BINDING AFFINITY OF BETA-ADRENOCEPTORS IN ATRIA OF STREPTOZOTOCIN-INDUCED DIABETIC AND NORMAL RATS

Diabetes mellitus is associated with a reduced responsiveness to catecholamines. The reduced responsiveness may be attributable to a reduction in beta-adrenoceptor sensitivity to catecholamines. Radioligand binding studies demonstrate that chemically-induced diabetes reduces beta-adrenoceptor number without altering betaadrenoceptor drug binding affinities. The present study re-examines the effect of chronic (10 weeks) streptozotocin-induced diabetes in the rat on beta-adrenoceptor drug binding affinity, using pharmacological techniques, to test the results of the binding studies. The study employed two different methods: partial irreversible receptor blockade and the use of a partial agonist, to determine betaadrenoceptor agonist binding affinity and the method of competitive antagonism to determine beta-adrenoceptor antagonist binding affinity. Diabetes produced no significant differences in the dissociation constants (l/affinity) for isoproterenol or for metaproterenol and no significant differences in the pAz values for timolol maleate which correlates to antagonist binding affinity. Therefore, the study confirms the results of radioligand binding studies by using intact tissue that diabetes does not alter beta-adrenoceptor drug binding affinity in cardiac tissue.

The study employed two different methods: partial irreversible receptor blockade and the use of a partial agonist, to determine betaadrenoceptor agonist binding affinity and the method of competitive antagonism to determine beta-adrenoceptor antagonist binding affinity.
Diabetes produced no significant differences in the dissociation constants (l/affinity) for isoproterenol or for metaproterenol and no significant differences in the pAz values for timolol maleate which correlates to antagonist binding affinity. Therefore, the study confirms the results of radioligand binding studies by using intact tissue that diabetes does not alter beta-adrenoceptor drug binding affinity in cardiac tissue.
ii ACKNOWLEDGEMENTS I wish to acknowledge the Department of Pharmacology and Toxicology for providing a graduate teaching assistantship and the American Heart Association -Rhode Island Affiliate for providing partial funding to carry out this research. The assistance, support, and advice of the faculty, staff, and students of the Department of Pharmacology and Toxicology of the University of Rhode Island is deeply appreciated.
Special gratitude is extended to Dr. Robert L. Rodgers for his advice, counsel, and patience during this research.
iii DEDICATION Special thanks are extended to my parents, Edgar G. and Doris L. McKenna for their support, encouragement, and constant concern throughout my educational training. iv          viii Diabetes mellitus is associated with a higher incidence of morbidity and mortality from cardiac disease than is present in the nondiabetic population (Kannel, 1979). The data suggest that diabetes is another discrete cause of congestive heart failure and that a cardiomyopathy is associated with diabetes, as a result of small vessel disease, metabolic disorders, or both. These disorders produce structural, functional, and biochemical changes in cardiac tissue (Ledet et al., 1979).
Vascular disease associated with diabetes mellitus is well known (Colwell~ al., 1979;Zoneraich ~al., 1980). However, diabetes may produce a cardiopathy which is independent of vascular abnormalities.
In some cases, pathological studies of diabetic human hearts revealed patent and atherosclerosis-free large coronary arteries (Hamby ~al, 1974). Autopsies failed to detect significant obstructive disease of the proximal arteries in some diabetic patients succumbing to heart failure (Regan~ al., 1977). Several experimental studies (Regan~ al., 1974;Miller, 1979;Fein~ al., 1980;Penpargkul et al., 1980;Vadlamudi et al., 1982) demonstrated impaired muscle function in both chronic (greater than two weeks duration) and acute (usually less than one week) chemically-induced diabetic rat hearts. Some of these studies, also, suggest that diabetes decreases diastolic ventricular compliance and the rate of relaxation (Regan :=..t al., 1974;Miller, 1979). Streptozotocin-induced diabetes in the rat produces no significant cardiac macrovascular disease (Chobanian ~al., 1982), but impairs cardiac performance (Vadlamudi et al., 1982). These results suggest that diabetes produces a direct alteration of the rat myocardium.
A possible consequence of diabetic cardiopathy might be the appearance of an ~ltered sensitivity to the positive chronotropic and positive inotropic effects of catecholamines (Cavaliere et al., 1980).
Experimental and clinical evidence suggests that diabetes may also alter autonomic control of the myocardium. Clinically, diabetic patients exhibit a supersensitivity to cholinomimetic agents and to catecholamines; the supersensitivity to the former is more pronounced than to the latter (Lloyd-Mostyn and Watkins, 1975). However, hearts of two week streptozotocin-diabetic rats are subsensitive to both acetylcholine and catecholamines (Foy and Lucas, 1978). Diabetic hearts are less sensitive than non-diabetic hearts to carbachol 100 days after . the induction of diabetes by streptozotocin, but they are supersensitive to carbachol at 180 days (Vadlamudi and McNeill, 1983). The supersensitivity observed both clinically and experimentally may be related to the well known autonomic neuropathy associated with chronic diabetes in humans (Watkins and Edmonds, 1983) and experimentally diabetic rats (Schmidt~ al., 1981). The possible mechanism for diabetes-induced cardiac subsensitivity is less clearly defined.
The effect of diabetes on cardiac responsiveness to catecholamines may vary with the severity of the diabetic state, although the evidence is not at all clear. In humans, tachycardia is often associated with chronic diabetes, primarily due to vagal dysfunction (Lloyd-Mostyn and Watkins, 1975). In the rat, experimental chronic diabetes most often produces bradycardia (Savarese and Berkowitz, 1980), but acute diabetes has no effect on heart rate. A reduction in beta-adrenoceptor sensitivity to catecholamines can explain the lowered heart rate of the diabetic rat. ~-adrenoceptor number or density and drug-binding affinity determine beta-adrenoceptor sensitivity to catecholamines. Radioligand binding techniques provide a method of assessing beta-adrenoceptor density (Bmax) and drug-binding affinity. Diabetes produced a 28% decrease in beta-adrenoceptor density accompanied a 24% in heart rate without any alterations in antagonist (H3-dihydroalprenolol) binding affinity (Savarese and Berkowitz, 1980). Heyliger~ al.(1982), and Ramanadham and Tenner (1983) confirm that diabetes decreased betaadrenoceptor density without altering antagonist affinity in rat hearts. No direct in vitro studies of the positive chronotropic responsiveness of diabetic hearts to catecholamines have been performed. Diabetes had no effect on competitive [3H]-DHA binding curves by isoproterenol in membrane homogenates suggesting that agonist binding affinity is unaltered . However, no studies on beta-adrenoceptor drug-binding affinities of intact tissue have been performed.
Diabetes alters the mechanical function of the heart, as previously indicated, but apparently not the inotropic responsiveness to catecholamines. Ingebretson ~al. (198lb) reported that acute alloxan diabetes had no effect on the inotropic response of isolated rat hearts to isoproterenol. Heyliger ~ al. (1982) report that diabetic papillary muscle exhibited a decreased ability to respond to beta-adrenergic stimulation, based upon the rates of tension development (positive dF/dt) and relaxation (negative dF/dt). In control preparations, isoproterenol produced marked increases in both the positive dF/dt and negative dF/dt, whereas, in diabetic preparations, the positive dF/dt was unresponsive to isoproterenol and the negative dF/dt responded only marginally.
The questions of cardiac beta-adrenoceptor sensitivity in experimental diabetes has not been fully resolved. The depressed responses of the papillary muscle preparations from diabetic rats discussed above may be explained as an alteration in beta-adrenoceptor sensitivity. A reduced beta-adrenoceptor sensitivity can also explain the bradycardia which accompanies experimental diabetes. As previously shown, diabetes reduces beta-adrenoceptor sensitivity by reducing beta-adrenoceptor density. The possibility of reducing betaadrenoceptor sensitivity by the reduction of agonist binding affinity needs to be reexamined. The functional differences between agonist and antagonist binding kinetics (Weiland~ al., 1979;1980) suggest that diabetes may alter agonist binding affinity without altering antagonist binding affinity. Agonist-induced desensitization of beta-adrenoceptors may reduce the apparent agonist binding affinity without altering antagonist binding affinity (Tse~ al., 1978;Hoffman and Lefkowitz, 1980;Harden, 1983). Response is a function of agonist binding, complete assessment of beta-adrenoceptor sensitivity requires characterization of agonist binding affinity (Wessels~ al., 1978).
The original hypothesis of the present study stated that diabetes reduces agonist affinity for beta-adrenoceptors in the rat heart.
The recent observations by ) that eight weeks of streptozotocin-induced diabetes had no effect on beta-adrenoceptor agonist binding lead to a re-evaluation of the hypothesis. The present study reexamines the recent findings that diabetes does not alter beta-adrenoceptor drug-binding affinity using alternative experimental techniques. A decrease in beta-adrenoceptor number plays a major role in the diabetic subsensitivy to catecholamines and can help to explain the bradycardia which accompanies experimental diabetes in the rat; however, the possible role of a reduction in agonist affinity remains unclear. This study provides a systematic determination of betaadrenoceptor agonist and antagonist affinities in normal and experimentally-diabetic rat atria. The use .of isolated atria allows characterization of both the positive ~ronotropic and positive inotropic effects of catecholamines.
Several pharmacological procedures exist for the determination of agonist affinity. The availability of the beta-adrenoceptor antagonist Ro 03-7894 (1-(5-chloroacetylaminobenzfuran-2-yl)-2-isopropylaminoethanol), which acts irreversibly and selectively with beta-adrenoceptors (Nicholson and Broadley, 1978;Rankin and Broadley, 1982) allows pharmacological characterization by the method of partial irreversible receptor blockade (Furchgott and Bursztyn, 1967). The method of partial agonists described by Waud (1969)   Alloxan and streptozotocin are the most extensively used agents for induction of diabetes because the diabetogenic dose is 1/4 to 1/5 times the lethal dose (Grodsky ~al., 1982). The dose varies considerably among species and with the age and metabolic state of the animal. Both alloxan and streptozotocin produce beta-cell necrosis in the rat (Ganda ~al., 1976). Streptozotocin appears to be more selective than alloxan (Rerup, 1970); the possible reason for this might be the high capacity of beta cells to accumulate this agent (Srivasta ~al., 1982). Streptozotocin models are thought to be more relevant to the human diabetic state than alloxan models, due to metabolic profiles, enzyme concentrations, and histopathology. Alloxaninduced diabetes is more ketotic than streptozotocin-induced diabetes (Mansford and Opie, 1968).
Due to its instability, streptozotocin is dissolved in 0.1 M citrate buffer, pH 4.5 just prior to injection. Streptozotocin is optimally stable at around pH 4 (Rerup, 1970) and its biological halflife is about 5 minutes, which necessitates an intravenous injection.
For induction of diabetes, streptozotocin is conventionally administered as a single injection . Maximal elevation of plasma glucose is achieved with 60 mg/kg (Ganda ~al., 1976), but significant increases in plasma glucose occur with 40 mg/kg.

Pharmacological vs. Radioligand Binding Techniques for Assessing Receptor Characteristics
Recent reviews (Furchgott, 1978;Tallarida 1981; compare the strengths and drawbacks of pharmacological and radioligand binding procedures. In summary, both procedures estimate drug-binding affinities for a specific receptor and the rate and extent of receptor inactivation by irreversible antagonists. However, only a radioligand yields an estimate of receptor number or density (Bmax), whereas, a pharmacological procedure permits evaluation of the relative efficacies of agonists acting upon a receptor to produce a response.
Radioligand binding procedures require the demonstration of specific binding to tissue sites and no effect is measured.  (Furchgott, 1972). The pA2 is important in classifying receptors (Kenakin, 1982). The pA2 is defined as the negative logarithm of the molar concentration of an antagonist which reduces the effect of a dose of agonist by half (Tallarida ~al., 1979).   As the antagonist concentration [B] increases, the dose-response curve shifts to the right. The degree of the shift is indicated by the dose ratio (DR) of A'/A. Thus, it takes a higher agonist concentration, A', to produce the same effect in the presence of antagonist than the concentration A, producing the same effect in the absence of antagonist. All curves achieve the same maximum effect, since the antagonism is surmountable and the curves should be parallel due to the competitiveness of the antagonism. The dose ratios (DR), converted to log (DR-1), for each antagonist concentration [B], expressed as -log [B] yield points for a Schild plot.
Based upon the occupation theory (Furchgott, 1972) stating that the effect produced by an agonist depends upon the concentration of the agonist-receptor complex, the following equation can be derived:  Curve II represents the dose-response curve for agonist after washout of the irreversible antagonist. Curve II does not achieve the maximum response of curve I (pre-antagonist). The irreversible antagonist displaces the dose-response curve to the right. Equiactive effects from the linear portion of each line E 1 and E 4 etc., yield pairs of agonist concentrations, (A_ ,A' 1 ), (A 7 , A 1 2 ), etc. which produce each effect. Plotting the reciptccaI valu~s of each pair of agonist concentration, (1/A 1 ,1/A' 1 ), etc. yields a straight-line. From these conditions, the following equation for the line obtained in the double reciprocal plot is derived: ..

Agonist Affinity (l/KA): Method of Partial Agonists
The method of partial agonists (Waud, 1969) makes use of the large cardiac spare receptor capacity (Venter, 1979) and the fact that partial agonists require greater receptor occupancy than full agonists. Irreversible antagonists make full agonists act like partial agonists, and thereby allow estimation of the full agonist's receptor affinity. Since a larger receptor occupancy is required, a partial agonist may or may not elicit the tissue's maximum response.
In this method, one constructs dose-response curve for a full agonist followed by a dose-response curve for the partial agonist, after restoration of resting levels. Figure 4 shows the concentration of each agonist which produce equal responses yield agonist concentration pairs, (A1, P1) ••• (AN, PN) with A1 representing the full agonist concentration and P1 representing the partial agonist concentration from only the linear portions of each line (Thron, 1970).
Plotting the reciprocal values of each agonist pair, (l/A, l/P), yields a straight-line (figure f) having the following equation: where A and P represents the full agonist and partial agonist concentration, respectively. The two terms KA and Kp represent equilibrium dissociation constants for the full and partial agonists. Curve I represents a dose-response curve to a full agonist. Agonist concentrations, A , Ar., etc. produces an effect, E , E etc. Partial agonist concentrations, P 1 , P 2 etc. from curve 11 1 yi~ld the same effects, E 1 , E 2 etc. 'Illus, for egch effect, there exists a pair of agonist concentrations which can produce the same effect.

Figure f: Double Reciprocal Plot of Equipotent Agonist Concentrations Using
Agonists with Different Efficacies 1/P Equiactive agonist concentrations (Ai, Pi) yield reciprocal pairs (l/A 1 , l/Pi) which yield the following line; The quotient obtained by dividing the slope , by the y-intercept yields the partial agonist dissociation constant (KP).

Animals
Male Sprague-Dawley rats (7 weeks old) were obtained from Charles River Breeding Labs (Wilmington, Mass.). Unless otherwise indicated, all animals were supplied with food (Purina Rat Chow) and water ad libitum and housed under identical conditions throughout the 10-week experimental period.

Experimental Grouping
The rats were divided into three main groups, designated control (CT), streptozotocin-diabetic (STZ), and food restricted control (FR) • . .
The CT and STZ groups were divided into eight subgroups (1-8) and the FR group into two subgroups (7-8) of at least 4 rats each. The food restricted rats received two pellets of rat chow daily. Five of the subgroups (1-5) were used for the study of competitive betaadrenoceptor blockade with each subgroup representing a different concentration of timolol maleate. The sixth subgroup (6) was used for the study of beta-adrenoceptor activation by the partial agonist, metaproterenol. The remaining subgroups (7-8) were used to study the effect of the irreversible beta-adrenoceptor antagonist, Ro 03-7894.
The number of rats within each subgroup are shown in Table 1.  This, table shows the grouping and number of rats used in the study of beta-adrenoceptors. CT = Age-matched controls fed ad libitum; STZ = streptozotocin-induced diabetic animals; FR age-matched food restricted controls fed 40 grams of rat chow daily. Subgroups 1-4 represent timolol concentrations used in the study of reversible antagonism. Subgroup S was used as controls for both the reversible antagonism and partial agonist study. The partial agonist, metaproterenol, was used with animals in subgroup 6. The remaining subgroups (7-8) were used for the study of irreversible antagonism, ascorbate control and Ro 03-7894. N .....

Induction of Diabetes
Diabetes was induced in rats with a single intravenous injection of streptozotocin or STZ (40 mg/kg) into a tail vein (Rerup, 1970;Like and Rossini, 1976;Ganda ~al, 1976).
The STZ was prepared in 0.1 M citrate buffer, pH 4.5 (40 mg/ml) immediately before injection.
About 75% of the animals injected with STZ became diabetic and exhibited glycosuria. Animals which were injected with STZ, but did not subsequently become diabetic, were omitted from the study. Agematched controls were injected with the citrate vehicle. Four to seven days after injection, weekly metabolism and blood pressure recordings were initiated. Ten weeks after injection, the animals were sacrificed.

Metabolism Studies
Beginning one week prior to injection, all rats were placed singly into metabolism cages once a week for a period of 24 hours.
The following measurements were recorded: urine output, the extent of glycosuria using enzymatic test strips (Tes-Tape®, Lilly) body weight, and food and water consumption.

Indirect Blood Pressure and Pulse Rate Measurement
Beginning two weeks prior to injection, systolic blood pressure and pulse rates were measured weekly using an indirect tail-cuff method. The measurements were made after warming the rat at 34°C for 20 minutes in a temperature-controlled box. An inflatable cuff was placed around the base and a small bulb was placed on the distal portion of the tail. The bulb was attached to a pneumatic pulse trans-23 ducer (MK III) which was coupled to an electrosphygmograph coupler (Narco 7211) and an E & M type 4 physiograph. Systolic arterial pressure was obtained by inflating the tail cuff at pressures exceeding 180 mm Hg, then noting the point at which the pulsations reappeared during slow pressure reduction. Pulse rates were recorded simultaneously by determining the number of pulses per centimeter at a set paper speed on the physiograph. The mean of at least three measurements was recorded for each animal. The pressure was calibrated at frequent intervals using a mercury column manometer.

Isolated Atria
Ten weeks after injection, the rats were killed by a blow to the head. The chest cavity was opened and the heart was rapidly removed and placed in oxygenated buffer at ·room temperature. A blood sa~ple was taken from the chest cavity for analysis of serum thyroxine and glucose. The blood sample was frozen in liquid nitrogen for later analysis. The left and right atria were surgically removed from the ventricles. The right atria were tied to tissue hangers by cotton thread and the left atria were clamped to stimulating electrodes.
The atria were connected by cotton thread to a tension transducer; Narco type A (0.1 -3 gram sensitivity), and Narco type B (0.1 -10 gram sensitivity) for right and left atria, respectively. The 24 resulting tensions were recorded on an E & M type six physiograph.
Initial diastolic resting tensions of 0.8 and 0.5 grams were applied to the left and the right atria, respectively. Resting tensions were determined from preliminary length-tension determinations. The left atria were driven at 2 HZ with square wave pulses (5 msec) at 1.5 times the threshold voltage by a Narco stimulator and the right atria were allowed to beat spontaneously. Tension (g) and rate (bpm) changes were measured from the left and right atria, respectively.

Serum Analysis
Blood samples were thawed at room temperature and were allowed to clot. The clot was sedimented by centrifugation at 5000 g for 5 minutes at 4°C. The supernatant (serum) was decanted for analysis of thyroxine and glucose. Hypothyroidism often accompanies the diabetic state. An AmerlexN T-4 RIA kit was used to determine serum thyroxine levels; the kit has a total range of 0 to 25 ug thyroxine/100 ml.
Serum was deproteinized with equinormal amounts of barium hydroxide and zinc sulfate solutions prior to glucose determination. The solution fraction was obtained by centrifugation at 5000 g for 10 minutes and used for analysis of glucose. Glucose was determined enzymatically with glucose oxidase and peroxidase (Sigma Kit No. 510). Serum glucose was used to estimate the degree of the diabetic state.

Drug Addition
Left and right atria from non-diabetic (CT) and diabetic (STZ) rats were allowed to stabilize for thirty minutes with frequent buffe! changes. The following protocol of drug administration was employed. Parallel controls were exposed to an equivalent volume of the solution used to dissolve the antagonist, but were otherwise treated iden- Responses were measured as the total rate or contraction frequency (bpm) and total developed tension (g) of the right and left atria, respectively, at each agonist concentration. Possible changes in the sensitivity of right or left atria to the agonist between the second and third curves which might occur in the absence of antagonist drugs were accounted for using parallel control atria which were not treated with the antagonist (Broadley and Nicholson, 1979). Mean responses of untreated atria to each agonist concentration during the generation of the third dose-response curve were expressed as a f raction of the mean responses to the equivalent agonist concentration occurring during the generation of the second dose-response curve.
These fractions were then applied as correction factors for antagonist treated atria. The response to each agonist concentration during the second dose-response curve was multipled by the appropriate correction factor, to yield a corrected dose-response curve. The effect of the antagonist on the third curve was then determined by comparison with the corrected second curve.
The responses were standardized to a percentage maximum response scale. Increases in rate or tension above the resting levels were expressed as a percentage of the maximum increase. This was calculated by dividing the individual increase in rate or · tension by the maximum rate or tension increase and multiplying the resulting quotient by roo%.
Estimations of agonist potency (EC50 values) were calculated as the negative log of the agonist concentration which produced half the maximum response. Arithmetic mean values of the EC50 are not normally distributed (Fleming et al., 1972), but the logarithmic values are. Therefore, pD2 values (-log EC50) were compared by the student's ttest for normal and diabetic atria. The values of slope, y-intercept, and x-intercept (pAz), with 95% confidence limits were calculated (Tallarida, 1979). The pAz values for both non-diabetic and diabetic rats were compared using an unpaired Student's t-test.
Estimation of AgOnist Affinity Using a Partial Beta-Adrenoceptor Agonist: 29 Dissociation constants (Kp) for metaproterenol were calculated for control (CT) and diabetic (STZ) rat atria by the method of partial agonists described by Waud (1969). The atria were exposed to cumulative concentrations of isoproterenol twice with a washing in drug-free buffer after each exposure to yield two dose-response curves. A third dose-response curve was constructed using cumulative concentrations of metaproterenol. The second dose-response curve was corrected as described earlier (see "plotting dose-response curve"). The increases in rate or tension in response to each isoproterenol concentration were plotted as a percentage of the maximum increase. The increases in rate or tension in response to each metaproterenol concentration were plotted as a percentage of the maximum possible increase, which was calculated by subtracting the resting level prior to the third dose-response curve from the corrected second dose-response curve maximum total rate or tension. Because constants for normal, diabetic, and food restricted rat atria were compared using one way analysis of variance (Daniel, 1978).

RESULTS
The Streptozotocin Diabetic Model* Differences between streptozotocin-induced diabetic foodrestricted, and control animals are shown in within three weeks after injection of streptozotocin and these changes persist throughout the ten week experimental period.
* "Diabetes" will be used to describe the diabetic condition induced by streptozotocin injection, with the implicit recognition that chemically induced diabetes may differ in some respects from the true diabetic state. Values represent mean ± 95% confidence intervals. Numbers in parentheses represent the sample size. Asterisk (*) means that the value is significantly different (P (0.05) than the control value; the double asterisks means that the value is significantly different than the food restricted value.
Abbreviations CT = Age-matched controls fed ad libitum; STZ = Streptozotocin-induced diabetes, 10 weeks after the induction of diabetes; FR = Age-matched controls on a food-restricted diet; BW = Body weight (grams; Glu = Serum glucose levels (mg/ml); T-4 = Serum thyroxine levels (µg/dl); Urine =Urine output (ml/g BW); Urine Glu =Urine glucose levels (%); BP = Systolic blood pressure (mm Hg); HR= Heart rate (beats per minute); pD2 =-log EC50 of the molar concentration of isoproterenol producing half the maximal response; RA = Right atria; LA = Left atria. Heart rates (± S.E.M.) taken weekly during the 10 week experimental period from control ( e ) and diabetic ( • ) rats. A typical age-dependent bradycardia developnent occurs in both groups with the diabetic rats showing significant reductions within three weeks after injection of streptozotocin. Asterisks represents significant differences from control at P ~ 0.05. Nunbers in parentheses represent sample size. Time (weeks) after streptozotocin or citrate vehicle lies on the abscissa. Heart rate expressed as beats per minute (bpn) is shown on the ordinate.        (Table 4). Neither diabetes nor food restriction significantly altered the effects of Ro 03-7894 rat atrial dissociation constants for isoproterenol (Table 4). Therefore, neither diabetes nor food restriction alters beta-adrenoceptor agonist binding affinity. The fraction of unoccupied receptors (q) were not significantly different (P > O.OS) between experimental groups, but exhibited a high degree of variability (Table 4). Mean dissociation constants (± S.E.M.) for isoproterenol of cardiac beta-adrenoceptors. No significant differences (P >0.05) exist between dissociation constants from control, food-restricted, and diabetic atria. The fraction of unoccupied receptors (q) within experimental groups are not significantly different (P >0.05). The post-antagonist maximum responses (% max) are significantly (P (0.05) reduced from the maximum possible response. Numbers in parentheses represent sample size.  The mean EC50 values of isoproterenol and metaproterenol for left and right atria are compared in Table 5. The EC50 values of metaproterenol are one hundred-fold higher than those of isoproterenol. Regression analysis of these lines provides values for the slope, yintercept, and correlation coefficient. The quotient of the slope divided by the y-intercept equals the dissociation constant for metaproterenol. Table 6 contains dissociation constants for metaproterenol from control and diabetic atria. No significant differences 57 (P > 0.05) were observed between control and diabetic atria. Diabetes does not alter metaproterenol binding affinity. Mean dissociation constants (± S.E.M.) for metaproterenol binding to beta-adrenoceptors from control and diabetic atria. No significant differences (P >0.05) exist between dissociation constants from right or left atria. Numbers in parentheses denote sample size.

DISCUSSION
The results of the present study did not support the original hypothesis that diabetes reduces beta-adrenoceptor agonist binding affinity in rat atria. These findings are in agreement with radioligand binding studies recently reported by Ingebretson ~al. (1983) and , showing that neither acute nor chronic diabetes affected cardiac beta-adrenoceptor binding curves.
STZ-induced diabetes did not alter the estimated dissociation constants for isoproterenol or metaproterenol in the present study.
The values for the dissociation constants for isoproterenol obtained in the present study agree closely with those reported for guinea pig atria ( Table 7) and those of , using control (51 ± 15 nM), food restricted (70 ± 11 nM), and STZ-induced diabetic (78 ± 24 nM) rat hearts. The value for the dissociation constants for metaproterenol closely correspond with published values using guinea• pig atria and different pharmacological techniques (Table 8).
.60 Kiso values are the means ± S.E. mean calculated through the use of irreversible receptor blockade by Ro 03-7894, (a) 7.6 x lo-4 M, (b) 6.4 x lo-4 M, or (c) 3.24 x lo-4 M with a 3 hour washout period. Numbers in parentheses denote sample size.  Broadley and Williams, 1982, 1983Williams and Broadley, 1983Broadley and Nicholson, 1980 KMeta values are the means ± S.E. mean calculated through the use of irreversible receptor blockade by Ro 03-7894 (7.6 x lo-4 M) with a 3 hour washout period. Asterisks signify the use of functional antagonism with carbachol to calculate the dissociation constant. Numbers in parentheses denote sample size.
Values for dissociation constants have a large variability (Broadley and Nicholson, 1978). This variability reduces the sensitivity of statistical tests in distinguishing differences between dissociation constants. The magnitude of the variability in the present study is similar to those reported in the studies cited in Table   8.
The estimations of the fraction of unoccupied receptors (q) varied, which limits the ability to detect changes in betaadrenoceptor number. There were no significant differences between any of the estimates of the fraction of unoccupied receptors.
Furthermore, no specific trend in the estimates occurs: for right atria, the food restricted group had the lowest estimates (0.02), but for left atria, the control group had the lowest estimates (0.057).
The irreversible antagonist, theoretically, inactivates the same number of beta-adrenoceptors and a reduction in beta-adrenoceptor number would be reflected by lower estimates of the fraction of unoccupied receptors. If diabetes does reduce beta-adrenoceptor number, then the estimations of the fraction of unoccupied receptors in diabetic atria should be lower than those in control atria. However, atrial size can influence the estimation of the fraction of unoccupied beta-adrenoceptors, since a larger tissue will contain greater amounts of beta-adrenoceptors, and atrial size was an uncontrolled variable.
The variations in atrial size limits the usefulness of the estimation of the fraction of unoccupied receptors.
Diabetes does not alter antagonist binding affinity of cardiac beta-adrenoceptors. Several radioligand binding studies (Savarese and Berkowitz, 1980;Ingebretson ~al., 1981;Heyliger~ al., 1982) demonstrate that chronic diabetes had no effect of [H3]-DHA binding affinity in · rat ventricular tissues. The present study confirms by alternative methodology that antagonist binding affinity of cardiac beta-adrenoceptors is not affected by diabetes. The pA2 values for timolol maleate from control and diabetic atria were not significantly different from each other ( Table 3). The estimates of the pA2 values are consistent with those found by other investigators (Dreyer and Offermeier, 1980). The pA2 value equals the negative logarithm of the antagonist's dissociation constant when the slope of the Schild plot is unity (Tallarida ~ al., 1979).
None of the slopes of the Schild plots differed significantly from each other. Thus, the antagonist binding affinities are the · same in control and diabetic atria.
Streptozotocin is void of cardiotonic effects at a tissue and subcellular level (Fein .!:E. al., 1980;1981). However, hypothyroid animals have been shown to contain a decreased number of cardiac beta-adrenoceptors with alterations in antagonist (DHA) binding affinity (Ciaraldi .!:E. al., 1977;Mcconnaughey .!:E. al., 1979;Chang .!:E. al., 1982). Ischac .!:E. al. (1983) found that hypothyroidism had not effect on agonist potency (pAz), produced bradycardia, and had no effect on maximum responses to isoproterenol. Furthermore, in general, hypertension can cause a reduction in beta-adrenoceptor number without altering [H3]-DHA binding affinity (Williams .!:E. al., 1977;Woodcock .!:E. al., 1979). Because diabetes reduces both alpha-and betaadrenoceptor density and hypothyroidism increases alpha-adrenoceptor density, Williams .!:E. al. (1983) discounts hypothyroidism as the primary cause of the reduction in beta-adrenoceptor density seen in diabetes. Furthermore,  found that diabetes had no effect on muscarinic receptor number which suggests that alterations in adrenergic receptors are specific. Fein .!:E. al. (1980) considers it unlikely that hypothyroidism produced the altered mecha-64 nics exhibited by diabetic papillary muscle because of the lack of correlation between the free T4 index and the altered mechanical properties. Bhalla~ al. (1980) found no difference in betaadrenoceptor number between control and spontaneously hypertensive rats, instead they found a reduced affinity of beta-adrenoceptors for isoproterenol. The altered metabolic status of diabetic rats may also produce changes in beta-adrenoceptor sensitivity. Increased plasma lipid content can reduce cardiac beta-adrenoceptor number without affecting antagonist binding affinity (Wince and Rutledge, 1981).
Certain cardiac disease states such as ischaemia (Feuvray ~al., 1979) and heart failure (Bristow~ al., 1982) have been associated with a reduction in beta-adrenoceptor number without alterations in [H3]-DHA binding affinity. It has already been noted that diabetics have a higher incidence of mortality from heart disease than the nondiabetic population (Kannel, 1979). Ischaemia and heart failure are often the end results of heart disease. Thus, diabetes may predispose the heart to congestive heart failure in part by reducing betaadrenoceptor number without altering affinity.
Food restriction does not reproduce the cardiac alterations which occur with diabetes (Fein~ al., 1980;Malhorta ~al., 1981). The present study shows that caloric deprivation induced by restricting food intake did not alter the dissociation constants for isoproterenol  also demonstrated that caloric deprivation did not produce the alterations in betaadrenoceptors that occur in the diabetic state.
In summary, the present study confirms that chronic diabetes does not affect cardiac beta-adrenoceptor binding affinities for agonists or antagonists. Diabetes lowered the basal heart rate, but had no affect on the maximum chronotropic responses to isoproterenol or metaproterenol. In addition, diabetes had no affect on agonist poten-  Mean EC50 values from corrected pre-antogonist dose-response curves to isoproterenol (see "Plotting Dose-Response Curves") and from post-antagonist dose-response curves to isoproterenol for each concentration of timolol maleate (M) for right and left atria. Corrected pre-antogonist and post-antogonist EC50 values yield a dose ratio (DR) for each timolol maleate concentration, which is converted to the log (DR-1) value. The table shows the mean log (DR-1) values (±S.E.M.), which represent the ordinate values for the Schild plot. The values for both control (CT) and diabetic (STZ) atria are shown. Numbers in parentheses denotes the sample size • . .