QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Submit a response Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Rybin, V. O. Articles by Steinberg, S. F. Search for Related Content PubMed PubMed Citation Articles by Rybin, V. O. Articles by Steinberg, S. F.

Developmental Changes in ₂-Adrenergic Receptor Signaling in Ventricular Myocytes: the Role of Gi proteins and Caveolae Microdomains

Departments of Pharmacology (V.O.R., E.P., S.A., S.F.S.) and Medicine (S.F.S.), College of Physicians and Surgeons, Columbia University, New York, New York

Received January 22, 2003; accepted February 27, 2003

	Abstract

Top Abstract Materials and Methods Results Discussion References

 
Cardiomyocyte ₂-adrenergic receptors (-ARs) provide a source of inotropic support and influence the evolution of heart failure. Recent studies identify distinct mechanisms for ₂-AR actions in neonatal and adult rat cardiomyocytes. This study examines whether ontogenic changes in cardiac ₂-AR actions can be attributed to altered Gi expression or changes in the spatial organization of the ₂-AR complex in membrane subdomains (caveolae). We show that ₂-ARs increase cAMP, calcium, and contractile amplitude in a pertussis toxin (PTX)-insensitive manner in neonatal cardiomyocytes. This is not caused by lack of Gi; Gi expression is higher in neonatal cardiomyocytes than in those of adult rats. ₂-ARs provide inotropic support without detectably increasing cAMP, in adult cardiomyocytes. This cannot be attributed to dual coupling of ₂-ARs to Gs and Gi, because ₂-ARs do not promote cAMP accumulation in PTX-pretreated adult cardiomyocytes. Spatial segregation of ₂-ARs, Gs/Gi, and adenylyl cyclase to distinct membrane subdomains also is not a factor, because all of these proteins copurify in caveolin-3-enriched vesicles isolated from adult cardiomyocytes. However, these studies demonstrate that enzyme-based protocols routinely used to isolate ventricular cardiomyocytes lead to proteolysis of -ARs. The functional consequences of this limited -AR proteolysis is uncertain, because truncated ₁-ARs promote cAMP accumulation and truncated ₂-ARs provide inotropic support in adult cardiomyocytes. Collectively, these studies indicate that components of the ₂-AR signaling complex compartmentalize to restricted membrane subdomains in adult rat cardiomyocytes. Neither compartmentalization nor changes in Gi expression fully explain the ontogenic changes in ₂-AR responsiveness in the rat ventricle.

	Materials and Methods

Top Abstract Materials and Methods Results Discussion References

 
Preparation of Cultured Neonatal and Acutely Disaggregated Adult Rat Ventricular Myocytes. Cardiomyocytes were isolated from the hearts of 2-day-old Wistar rats by a trypsin dispersion procedure according to a protocol that incorporates a differential attachment procedure to enrich for cardiomyocytes followed by irradiation as described previously (Kuznetsov et al., 1995). The yield of cardiomyocytes typically is 2.5 to 3 x 10⁶ cells per neonatal heart. Cells were plated at a density of 0.5 x 10⁶ cells/ml on protamine sulfate-coated culture dishes (for biochemical assays) or glass cover-slips (for calcium and cell motion experiments). Experiments were performed after 5 to 6 days of culture in minimal essential medium (Invitrogen, Carlsbad, CA) with 10% fetal calf serum, 5 x 10^–6 M hypoxanthine, and 12 mM NaHCO₃. Adult rat ventricular myocytes were disaggregated according to methods described in detail previously (Kuznetsov et al., 1995). Cells were used within 1 to 6 h of isolation.

Measurements of cAMP. Intracellular cAMP accumulation was measured in neonatal myocytes grown in 22.1-mm multiwell dishes or in rod-shaped adult myocytes as described previously (Kuznetsov et al., 1995). Preincubation was with 10 mM theophylline for 60 min at room temperature and was followed by assays performed for 5 min at room temperature in the absence or presence of isoproterenol (10^–7 M), carbachol (10^–5 M, starting 5 min before isoproterenol), or forskolin (10^–7 M) as indicated. Assays were terminated by removal of the incubation buffer and addition of 1 ml of ethanol, with each condition performed on three wells and assayed for cAMP in quadruplicate. Aliquots of the alcohol-fixed cell extracts were dried under a stream of nitrogen, and cAMP in the residue was determined by radioimmunoassay (PerkinElmer Life Sciences, Boston, MA).

Pertussis Toxin-Dependent ADP-Ribosylation of G Proteins. A crude membrane fraction was prepared by scraping cells into a buffer containing 0.25 M sucrose, 1 mM EDTA, 0.1 mM phenylmethylsulfonyl fluoride, and 50 mM Tris, pH 7.6, followed by homogenization with a Polytron homogenizer and centrifugation at 43,000g for 45 min. Membranes were resuspended in 50 mM Tris buffer, pH 7.6, containing 2 mM MgCl₂ and 1 mM EDTA at a concentration of 2 to 3 mg/ml. ADP-ribosylation assays were performed as follows: membranes were incubated in 20 µl of a 50 mM Tris-chloride buffer, pH 8.0, containing 2 mM MgCl₂, 1 mM EDTA, 10 mM dithiothreitol, 0.1% Lubrol PX, 10 mM thymidine, 10 µM [³²P]NAD (1.5 µCi per assay), and 20 µg/ml PTX for 1 h at 37°C. The reaction was terminated by addition of SDS-PAGE sample buffer and boiling for 5 min. Electrophoresis was performed on vertical slab gels (resolving gel, 12%; stacking gel, 3.9% acrylamide) and was followed by autoradiography.

Preparation of Caveolae Membranes. Fractions enriched in the muscle-specific caveolin-3 isoform were prepared according to a detergent-free purification scheme essentially as described previously (Rybin et al., 2000). All steps were carried out at 4°C. Briefly, cells from five 100-mm diameter dishes were washed twice with ice-cold phosphate-buffered saline and then scraped into 0.5 M sodium carbonate, pH 11.0 (0.5 ml/dish). Cells from five dishes were combined (total volume, 2.5 ml) for each preparation. The extract was sequentially disrupted by homogenization with a loose-fitting Dounce homogenizer (10 strokes), a Polytron tissue grinder (three 10-s bursts), and a tip sonicator (three 20-s bursts). The homogenate was then adjusted to 40% sucrose by adding an equal volume of 80% sucrose prepared in MES-buffered saline (25 mM MES, pH 6.5, and 0.15 M NaCl), placed on the bottom of an ultracentrifuge tube, overlaid with a 5 to 30% continuous sucrose gradient, and centrifuged at 38,000 rpm for 16 to 18 h in a SW40 rotor (Beckman Coulter, Palo Alto, CA). Preliminary studies established that caveolin-3–enriched vesicles are optimally isolated from adult rat cardiomyocytes using a 5 to 30% continuous sucrose gradient. This was based upon the observation that the floating vesicle fractions isolated by 5 to 30% continuous and 5 to 35% discontinuous sucrose gradient centrifugation contain identical amounts of caveolin-3 immunoreactivity, but fractions 4 and 5 were isolated using the 5 to 35% discontinuous sucrose gradient centrifugation containing 3- to 4-fold more protein (i.e., they are contaminated with a substantial amount of noncaveolin-3-containing protein material). Contaminating proteins are separated into the heavier fractions 7 to 9 using a 5 to 30% continuous sucrose gradient centrifugation scheme. Although the recovery of caveolin-3 immunoreactivity and total protein from neonatal rat cardiomyocytes was similar using either discontinuous 5 to 35% or continuous 5 to 30% sucrose gradient centrifugation schemes, the continuous sucrose gradient centrifugation purification scheme (required to optimize purification of caveolin-3-containing membrane from adult cardiomyocytes) was used in studies on neonatal cardiomyocytes to maintain a consistent method throughout the study and facilitate comparisons. After centrifugation, thirteen 1-ml fractions were collected and aliquots were subjected to SDS-PAGE and immunoblotting.

Immunoblot Analysis. Samples were separated by SDS-PAGE (10% acrylamide) and transferred to nitrocellulose for immunoblot analysis with antibodies for ₁-ARs, ₂-ARs, G_s, G_i1/2, G_o/i3, Type V/VI AC, caveolin-3, and clathrin according to methods described in detail previously (Rybin et al., 2000). The specificity of immunoblot analysis with each of these antibodies was established in the previous study (Rybin et al., 2000). Anti-ryanodine receptor monoclonal antibodies were from Affinity BioReagents, Inc. (Goldne, CO). In some experiments, samples were treated with PNGF, to deglycosylate -ARs, according to methods published previously (Rybin et al., 2000). Bound primary antibodies were visualized with enhanced chemiluminescence according to manufacturer's instructions, with each in each figure representing results from a single gel exposed for a uniform duration.

Immunoprecipitation. Cardiomyocytes from four 100-mm diameter dishes (4 mg of protein) were rinsed with ice-cold phosphate-buffered saline and harvested by addition of 3.6 ml of extraction buffer (50 mM Tris-Cl, pH 8, 150 mM NaCl, 50 µg/ml aprotinin, 50 µg/ml leupeptin, 50 µg/ml benzamidin, 2 mM phenylmethylsulfonyl fluoride, 5 µM pepstatin A, and 1% Igepal) followed by centrifugation at 4°C for 15 min at 10,000g. For immunoprecipitation, supernatants were incubated with 15 µg of anti-caveolin-3 antibodies for 1 h at 4°C followed by addition of 150 µl of Protein A/G PLUS-Agarose beads (Santa Cruz Biotechnology, Santa Cruz, CA) and incubation overnight at 4°C. The beads were washed three times with extraction buffer and bound proteins were eluted with SDS sample buffer and boiled for 5 min. Samples were subjected to SDS-PAGE and immunoblotting with caveolin-3- and G_s-specific antibodies.

Measurements of Intracellular Calcium and Contraction. Simultaneous measurements of cytosolic calcium and cell motion in fura-2–loaded cardiomyocytes were performed as published previously (Kuznetsov et al., 1995). In brief, myocytes were loaded with the acetoxymethyl ester form of fura-2 by incubation with 3 µM fura-2 AM and 1.5 µl of 25% (w/w in dimethyl sulfoxide) Pluronic F-127 (BASF Wyandotte Corp., Wyandotte, MI) dissolved in 1.0 ml of Tyrode's solution for 20 min at 37°C. Cytosolic free calcium ion concentration is reported as the fura-2 fluorescence ratio, which was measured with a device (Photon Technologies, Inc., Princeton, NJ) that alternately illuminates the cells with 340- and 380-nm light while measuring emission at 520 nm (100-Hz sampling rate). To monitor cell motion, cells were simultaneously illuminated with red light and a dichroic mirror (630-nm cut-off) in the emission path deflected the cell image to a video optical system (Crescent Electronics) which tracked motion of a high contrast microsphere attached to the myocyte surface (for neonatal cardiomyocytes) or the motion of the cell edges (for adult cardiomyocytes) along raster line segments of the image during electrically stimulated contractions. The analog voltage output from the motion detector was calibrated to convert to microns of motion. The motion signal was obtained at a rate of 60 Hz and reflected the motion of the same myocyte simultaneously monitored with fura-2 for calcium. The signal was digitized and stored along with the fluorescence data. For measurements of calcium and cell motion, six successive transients were superimposed and averaged. The amplitude of the calcium transient is defined as the difference between the peak systolic and diastolic fura-2 fluorescence rations. The amplitude of the twitch is defined as the difference in cell length before electrical stimulation and at peak contraction.

	Results

Top Abstract Materials and Methods Results Discussion References

 
Development Is Associated with Changes in Cardiomyocyte G Protein Subunit Expression. Initial studies explored whether developmental changes in Gi expression could contribute to the maturation of ₂-AR responsiveness in cardiomyocytes. G protein subunit expression was compared in neonatal and adult ventricular myocardium as well as in cardiomyocytes isolated from neonatal and adult ventricles [using the preparations typically used by us and others for signaling studies (Xiao and Lakatta, 1993; Kuznetsov et al., 1995; Chesley et al., 2000), because contaminating cell types in intact ventricular myocardium could contribute to G subunit measurements and themselves may be subject to important regulatory control during development. The studies focused on G_i2 and G_i3, the PTX-sensitive G_i subunits expressed by cardiomyocytes, using a selective anti-G_i2 antibody and an antibody that recognizes G_i3 but also crossreacts with G_o. Figure 1 shows that G_i2 and G_i3/o are readily detected in neonatal cardiomyocyte cultures; both PTX-sensitive G subunits are detected at substantially lower levels in acutely isolated adult cardiomyocytes preparations. The developmental decline in G_i2 immunoreactivity is also identified in intact myocardial tissue. In contrast, levels of G_i3/o immunoreactivity are similar in intact neonatal and adult ventricular preparations. The discrepancy between the relative abundance of G_i3/o in tissue and cell preparations is probably attributable to the presence of G_o-enriched neuronal contaminants in tissue samples and/or the induction of G_o in neonatal ventricular myocytes during culture (Foster and Robishaw, 1991).

View larger version (41K): [in this window] [in a new window]   Fig. 1. G subunit expression in neonatal and adult cardiomyocytes and ventricles. Particulate fractions from neonatal (N) cardiomyocyte cultures, isolated adult (A) cardiomyocytes, neonatal ventricular myocardium, and adult ventricular myocardium (25 µg each) were subjected to immunoblot analysis with anti-Gs, anti-Gi2, and anti-Gi3/o antibodies. Arrowheads denote the long and short splice variants of Gs, Gi2 and Gi3/o. In this particular experiment, the levels of Gi2 are decreased by 42% (whereas Gi3 is decreased only by 11%) in the adult, compared with the neonatal ventricular myocardial tissue. However, Gi2 and Gi3/o levels are 91 and 87% lower in isolated adult cardiomyocytes, compared with neonatal cardiomyocytes. Similar results were obtained in two separate experiments performed on separate preparations.

Gs exists as short (G_sS) and long (G_sL) splice variants; G_sL contains an additional 15 amino acid insert in the first of two linker sequences connecting the Ras-like and -helical domains [a region close to two of the three "switch" regions in the Gs nucleotide binding pocket (Seifert et al., 1998)] (Walseth et al., 1989). G_sL expression is prominent in neonatal cardiomyocytes, whereas Gs is expressed largely as G_sS in adult cardiomyocytes; little age-dependent difference in total Gs expression is detected. Hence, the developmental change in G subunit expression (in intact tissue and isolated cardiomyocytes) is predicted to result in an age-dependent decrease in Gi function.

PTX-Sensitive G Proteins Do Not Modulate ₂-AR Responses in Neonatal Cardiomyocytes. The marked developmental change in G_i expression in rat ventricular myocytes provides a convenient assay system to examine the relationship between G_i expression and G_i-dependent ₂-AR signaling. To address this issue, we examined the functional activity of G_i, and its role in ₂-AR responses, in neonatal rat cardiomyocytes.

Muscarinic cholinergic receptor (mAChR)-dependent inhibition of isoproterenol-activated cAMP accumulation was used as an assay to compare G_i function in neonatal and adult cardiomyocytes. Table 1 shows that -AR activation with isoproterenol elicits a similar large increase in cAMP formation in neonatal and adult rat cardiomyocytes. In each case, the stimulatory -AR pathway is subject to inhibitory modulation by mAChRs; mAChR inhibition is mediated via Gi and it is inhibited by PTX. These results establish the functional competence of G_i subunits in neonatal and adult cardiomyocytes as well as the efficacy of the PTX pretreatment protocol used to ADP-ribosylate/inactivate G_i subunits. The adequacy of Gi subunit ADP-ribosylation also was validated directly using a biochemical approach that relies upon the inverse relationship between the extent of ADP-ribosylation during the PTX pretreatment in vivo and the amount of [³²P]ADP-ribose that subsequently can be incorporated into control and PTX-pretreated cell membranes in an in vitro ADP-ribosylation reaction. Figure 2 shows that PTX pretreatment results in complete ADP-ribosylation/inactivation of G_i subunits in neonatal and adult cardiomyocytes.

View this table: [in this window] [in a new window]   TABLE 1 Gi-mediated carbachol-dependent inhibition of isoproterenol-stimulated cAMP accumulation in neonatal and adult cardiomyocytes

View larger version (27K): [in this window] [in a new window]   Fig. 2. PTX-dependent ADP-ribosylation of G subunits in neonatal and adult cardiomyocytes. Membranes from neonatal and adult cardiomyocytes pretreated for 24 h at 37°C with vehicle or PTX (100 ng/ml for the neonate and 1 µg/ml for the adult) were subjected to ADP-ribosylation.

To explore the role of G_i in the regulation of ₂-AR signaling to AC, it was necessary to first resolve the ₁- and ₂-AR pathways leading to cAMP accumulation. Figure 3A shows that isoproterenol (nonselective -AR agonist) induces a robust increase in cAMP accumulation by activating both ₁- and ₂-ARs in neonatal rat cardiomyocytes. At low concentrations, isoproterenol primarily occupies the more abundant ₁-AR subtype [shown in a previous study, using radio-ligand binding techniques, to comprise 84% of total -ARs (Kuznetsov et al., 1995)]; cAMP accumulation in response to 10^–9 M isoproterenol is completely blocked by the ₁-AR inhibitor CGP20712A. As the concentration of isoproterenol increases, a ₂-AR-component becomes more prominent; at 10^–7 M [a concentration that maximally activates cAMP accumulation (Kuznetsov et al., 1995)], the effect of isoproterenol to increase cAMP levels is partially blocked by either CGP20712A or ICI 118,551 (₁- and ₂-AR antagonists, respectively); 10^–7 M isoproterenol induces only a trivial increase in cAMP accumulation in the combined presence of CGP20712A and ICI 118,551 (3.2% of the response to 10^–7 M isoproterenol without -AR inhibitors, n = 5, p > 0.5). The ₂-AR-dependent pathway for cAMP accumulation in neonatal cardiomyocytes also is exposed by zinterol (₂-AR-selective agonist). Zinterol promotes a 7-fold increase in cAMP levels by selectively activating ₂-ARs at 10^–7 M (response completely inhibited by ICI 118,551); zinterol's selectivity for ₂-ARs progressively decreases at concentrations exceeding 10^–7 M. Consistent with calculations (based upon published values for the affinity of zinterol at individual -AR subtypes) that predict that 10^–5 M zinterol occupies 100% of the ₂-ARs, but also 90% of the predominant ₁-AR population, approximately one third of the rise in cAMP levels induced by 10^–5 M zinterol is blocked by the ₁-AR antagonist CGP20712A (not by ICI 118,551).

View larger version (30K): [in this window] [in a new window]   Fig. 3. 2-ARs promote cAMP accumulation in neonatal, but not adult, cardiomyocytes; 2-AR signaling to cAMP is not influenced by PTX-sensitive G proteins. A, neonatal or adult cardiomyocytes were challenged for 5 min with the indicated concentrations of isoproterenol or zinterol; cAMP accumulation (in the absence or in the presence of 10–7 M CGP20712A or 10–7 M ICI118,551, each starting 5 min before -AR stimulation) was determined as described under Materials and Methods. Results are expressed as mean ± S.E.M. cAMP accumulation over basal (13.7 ± 2.3 pmol/well or 17.6 ± 0.7 pmol/10,000 cells for neonatal [n = 10] or adult cells [n = 9], respectively). Errors fall within some data points. B, neonatal or adult cardiomyocytes were pretreated for 24 h at 37°C with PTX (100 ng/ml for the neonate and 1 µg/ml for the adult) and then challenged for 5 min with the indicated concentrations of isoproterenol, zinterol, or forskolin before measurements of cAMP accumulation. Results are expressed as mean ± S.E.M. cAMP accumulation over basal cAMP accumulation (which was not altered by the PTX-treatment). Errors fall within some data points; on the scale used for this figure, the results for zinterol alone and zinterol + PTX are superimposed.

To determine whether the ₂-AR pathway leading to cAMP formation in neonatal rat cardiomyocytes is subject to inhibitory modulation by Gi proteins, cAMP was measured in parallel in neonatal cardiomyocyte cultures pretreated with vehicle or PTX. Figure 3B shows that PTX pretreatment neither increases nor decreases cAMP accumulation in response to isoproterenol, zinterol, or forskolin. In the context of control studies that establish functional Gi expression (and complete Gi inactivation by PTX; Fig. 2 and Table 1), these results exclude a significant role for Gi proteins in -AR signaling to cAMP in neonatal cardiomyocytes.

Similar protocols were performed on adult cardiomyocytes. Here, isoproterenol promotes cAMP accumulation exclusively by activating ₁-ARs; the response is completely blocked by CGP20712A (Fig. 3A). cAMP levels also are not appreciably elevated by zinterol (although the absence of substantial promiscuous activation of ₁-ARs by high concentrations of zinterol is puzzling). The failure to detect a ₂-AR-dependent increase in cAMP levels in adult cardiomyocytes is not caused by dual ₂-AR actions via stimulatory Gs and inhibitory Gi proteins, with the stimulatory pathway obscured by a ₂-AR-Gi linkage. PTX pretreatment of adult cardiomyocytes (according to a regimen that reverses carbachol-dependent inhibition of -AR-dependent cAMP formation) has no effect on cAMP accumulation in response to 10^–9 to 10^–8 M isoproterenol (Fig. 3B). Whereas PTX-treated cultures display slightly increased cAMP accumulation in response to 10^–7 M isoproterenol, these actions of isoproterenol remain completely sensitive to inhibition by CGP20712A [not ICI 118,551 (data not shown)] and are accompanied by a significant increase in forskolin-dependent cAMP accumulation; zinterol does not elevate cAMP in PTX-treated adult cardiomyocytes. Collectively, these results suggest that there is a developmental increase in the inhibitory regulation of AC by PTX-sensitive G proteins (which can not be attributed to a developmental change in G_i expression). This could be caused by a developmental increase in the expression of AC type V, which is more susceptible than AC type VI to inhibitory regulation by G_i proteins (Okumura et al., 2002). Importantly, the results also indicate that a putative ₂-AR-Gi linkage does not obscure or attenuate cAMP accumulation through a stimulatory -AR-Gs pathway.

₂-ARs are reported to modulate intracellular calcium and contractile function via Gi proteins in adult cardiomyocytes, but the contribution of Gi proteins to inotropic support by ₂-ARs in neonatal cardiomyocytes has not been examined. Figure 4 shows typical records from control and PTX-treated cardiomyocytes electrically stimulated at 1 Hz during continuous monitoring of intracellular calcium and cell motion transients. In each case, superfusion with 10^–7 M zinterol [a concentration previously identified as selective for ₂-ARs (Kuznetsov et al., 1995)] increases the amplitude and accelerates the kinetics of relaxation of the calcium transient and the contraction. The averaged results from a series of control and PTX-treated cells studied according to this protocol failed to reveal any difference between cells that had or had not been pretreated with PTX (Table 2). ₂-AR stimulation with isoproterenol in the presence of CGP20712A (in the ₂-mode) also is not influenced by PTX (data not shown). Collectively, these studies are most consistent with a Giindependent mechanism for inotropic support by ₂-ARs in neonatal cardiomyocytes.

View larger version (19K): [in this window] [in a new window]   Fig. 4. 2-AR modulation of calcium transients and cell shortening is not influenced by PTX in neonatal rat cardiomyocytes. Representative tracings from cardiomyocytes cultured with vehicle or PTX (100 ng/ml) for 24 h and then electrically driven at 1 Hz during challenge with 10–7 M zinterol. Cardiomyocyte shortening was recorded as micrometers of motion of a glass microsphere on the cell surface. Because the motion of only one portion of the neonatal myocytes is monitored (rather than total cell length), the position of the microsphere before electrical stimulation is set to zero (diastole) and motion relative to the diastolic position is reported. Signal averaged motion transients during the control interval and at the peak of the response to zinterol (which was at 3 min in both vehicle and PTX-treated cardiomyocytes) are superimposed for comparison.

View this table: [in this window] [in a new window]   TABLE 2 2-AR-dependent modulation of calcium transients and cell shortening is not influenced by PTX in neonatal cardiomyocytes

Components of the ₂-AR Signaling Complex Are Organized in Caveolae/Lipid Rafts in Neonatal and Adult Cardiomyocytes. Previous studies established that ₂-ARs cofractionate with AC in caveolin-3-enriched vesicles isolated from neonatal cardiomyocytes. Neonatal cardiomyocytes represent a preparation in which ₂-ARs promote cAMP accumulation. We reasoned that the colocalization of ₂-ARs and AC in caveolae might be required to `launch' ₂-AR signaling to cAMP. According to this formulation, the failure of ₂-ARs to promote a detectable increase in cAMP accumulation in adult cardiomyocytes could result from a change in the localization of either ₂-ARs or the AC enzyme, restricting interactions between these two molecules. To compare -AR and AC enzyme partitioning between caveolae and other cellular membranes in neonatal and adult rat ventricular preparations, caveolae were isolated by a protocol that relies upon their resistance to solubilization in detergent-free alkaline sodium carbonate buffer at low temperatures and buoyancy in sucrose density gradients. Figure 5A shows a representative example of this fractionation scheme applied to neonatal and adult rat cardiomyocytes. In each case, the light sucrose gradient fractions contain a minor fraction of total cell protein (no detectable contamination with clathrin or ryanodine receptor immunoreactivity; Fig. 5B), but the bulk of cellular caveolin-3 immunoreactivity. Over a series of experiments, protein recovery in the light vesicle fractions tended to be somewhat higher in neonatal cultures than in their adult counterparts. This result is consistent with a somewhat higher level of caveolin-3 expression in neonatal cultures than in isolated adult cardiomyocytes (Rybin et al., 2003), because caveolin-3 drives caveolae biogenesis (Smart et al., 1999; Galbiati et al., 2001).

View larger version (38K): [in this window] [in a new window]   Fig. 5. 2-ARs, G protein subunit, and AC partitioning between caveolin-3–enriched vesicles and heavy sucrose fractions in neonatal and adult cardiomyocytes. Neonatal or adult cardiomyocytes were homogenized in sodium carbonate buffer and fractionated by relative buoyancy in sucrose gradients as described under Materials and Methods. Fractions were collected from the top of the gradient and analyzed by SDS-PAGE and immunoblot analysis with anti-AC V/VI, anti-2-AR, anti-Gs, anti-Gi2, and anti-caveolin-3 antibodies (A). A separate profile from adult rat cardiomyocytes was used to document that ryanodine receptors and clathrin are not recovered in the caveolin-3-enriched fractions (B). Protein recovery in gradient fractions is reported as a percentage of total protein. Fractions 1 to 3 and 4 to 6 (in A, or fractions 1 to 3 and 4 to 5 in B) in profiles from adult cardiomyocytes were pooled, because of the limiting amounts of protein recovered in these fractions. Results are representative of data from three separate experiments.

Immunoblot analysis reveals an unanticipated difference in the mobility of ₂-ARs (but not such proteins as G subunits or caveolin-3, which are not exposed on the extracellular surface) in neonatal cardiomyocyte cultures versus isolated adult cardiomyocytes (Fig. 5A). The antibody to the ₂-AR intracellular C terminus detects a broad 66-kDa, epitope-specific immunoreactive species in caveolae fractions from neonatal cardiomyocyte cultures; this diffuse immunoreactivity is blocked by competing antigen peptide [whereas the sharp band that comigrates with ₂-ARs and persists in the presence of competing peptide is nonspecific (Fig. 6)]. The diffuse appearance of ₂-ARs can be attributed to glycosylation, because ₂-ARs migrate as distinct smaller species after treatment with PNGF (Fig. 6). In contrast, ₂-ARs are recovered from acutely isolated adult rat cardiomyocyte caveolae as smaller 55- and 40-kDa bands, whose migration is not altered by PNGF treatment; both of these diffuse bands are epitope-specific [i.e., blocked when immunoblot analysis is performed with competing antigen peptide (Fig. 6, right)]. Figure 6 shows that ₂-ARs are recovered from caveolin-enriched fractions isolated from intact adult ventricular myocardial tissue as diffuse 66-kDa (PNGF-sensitive) species, similar to the pattern observed in neonatal cardiomyocyte cultures; the more rapidly migrating species are not detected. These results suggest that the enzyme-based procedure used to isolate adult cardiomyocytes from the intact ventricle results in limited proteolysis of ₂-ARs (with cleavage of the N-terminal glycosylation sites). Although it is possible that the impairment in ₂-AR signaling to cAMP (identified in this and previous studies) represents an artifact, resulting from limited proteolysis of ₂-ARs during the enzyme-based cell isolation procedure, this is considered unlikely for three reasons:

View larger version (45K): [in this window] [in a new window]   Fig. 6. Immunoblot analysis of -ARs in neonatal and adult cardiomyocyte and cardiac tissue preparations. Membranes from neonatal cardiomyocyte cultures (lanes 1, 2, and 7), acutely isolated adult cardiomyocytes (lanes 3, 4, and 8), and adult ventricular myocardial tissue (5, 6, and 9) were subjected to immunoblot analysis with antibodies that recognize the C terminus of the 2- and 1-ARs (without and with preblocking with antigen peptide to distinguish -ARs from nonspecific immunoreactivity). Samples were treated with PNGF to deglycosylate the receptor as indicated.

1. The cleaved ₂-ARs can be activated by zinterol; it provides effective inotropic support. Figure 7 replicates results previously published demonstrating that ₂-AR activation with zinterol leads to an increase in the amplitude of the calcium transient and the twitch in adult rat ventricular myocytes (Kuznetsov et al., 1995; Laflamme and Becker, 1998).
   
2. Limited proteolytic cleavage is not limited to ₂-ARs; Fig. 6 shows that ₁-ARs, which show no obvious defect in signaling to cAMP, also are detected as truncated proteins in isolated adult cardiomyocytes, but not in intact adult ventricular myocardium. These results indicate that proteolysis does not necessarily lead to a defect in coupling to AC and the generation of cAMP and that the failure to observe an increase in cAMP should not be dismissed as an artifact of ₂-AR receptor cleavage during cell isolation.
   
3. Our previous studies identify ₂-ARs as full-length glycosylated proteins—which do not detectably increase cAMP accumulation—in embryonic mouse cardiomyocyte cultures (Sabri et al., 2000). Although preparative cleavage could induce subtle changes in -AR responsiveness [perhaps by impairing higher-order events such as receptor oligomerization, which has been attributed to glycosylation and/or structural determinants in the extracellular domain of GPCRs (Rios et al., 2001; He et al., 2002)] that could represent somewhat of a liability for functional studies, the isolated cardiomyocyte preparation (freed of ₂-AR-containing cardiac fibroblasts) remains the preferred preparation to resolve catecholamine actions at cardiomyocyte ₂-ARs.
   

View larger version (18K): [in this window] [in a new window]   Fig. 7. 2-ARs provide inotropic support in cardiomyocytes isolated from the adult ventricle. Representative tracings demonstrating that superfusion with zinterol (10–5 M, in the presence of 10–7 M CGP20712A) increases the amplitude of the calcium transient and twitch in adult rat cardiomyocytes.

In both neonatal and adult cardiomyocyte preparations, ₂-ARs are confined to caveolae fractions; ₂-ARs are not detected in heavy fractions, even with 10-fold greater protein loading and long exposures of the gel. AC V/VI is recovered from light vesicular and (to a lesser extent) heavy sucrose fractions in the neonate but is much more confined to the light vesicular fractions in the adult (Fig. 5A). Hence, in both neonatal cardiomyocyte cultures (in which ₂-ARs promote cAMP accumulation) and adult cardiomyocytes (in which ₂-ARs do not promote a global increase in cAMP levels) ₂-ARs copurify with the bulk of the AC enzyme in caveolin-3–enriched light vesicles.

Because ₂-ARs and AC colocalize to caveolin-3–enriched light vesicles in neonatal and adult cardiomyocytes, subcellular compartmentalization of the G protein subunits that couple receptors to the regulation of AC activity was considered. As reported previously, PTX-sensitive G_i2 subunits are highly localized to the caveolae fractions of neonatal cardiomyocytes (Fig. 5A; Rybin et al., 1999). G_i2 localizes similarly to caveolin-3-enriched light vesicles isolated from adult cardiomyocytes. In contrast, the two molecular species of G_s partition differently across the gradient. In neonatal cardiomyocytes that express G_sS and G_sL, G_sL is recovered in both caveolae and heavy sucrose fractions, whereas G_sS is much more localized to caveolae. In adult cardiomyocytes, Gs is expressed largely as the short splice variant and is recovered largely in light vesicular fractions. These results suggest that the developmental shift from G_sL to G_sS, in the context of the difference in G_sL versus G_sS partitioning across the gradient, contribute to the striking age-dependent difference in the subcellular G_s targeting.

A mechanism for differences in G_sS versus G_sL targeting is not obvious. Because G_s subunits are variably reported to interact with caveolin proteins (Huang et al., 1997; Oh and Schnitzer, 2001), we tested the hypothesis that caveolin-3 discriminates G_sS versus G_sL (perhaps explaining differences in G_s splice variant targeting as well as inconsistencies regarding caveolin-3-G_s interactions in the literature). Figure 8 shows that extracts prepared with an Igepal-containing buffer from neonatal rat cardiomyocytes contain all of the cellular caveolin-3 and a substantial fraction of total cellular G_sS and G_sL. Immunoprecipitation of caveolin-3 from these extracts (according to conditions that essentially clear caveolin-3 immunoreactivity from the Igepal extract) leads to the coimmunoprecipitation of a minor portion of total G_sL but not G_sS. These results indicate that the distinct G_s splice variants differ in their interactions with caveolin-3 (which could potentially explain discrepancies in previous literature on this issue), but these immunoprecipitation experiments do not provide an explanation for the apparent preferential targeting of G_sS to caveolin-enriched vesicles.

View larger version (52K): [in this window] [in a new window]   Fig. 8. Immunoprecipitation of caveolin-3 and Gs. Cardiomyocytes were lysed and subjected to immunoprecipitation with anti-caveolin-3 IgG. The starting lysate, Igepal extract before and after immunoprecipitation (IP) with anti-caveolin-3 and the anti-caveolin-3 immunoprecipitate was probed by Western analysis with anti-Gs and anti-caveolin-3 IgGs. Under conditions that clear caveolin-3 from the Igepal-extract, we detect coimmunoprecipitation of GsL but not GsS. The larger Gs immunoreactive species also coimmunoprecipitates with caveolin-3; the significance of this observation is uncertain. The result was replicated in three separate experiments.

	Discussion

Top Abstract Materials and Methods Results Discussion References

 
The mechanisms whereby ₂-ARs provide inotropic support and influence the evolution of cardiac hypertrophy and failure remain incompletely understood. This study expands upon current models of ₂-AR signaling to show that ₂-ARs promote cAMP accumulation, modulate calcium cycling, and enhance contractile function in a completely PTX-sensitive, Gi protein-independent manner in neonatal cardiomyocytes. ₂-ARs also increase the amplitude of contraction but without a detectable increase in cAMP levels in adult rat cardiomyocytes. The failure to detect a ₂-AR–dependent increase in cAMP in adult rat cardiomyocytes can not be attributed to spatial segregation of ₂-ARs, G_s/G_i subunits, and AC to distinct membrane subdomains; ₂-ARs, G subunits, and AC all copurify with buoyant caveolin-3–enriched membranes isolated from neonatal and adult rat cardiomyocytes. Dual coupling of ₂-ARs to Gs and Gi also does not provide an obvious mechanism to preclude ₂-AR signaling to AC, because ₂-ARs do not promote cAMP accumulation in PTX-pretreated adult rat cardiomyocytes. These later results are noteworthy, because there is still only limited information on the nature of the Gi-dependent pathway(s) emanating from ₂-ARs in adult rat cardiomyocytes. Because Gi proteins typically link GPCRs to inhibition of AC, some investigators have inferred that ₂-ARs exert dual regulation on AC activity through reciprocal actions mediated by Gs and Gi proteins. However, these studies provide clear evidence that G_i subunits do not mitigate cAMP formation by either ₁- or ₂-ARs. The mechanism whereby ₂-ARs provide inotropic support in adult rat cardiomyocytes, without detectably elevating cAMP levels, remains uncertain. Our previous study identified intracellular alkalinization (which would enhance myofibrillar calcium sensitivity) as a cAMP-independent inotropic mechanism for ₂-ARs in adult cardiomyocytes (Jiang and Steinberg, 1997). However, an additional contribution from a cAMP signal that is too small (or too highly localized to the plasma membrane) to be detected by a typical cAMP radioimmunoassay that measures a rise in cAMP in the bulk cytosol cannot be excluded. Future studies with newer technologies that resolve localized changes in cAMP levels in subcellular microdomains in cardiomyocytes should be revealing (Zaccolo and Pozzan, 2002).

This study demonstrates that the standard enzyme-based procedure typically used to isolate cardiomyocytes from intact rat ventricles results in limited proteolysis of -ARs. This has not been appreciated previously but might have been predicted based upon previous results suggesting that proteolysis of the extracellular tethered-ligand domain of protease-activated receptor-1 (PAR-1, the receptor for thrombin) during enzyme-based cell disaggregation renders adult rat cardiomyocytes relatively refractory to thrombin's actions (Jiang et al., 1998). Proteolytic cleavage of GPCRs is probably an intrinsic (and heretofore unnoticed) feature of standard adult rat cardiomyocyte preparations used for biochemical and electrophysiologic studies. However, the significance of limited -AR proteolysis during cell isolation is uncertain, because truncated ₁-ARs promote a robust increase in cAMP accumulation and truncated ₂-ARs effectively provide inotropic support. Nevertheless, recent evidence that higher-order events such as ₁-AR receptor oligomerization may be influenced by N-terminal glycosylation events suggests that limited -AR proteolysis could introduce subtle changes in the regulation of -AR responsiveness that deserve further study (He et al., 2002).

Recent efforts to decipher ₂-AR subtype function have relied heavily on the ₂-AR-selective agonist zinterol. This study presents evidence that zinterol's actions are faithfully mimicked by isoproterenol plus CGP (the more traditional -AR agonist acting in the ₂-mode), largely eliminating significant concerns that zinterol exerts unusual properties as an agonist. However, the pharmacologic studies with -AR subtype selective antagonists emphasize that zinterol displays only a modest level of selectivity for ₂-ARs over ₁-ARs. In neonatal cardiomyocytes, where ₂-AR responses are elicited by 10^–7 M zinterol (a relatively low concentration that retains full ₂-AR selectivity), the use of zinterol without a ₁-AR antagonist is justified. However, for reasons that are still obscure, ₂-AR-dependent responses are optimally detected only at high zinterol concentrations in the adult cardiomyocytes [2 log orders higher than in the neonate (Steinberg, 1999; Xiao et al., 1999)]. The evidence that zinterol's ₂-AR-selectivity declines rapidly at concentrations >10^–7 M suggests caution in interpreting previous studies that used 10^–5 M zinterol (without a ₁-AR inhibitor) to selectively activate ₂-ARs in adult rat cardiomyocytes.

This study identifies a developmental switch in G_s subunits expression from G_sL to G_sS. The functional significance of the distinct molecular forms of G_s has never been adequately explained. Previous efforts to distinguish the signaling properties of G_sS and G_sL have relied largely on in vitro assays, where G_sS reportedly shows a somewhat increased affinity for GDP [and activates AC somewhat more effectively than G_sL (Seifert et al., 1998)]. However, this study shows that G_sL and G_sS differ markedly in their subcellular targeting; this property would be pertinent to studies in intact cells, but probably would be missed in studies that rely on in vitro assays to discriminate G_s splice variant function. A mechanism that would permit G_sL to distribute widely (in caveolae as well as other membrane domains) but restrict G_sS to caveolae is not obvious. Although G_s splice variants differ in their propensity to form complexes with caveolin-3, G_sL (rather than G_sS, which is preferentially recovered in caveolin-3–enriched membranes) coimmunoprecipitates with caveolin-3. G_s subunits anchor to plasma membranes (and partitioning to caveolae/lipid rafts) because of covalent lipid modifications and stable interactions with dimers (Moffett et al., 2000; Zaccolo and Pozzan, 2002). However, there is no reason to expect differences in N-terminal palmitoylation of G_sL versus G_s. Rather, distinct interactions with dimers (with distinct subcellular distributions) are more likely, particularly as dramatic age-dependent differences in dimer expression have been reported (Hansen et al., 1995).

On the basis of studies that implicate PTX-sensitive G proteins in the regulation of cardiac ₂-AR responsiveness, current models generally have assumed that altered G_i expression (as occurs in heart failure) would functionally affect ₂-AR signaling (Xiao, 2001). However, the relationship between G_i expression and ₂-AR signaling via G_i proteins has not been subject to formal scrutiny in previous studies. This study identifies a developmental decline in PTX-sensitive G protein subunit expression in rat ventricular myocytes [similar to the results previously reported for rat and rabbit ventricle (Luetje et al., 1988; Kumar et al., 1994; Bartel et al., 1996)], identifying rat ventricles as a convenient assay system to test the hypothesis that changes in G_i expression translate directly into changes in G_i-dependent ₂-AR signaling. The observation that ₂-AR responses are not modulated by PTX-sensitive G proteins in neonatal rat cardiomyocytes that express ample levels of functional G_i proteins deserves emphasis. These results suggest that the intuitive assumption that up-regulation of G_i proteins in human heart failure syndromes might be sufficient to shift the balance of ₂-AR signaling toward Gi pathways [and thereby constitute a cardioprotective mechanism (Xiao et al., 1999)] deserves more direct and careful scrutiny. The results also emphasize that even sophisticated models of signal transduction mechanisms that allow for localized differences in signaling molecule stoichiometry (imposed by structures such as caveolae/lipid rafts) may be inadequate to fully explain the complexity of signaling networks in differentiated cells such as cardiomyocytes.

	Footnotes

 
This work was supported by National Heart, Lung, and Blood Institute grant HL28958.

ABBREVIATIONS: AR, adrenergic receptor; AC, adenylyl cyclase; PTX, pertussis toxin; PAGE, polyacrylamide gel electrophoresis; mAChR, muscarinic acetylcholine receptor; CGP20712A, [2-(3-carbamoyl-4-hydroxyphenoxy)-ethylamino]-3-[4-(1-methyl-4-trifluormethyl-2-imidazolyl)-phenoxy]-2-propanolmethanesulfonate; ICI 118,551, erythro-(±)-3-isopropylamino-1-(7-methylindan-4-yloxy)butan-2-ol; PNGF, peptide-N-glycosidase F.

Address correspondence to: Dr. Susan F. Steinberg, Department of Pharmacology, College of Physicians and Surgeons, Columbia University, 630 West 168th Street, New York, NY 10032. E-mail: sfs1{at}columbia.edu

	References

Top Abstract Materials and Methods Results Discussion References

 
Bartel S, Karczewski P, and Krause EG (1996) G Proteins, adenylyl cyclase and related phosphoproteins in the developing rat heart. Mol Cell Biochem 163–164: 31–38.

Chesley A, Lundberg MS, Asai T, Xiao RP, Ohtani S, Lakatta EG, and Crow MT (2000) The ₂-adrenergic receptor delivers an anti-apoptotic signal to cardiac myocyte through G_i-dependent coupling to phosphatidylinositol 3'-kinase. Circ Res 87: 1172–1179.[Abstract/Free Full Text]

Foster KA and Robishaw JD (1991) Effect of calcium and cAMP on Go expression in neonatal rat cardiac myocytes. Am J Physiol 261(4 Suppl): 15–20.

Galbiati F, Razani B, and Lisanti MP (2001) Emerging themes in lipid rafts and caveolae. Cell 106: 403–411.[CrossRef][Medline]

Hansen CA, Schroering AG, and Robishaw JD (1995) Subunit expression of signal transducing G proteins in cardiac tissue: implications for phospholipase C- regulation. J Mol Cell Cardiol 27: 471–484.[Medline]

He J, Xu J, Castleberry AM, Lau AG, and Hall RA (2002) Glycosylation of ₁-adrenergic receptors regulates receptor surface expression and dimerization. Biochem Biophys Res Commun 297: 565–572.[CrossRef][Medline]

Huang C, Hepler JR, Chen LT, Gilman AG, Anderson RGW, and Mumby SM (1997) Organization of G proteins and adenylyl cyclase at the plasma membrane. Mol Biol Cell 8: 2365–2378.[Abstract/Free Full Text]

Jiang T, Danilo P, and Steinberg SF (1998) The thrombin receptor elevates intracellular calcium in adult rat ventricular myocytes. J Mol Cell Cardiol 30: 2193–2199.[CrossRef][Medline]

Jiang T and Steinberg SF (1997) ₂-Adrenergic receptors enhance contractility by stimulating HCO₃^–-dependent intracellular alkalinization. Am J Physiol 273: H1044–H1047.

Kilts JD, Gerhardt MA, Richardson MD, Sreeram G, Mackensen GB, Grocott HP, White WD, Davis RD, Newman MF, Reves JG, et al. (2000) ₂-Adrenergic and several other G protein-coupled receptors in human atrial membranes activate both G_s and G_i. Circ Res 87: 705–709.[Abstract/Free Full Text]

Kumar R, Joyner RW, Hartzell HC, Ellingsen D, Rishi F, Eaton DC, Lu C, and Akita T (1994) Postnatal changes in the G-proteins, cyclic nucleotides and adenylyl cyclase activity in rabbit heart cells. J Mol Cell Cardiol 26: 1537–1550.[CrossRef][Medline]

Kuschel M, Zhou YY, Cheng H, Zhang SJ, Chen Y, Lakatta EG, and Xiao RP (1999) G_i Protein-mediated functional compartmentalization of cardiac ₂-adrenergic signaling. J Biol Chem 274: 22048–22052.[Abstract/Free Full Text]

Kuznetsov V, Pak E, Robinson RB, and Steinberg SF (1995) ₂-adrenergic receptor actions in neonatal and adult rat ventricular myocytes. Circ Res 76: 40–52.[Abstract/Free Full Text]

Laflamme MA and Becker PL (1998) Do ₂-Adrenergic receptors modulate Ca²⁺ in adult rat ventricular myocytes? Am J Physiol 274: H1308–H1314.

Luetje CW, Tietje KM, Christian JL, and Nathanson NM (1988) differential tissue expression and development of guanine nucleotide binding regulatory proteins and their messenger RNAs in rat heart. J Biol Chem 263: 13357–13365.[Abstract/Free Full Text]

Moffett S, Brown DA, and Linder ME (2000) Lipid-dependent targeting of G proteins into rafts. J Biol Chem 275: 2191–2198.[Abstract/Free Full Text]

Oh P and Schnitzer JE (2001) Segregation of heterotrimeric G proteins in cell surface microdomains. G_q Binds caveolin to concentrate in caveolae, whereas G_i and G_s target lipid rafts by default. Mol Biol Cell 12: 685–698.[Abstract/Free Full Text]

Okumura S, Kawabe J, Takagi G, Lee M, Hong C, Liu J, Honda R, Takagi I, Yatani A, Vatner DE, et al. (2002) -Adrenergic, muscarinic and calcium-Mediated regulation of the heart; effects of target disruption of type 5 adenylyl cyclase gene (Abstract). Circulation 106: 334.

Ostrom RS, Gregorian C, Drenan RM, Xiang Y, Regan JW, and Insel PA (2001) Receptor number and caveolar co-localization determine receptor coupling efficiency to adenylyl cyclase. J Biol Chem 276: 42063–42069.[Abstract/Free Full Text]

Ostrom RS, Liu X, Head BP, Gregorian C, Seasholtz TM, and Insel PA (2002) Localization of adenylyl cyclase isoforms and G protein-coupled receptors in vascular smooth muscle cells: expression in caveolin-rich and noncaveolin domains. Mol Pharmacol 62: 983–992.[Abstract/Free Full Text]

Rios CD, Jordan BA, Gomes I, and Devi LA (2001) G-protein-coupled receptor dimerization: modulation of receptor function. Pharmacol Ther 92: 71–87.[CrossRef][Medline]

Rybin VO, Grabham PW, Elouardighi H, and Steinberg SF (2003) Caveolae-associated proteins in cardiomyocytes: caveolin-2 expression and interactions with caveolin-3. Am J Physiol, in press.

Rybin VO, Xu X, Lisanti MP, and Steinberg SF (2000) Differential targeting of -adrenergic receptor subtypes and adenylyl cyclase to cardiomyocyte caveolae: a mechanism to functionally regulate the cAMP signaling pathway. J Biol Chem 275: 41447–41457.[Abstract/Free Full Text]

Rybin VO, Xu X, and Steinberg SF (1999) Activated protein kinase C isoforms target to cardiomyocyte caveolae: stimulation of local protein phosphorylation. Circ Res 84: 980–988.[Abstract/Free Full Text]

Sabri A, Pak E, Alcott SA, Wilson BA, and Steinberg SF (2000) Coupling function of endogenous ₁- and -adrenergic receptors in mouse cardiomyocytes. Circ Res 86: 1047–1053.[Abstract/Free Full Text]

Seifert R, Wenzel-Seifert K, Lee TW, Gether U, Sanders-Bush E, and Kobilka B (1998) Different effects of G_s splice variants on ₂-adrenoreceptor-mediated signaling: the ₂-adrenoreceptor coupled to the long splice variant of G_s has properties of a constitutively active receptor. J Biol Chem 273: 5109–5116.[Abstract/Free Full Text]

Smart EJ, Graf GA, McNiven MA, Sessa WC, Engelman JA, Scherer PE, Okamoto T, and Lisanti MP (1999) Caveolins, liquid-ordered domains and signal transduction. Mol Cell Biol 19: 7289–7304.[Free Full Text]

Steinberg SF (1999) The molecular basis for distinct -adrenergic receptor subtype actions in cardiomyocytes. Circ Res 85: 1101–1111.