Introduction

Top Summary Introduction Procedures Results Discussion References

Cysteinyl-LTs are a family of biologically active compounds derived from arachidonic acid via the 5-lipoxygenase pathway (Murphy et al., 1979); they have been shown to be potent bronchoconstrictors (Dahlén et al., 1980) in both normal people and asthma patients (Barnes et al., 1984). Therefore, the main pathophysiological role of cysteinyl-LTs lies in asthma (Piper et al., 1991), which is considered a chronic inflammatory condition characterized not only by bronchial constriction but also by bronchial hyperresponsiveness, mucus hypersecretion, and plasma extravasation. In the latter phenomena, the lung parenchyma plays a fundamental role (Chanarin and Johnston, 1994). For these reasons, an increasing number of studies aimed at the pharmacological characterization of Cys-LT receptors have been performed in the respiratory system of different species.

It has been demonstrated that at least two different Cys-LT receptors exist in guinea pig airways: one predominantly activated by LTD₄ and LTE₄, and a second predominantly activated by LTC₄. The former is sensitive to the so-called classic antagonists, the most studied of which are SKF 104353, MK 571, ICI 198,615, and Ro 24-5913 (Salmon and Garland, 1991), and has been named Cys-LT₁, whereas the latter, Cys-LT₂, is insensitive to the same antagonists (Coleman et al., 1995). The nature of Cys-LT receptors in human airways has been a matter of debate; it is widely accepted that LTC₄ binds to a variety of nonreceptor sites (i.e., enzymes involved in its synthesis and metabolism and transporters) (Keppler, 1992; Nicholson et al., 1992; Metters et al., 1994). So far, such binding proteins have impaired the identification of its specific receptor by ligand-binding studies.

The aim of the current study was to characterize the nature and number of cysteinyl-LT binding sites in HLPM. In fact, it has already been demonstrated in this tissue that under controlled metabolic conditions, both LTC₄ and LTD₄ are able to contract isolated human lung strips (Gardiner and Cuthbert, 1988). We used a variety of experimental protocols for ligand binding studies in addition to computer modeling of binding data and GTPase activation. We report the identification of two distinct high affinity binding sites for LTC₄ and LTD₄ that bear the characteristics of specific receptors.

	Experimental Procedures

Top Summary Introduction Procedures Results Discussion References

Materials

[³H]LTC₄ (127-173 Ci/mmol) and [³H]LTD₄ (127-173 Ci/mmol) were purchased from DuPont-New England Nuclear (Boston, MA). [-³²P]GTP (>5000 Ci/mmol) was purchased from Amersham International (Little Chalfont, Buckinghamshire, UK). LTC₄, LTD₄, and LTE₄ were obtained from Cayman Chemical (Ann Arbor, MI). SKF 104353 was kindly provided by SmithKline and Beecham (King of Prussia, PA). ICI 198,615 was kindly provided by Zeneca (Basiglio, Italy). Gpp(NH)p, S-decyl-GSH, cysteine, glycine, boric acid, serine, HEPES, and the reagents used for GTPase assay were purchased from Sigma Chemical (St. Louis, MO). Filtercount and Ultima Gold were from Packard Instruments (Meriden, CT). All the reagents used in HPLC analysis were of analytical grade and purchased from Carlo Erba (Milan, Italy), as were GF/C Whatman fiber-glass filters.

Preparation of Human Lung Membranes

Crude membranes were prepared from macroscopically normal human lung specimens that had removed during thoracotomy for lung cancer as described previously (Rovati et al., 1985). Briefly, specimens were minced and homogenized at 4° in 10 mM HEPES buffer, pH 7.4 (1:24, w/v), with a Polytron homogenizer (Brinkmann Instruments, Westbury, NY). The homogenate was centrifuged at 770 × g for 10 min, and the supernatant was centrifuged at 27,000 × g for 20 min. The pellet was resuspended in the same buffer, centrifuged under the same condition and resuspended in 1/20th of the homogenization volume. The membrane aliquots were frozen at 80° and stored for no longer than 3 months. Protein content was determined with the Bradford dye-binding protein assay (Pierce, Rockford, IL). Before use, serine/borate complex (40 mM), prepared as an equimolar solution of serine and boric acid, cysteine (10 mM), and glycine (10 mM) were added to the membrane suspension to avoid cysteinyl-LT metabolism.

Reversed-Phase HPLC

Before use, labeled and unlabeled leukotriene purity always was assessed by reversed-phase HPLC. Only leukotrienes with a purity grade 90% were used. The Beckman Instruments (Columbia, MD) HPLC system was equipped with a 110B Solvent Delivery Module, ODS Ultrasphere C18 column (5 µm, 4.6 mm × 25 cm), and a Programmable Detector Module 166 set at 280 nm. Both labeled and unlabeled leukotrienes were eluted isocratically with a filtered and degassed mixture of CH₃OH/H₂O/CH₃COOH (65:35:0.02), adjusted at pH 5.8 with NH₄OH, at a flow rate of 1 ml/min. To check the purity of tritiated leukotrienes, fractions were collected every 30 sec, and the radioactivity profile was assessed by liquid scintillation counting (Ultima Gold; Packard, Meriden, CT).

Binding Studies

Equilibrium binding studies were performed at 25° for 30 min with 0.02-0.5 nM [³H]LTC₄ or [³H]LTD₄ and unlabeled homologous and heterologous ligands at the indicated concentrations. Heterologous competition curves were performed with 0.5 nM labeled ligand. HLPM (0.25 mg/sample), 10 mM HEPES-KOH, pH 7.4, 1 mM CaCl₂, and 10 µM S-decyl-GSH (unless otherwise indicated) were added to the incubation mixture to achieve a final volume of 250 µl. Gpp(NH)p was used at a concentration of 30 µM where indicated. Unbound ligand was separated from bound ligand by rapid vacuum filtration (Brandel Cell Harvester) onto glass-fiber GF/C filters soaked in 2.5% polyvinyl alcohol, and the filters were washed twice with 4 ml of HEPES buffer at 4°. Radioactivity was then measured in a liquid scintillation counter (Filter Count; Packard). Nonspecific binding was 35-50% and 25-30% of the total binding (at 0.5 nM labeled ligand) for LTD₄ and LTC₄, respectively. It was calculated by LIGAND as one of the unknown parameters of the model. Each experiment was performed at least three times in triplicate.

Protocol Optimization and Computer Analysis

The program DESIGN (Rovati et al., 1988) was used to optimize the binding protocols by selecting the lowest number of most appropriate concentrations in mixed-type curves (Rovati et al., 1991) and multiligand experiments (Rovati et al., 1990).

Mixed-type curves. The first three to five concentrations (0.02-0.5 nM) in the curves were obtained using increasing concentrations of labeled ligand (saturation part of the curve), whereas the last three to five concentrations (1 nM to 1 µM) were obtained by adding increasing concentrations of unlabeled ligand to a fixed concentration of labeled ligand (homologous competition part of the curve). By effectively combining both saturation and competition protocols in a single curve, high concentrations of the ligands can be reached without consumption of excessive amounts of labeled ligand (competition part of the curve), yet there will be adequate radioactivity in the lower concentration range (saturation part of the curve). This type of protocol is useful when dealing with labeled ligands with a low specific activity (typically tritiated ligands) that interact with high affinity binding sites (Rovati, 1993).

Multiligand protocols. With this type of protocol, the concentrations of both labeled and unlabeled ligands can be varied simultaneously. By definition, these designs allow the use of any combination of two (or more) ligands in each reaction tube; therefore, a multiligand design potentially includes all possible combinations of concentrations of labeled and unlabeled ligand. This three-dimensional binding "surface" may be regarded either as a family of dose-response curves for ligand 1 in the presence of increasing concentrations of ligand 2 or as a family of dose-response curves for ligand 2 in the presence of increasing concentrations of ligand 1. We performed a series of saturation or mixed curves (e.g., with [³H]LTD₄), each in the presence of a fixed concentration of a second unlabeled compound (LTC₄ or an antagonist). In fact, to study an unlabeled ligand (regardless of its K_i value) with a classic heterologous competition curve, a concentration of the labeled ligand must be used that is lower than its K_d, or the so-called self-displacement by the same labeled ligand occurs, thus preventing interaction of the unlabeled ligand with that site (Rovati, 1993). On the other hand, such a low concentration of labeled ligand might yield an amount of bound radioactivity extremely low. To overcome this limit, we extensively used the multiligand protocols.

Computer analysis. Analysis of binding data was performed using the program LIGAND (Munson and Rodbard, 1980). The computerized analysis of the data through the use of LIGAND has several advantages: it allows (1) analysis of the mixed-type curves and multiligand designs; (2) simultaneous analysis of several experiments, thus pooling information from different curves; (3) testing of different models of increasing complexity (i.e., one-site, two-site, possible cross-reactivity, and so on) and selection of the most appropriate model using the F test for the extra sum of square principle (Draper and Smith, 1966); (4) direct calculation of K_d and K_i values without any further approximation; and (5) generation of curves according to the model selected. The final model of leukotriene interaction was chosen among several others of lesser or greater complexity according to this principle. A value of p < 0.05 was accepted as indicating statistical significance.

GTPase Activity

The following standard conditions were used to measure the release of ³²P_i from [-³²P]GTP on the basis of the method of Cassel and Selinger (1976). The assay system contained 1 µM [-³²P]GTP (5 Ci/mmol), 2 mM MgCl₂, 1 mM 5'-adenylylimidophosphate, 0.5 mM ATP, 10 mM creatine phosphate, 10 mM creatine phosphokinase, 1 mM dithiothreitol, 0.1 mM EDTA, 0.5 mM ouabain, 150 mM NaCl, and 10 mM Tris·HCl, pH 7.4, in a final volume of 0.1 ml. Low affinity hydrolysis of [-³²P]GTP (high K_m GTPase activity) was assessed by incubating parallel tubes in the presence of 50 mM GTP and was subtracted from the total GTPase activity to calculate high affinity GTPase hydrolysis. The reaction was initiated by the addition of HLPM (0.005 mg of protein) to the mixture at 37° and terminated after a 20-min incubation through removal of the tubes to ice for 3 min and the addition of 0.5% (w/v) active charcoal in H₃PO₄, pH 2.2-2.3, giving a total volume of 1 ml. The ³²P_i formed was separated from the nonhydrolyzed nucleotide-bound phosphate by centrifugation at 11,000 × g for 5 min. Aliquots (0.5 ml) of the supernatant were removed form each tube and added to scintillation liquid (Ultima Gold; Packard) for radioactivity counting.

	Results

Top Summary Introduction Procedures Results Discussion References

[³H]LTC₄ and [³H]LTD₄ binding. [³H]LTC₄ and [³H]LTD₄ mixed-type curves (see Experimental Procedures) were performed in the absence and presence of 10 µM S-decyl-GSH (Fig. 1). In the absence of S-decyl-GSH, only one low affinity class of sites was detectable in most of the experiments with [³H]LTC₄ (Fig. 1A), whereas in the presence of S-decyl-GSH, at least two classes of sites were present: one with high affinity and low capacity and a second with low affinity and high capacity (Table 1). At variance with [³H]LTC₄, [³H]LTD₄ binding was practically unaffected by the presence of S-decyl-GSH (Fig. 1B and Table 1). On the basis of these data, 10 µM S-decyl-GSH was included routinely in LTC₄ receptor binding assay.

View larger version (16K): [in this window] [in a new window]   Fig. 1.   Effect of S-decyl-GSH on [3H]LTC4 and [3H]LTD4 binding. A, Mixed-type curve of [3H]LTC4 in the absence () and presence () of 10 µM S-decyl-GSH. B, Mixed-type curve of [3H]LTD4 in the absence () and presence () of 10 µM S-decyl-GSH. Dotted lines, ±95% confidence limits. Data are mean of three replicates from a single experiment, representative of at least two other experiments.

View larger version (17K): [in this window] [in a new window]   Fig. 2.   Homologous and heterologous curves of [3H]LTC4 and [3H]LTD4. A, Mixed-type curve of [3H]LTC4 () and heterologous competition curves of LTD4 () and LTE4 (). B, Mixed-type curve of [3H]LTD4 () and heterologous competition curves of LTC4 () and LTE4 (). Dotted lines, ±95% confidence limits and are shown only for the control curve, for the sake of clarity. Data are mean of three replicates from one to three different experiments.

View larger version (16K): [in this window] [in a new window]   Fig. 3.   Multiligand curves of [3H]LTC4 and [3H]LTD4. For the sake of clarity, only the effect of the highest concentration of unlabeled ligand (10 nM) is shown. A, Multiligand curve of [3H]LTC4 in the absence () and presence () of 10 nM LTD4. B, Multiligand curve of [3H]LTD4 in the absence () and presence () of 10 nM LTC4. Dotted lines, ±95% confidence limits. Data are mean of three replicates from a single experiment, representative of at least two other experiments.

View larger version (19K): [in this window] [in a new window]   Fig. 4.   Theoretical model for the interaction of a ligand, with the binding characteristic of LTD4 (Ki1 = 7.7 nM; Ki2 = 46 nM), with the high and low affinity binding sites labeled by another ligand with the binding characteristics of [3H]LTC4 (Kd1 = 0.015 nM; Kd2 = 105 nM).

View larger version (15K): [in this window] [in a new window]   Fig. 5.   Effect of Gpp(NH)p on [3H]LTC4 and [3H]LTD4 binding. A, Mixed-type curve of [3H]LTC4 in the absence () and presence () of 30 µM Gpp(NH)p. B, Mixed-type curve of [3H]LTD4 in the absence () and presence () of 30 µM Gpp(NH)p. Dotted lines, ±95% confidence limits. Data are mean of three replicates from a single experiment, representative of at least two other experiments.

Antagonist binding studies. Competition curves were performed with two antagonists of different structural classes, ICI 198,615 (indole derivative) and SKF 104353 (leukotriene analog), using [³H]LTD₄ as labeled ligand (Fig. 6A). The responses, albeit varying in potencies, are monophasic, suggesting an interaction with a homogeneous class of binding sites (Table 4). We also performed multiligand experiments with [³H]LTD₄ in the absence and presence of fixed concentrations of SKF 104353, revealing that this compound is able to interact with both the high and low affinity binding sites labeled by [³H]LTD₄ with the same affinity (Fig. 6B and Table 4).

View larger version (16K): [in this window] [in a new window]   Fig. 6.   Antagonist binding curves versus [3H]LTD4. A, Heterologous competition curves of SKF 104353 () and ICI 198,615 (). B, Multiligand curves of [3H]LTD4 in the absence () and presence () of SKF 104353 at the indicated concentrations. Dotted lines, ±95% confidence limits are shown only for the control curve, for the sake of clarity. Data are mean of three replicates from a single experiment, representative of at least two other experiments.

View larger version (17K): [in this window] [in a new window]   Fig. 7.   Antagonist binding curves versus [3H]LTC4. A, Heterologous competition curves of SKF 104353 () and ICI 198,615 (). B and C, Multiligand curves of [3H]LTC4 in the absence () and presence () of either ICI 198,615 (B) or SKF 104353 (C) at the indicated concentrations. Dotted lines, ±95% confidence limits and are shown only for the control curve, for the sake of clarity. Data are mean of three replicates from a single experiment, representative of at least two other experiments.

GTPase activity. GTPase activity was assayed in HLPM in response to 100 nM LTC₄ or LTD₄. LTD₄, but not LTC₄, was able to significantly (p < 0.05) stimulate the receptor-induced hydrolysis of GTP (Fig. 8A). Furthermore, the receptor antagonist SKF 104353 (10 µM) was able to completely inhibit LTD₄-induced stimulation (Fig. 8A), and LTD₄ stimulated the GTPase activity in a dose-dependent manner (Fig. 8B).

View larger version (32K): [in this window] [in a new window]   Fig. 8.   Effect of LTD4 and LTC4 on GTPase activity in human lung parenchyma membranes. A, Effect of 0.1 µM LTD4, 0.1 µM LTC4, and 10 µM SKF 104353 plus 0.1 µM LTD4. B, Dose-response curve for LTD4. Data are expressed as mean ± standard deviation of seven replicates from a single experiment, representative of at least two other experiments. *, p < 0.05 versus basal GTPase activity (one-way analysis of variance).

	Discussion

Top Summary Introduction Procedures Results Discussion References

Despite the efforts by many researchers, no Cys-LT receptor has been successfully purified or cloned. This might be due to the fact that these receptors are present in such a scarce amount that even their detection by binding techniques is sometimes difficult. Moreover, it is well known that in cellular membranes, LTC₄ binds predominantly to a number of nonreceptor sites, such as glutathione-S-transferase (Metters et al., 1994; Sun et al., 1986), LTC₄ synthase (Nicholson et al., 1992), or transport proteins (Keppler, 1992). Therefore, we used S-decyl-GSH, a ligand structurally similar to LTC₄ but devoid of either agonist or antagonist activities (Norman et al., 1987; Sala et al., 1990), to inhibit binding to such nonreceptor sites and thus unmask a putative high affinity LTC₄ receptor. In fact, because LTC₄ is able to contract human isolated lung strips with a potency similar to LTD₄ (Gardiner and Cuthbert, 1988), it is likely that LTC₄, as well as LTD₄, should have a high affinity binding site.

Although only one low affinity/high capacity site was detectable in most experiments in the absence of S-decyl-GSH (Fig. 1A), in its presence two sites became detectable. One of them is a previously undetected high affinity site for [³H]LTC₄, with a K_d value in the picomolar range (as the one for [³H]LTD₄), compatible with the hypothesis that this binding protein is a receptor. It is unlikely that this site is one of the enzymes that usually are abundant and for which LTC₄ has a K_d values in the range of tens of nM (Metters et al., 1994), as reported in practically all the previous [³H]LTC₄ binding studies (Pong et al., 1983; Nicosia et al., 1984; Rovati et al., 1985; Civelli et al., 1987; Norman et al., 1987). Indeed, it should be considered that the low affinity binding site for LTC₄ could correspond to the ATP-dependent LTC₄ export pump. This MRP gene-encoded protein is strongly expressed in normal lung parenchyma (Narasaki et al., 1996) and has a K_m value of 97 nM (Leier et al., 1994) that is compatible with the K_d value of 105 nM (Table 2) estimated for the low affinity binding site of LTC₄. Moreover, this site has a B_max value that is 14% of the value obtained in the absence of S-decyl-GSH, indicating that S-decyl-GSH is able to inhibit most of the nonreceptor binding. However, we cannot completely rule out the hypothesis that this site might be a mixture of LTD₄ receptors and nonreceptor proteins.

As expected for a ligand with low affinity for enzymes such as glutathione-S-transferase (Sun et al., 1986), [³H]LTD₄ binding was basically unaffected by the presence of S-decyl-GSH. Indeed, in both the absence and presence of S-decyl-GSH, a two-site model was identified without any statistical difference in the K_d and B_max values (Fig. 1B and Table 1). Therefore, we can conclude that [³H]LTC₄ and [³H]LTD₄ each recognizes two classes of binding sites.

To investigate whether these sites coincide or are separate entities, a series of heterologous competition curves using LTC₄ and LTD₄ were performed. Both LTC₄ and LTD₄ heterologous competition curves appear to be monophasic (Fig. 2), suggesting that each ligand is able to compete with comparable affinities with the low affinity sites labeled by the other one (Table 2). However, in the case of low specific activity ligands (see Experimental Procedures), this type of experimental protocol does not allow the study of the influence of the nonlabeled heterologous ligand on the high affinity sites labeled by the labeled ligand (Rovati, 1993). Therefore, heterologous competition curve can provide information only on the low affinity site.

Hence, we applied a multiligand protocol to study the possible influence of LTC₄ on the high affinity sites labeled by [³H]LTD₄ and vice versa (Fig. 3). Computerized analysis of the data indicated that LTC₄ is able to inhibit [³H]LTD₄ binding to its high affinity sites and allowed calculation of of a K_i value 520-fold higher than the K_d of [³H]LTC₄ for its own high affinity site (Table 2). If LTC₄ and LTD₄ share the high affinity site, then the K_d and K_i values of LTC₄ should have been the same. In the same way, LTD₄ is able to inhibit [³H]LTC₄ binding to its high affinity sites, with a K_i value 330-fold higher than the K_d of [³H]LTD₄ for its own high affinity sites. Interestingly, computer simulation of the proposed model yielded theoretical curves (Fig. 4) that were almost superimposable with the experimental ones (Fig. 3), thus validating the conclusion drawn from analysis of the real data.

These results strongly suggest that the high affinity site labeled by [³H]LTC₄ is a different entity from the site labeled by [³H]LTD₄ and that there is cross-reactivity between LTD₄ and LTC₄ high affinity sites. Indeed, it is known from functional studies that in some tissues, so-called classic LTD₄ antagonists failed to block the effect of high doses of LTD₄ (Gardiner et al., 1990). This provides evidence that LTD₄ response includes an antagonist resistant component in some tissues (Gardiner et al., 1994) that might represent LTD₄ interaction with LTC₄ receptor.

Heterologous and multiligand curves of LTE₄ were also performed using [³H]LTC₄ and [³H]LTD₄ as labeled ligands. Although LTE₄ is able to compete for both LTD₄ binding sites, it seems to recognize only the low affinity LTC₄ binding site. Moreover, there is 1 order of magnitude difference in the K_i values for LTD₄ and LTC₄ (Table 2); therefore, the LTE₄ response also is in agreement with the hypothesis that two distinct receptors exist in HLPM: one more specific for LTD₄/LTE₄ and one more specific for LTC₄.

To investigate the nature of LTC₄ and LTD₄ putative receptors in HLPM, we studied their potential coupling to a G protein by performing mixed binding curves in the absence and presence of Gpp(NH)p, a stable GTP analog. Gpp(NH)p was able to completely shift [³H]LTD₄ binding from the high affinity site to its low affinity site (Fig. 5B), suggesting that these sites represent two interconvertible affinity states of the same receptor and confirming that [³H]LTD₄ binds to a G protein-coupled receptor (Mong et al., 1986; Crooke et al., 1989). On the other hand, the same Gpp(NH)p concentration was not able to substantially affect [³H]LTC₄ binding (Fig. 5A). The tiny but reproducible effect of Gpp(NH)p on [³H]LTC₄ binding to its high affinity sites might be due to modulation by the GTP analog of the binding of [³H]LTC₄ to LTD₄ receptor.

To confirm further these data, GTPase activity was assayed in response to both LTC₄ and LTD₄. LTD₄, but not LTC₄, is able to significantly stimulate GTP hydrolysis over the basal value (Fig. 8A). Furthermore, LTD₄ response is concentration dependent (Fig. 8B), and the receptor antagonist SKF 104353 is able to significantly inhibit LTD₄-induced stimulation of the enzyme (Fig. 8A). Therefore, these data indicate that LTD₄ receptor is indeed coupled to a G protein, whereas the putative specific LTC₄ receptor in HLPM is not coupled to any G protein. A further, but unlikely, hypothesis is that LTC₄ binds to a G protein-coupled receptor (different from that of LTD₄) but behaves as an antagonist, thus being unable to activate GTPase and displaying a GTP-insensitive binding.

A number of antagonists are available for research, and some are undergoing clinical evaluation. We used two of them to study the pharmacological profile of the putative LTD₄ and LTC₄ receptors. As pure antagonists, they should not distinguish between the high and low affinity state of a G protein-coupled receptor, such as that for LTD₄ (De Lean et al., 1980; Lefkowitz et al., 1993). In fact, the multiligand experiment performed with SKF 104353 (Fig. 6B) confirmed that as expected, this compound interacts with both the high and low affinity states of the LTD₄ receptor with the same affinity. Furthermore, the dissociation constant computed only from the competition curve (Fig. 6A) and the one computed from the multiligand curves are substantially identical (Table 4), making multiligand experiments unnecessary in the case of pure antagonists.

When tested versus [³H]LTC₄ with the classic competition curves (Fig. 7A), only SKF 104353 (not ICI 198,615) has a K_i value lower than 10 µM (Table 4), the value we consider the upper limit for a biologically relevant interaction. For the reasons we discussed, we used a multiligand protocol to study the possible influence of the antagonists on the high affinity sites labeled by [³H]LTC₄ (Fig. 7, B and C). This experiment allowed us to reveal that SKF 104353 is able to compete with both the sites labeled by [³H]LTC₄ with markedly different affinities. Moreover, the affinity for the low affinity site computed from the multiligand experiment is not significantly different from that calculated from the pure competition curve (Table 4). On the contrary, ICI 198,615 is unable to compete with either of the sites labeled by [³H]LTC₄ (Fig. 7B). It is clear from Figs. 6 and 7 that there is a striking difference in the behavior of such antagonists versus [³H]LTD₄ or [³H]LTC₄, again supporting the idea that two different receptors might exist in HLPM.

Taken together, these data indicate that in human lung parenchyma (Fig. 9), (1) [³H]LTD₄ and [³H]LTC₄ recognize two different binding sites, a high affinity one (R_1H and R_2H) and a low affinity one (R_1L and S_2L); (2) the two high affinity binding sites are different entities, albeit there is cross-reactivity between LTD₄ and LTC₄; (3) for [³H]LTD₄, the two binding sites represent the interconvertible high and low affinity states of a G protein-coupled receptor; (4) for [³H]LTC₄, the low affinity binding site probably consists of a mixture of proteins, including the low affinity state of the LTD₄ receptor (R_1L), specific enzymes, and transport systems (S_2L). The high affinity binding site (R_2H) might be a specific LTC₄ receptor. Indeed, as mentioned, functional data indicate the existence of a receptor for LTC₄ in human lung parenchyma (Gardiner and Cuthbert, 1988), to which a high affinity binding site must correspond.

View larger version (15K): [in this window] [in a new window]   Fig. 9.   Final model for Cys-LT receptors in human lung parenchyma.

In conclusion, the characterization of the Cys-LT receptors in human lung parenchyma, which is important in the pathogenesis of asthma, may foster the identification of a novel cysteinyl-LT antagonist with an ideal pharmacological profile to be used in the therapy of this chronic disease. The existence of a specific receptor for LTC₄ in this tissue implies that an ideal antagonist should be able to recognize and antagonize not only the LTD₄ receptor but also that specific for LTC₄. In fact, all the antagonists developed up to now, including the dual antagonist BAY u9773, are unable to completely reverse the effect of high doses of LTD₄ or LTC₄ in some tissues (Gardiner et al., 1994), such as human lung, thus limiting the therapeutical efficacy of this class of drugs.