Sarcoplasmic Reticulum-associated Cyclic Adenosine 5'-Monophosphate Phosphodiesterase Activity in Normal and Failing Human Hearts Matthew A. Movsesian,* Carolyn J. Smith,* Judith Krall,* Michael R. Bristow,* and Vincent C. Manganiello$ *Cardiology Division, University of Utah Medical Center, Salt Lake City, Utah 84132; and $Laboratory of Cellular Metabolism, National Heart, Lung, and Blood Institute, National Institutes ofHealth, Bethesda, Maryland 20814 Abstract often referred to as PDE III), which are characterized by their high affinity for cAMP, insensitivity to rolipram, and suscepti- Sarcoplasmic reticulum-associated cAMP phosphodiesterase bility to competitive inhibition by low concentrations of activity was examined in microsomes prepared from the left cGMP, cilostamide derivatives, and some of the newer inotro- ventricular myocardium of eight heart transplant recipients pic agents such as milrinone, amrinone, and enoximone (2-8). with end-stage idiopathic dilated cardiomyopathy and six un- In addition, the association of cGI PDE with the sarcoplasmic matched organ donors with normal cardiac function. At cAMP reticulum appears to be an important determinant of the ino- concentrations < 1.0 MM, sarcoplasmic reticulum-associated tropic efficacy of these agents. In dog and rhesus monkey, cAMP phosphodiesterase activity was functionally homoge- whose cGI PDE activity is localized to the sarcoplasmic reticu- neous. cAMP phosphodiesterase activity was inhibited compet- lum, cGI PDE inhibitors are very effective as inotropic agents. itively by cGMP (K; = 0.031±0.008 MM) and the cilostamide These inhibitors are much less effective inotropic agents in derivative OPC 3911 (K; = 0.018±0.004 gM), but was essen- hamster, guinea pig, and rat, whose cGI PDE activity is pre- tially insensitive to rolipram. V.,,. and K. were 781.7±109.2 dominantly cytoplasmic (9, 10). nmol/mg per min and 0.188±0.031,MM, respectively, in micro- Studies of tension development in muscle strips obtained somes prepared from nonfailing hearts and 793.9±68.9 nmol/ from normal and failing human hearts have demonstrated a mg per min and 0.150±0.027MgM in microsomes prepared from decreased inotropic response to phosphodiesterase inhibitors failing hearts. Microsomes prepared from nonfailing and fail- in the latter (11, 12). As suggested in these reports, this dimin- ing hearts did not differ with respect to either the ratio of ished augmentation of tension development may reflect im- cAMP phosphodiesterase activity to ATP-dependent Ca2" ac- paired cAMP generation. It is also plausible, however, that alter- cumulation activity or the sensitivity of cAMP phosphodiester- ations in sarcoplasmic reticulum-associated cGI PDE activity ase activity to inhibition by OPC 3911. These data suggest that contributed to the diminished response. To investigate this pos- the diminished inotropic efficacy of phosphodiesterase inhibi- sibility, we examined and compared the characteristics of sar- tors in failing human hearts does not result from changes in the coplasmic reticulum-associated cAMP phosphodiesterase activ- level, kinetic properties, or pharmacologic sensitivity of sarco- ity in microsomes prepared from normal and failing human plasmic reticulum-associated cAMP phosphodiesterase activ- hearts. ity. (J. Clin. Invest. 1991.88:15-19.) Key words: phosphodies- terase inhibitors - cyclic guanosine 5'-monophosphate - cilosta- mide * rolipram - Ca2+ transport Methods Introduction Procurement ofhuman cardiac tissue. Left ventricular free wall myo- cardium was obtained from the excised failing hearts ofeight transplant Certain drugs that inhibit cAMP hydrolysis by cyclic nucleo- recipients (age 49.1±5.4 yr) with class IV heart failure resulting from tide phosphodiesterases increase myocardial contractility, and idiopathic dilated cardiomyopathy, and the nonfailing hearts of six unmatched organ donors (age 40.3±6.7 yr). The sarcolemmal fl-adren- there is interest in the potential usefulness of such agents in the ergic receptor density, measured as previously described (13, 14), was treatment of heart failure. Five distinct families of cyclic nu- 99.8±14.4 fmol/mg in the nonfailing hearts and 49.8±5.0 fmol/mg in cleotide phosphodiesterases have been identified and charac- the failing hearts (P < 0.01). terized (1). Studies in animals suggest that the inotropic effi- Preparation ofhuman cardiac microsomes. Microsomes were pre- cacy of phosphodiesterase inhibitors is most closely related to pared from left ventricular free wall myocardial tissue as previously their ability to inhibit the activity of enzymes belonging to the described (15), save that the pH of the buffer solutions used in the cGMP-inhibited cAMP phosphodiesterase family (cGI PDE', homogenization and differential sedimentation steps was adjusted to 6.0 rather than 7.0. This modification increased the specific cAMP phosphodiesterase activity without decreasing total cAMP phospho- diesterase activity. Address correspondence to Dr. Matthew A. Movsesian, Cardiology Measurement ofcAMPphosphodiesterase activity. The method for Division, Room 4A-100, University of Utah Health Sciences Center, measuring cAMP phosphodiesterase activity was adapted from the 50 North Medical Drive, Salt Lake City, UT 84132. protocol described by Kincaid and Manganiello (16). Unless otherwise Received for publication 25 October 1990 and in revisedform 16 stated, microsomes were suspended at 0.012 mg/ml in a reaction mix- January 1991. ture comprised of 0.1 mM EGTA, 8.3 mM MgCl2, and 50 mM Hepes (pH 7.5, 30°C). cAMP hydrolysis was initiated by addition of [3H]- 1. Abbreviation used in this paper: cGI-PDE, cGMP-inhibited cAMP cAMP (New England Nuclear, Boston, MA); each assay was performed phosphodiesterase. in duplicate. The reaction was stopped by the addition of 10.1 mM unlabelled cAMP and 5 mM unlabelled 5'AMP in 0.25 N HC. The The Journal of Clinical Investigation, Inc. reaction mixture was neutralized with NaOH, and [3H]5'AMP was Volume 88, July 1991, 15-19 converted to [3H]adenosine by incubation with Crotalus atrox venom Human Cardiac Sarcoplasmic Reticulum-associated Phosphodiesterase 15 (30 mg/ml) at 30'C for 30 min. The reaction mixture was then applied 100 Figure 2. Inhibition of to QAE-Sephadex columns, and adenosine was eluted with H20 at Z g X cAMP phosphodiester- neutral pH. cAMP hydrolysis was determined by measuring [3H]- Q80 - 5 i rolipram ase activity by OPC adenosine in the eluent by scintillation spectrometry. Values for V,,[, 3911 (open circles), Ki, and Ki were determined by weighted nonlinear least squares regres- e60 60Y~cOMP \ cGMP (filled circles), 3911 sion analysis (Gauss-Newton method) using published computer pro- 0. OPC o 40 and rolipram (open tri- grams (17). angles). Measurements Microsomal cAMP phosphodiesterase activity was also measured E 20 were made at 0.2 iM using the same technique in buffer comprised of 0.102 mM KCI, 5.24 10E. cAMP. The sarcoplas- mM MgCl2, 1.00 mM EGTA, 0.40 mM CaC12, 5 mM ATP, and 20 o mic reticulum prepara- 0.001 0.01 0.1 I 10 100 1000 104 mM [N-morpholino]propanesulfonic acid (MOPS) (pH 7.05, 370C). tionto wasa theh samea eaas The free Mg2+ concentration was calculated to be 0.8 mM and the free [drug], ;iM that used in Fig. 1. Each Ca2+ concentration was calculated to be 0.2 uM (18). point represents the mean±standard error of four to six determina- Miscellaneous. Protein was measured according to the method of tions. Bradford using bovine serum albumin as a standard (19). Lipids were extracted from microsomes by the method of Bligh and Dyer (20), and the phosphorus contents of the extracts were quantified by the method of Chen et al. (21). Oxalate-supported, ATP-dependent Ca2` accumu- pram was observed only at very high concentrations ofthe drug lation by human cardiac microsomes was measured as previously de- (IC50 = 76.9 ,M). When inhibition of cAMP phosphodiester- scribed (15) in the presence of 0.102 M KCl, 5.05 mM MgCl2, 1.0 mM ase activity by cGMP and OPC 391 1 was examined as a func- EGTA, 0.99 mM CaCl2, 5.0 mM oxalic acid, 5.0 mM NaN3, 0.5 mM tion of cAMP concentration, it was seen to be competitive, ryanodine, 5.0 mM ATP, and 20 mM MOPS (pH = 7.05, 37°C). The with K1 values of0.031±0.008 MM for cGMP and 0.018±0.004 free Ca2+ concentration was calculated to be 5.0,uM (18). OPC 3911 ,MM for OPC 391 1 (Fig. 3). This pattern of inhibition is typical was supplied by Otsuka Chemicals, Ltd., Osaka, Japan. of cGI PDE. These observations demonstrated the feasibility of deter- Results mining values for the steady-state kinetic parameters of sarco- plasmic reticulum-associated cAMP phosphodiesterase in mi- Sarcoplasmic reticulum-associated cAMP phosphodiesterase crosomes prepared from normal and failing human hearts. Be- activity was measured under standard conditions in micro- fore comparing these values, however, it was necessary to somes prepared from left ventricular free wall myocardium demonstrate that our preparative methods resulted in the isola- obtained from failing human hearts. In Fig. 1, cAMP phospho- tion of comparable microsomal fractions in the two popula- diesterase activity is plotted as a function of cAMP concentra- tions. As shown in Table I, microsomes prepared from normal tion in double reciprocal form. The linearity of the double re- and failing hearts were similar with respect to yield and phos- ciprocal plot indicates that the cAMP phosphodiesterase activ- pholipid:protein ratio (expressed as nanomoles of extractable ity associated with the sarcoplasmic reticulum follows the phosphorus per milligram protein). In addition, oxalate-sup- Michaelis-Menton steady-state model, with Vm,, = 533±18 ported, ATP-dependent Ca2" accumulation activity was as- pmol/mg-min and Km = 0.133±0.013 MM in this preparation. sayed as an index of sarcoplasmic reticulum function. As in The high affinity for cAMP suggested that the cAMP phos- phodiesterase activity associated with human cardiac sarco- plasmic reticulum reflected the presence of either cGMP- or a 0.016 rolipram-inhibited phosphodiesterases.