Unexpected Binding of an Octapeptide to the Angiotensin II Receptor (Renin-Angiotensin System/Hypertension/Angiotensin Antagonist) R

Proc. Natl. Acad. Sci. USA Vol. 84, pp. 9219-9222, December 1987 Medical Sciences Unexpected binding of an octapeptide to the angiotensin II receptor (renin-angiotensin system/hypertension/angiotensin antagonist) R. L. SOFFER*, S. BANDYOPADHYAY*, E. ROSENBERG*, P. HOEPRICHt, A. TEITELBAUMt, T. BRUNCKt, C. B. COLBYt, AND C. GLOFFtt *Department of Medicine, Cornell University Medical College, New York, NY 10021; and tTriton Biosciences Inc., 1501 Harbor Bay Parkway, Alameda, CA 94501 Communicated by William A. Goddard III, August 17, 1987 (receivedfor review April 24, 1987)

ABSTRACT An octapeptide, TBI-22 (Lys-Gly-Val-Tyr- MATERIALS AND METHODS Ile-His-Ala-Leu), inhibited binding of angiotensin II by a III-labeled angiotensin II ('l25-angiotensin II; 1400 puCi/nmol; solubilized angiotensin receptor partially purified from rabbit 1 Ci = 37 GBq) was purchased from New England Nuclear. liver. This inhibition appears to result from competition for t-Butoxycarbonyl (Boc) derivatives of amino acids were from binding to the same receptor. Radioiodinated TBI-22, like Peninsula Laboratories (Belmont, CA). Diisopropylcarbodii- angiotensin I, bound to the solubilized receptor with an mide and 1-hydroxybenzotriazole were products of Aldrich. affinity such that the binding was inhibited 50% by unlabeled Peptides were synthesized by the solid-phase procedure TBI-22 or angiotensin HI at nanomolar concentrations. The (12), using an automated synthesizer, model 9500 from binding reaction, like that for angiotensin HI, required p- Biosearch (San Rafael, CA). The first amino acid was chloromercuriphenylsulfonic acid and was reversed in the esterified to chloromethylated polystyrene-divinylbenzene presence of dithiothreitol. TBI-22 and angiotensin II share the copolymer (Bio-Rad; 1% cross-linked, 1.34 meq/g) as de- sequence Val-Tyr-Ile-His; this tetrapeptide alone, however, did scribed (13). Subsequent amino acids were coupled twice by not inhibit binding of angiotensin II. Replacement of the using diisopropylcarbodiimide. In general, the Boc group was tyrosine residue by aspartic acid in TBI-22 greatly reduced the removed by treatment of the resin containing protected ability of the peptide to compete with angiotensin H for peptide for 20 min with 45% trifl uoroacetic acid in dichloro- binding, suggesting an important contribution of this residue to methane (vol/vol) followed by two neutralizations of 5 min the configuration required for revognition by the receptor. with 5% diisopropylethylamine in dichloromethane (vol/ vol). The resin was washed with appropriate solvents before The is involved in the and after each step of deprotection, neutralization, and renin-angiotensin system regulation of coupling. Removal of the Boc group and completeness of blood pressure (1). Angiotensin II is the biologically active coupling were monitored qualitatively by a ninhydrin color component of this physiological system (2). Angiotensin II is test (14). All peptides were cleaved from the resin with a potent vasopressor agent as a consequence of its contrac- simultaneous removal ofthe protecting group by exposure to tive effect on blood vessels (2), its stimulation of aldosterone anhydrous hydrogen fluoride/10% anisole (vol/vol) for 40 secretion by the adrenal cortex (3), and its action on the min at 0-40C. Hydrogen fluoride and anisole were removed central nervous system (4). These and other effects of from the resin by evaporation under reduced pressure pro- angiotensin II are mediated by a specific receptor(s) on the vided by a water aspirator, followed by washing with anhy- surface of its target cells (reviewed in ref. 5). drous diethyl ether. The peptide was then obtained with Angiotensin-converting enzyme (reviewed in ref. 6) is a sequential washes ofdimethylformamide, 50% dimethylform- dipeptidyl carboxypeptidase that cleaves angiotensin I to amnide/10% acetic acid/40% water (vol/vol), 10% acetic angiotensin II. The use of drugs that inhibit angiotensin- acid/90% water (vol/vol), and distilled water. The combined converting enzyme (7, 8) has been shown to be an effective washings were finally lyophilized. antihypertensive therapy. Angiotensin-converting enzyme The peptides were then purified by reverse-phase high- possesses broad substrate specificity, and therefore its inhi- performance liquid chromatography. The various peptides bition is likely to influence the metabolism of peptides were chromatographed on a preparative column (Whatman; unrelated to the renin-angiotensin system. An antagonist Partisil 10 ODS-3 Magnum 20, 2.2 x 50 cm); the elution was targeted at the angiotensin receptor might represent a more 65 min and the gradient ranged from 83% solvent A (0.05% physiologically specific agent for antihypertensive therapy. trifluoroacetic acid/water) and 17% solvent B (0.05% triflu- Analogs of angiotensin II with substitutions of the carboxyl- oroacetic acid/acetonitrile) to 50% A and 50% B. A portion terminal residue have shown promise in this regard (9-11). of each purified peptide was hydrolyzed in 5.7 M HCO and its However, jt may be that identification of determinants amino acid composition was determined (15). actually recognized by the receptor will be necessary to TBI-22 was radiolabeled by a modification of the method expedite development of the desired antagonist. To obtain of Greenwood et al. (16). The peptide (10 nmol) was incu- this type of information, examination of the binding of bated at 20'C for 30 sec with 4 mCi of Na125I (Amersham) in putative ligands by isolated receptor may be useful. To date, 70 Al of0.5 M potassium phosphate buffer, pH 7.4, containing there have been few such studies, and virtually all of them 10 ,ug of chloramine T. The reaction was stopped by addition have been carried out with intact membrane fractions (5). of 1 mg of 2-mercaptoethanolamine (Sigma) followed by 0.5 Here, we report the binding of an octapeptide and some ofits ml of0.1 M potassium phosphate buffer, pH 7.4. The solution analogs and derivatives to a solubilized angiotensin II recep- was then transferred to a C8 Sep-pak column (J. T. Baker tor partially purified from rabbit liver. Chemical, Phillipsburg, NJ) and unincorporated 1251 was washed off with 5 ml of 0.1 M phosphate buffer. Radioiodi- The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" Abbreviation: PCMS, p-chloromercuriphenylsulfonic acid. in accordance with 18 U.S.C. §1734 solely to indicate this fact. fTo whom reprint requests should be addressed.

9219 9220 Medical Sciences: Soffer et al. Proc. Natl. Acad. Sci. USA 84 (1987) nated TBI-22 (125I-TBI-22; 120 ,uCi/nmol) was eluted with Table 1. Binding of radioiodinated TBI-22 by solubilized glacial acetic acid and lyophilized. angiotensin II receptor The angiotensin II receptor was isolated as described (17) Binding, Residual binding, except that the SE-Sephadex and heat treatment steps were System cpm % replaced by chromatography on DEAE-cellulose. The DEAE- cellulose was equilibrated in 20 mM Tris-HCl, pH 7.6, Complete 32,764 100 containing 1 mM K2EDTA and 0.5% Brij 99 (Emulsion + 10 ,uM TBI-22 212 1 Engineering, Sanford, FL). After application of protein (10 + 10,M angiotensin II 290 1 mg/ml of bed volume), the column was washed with 20 bed - PCMS 596 2 volumes of a linear gradient ranging from 0.0 to 0.3 M KCI. + 200 ,uM dithiothreitol* 1,162 4 The receptor emerged as a symmetrical peak at about 0.13 M The complete system is described in Materials and Methods. The KCl. The active fractions were pooled and dialyzed against value obtained in the absence of receptor (277 cpm) has been equilibration buffer. The binding activity of the partially subtracted from all data. purified receptor, estimated by Scatchard analysis, was 18 *Added after standard 60-min incubation and reaction mixture pmol of angiotensin II per mg of protein. processed after additional 3 min. Reaction mixtures (150 ,u1) for binding assays contained 30 mM Tris-HCl at pH 7.6, 2.5 mM K2EDTA, 0.2 mM p- II to the partially purified receptor in a dose-dependent chloromercuriphenylsulfonic acid (PCMS), 100 ,ug of bovine manner (Fig. lA). The IC50 is approximately 8 nM. By serum albumin, 30 ,ug of receptor preparation, 0.25% Brij 99, comparison, unlabeled angiotensin II has an IC50 equal to 12 and 100,000 cpm of 125I-angiotensin II (0.25 nM) or 115I-TBI-22 nM in the same system. To determine the biochemical (3 nM). After incubation at 20°C for 60 min, bound radioactive mechanism for this effect ofTBI-22, a radioiodinated sample peptide was estimated as described (18) except that 0.05% of this octapeptide was prepared and its binding by the dextran/0.5% charcoal was employed. ICso values (concentra- receptor preparation was examined (Table 1). The receptor- tion yielding 50% inhibition) for competing unlabeled peptides dependent binding was completely blocked by 10 ,uM un- were estimated from inhibition dose-response curves com- labeled TBI-22 or arigiotensin II. The binding reaction re- posed of at least four points. Results were corrected for quired the presence of PCMS, and the receptor-ligand "nonspecific" binding-i.e., cpm bound in the presence of 10 complex was dissociated by dithiothreitol. These unusual ,uM unlabeled angiotensin II, which was essentially identical to characteristics are identical to those we previously described that measured in the absence of receptor and less than 2% of for the binding of angiotensin II to this solubilized rabbit that obtained with the complete system. Peptides that did not hepatic receptor (17). The data on binding of radioiodinated inhibit at 10 p.M were not further examined. TBI-22 thus provide strong evidence that this peptide is Rabbit antiserum to angiotensin II was developed as bound by the angiotensin II receptor and that it inhibits the described by Gocke et al. (19). Immunoglobulins were binding of angiotensin II by competing for the same binding prepared from it by heat treatment, precipitation at 50% site. saturation with ammonium sulfate, and passage through As anticipated from these results, unlabeled TBI-22 inhib- DEAE-cellulose (20). Rabbit anti-TBI-22 immunoglobulins ited the binding of 3 nM radioiodinated TBI-22 at somewhat were obtained from J. E. Blalock (University of Alabama, lower concentrations than did angiotensin II (Fig. 1B), with Birmingham) and prepared according to Bost et al. (21). an IC50 for TBI-22 of 95 nM compared with a value for angiotensin II of 140 nM. The ratio of these IC50 values is RESULTS essentially identical to the ratio of IC50 values obtained when Binding of TBI-22 by the Angiotensin II Receptor. TBI-22 using the two molecules to inhibit radioiodinated angiotensin prevented the binding of 0.25 nM radioiodinated angiotensin II binding. In the radioiodinated TBI-22 binding assay, the A B 100

-- 0 0- cr c

-5 Q) V0n

._0 nf a1

10-7 10-8 10-6 Peptide (M) FIG. 1. Competition by unlabeled TBI-22 (o) and angiotensin II (o) with 0.25 nM radioiodinated angiotensin II (A) or 3.0 nM radioiodinated TBI-22 (B) for binding to the partially purified receptor. Medical Sciences: Soffer et al. Proc. Natl. Acad. Sci. USA 84 (1987) 9221

Table 2. Inhibition by TBI series peptides of the binding of radioiodinated angiotensin II to its receptor Peptide Structure IC50,* nM Angiotensin II Asp-Arg-Val-Tyr-Ile-His-Pro-Phe 12 TBI-22 Lys-Gly-Val-Tyr-Ile-His-Ala-Leu 8 TBI-33 A11-D Lys-Gly-Val-Tyr-Ile-His-Ala-Leu >>10,000 TBI-48 Leu-Ala-His-Ile-Tyr-Val-Gly-Lys >>10,000 TBI-27 Lys-Gly-Val-Tyr-Ile-His-Ala-Leu-Pro 7 TBI-30 Ac-Lys-Gly-Val-Tyr-Ile-His-Ala-Leu-Pro 300 TBI-29 Lys-Gly-Val-Tyr-Ile-His-Ala-Leu-NH2 40 TBI-26 Ac-Lys-Gly-Val-Tyr-Ile-His-Ala-Leu-NH2 2,400 TBI-75 Val-Tyr-Ile-His >>10,000 TBI-76 Ile-His-Ala-Leu >>10,000 TBI-77 Val-Tyr-Ile-His-Ala-Leu 2,500 TBI-31 Lys-Gly-Val-Asp-Met-His-Ala-Leu 5,000 TBI-61 Lys-Gly-Val-Tyr-Met-His-Ser-Leu 6 *Determined from an inhibition dose-response curve composed of at least four concentrations; >>10,000 denotes complete lack of inhibition at 10 MM. required concentrations of both unlabeled peptides were binding of the latter was detectable. The antigenic determi- proportionately higher, because of the higher concentration nants of these peptides thus seem to differ from those of labeled TBI-22 used. Use of the higher concentration of recognized by the receptor. labeled material was required due to a lower specific radioactivity of the labeled TBI-22. DISCUSSION Competition for Binding to the Angiotensin Receptor by Peptides Related to TBI-22. Table 2 presents IC50 values of There is limited information describing the effect of substi- various unlabeled peptides; respective abilities to compete tutions within the angiotensin II molecule on its binding by with 0.25 nM radioiodinated angiotensin II for binding to its cell-free receptor preparations. Studies with intact mem- receptor were measured. TBI-33, the all-D analog of TBI-22, branes from adrenal cortex have suggested that the requisite and TBI-48, which contains its residues in reversed se- peptide conformation depends on the carboxyl-terminal quence, did not inhibit binding, indicating biological speci- tetrapeptide of the molecule (5). TBI-22 shares only two of ficity for the binding of TBI-22. Several derivatives contain- these four residues with angiotensin II, and TBI-61 shares ing groups that might be expected to enhance metabolic only one. TBI-27, the nonapeptide analog of TBI-22 with a stability were examined. TBI-27, the nonapeptide resulting carboxyl-terminal proline residue, is bound as well as the from prolyl addition at the carboxyl terminus of TBI-22, parent compound. Although TBI-22 shares residues 3 exhibited the same IC50 value as TBI-22. However, the through 6 with angiotensin II, these residues as a tetrapeptide acetylated derivative of the nonapeptide, TBI-30, was ap- do not contain the structural signal required for recognition- proximately 1/40th as inhibitory. Similarly, whereas amida- i.e. it is inactive in the binding reaction. tion of TBI-22 yielded a peptide (TBI-29) with an IC50 value It seems clear, however, that one of these residues, only 5-fold higher than the parent compound, further deriv- tyrosine in position 4, is important for the appropriate atization by acetylation was associated with an additional conformation, since replacement by aspartic acid yields an 60-fold increase in the IC50 value for the resulting doubly octapeptide that is less tightly bound. Replacement of the blocked peptide (TBI-26). tyrosine residue in position four of angiotensin II with Other peptide fragments and analogs of TBI-22 were phenylalanine as in [Phe4]- and [Phe4,Tyr8]angiotensin II analyzed to identify parts of its structure that might be results in decreased binding by particulate preparations (22, required for recognition by the receptor. Although TBI-22 23). Our binding data with the solubilized partially purified and angiotensin II possess the same residues in positions 3 hepatic receptor and fragments of TBI-22 are also consistent through 6, a tetrapeptide composed of this sequence (TBI- with those described for angiotensin II fragments binding to 75), failed to inhibit. The carboxyl-terminal tetrapeptide of adrenal membranes (3, 5), suggesting that most of the length TBI-22 (TBI-76) was similarly inactive, and even the car- of the octapeptide is required for assumption of the recog- boxyl-terminal hexapeptide (TBI-77) was relatively inactive, nized configuration. with an IC50 value of 2500 nM. These data resemble qualitatively those reported for inhibition of angiotensin II binding Table 3. Immunological relationship of TBI-22 and angiotensin II to canine adrenal membranes by carboxyl-terminal fragments Immunoglobulin, Mg/ml Bound of angiotensin 11 (3). TBI-31, approximately 1/600th as inhibitory as TBI-22, contains Asp-Met instead of Tyr-Ile at 1251I-labeled Pre- Anti- Anti- radioactivity,* positions 4 and 5. TBI-61, which also contains a methionine ligand immune TBI-22 angiotensin II cpm residue in position 5, was as potent as TBI-22 at inhibiting TBI-22 20 - 14 angiotensin II binding. Thus, the decrease in inhibitory 20 1875 activity of TBI-31 relative to TBI-22 is apparently the con- 100 398 sequence of substitution of an aspartic residue for tyrosine in Angiotensin II 67 90 position 4; this suggests that tyrosine is important in binding - 670 74 of ligand to the receptor. 0.5 1522 Immunological Studies on TBI-22 and Angiotensin II. An- *Reaction mixtures (150 p1) contained 30 mM Tris HCl at pH 7.6, tibodies developed against either TBI-22 or angiotensin II 0.15 M NaCl, 2.5 mM K2EDTA, 100 ,ug of bovine serum albumin, were incapable of binding both peptides with comparable either 'l25-TBI-22 (3700 cpm) or 1251I-angiotensin 11 (2100 cpm), and avidity. As shown in Table 3, antibodies against TBI-22 failed rabbit immunoglobulins as indicated. Incubations were at 4°C for 16 to bind angiotensin II, and those against angiotensin II bound hr, after which bound radioactivity was determined by the dex- more tightly to angiotensin II than TBI-22, although some tran/charcoal technique. 9222 Medical Sciences: Soffer et al. Proc. Natl. Acad. Sci. USA 84 (1987) An angiotensin antagonist acting at the receptor level may (1981) in Biochemical Regulation ofBlood Pressure, ed. Sof- represent the most physiologically specific agent for phar- fer, R. L. (Wiley, New York), pp. 205-262. macological blockage of angiotensin II. Inhibition of renin 6. Soffer, R. L. (1981) in Biochemical Regulation ofBlood Pres- and converting enzyme, while diminishing production of sure, ed. Soffer, R. L. (Wiley, New York), pp. 123-164. 7. Ondetti, M. A., Rubin, B. & Cushman, D. W. (1977) Science angiotensin II, might alter the metabolism of other proteins 196, 441-444. and peptides. A receptor antagonist would also offer the 8. Patchett, A. A., Harris, E., Tristram, E. W., Wyvratt, M. J., advantages of preventing the effects of angiotensin II poten- Wu, M. T., Taub, D., Peterson, E. R., Ikeler, T. J., Ten tially generated by other enzymes such as tonin (24), and of Broecke, J., Payne, L. G., Ondeyka, D. L., Thorsett, E. D., angiotensin I, which, although only weakly bound by the Greenlee, W. J., Lohr, N. S., Hoffsommer, R. D., Joshua, H., receptor (5, 17), is markedly elevated in the circulation when Ruyle, W. V., Rothrock, J. W., Aster, S. D., Maycock, A. L., converting enzyme is inhibited (25). The promise of an Robinson, F. M., Hirschmann, R., Sweet, C. S., Ulm, E. H., angiotensin receptor antagonist stimulated much research in Gross, D. M., Vassil, T. C. & Stone, C. A. (1980) Nature the 1960s and early 1970s (9-11). Recent technological (London) 288, 280-283. 9. Marshall, G. R., Vine, W. & Needleman, P. (1970) Proc. Natl. progress, including relative ease of peptide synthesis, avail- Acad. Sci. USA 67, 1624-1630. ability of partially purified receptor preparations for binding 10. Pals, D. J., Massuci, F. D., Sipos, F. & Denning, G. S., Jr. studies, and advances in computer modeling, suggest that (1971) Circ. Res. 29, 664-672. development of a drug that acts at the receptor level has 11. Khosla, M. C., Smeby, R. R. & Bumpus, F. M. (1974) in become a feasible goal. These studies on the binding of Angiotensin, eds. Page, I. H. & Bumpus, F. M. (Springer, TBI-22 and related analogs by the solubilized receptor, and New York), pp. 126-161. investigations on the binding of angiotensin II and its frag- 12. Erickson, B. W. & Merrifield, R. B. (1976) in The Proteins, eds. Neurath, H. & Hill, R. L. (Academic, New York), 3rd ments by particulate fractions (5), suggest that the receptor Ed., Vol. 2, pp. 256-527. recognizes a specific configuration. This conformation is 13. Hoeprich, P. D. & Doolittle, R. F. (1983) Biochemistry 22, probably generated from at least seven amino acid residues. 2049-2055. The tyrosine in position four of both angiotensin II and 14. Kaiser, E., Colescott, R. L., Bassinger, C. D. & Cook, P. I. TBI-22 is critical. Identification of the preferred conforma- (1970) Anal. Biochem. 34, 595-598. tion and its simulation by a compound resistant to peptidase 15. Henrikson, R. L. & Meredith, S. C. (1984) Anal. Biochem. action might represent an important step in development of 136, 65-74. the desired antagonist. A compound that exhibits the desired 16. Greenwood, F. C., Hunter, W. M. & Glover, J. S. (1963) biochemical property ofbinding to the isolated receptor must Biochem. J. 89, 114-123. also be examined in more integrated systems (such as smooth 17. Sen, I., Bull, H. G. & Soffer, R. L. (1984) Proc. Natl. Acad. muscle preparations and intact animals) to determine its Sci. USA 81, 1679-1683. pharmacological properties, including its antagonistic and 18. Sen, I., Jim, K. E. & Soffer, R. L. (1983) Eur. J. Biochem. agonistic actions. 136, 41-49. 19. Gocke, D. J., Gerten, J., Sherwood, L. M. & Laragh, J. H. Work done at Cornell University Medical College was supported (1969) Circ. Res. Suppl. 1, 24-25, 131-146. in part by Grant 3P50 HL18323 from the National Institutes ofHealth 20. Das, M. & Soffer, R. L. (1976) Biochemistry 15, 5088-5094. and by a grant from Triton Biosciences Inc. 21. Bost, K. L., Smith, E. M. & Blalock, J. E. (1985) Proc. Natl. Acad. Sci. USA 82, 1372-1375. 1. Skeggs, L. T., Dorer, F. E., Kahn, J. R., Lentz, K. E. & 22. Lin, S. Y. & Goodfriend, T. L. (1970) Am. J. Physiol. 218, Levine, M. (1981) in Biochemical Regulation of Blood Pres- 1319-1328. sure, ed. Soffer, R. L. (Wiley, New York), pp. 3-38. 23. Glossman, H., Baukal, A. J. & Carr, K. J. (1974) J. Biol. 2. Helmer, 0. M. (1957) Am. J. Physiol. 188, 571-577. Chem. 249, 825-834. 3. Capponi, A. M. & Catt, K. J. (1979) J. Biol. Chem. 254, 24. Demassieux, S., Boucher, R., Grise, C. & Genest, J. (1976) 5120-5127. Can. J. Biochem. 54, 788-795. 4. Phillips, M. I. (1978) Neuroendocrinology 25, 354-377. 25. Ferguson, R. K., Brunner, H. R., Turini, G. A., Gavras, H. & 5. Capponi, A. M., Aguilera, G., Fakunding, J. L. & Catt, K. J. McKinstry, D. N. (1977) Lancet i, 775-779.