IMPROVED MODELS FOR PHARMACOLOGICAL NULL EXPERIMENTS - CALCULATION OFDRUG EFFICACY AT RECOMBINANT D1A DOPAMINE-RECEPTORS STABLY EXPRESSED IN CLONAL CELL-LINES

Modern drug discovery demands accurate knowledge of the drug propertie s of affinity and efficacy at specific receptor proteins. Furthermore, drugs with well defined properties make better tools with which to ex plore and understand receptor regulation. The use of clonal cell lines stably expressing a given recombinant receptor may provide a highly u seful model in which drug effects may be studied on one receptor subty pe at a time. The present report was designed to evaluate the utility of a general method in which a clonal cell line stably expressing a re combinant D1A dopamine receptor was used as a model system for studyin g drug actions by null models. The null model for receptor occlusion ( to calculate agonist K-a) and the null model for relative efficacy (to calculate test agonist affinity and epsilon(r)) were evaluated in the se studies. To initiate these studies, rat C6 glioma cells that do not normally express DA receptors have been modified by stable transfecti on with the primate D1A DA receptor [Machida et al., 1992 (Molec. Phar macol. 41: 652-659)] to a density of approximate to 200 fmol/mg protei n. The recombinant receptors show robust stimulation of cAMP in the st ably transfected C6 cells. Calculation of agonist K-a from dose-respon se data requires that a portion of the cell's receptors be occluded in the absence of changes in post-receptor events leading to the respons e. Receptor reserve is typically reduced by alkylation, thereby loweri ng maximal response. Unfortunately, most of the currently available al kylating agents are not selective either for a particular receptor or for receptors vs other proteins within a signaling pathway. Short-term agonist treatment offers a possible complement to the use of non-sele ctive or poorly characterized alkylating drugs for reducing maximum re sponse in appropriate cell systems. The null method of receptor occlus ion was used to determine the K-a for dopamine when maximum response w as decreased by alkylation vs short-term agonist treatment. Direct non -linear curve fitting was used to analyze the data. In addition to DA, two other compounds were used to reduce receptor reserve to validate the method: fenoldopam (relatively high efficacy) and SKF38393 (low ef ficacy). Analyses indicated that the affinity of DA was similar whethe r calculated by alkylation (1.1 +/- 0.58 mu M), 75 min DA treatment (0 .57 +/- 0.16 mu M) or 45 min treatment with DA (0.86 +/- 0.11 mu M). S hort-term agonist treatment experiments using multiple concentrations of DA, fenoldopam, or SKF38393 to decrease receptor reserve provided a dditional support for the validity of the K-a determinations using thi s procedure. Other experiments were conducted according to the null mo del for relative efficacy in which the affinity for DA is calculated b y comparing the DA response before and after receptor occlusion, and t he affinity and relative intrinsic efficacy of the test agonist are de termined as a function of its actions relative to DA. We used the foll owing four test drugs: + Br-APB, a novel agent with potential dopamine agonist properties, and three high-affinity DA agonists, fenoldopam, R-(-)-apomorphine (APO), and SKF38393. Intrinsic efficacy values relat ive to that of DA (1.0) were as follows: fenoldopam, 0.46 +/- 0.11; AP O, 0.19 +/- 0.13; SKF38393, 0.07 +/- 0.01; and +Br-APB, 0.26 +/- 0.40. The agonist affinities (K-a) were: fenoldopam, 0.018 +/- 0.008 mu M; APO, 0.80 +/- 0.18 mu M; SKF38393, 0.16 +/- 0.04 mu M; Br-APB, 0.43 +/ - 0.29 mu M; and DA, 0.58 +/- 0.17 mu M EC(50)/K-a ratios were consist ent with relative intrinsic efficacies and K-a values were similar to K-L values reported for membrane binding studies. Finally, Monte Carlo simulations were conducted to determine the precision of the paramete r estimates. These simulations of 1000 experiments each, using the emp irical variance of approximate to 3% of the maximum response, indicate d a precision for the DA K-a of approximate to 16%, for the independen tly determined fenoldopam K-a of approximate to 24%, and of the fenold opam epsilon(r) of approximate to 16% of the parameter values. To our knowledge, this represents the first report of agonist affinity plus r elative intrinsic efficacy measurement at any recombinant receptor. Th e present results also provide the first report of the precision of th ese important drug parameter measurements using Monte Carlo methods. T ogether, these results suggest that clonal cell lines which stably exp ress a recombinant receptor of interest may offer a unique system with which to determine essential drug properties. The method is able to a ccommodate agonists with very high potency and intrinsic activity, ext remely low relative intrinsic efficacy, and can determine whether or n ot a drug is behaving as expected for a classical partial agonist. Cop yright (C) 1996 Elsevier Science Ltd