signal transduction system for intracellular andintracellular communication. Rec Prog Hormone Res 49:239–248

16. Nawrath H (1977) Does cyclic GMP mediate the negative inotropic effect of acetyl-choline in the heart? Nature 267:72–74

17. Rose H, Schnitzler N, Kammermeier H (1987) Electrically stimulated cardiomyocy-tes: evaluation of shortening, oxygen consumption and tracer uptake. Biomed BiochimActa 9:622–627

18. Schmidt HHHW, Lohmann SM, Walter U (1993) The nitric oxide and cGMP signaltransduction system: regulation and mechanism of action. Biochim Biophys Acta1178:153–175

19. Shah, AM, Spurgeon HA, Sollott SJ, Talo A, Lakatta EG (1994) 8-bromo-cGMP reduces the myofilament response to Ca2+ in intact cardiac myocytes. Circ Res 74:970–978

20. Sperelakis N (1994) Regulation of calcium slow channels of heart by cyclic nucleo-tides and effects of ischemia. Adv Pharmacol 31:1–24

21. Sperelakis N, Katsube Y, Yokoshiki H, Sada H, Sumii K (1996) Regulation of the slowCa++ channels of myocardial cells. Mol Cell Biochem 64:85–98

22. Sperelakis N, Tohse N, Ohya Y, Masuda H (1994) Cyclic GMP regulation of calciumslow channels in cardiac muscle and vascular smooth muscle cells. Adv Pharmacol26:217–252

23. Straznicka M, Gong G, Tse J, Scholz PM, Weiss HR (1997) The cyclic GMP level whichreduces cardiac myocyte O2 consumption is altered in renal hypertension. Am J Phys-iol 273:H1949–H1955

24. Sumii K, Sperelakis N (1995) cGMP-dependent protein kinase regulation of the L-typeCa2+ current in rat ventricular myocytes. Circ Res 77:803–812

25. Tohse N, Nakaya H, Takeda Y, Kanno M (1995) Cyclic GMP-mediated inhibition ofL-type Ca2+ channel activity by human natriuretic peptide in rabbit heart cells. Br JPharmacol 114:1076–1082

26. Trautwein W, Taniguchi J, Noma A (1982) The effect of intracellular cyclic nucleo-tides and calcium on the action potential and acetylcholine response of isolated cardiaccells. Pflugers Arch 392:307–314

27. Wahler GM, Dollinger SJ (1995) Nitric oxide donor SIN-1 inhibits mammalian cardiaccalcium current through cGMP-dependent protein kinase. Am J Physiol 268:C45–C54

28. Weiss HR, Rodriguez E, Tse J, Scholz PM (1994) Effect of increased myocardial cy-clic GMP induced by cyclic GMP-phosphodiesterase inhibition on oxygen consump-tion and supply of rabbit hearts. Clin Exp Physiol Pharmacol 21:607–614

29. Wittenberg BA, Robinson TF (1981) Oxygen requirements, morphology, cell coat andmembrane permeability of calcium-tolerant myocytes from hearts of adult rats. CellTissue Res 216:231–251

30. Yan L, Gong GX, Tse J, Scholz PM, Weiss HR (1998) Reduced function and metab-olism caused by cGMP are altered by L-type calcium channels only at high adenylatecyclase activity. FASEB J 12:A73 (abstract)

121