Transmembrane Helix 7 of the Endothelin B Receptor Regulates Downstream Signaling*

Endothelin is a 21-amino acid peptide with a striking diversity of important biological responses, including, vasoconstriction, bronchoconstriction, and mitogenesis. Endothelin-1 binding to the endothelin B receptor (ETB), a member of the superfamily of G-protein-coupled receptors, was associated with catalytic activation of the extracellular-regulated kinase 2 (ERK2) and stimulation of AP-1 transcriptional reporter activity. A panel of single point mutations in transmembrane helix 6 (TM6), intracellular loop 3, and transmembrane helix 7 (TM7) were developed to study the structural requirements for ETB activation. Point mutations within highly conserved regions of TM6 and intracellular loop 3 were without effect on agonist-stimulated ERK activation. However, mutations within TM7 of the ETB significantly impacted ligand-stimulated downstream signaling. For example, nine point mutations within TM7 of the ETB were identified that prevented endothelin-stimulated ERK activation. Interestingly, the TM7 mutants fell into two classes; several exhibited greatly decreased AP-1 activity, relative to wild type ETB, whereas others displayed augmented endothelin-stimulated AP-1 transcriptional activity relative to wild type ETB. Our results suggest that TM7 of the ETB is involved in its activation mechanism and regulates agonist-stimulated ERK activation.

The endothelin receptors are members of a superfamily of G-protein-coupled receptors that are thought to activate downstream signaling networks through their dynamic interaction with heterotrimeric G-proteins. Mutational studies of the endothelin receptors have defined extracellular loop 2 as being critical for ligand contact and binding, whereas the cytoplasmic tail has been implicated as important for coupling to G-proteins (15)(16)(17).
The relationship between ligand binding affinity and receptor-G-protein interaction is described by the ternary complex model of receptor activation (18). It is widely thought that a small fraction of receptors reside in the activated conformation and that agonist binding stabilizes this conformation, which results in activation of the associated G-protein. The dynamic interaction of ligand-receptor-G-protein is thought to result in high affinity agonist binding when the receptor is in the active, G-protein bound state. Therefore, mutations that diminish or prevent receptor-G-protein binding in some cases are associated with decreased agonist binding. In contrast, several constitutively active GPCR mutants have been identified that result in higher affinity ligand binding (17,19,20).
Although the molecular interactions that govern GPCR activation have not been completely defined, mutagenesis studies, biophysical analysis, and cryoelectron microscopy of rhodopsin, a well characterized GPCR, have provided a comprehensive model for GPCR activation (21)(22)(23). Transmembrane helices 3 and 6 are thought to be packed into close proximity in threedimensional space, and ligand binding results in rigid body movement of TM6 relative to TM3. This displacement of helix 6 relative to helix 3 is hypothesized to induce a conformational change in the loop connecting TM5 and TM6, which is where G T is associated with the receptor (24).
The importance of TM6 has been confirmed through mutagenesis studies of other GPCRs. For example, the ␣ 1B -adrenergic receptor can be constitutively activated by mutation of alanine 293, located on the intracellular side of TM6 (25), and the m5 muscarinic receptor can be constitutively activated by mutations at the extracellular side of TM6 (20).
Taken together the data imply that TM6 is important in the activation mechanism of the adrenergic and muscarinic GPCRs. We have therefore examined the TM6 region of the ETB to define the importance of this domain in ERK activation and AP-1 transcriptional control. Our results indicate that the ETB is distinct from the muscarinic and adrenergic receptors in this regard, because mutation of several highly conserved amino acids within or near TM6 had little or no effect on activation of ERK mitogen-activated protein kinase. In contrast, we have observed that TM7 of the ETB is critical for proper signaling functions of the receptor. Point mutations within the TM7 region of the ETB interfered with ligand-stimulated ERK activation and had dual effects on AP-1 transcriptional activation.

Cell Culture and Transfection
COS cells were grown in Dulbecco's modified Eagle's medium (DMEM) containing 10% fetal bovine serum (FBS). For kinase assays cells approximately 50% confluent in 100-mm plates were transfected with the indicated plasmids (5 l/g DNA) using LipofectAMINE (Life Technologies, Inc.). Cells were exposed to the LipofectAMINE/DNA mixture for 6 h in DMEM, the medium was changed to DMEM/10% FBS, and the cells were incubated for 48 h. Cells were serum starved by changing the medium to DMEM/0.1% FBS for 16 h, and the cells were treated with endothelin and harvested for immunoprecipitation. For immunofluorescence COS cells were grown and transfected in 0.2-m glass slides. For AP-1 reporter assays the AP-1/luciferase plasmid was co-transfected with wild type and mutant ETB plasmids and pCMV-␤gal, which encodes ␤-galactosidase at a ratio of 5:1.

Plasmid Construction
ERK2 containing a 6-fold repeated Myc epitope fused in frame to the 5Ј end was used as described previously (17). Wild type rat endothelin B receptor (GenBank TM accession number X57765) was subcloned into pCDNA3 containing a FLAG epitope fuse in frame at the 3Ј end as described previously (17). Single point mutations were made in TM6, TM7, and the third intracellular loop of the rat ETB using the Transformer site-directed mutagenesis kit (CLONTECH) following the manufacturer's instructions. All mutations were verified by sequencing. The AP-1/lucierase plasmid contains the AP-1 transcriptional promoter element cloned upstream of the luciferase gene (kindly provided by Dr. Mercedes Rincon, University of Vermont).

Mutagenesis
Single point mutations were introduced into the wild type rat ETB using the Transformer site-directed mutagenesis kit following the manufacturer's instructions (CLONTECH). Briefly, a selection primer was engineered that changed a unique SmaI site in the pcDNA3 vector at nucleotide 2093 to a XhoI site that was also unique. Additional primers were designed that changed the desired amino acids in the ETB. After synthesizing the mutant strand in vitro the plasmids were transformed into a mutS Escherichia coli strain. Parental plasmids were eliminated by digestion with SmaI, and the pool of plasmids was transformed again. Colonies were picked, and the mutations were confirmed by sequencing.

Luciferase Reporter Assay
COS cells were grown in 6-well plates in DMEM containing 10% FBS. At 50% confluence, cells were transfected with 0.5 g of ETB DNA, 0.25 g of AP-1/luciferase reporter plasmid, and 0.1 g of ␤-gal using LipofectAMINE in serum-free DMEM. After 5 h the medium was changed to growth medium, and the cells were grown for 24 h. Cells were changed to 0.1% FBS serum, incubated overnight, and then stimulated with endothelin for 6 h. The receptors were transfected into three individual culture dishes, and the values are the means Ϯ S.D. of the three independent measurements. Cell lysates were made, and luciferase activity was measured in a scintillation counter according to the manufacturer's specifications (Promega Corp., Madison, WI). Luciferase activity was normalized to ␤-gal activity.

Immunofluorescence
Cells were grown in 2-well Nunc chambered slides until approximately 50% confluent and transfected with 1 g of ETB DNA and 3 l of LipofectAMINE, and the medium was changed to 10% FBS 5 h later. Cells were rinsed twice with phosphate-buffered saline, fixed with 4% paraformaldehyde, and permeabilized with MeOH. Nonspecific binding was blocked with 5% normal goat serum, and the cells were incubated for 1 h at room temperature in the anti-flag antibody (M2, Kodak Corp. New Haven, CT) at 2 g/ml. The cells were rinsed and blocked again for 30 min in 5% normal goat serum. Cells were washed five times over 30 min in TBS-T (Tris-buffered saline with 0.1% Tween 20), incubated for 45 min in the dark at room temperature in a CY3-conjugated, affinity purified, goat anti-mouse secondary antibody (1:2000, Jackson Immu-noResearch), and mounted with VECTASHIELD (Vector laboratories Inc., Burlingame, CA) containing DAPI.

Kinase Assays
COS cells co-transfected with wild type or mutant ETBs and MT-ERK2 were treated with endothelin (100 nM) for 10 min. Cells were scraped into 600 l of 1% Nonidet P-40, 10 mM Hepes, 2 mM EDTA, 50 mM NaF, 0.1% ␤-mercaptoethanol, 1% aprotinin, 50 M microcystin, 0.2 mM sodium orthovanadate. Insoluble material was removed by centrifugation at 15,000 rpm for 10 min at 4°C. The Myc-tagged ERK2 was immunoprecipitated using monoclonal antibody 9E10. The immunoprecipitates were washed and incubated in 20 mM Hepes, pH 7.5, 10 mM MgCl 2 , 2 mM dithiothreitol) in the presence of myelin basic protein (MBP) and [␥ 32 P]ATP (5 Ci) for 15 min. The extent of phosphorylation of MBP was analyzed by running the immunoprecipitates on a 15% SDS gel and subsequent autoradiography.

Ligand Binding
Binding to Purified Membranes-COS cells transfected with empty vector, wild type ETB, or ETB mutants were lysed in 10 mM Hepes, pH 7.5, 2 mM MgCl 2 , 5 mM KCl, and the post-nuclear supernatants were centrifuged at 100,000 ϫ g for 1 h to isolate the membrane fraction. COS cell membranes (10 g) were incubated with 50 pM 125 I-endothelin-1 for 3 h. The membranes were applied to Whatman GF/B filters and washed extensively prior to counting. Nonspecific binding was estimated by incubating the 125 I-endothelin-1 in the presence of excess (1 M) unlabeled endothelin. Nonspecific binding averaged 15-25% of specific binding. Binding of endothelin to ETB mutant receptors is expressed relative to the binding observed in the membranes expressing the wild type receptor.

Model of ETB topology and alignment of TM6 and TM7.
A, the ETB is a member of the GPCR superfamily that is thought to have seven membrane spanning helices joined by alternating intracellular and extracellular loops. B, TM6 of the ETB has several residues that are found to be highly conserved among different classes of GPCRs. TM7 of these GPCRs does not have highly conserved amino acid residues. The Swiss-Prot accession numbers for the aligned sequences are: rat dopamine D 1 receptor (D1), P35406; rat muscarinic receptor (M2), P10980; rat endothelin B receptor (ETB), P21451; and rat ␤ 1 adrenergic receptor (B1AR), P18090. and lysed in 0.1 N NaOH. Lysates were collected, and 125 I-ET content determined by liquid scintillation ␥ counting using a Packard B5005 ␥ counter.

RESULTS
Because direct structural information is not available for GPCRs, models regarding the activation mechanism of this family of receptors have relied primarily on mutagenesis and biophysical studies. Secondary structure models have provided insight into the topology of rhodopsin, including the packing of the transmembrane helices (21,22). Recent work on rhodopsin indicates that the activation mechanism involves rigid body movement of TM6 (23). The primary sequence of the GPCR superfamily reveals several highly conserved amino acids in this helix that are also present in the ETB (Fig. 1). We wished to determine whether the activation mechanism of ETB involves TM6. Therefore, we made substitutions of several highly conserved residues in this region.
TM6 Mutants of the ETB-Our results indicate that although several residues within TM6 are almost invariably conserved among the GPCR superfamily, mutation of these residues in the ETB appeared to be without effect on ERK mitogenactivated protein kinase activation (Fig. 2). ETB mutants F331A, C334A, W335A, P337A, S341K, and K345A were cotransfected into COS cells with MT-ERK2. Endothelin stimulated ERK2 catalytic activation in the wild type ETB and mutant receptors at approximately similar levels. The basal level of ERK activity for unstimulated mutant receptors was also unchanged relative to wild type receptor (Fig. 2B). Because TM6 of the ETB is not sensitive to structural changes introduced by mutagenesis, the data infer that this helix is not integrally involved in the activation mechanism of this receptor.
Intracellular Loop 3 Mutants of the ETB-Several studies have implicated the third intracellular loop of GPCRs, such as the adrenergic and muscarinic receptors, as playing a role in downstream effector function (20,25). Therefore, we also examined the third loop of the ETB by introducing amino acid substitutions into this area and expressing the mutant receptors in COS cells with MT-ERK2. The third loop of the ETB appeared to be relatively unaffected by mutation. ETB mutants M299A, T323A, M306A, and D312A had essentially wild type levels of ERK activation, whereas ETB mutants Q316A and K322A were slightly reduced in their ability to activate ERK ( Fig. 3) As with the TM6 region of the ETB, analysis of the basal activity of ERK in cells transfected with mutant receptors revealed levels of activity comparable with those of unstimulated wild type receptors (Fig. 2B).
TM7 Mutants of the ETB-Transmembrane helix 7 of the ETB shares relatively little homology with other GPCRs, and in fact the GPCR superfamily has little overall homology in this region (Fig. 1). However, TM7 of the ETB was very sensitive to mutation. A panel of TM7 mutants were engineered and coexpressed in COS cells with MT-ERK2. ETB mutants M373K, N377P, S378K, I380A, P382K, L385A, V388K, and K390A were The wild type and mutant receptors (1 g) were co-transfected into COS cells with MT-ERK2 (1 g). The transfected cells were stimulated with endothelin (100 nM) for 10 min prior to immunoprecipitating MT-ERK2. An in vitro kinase assay was done using MBP as an in vitro substrate to measure catalytic activity. The substitutions did not impair endothelin-stimulated ERK activation. B, basal ERK activity in cells expressing mutant ETBs. COS cells were transfected as above with 1 g of Myc-tagged ERK2 and either wild type or mutant ETB containing site-specific mutations in loop 3, TM6, or TM7. ERK activity was determined by immune complex kinase assay, in vitro. Mutations in ETB did not significantly alter observed basal ERK activity.
all severely compromised in their ability to activate ERK, in comparison with wild type (Fig. 4). Western blotting of the COS cell lysates demonstrated that MT-ERK2 was expressed at similar levels across the experiment, confirming that the decrease in catalytic activity was due to decreased activation of MT-ERK2 and not lower expression of the kinase. Although several TM7 mutant receptors revealed slightly elevated levels of basal ERK activity relative to wild type receptor (Fig. 2B), this was not reflected in an increase in ET-stimulated ERK activity relative to wild type receptors. This suggests that introduced mutations did not produce catalytically active mutants. FIG. 3. Loop 3 of the ETB is not sensitive to mutation. Amino acid substitutions were introduced into the intracellular loop 3 region of the ETB and expressed in COS cells as in Fig. 2. The mutant receptors were activated by ligand binding, and ERK kinase activity was measured by immune complex kinase assay using MBP as an in vitro substrate. Basal activity of Loop 3 mutants is presented in Fig. 2B. The results indicate that the mutant receptors stimulate ERK in response to ET to a similar degree as do the wild type receptors (ETB-WT), suggesting that loop 3 is not sensitive to structural changes with respect to ERK activation.

FIG. 4. TM7 regulates receptor-mediated activation of ERK.
Amino acid substitutions were introduced into the TM7 region of the ETB, and the wild type (ETB-WT) and mutant receptors were co-transfected with MT-ERK2 into COS cells. Kinase assays were done using MBP as an in vitro substrate to measure catalytic activity. Basal activity is presented in Fig. 2B. Several of the point mutations significantly decreased endothelin-stimulated ERK activation. Western blots were done to document that decreases in ERK kinase activity reflected changes in the activation of the enzyme rather than its expression level.

ETB Mutant Receptors Are Expressed and Properly
Localized-Introducing nonconserved amino acid changes in the ETB could reduce its downstream signaling capabilities as a result of a gross alteration in the topology of the receptor. Therefore, we investigated the expression levels and localization of the ETB mutant receptors by immunofluorescence. Receptors bearing a FLAG tag on their carboxyl terminus were expressed in COS cells and processed for immunofluorescence using the anti-FLAG monoclonal antibody (Fig. 5). The mutant receptors were expressed and localized in essentially the same manner as the wild type receptor. The data indicate that the inability of the TM7 mutations to activate ERK resulted from a defect in the activation mechanism, rather than a mutationinduced effect on expression or localization of the mutant receptors.
TM7 of the ETB Regulates AP-1 Transcriptional Activation-Ligand binding to the ETB triggers a diverse spectrum of biological responses, depending on the target tissue in which the receptor is expressed. ETBs stimulate mitogenesis in selective cell types, and this response is thought to involve the activation of key transcriptional events, including AP-1 activity (14). Furthermore, catalytically activated ERK is known to phosphorylate transcription factors involved in AP-1 activity (26) and stimulate transcription. Therefore, we wished to determine whether ERK activation is required for AP-1 transcrip-tional activation. To this end we have examined transcriptional activation of an AP-1 luciferase reporter construct by wild type and mutant ETBs (8). The mutants fell into two distinct classes based on their ability to stimulate AP-1 transcriptional activation. Several of the mutations in TM7 of the ETB significantly decreased endothelin-stimulated AP-1 transcriptional activation. For example, N377P, S378K, P382K, A384K, L385A, K390A, and F392A were compromised in their ability to stimulate AP-1 reporter activity when exposed to ligand. Interestingly, a second class of TM7 mutant was identified that displayed augmented endothelin-stimulated AP-1 transcriptional activity. TM7 mutants D367A, N372A, I383A, V388K, and I380A when stimulated with ligand demonstrated augmented AP-1 activity compared with wild type ETB (Fig. 6). The I380A TM7 mutant had the largest increase in ligand-stimulated AP-1 activity. These mutants were consistently observed to have increased endothelin-stimulated AP-1 activity; however, the basal AP-1 activity was essentially the same as wild type (data not shown). The results suggest that TM7 of the ETB is integrally involved in ligand-induced activation of AP-1 activity because mutations were identified that both prevented and augmented AP-1 responses.
Ligand Binding Properties of the Mutant ETBs-The ternary complex model defines the dynamic interactions between receptor-G-protein interactions and ligand binding (18). High affinity agonist binding is thought to be dependent on efficient G-protein interactions with the receptor. Furthermore, several constitutively activated GPCRs are found to have increased affinity for agonist (19,20,25). The ternary complex model suggests that receptor mutations that decrease or interfere with proper receptor-G-protein physical interactions will result in altered agonist binding.
We conducted binding studies on both purified membranes and transfected, whole cell preparations for each receptor mutant. Our results are consistent with this model. Several of the TM7 mutants displayed a range of 125 I-endothelin binding, which was generally lower than the wild type receptor, suggesting that the mutant receptors were impaired in their ability to couple with G-protein (Table I and Fig. 7). Other mutants displayed binding closer to wild type levels but were unable to efficiently activate ERK or stimulate AP-1 luciferase activity (N377P and S378K mutants; Figs. 6 and 7 and Table I). Inefficient receptor-G-protein interactions may therefore account for the inability of ETB mutants to activate ERK. Furthermore, any decreased binding was not a result of a lack of receptor expression or localization, because the mutant ETBs were expressed and localized in a manner similar to that of the wild type receptor (Fig. 5). DISCUSSION GPCRs form an important and diverse group of cell surface receptor proteins that have yet to be crystallized for structural studies. Mutagenesis experiments involving the major classes of GPCRs have provided a comprehensive picture regarding the molecular details of receptor-ligand interactions and receptor-G-proteins interactions. Studies with rhodopsin suggest a model of activation that involves interaction of TM3 and TM6. In the case of rhodopsin, rigid body movement of TM6 is thought to be induced by agonist binding and to alter the conformation of the binding site for transducin, the associated heterotrimeric G-protein (23).
In our experiments we have introduced a series of point mutations in TM6, intracellular loop 3, and TM7 to define the regions of the ETB that are involved in downstream signaling. Our results provide insight into the activation mechanism of the ETB. The TM6 domain of the ETB has several residues that are highly conserved across a broad spectrum of GPCRs, and studies with rhodopsin implicate this region in the ligandinduced activation mechanism. Therefore, we postulated that these residues may be important in regulating the coupling of the ETB with downstream effectors, such as ERK activation. Surprisingly, systematic substitution of the highly conserved residues of TM6 had no effect on activation of ERK (Fig. 2). Although these residues are highly conserved, this area of the ETB does not appear to be sensitive to structural changes. The FIG. 6. AP-1 transcriptional activation by wild type and mutant ETBs. COS cells were co-transfected with wild type (WT) and mutant ETBs (1 g), AP-1/luciferase reporter plasmid (1 g), and CMV-␤-gal (0.25 g). The transfected cells were serum starved for 48 h followed by endothelin treatment (100 nM) for 6 h. Lysates were made, and the luciferase activity was measured and is expressed as arbitrary units. Each receptor was transfected into three separate dishes, and each bar represents three independent transfections and measurements. The data presented are representative of five independent experiments. The activity of the ␤-gal reporter was measured in each lysate, and the values were normalized to ␤-gal activity, which usually varied between 5 and 15% among replicates. results argue against TM6 as being critically involved in ligand-induced ERK activation. The ETB appears to be distinct from at least two other GPCRs in which TM6 is involved in the activation mechanism. For example, in the case of the m5 muscarinic receptor, substitution of Ser 465 at the extracellular junction of TM6 resulted in activation of the receptor (20). Substitution of Ala 293 of the ␣ 1B -adrenergic receptor, which is in loop 3 close to the junction of TM6, resulted in constitutive activation of the receptor (25). Amino acid residues throughout TM6 and loop 3 of the ETB were mutated without effect on basal or ligand-induced ERK activation. Although these regions in the muscarinic and adrenergic receptors are involved in the activation mechanism of the receptors, this theme is not applicable to the ETB. The intracellular loops of GPCRs form a structure in the cytosol of the cell that may be important in mediating proteinprotein interactions with heterotrimeric G-proteins. In fact, introducing mutations into this area can impair downstream signaling functions (25). In the case of the ETB, the putative intracellular loops are very small, relative to other GPCRs. The intracellular loop 3 of the ETB is the largest and was therefore hypothesized as most likely to be involved in coupling to Gproteins and downstream signaling functions. We therefore systematically introduced point mutations into the ETB loop 3 area and evaluated the ability of the mutant receptors to activate the ERK pathway. Our results indicate that, similar to the TM6 region, loop 3 is not sensitive to the introduction of nonconserved amino acid substitutions. Several mutations were introduced into loop 3 of the ETB with no effect on ligandstimulated ERK activation.
In contrast, TM7 of the ETB proved to be very sensitive to mutation. Nine residues in TM7 were identified that, when substituted with a nonconserved residue, resulted in a significant decrease in ligand-stimulated ERK activation (Fig. 4). Our results differ from the rhodopsin model of GPCR activation to the extent that TM6 is postulated to be critically involved in ligand-induced activation of rhodopsin, whereas this region of the ETB does not appear to be important for ERK activation. In contrast, TM7 of the ETB was very sensitive to mutation, with respect to agonist-induced ERK activation. In the three-dimensional model of helix packing of rhodopsin, helix 7 is close to helix 6. Because helix 6 is thought to move during activation, it is possible that mutations of helix 7 sterically interfere with the ability of helix 6 to be displaced during ligand binding. However, if helix 6 of the ETB were critically involved in activation, a reasonable prediction would be that it should be sensitive to mutagenesis. Therefore, the data suggest the possibilities that the activation mechanism of the ETB is distinct from that of rhodopsin and that helix 7 of the ETB is involved in ligandinduced activation.
Some of the biological effects of ETB activation are thought to involve activation of AP-1 mediated transcription (14). Be- FIG. 7. 125 I-Endothelin-1 binding to COS cells. Ligand binding to whole cells expressing wild type (WT) or mutant ETBs for the TM7 domain was determined by analysis of bound radiolabeled ET at 4°C. COS cells transfected with 1 g of wild type or mutant TM7 receptors were exposed in duplicate to either 50 or 150 pM 125 I-ET for 3 h at 4°C. Cells were washed, and bound ET was determined by scintillation counting. Values are presented as the mean Ϯ S.E. Nonspecific binding was measured on vector transformed cells in the presence of 1 M cold ET. Although most receptors examined displayed reduced ERK activity, which correlated with reductions in ET binding, several receptors maintained ET binding similar to wild type levels.
cause ERK is known to phosphorylate transcription factors leading to increased AP-1 activity (26), we examined the relationship between ETB-stimulated ERK activation and AP-1 transcriptional activation. Seven of the residues in TM7 that were sensitive to mutation, with respect to ERK activation, demonstrated reduced AP-1/luciferase reporter activity, suggesting that AP-1 activation and ERK activity are positively correlated. Interestingly, there were five notable exceptions to this correlation. ETB TM7 mutants D367A, N372A, I383A, V388K, and I380A consistently demonstrated increased endothelin-stimulated AP-1 activity. The I380A TM7 mutant was especially notable because it consistently demonstrated augmented ligand-stimulated AP-1 activation. The basal ERK activity of the receptor mutants was similar to wild type levels with minor elevations in some TM7 mutants (F392A, S378K, p382K, and K390A). However, no positive correlation could be shown between increased basal ERK activity and increased AP-1 transcription (compare Fig. 2B with Fig. 6). With the exception of D367A, the mutants that had increased endothelin-stimulated AP-1 activity were unable to stimulate ERK kinase activity (Figs. 4 and 6). The results suggest that ERK activation and AP-1 activation are separable, and that ERK activity is not an absolute requirement for AP-1 activation. The mechanism whereby AP-1 is activated may involve one or more of the several signal transduction systems known to be activated by endothelin. Additional studies will be required to dissect the layers of interaction between the components of the endothelin signaling pathway.