Abstract
The Ang IV receptor, AT4, has been shown to play important roles in various mammalian tissues. In this study, structural properties of the AT4 receptor from bovine adrenals are described using a novel photoactive analog of Ang IV, [125I]Benzoylphenylalanine-Ang IV (BP-Ang IV), recently developed in our laboratory. [125I]BP-Ang IV is identical to Ang IV with regards to binding specificity and affinity and is easily cross-linked to the AT4 receptor under UV light, thus greatly facilitating the structural analysis of the AT4 receptor by SDS-PAGE. Comparisons between the native, reduced and nonreduced forms of the AT4 receptors by SDS-PAGE revealed that this receptor consists of multiple subunits. The subunit containing the Ang IV binding site (designated as thealpha subunit) has a molecular weight of ∼165 kDa and contained ∼20% N-linked carbohydrates. A subunit similar to the adrenal alpha subunit of the AT4 receptor was identified in all of the bovine tissues examined. Hippocampus and aorta contained additional [125I]BP-Ang IV bound protein bands with molecular weights of 150 and 125 kDa, respectively. Further, the alpha subunit was purified to homogeneity using a method that integrates electrofractionation with conventional protein purification techniques.
Ang IV is a six-amino-acid peptide belonging to the angiotensin family of peptides. Angiotensins are best recognized as major regulators of cardiovascular function and body-water homeostasis (Peach and Dostal, 1990; Wright et al., 1995). Classically, the octapeptide Ang II and the heptapeptide Ang III have been considered to be biologically active while angiotensin fragments smaller than Ang III were believed to be inactive (Blair-West et al., 1971; Timmermans et al., 1991; Tonnaer et al., 1982a, 1982b). Recent studies, however, suggest that smaller angiotensin fragments may mediate unique physiological functions distinct from those ascribed to Ang II and Ang III (Ferrario et al., 1991; Wright et al., 1995). While the classic angiotensin-dependent functions are primarily the result of activation of AT1receptors and to a lesser extent, AT2 receptors, the actions of Ang IV and related compounds result from interaction with the AT4 receptor (Swanson et al., 1992). The AT1 receptor has been defined by its ability to bind Ang II, Ang III and losartan, whereas the AT2 receptor, which also binds Ang II and Ang III, shows higher affinity for CGP4211A, PD123177, and related substances. The AT4 receptor, however, binds Ang IV and shows poor affinity for Ang II, losartan, CGP4211A and PD123177 (Swanson et al., 1992). AT4 ligand structure-binding studies have revealed that the N-terminal sequence of the Ang IV peptide plays a pivotal role in the recognition of the ligand by the AT4 receptor as well as being the primary determinant of binding affinity (Sardinia et al., 1993, 1994).
The AT4 receptor has been identified in all mammalian species so far examined, including but not limited to rat, guinea pig, monkey, human and bovine (Wright et al., 1995;Moeller et al., 1996). The AT4receptor also exhibits a wide tissue distribution with high concentrations in heart, kidney vasculature, adrenals and many regions of the brain (Hanesworth et al., 1993; Harding et al., 1994; Hall et al., 1993, 1995; Bernier et al., 1994; Miller-Wing et al., 1993; Moeller et al., 1995). Bovine adrenal gland, which has the highest known concentration of AT4 receptors, was the tissue chosen for this initial structural and chemical characterization of the AT4 receptor (Swanson et al., 1992;Bernier et al., 1995; Jarvis et al., 1992; Krebset al., 1996).
A growing list of physiological functions has been attributed to the AT4 receptor. Intracerebroventricular infusions of Ang IV to rats enhances both memory retrieval and retention in passive avoidance tests (Brazko et al., 1988; Wrightet al., 1993). In addition, spatial learning tasks show improvement with administration of an AT4 agonist in congitively compromised rats (Pederson et al., 1996;Stubley-Weatherly et al., 1996). Ang IV, working through AT4 receptors, has also been shown to increase blood flow in various vascular beds including renal cortical (Swansonet al., 1992; Coleman, et al., 1998) and cerebral (Kramar et al., 1997). In addition to mediation of cortical blood flow in the kidney, Ang IV affects Na+excretion and associated O2 consumption in rat proximal tubules (Handa et al., 1998). Ang IV binding sites have also been identified in guinea pig and rabbit heart, where Ang IV has been shown to antagonize Ang II-induced hypertrophic changes in cultured chick myocytes (Baker and Aceto, 1990), regulate immediate early gene expression (Yang et al., 1997) and modulate diastolic function. Additionally, Ang IV has been shown to effect endothelial cell function in several ways. Treatment of bovine CVEC with Ang IV enhances fibroblast growth factor-β-induced [3H]thymidine incorporation (Hall et al., 1995). Ang IV has also been found to be a potent stimulator of PAI1 expression in bovine aortic endothelial cells (Kerins et al., 1995).
Despite the growing body of information concerning AT4 pharmacology and physiology, little is known about the structure of the AT4 receptor and its method of intracellular signaling. To further advance the knowledge of the AT4 receptor system, it is therefore important to elucidate its structure and signaling mechanisms. Thus, the goal of this investigation was to acquire initial information concerning the structure of the AT4 receptor. The method of approach entails using a specific125I-labeled photoactive analog to covalently label the receptor followed by subsequent SDS-PAGE analysis. The analog employed in the study incorporated Bp into the Ang IV-type structure in the number six position. The Bp moiety has been included successfully in several photoprobes that have been employed to covalently label and characterize membrane-bound receptors (Thiele and Fahrenholz, 1993;Macdonald et al., 1996; McNicoll et al., 1996;Boss’e et al., 1993).
In the present study, we first solubilized the bovine adrenal AT4 receptor using the zwitterionic detergent CHAPS and compared its binding characteristics with that of the membrane-bound receptor. A Bp-substituted ligand, BP-Ang IV, was synthesized, characterized to demonstrate its suitability as a specific probe for the AT4 receptor and then used to covalently label the AT4 receptor. The covalently labeled receptor was analyzed by SDS-PAGE to estimate the molecular weight of the receptor as well as to identify the existence of possible subunits. The effects of sulfhydral reducing agents such as β-ME or DTT on the receptor structure were investigated to give an indication of possible critical disulfide bonds associated with the receptor. Additional studies were carried out to quantitate the level of glycosylation of the receptor and to establish its isoelectric point. Finally, the photoaffinity label was employed to guide the purification of the receptor.
Materials and Methods
Tissues and chemicals.
Fresh bovine adrenals were obtained locally and frozen at –80°C in 15-g aliquots. Ang IV [VYIHPF] and divalinal-Ang IV [Vψ(CH2NH)YVψ(CH2NH)HPF] were synthesized in the author’s laboratory (J.W.H.). DuP753 was a gift from Dr. Ron Smith of DuPont/Merck (Wilmington, DE). PD123319 was a gift from Dr. David Taylor of Warner Lambert-Park-Davis. CGP42112A was as gift from Mark DeGasparo of Ciba-Geigy. Ang II and Sar1-Ile8-Ang II were purchased from Sigma (St. Louis, MO). Bioampholyte(1–10) was purchased from BioRad (Hercules, CA). All other chemicals and buffers used were reagent grade from Sigma unless specified.
Synthesis of BP-Ang IV.
BP-Ang IV was synthesized on methylated 1% divinylbenzene cross-linked polystyrene resins substituted with the protected carboxy-terminal amino acid using a semiautomated peptide synthesizer (Vega 250c coupler). Thetert-butoxycarbonyl protection group was used for the α-amine of all amino acids. Other protecting groups employed were 2,6-dichlorobenzyl for Tyr and tosyl for His. Amino acids were activated with EDC, and coupling was monitored for completeness with the Kaiser ninhydrin test (Kaiser et al., 1970). Coupling reactions that tested <99.4% complete were repeated. The crude peptide was cleaved from the resin and deprotected utilizing anhydrous HF containing 10% anisole at 0°C for 45 min. The HF and anisole were removed under vacuum and the resin was washed with anhydrous diethyl ether. The peptide was extracted with 10% acetic acid and lyophilized.
The crude peptide was purified by reverse-phase HPLC in two steps on a preparative Dynamax C-18, 21.4 mm × 25 cm column (Rainin Instrument) at 10 ml/min. In the first separation, buffer A consisted of 80 mM triethylamine phosphate, pH 3.0. Solvent B was acetonitrile. The absorbance of the eluted peptide was monitored at 210, 254 and 280 nm. The collected peak was concentrated under a stream of nitrogen and further purified using a gradient elution: 0% to 30% B over 26 min, 30% to 30% B over 15 min, 30% to 50% B over 15 min at 10 ml/min (buffer A: 0.1% TFA in H2O, solvent B: 0.1% TFA acetonitrile). Purity of the collected peaks was assessed by monitoring the ratio of 280 and 210 nm absorbance across the collected peaks. A constant ratio was indicative of a pure compound. All analogs employed in this study eluted with a constant ratio. In addition, UV spectra from 200 to 300 nm were recorded for each peak and found to be consistent with the peptide sequence. The purified peptide was lyophilized and stored desiccated at –20°C. Final peptide purity was determined by compositional amino acid and mass spectrometic analysis. Typical BP-Ang IV purity was >98%.
Iodination of Ang IV and BP-Ang IV.
Peptides were iodinated using the chloramine T method. Peptide (50 μg) was incubated in a total volume of 150 μl of 0.2 M sodium phosphate buffer (pH 7.2), containing 80 μg of chloramine T and 2 mCi Na125I (DuPont) for 25 sec at room temperature. The reaction was stopped by the addition of 100 μg of Na2S2O5in 50 μl of sodium phosphate buffer. The mono-[125I] peptides were separated from unlabeled and diiodinated peptides by HPLC (Beckman) using a reverse phase C18 column (5 μm × 250 mm, Microsorb-MV, Rainin Instrument). Solvent A was 80 mM triethylamine phosphate (pH 3.0), whereas acetonitrile was solvent B. [125I] Ang IV and [125I]BP-Ang IV were purified using a linear gradient of 9% B/25% B and 15% B/30% B over 90 min, respectively.
Tissue preparation and solubilization of the AT4 receptor from bovine adrenal membranes.
For each membrane preparation containing AT4 receptor, 15 g of bovine adrenals was homogenized in 10 ml of hypotonic buffer at 4°C as described byHanesworth et al. (1993). After homogenization, the concentration of total membrane protein was adjusted to 10 mg/ml. Membrane proteins were solubilized in buffer containing 1% CHAPS, 50 mM Tris (pH 6.8) and 5 mM EDTA at 4°C for 2 hr. Nonsolubilized material was removed by passing the solubilized mixture through a 0.2-μm filter (Sigma) following centrifrigation at 100,00 g for 60 min. The amount of AT4 receptor in each preparation was determined by a radioligand ([125I] Ang IV) binding assay (Hanesworthet al., 1993). For each preparation, ∼200 μl of sample was withdrawn, cross-linked with [125I]BP-Ang IV as described below and added back to the preparation.
AT4 receptor cross-linking with [125I]BP-Ang IV.
To cross-link [125I]BP-Ang IV to AT4specifically, the AT4-bound [125I]BP-Ang IV was first separated from free [125I]BP-Ang IV by using a P-6 column. The P-6 column was prepared by transferring 1 ml of P-6 resin (BioRad) into a 1 ml disposable syringe, which was then placed in a 12 × 75 mm plastic centrifuge tube and centrifuged at 232 × g for 10 min at room temperature. The binding reaction between the solubilized AT4 receptor and [125I]BP-Ang IV was as described above. After elution from the P-6 column the effluent was transferred to a weighing dish and exposed to UV light (312 nm variable intensity Transilluminator FBTIV-88 (Fisher) on ice for 30 min.
Receptor binding assay procedures.
The AT4 receptor binding assay was performed at 37°C for 2 hr in a total volume of 250 μl of isotonic assay buffer as described by Hanesworth et al. (1993). At the end of each incubation, the bound and free ligands in membrane preparations were separated by vacuum filtration (Brandel Cell Harvester, Gathersberg, MD.) using #32 glass fiber filters (Schleicher and Schuell, Keene, NH) and washed with 4 ml of PBS (pH 7.2). The PBS washes were repeated a total of 4 times. The radioactivity retained by the filters was determined by gamma counter (Isomedic, 10/880, ICN, Costa Mesa, CA). With the solubilized receptor, the bound and free ligands were separated using a P-6 column. Nonspecific binding was ascertained in the presence of 1 μM unlabeled Ang IV. The apparent binding equilibrium constant (KD ) was derived from the equation KD =k–1/k1 wherek–1 is the dissociation rate constant andk1 is the association rate constant.k1 = (kobs −k–1)/[L] wherekobs is the pseudo-first-association rate constant and [L] is the radioligand concentration. To obtain thekobs value, the binding assays for association experiments were performed at 37°C in the presence of 0.6 nM [125I]Ang IV for 240 min. Duplicate samples were taken at 11 time points. The kobsvalue was determined using the curve fitting program, INPLOT4 (GraphPAD, San Diego, CA). For dissociation experiments, the binding assays were performed by preincubation of the tissue preparation with 0.6 nM radioligand for 120 min, followed by the addition of 1 μM unlabeled ligand (final concentration). Duplicate samples were taken over a 330-min time period at 13 time points. The dissociation rate constant was determined using the same INPLOT4 program. Saturation equilibrium binding and competition studies were carried out for a 120-min incubation period at 37°C in the presence of various concentrations of radioligand and/or competing ligands. The half-log dilutions of competitor with concentrations ranging from 10−11 M and 10−4 M were used to generate the competition curves. The INPLOT4 program was used to analyze the saturation and competition data to determine the maximum number of binding sites (Bmax), the affinity of the labeled ligands (KD ) and the affinity of the competing ligands (Ki ).
SDS-PAGE analysis of the binding specificity between the AT4 receptor and [125I]BP-Ang IV.
The binding specificity between the AT4 receptor and [125I]BP-Ang IV was examined by cross-linking the [125I]BP-Ang IV to the receptor in the presence of AT1, AT2 or AT4 specific competitors. The binding conditions and the separation of the bound and free ligands were described as above. After binding and cross-linking of the [125I]BP-Ang IV with the AT4 receptor, 100 μl of each sample was methanol/chloroform precipitated. The protein pellet was dissolved in 40 μl of the Laemmli sample buffer, heated at 57°C for 20 min and analyzed by SDS-PAGE using the Laemmli buffer system at constant current of 20 mA for 100 min at room temperature. After electrophoresis, the gel was vacuum dried and autoradiography was performed using Kodak X-ray film (Sigma) at –80°C. The compounds (1 μM) used as competitors of [125I]BP-Ang IV included PD123319 (an AT2 antagonist), DuP753 (an AT1 antagonist), Ang II, CGP42112A (an AT2 antagonist) and divalinal-Ang IV (an AT4 antagonist). Nonradiolabeled Ang IV was also included as AT4 specificity control.
Elucidation of the presence of disulfide-bonds in the AT4 receptor complex.
The existence of disulfide bonds in the AT4 receptor complex was studied by comparing the molecular weights of the reduced and nonreduced AT4 receptor as determined by SDS-PAGE. The solubilized AT4 receptor was cross-linked with [125I]BP-Ang IV and separated from the free [125I]BP-Ang IV using the P-6 column. When visualizing the nonreduced AT4receptor, 20 μl of the cross-linked sample was mixed with 20 μl of Laemmli loading buffer without DTT. For the visualization of reduced AT4 receptor, 20 μl of the cross-linked sample was mixed with 20 μl of Laemmli loading buffer containing 0.1 M DTT. The samples were heated at 57°C for 20 min and analyzed by SDS-PAGE using standard Laemmli buffer system (see above). After electrophoresis, the gel was vacuum dried and autoradiography was performed.
Glycosylation analysis of the AT4receptor.
The presence of the N-linked carbohydrate side chains on the AT4 receptor was determined by deglycosylation analysis of the solubilized AT4receptor cross-linked with [125I]BP-Ang IV. To a 1.5 ml microcentrifuge tube, 100 μl of the sample was added and precipitated using the methanol/chloroform method. The protein pellet was then dissolved in 35 μl of a buffer containing 0.5% SDS and 0.1% βME. After the proteins went into the solution, the following components were added in sequence: 25 μl if 0.5 M Tris-Cl, pH 8.0; 10 μl of 0.1 M 1,10-phenanthroline in methanol; 10 μl of 10% of NP-40; and 5 μl of PNGase F (Sigma). The deglycosylation reaction was performed overnight at 37°C. When the reaciton was finished, 15 μl of distilled water was added and methanol/chloroform precipitation was performed. The pellet was then dissolved in 40 μl of Laemmli loading buffer and subject to SDS-PAGE analysis. The gel was then dried, and autoradiography was performed.
Comparison of the AT4 receptors from selected bovine tissues.
The AT4 receptors prepared from different bovine tissues (adrenal, hippocampus, kidney, aorta and thymus) according to the standard method (Hanesworth et al., 1993) were cross-linked with [125I]BP-Ang IV under UV light. The separation of bound from free ligands was carried out using the P-6 column prepared as described above. From each protein preparation, 200 μl of sample was withdrawn and precipitated by methanol/chloroform. The pellet was then dissolved in 40 μl of Laemmli loading buffer and analyzed by SDS-PAGE. The gel was vacuum dried and autoradiography was carried out.
Large-scale purification of the AT4receptor from bovine adrenal.
The specific photolabel, [125I]BP-Ang IV, was used to guide the purification of the AT4 receptors from bovine adrenals as summarized in table 4. Total membrane proteins from bovine adrenal glands were prepared as described previously. For every 15 g of the bovine adrenal tissue, ∼30 ml of solubilized membrane protein at a concentration of 1 mg/ml was obtained. To each preparation, 1 ml of a solution containing [125I]BP-Ang IV cross-linked AT4 receptor was added to monitor the purification process. The protein sample was desalted and concentrated using a MacroSep centrifugal concentrator (Filton, Northborough, CA). Briefly, 30 ml of the sample prepared above was divided into 6-ml aliquots and mixed with 6 ml of distilled water. Each 12 ml aliquot was then loaded into the concentrator and centrifuged at 5000 × g for 2 to 3 hr at 4°C or until the volume was reduced to 3 to 4 ml. The volume was then brought up to 12 ml with distilled water and centrifuged again. This process was repeated one more time for a total of 3 centrifugations and then mixed with the isoelectro-focusing buffer (3% CHAPS, 18% glucose, 50 mM DTT, 1.2% ampholyte). Proteins were then separated by isoelectro-focusing using the Rotofor Cell (BioRad, Hercules, CA) as directed by the accompanying instruction manual. For each Rotofor run, ∼6 ml of the protein sample containing cross-linked [125I]BP-Ang IV was separated at 4°C for 4 hr on a pH gradient created by the added ampholytes (1–10) (BioRad). Following focusing, 20 fractions were harvested and the radioactivity of each fraction was counted using a Gamma counter (ICN). The pH value of each fraction was measured and the 1 to 2 fractions with the highest counts and the expected pH value (∼4.8) were collected. Proteins in the peak fractions were precipitated by the methanol/chloroform method. The pellets were dissolved in Laemmli loading buffer without dye and subject to size fractionation using an Electro-Fractionator (patent pending) through which the majority of the proteins with molecular weight less than 100 kDa were removed. After reducing the volume to 1 to 1.5 ml using the Macrosep spin column (Filtron), samples were precipitated and subjected to SDS-PAGE (6–12% gradient). The polyacrylamide gel was stained with Coomassie blue R250 (0.2% in water). After destaining, the protein bands were excised and those containing the AT4 receptor band were identified by the radioactivity detected using the Gamma counter (ICN). A major protein band with molecular weight of ∼165 kDa that contained the radioactivity was eluted out from the gel using an Electro-Eluter (Bio-Rad) following the manufacturer instructions. The eluants were then concentrated and precipitated as described above. The eluants accumulated from 15 Rotofor runs were combined and subject to a second SDS-PAGE (5%). The gel was stained with 0.2% Coomassie blue R250 solution containing 5% acetic acid and 45% methanol for 15 to 20 min and destained in the same solution, but without dye, until the background was clear (about 2 hr with 3 changes of destaining solution). Autoradiography of the dried polyacrylamide gel was performed to confirm the identity of the protein band as the AT4 receptor.
Results
Comparison of membrane-bound and solubilized AT4 receptors from bovine adrenals.
The AT4 receptor from bovine adrenal gland membranes were solubilized with the zwitterion detergent CHAPS and compared to the membrane bound receptor in terms of binding characteristics and ligand specificity. Solubilized receptors bound [125I]Ang IV in a reversible manner yielding a calculated kinetic KD of 1.18 × 0.14 × 10−10 M (table1), which was similar to that observed for the membrane-bound receptor (table 1). In addition to binding [125I]Ang IV reversibly, the solubilized receptor bound saturably (fig. 1) as has been previously shown for the membrane-bound adrenal receptor (Krebset al., 1996). The binding of [125I]Ang IV to the solubilized receptor exhibited high affinity similar to the membrane bound receptor. In order to further compare the solubilized and membrane-bound receptor the binding affinity of several angiotensin ligands for the receptor was examined (table 2). Both receptor preparations exhibited similar structure-binding profiles. The binding profile for both the membrane-bound and solubilized receptors was “AT4-like” as compared to previously characterized receptors (Wright et al., 1995). Namely, neither receptor bound the AT1/AT2 ligands Ang II (AT1 and AT2), Sar1Ile8-Ang II (AT1 and AT2), losartan (AT1) and CGP42112A (AT2). As expected (Sardinia et al., 1993), neither receptor preparation bound N-terminal modified Ang IV analogd-Val1-Ang IV or des-Val Ang IV. This loss of binding is consistent with the critical role played by the N-terminal l-Val in high affinity AT4 binding (Sardinia et al., 1994). The angiotensin related peptide, Ang II(1–7), which has been suggested to mediate unique physiological actions via its own specific receptor, bound with low affinity to both receptor preparations. The only ligands tested that bound with high affinity were Ang IV and Ang III. The binding of the highly labile Ang III was previously shown to be dependent upon its prior conversion to Ang IV (Swanson et al., 1992). Together these data indicate that the solubilized AT4 receptor possesses the same binding characteristics as the membrane-bound receptor and thus validate its use as a source of AT4 receptor for further characterization and purification.
Specificity of [125I]BP-Ang IV for the AT4 receptor.
To characterize the AT4 receptor in terms of subunit structure, glycosylation content, disulfide composition and to guide purification, it was necessary to develop a tool that would allow the specific visualization of the AT4 receptor in polyacrylamide gels. The visualization tool that was chosen to use was an 125I-labeled angiotensin IV analog that contained the photoreactive moiety, p-benzoylphenylalanine (BP). Before 125I-I1, BP6, G-Ang IV (BP-Ang IV) could be used for the stated purpose it was necessary to verify that BP-Ang IV bound specifically to the AT4 receptor and that it could be photocross-linked yielding an AT4 receptor with a covalently attached 125I label.
The initial validation studies focused on establishing the binding specificity of the [125I]BP-Ang IV photoprobe. Saturation isotherms were developed for the binding of [125I]BP-Ang IV and [125I]-Ang IV to identical membrane preparations from bovine adrenals. As can be seen in figure2, which shows typical binding curves, the Bmax for the binding of both labeled ligands was very similar, consistent with the notion that both ligands bound the same site exclusively. Combined data (n = 4) yielded similar results ([125I]BP-Ang IV:Bmax = 967 (100.9 fmol/mg protein,KD = 0.72 (0.12 nM; [125I]Ang IV: Bmax= 1004.1 (98.4 fmol/mg protein; KD = 0.89 [0.22 nM; mean (S.D.)]. The suggestion that both labeled ligands bound the AT4 receptor exclusively was further strengthened by a comparison of competition curves in which a selection of angiotensin ligands were employed as competitors. The inhibitory binding constants of the angiotensin ligands for [125I]-Ang IV and [125I]BP-Ang IV binding are summarized in table3. The data demonstrated that the rank-order binding affinity of the tested ligands was identical for both [125I]Ang IV and [125I]BP-Ang IV binding. In addition, the absolute Ki values of the ligands for each labeled ligand was also found to be similar. Another important outcome of the competition studies was the obsevation that BP-Ang IV could completely compete for [125I]Ang IV binding and conversely Ang IV could completely compete for [125I]BP-Ang IV binding.
A final test for the specificity of [125I]BP-Ang IV for the AT4 receptor was the direct examination of the ability of several angiotensin analogs (including AT1, AT2 and AT4 specific ligands) to compete for the covalent labeling of protein from solubilized bovine adrenal membranes. Thus, the [125I]BP-Ang IV-AT4binding reaction was carried out under the following conditions: (1) with no competitor, (2) in the presence of the AT4 specific ligands Ang IV and divalinal-Ang IV (Krebs et al., 1996), which would be expected to compete with [125I]BP-Ang IV if it did indeed bind to the AT4 receptor and (3) in the presence of the AT1 and AT2 specific ligands losartan, Ang II, Sar1-Ile8-AngII, DuP753, CGP42112A and PD123319, respectively, which would be expected to be ineffectual competitors. Following the binding reaction and the separation of bound and free ligand by gel filtration, the bound [125I]BP-Ang IV was photocross-linked with UV light. The samples containing the solubilized and potentially labeled AT4 receptor were then subjected to SDS-PAGE analysis under reducing conditions. Autoradiography of the polyacrylamide gels (fig. 3A and 3B) revealed a single 165-kDa band characteristic of the AT4 receptor in figure 3A, lanes 1, 3 to 6, and 3B, lanes 1 to 2 and 5. No bands were observed in figure 3A, lane 2, and 3B, lanes 3 and 4. The samples in figure 3A, lanes 3 to 6, and 3B, lanes 1 to 2, contained either AT1 or AT2 competitors, whereas figure 3A, lane 1, and 3B, lane 5, contained no competitors. Figure 3A lane 2, and 3B, lanes 3 and 4, contained AT4 specific competitors. The results indicated that the AT1 specific ligand, DuP753 (fig. 3B, lane 2) and the AT2 specific ligands including PD123319 (fig. 3B, lane 1) and CGP42112A (fig. 3A, lane 6), and the AT1/AT2specific ligands, Sar1-Ile8-Ang II (fig. 3A, lane 4), and Ang II (fig. 3A, lane 3) were not able to block the binding and ultimate cross-linking of [125I]BP-Ang IV with the AT4 receptor while the AT4specific ligands Ang IV (fig. 3A, lane 2, and 3B, lane 4) and divalinal-Ang IV (fig. 3B, lane 3) completely blocked the binding of [125I]BP-Ang IV to the AT4 receptor.
While it is likely that the ability of [125I]BP-Ang IV to specifically label a 165-kDa protein that withstands the harsh denaturing environment of SDS-PAGE indicates a covalent linkage, the covalency of the association between [125I]BP-Ang IV and the 165 kDa AT4 receptor was further verified by data shown in figure 4. The data in figure 4indicate that the efficiency of the association between [125I]BP-Ang IV and the AT4 receptor was, as predicted, dependent on the UV exposure time. No UV exposure (fig. 4, lane E) leads to no stablely labeled 165-kDa protein on the gel. Together these data indicated that [125I]BP-Ang IV covalently labels the AT4 receptor exclusively, thus making it an ideal tool to visualize AT4 receptors on gels and mark the receptor during purification.
The AT4 receptor contains more than one subunit.
To elucidate the subunit structure of the AT4 receptor, the solubilized membrane proteins containing the AT4 receptor were cross-linked with [125I]BP-Ang IV and the mobility of cross-linked receptors examined under both reducing and nonreducing conditions. There were no labeled protein bands observed in lanes A and C, the nonspecific binding controls for both reduced and nonreduced AT4 receptors (fig.5). Lane B containing the nonreduced sample revealed two bands. The major band had the molecular weight of ∼165 kDa, whereas the minor band had the molecular weight of ∼ 225 kDa. The relative abundance of the 165- and 225-kDa bands was ∼9:1. In contrast, lane D containing the reduced sample only revealed a single ∼ 165-kDa protein band which is identical to the lower protein band in lane B. It appeared that some portion of the AT4 receptor was associated with one or more other proteins by disulfide links. The total mass of protein(s) associated with the Ang IV-binding protein was estimated as 60 kDa. More interestingly, protein samples run under native conditions (Sambrook et al., 1989), without SDS and DTT added in the buffer and gel system, exhibited only a single protein band of larger than 300 kDa (Data not shown). This suggests that the native AT4 receptor may contain additional proteins that were associated through mechanisms other than disulfide bonds. The 165 kDa protein that contained the Ang IV binding site was designated as the alpha subunit of the AT4 receptor complex.
The alpha subunit of the AT4receptor contains a large proportion of carbohydrate side chains.
Autoradiography of the 165-kDa subunit of AT4receptor revealed a broad protein band suggesting that this subunit may contain carbohydrates or other post translational modifications. As an initial attempt to determine if carbohydrates were associated with thealpha subunit, N-linked glycosylation analysis of thealpha subunit of the AT4 receptor was conducted using PNGase F digestion followed by SDS-PAGE. The apparent molecular weight of the alpha subunit of the AT4 receptor was decreased by PNGase F to ∼130 kDa from 165 kDa as displayed in figure6. Therefore, N-linked carbohydrates constituted approximately 20% of the mass of the alphasubunit of the AT4 receptor based on SDS-PAGE analysis. By contrast, incubation of the receptor at 37°C overnight without addition of PNGase F had no effect on its molecular weight.
Structural variation of the AT4 receptor from selected bovine tissues.
Autoradiographic studies have shown that the AT4 receptor is present in many different tissues (Wright et al., 1995). To begin to compare the structural features of the AT4 receptors from different tissues, five bovine organs (adrenal, hippocampus, aorta, kidney and thymus) were examined. The AT4receptors were solubilized from these tissues, cross-linked with [125I]BP-Ang IV and analyzed by SDS-PAGE under reducing conditions using the standard Laemmli buffer system as described above. An autoradiograph of the polyacrylamide gel is shown in figure 7. All the tested samples contained a protein band of ∼165 kDa, which corresponded to the alpha subunit of the adrenal AT4 receptor. However, the hippocampus and aorta contained additional protein bands with molecular weights of 150 (fig.7, lane D) and 125 kDa (fig. 7, lane B) respectively. Because the total membrane protein in each sample preparation was normalized, the intensity of the bands reflected the relative abundance of the AT4 receptors in different bovine tissues. Interestingly, in hippocampus the 165- and 150-kDa protein bands were of similar intensity suggesting similar abundance, whereas in aorta ∼90% of the [125I]BP-Ang IV cross-linkable protein was at 165 kDa. These results suggested that thealpha subunit of the adrenal AT4receptor represents the major binding protein for Ang IV in all tested tissues but that additional binding proteins, including possible receptor subtypes are present in hippocampus and aorta.
Purification of the AT4 receptor.
The radiolabeled Ang IV analog, [125I]BP-Ang IV, which was highly specific for the AT4 receptor was used to monitor the purification process (table4). Protein preparations containing a “spike” of AT4 receptor cross-linked with [125I]BP-Ang IV were desalted and separated by isoelectro-focusing using the Rotorfor (BioRad). The fraction with the highest level of radioactivity, which was typically fraction 7, had a pH of ∼4.82 (± 0.15, n = 19, mean ± S.D.) (fig. 8). Additional peaks of radioactivity seen in fractions 1 to 2 and fractions 12 to 15 were determined to represent [125I]BP-Ang IV associated noncovalently to denatured proteins or free [125I]BP-Ang IV, respectively. To definitively establish which fraction(s) contained the AT4 receptor, 100 μl of each fraction was precipitated using methanol/chloroform and analyzed by SDS-PAGE. Autoradiography of the polyacrylamide gel revealed a 165-kDa protein band only in fraction 7, whereas the rest of the labeled fractions contained no radioactively tagged protein bands (data not shown). The peak fraction identified by both radioactivity and pH value for each separation was then precipitated by methanol/chloroform, redissolved, and size fractionated through a 100-kDa cutoff membrane using the Electro-Fractionator. After concentration, the <100-kDa fraction was separated by SDS-PAGE (6–12% gradient). The polyacrylamide gel was stained with Coommassie blue R250 and the protein bands were excised. The radioactivity of each protein band was counted using a gamma counter. It was found that the protein band with a molecular weight of 165 kDa contained the highest radioactivity. The 165-kDa protein band from several gels was collected and separated a second time by SDS-PAGE (5%). The gel revealed two protein bands with a molecular weight of ∼165 and ∼70 kDa (fig. 9, lane B). To confirm that the 165-kDa protein band was the AT4 receptor, the gel was vacuum dried and autoradiography performed. The autoradiograph of the dried gel is displayed in figure 9, lane A, thus demonstrating that the 165-kDa Commassie blue-stained protein band was the AT4receptor. It was estimated that ∼5 μg of the alphasubunit of the AT4 receptor could be obtained from 50 g of bovine adrenals. The lower 70-kDa band was subsequently determined to be a receptor degradation product that was generated either during or in preparation for running of SDS-PAGE. A final SDS-PAGE (5%) separation of the pure 165-kDa protein again revealed the 70-kDa fragment (data not shown).
Discussion
The renin-angiotensin system has been the subject of intensive investigation for several decades. These studies have, in general, focused on the critical and integrative role of the angiotensin peptides Ang II and Ang III in cardiovascular regulation. One outcome of these investigations has been the impression that angiotensin peptides smaller than the heptapeptide Ang III are biologically inactive. This conclusion is not surprising given that smaller angiotensin fragments neither bind AT1 and AT2 receptors nor mimic the action of Ang II and Ang III with regards to classic cardiovascular and body water homeostatic physiologies. This long held perception, however, has recently been challenged by the discovery of at least two angiotensin fragments Ang II(1–7) (Ferrario et al., 1991) and Ang IV (Wright et al., 1995) that mediate unique physiological responses.
A novel receptor that binds Ang IV, termed the AT4 receptor, was subsequently discovered (Swanson et al., 1992) and found to be present in many different tissues including brain, heart, kidney, thymus and adrenal (Wright et al., 1995). To aid in the characterization and purification of the receptor, a radioactive photoactivated analog of Ang IV, [125I]BP-Ang IV, was developed in our laboratory. The key features of an effective photolabel includes the ability to participate in a stable covalent linkage to the target protein and absolute specificity for the target protein. To validate the utility of the [125I]BP-Ang IV for the AT4 receptor characterization, a series of competition binding assays were carried out using various compounds as competitors, The results revealed that only Ang IV and its analogs, including both agonists and antagonists, were able to block the binding and ultimate cross-linking of [125I]BP-Ang IV to the AT4 receptor. SDS-PAGE analysis of the AT4 receptor cross-linked with [125I]BP-Ang IV confirmed the specificity of [125I]BP-Ang IV, yielding a single protein band with a molecular weight of ∼165 kDa. This proposed size for the AT4 receptor is in close agreement with that suggested by Bernier (1995) who identified a 185-kDa protein following nonspecific chemical cross-linking of [125I]Ang IV. [125I]BP-Ang IV was readily cross-linked to the AT4 receptor forming a highly stable covalent linkage. Thus, [125I]BP-Ang IV possessed the critical characteristics that make it an effective tool for visualizing the AT4 receptor for the purposes of analysis and purification.
This investigation includes initial studies on the structural features of the AT4 receptor from bovine adrenal tissue as defined using [125I]BP-Ang IV labeling. Comparisons of the molecular size of reduced and nonreduced AT4 receptors by SDS-PAGE revealed that under nonreducing conditions, a ∼225-kDa protein band appeared in addition to the 165-kDa protein band (fig. 5) that is present in both reducing and nonreducing conditions. The ratio of the intensities between the 165- and 225-kDa proteins is ∼9:1. The significance of the fact that only 10% of the alpha subunits appear to be linked to other polypeptides by disulfide bonds is unknown. Conceivably this could be a manifestation of either multiple receptor subtypes, some of which are multimeric; receptors in activated and nonactivated states; or an artifact of the tissue and solubilization procedure that results in disulfide reduction. The later is probably not a viable explanation since preparation of tissue and solubilization of receptor in the presence of a mild oxidizing agent (NaNO2) had little effect on the SDS-PAGE pattern (data not shown). Although the total molecular weight of the protein(s) associated with the 165-kDa protein is ∼60 kDa, the number of associated proteins is unknown since there is no convenient way to monitor these proteins after reduction. More interestingly, under native conditions, the PAGE analysis of the AT4 receptor revealed only a single protein band with molecular weight of more than 300 kDa, indicating that the AT4 receptor may also be associated with additional proteins via forces other than disulfide bonds. Therefore, it appears that the native AT4receptor consists of several subunits that are associated with each other through different linkages. Nevertheless, the alphasubunit of the AT4 receptor was able to bind Ang IV and [125I]BP-Ang IV in the presence of 0.1% SDS and 1 mM DTT, indicating that the other subunits of the AT4 receptor are not required for the binding of Ang IV.
Many membrane-associated receptors are reported to contain post-translational modifications that affect the binding specificity, affinity and tissue distribution of the receptors. Some receptors have been found to have several subtypes that vary structurally only in their post-translational modifications, such as glycosylation. Glycosylation analysis of the alpha subunit of the AT4 receptor from bovine adrenals indicated that N-linked carbohydrate chains contributed ∼20% of the apparent molecular mass. However, further study is required to quantitate the contents of glycosylation of the AT4-alpha subunit. Comparisons of the AT4 receptor subunits of different bovine tissues revealed that all the tested tissues contained a labeled band similar in size to the alpha subunit of the AT4 receptor of the adrenal. However, hippocampus and aorta contained additional bands with molecular weights of 150 and 125 kDa, respectively. Whether these two smaller proteins are the result of glycosylation difference, heterogenity or more basic structural differences needs to be investigated further. Nevertheless, the notion that the AT4 receptor may exist as a family of subtypes is not unexpected and is further supported by three additional unpublished observations that have been seen in our laboratory. First, the affinity of various AT4ligands is tissue dependent. Second, certain AT4ligands exhibit distinct multisite binding curves in particular tissues. And third, the apparent physiological potency or activity (agonism or antagonism) of a particular ligand is dependent on the physiological process that is being examined. Cloning of the AT4 receptor and subsequent studies of the AT4 receptor-protein interactions using the yeast two-hybrid system and immunoprecipitation methods should help to clarify the structure and functional identify of the AT4 receptor.
Acknowledgments
We are very grateful to Bea O’Neill, Laureen Poesy and Jeff Helm for their technical assistance.
Footnotes
-
Send reprint requests to: Dr. Joseph W. Harding, Department of Veterinary and Comparative Anatomy, Physiology and Pharmacology, Washington State University, Pullman, WA 99164-6520.
- Abbreviations:
- Ang
- angiotensin
- AT4
- receptor of angiotensin IV
- BP-Ang IV
- Ile1,125I-Tyr-benzophenone6,Gly7-angiotensin IV
- SDS-PAGE
- sodium dodecyl sulfate-polyacrylamide gel electrophoresis
- β-ME
- β-mercaptoethanol
- DTT
- dithiolthrietol
- EDC
- 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide
- HF
- hydrogen fluoride
- Received November 12, 1997.
- Accepted May 19, 1998.
- The American Society for Pharmacology and Experimental Therapeutics