INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

IN RESPONSE TO CHRONIC HYPOXIA, a general increase in adrenergic drive occurs that is reflected, in part, in elevated plasma and urine catecholamine concentrations (8, 22). That process may be considered as homeostatic because it produces subsequent cardiovascular adaptations to offset a global decrease in tissue oxygen supply. The myocardial alterations of adrenergic neurotransmission due to chronic hypoxia are complex and somewhat paradoxical because some findings suggest an increased neuronal activity (32), whereas others indicate a postsynaptic -adrenergic desensitization (39). Many studies have demonstrated that chronic hypoxia induces a -adrenergic receptor downregulation with a progressive decline in cell surface -receptors and a decline in adenylate cyclase activity in response to agonists (15, 39). Limited information is available on the intracellular mechanism of this myocardial -adrenergic desensitization, but changes in G-protein regulatory subunits may be hypothesized (36, 38).

The myocardial -adrenergic desensitization due to hypoxia indicates a chronic elevation in norepinephrine (NE) concentration in the synaptic cleft (18, 31). This elevation in NE concentration may be a result of elevated plasma NE concentration (8), increased NE production and release, or decreased neuronal NE reuptake function. Richalet et al. (32) found in humans a decrease in myocardial ¹²³I-labeled metaiodobenzylguanidine (MIBG) uptake assessed scintigraphically, after an 8-day stay at high altitude (4,350 m), that was reversible after 6-8 wk, suggesting that hypoxia might alter adrenergic neurotransmitter reuptake function. More than 80% of the NE released in the synaptic cleft is taken back in neurons by the uptake-1 transport mechanism (2). An impairment of this transport may be hypothesized in the mechanism of -receptor desensitization. The neuronal NE reuptake process of released NE relies on an NE transporter, the uptake-1 carrier protein, which is energically driven by sodium and chloride ion cotransport (14, 35). No information is available on the potential influence of exposure to hypoxia on uptake-1 carrier protein density.

The present study was performed to evaluate the hypothesis that a 5-day exposure to hypoxia may induce -receptor densensitization as a consequence of an altered neuronal NE due to an impairment in the uptake-1 transport mechanism.

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

All animal procedures were in accordance with the recommendation of the European Economic Community (86/609/CEE) and the French National Committee (87/848) for the care and use of laboratory animals.

In 32 male Wistar rats kept in a hypobaric pressure chamber (PO₂ = 40 Torr, atmospheric pressure = 450 Torr) for 5 days, uptake 1, the NE reuptake function during hypoxia, was assessed by using the cardiac fixation of intravenously injected [³H]NE, and uptake-1 carrier protein density was assessed by using the in vitro binding technique with [³H]mazindol. The postsynaptic modifications were evaluated in myocardial membrane preparations by the density and affinity of cardiac -receptors and by adenylate cyclase activity in either the basal state or after stimulation by isoproterenol, sodium fluoride, forskolin, 5'-guanylylimododiphosphate [Gpp(NH)p], and MnCl₂. The levels of myocardial inhibitory and stimulatory G-protein -subunits (G_i and G_s, respectively) were also determined by immunochemical analysis.

Tissue procurement. Male Wistar rats (200-300 g) were kept for 5 days in a hypobaric pressure chamber at a simulated altitude of 4,350 m (atmospheric pressure = 85 mmHg, barometric pressure = 450 Torr, leading to a predicted alveolar PO₂ of 51 Torr). Control animals were maintained at the altitude of the laboratory (300 m). Both groups of animals received food and water ad libitum and were subjected to a 12:12-h dark-light cycle. After 5 days, the experiments were conducted in 32 rats (16 hypoxic animals and 16 normoxic animals). The first series of rats was removed (n = 6) from the chamber, injected in the tail vein with 17 µCi of [³H]NE, and replaced in the chamber. Control animals (n = 6) also received 17 µCi of [³H]NE. All rats were unanesthetized when receiving the injections. Animals were killed by guillotine 10 min later, an arterial blood sample was drawn, and heart, lungs, liver, and muscles (soleus and extensor digitorum longus) were excised. Right and left ventricles were separated, and all samples were weighed and immersed for one night at 55°C in a liquid solution (Tissue Solubilizer-450, Beckman, Fullerton, CA) to dissolve tissues. After dissolution of tissues, a beta-scintillation cocktail (Ready solv MP, Beckman) was added. The [³H]NE activity was measured in a beta-counter (80% efficiency, Packard SL 2000, Zurich, Switzerland). Tissue concentrations are expressed in percentage of kilogram dose per gram to normalize the differences in the body weights of the rats (27).

s

i

Plasma catecholamine concentration. Plasma NE and epinephrine levels were measured by HPLC adapted from the 1962 method of Anton and Sayre (1) by using the ESA plasma catecholamine analysis kit (ESA, Bedford, MA).

Uptake-1 carrier protein measurement. The membrane preparation and receptor assay were similar to the method used by Fowler et al. (10). We slightly decreased the intensity and length of the polytron bursts, the original method being too strong for the cardiac samples. The frozen right and left ventricles were homogenized in 10 ml of ice-cold buffer (50 mM Tris · HCl, pH 7.7) with a polytron (PCU Kinematica, Bioblock, Lucerne, Switzerland) at a setting of seven for 5 s and centrifuged at 40,000 g for 15 min. The pellet was resuspended in 50 mM Tris · HCl, 120 mM NaCl, and 5 mM KCl, pH 7.4, to obtain a protein concentration of 0.3 mg/ml. [³H]mazindol was used to bind to the uptake-1 carrier protein (14, 35). The incubation buffer contained 0.1 ml of 50 mM Tris · HCl, 120 mM NaCl, 5 mM KCl, pH 7.4, or 0.1 ml of 100 µM desipramine (uptake-1 inhibitor), 0.1 ml [³H]mazindol (concentration 3-60 nM), and 0.3 ml of membrane preparation. After incubation for 20 min at 4°C, membrane-bound radioactivity was separated from free radioactivity by filtration on Whatman GF/C filters (Millipore, Bedford, MA). Each filter was rinsed with 30 ml of buffer. The filters were counted in a liquid scintillation counter (Packard SL 2000) to determine the amount of bound ³H. Specific binding was defined as maximal binding minus binding in presence of desipramine. The degree of nonspecific binding estimated at a concentration of 30 nM of [³H]mazindol was calculated to be 40 and 45% of the total binding in the right ventricle and left ventricle, respectively. The maximum number of binding sites and [³H]mazindol equilibrium dissociation constant (K_d) were calculated from Scatchard plots (34) of the binding data. Calculated variables as well as parameter estimates from radioligand-binding data are given as means ± SD of n experiments.

Myocardial -adrenergic receptor measurement. The ventricles were thawed, placed in 10 ml of ice-cold buffer (20 mM Tris, 1 mM EDTA, pH 8), and minced finely with scissors. After the preparation was homogenized with a polytron with one 5-s burst at a setting of seven, 1 ml of ice-cold 2.5 M KCl was added to extract contractile proteins, followed by stirring at 4°C for 15 min. The suspension was then centrifuged at 40,000 g for 20 min at 4°C. The pellet was resuspended in 20 mM Tris-1 mM EDTA ice-cold buffer with a dounce (homogenizer) to achieve a protein concentration of 0.3 mg/ml. Protein measurements were made by Lowry's method (21) adapted by Hartree (11), using bovine serum albumin as standard. Two radioligands were used to identify whole-population and externalized -adrenergic receptors: the lipophilic ligand, ¹²⁵I-labeled iodocyanopindolol (ICYP) (4, 15, 39), and the hydrophilic ligand, [³H]CGP-12177 (12, 37). The incubation buffer contained 0.1 ml of 50 mM Tris · HCl, 10 mM MgCl₂, pH 7.4, or 0.1 ml of 10 µM L-propranolol (nonspecific binding), 0.1 ml [¹²⁵I]ICYP (concentration 15-150 pM), or 0.1 ml [³H]CGP-12177 (0.1-8 nM), and 0.3 ml of membrane preparation. The assay was incubated for 60 min at 37°C and filtered on Whatman GF/C filters. Each filter was rinsed with 30 ml of buffer to remove free [¹²⁵I]ICYP or [³H]CGP-12177. The filters were counted in a liquid scintillation counter to determine the amount of [¹²⁵I]ICYP or [³H]CGP-12177. Specific binding of [¹²⁵I]ICYP or [³H]CGP-12177 was defined as the amount of [¹²⁵I]ICYP or [³H]CGP-12177 in the absence of competing ligand minus the amount of [¹²⁵I]ICYP or [³H]CGP-12177 in the presence of 10 µM L-propranolol. The degree of nonspecific binding estimated at a concentration of 100 pM of [¹²⁵I]ICYP was 15 and 10% of the total binding in the right ventricle and left ventricle, respectively. The degree of nonspecific binding estimated at a concentration of 4 nM of [³H]CGP-12177 was 20% of the total binding in both ventricles. The maximum number of specific binding sites and [¹²⁵I]ICYP or [³H]CGP-12177 K_d was calculated from Scatchard plots (34).

Competition binding experiments. Determination of ₁- and ₂-receptor subtypes was obtained through competition experiments. A fixed concentration of [³H]CGP-12177 (2 nM) was added to 0.3 ml of membrane samples (0.3 mg/ml) in the presence of increasing concentrations (10² to 10¹⁴ M) of displacing drug. Practolol, a ₁-adrenoceptor-selective antagonist, was used (7). After incubation for 60 min at 37°C, the separation of free and bound ligand was carried out by filtration through Whatman GF/C filters under vacuum. The filters were washed twice with 10 ml of buffer and then placed in vials. The radioactivity retained on the filters was counted in a liquid scintillation counter. Analyses of competition curves were performed on a computer system by using the specialized nonlinear regression program Ligand (26).

Adenylate cyclase assay. Adenylate cyclase activity was measured in cardiac homogenate membranes according to the method described by Salomon (33). After incubation with [³²P]ATP-Mg²⁺, the level of [³²P]cAMP produced was measured. Briefly, the frozen ventricles were dissected and placed in 10 ml of ice-cold buffer (50 mM Tris, 10 mM MgCl₂, pH 7.4). After the preparation was homogenized in a polytron with one 5-s burst, at a setting of seven, the preparation was centrifuged at 40,000 g for 20 min at 4°C. The pellet was resuspended in the same buffer with a dounce to obtain a protein concentration of 2.5-5 mg/ml (11, 21). The reaction was carried out in an assay medium (final volume of 60 µl) containing 50 mM Tris · HCl, 5.0 mM MgCl₂, 1.0 mM 3-isobutyl-1-methylxanthine (phosphodiesterase inhibitor), 1.0 mM EDTA, an ATP-regenerating system, 1 mg/ml creatine kinase and 1 mM phosphocreatine, 1 mM [³²P]ATP (10⁶ counts/min), 1 mM cAMP to inhibit endogenous phosphodiesterase activity, and ~10⁴ counts/min of tritiated cAMP were also added to calculate the efficiency of the alumina column (column recovery ranged from 70 to 90%). The incubation was started by addition of membrane fractions (50-100 µg protein). After a 10-min incubation period in a shaking water bath at 37°C, the reaction was terminated by adding 200 µl of 0.5 M HCl followed by immediate boiling for 7 min. The pH of the assay mixture was adjusted to 7.6 with 250 µl of 1.5 M imidazole. Samples were then eluted with 3 ml of 10 mM imidazole-HCl (pH = 7.6) through an alumina column that retains [³²P]ATP. The ³H and ³²P activities of the eluate were then counted after addition of a scintillation cocktail in a beta-scintillation counter. Adenylyl cyclase activity was determined with and without addition of stimulant. The adenylyl cyclase activity was stimulated by using different drugs: 5 mM sodium, 0.1 mM L-isoproterenol + 0.1 mM GTP, 100 µM forskolin, and 0.1 mM or 5 mM Gpp(NH)p. To assess Mn²⁺ stimulation, a special reaction buffer was used and consisted of the buffer compositions described previously without MgCl₂. All determinations were performed in triplicate.

Quantitation of G_s and G_i by immunochemical analysis. SDS-PAGE was performed according to Laemmli and Faure (17) on 12% acrylamide slab gels (0.1% bisacrylamide). Membranes were prepared for SDS-PAGE by suspension in sample buffer (0.1 M Tris, 8 M urea, 1% SDS, 10% -mercaptoethanol, pH 6.8) and boiled for 5 min. The protein content of each sample was determined in triplicate according to Lowry's (21) method. Equal amounts of total protein (100 µg) were submitted to SDS-PAGE. For immunoblotting, the proteins were transferred from gel to nitrocellulose (Millipore) by using a Tris-glycine buffer (25 mM Tris, 114 mM glycine, pH 8.3). The transfer was carried out at a constant voltage of 50 V for 1 h at 4°C in a Bio-Rad transblot electrophoretic transfer cell. Proteins on the blot were visualized with Ponceau red (0.2% in 3% TCA, Serva). Nonspecific binding was blocked by incubating the blot in 3% BSA in Tris-buffered saline with Triton X-100 (TBST; 50 mM Tris, pH 7.4, 155 mM NaCl, 0.5% vol/vol Triton X-100) for 60 min. The blots were rinsed with TBST and then incubated with the primary antibody (AS/7 or RM/1) in 1% BSA in TBST (0.5% sodium azide, NaN₃). The antiserum AS/7 specifically labels the isoforms G_i1 and G_s2. G_i2 is the cardiac-specific isoform. After washing, the antibody was visualized by using a biotinylated secondary antibody (Extravidin staining kit). Peroxidase activity was visualized by using 3,3'-diaminobenzidine as a chromogen. Color development was stopped after 10 min by rinsing with water, and nitrocellulose was dried between two sheets of Whatman paper. All steps of the immunodetection procedure were carried out at room temperature with agitation, except for incubation with the primary antibody, which was performed overnight at 4°C. G_i antiserum (AS/7) and G_s antiserum (RM/1) were used at a 1:1,000 dilution. Amounts of G_i in the 40-kDa band and amounts of G_s in the 45-kDa bands were analyzed by computer-assisted densitometry and image analysis (transmittance/reflectance, GS 300, Hoefer, San Francisco, CA) determined for each band, an integrated optical density. The background was estimated in the protein-free region nearest to the band, and then the background was substracted from the band of interest.

Chemicals. [³H]NE, [³²P]ATP-Mg²⁺, and tritiated cAMP were obtained from Amersham (Buckinghamshire, UK). [³H]mazindol, [¹²⁵I]ICYP, [³H]CGP-12177, and antibodies AS/7 and RM/1 were obtained from New England Nuclear (Wilmington, DE). Desipramine, BSA, propranolol, practolol, 3-isobutyl-1-methylxanthine, creatine kinase, phosphocreatine, cAMP, NaF, isoproterenol, GTP, forskolin, Gpp(NH)p, and Extravidin staining kit were obtained from Sigma Chemical (St. Louis, MO). EDTA, HCl, imidazole, and MnCl₂ were obtained from Merck (Darmstadt, Germany).

Data analysis. Results are expressed as means ± SD, except when otherwise indicated. Differences between the hypoxic and the normoxic rat groups were determined by using two-way analysis of variance and unpaired Student's t-test. Correlation coefficients, assuming a linear regression, were calculated for paired variables. A P value <0.05 was considered statistically significant.

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Body weight and right and left ventricle weights of euthanized animals. Data are shown in Table 1. Animals in the normoxic group (n = 8) weighed 258 ± 25 g, and animals in the hypoxic group (n = 8) weighed 244 ± 29 g [P = not significant (NS)]. The left ventricle-to-body weight ratio (2.14 ± 0.07 mg/g for normoxic rats vs. 2.18 ± 0.16 mg/g for hypoxic rats, P = NS) as well as the right ventricle-to-body weight ratio (0.84 ± 0.06 mg/g for normoxic rats vs. 0.77 ± 0.07 mg/g for hypoxic rats, P = NS) were not statistically different between the two groups.

Plasma catecholamine concentration. Plasma NE levels were higher in the hypoxic group than in the normoxic group (2.1 ± 0.7 vs. 0.6 ± 0.2 ng/ml, respectively, P < 0.05). No difference has been found in the plasma epinephrine levels between the hypoxic group and the normoxic group (2.2 ± 0.7 vs. 1.8 ± 0.3 ng/ml, respectively).

Myocardial tritiated NE fixation. A significant decrease in [³H]NE myocardial uptake was found in hypoxic rats. A 33% decrease in [³H]NE uptake was found (P < 0.01) in the whole heart. A decrease in [³H]NE uptake was found in both ventricles: 32% (P < 0.05) in the right ventricle and 35% (P < 0.01) in the left ventricle. No significant difference in [³H]NE fixation was found in other organs when hypoxic rats were compared with normoxic rats. Results are listed in Table 2.

Uptake-1 carrier protein measurement. Figure 1 depicts a typical example of ventricular binding with [³H]mazindol by use of tissue from a hypoxic rat. A 38% decrease in uptake-1 carrier protein density was found in the right ventricle (290 ± 18 vs. 470 ± 47 fmol/mg protein, P < 0.01), whereas a 41% decrease was found in the left ventricle (308 ± 75 vs. 523 ± 48 fmol/mg protein, P < 0.01). K_d of [³H]mazindol in hypoxic rats was not significantly different from normoxic rats (right ventricle: 11.8 ± 2.6 vs. 14.8 ± 3.6 nM, respectively, P = NS; and left ventricle: 16.5 ± 6.3 vs. 16.7 ± 5.6 nM, respectively, P = NS).

View larger version (17K): [in this window] [in a new window]   Fig. 1.   Effects of 5 days of hypoxia on norepinephrine (NE) uptake-1 carrier protein density (A) and dissociation constant (B) in left and right ventricles measured with [3H]mazindol-binding techniques. Density of uptake-1 carrier protein and dissociation constant were determined by Scatchard analysis in normoxic (n = 7) and hypoxic (n = 6) rats. * P < 0.01.

Density and distribution of -receptor. When [¹²⁵I]ICYP is used as a ligand, myocardial -receptor density decreased by 15% (P < 0.05) in the right ventricle and by 8% (P = NS) in the left ventricle. No significant difference was found in K_d of receptors for [¹²⁵I]IYCP between normoxic and hypoxic rats. When [³H]CGP-12177 is used as a ligand, a 40% decrease in myocardial -receptor density was found in the right ventricle (P < 0.01), whereas a 32% decrease was found in the left ventricle (P < 0.01). No significant difference was found in the affinity of [³H]CGP-12177 for -receptor between normoxic and hypoxic rats. Data are shown in Table 3.

View this table: [in this window] [in a new window]   Table 3.   Effects of 5-day hypoxia on -receptor density

Relationship between -receptor density and NE uptake-1 carrier density. In hypoxic rats, -receptor density measured with [³H]CGP-12177-binding technique correlated with NE uptake-1 carrier density measured with [³H]mazindol in right and left ventricles (r = 0.820, P = 0.04 and r = 0.870, P = 0.023, respectively; see Fig. 2).

View larger version (21K): [in this window] [in a new window]   Fig. 2.   Correlation between -adrenergic receptor density ([3H]CGP-12177) and mazindol-binding sites (NE uptake carrier density) on right (A) and left (B) ventricles of hypoxic rats (n = 6).

-Adrenergic receptor subtypes. Figure 3 (top) depicts a typical example of a competition experiment using [³H]CGP-12177 with the ₁-selective antagonist (practolol). Figure 3, A and B, shows the respective proportions of ₁- and ₂-adrenoceptors evaluated in right and left ventricles. In normoxic rats, the right ventricle showed a density in ₁- vs. ₂-adrenoceptors of 23 ± 1.9 vs. 2 ± 0.2 fmol/mg protein, respectively. In the left ventricle, the density of ₁-adrenoceptors was 30.7 ± 2.7 fmol/mg protein and the density of ₂-adrenoceptors was 2.3 ± 0.2 fmol/mg protein. In hypoxic rats, the right ventricle showed a density of ₁- vs. ₂-adrenoceptors of 11.4 ± 1.2 vs. 3.6 ± 0.4 fmol/mg protein, respectively. In the left ventricle, the density of ₁-adrenoceptors was 16.5 ± 1.8 fmol/mg protein and the density of ₂-adrenoceptors was 5.5 ± 0.6 fmol/mg protein.

View larger version (23K): [in this window] [in a new window]   Fig. 3.   Top: representative 2-site displacement curves. Left and right ventricular membranes from 5-day hypoxic rats (n = 5) and control rats (n = 5) were incubated with various concentrations of 1-receptor-selective-antagonist practolol in presence of 2 nM of [3H]CGP-12177. Bottom: proportion of 1- and 2-receptors in membranes prepared from right (A) and left (B) ventricular myocardium of 5-day hypoxic rats (n = 5) and control rats (n = 5). * P < 0.05.

Adenylate cyclase activity. Basal and stimulated adenylate cyclase activity were decreased in both ventricles of hypoxic rats. After a 5-day hypoxia exposure, basal adenylate cyclase activity was decreased by 25% in the right ventricle (P < 0.05), and stimulated adenylate cyclase activity was reduced by 36% in the presence of isoproterenol, by 40% in the presence of Gpp(NH)p, and by 39% in the presence of forskolin (P < 0.01 for the 3 stimuli, respectively). The left ventricle showed a 24% reduction in basal adenylate cyclase activity (P < 0.05), a 41% decrease in isoproterenol-stimulated adenylate cyclase activity, a 42% decrease in Gpp(NH)p-stimulated adenylate cyclase activity, and a 41% decrease in forskolin-stimulated adenylate cyclase activity (P < 0.01 for the 3 stimuli, respectively). In all experiments, NaF- and MnCl₂-sensitive adenylate cyclase activity was not significantly altered. These data are shown in Table 4.

Quantitation of G_i and G_s. Densitometric quantitation of the Western blot showed an increase in the cardiac-specific isoform G_i2 by 46% in hypoxic rats compared with normoxic rats (29 ± 6.5 vs. 16 ± 2.8 integrated optical density, respectively; P < 0.05; n = 3). No significant difference was found in G_s level when hypoxic rats were compared with normoxic rats.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

The present data show that, in rats, a 5-day period of exposure to hypoxia leads to statistically significant changes in pre- and postsynaptic components of myocardial adrenergic function. The study shows for the first time in a model of hypoxia a decrease in the myocardial uptake-1 carrier sites. Moreover, this decrease in uptake-1 carrier sites appeared to be related to a reduction in -adrenergic receptor density.

NE uptake-1 function. The decrease in [³H]NE cardiac uptake found in rats is consistent with the decreased cardiac [¹²³I]MIBG uptake found by Richalet (32) by using scintigraphy in humans. The present study shows by the use of in vitro binding assays with [³H]mazindol that hypoxia induces a loss in uptake-1 carrier protein. These abnormalities of NE reuptake function have a physiological importance because this function is the principal means of terminating the action of the neurotransmitter. Indeed, in animals, blockade of neuronal NE uptake increases postsynaptic exposure to NE, and cardiac response to exogenous NE is prolonged by the uptake-1 (neuronal) blockade but is unchanged by uptake-2 (extraneuronal) blockade (19, 24). The cause of impaired myocardial uptake-1 is unclear. The uptake-1 mechanism is Na⁺-K⁺-pump-dependent. A specific decrease in energy supply in this active mechanism in response to the oxygen decrease in a working muscle may perturb this myocardial mechanism. On the other hand, a recent study in dogs chronically infused with NE has suggested that elevated interstitial NE concentrations per se could lead, by an unknown mechanism, to a decrease in NE reuptake function (13). A depressed NE uptake was found associated with lesions in the adrenergic nerve terminal as assessed by catecholaminergic histofluorescence and tyrosine hydroxylase-immunostained profiles; these alterations were similar to that found in dogs with heart failure related to rapid ventricular pacing (13). Finally, in rats infused with NE for 5 days, a decrease in left ventricular tritiated NE uptake was related to a loss of uptake-1 carrier as assessed by tritiated mazindol in vitro binding (23). These data suggest that a "downregulation" process of the uptake-1 carrier protein is similar to that observed for postsynaptic -adrenergic receptors in response to an increased level of NE (8, 6, 9, 20). This downregulation of uptake-1 carrier protein would be a consequence of either increased circulating NE or increased myocardial NE release or both. In the present study, an increased plasma NE concentration was found, but no information was available on myocardial NE release.

-adrenergic receptor downregulation. The present study shows that hypoxia induces a decrease in cell surface -adrenergic receptor density measured with [³H]CGP-12177-binding techniques. This finding is concordant with previous reports. Voelkel et al. (39) found in Wistar rats after 5 wk of hypobaric hypoxia (barometric pressure = 450 Torr) a diminished -receptor density by using [¹²⁵I]ICYP as a ligand associated with a decrease in basal and isoproterenol-stimulated adenylate cyclase activity. In newborn lambs exposed to chronic hypoxia by using a model of cyanotic heart disease, Bernstein et al. (4) reported a 55% decrease in density of left ventricular -adrenergic receptors measured with [¹²⁵I]ICYP in comparison with control lambs. However, Kacimi et al. (15), by using [¹²⁵I]ICYP as a ligand, found no change in left ventricular -receptor density in Wistar rats after 2 wk of hypobaric hypoxia (380 Torr) but a selective downregulation of left ventricular -receptors after 3 wk. These discrepancies with our data may be due, in part, to the different techniques used for measuring -receptor density. A lipophilic radioligand such as ICYP determines total cellular (externalized membrane-bound, internalized membrane-bound, and cytosolic) -receptor-binding sites, and it may not enable the detection of a shift in -receptors between different cellular compartments. Conversely, the hydrophilic ligand [³H]CGP-12177 binds mainly to externalized -receptors, which are thought to be mainly coupled to the adenylate cyclase complex (12, 37). However, in myocardial membrane homogenates, where some of the vesicles are right-side out and some are inside out, -receptors on the inside-out vesicles are hidden from the hydrophilic ligand and do not permit an accurate measurement of cell surface -receptors. Nevertheless, in the present study, the combined use of the two ligands suggests strongly that after a 5-day hypoxic episode, the externalized -receptor fraction was reduced.

Relationship between -adrenoceptor downregulation and decreased uptake-1. In the present study, the downregulation of myocardial -adrenergic receptors was on the same order of magnitude as that of both decreased tritiated NE fixation and decreased number of sites of the uptake-1 carrier protein. Moreover, a statistically significant correlation was found between -receptor downregulation and decrease in uptake-1 carrier protein density. This finding suggests that -adrenergic receptor downregulation may be due to impaired NE uptake function because a decrease in NE reuptake leads to an increase in synaptic cleft NE concentration. Data obtained from other models are not congruent with this hypothesis. In vitro studies using several cell lines and a reconstituted cell membrane system have exhibited that short hypoxic exposure may promote -receptor downregulation in cultured myocytes, without any role of sympathetic stimulation. Krishna et al. (16) demonstrated in cultured neonatal rat ventricular myocytes, by using [³H]CGP-12177-binding techniques, that a 2-h hypoxia induced a downregulation of cell surface -receptors by translocation to a cytosolic fraction. Conversely, by using [¹²⁵I]ICYP-binding techniques, Thandroyen et al. (38) showed that myocytes exposed to 120-150 min of hypoxia were associated with a 64% increase in -receptor density. These discrepancies may be due to differences in models, especially differences in the time course of the exposure to hypoxia, or to differences in binding techniques.

Adrenoceptor signal transduction. Hypoxia may act directly on the signal transduction between -adrenergic receptors and the adenylate cyclase complex. In the present study, an increase in G_i2 protein was found instead of a decrease in G_s, assessed by immunoblotting. The immunoblotting technique only provides a semiquantitative estimation of G-protein concentrations, but this finding is consistent with the absence of a significant decrease in NaF-stimulated adenylate cyclase activity. NaF activates adenylate cyclase by enhancing the interaction of the regulatory proteins with the catalytic unit without dissociating the G protein into its subunits. Inhibition of adenylate cyclase by G_i requires the dissociation of free -subunits from G_i and subsequent coupling of the free -subunits to activated G_s. Therefore, activation of G_s by NaF may be insensitive to changes in G_i protein. The absence of decreased Mn²⁺-stimulated adenylate cyclase activity provides evidence that the catalytic subunit of adenylate cyclase is not modified. To explain the decrease in forskolin-stimulated adenylate cyclase activity, several reports have shown that, although forskolin increases cellular cAMP levels by activating the catalytic subunit of adenylate cyclase, the maximal forskolin effect also requires an interaction with the intact stimulatory regulatory subunit G_s of adenylate (30). The present alteration in the cardiac-specific isoform G_i2 protein as a mechanism of -adrenergic desensitization is in accordance with data published by Reithmann et al. (29), who showed an increase in G_i in rat cardiac cells treated with NE and with those published by Newmann et al (28), who also found such an increase in heart failure. The increase in G_i protein could explain, in part, the -adrenergic receptor desensitization observed in the in vitro hypoxic cell lines in which the sympathetic stimulation has no effect.