Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Endogenous adenosine elicits a large variety of effects through interaction with cell-surface adenosine receptors, which are heterogeneous and widely spread throughout the body. This large variety of physiological effects elicited by adenosine provides a potential for therapeutic application of analogs of this purine. Ligands with the appropriate selectivity for A₁, A_2A, A_2B or A₃ adenosine receptors form an interesting class of drugs for use in several metabolic, cardiovascular, central nervous system or immunological disorders. This interest has led to the synthesis of a vast number of A₁ and A_2A selective ligands (see Jacobson et al., 1992 for overview) and recently to the identification of selective agonists (Kim et al., 1994) and antagonists (Van Rhee et al., 1996) for the adenosine A₃ receptor. Despite considerable interest in compounds acting at A_2B receptors no selective ligands for this subtype have been identified so far.

Inherent in the widespread distribution of adenosine receptor subtypes is the difficulty in obtaining desirable drug actions without concomitant side effects. For example, the profound hemodynamic disturbances observed with selective adenosine A₁ and A_2A agonists have limited their use for other therapeutic targets. In this respect, application of agonists with reduced intrinsic activity may be beneficial, because activity of these drugs not only depends on receptor subtypes but on tissue differences as well (Kenakin, 1993). This may result in less pronounced cardiovascular actions and a potential increase in selectivity of action.

Recently, two series of CPA analogs have been synthesized in a search for partial agonists. Deoxyribose analogs of CPA have been developed and tested in vitro and in vivo (Mathôt et al., 1995; Van der Wenden et al., 1995a). Removal of the 2- and 3-hydroxyl group resulted in partial agonism in combination with a dramatic decrease in affinity for the adenosine A₁ receptor. Recently, a series of 8-alkylamino derivatives of CPA has been synthesized (Roelen et al., 1996). These compounds were tested in radioligand binding studies on rat brain and were shown to be selective and to have moderate affinity for adenosine A₁ receptors. The intrinsic activity of these compounds was evaluated in vitro by determination of the ratio between the affinities on rat brain membranes in the presence and absence of 1 mM GTP (the GTP shift). All GTP shifts were lower than the GTP shift of the full agonist CPA, which indicates that these compounds may behave as partial agonists for the adenosine A₁ receptor.

The present study characterized the hemodynamic actions of these 8-alkylamino derivatives of CPA (for structures see fig. 1) in vivo in normotensive rats by use of an integrated pharmacokinetic-pharmacodynamic modeling approach. This approach has been useful in the characterization of partial agonistic properties of the deoxyribose analogs of CPA in vivo (Mathôt et al., 1995). By quantification of the relationship between blood concentrations and HR, estimates of intrinsic activity and potency were obtained in vivo. The concentration-HR relationships of the 8-alkylamino derivatives of CPA were determined in individual rats. The observed in vivo potencies of the compounds were correlated to their A₁ receptor affinity as determined in radioligand binding experiments. Furthermore, the intrinsic activities for the cardiovascular effect were compared with the GTP shift in vitro.

View larger version (16K): [in this window] [in a new window]   Fig. 1.   Chemical structures of the 8-alkylamino derivatives of CPA.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Chemicals. The 8-alkylamino derivatives of N⁶-cyclopentyladenosine (8MCPA, 8ECPA, 8PCPA, 8BCPA, 8CPCPA) were synthesized as described previously (Roelen et al., 1996). 1-Deaza-2-chloro-N⁶-cyclopentyladenosine and 1-deaza-2-chloro-2deoxy-N⁶-cyclohexyladenosine were kindly provided by Dr. G. Cristalli (Camerino, Italy). Ethyl acetate was purchased from Baker Chemicals (Deventer, The Netherlands) and distilled before use. Acetonitrile (HPLC grade) was obtained from Westburg (Leusden, The Netherlands). All other chemicals were of analytical grade (Baker, Deventer, The Netherlands). Water was used from a Milli-Q system (Millipore SA, Molsheim, France). Polyvinylpyrrolidone was from Brocacef, Maarssen, The Netherlands, and heparin was from Hospital Pharmacy, Leiden, The Netherlands.

Animals and surgical procedures. Male, normotensive SPF rats (Wistar WU, Sylvius Breeding Facilities, Leiden, The Netherlands), weighing 200 to 250 g, were used throughout the study. The animals were housed individually in plastic cages with a normal 12-h light-dark cycle. They had free access to laboratory chow (Standard Laboratory Rat, Mouse and Hamster Diets, SMR-A, Hope Farms, Woerden, The Netherlands) and tap water.

Experimental protocol. All experiments were started between 9.00 and 10.00 A.M. to minimize influence of diurnal rhythm in the hemodynamic measurements. Arterial blood pressure was measured from the cannula in the left femoral artery with a miniature strain gauge P10EZ transducer connected to a plastic diaphragm dome (TA1017, Disposable Critiflo Dome) (both Viggo-Spectramed B.V., Bilthoven, The Netherlands). The pressure transducer was connected to a polygraph amplifier console (RMP6018, Nihon Kohden Corporation, Tokyo, Japan). A tachograph, triggered by the blood pressure signal, provided measures for HR. HR, blood pressure and MAP signals were passed through a CED 1401 interface (Cambridge Electronic Design LTD, Cambridge, England) into a 80486 computer and the Spike 2 program (Spike 2 Software, Version 3.1, Cambridge, England) was used for data acquisition and off-line data reduction. During the experiments, the cannula connected to the pressure transducer was flushed continuously with heparinized saline (20 IU/ml) at a flow rate of 500 µl/hour (Syringe infusion pump 22, Harvard apparatus, Plato B.V., Diemen, The Netherlands) to prevent disturbances of the blood pressure by obstruction of the cannula.

Plasma protein binding. The P/B and the free fraction in plasma (f_u) of the compounds were determined in the same group of animals. Total blood concentration was determined in a 20-µl aliquot of the 350 µl of sample. This blood sample was hemolyzed in 400 µl millipore water (0°C). The remaining blood was centrifuged at 4°C to separate the plasma. A sample of 20 µl was retained for analysis and the remaining plasma was subjected to ultrafiltration. Free compound was separated from plasma protein-bound compound by ultrafiltration of the supernatant at 1090 g at 37°C using the Amicon Micropartition System in combination with an YMT ultrafiltration membrane (Amicon Divisions, Danvers, MA). The unbound concentration was determined in 50 µl of the ultrafiltrate. The samples were stored at 20°C until analysis by HPLC.

Drug analysis. The concentrations of the CPA derivatives in blood, plasma and ultrafiltrate were assayed by reversed phase HPLC. For all five CPA derivatives a similar HPLC assay was used with only small adjustments between the compounds. Calibration standards were prepared by addition of aqueous solutions of the compounds to a mixture of 100 µl blood and 400 µl water, resulting in blood concentrations of 0 to 3000 ng/ml (8MCPA, 8ECPA, 8BCPA and 8CPCPA) and 0 to 2500 ng/ml (8PCPA). After addition of 50 µl internal standard (DCCA for 8MCPA and 8ECPA, and 1-deaza-2-chloro-2dCHA for the other compounds, respectively), the blood samples were subjected to liquid-liquid extraction using 5 ml ethyl acetate and shaking on a vortex. After centrifugation for 10 min at 2000 × g the organic layer was transferred to a clean tube and 200 µl water and 50 µl 3 M sodium hydroxide were added. The samples were extracted for the second time and the aqueous layer was removed from underneath. The remaining organic layer was evaporated to dryness under reduced pressure at 40°C. The residue was dissolved in 150 µl water and a volume of 100 µl was injected into the HPLC system.

Data analysis. In individual animals the blood concentration-time profiles of the 8-alkylamino derivatives of CPA were fitted to a polyexponential equation for intravenous infusion (Gibaldi and Perrier, 1982):

(1)

(2)

In these equations C(t) is the concentration at time t, T is the infusion duration and C_i and _i are the coefficients and exponents of the equation, respectively. Various exponential models were investigated and the most suitable model was chosen based on the Akaike information criterion (Yamoaka et al., 1978). The area under the concentration-time curve (AUC), the systemic clearance (Cl), the elimination half-life (t_1/2,n) and the volume of distribution at steady state (V_dss) were calculated by use of standard equations (Gibaldi and Perrier, 1982). The pharmacokinetic functions were fitted to the data with weight 1/y² with the nonlinear least squares regression program Siphar (Simed SA, Creteil, France). In each rat the fitted function of the concentration-time profile was used to calculate the concentrations at the measured effect-time points.

(3)

Statistical analysis. The pharmacokinetic and pharmacodynamic parameter estimates were compared statistically by the parametric one-way ANOVA or a nonparametric Kruskall-Wallis test, if more appropriate. All data are reported as mean ± S.E., unless indicated otherwise.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Pharmacokinetics. The time course of drug concentrations after intravenous infusion of the six adenosine agonists for 15 min is shown in figure 2. The solid lines represent the best fits of the pharmacokinetic model to the pooled data of each treatment group. A biexponential equation was found to best describe the concentration-time profiles of the six compounds.

View larger version (29K): [in this window] [in a new window]   Fig. 2.   Blood concentration versus time profiles of all individual animals given an i.v. infusion (black bar) of 0.2 mg/kg CPA (n = 6), 4.8 mg/kg 8MCPA (n = 6), 4.8 mg/kg 8ECPA (n = 6), 4.8 mg/kg 8PCPA (n = 6), 8.0 mg/kg 8BCPA (n = 6) and 8.0 mg/kg 8CPCPA (n = 5) for 15 min. The solid lines represent the best fits to the data on the basis of averaged estimates of pharmacokinetic parameters. The CPA data are from Mathôt et al. (1994).

Pharmacodynamics. The time profiles of HR and MAP after i.v. administration of the CPA derivatives to conscious normotensive rats are shown in figures 3 and 4. The effects observed for CPA are also included in the figures. The figures depict the average HR and MAP for all animals within the treatment groups. For clarity the S.E. bars were omitted. Administration of the vehicle in conjunction with blood sampling did not affect the HR and blood pressure of the animals (data not shown). CPA caused a very rapid decrease in both HR and MAP. During the infusion this reduction reached a maximum which was maintained for several minutes. The derivatives of CPA produced smaller reductions in the HR and blood pressure. Higher doses of the compounds did not result in larger decreases in MAP and HR (E. A. van Schaick, A. P. IJzerman and M. Danhof, unpublished observations). Furthermore, during the infusion maximal reductions were reached which did not change or augment, whereas blood concentrations were still increasing, which indicated that the maximal effect had been reached in the experiment. All compounds produced smaller reductions in HR and MAP than CPA. The order of magnitude was CPA > 8MCPA > 8ECPA = 8PCPA > 8BCPA = 8CPCPA. Both 8BCPA and 8CPCPA did not seem to produce significant changes in HR and MAP as compared with the vehicle-treated group. The decrease in HR observed during the infusion of 8BCPA was small and within the 10% variation in the base-line levels.

View larger version (24K): [in this window] [in a new window]   Fig. 3.   Averaged HR versus time profiles of all individual animals that had received an i.v. infusion (black bar) of 0.2 mg/kg CPA (), 4.8 mg/kg 8MCPA (), 4.8 mg/kg 8ECPA (), 4.8 mg/kg 8PCPA (), 8.0 mg/kg 8BCPA () or 8.0 mg/kg 8CPCPA (+) for 15 min. The standard errors of the mean have been omitted for clarity.

View larger version (26K): [in this window] [in a new window]   Fig. 4.   Averaged MAP versus time profiles of all individual animals that had received an i.v. infusion (black bar) of 0.2 mg/kg CPA (), 4.8 mg/kg 8MCPA (), 4.8 mg/kg 8ECPA (), 4.8 mg/kg 8PCPA (), 8.0 mg/kg 8BCPA () or 8.0 mg/kg 8CPCPA (+) for 15 min. The standard errors of the mean have been omitted for clarity.

View larger version (26K): [in this window] [in a new window]   Fig. 5.   Averaged concentration-HR relationships of all individual animals that had received an intravenous infusion of 0.2 mg/kg CPA, 4.8 mg/kg 8MCPA, 4.8 mg/kg 8ECPA, 4.8 mg/kg 8PCPA, 8.0 mg/kg 8BCPA or 8.0 mg/kg 8CPCPA for 15 min. The concentrations are based on total blood concentrations. The solid lines represent the fit to the sigmoidal Emax model with use of the parameters in table 2; the no-drug response (E0) values were 355 (CPA), 362 (8MCPA, 8ECPA and 8PCPA), 341 (8BCPA) and 318 bpm (8CPCPA), respectively.

View larger version (15K): [in this window] [in a new window]   Fig. 6.   Correlation (r = 0.97) between the GTP shift in vitro (n = 3) and the maximal effect for HR in vivo (n = 6) of CPA and the five 8-alkylamino derivatives of CPA. Data are presented as mean ± S.E.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

In the present study we have shown that the novel 8-alkylamino-substituted CPA analogs are partial agonists for the cardiovascular adenosine A₁ receptors in conscious, normotensive rats in vivo. Estimates of potency and intrinsic activity of the compounds were obtained by quantification of the relationship between blood concentrations and the reduction in HR based on an integrated PK/PD approach. Because compounds can differ in pharmacokinetic properties it is important to quantify effects on the basis of concentrations rather than dose (Van Schaick, submitted for publication). Differences in rate of metabolism, distribution or plasma protein binding influence the concentrations of free compound at the site of the receptor. Consequently, this difference in effect-site concentration can alter the observed effect. Recently, an integrated PK/PD approach has been useful in the characterization of the hemodynamic effects of deoxyribose analogs of CPA in vivo in rats (Mathôt et al., 1995). The partial agonistic behavior of some of the analogs and the lower adenosine A₁ receptor affinity were reflected in the concentration-effects relationship of these compounds.

The blood concentrations after i.v. infusion of the CPA analogs were determined in individual animals. The concentration-time profiles were described best by a biexponential model. The pharmacokinetic parameters estimates were very similar among the 8-alkylamino derivatives of CPA. In comparison with CPA the compounds had 3- to 4-fold longer elimination half-lives. These longer half-lives were likely caused by a significantly larger volume of distribution at steady state (V_dss), because clearance values were approximately equal to those of CPA. The larger V_dss of the 8-alkylamino-substituted CPA analogs may be the result of increased lipophilicity. The increased lipophilicity of the compounds was also indicated by their order of retention on reversed phase chromatography. Increasing alkyl chain lengths resulted in longer retention times on HPLC.

Intravenous administration of the 8-alkylamino analogs of CPA resulted in a transient decrease in both MAP and HR (figs. 3 and 4). Administration of CPA has been shown to cause severe bradycardia and hypotension (Mathôt et al., 1994). Intravenous infusion of this compound at a dose of 0.2 mg/kg for 15 min resulted in an almost instantaneous decrease in HR of approximately 200 bpm (54%). Similarly, MAP was reduced to 43 mm Hg after administration of CPA, which is a decrease of 61% from base line. All 8-alkylamino derivatives of CPA caused a less pronounced reduction in HR and MAP. Moreover, the maximal reductions elicited by the various derivatives depended on the chain length of the 8-alkylamino group. The methyl-substituted analog caused a decrease in HR of approximately 150 bpm, whereas the cyclopentyl-substituted analog was virtually inactive on HR.

HR has been shown to be a sensitive and realistic pharmacodynamic parameter to investigate the hemodynamic actions of adenosine A₁ receptor agonists in vivo (Coffin and Spealman, 1986; Mathôt et al., 1994, 1995). The A₁-receptor-mediated effects on HR include: slowing of the HR (negative chronotropic), impairment of AV conduction (negative dromotropic) and reduction in contraction force (negative inotropic) (Belardinelli et al., 1989). Furthermore, activation of adenosine A₁ receptors can antagonize the stimulatory effects of catcholamines (anti-beta-adrenergic effect), and as a result influence HR indirectly (Belardinelli and Isenberg, 1983). In in vivo studies in rats, Mathôt and colleagues have shown that the reduction in HR is directly proportional to agonist concentrations in blood. Quantification of the relationship between concentration and HR was shown to be relevant for activation of A₁ receptors in vivo (Mathôt et al., 1994, 1995).

The hypotensive effect observed for adenosine A₁ receptor agonists has been shown to be a less suitable pharmacological effect parameter for A₁ receptor activation (Coffin and Spealman, 1986; Abiru et al., 1991; Appel et al., 1995). This effect is mainly caused by a reduction in cardiac output (Webb et al., 1990; Merkel et al., 1993) and is under regulation of complex homeostatic control mechanisms (Struyker Boudier, 1992). Additionally, activation of A_2A and A₃ receptors may contribute to a decrease in blood pressure through vasodilation (Mullane and Williams, 1990; Fozard and Carruthers, 1993) and mediator release from mast cells (Hannon et al., 1995; Van Schaick et al., 1996a). The derivatives of CPA are slightly less selective for A₁ receptors than CPA itself. Activation of A_2A receptors is only possible at high concentrations and may therefore only occur at the end of the 15-min infusion. Recently, administration of the mixed A₁/A_2A agonist 8-butylaminoadenosine to rats was shown to result in a biphasic response on HR because of a reflex tachycardia that was activated by the A_2A-receptor-mediated hypotension (Mathôt et al., 1996). In the present study, however, such a biphasic response did not occur, which suggests that activation of A_2A receptor was limited. All compounds (except 8CPCPA) produced significant reductions in MAP. However, for 8MCPA and 8BCPA these reductions in MAP were relatively larger than for HR. 8MCPA elicited a hypotensive response similar to CPA (50% and 61% reduction, respectively), whereas the reduction in HR was much less pronounced (34% and 54% reduction for 8MCPA and CPA, respectively). These results may indicate that a decrease in cardiac output is not the only determinant of the hypotensive response of these agonists.

The individual pharmacokinetic fits were used to calculate the blood concentrations of the compounds at the time points of the effect measures. The HR was directly related to blood concentrations because no delay between the concentrations in blood and the effect was observed. Unfortunately, in most treatment groups it was not possible to determine blood concentrations during the entire duration of the experiment. In the description of the concentration-effect relationship only the HR measures during the time span of blood sampling (0-90/120 min) were included. The no-drug response values (E₀) were fixed to post-dose values. It has been demonstrated that in this way it is possible to obtain realistic estimates of pharmacodynamic parameters (Mathôt et al., 1994). The individual concentration-HR relationships of 8MCPA, 8ECPA and 8PCPA were described successfully based on the sigmoidal E_max model (fig. 5). These relationships show that the maximal effect on HR was indeed reached (a plateau level that is maintained at high concentrations), and that estimates of intrinsic activity of the compounds could be obtained after a single dose. The concentration-effect relationships of 8BCPA and 8CPCPA could not be described in individual rats. The marginal reduction in HR elicited by 8BCPA could be modeled with use of the average of the pooled data. 8BCPA produced a maximal effect (E_max) of 31 bpm (11% change from base line). This decrease is small relative to the base-line variation observed in this group (6%). 8CPCPA did not affect HR across the entire concentration range (fig. 5).

The concentration-HR relationships showed that the maximal effects elicited by the 8-alkylamino derivatives were significantly less than the maximal effect elicited by the full agonist CPA, which indicates partial agonism of these compounds in vivo (fig. 5). The intrinsic activity of the compounds was reflected in the E_max value, which ranged from 208 bpm for the most active agonist (CPA) to 31 bpm for the least active (8BCPA). These in vivo estimates of intrinsic activity correlated very well (r = 0.97) to the GTP shift as observed in receptor binding experiments with rat brain membranes (fig. 6). The GTP shift is defined as the ratio between the K_i in the presence and in the absence of GTP. A shift is only observed for agonists, whereas antagonists do not show a GTP shift, resulting in a ratio of 1. All 8-substituted derivatives of CPA had significantly lower GTP shifts (ranging from 1.1 to 3.8) than the GTP shift of CPA (approximately 6) (Roelen et al., 1996). These in vitro values were in the same order of magnitude as the observed intrinsic activity in vivo (table 3). The lack of intrinsic activity of the 8-cyclopentyl derivative (8CPCPA) corresponds to its GTP shift of 1.2, which is similar to a value observed for antagonists (Stiles, 1988).

Strong correlations between the GTP shift and intrinsic activity have been described for several G-protein-coupled receptors such as beta adrenoceptors (Kent et al., 1980) and muscarinic receptors in vitro (Kenakin, 1993). The correlation between the rank order of the GTP shift and the order of the E_max values for the bradycardic effect in vivo corroborates the usefulness of the GTP shift to investigate intrinsic activity in vitro. Additional data to support the functional responses have been provided by determination agonist efficacies of the compounds in [³⁵S]GTPS binding studies and adenylate cyclase studies in vitro (Lorenzen et al., 1996). Usually, drug-receptor interactions are studied in isolated systems in vitro (e.g., isolated organs, cellular systems). However, the combined quantification of drug concentrations and the effects on HR, as applied in the present study, has provided reliable estimates of receptor activation in vivo (Mathôt et al., 1994; Appel et al., 1995).

The values for E_max and GTP shift are indicative of the effects on cardiac adenosine A₁ receptors. The observed intrinsic activity for HR is a combination of reduced intrinsic efficacy of the drugs and the efficiency of receptor coupling in the tissue. The effect profile of the agonists may, therefore, be totally different in tissues in which the receptor is coupled to other G proteins or effector systems (Kenakin, 1993). Recently, we observed that the adenosine A₁ agonist N⁶-(p-sulfophenyl)adenosine was 6-fold more potent on the antilipolytic effect than on the bradycardiac effect in rats (Van Schaick et al., in press). This selectivity is likely caused by a difference in receptor-effector coupling between cardiac and adipose tissue (Dennis et al., 1989; Lohse et al., 1986). The selectivity of some 8-alkylamino derivatives of CPA is currently explored by comparing PK/PD relationships for the effect on lipid metabolism with the relationships for the bradycardiac effect. Preliminary results show that the partial agonists 8MCPA, 8ECPA and 8BCPA act as full agonists in adipose tissue (Van Schaick et al., 1996b).

The potency of the 8-alkylamino-substituted CPA analogs, e.g., the concentration required to produce 50% of the maximal reduction in HR (EC_50,u), is in the same range as the receptor affinity in radioligand binding experiments (table 3) (Roelen et al., 1996). These correlations have also been shown for other adenosine A₁ receptor agonists and antagonists (Mathôt et al., 1995; Appel et al., 1995). The 8-substituted analogs are less potent than CPA. Clearly, substitution of an alkylamino group at the 8-position results not only in a decrease in intrinsic activity but in potency as well. In the past, 8-substituted adenosine analogs have been neglected because of their low affinity and activity (Bruns, 1980; Olsson et al., 1979; Jacobson, 1990). However, substitution at the C8-position was found not only to affect affinity but to cause a favorable decrease in intrinsic activity as well (Bruns, 1980; Van der Wenden et al., 1995b). These observations have led to the development and in vivo characterization of 8-butylaminoadenosine (Van der Wenden et al., 1995b; Mathôt et al., 1996). Despite its poor affinity and selectivity this compound was shown to be a partial agonist for the adenosine A₁ receptor.

Interestingly, the N⁶,C8-disubstituted adenosine analogs investigated in the present study have higher potency and A₁ selectivity than mono-C8-substituted adenosines. Substitution of the cyclopentyl group at the N⁶-position greatly enhances affinity for the adenosine A₁ receptor while the reduction in intrinsic activity is kept intact. Moreover, our results show that by substitution at the 8-position a differentiation between reduction in potency and reduction in intrinsic activity can be obtained, which is in contrast to the correlations between affinity and activity observed in other studies (Borea et al., 1994; Mathôt et al., 1995). The analogs 8MCPA, 8ECPA, 8PCPA and 8BCPA have similar EC₅₀ values, whereas their intrinsic activities are differing (149, 81, 101 and 31 bpm, respectively).

In conclusion, the present series of compounds may be useful as pharmacological tools, because of their relative high potency and controllable intrinsic activity for cardiac adenosine A₁ receptors in vivo. To our knowledge these compounds are the most potent partial adenosine A₁ agonists reported to date. Therapeutic application of partial agonists may be advantageous because this may lead to increased tissue or organ selectivity and reduced receptor desensitization and down-regulation (IJzerman et al., 1994). The magnitude of the response mediated by partial agonists depends largely on the amplification caused by receptor-effector coupling (Kenakin, 1993). Because of tissue differences in receptor-effector coupling this may lead to increased selectivity in effects. The partial agonists with low cardiovascular activity (e.g., 8BCPA) may still have pronounced effects on other physiological processes such as lipid metabolism.