Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

ET, originally isolated from cultured porcine aortic endothelial cells, is a highly potent vasoconstricting peptide with 21-amino acid residues (Yanagisawa et al., 1988). Three distinct members of the ET family, namely ET-1, ET-2 and ET-3, have been identified in humans through cloning (Inoue et al., 1989). The effects of ETs on mammalian organs and cells are initiated by their binding to high-affinity G-protein-linked receptors. ET receptors are found in various tissues and cells, such as brain, lung and mesangial cells (Sokolovsky, 1992). Two types of ET receptors, ET_A and ET_B, have been characterized, isolated (Kozuka et al., 1991; Wada et al., 1990) and their cDNA cloned (Arai et al., 1990; Sakurai et al., 1990). ET_A receptors are selective for ET-1 and ET-2, whereas ET_B receptors bind to ET-1, ET-2 and ET-3 with equal affinity. Several antagonists and agonists for ET receptors have been developed (Opgenorth, 1995). Among them, FR139317 (Sogabe et al., 1993), PD-156707 (Reynolds et al., 1995), L-749329 (Walsh, 1995), Ro-47-0203 (Clozel et al., 1994) and A-127722 (Opgenorth et al., 1996) are potent ET receptor antagonists.

Since endothelins were discovered in 1988 (Yanagisawa et al., 1988), research in this field has become one of the most rapidly developing areas in the biological sciences. Since the development of potent antagonists for ET receptors, there has been keen interest in developing these ET receptor antagonists for clinical utilization. Information on the interaction between ET receptor antagonists and plasma proteins is rather limited, despite the potential impact of protein binding on the in vivo efficacy. Previously we have shown that ET-1, ET-3 and ET receptor antagonists (PD-156707, L-749329, Ro-47-0203 and A-127722) exhibit high degrees of binding to human plasma proteins, especially serum albumin (Wu-Wong et al., 1996). Addition of human serum albumin (HSA, 5%) decreases the potency of ET receptor antagonists in inhibiting ET binding to the receptors (Wu-Wong et al., 1996). Because BSA is frequently included in in vitro assays for determination of the potency and efficacy of ET receptor antagonists, it is important to know whether BSA will have effects similar to those previously shown for HSA. In addition, because investigators do not always use the same concentrations of BSA in their studies, it may be important to know whether different amounts of BSA affect the potency to different degrees.

The purposes of this report are 1) to examine whether the presence of BSA affects the potency of ET receptor antagonists and 2) to compare ET receptor antagonists in various assays using a fixed concentration of BSA. We first show that increasing concentrations of BSA in the assay buffer incrementally decrease the potencies of test antagonists in inhibiting ET binding to the receptors. Furthermore, antagonists of distinct structures are affected by BSA to different degrees. In functional assays, BSA also decreases the potency of antagonists in inhibiting ET-1-stimulated PI hydrolysis. Five structurally distinct ET receptor antagonists (FR139317, PD-156707, L-749329, Ro-47-0203 and A-127722) were then compared in binding and functional assays under identical conditions with 0.2% BSA.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Materials. [¹²⁵I]ET-1 (2200 Ci/mmol) and [¹²⁵I]ET-3 (2200 Ci/mmol) were obtained from Du Pont, NEN (Boston, MA). ET-1 and ET-3 were purchased from American Peptide Company (Sunnyvale, CA). FR139317 (cC6N-L-Leu-D-Trp-(Me)-D-2Pya-OH), A-127722 (trans-trans-2-(4-methoxyphenyl)-4-(1,3-benzodioxo-5-yl)-1-((N,N-dibutylamino)carbonylmethyl)pyrrolidine-3-carboxylic acid), L-749329 (3,4-methylenedioxy-1-(2-propyl-4-carboxyphenoxy)-N-(4-isopropyl-phenysulfonyl)-benzene acetamide), Ro-47-0203 ((4-tert-butyl-N-[6-(2-hydroxy-ethoxy)-5-(2-methoxy-phenoxy)-2,2-bipyrimidin-4-yl]-benzenesulfonamide)) and PD-156707 ({sodium 2-benzo[1,3]dioxol-5-yl-4-(4-methoxy-phenyl)-4-oxo-3-(3,4,5-trimethoxy-benzyl)-but-2-enoate}) were synthesized in house. Other reagents were of analytical grade. Figure 1 shows the chemical structures of the five ET receptor antagonists evaluated.

View larger version (19K): [in this window] [in a new window]   Fig. 1.   Structures of ET receptor antagonists.

Cell culture. MMQ cells were licensed from University of Virginia and were cultured as described previously (Judd et al., 1988). Smooth muscle cells prepared from human pericardium were provided by Dr. Maria J. Vidal (Department of Cardiology, Instituto Scientifico San Raffaele, Milano, Italy) and were cultured as described previously (Wu-Wong et al., 1994a). Chinese hamster ovary (CHO) cells permanently expressing human ET_A or ET_B receptor (Magnuson et al., 1994) were cultured in F-12 medium containing 10% fetal bovine serum (FBS) and 500 µg/ml G418 (geneticin). Cell viability was examined by the trypan blue exclusion method.

Preparation of membranes. Membranes were prepared from porcine cerebellum, rat pituitary MMQ cells or CHO cells as previously described (Wu-Wong et al., 1993; 1996). Briefly, cerebella or cells were homogenized in 25 volumes (w/v) of 10 mM HEPES (pH 7.4) containing 0.25 M sucrose and protease inhibitors (3 mM EDTA, 0.1 mM phenylmethyl sulfonyl fluoride and 5 µg/ml Pepstatin A) by 3 to 10 s polytron at 13,500 rpm with 10-s intervals (for cerebella) or by a micro ultrasonic cell disruptor (Kontes, Vineland, NJ) (for cells). The mixture was centrifuged at 1000 × g for 10 min. The supernatant was collected and centrifuged at 30,000 × g for 30 min (for cerebella) or at 60,000 × g for 60 min (for cells). The precipitate was resuspended in Buffer B-1 (20 mM Tris, 100 mM NaCl, 10 mM MgCl₂, pH 7.4) containing the aforementioned protease inhibitors and then centrifuged again. The final pellet was resuspended in Buffer B-1 containing protease inhibitors and stored at 80°C until used. Protein content was determined by the Bio-Rad dye-binding protein assay.

Radioligand binding to membranes. Binding assays were performed in 96-well microtiter plates precoated with 0.1% BSA unless otherwise indicated. Membranes were diluted in Buffer B (Buffer B-1 with the aforementioned protease inhibitors plus 0.025% bacitracin and 0.2% BSA) to a final concentration of 0.05 mg/ml of protein. In some experiments, the concentrations of BSA in Buffer B were different as indicated. In competition studies, membranes were incubated with 0.1 nM of [¹²⁵I]ET in Buffer B (final volume: 0.2 ml) in the presence of increasing concentrations of unlabeled test ligands for an indicated period of time at 25°C. In saturation studies, membranes were incubated with increasing concentrations of [¹²⁵I]ET in Buffer B (final volume: 0.2 ml) in the presence or absence of unlabeled test ligands for 4 h at 25°C. After incubation, unbound ligands were separated from bound ligands by vacuum filtration using glass-fiber filter strips in PHD cell harvesters (Cambridge Technology, Inc., Watertown, MA), followed by washing of the filter strips with saline (1 ml) three times. Nonspecific binding was determined in the presence of 1 µM ET-1 or ET-3.

d

d

d

Measurement of PI hydrolysis. The procedure has been reported previously (Wu-Wong et al., 1993). Briefly, MMQ cells (0.4 × 10⁶ cells/ml) prelabeled with 1 µCi/well of [³H]myo-inositol for 16 to 24 h were washed with PBS and then incubated with Buffer A (Earle's solution: 140 mM NaCl, 5 mM KCl, 1.8 mM CaCl₂, 0.8 mM MgSO₄, 5 mM glucose, buffered with 25 mM HEPES, pH 7.4, with or without 0.2% BSA) containing protease inhibitors (3 mM EDTA, 0.1 mM PMSF, and 5 µg/ml Pepstatin A) and 10 mM LiCl for 60 min before being challenged with ET-1 for an additional 45 min. ET challenge was terminated by the addition of 50 µl of 1 N NaOH, and the mixture was immediately neutralized by adding 50 µl of 1 N HCl. The samples were treated with 1.5 ml of chloroform/methanol (1:2, v/v). Total inositol phosphates were extracted after adding chloroform and water to give final proportions of chloroform/methanol/water of 1:1:0.9 (v/v/v) as described by Berridge et al. (1982). The upper aqueous phase (1 ml) was retained, and a small portion (100 µl) was counted. The rest of the aqueous sample was analyzed by batch chromatography using the anion-exchange resin AG1-X8 (Bio-Rad, Hercules, CA). Total water-soluble inositol phosphates were eluted from the resin by 6 ml of 1 M ammonium formate with 0.1 N formic acid after the resin was washed with 6 ml of 60 mM sodium formate with 5 mM sodium tetraborate.

Measurement of AA release. Human pericardial smooth muscle cells (HPSMC) in 48-well culture plates at 80% to 100% confluency were labeled with 0.4 µCi/well of [³H]AA in Dulbecco's modified Eagle's medium (DMEM) with 10% FBS for 16 to 24 h. To assay AA release, cells were incubated with DMEM plus 0.2% BSA (0.5 ml/well) for 30 min. After the incubation, the medium was removed, and 0.3 ml DMEM with 0.2% BSA was added to each well. Various test agents at different concentrations were added, and finally ET-1 ranging from 10¹¹ to 10⁶ M was added. Cell were incubated at 37°C for another 30 min, and the incubation medium was collected for determination of radioactivity.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Competition binding studies were conducted in the presence of different amounts of BSA for A-127722 and L-749329, two antagonists that have similar potencies for ET_A receptor but are different in potencies for ET_B receptor as reported in the literature (Walsh, 1995; Opgenorth et al., 1996). Figure 2 shows that the IC₅₀ value for A-127722, when assayed using membranes prepared from MMQ cells with predominantly ET_A receptor, shifted from 0.09 nM in the absence of BSA to 0.44, 1.26 and 4.30 nM in the presence of 0.2, 1 and 5% BSA, respectively. A similar shift was observed when the binding assay was done using membranes prepared from porcine cerebella with predominantly ET_B receptor; the IC₅₀ value for A-127722 changed from 0.10 µM in the absence of BSA to 0.34, 1.21 and 2.96 µM in the presence of 0.2, 1 and 5% BSA, respectively. Interestingly, when similar studies were conducted for L-749329 using membranes prepared from MMQ cells, the addition of BSA had a much greater impact on the potency of the antagonist. The IC₅₀ value for L-749329 shifted from 0.09 nM in the absence of BSA to 45, 181 and 820 nM in the presence of 0.2, 1 and 5% BSA, respectively. When the binding assay was done using membranes prepared from porcine cerebella, the IC₅₀ value for L-749329 shifted from 0.026 µM in the absence of BSA to 12, 33 and >100 µM in the presence of 0.2, 1 and 5% BSA, respectively. Table 1 summarizes the inverse relationship between IC₅₀ values and BSA concentrations for A-127722 and L-749329.

View larger version (33K): [in this window] [in a new window]   Fig. 2.   The effect of BSA on the potency of A-127722 (panels A and B) and L-749329 (panels C and D) in inhibiting [125I]ET-1 binding to human ETA (panels A and C) and ETB (panels B and D) receptors. Receptor binding was performed as described except that the assay plates were not BSA-coated. Nonspecific binding (ranging from 3.3 to 8.6 fmol/mg), determined in the presence of 1 µM ET-1, was subtracted from total binding to give specific binding. The results are expressed as percent of control, with binding in the absence of the antagonist as 100%. In the absence of antagonists, specific binding for ETA receptor was 50.1, 45.5, 47.3 and 38.1 fmol/mg at 0, 0.2%, 1% and 5% of BSA, respectively. Specific binding for ETB receptor was 434, 392, 390 and 363 fmol/mg at 0, 0.2%, 1% and 5% of BSA, respectively. Each value represents the mean of two (panels A and B) and three (panels C and D) determinations. Results shown are representative of two different experiments. The active enantiomer of A-127722 was used in this study.

The above observation suggests that the potency of an ET receptor antagonist is critically dependent on the concentration of BSA used in the assay. This led us to compare directly a number of ET receptor antagonists in various assay systems at a fixed BSA concentration. We selected five ET receptor antagonists for comparison: A-127722, L-749329, FR139317, Ro-47-0203 and PD-156707. Figure 3 shows the results from competition binding studies using cloned human ET_A and ET_B receptors permanently expressed in CHO cells. In this binding assay system, 0.2% BSA was used in a buffer containing 20 mM Tris (pH 7.4), 100 mM NaCl, 10 mM MgCl₂, and protease inhibitors 0.1 mM PMSF, 5 µg/ml Pepstatin A, 0.025% bacitracin and 3 mM EDTA. The IC₅₀ values are summarized in table 2. The results suggest that the potencies of the five antagonists are in the order of A-127722 > PD-156707 > FR139317 > Ro-47-0203 > L-749329 for ET_A receptor, and A-127722 > Ro-47-0203 > L-749329 > PD-156707 > FR139317 for ET_B receptor. All the antagonists tested exhibit a preference for ET_A receptor, although the selectivity for ET_A vs. ET_B ranges from 42-fold for L-749329 to 10,686-fold for PD-156707.

View larger version (29K): [in this window] [in a new window]   Fig. 3.   Comparison of antagonists in competition binding studies. Membranes (20 µg) were prepared from CHO cells stably transfected with the human ETA receptor (panel A) or the human ETB receptor (panel B). Membranes were incubated with 0.1 nM [125I]ET-1 (panel A) or [125I]ET-3 (panel B) in the presence of increasing concentrations of test ligand for 3 h at 25°C. Nonspecific binding, determined in the presence of 1 µM ET-1 or -3, was subtracted from total binding to give specific binding. The results are expressed as percent of control, with binding in the absence of the antagonist as 100%. Each value represents the mean (± S.E.) of n determinations as indicated in Table 2.

To examine further the nature of the interaction between the antagonists and ET receptor, we performed [¹²⁵I]ET-1 saturation binding studies using membranes prepared from CHO cells stably expressing human ET_A and ET_B receptors. As shown in figure 4, A and B, increasing concentrations of L-749329 affected [¹²⁵I]ET-1 binding by causing successive increases in the apparent K_d values without having a significant effect on the B_max values. This result suggests that L-749329 is a competitive inhibitor for ET-1 binding to the ET_A receptor. The K_i value of L-749329 was determined to be 33.6 nM from figure 4C. Similar experiments were performed for the other four antagonists, and the results are compared in figure 4C. All five antagonists exhibited competitive behavior for inhibiting ET-1 binding. K_i values for A-127722 and Ro-47-0203 against human ET_B receptor were also determined using [¹²⁵I]ET-3 binding to the cloned human ET_B receptor, and again, competitive inhibition was observed. Because of the weak potencies of PD-156707, FR139317 and L-749329 for the ET_B receptor, K_i values for these three compounds against human ET_B receptor were not determined. The K_i values are summarized in table 3. Again, the rank order of potency of the five antagonists is A-127722 > PD-156707 > FR139317 > Ro-47-0203 > L-749329 for ET_A receptor.

View larger version (20K): [in this window] [in a new window]   Fig. 4.   Saturation binding studies. A) Membranes (20 µg) with the human ETA receptor were incubated with increasing concentrations of [125I]ET-1 in the presence of 0, 20, 40 and 60 nM L-749329. Binding was performed as described. Nonspecific binding, determined in the presence of 1 µM ET-1, was subtracted from total binding to give specific binding. B) Scatchard analysis of panel A. C) Ki value was determined from panel B as described in "Materials and Methods." Similar studies were done for the other four antagonists against human ETA receptor, and the final Ki results are shown in panel C for comparison. Results shown are representative of n different experiments as indicated in Table 3.

The potency of an antagonist in inhibiting binding is generally expected to translate into comparable potency for antagonizing a functional response. To confirm this notion, we compared these ET receptor antagonists in functional assays. First, the effect of BSA on the potency of receptor antagonists was tested in ET-1-stimulated PI hydrolysis in MMQ cells. We have previously shown that ET-1-stimulated PI hydrolysis in MMQ cells is mediated through the ET_A receptor (Wu-Wong et al., 1993). Figure 5 shows that all the test antagonists inhibited ET-1-evoked PI hydrolysis in a dose-dependent manner in the presence or absence of BSA (0.2%). None of the compounds alone showed any effect, which suggests that none has agonist activity. The IC₅₀ values were significantly affected by the addition of BSA (table 4); structurally distinct antagonists were affected differently. In the absence of BSA, the potency is in the order of L-749329 > A-127722 > PD-156707 > FR139317 > Ro-47-0203; in the presence of BSA, the rank order of potency is A-127722  PD-156707 > FR139317 > L-749329  Ro-47-0203.

View larger version (26K): [in this window] [in a new window]   Fig. 5.   PI hydrolysis with or without BSA. Rat pituitary MMQ cells were prelabeled with [3H]myo-inositol (1 µCi/well) for 16 h. Cells were incubated with or without 1 nM ET-1 ± antagonists for 45 min at 37°C. Control: no addition of ET-1 or test agents. A) Without BSA. B) With 0.2% BSA in the buffer. Each value represents the mean ± S.E. of n determinations as indicated in Table 4.

ET-1 has also been shown to stimulate AA release in human pericardial smooth muscle cells, an effect mediated through the ET_A receptor (Wu-Wong et al., 1994a). The antagonists were compared in the AA release assay with 0.2% BSA included in the buffer. Figure 6A shows that all the test antagonists inhibited ET-1-evoked AA release in a dose-dependent manner. Figure 6B shows that ET-1 stimulated AA release in a dose-dependent manner with an EC₅₀ value of 0.6 nM. Increasing concentrations of A-127722 shifted the concentration-dependent curves of ET-1 to the right in a parallel manner without a significant effect on the maximal response, which suggests that A-127722 is a competitive inhibitor of ET-1-induced AA release. The pA₂ value for A-127722 was determined to be 10.5 ± 0.3 (n = 3) by Schild analysis (fig. 6C). Similar experiments were performed for the other four antagonists, and the results are compared in Figure 6C and table 5. All five antagonists exhibited competitive antagonism of the ET-1-stimulated AA release. The rank order of potency is PD-156707  FR139317  A-127722 > Ro-47-0203 > L-749329.

View larger version (24K): [in this window] [in a new window]   Fig. 6.   AA release. A) IC50 determination: Human pericardial smooth muscle cells were prelabeled with [3H]AA (0.4 µCi/well) for 16 h. Cells were incubated with or without 1 nM ET-1 ± antagonists for 30 min at 37°C in the presence of 0.2% BSA as described in "Materials and Methods." Control: no addition of ET-1 or test agents. B) The effect of A-127722: Human pericardial smooth muscle cells in 48-well plates prelabeled with [3H]AA were challenged with increasing concentrations of ET-1 in the presence of A-127722 at 0, 0.1, 0.5, 2 and 10 nM for 30 min at 37°C as described in "Materials and Methods." Radioactivity released was expressed as percent of control (cells not treated with ET-1 or test agents). C) Schild analysis of panel B for determining the pA2 value. Similar studies were done for other antagonists, and the final Schild analysis results are shown in panel C for comparison. Results shown are representative of n different experiments as indicated in Table 5. CR: concentration ratio; I: antagonist.

To evaluate further the data from binding studies and functional assays for the five antagonists, we plotted the IC₅₀ values from [¹²⁵I]ET-1 binding to human ET_A receptors (table 2) against the IC₅₀ values from ET-1-stimulated PI hydrolysis and AA release (tables 4 and 5) as shown in figure 7.

View larger version (14K): [in this window] [in a new window]   Fig. 7.   Correlation between binding studies and functional assays. The IC50 values from [125I]ET-1 binding to human ETA receptors were plotted against the IC50 values from ET-1-stimulated PI hydrolysis () and AA release () for the five antagonists. The correlation coefficient (R) between binding studies and functional assays is 0.82. BSA (0.2%) was present in all assays.

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

The studies presented here clearly demonstrate that BSA affects the potency of ET receptor antagonists. BSA affects not only the potency of antagonists in inhibiting ET binding to receptors but also the potency of antagonists in inhibiting ET-1-stimulated PI hydrolysis. Interestingly, the effect of BSA is remarkably different for different antagonists. As shown in figure 2 and table 1, L-749329 and A-127722 exhibit identical potency for inhibition of ET-1 binding to ET_A receptor when no BSA is added in the assay buffer. However, when 5% BSA is added, an approximately 9000-fold drop in potency for ET_A receptor is observed for L-749329, whereas only a 47-fold decrease in potency for ET_A receptor is observed for A-127722. A similar observation was made in functional assays. When BSA was not added, the IC₅₀ value for L-749329 was 0.01 nM in the PI hydrolysis assay (fig. 5 and table 4). However, in the presence of 0.2% BSA, the IC₅₀ value for L-749329 was more than 4000-fold greater (fig. 5 and table 4). These results are consistent with our previous observation with human serum albumin (Wu-Wong et al., 1996). The reason why the activity of L-749329 is more affected by BSA than that of A-127722 is not immediately apparent.

Why does BSA have such an impact on the potency of ET receptor antagonists? Our previous studies show that ET receptor antagonists exhibit strong binding to plasma proteins, especially serum albumin (Wu-Wong et al., 1996). The finding is not surprising, because albumin readily binds to lipophilic acids, a common characteristic of the ET antagonists tested in this report. Thus it is likely that BSA acts as a "pseudo-receptor" to bind ET receptor antagonists and that antagonists that bind to BSA are no longer free to bind to ET receptors. As a result, a decrease in the potency is observed. The hypothesis that plasma proteins, such as serum albumin, act as "pseudo-receptors" for ET receptor antagonists may also explain why a disparity between the in vitro and in vivo potencies is often observed for ET antagonists (Sogabe et al., 1993; Clozel et al., 1994; Reynolds et al., 1995; Opgenorth et al., 1996). The following example is an exercise to demonstrate this point. From in vitro binding assays, the IC₅₀ value at 0% BSA for A-127722 is ~0.1 nM. A concentration of 100 nM is required to inhibit ET-1 binding completely. On the basis of the plasma elimination half-life of 3.5 h and the peak plasma concentration of 1.1 µg/ml at 5 mg/kg in the rat (Opgenorth et al., 1996), we can estimate that the dose of A-127722 required to reach a plasma concentration of 100 nM is ~0.25 mg/kg. At this dosage, A-127722 should completely inhibit the increase in arterial blood pressure induced by ET-1. However, we have shown that approximately 10 mg/kg of A-127722 is required to completely inhibit an ET-1-induced increase in arterial blood pressure (Opgenorth et al., 1996). The potency difference (40-fold) between in vitro and in vivo is similar to the IC₅₀ difference observed in binding assays at 0% and 5% BSA. It is worth mentioning that the protein concentration in human blood is 7% and that 55% to 60% of that is serum albumin. Similar calculations can be done for other antagonists. Clearly, serum albumin, and perhaps other plasma proteins, greatly impacts the potency of ET receptor antagonists.

Both this report and our previous studies (Wu-Wong et al., 1996) show that the effect of serum albumin on antagonist pharmacology is different for different antagonists. Why do the antagonists behave differently in their sensitivity to serum albumin? It is unlikely that these differences can be explained by different inhibiting modes (e.g., competitive vs. non-competitive), because the K_i and pA₂ studies in this report suggest that all these compounds are fully competitive inhibitors. A second possibility is that each antagonist interacts with different regions of the receptor. Indeed, it has been shown that binding sites for Ro-47-0203 are different from those for BQ-123 (Breu et al., 1995). The advent of radiolabeled antagonists may make it possible to investigate the particular interacting sites on the receptor for each antagonist. A third explanation is that the more reversible the binding of an antagonist to ET receptors is, the greater the impact BSA has on its potency. We have previously shown that ET receptor agonists and antagonists, once bound to the receptor, are difficult to wash away and that the binding of some antagonists is more reversible than that of others (Wu-Wong et al., 1994b, c). The difference in the "stickiness" of binding may shed some light on why different ET receptor antagonists respond to BSA differently. In fact, our preliminary results show that the binding of L-749329 to ET receptors is more reversible than that of A-127722, an outcome coincidental with results in the current study that the effect of BSA on L-749329 is more profound than its effect on A-127722. Finally, because these antagonists have distinct structures, it is possible that the "stickiness" of binding is linked to the chemical features of each compound. So far, development of small-molecule ET receptor antagonists has concentrated on improving potency and pharmacokinetic profiles, and there has been no reported attempt to evaluate the structure-activity relationship of protein binding.

Reports in the literature on ET receptor antagonists utilize various serum albumin concentrations in the assays. For example, Williams et al. (1995) and Sogabe et al. (1993) used 0.01% HSA and BSA, respectively, in the binding assay. BSA at 0.1% was used in the assays by Webb et al. (1995) and Reynolds et al. (1995), and 0.5% BSA was used by Clozel et al. (1994). If BSA at different concentrations indeed affects the potency of antagonists to different degrees, and if the effect of BSA is different for different antagonists, it is of significant importance to compare ET receptor antagonists directly in assays using a fixed concentration of BSA. Because of our previous studies on ET receptor antagonists (Opgenorth et al., 1996), we chose to continue the practice of using 0.2% BSA in the assays for the comparison study. It would be ideal to compare all known ET receptor antagonists directly, but because of practical considerations, five antagonists were chosen on the basis of the following information: FR139317 is one of the earliest ET receptor antagonists and has been tested in various animal disease models. Ro-47-0203 (Bosentan) is furthest along in clinical development and therefore is of particular interest to researchers in the ET field. And A-127722, PD-156707 and L-749329 are among the most potent antagonists reported in the literature and appear to have clinical utility.

In summary, we show that BSA affects the potency of ET receptor antagonists in binding and functional assays and that the effect of BSA is remarkably different for different antagonists. The comparison study shows that although there is some variation among assays, in general, the rank order of potency of the five antagonists tested in various assays is A-127722  PD-156707 > FR139317 > Ro-47-0203 > L-749329 when 0.2% BSA is included in the assay buffer. The correlation between binding studies and functional assays is reasonably good with a correlation coefficient of 0.82.