Functional endothelin receptors are present on nuclei in cardiac ventricular myocytes.

Endothelins are thought to act through two specific, plasmalemmal G protein-coupled receptor subtypes, ETAR and ETBR. However, in subfractionated cardiac membranes, ETAR immunoreactivity was detected only in the plasma membrane whereas ETBR immunoreactivity was detected predominantly in membranes of intracellular origin. Confocal microscopy demonstrated the presence of intracellular ETAR and ETBR in ventricular myocytes. ETAR were primarily on plasma membrane (surface membranes and transverse-tubules) and to a lesser extent on the nucleus while ETBR localized primarily to the nuclei. Western blot analysis of nuclei isolated from the heart indicated the presence of endothelin receptors: both ETAR and ETBR copurified with nucleoporin 62, whereas markers of endoplasmic reticulum and Golgi membranes were depleted. Radioligand binding studies revealed that isolated nuclei contain specific [125I]ET-1 binding sites. Specific [125I]ET-1 binding was reduced by 70-80% using the ETAR-selective antagonist BQ610 and 20-30% using the ETBR-specific antagonist BQ788. IRL-1620, an ETBR-specific agonist, also reduced [125I]ET-1 binding. Furthermore, ET-1 and IRL-1620 altered the incorporation of 32P into nuclear proteins and caused a transient increase in nuclear Ca2+ concentration. Hence, cardiac nuclei possess both ETAR and ETBR subtypes, which are functional with respect to ligand binding and are coupled to signaling mechanisms within the nuclear membrane.

thought to be mediated through its interactions with specific receptors located on the cell surface. However, there is now evidence demonstrating insulin (46), epidermal growth factor (47), nerve growth factor (48), ␤-interferon (49), angiotensin II (50), and prostaglandin (51,52) receptors localize to the perinuclear region or nuclear membrane. In addition, endocytosed ET-1⅐ET B R reduces prepro-ET-1 mRNA levels (53) and cytosolic application of exogenous ET-1 increases nuclear Ca 2ϩ levels (54), raising the possibility of intracellular endothelin receptors regulated by an "intracrine" process. Considering the major role the ET-1 system has in regulating heart function and dysfunction, the objectives of the present study were to determine the existence, localization, and subtype of intracellular endothelin receptors in ventricular myocytes. Our findings reveal: 1) nuclear/perinuclear localization of ET A R and ET B R immunoreactivity; 2) nuclear endothelin receptors are capable of specific ligand binding; 3) stimulation of the nuclear endothelin receptors induce a transient increase in nuclear cisternal Ca 2ϩ content; and 4) stimulation of nuclear endothelin receptors activate endogenous nuclear protein kinase activities. This study establishes the presence of functionally coupled nuclear endothelin receptors in the ventricular myocardium. Fura-2/AM was from Molecular Probes (Eugene, OR). Antisera specific for the ET A R were from Alomone Labs (Jerusalem, Israel) (AER-001, rat ET A R amino acids 413-426, NHNTERSSHKDSMN) and ABCAM (ab1919, rat ET A R amino acids 31-45, SSHVEDFTPFPGTEF). Antisera specific for the ET B R were from Alomone Labs (AER-002, rat ET B R amino acids 299 -314, CEMLRKKSGMQIALND) and Biogenesis Ltd. (Poole, United Kingdom) (number 4113-3059, rat ET B R amino acids 405-417, QTFEEKQSLEEKQ). Anti-nucleoporin p62 (number N43620), annexin II (number A14020), BiP/GRP78 (number G73320), and GM130 (number G65120) were from Transduction Laboratories (Lexington, KY). Anti-caveolin-3 (N-18) was from Santa Cruz Biotechnology (Santa Cruz, CA). TRITC-, Cy5-, and horseradish peroxidaseconjugated secondary antibodies were from Jackson ImmunoResearch Laboratories (West Grove, PA). The chemiluminescence reagent Renaissance Plus was from PerkinElmer Life Sciences. Triton X-100, leupeptin, and phenylmethylsulfonyl fluoride were from Roche Diagnostics. SDS-polyacrylamide gel electrophoresis reagents, polyvinylidene difluoride, nitrocellulose (0.22 m), and Bradford protein assay reagents were from Bio-Rad. Essentially fatty acid-free bovine serum albumin was from Sigma. Unless otherwise stated, all reagents were of analytical grade and were purchased from VWR Canlab (Ville Mont-Royal, Quebec) or Fisher.
Isolation of Ventricular Myocytes-Calcium-tolerant cardiomyocytes were isolated by Langendorff perfusion as described previously (55). This preparation provided 8 to 10 million cells/heart with 70 to 85% viability, as assessed by the presence of quiescent cells with rod-shaped morphology. No other cell types were detected.
Preparation of Canine Cardiac Sarcoplasmic Reticulum-Cardiac microsomes were prepared from fresh ventricular myocardium and subfractionated as described previously (56). Fresh canine ventricles were homogenized in 5 volumes of 20 mM imidazole/HCl buffer (pH 6.8 at 5°C) containing 0.3 M sucrose, 2 mM EGTA, 2 mM EDTA, 5 mM DTT, 5 mM sodium azide, 2 mM phenylmethylsulfonyl fluoride, 10 g/ml leupeptin, 1.4 g/ml pepstatin A, and 5 g/ml aprotinin. Crude membrane vesicles (MV) were then prepared (57) and subfractionated into plasma membrane, longitudinal sarcoplasmic reticulum (SR), and junctional SR by loading with calcium phosphate followed by centrifugation into discontinuous sucrose density gradients (58). Plasma membrane, junctional SR, and longitudinal SR, also referred to as fractions 1, 3, and 5, respectively, were harvested from the 0/0.6 M sucrose interface, 0.8/1.1 M interface, and the pellet at the bottom of the tube. Hence, fractions 2-5 comprise different membranes of intracellular origin. Each fraction was centrifuged for 30 min at 42,000 rpm in a Beckman 50.2 Ti rotor then resuspended in 1 ml of 20 mM histidine/HCl buffer (pH 6.8 at 5°C) containing 0.3 M KCl, 2 mM DTT, and 10 M leupeptin, frozen using liquid nitrogen, and stored at Ϫ70°C.
Isolation of Nuclei-Rat cardiac nuclei were isolated according to a modified version of a previously described method (59). Briefly, rat hearts were pulverized under liquid nitrogen, resuspended in cold PBS and homogenized (Polytron, 8000 rpm; 2 ϫ 10 s). The homogenate is referred to as the total extract (Fraction 1). All subsequent steps were carried out on ice or at 5°C. Homogenates were centrifuged for 15 min at 500 ϫ g and the supernatants, referred to as Fraction 2, were diluted 1:1 with buffer A (10 mM K-HEPES (pH 7.9), 1.5 mM MgCl 2 , 10 mM KCl, 1 mM DTT, 25 g/ml leupeptin, 0.2 mM Na 3 VO 4 ), incubated 10 min on ice, and centrifuged for 15 min at 2000 ϫ g. The resulting supernatant was discarded and the pellet, referred to as the crude nuclei (Fraction 3), was resuspended in buffer B (0.3 M K-HEPES, pH 7.9, 1.5 M KCl, 0.03 M MgCl 2 , 25 g/ml leupeptin, 0.2 mM Na 3 VO 4 ), incubated on ice for 10 min, and centrifuged for 15 min at 2000 ϫ g. The supernatant was removed and the pellet, referred to as the enriched nuclear fraction (Fraction 4), was resuspended in buffer C (20 mM Na-HEPES (pH 7.9), 25% (v/v) glycerol, 0.42 M NaCl, 1.5 mM MgCl 2 , 0.2 mM EDTA, 0.2 mM EGTA, 0.5 mM phenylmethylsulfonyl fluoride, 0.5 mM DTT, 25 g/ml leupeptin, 0.2 mM Na 3 VO 4 ) and either used fresh or aliquoted, snapfrozen using liquid nitrogen, and stored at Ϫ80°C. Sheep nuclei were isolated as described previously (60).
Preparation of Rat Brain Membranes-Membranes were prepared from fresh rat brains by homogenizing in 5 volumes of 20 mM imidazole/ HCl buffer (pH 6.8 at 5°C) containing 0.3 M sucrose, 2 mM EGTA, 2 mM EDTA, 5 mM NaN 3 , 2 mM phenylmethylsulfonyl fluoride, 10 g/ml leupeptin, 1.4 g/ml pepstatin A, and 5 g/ml aprotinin using a 25-ml Potter-Elvehjem tissue grinder (10 strokes). Homogenates were cleared of cellular debris by centrifugation for 10 min at 1000 ϫ g and 5°C and then membranes were pelleted by centrifugation for 60 min at 100,000 ϫ g and 5°C. Membranes were resuspended in 1 volume of homogenization buffer, aliquoted, snap-frozen using liquid nitrogen, and stored frozen at Ϫ80°C.
Confocal Microscopy-The intracellular localization of the ET A R and ET B R subtypes was studied using a scanning confocal fluorescence microscope (LSM 510 Carl Zeiss, Oberkochen, Germany). Freshly isolated adult ventricular myocytes were plated on laminin-coated coverslips for 1 h (37°C, 95% O 2 , 5% CO 2 ) and fixed for 20 min in 2% paraformaldehyde solution containing 0.1% (v/v) Triton X-100. Coverslips containing fixed myocytes were rinsed three times in PBS and incubated overnight at 4°C in PBS containing 10% donkey serum. Excess serum was removed and the cells were incubated for 1 h at room temperature with primary antibody diluted 1:100 in PBS containing 1.5% donkey serum. The coverslips were then rinsed gently three times with PBS, drained, and incubated for 1 h at room temperature with a 1:500 (v/v) dilution of the appropriate secondary antibody (TRITCconjugated anti-rabbit antibody, or Alexa 488-conjugated anti-sheep antibody). The coverslips were gently washed three times with PBS, drained, and mounted onto glass slides using a drop of 0.1% diazabicyclo(2.2.2)octane/glycerol medium. Fluorescent myocytes were visualized as serial 0.5-m thick optical sections in the z-axis plane of each cell. To obtain confocal images of dual-labeled cells, Alexa 488 and TRITC emissions were collected simultaneously and composite images were created using the Zeiss software. For each secondary antibody, control experiments were performed in the absence of primary antibody. Where indicated, image stacks were digitally deconvolved (Huygens Professional, Scientific Volume Imaging).
Receptor Binding Assays-Radioligand binding assays were performed essentially as described previously (45) in a binding buffer comprising 50 mM Tris-HCl (pH 7.4 at 24°C), 100 mM NaCl, 10 mM MgCl 2 , 0.1% bovine serum albumin, 1.0 mM phenylmethylsulfonyl fluoride, and 40 pM [ 125 I]ET-1. Binding was at room temperature for 2 h in a reaction volume of 250 l. Non-radioactive ET-1, BQ610, BQ788, and IRL-1620 were prepared as 10 mM stocks in 100% Me 2 SO. Buffer, ligand, and antagonists were combined in the tubes (12 ϫ 75-mm polypropylene). Preliminary experiments were performed to ascertain that [ 125 I]ET-1 binding was linear with respect to protein concentration and that ligand binding was less than 10% of total added ligand. Binding was initiated by the addition of either 25 (sheep) or 100 (rat) g of isolated nuclei. The reaction was stopped by rapid filtration under reduced pressure through Whatman glass microfiber filters (GF/C) with a cell harvester (Brandel). Tubes and filters were rinsed three times with cold 25 mM Tris-HCl, pH 7.5. Before filtration, filters were blocked by incubation for 60 min in 50 mM Tris-HCl (pH 7.5) containing 5% (w/v) skim milk powder (Carnation). Radioactivity on the filters was quantified using a ␥-counter (LKB Wallac). To determine the specific activity of the [ 125 I]ET-1 solution, 10-l aliquots were quantified in quadruplicate by ␥-counting during each binding assay. Data were analyzed using a non-linear curve-fitting program (Prism 3.0cx, GraphPad Software).
SDS-PAGE-[ 125 I]ET-1 binding to ET A R and ET B R is essentially irreversible (45) and for ET B R, but not ET A R, binding is stable during SDS-PAGE at reduced temperature (61)(62)(63)(64). To determine the apparent molecular mass of proteins binding [ 125 I]ET-1 and [ 125 I]IRL-1620, isolated nuclei were incubated with 2.0 nM [ 125 I]ET-1 or 2.2 nM [ 125 I]IRL-1620 plus 1 M unlabeled ET-1, 1 M BQ610, or 1 M BQ788, where indicated, for 1 h on ice in buffer comprising 50 mM Tris-HCl (pH 7.4), 2 mM EGTA, 10 mM MgCl 2 , 1 g/ml aprotinin, 1 g/ml leupeptin, and 0.02% sodium azide. Samples were then solubilized with Laemmli sample buffer and resolved, along with prestained molecular mass markers (Bio-Rad), on 10 -20% (w/v) acrylamide-gradient SDS-PAGE. The electrophoresis cells were cooled to 15°C using an external, circulating refrigeration system. Gels were dried between two cellophane sheets, and exposed to Bio-Max BMR films for 48 h at Ϫ80°C in the presence of an intensifying screen (Kodak Bio-Max TranScreen-HE).
Fluorometric Measurement Calcium-The effects of endothelin receptor agonists upon the permeability of the nuclear membrane to calcium ions were measured by fluorometry using the membrane-permeable, ratiometric calcium indicator, fura-2/AM. Fura-2/AM was prepared fresh daily as a 5 mM stock solution in anhydrous 100% Me 2 SO. Rat heart nuclei were employed immediately after isolation. Aliquots of isolated ventricular nuclei were resuspended in 0.5 ml of a loading buffer comprising 25 mM HEPES (pH 7.0), 100 mM KCl, 2 mM K 2 HPO 4 , and 4 mM MgCl 2 . Fura-2/AM was added to the suspension of nuclei to a final concentration of 7.5 M and incubated for 45 min on ice. The fura-2-loaded nuclei were then diluted 1:2 and washed free of extranuclear fura-2/AM by centrifugation for 5 min at 2,700 ϫ g and 4°C, resuspended once in 1 ml of loading buffer, and then centrifuged and resuspended in 50 l of loading buffer. Fura-2-loaded nuclei were incubated for 15 min on ice for final hydrolysis. Nuclei were then added to 50 l of loading buffer containing 800 nM CaCl 2 and seeded onto glass slides. Where indicated ET-1, IRL-1620, or vehicle (Me 2 SO) were added directly to the nuclei. Fluorescence was measured using an Ion Optix microspectrofluorimeter using excitation wavelengths of 340 and 380 nm and emission at 509 nm. Calibration of the fluorescent signal was determined by sequential addition of 10 M ionomycin plus 1 mM CaCl 2 to obtain the maximal fluorescence ratio (R max ) and 4 mM EGTA to obtain the minimum fluorescence ratio (R min ). Autofluorescence was determined by measuring fluorescence from non-loaded nuclei and subtracting it from the fluorescence produced by fura-2-loaded nuclei.
Determination of Protein Kinase Activity in Isolated Nuclei-The effect of ETR agonists upon protein kinase activities in isolated cardiac nuclei was studied by incubating freshly isolated nuclei with [␥-32 P]ATP in the presence or absence of ET-1 or IRL-1620 and then subjecting the phosphorylated nuclear proteins to SDS-PAGE and autoradiography. The reaction medium (30 l) consisted of 20 mM Tris-HCl buffer (pH 7.5 at 30°C), 20 mM ␤-glycerophosphate, 20 mM NaF, 10 mM DTT, 10 mM MgCl 2 , 100 M GTP, 10 M [␥-32 P]ATP (10 Ci/tube, 30 -40 Ci/mmol), 100 g/ml leupeptin, 5 M microcystin LR, and 1.4 g/ml pepstatin A plus nuclei (30 g). Agonists were present as indicated. Phosphorylation was initiated by the addition of 10 l of nuclei suspended in reaction media minus ATP. Reactions were terminated after incubating 30 min at 30°C by the addition of 10 l of 4ϫ Laemmli sample buffer, heated at 70°C for 90 s, and resolved by electrophoresis on 10 -20% acrylamide-gradient SDS-PAGE. Following electrophoresis, gels were stained in 45% acid. Destained gels were dried between two sheets of cellophane (BioDesign, Inc.) and exposed to Kodak Bio-Max MR film for 24 h at Ϫ80°C in cassettes fitted with Kodak TranScreen-HE intensifying screens. Following autoradiography, gels were exposed to Molecular Imaging screens for 48 h and 32 P incorporation was digitized and quantified by phosphorimaging (Bio-Rad GS 525 molecular analyzer).
Statistical Analysis-Data are represented as the mean Ϯ S.E. The significance of differences between groups was estimated by one-way analysis of variance followed by Tukey's multiple comparison test (Prism 3.0cx, GraphPad Software). Differences were considered significant when p Ͻ 0.05.
Miscellaneous Methods-Protein concentrations were measured according to the method of Bradford (65) using bovine ␥-globulin as a standard. Molecular masses of proteins on Western blots were determined by interpolation from a plot of log M r versus R F for Bio-Rad prestained molecular mass markers to which a second order polynomial curve was fitted by iterative curve-fitting using a non-linear curvefitting program (Prism 3.0cx, GraphPad Software).

Subcellular Localization of Endothelin Receptor Subtypes-
Vascular endothelial cells internalize ET-1 in an ET B R-dependent manner and, once taken up, ET-1 induces a decrease in the amount of prepro-ET-1 mRNA (53). Similarly, nuclear Ca 2ϩ levels increase in response to a cytosolic application of exogenous ET-1 (54). We recently demonstrated the presence of ET A R and ET B R on adult rat ventricular myocytes (45) and the objective of the present study was to determine whether functional ETRs exist intracellularly and, if so, determine the receptor subtype and subcellular localization of these receptors within the cardiac ventricular myocyte. Initially, canine cardiac MV were isolated and subfractionated into membranes enriched in plasma membranes (Fig. 1A, fraction 1), and membranes of intracellular origin (Fig. 1A, fractions 2-5) (56,58). Immunoblotting using antisera specific for the ET A R identified bands of 49 and 65 kDa present in the non-fractionated membrane vesicles (Fig. 1A, MV). The 65-kDa immunoreactive band was present in fractions 1 and 2, showing markedly enrichment in fraction 1. Similarly, bands of 60 and 65 kDa were detected in rat brain membranes (Fig 1A, RB). In contrast, antisera against the ET B R subtype revealed a single band of 57 kDa in MV and fractions 1-5 (Fig. 1B). Fractions 3-5, comprising membranes of intracellular origin, were visibly enriched in ET B R immunoreactivity. In rat brain membranes, ET B R-specific antisera revealed a 46-kDa band. Thus in canine heart, ET A R immunoreactivity was detected primarily in plasma membranes whereas ET B R immunoreactivity was observed in fractions corresponding to intracellular membranes.
Confocal Imaging of Endothelin Receptor Subtypes in Ventricular Myocytes-The subcellular localization of ET A R and ET B R subtypes was studied further in isolated adult rat ventricular cardiomyocytes using confocal immunofluorescence microscopy along with ET A R-and ET B R-selective antibodies. As the selectivity of an antibody may differ in immunoblotting versus immunocytochemistry and endothelin receptors may exist in a folded state or complex that hinders access to a given epitope, two different antibodies, raised against peptides corresponding to distinct regions in the primary sequence of the receptor, were employed for each receptor subtype. Antisera for ET A R were raised against synthetic peptides corresponding to amino acids 31-45 (ET A R [31][32][33][34][35][36][37][38][39][40][41][42][43][44][45] ; ABCAM) and 413-426 (ET A R 413Ϫ426 ; Alomone Labs) in the primary sequence of the rat ET A R. Antisera for ET B R were raised against synthetic peptides corresponding to amino acids 299 -314 (ET B R 299 -314 ; Alomone Labs) and 405-417 (ET B R 405-417 ; Biogenesis Ltd.) in the primary sequence of the rat ET B R.
The distribution of ET A R was studied using antisera raised against peptides corresponding to amino acid residues 31-45 (Fig. 2, A, B, F, H, J, and K) or 413-426 (Fig. 2C) in the primary sequence for the rat ET A R. Both antibodies decorated the cell surface and produced a striated pattern within the cell (Fig. 2, A-C) indicating that ET A R exists in plasma membrane structures including both the surface membranes and the T-tubular network. In addition, nuclear or perinuclear staining was observed that appeared to extend outwards from the apex of the nuclei (e.g. Fig. 2C). Adult cardiac ventricular myocytes are generally binucleate. No staining was observed in the absence of the primary antibody (Fig. 2, D and E). To further examine the possible nuclear localization of the ET A R subtype, cells were decorated with antibodies to both ET A R and nucleoporin p62 (Nup62), a component of the central plug of the nuclear pore complex. Nup62 immunoreactivity was concentrated in a sharp ring surrounding the nucleus (Fig. 2, G, H, J, and K) and showed little non-nuclear staining. ET A R appeared to localize to a structure surrounding, but lying outside of, that decorated by the Nup62 antibody.
The distribution of ET B R was studied using antisera raised against peptides corresponding to amino acid residues 298 -314 (Fig. 3, A, B, F, L, and H) or 405-417 ( Fig. 3, C, G, and H) in the primary sequence for the rat ET B R. Both antibodies produced a striated pattern within the cell along with strong staining of the nucleus (Fig. 3, A-C, F, G, and L). Three-dimensional digital deconvolution of the image stack revealed that the intense nuclear staining was because of staining of the nuclear membrane with some punctate staining within the nucleoplasm ( Fig. 3U and inset). Staining of the surface membranes For control experiments, myocytes from the same heart were prepared according to the labeling procedure except that primary antibodies were omitted. Confocal microscope settings for control images were identical to those employed in the presence of the respective primary antibody. The absence of signal in these control images indicates that the blocking conditions were sufficient to prevent nonspecific interactions between the myocytes and the secondary antibodies. The arrows in panels A and C indicate the position of the nuclei: adult ventricular myocytes are generally binucleate.
was much less obvious than observed for ET A R, with ET B R showing a more striated distribution suggestive of localization to the transverse tubules. When cells were simultaneously decorated with both ET B R antibodies the pattern of distribution indicated that they co-localized (Fig. 3, F-H). To further examine the nuclear localization of the ET B R subtype, cells were decorated with antibodies to both ET B R and Nup62 (Fig.  3, L-Q). Nup62 immunoreactivity was concentrated in a sharp ring surrounding the nucleus (Fig. 3, M and P) and showed little association with non-nuclear structures. The distribution of ET B R immunoreactivity on the nucleus was strikingly similar to that observed with Nup62, suggesting that ET B R are also located on the nuclear envelope. Control experiments using only the secondary antibodies detected minimal nonspecific fluorescence under the imaging conditions employed (Fig. 3,

R-T).
Biochemical Characterization of Nuclear Endothelin Receptors-Imaging of cells using confocal microscopy indicated that ET A R and ET B R are located on or around the nuclei in ventricular myocytes. To further study and characterize the ETRs associated with the nucleus, nuclei were isolated from rat hearts. Four fractions were recovered from this purification: the total homogenate (Fraction 1), 500 ϫ g supernatant (Fraction 2), crude nuclear fraction (Fraction 3), and enriched nuclear fraction (Fraction 4). This preparation was characterized by determining the distribution of marker proteins of known subcellular origins including caveolin-3 (rafts-plasma membrane), annexin II (plasma membrane, subcellular vesicles), BiP/GRP78 (endoplasmic reticulum), GM130 (Golgi apparatus), and Nup62 (nuclear envelope) (Fig. 4A). Both ET A R and ET B R immunoreactivity was detected in Fraction 1 as were caveolin-3 (not shown), annexin II, BiP/GRP78, and GM130. ET A R, annexin II, BiP, and GM130 were detected in the 500 ϫ g supernatant (Fraction 2). This fraction was further purified. Nup62, annexin II, and GM130 immunoreactivity were detected in the crude nuclear fraction (Fraction 3). Relative to Fraction 3, the final nuclear fraction (Fraction 4) was enriched in Nup62, ET A R, and ET B R. In contrast, the levels of GM130 and annexin II were reduced in Fraction 4 relative to Fraction 3. The immunoreactive bands detected in the rat heart nuclei For control experiments, myocytes from the same heart were prepared according to the labeling procedure except that primary antibodies were omitted. Confocal microscope settings for control images were identical to those employed in the presence of the respective primary antibody. The absence of signal in these control images indicates that the blocking conditions were sufficient to prevent nonspecific interactions between the myocytes and the secondary antibodies. The arrows in panels A and C indicate the position of the nuclei.
using ET A R-or ET B R-specific antibodies were 48 and 57 kDa (Fig. 4C), respectively. A faint 48-kDa ET A R-immunoreactive band was also detected in canine cardiac vesicles (Fig. 1A, MV) and rat brain membranes (Fig. 1A, RB). Hence, both ET A R and ET B R-immunoreactive proteins copurified with a nuclear pore protein, Nup62. To further characterize the nuclear localization of ETRs, nuclei were isolated from sheep heart. An immunoreactive band of 48 kDa was detected in sheep nuclei using the ET A R-specific antibody from Alomone (Fig. 4B) and AB-CAM (not shown), whereas both ET B R-specific antibodies revealed a 57-kDa immunoreactive band (Fig. 4B) similar to that in rat nuclei and canine heart. Thus, in nuclei isolated from both rat and sheep hearts, immunoreactive bands with molecular mass similar to that predicted for the ET A R and ET B R were detected using antibodies directed against two different epitopes in the sequence of the receptors.
Nuclear [ 125 I]ET-1 Binding-To determine whether nuclear ETRs immunoreactivity was representative of the presence of functional, ligand-binding receptors, radioligand binding assays were performed using isolated nuclei. Nuclei isolated from both rat (Fig. 5A) and sheep heart (Fig. 5B) bound [ 125 I]ET-1 and this binding was displaced by unlabeled ET-1. The total specific ET-1 binding was 23.9 Ϯ 5.7 and 5.1 Ϯ 1.5 fmol/mg in nuclei isolated from sheep and rat heart, respectively. This difference is consistent with the fact that the nuclei isolated from sheep heart are of higher purity than those from rat heart. Nuclear ET-1-binding sites were pharmacologically classified using receptor subtype-specific antagonists. BQ610, an ET A Rselective antagonist displaced 17.  (63). Hence, to further characterize nuclear ETRs and compare the apparent molecular mass of the nuclear ETRs with those in other subcellular fractions, binding and SDS-PAGE was performed using [ 125 I]IRL-1620 (2.5 nM; Fig.  6). This compound is an ET B1 R-specific ligand characterized by a 100,000-fold selectivity for ET B R over ET A R (68). A band of 60 kDa was observed in all samples and was most abundant in sheep heart nuclei (Fig. 6, lane 7). This molecular mass is similar to that for ET B R detected by immunoblotting. Although ET B R can undergo an NH 2 -terminal proteolytic cleavage during cellular lysis and membrane isolation (69), no bands of lower molecular mass were detected in the present study. To confirm the specificity of [ 125 I]IRL-1620 binding in the enriched nuclei fraction, [ 125 I]IRL-1620 binding was examined in the presence of excess (1 M) unlabeled ET-1, BQ788 (ET B R antagonist), or BQ610 (ET A R antagonist). [ 125 I]IRL-1620 binding was reduced when the nuclei were preincubated either with unlabeled ET-1 or BQ788 (Fig. 6, lanes 2 and 4). BQ610 did not displace [ 125 I]IRL-1620. Thus, ET B1 R are present on cardiac nuclear membranes and display the same apparent molecular mass on SDS-PAGE as ET B1 R in other membranes.
Functional Changes Associated with Nuclear Actions of ET-1-The studies described above demonstrate localization of endothelin receptors immunoreactivity and [ 125 I]ET-1 binding in ventricular membranes, isolated myocytes, and isolated cardiac nuclei. At the plasma membrane, ETRs are coupled to activation of PKC and numerous downstream signaling events including activation of the ERK mitogen-activated protein kinase pathway. Nuclei have been shown to contain phosphatidylinositols (70, 71), 1,2-diacylglycerol (72), plus phosphatidyl- FIG. 4. Endothelin receptor immunoreactivity in isolated cardiac nuclei. A, nuclei were isolated from rat heart as described under "Experimental Procedures." Lanes contain 25 g each of homogenate (F.1), 500 ϫ g supernatants (F.2), crude nuclei (F.3), or enriched nuclear fraction (F.4). Proteins were separated on 10 -20% acrylamide-gradient SDS-PAGE gels and transferred onto filters. Following blocking, membranes were incubated with antisera specific for BiP/GRP78, nucleoporin p62 (Nup62), annexin II, GM130, ET A R 413-426 , or ET B R 298 -314 . B and C, nuclei were isolated from sheep heart and 50 g were separated on 10 -20% acrylamide-gradient SDS-PAGE gels and transferred onto filters. Following blocking, membranes were incubated with antisera specific for ET A R (B) or ET B R (C). Immunoreactive bands were visualized by ECL. In panels B and C, the molecular mass markers (in kDa) are indicated on the left.
inositol-specific phospholipase C␤ 1 (73) required to produce 1,2-diacylglycerol. In addition, EGF regulates protein phosphorylation in isolated nuclei (74) suggesting that the nuclear membrane contains sufficient elements to support receptoractivated protein kinase activities. Hence, the effect of ETR agonists upon nuclear phosphorylation was examined in isolated nuclei. Incubation of isolated nuclei in the presence of [␥-32 P]ATP resulted in the incorporation of 32 P into several proteins (Fig. 7A) including two intense bands of 42 kDa (Fig.  7B, panel 3) and 46 kDa (Fig. 7B, panel 4). ET-1, IRL-1620, and PMA induced different patterns of nuclear protein phosphorylation and both increases and decreases in phosphate incorporation were observed (Fig. 7B, panels 1-5). ET-1 decreased 32 P incorporation into the proteins shown in panels 1 and 4, whereas phosphorylation of a 65-kDa protein, shown in panel 2, was increased. IRL-1620 increased 32 P incorporation to a 30-kDa nuclear protein (panel 5). In contrast, the incorporation of 32 P into the phosphoprotein shown in panel 3 was unaffected by PMA or ETR agonists. It is also worthy of note that proteins where 32 P incorporation increased in response to ETR agonists also demonstrated an increased level of 32 P incorporation following treatment with PMA, suggesting the involvement of PKC downstream of nuclear endothelin receptors. As the ERK mitogen-activated protein kinase cascade is downstream of PKC, the effect of ET-1 and IRL-1620 upon the phosphorylation of ERK1/2, p90 rsk , and Elk1 was examined. Phosphate incorporation was determined using antisera specific for the phosphorylated form of these proteins. ETR agonists produced no detectable change in phosphorylation of these proteins while antisera against total ERK1/2 clearly demonstrated the presence of both proteins in isolated nuclei (data not shown).
The nuclear envelope is a double membrane separating the nucleus from the cytoplasm. The lumenal space between these membranes, referred to as the nuclear cisterna, can accumulate Ca 2ϩ and it has been proposed that Ca 2ϩ can be released directly from the nuclear cisterna into the nucleoplasm via Ca 2ϩ channels associated with inositol trisphosphate and/or ryanodine receptors (75,76). In the present study, the effect of ETR agonists ET-1 and IRL-1620 upon nuclear cisternal Ca 2ϩ concentration [Ca 2ϩ ] n was examined in freshly isolated nuclei using the ratiometric fluorescent Ca 2ϩ indicator, fura-2/AM. Addition of ETR agonists to the fura-2-loaded nuclei, in the presence of 400 nM external Ca 2ϩ , resulted in an increase in the A 340 /A 380 ratio. ET-1 caused a transient increase in [Ca 2ϩ ] n from a basal value of Ϸ215 nM Ca 2ϩ to Ϸ280 nM (Fig. 8). IRL-1620 also caused a similar increase in [Ca 2ϩ ] n . However, the increase in [Ca 2ϩ ] n induced by IRL-1620 was transitory and returned to basal level at 20 s. As there was no exogenous ATP present during these experiments, these results suggest that endothelin receptors are coupled to, and capable of opening, a calcium channel within the nuclear membrane.

DISCUSSION
To date, endothelin-related effects have been thought to be mediated via G protein-coupled receptors located on the plasma membrane. However, recent data imply that endothelins may also act intracellularly. ET-1 internalized via binding to the ET B R has been shown to decrease the level of prepro-ET-1 mRNA (53). Furthermore, cytosolic application of ET-1 is sufficient to evoke changes in nuclear calcium content (54). However, the existence and subcellular location of the ETRs has not been demonstrated. Thus, this study presents novel evidence for the presence of functional endothelin receptors on the nuclear envelope in cardiac myocytes.
Nuclear or perinuclear localization of ETRs can be attributed to translocation from the cell surface or de novo synthesis. Because the cells are exposed to circulating ET-1 in vivo, ligand-mediated receptor internalization and translocation to the perinuclear region is a possibility. ET A R localize primarily to the plasma membrane (77) and internalize in a ligand-dependent manner (78). Once internalized, ET A R follow a recycling pathway through the pericentriolar recycling compartment and then back to the plasma membrane (79). In contrast, in Chinese hamster ovary cells expressing a fusion protein comprising green fluorescent protein (GFP) and ET B R (GFP-ET B R), GFP-ET B R was observed to internalize constitutively, without a requirement for ligand binding, and the internalized receptor was targeted to the lysosomes for degradation (64,79). The fate of each receptor subtype is under the control of specific elements regulating protein trafficking. Chimeric ET A R and ET B R constructs reveal that the cytoplasmic COOH-terminal domain of ET A R is sufficient to specify receptor recycling whereas the comparable domain in ET B R specifies delivery to lysosomes (77,80). When stably expressed in HEK293 cells, ET B R-GFP localizes primarily to the plasma membrane and ET-1 induces a metalloprotease-mediated cleavage between amino acid residues 64 and 65 in the extracellular NH 2 -terminal domain that precedes internalization of the receptor and its sorting into endosomes (81). In addition, an NH 2 -terminal deletion mutant [⌬2-64]ET B R-GFP is functionally normal but  showed a 15-fold decrease in expression at the cell surface when expressed in HEK293 cells (81). In ventricular myocytes, ET B R was primarily intracellular (Figs. 1 and 3) and no differences in the molecular mass were observed for ET B R associated with plasma membrane versus nuclei (Fig. 6). Hence, both amino-and carboxyl-terminal domains contain elements that regulate the trafficking of endothelin receptors.
Endothelin-1 binding to either ET A R or ET B R results in receptor-ligand complexes that dissociate very slowly with reported half-lives of up to 20 h (82). Following internalization, both ET A R and ET B R remain bound to ligand for more than 2 h (64, 78). One could postulate that the ligand-bound and unoccupied forms of the ETRs follow different internalization pathways; however, GFP-ET B R bound with cyanin 3-conjugate ET-1 (Cy3-ET-1) cotraffic from the cell surface to lysosomes, remaining as a complex within the cell for up to 4 h (64). As ET A R is internalized in a ligand-dependent manner and subsequently recycled to the cell surface, either there is a means within the cell to induce dissociation or the time required for recycling exceeds the half-life of the receptor-ligand complex. Interestingly, salicylic acid has been reported to induce the dissociation of ET-1 from ET A R (83) whereas antagonists reduce the rate of dissociation (84) or increase ET-1 binding (45); hence, there may exist allosteric regulatory sites that regulate the dissociation of the receptor-ligand complex. Figs. 6 and 7 demonstrate that nuclear ETRs were available for ligand binding. Taken together, these results suggest it is likely that nuclear ETRs result from de novo synthesis, rather than liganddependent internalization, and may represent a distinct and novel subcellular target for endothelins.
Evidence is emerging that demonstrates growth factor receptors and G protein-coupled receptors are present on nuclear or perinuclear membranes. Studies have localized type 1 angiotensin II (AT 1 ) (50, 85), epidermal growth factor (EGF-R) (47), c-erbB-4 (86), insulin (46,87), interferon ␤ (49), muscarinic cholinergic (88), nerve growth factor (48), prostaglandin E 2 (51,52), and opioid (89) receptors, in addition to many polypeptide ligands (e.g. epidermal growth factor, insulin, platelet-derived growth factor, nerve growth factor, parathyroid hormone-related protein, prolactin, interleukin 1, somatostatin, fibroblast growth factors 1 and 2, and transforming growth factor ␣) to the nucleus. Receptor ligands may be derived from the extra-cellular milieu or synthesized within the cell. Ligand-mediated receptor internalization and translocation to the nucleus has been demonstrated for AT 1 (90) and EGF-R (91). Conversely, the pro-form of transforming growth factor ␣ (pro-TGF-␣) may exert a mitogenic effect by interacting with nuclear EGF-R (92). The nuclear EGF-R may act as a "stop" signal, following an initial mitogenic response, to prevent unregulated cell proliferation (see Ref. 93). In contrast, prostaglandin E 2 is taken up via prostaglandin transporters after which it acts upon nuclear receptors (94). For intracellular ETRs to be of functional relevance, there must also be a source of intracellular ligand and both prepro-ET-1 mRNA 2 and ET-1 immunoreactivity (17,95) have been demonstrated in adult ventricular myocytes.
At present there is little known concerning the function of nuclear growth factor receptors or G protein-coupled receptors. Potential roles for nuclear signaling pathways include regulation of nuclear transport, gene expression, and nuclear envelope formation. Key components of various signaling pathways are present at the nuclear envelope or within the nucleus itself, supporting the presence of nuclear signaling cascade(s). Evidence suggests the presence of nuclear heterotrimeric G-proteins (see Ref. 96). Nuclear AT 1 receptors display guanine nucleotide-dependent ligand binding, indicating the presence of and coupling to heterotrimeric G-proteins (50). Biophysical and microscopy studies support the presence of effectors including adenylyl cyclase (97), phosphodiesterase (98), diacylglycerol kinase-(99), phospholipase A 2 (100), phospholipase C (73), phospholipase D (101), and phosphatidylinositol 3-kinase C2␣ (102). The nuclear envelope comprises two membranes, the inner and outer nuclear membranes: the lumenal space between these membranes is referred to as the nuclear cisterna or perinuclear space (103). Nuclear membranes contain sarco/ endoplasmic Ca 2ϩ -ATPase pumps (104,105) as well as ryanodine-sensitive Ca 2ϩ channels (60,105) and inositol 1,4,5trisphosphate-sensitive (60) Ca 2ϩ channels. The signaling processes associated with the nuclear envelope include the products of nuclear lipid metabolism, such as 1,2-diacylglycerol and inositol 1,4,5-trisphosphate (see Refs. 106 and 107). 1,2-2 N. Farhat, H. Farhat, E. Thorin, and B. G. Allen, unpublished data.

FIG. 8. Effects of endothelin receptor agonists on Ca 2؉ transients in isolated cardiac nuclei.
Freshly isolated nuclei were resuspended in a buffer comprising 25 mM HEPES (pH 7.0), 100 mM KCl, 2 mM K 2 HPO 4 , and 4 mM MgCl 2 (loading buffer) plus 7.5 M fura-2/AM and incubated for 45 min on ice. Dyeloaded nuclei were washed free of extra nuclear fura-2/AM, incubated for 15 min on ice for final hydrolysis, added to an equal volume of loading buffer containing 800 nM CaCl 2 , and seeded onto glass slides. Where indicated (arrow), ET-1, IRL-1620, or vehicle (Me 2 SO) was added. The nuclear cisternal Ca 2ϩ concentration [Ca 2ϩ ] n was measured as described under "Experimental Procedures" using an Ion Optix microspectrofluorimeter ( ex ϭ 340 and 380 nm; em ϭ 509 nm). Typical tracings of intracellular Ca 2ϩ are shown. Results are representative of six determinations performed using three independent nuclei preparations. Diacylglycerol recruits and/or activates protein kinase C to phosphorylate intranuclear proteins: generation of inositol 1,4,5-trisphosphate releases Ca 2ϩ stored within the nuclear cisterna, increasing nucleoplasmic Ca 2ϩ . Nuclear fibroblast growth factor-2 activates casein kinase II and thereby induces transcription of ribosomal genes and phosphorylation of nucleolin (108). Similarly, EGF induces internalization and nuclear localization of the EGF-R and enhanced tyrosine phosphorylation of nuclear proteins (109). Stimulation of nuclear ETRs (Fig. 8) or prostaglandin E 2 receptors (51) causes an influx of Ca 2ϩ into the nuclear cisterna or nucleoplasm. Nucleoplasmic Ca 2ϩ regulates key nuclear functions including gene transcription, apoptosis, gene repair, topoisomerase activation, and polymerase unfolding. In addition to the regulatory effects mediated by the release of Ca 2ϩ into nucleoplasm, the filling status of the nuclear cisterna alters the conformational state of the nuclear pore complex, inhibiting diffusion across the nuclear envelope and hence controlling the transport of molecules between the cytosol and nucleoplasm (103,110).
To date, endothelin-related effects have been thought to be mediated via specific ETR subtypes located on the plasma membrane. The results of the present study provide novel evidence for the presence of functional ETRs on the nuclear envelope in the ventricular myocardium. This is based upon the following: 1) nuclei isolated from both sheep and rat ventricular myocardium contained both ET A R and ET B R immunoreactive bands; 2) ET A R and ET B R immunoreactivity copurified with a marker of the nuclear pore complex (nucleoporin 62) and not with markers of the endoplasmic reticulum and Golgi membranes; 3) radioligand binding using receptor subtype-specific antagonists confirmed the presence of both ET A R and ET B R in nuclei isolated from both sheep and rat ventricular myocardium; 4) confocal fluorescence microscopy localized ETRs to the nuclear membrane in adult ventricular myocytes; 5) ETR agonists altered the pattern of protein phosphorylation and elicited an increase in nuclear cisternal Ca 2ϩ concentration. Hence, cardiac nuclei possess ETRs that are functional with respect to ligand binding and coupled to signaling mechanisms within the nuclear membrane. For intracellular ETRs to be of functional relevance, there must also be an intracellular source of ligand. One potential source is via uptake of endothelin from the extracellular milieu. Alternatively, ET-1 may be produced by the myocyte and act intracellularly as an "intracrine" mediator.