Proc. Nati Acad. Sci. USA Vol. 78, No. 2, pp. 757-760, February 1981 Biochemistry

Synthesis of [a-methyltyrosine-4] II: Studies of its conformation, pressor activity, and mode of enzymatic degradation (angiotensin II analog/resistance to a-chymotrypsin degradation/full pressor activity/NMR and circular dichroism spectra similar to those of angiotensin II) M. C. KHOSLA*, K. STACHOWIAK, R. R. SMEBY*, F. M. BUMPUS*, F. PIRIOUt, K. LINTNERt, AND S. FERMANDJIANt *Research Division, The Cleveland Clinic Foundation, Cleveland, Ohio 44106; and tService de Biochimie, Department de Biologie, Centre d'Etudes Nucleaires de Saclay, P. B. No. 2, F-91190 Gif-sur-Yvette, France Communicated by Irvine H. Page, October 15, 1980

ABSTRACT Modifications in angiotensin II and its antago- full biological activity (agonist or antagonist), the backbone and nistic peptides that should have increased in vivo half-lives but not side-chain structure of the analog should resemble that of the reduced biological activity were studied by determining the effect hormone, All (2, 4). These observations may be helpful in the of a-methylation of the in position 4. [a-Methyltyrosine- 4]angiotensin II, synthesized by the solid-phase procedure, design of new potent analogs. showed 92.6 ± 5.3% pressor activity of angiotensin H. Incubation with a-chymotrypsin for 1 hr indicated absence of degradation EXPERIMENTAL although, under the same conditions, angiotensin H was com- O-Benzyl-a-Methyl-L-Tyrosine. The procedure used for the pletely degraded to two components. Comparison of the 'H NMR synthesis ofthis compound is a modification ofthe one used by spectra in aqueous solution and the circular dichroism spectra in trifluoroethanol of angiotensin II and [a-methyltyrosine- Khosla et aL (2) for the synthesis of 0-(2,6-dichlorobenzyl)-L- 4]angiotensin II suggested that a methylation of the tyrosine res- tyrosine. idue in angiotensin HI does not lead to major changes in the overall a-Methyl-L-tyrosine (3.904 g; 20 mmol) and 1.6 g of NaOH solution conformation. These results are in contrast to those ob- (20 mmol) in 10 ml of H20 were treated with 2.5 g of tained with N-methylation in position 4, which drastically reduced CuSO4-5H20 (10 mmol) in 10 ml of H20. The mixture was the biological activity and produced remarkable changes in the shaken on a wrist shaker for 10 min, and 60 ml of MeOH was peptide backbone and a severe limitation in rotational freedom of the side chains in tyrosine. Thus, it may be possible to synthesize added, followed by 2.38 ml of C6H5CH2Br (20 mmol). The flask potent angiotensin II analogs that have greater resistance to en- was stoppered, and the mixture was shaken for 20 hr at room zymatic degradation by a-methylation in position 4 (or 5) and si- temperature. The copper complex was removed by filtration, multaneous suitable modification at the NH2 and COOH termini. washed with three 10-ml portions of H20, two 10-ml portions of MeOH, and two 10-ml portions of Et2O. The semi-air-dried The antagonists for the pressor hormone angiotensin II (Asp- product (6.2 g) was suspended in 40 ml of 1 M HCl, pulverized, Arg-Val-Tyr-Ile-His-Pro-Phe; All) have proved useful in the and filtered out. The precipitate was repeatedly macerated with study ofexperimental hypertension and as possible new clinical 1 M HC1 and filtered out until the filtrate was almost colorless. diagnostic tools (1). However, their long-term application has The residue was then stirred with 20 ml of 16.4% AcONa, fil- been limited because they have a short in vivo half-life and, due tered out, washed with hot water, and dried at reduced pressure to degradation bypeptidases, are not orally active. We therefore over P205. O-Benzyl-a-methyl-L-tyrosine thus obtained was attempted to make these peptides resistant to this enzymatic homogeneous: thin layer chromatography (TLC) on silica gel, degradation by replacing the natural amino acid residues with RF 0.43 in n-BuOH/AcOH/H20 (4:1:5); RF 0.57 in n-PrOH/ N-methylamino (2) or /3homoamino (3) acid residues. These H20 (1:1). modifications drastically reduced the biological activity ofthese tert-Butyloxycarbonyl-O-Benzyl-a-Methyl-L-Tyrosine. A analogs, and conformation studies suggest that N-methylation suspension of 1.42 g ofO-benzyl-a-methyl-L-tyrosine (5 mmol) in positions 4 or 5 results in remarkable changes in the peptide in 30 ml of Me2SO was treated with 0.72 ml of NEt3 (5 mmol) backbone and a severe limitation in the rotational freedom of and 0.3 g oftert-butyloxycarbonyl azide (0.3 g), and the mixture the side chains in tyrosine, , and residues was stirred at 40°C under anhydrous conditions for 24 hr. The (4). The replacement ofan a proton by an a methyl group, how- addition ofBoc-N3 (0.3 g) and NEt3 (0.72 ml) was repeated every ever, is thought to produce minimal changes in backbone and 24 hr until the solid went into solution (1 week or more). The side-chain structures. The analogs thus obtained might mimic solution was diluted with 40 ml of H20, treated with NaOH to the parent hormone in recognizing and binding with the re- pH 9, and extracted with three 25-ml portions of Et2O. The ceptor on the cell membrane and, in addition, be stable to en- aqueous layer was evaporated to dryness at 20°C at reduced zymatic degradation (4, 5). Based on this hypothesis, we have pressure on a rotary evaporator. The residue was dissolved in synthesized [a-methyltyrosine4]angiotensin II ([a-MeTyr4]AII). 7 ml ofH20, and the solution was cooled to 0°C and treated with And, indeed, this peptide is resistant to chymotrypsin degra- ice-cold citric acid solution to pH 4.5. The solution was then dation and yet retains almost the full pressor activity of All. saturated with solid NaCl and extracted with AcOEt; the organic Conformation studies suggest minimum changes in backbone phase was washed with H20 and then with saturated NaCl so- and side-chain structures. These results again suggest that, for lution, dried with Na2SO4, and evaporated to yield 2.36 g of

The publication costs ofthis article were defrayed in part by page charge Abbreviations: AII, angiotensin II; [a-MeTyr4]AII, [a-methyltyrosine- payment. This article must therefore be hereby marked "advertise- 4]angiotensin II; TLC, thin layer chromatography; CD, circular ment" in accordance with 18 U. S. C. §1734 solely to indicate this fact. dichroism. 757 Downloaded by guest on September 29, 2021 758 Biochemistry: Khosla et aL Proc. Natt Acad. Sci. USA 78 (1981) product: TLC on silica gel,RF0.67 in EtOH; RF0.81 in CHC13/ RESULTS AcOH (95:5); RF0.73 in CHClJMeOH/AcOH (85:10:5). [a-MeTyr4]AII. COOH-Terminal tert-butyloxycarbonyl-L- 'H NMR. In addition to the conformational information that phenylalanine was attached to 2% crosslinked chloromethyl can be extracted from the 'H NMR spectrum, one can ascertain polymer as described, and chain elongation was performed on the correct chemical nature of the analog studied. Comparison a manual nitrogen-stirred apparatus (6) by using the protocol of the spectrum of[a-MeTyr4]AII with that of All shows (i) the described (7). In general, the coupling reaction was carried out appearance of a methyl group single peak at 1.31 ppm; (ii) the twice for 6-12 hr each time. 1-Hydroxybenzotriazole (2 mmol disappearance of the tyrosine a-proton octuplet (4.52 ppm in per mmol of tert-butyloxycarbonylamino acid) was used as an All), (iii) the transformation of the amide proton doublet of ty- additive during coupling with N,N-dicyclohexylcarbodiimide. rosine (8.07 ppm in All) into a single peak at 7.47 ppm; (iv) an Apart from facilitating the coupling reaction, this procedure also upfield shift of the aromaticmeta protons oftyrosine (0.26 ppm avoids racemization of the histidine residue during coupling with respect to All); and (v) that all the othera-protons exhibit with tert-butyloxycarbonyl-Nim-benzylhistidine (8). Coupling the same chemical shifts and patterns as in the native peptide with tert-butyloxycarbonylvaline was difficult and was repeated (Fig. 1). These data leave no doubt as to the correct chemical three times by using a 3-fold excess of this derivative and a 6- structure of the peptide [a-MeTyr4]AII. fold excess of 1-hydroxybenzotriazole each time. At the end of The majority ofthe proton resonances are identical in All and the synthesis, the peptide was removed from the polymer with la-MeTyr4]AII, but differences in the chemical shifts are de- HBr/CF3COOH and hydrogenated over 5% palladium/BaSO4 tected on the NH and sidechain protons of the tyrosine-neigh- in MeOH/AcOH/H20 (5:1:1). The crude product was purified bor residues -3 and isoleucine-5. It is reasonable to at- on a column of Bio-Rad anion exchange resin (AG-1 X 2, tribute these differences to the inductive effect ofthe a-methyl 200-400 mesh, acetate form) by eluting with ammonium acetate group and to a slight reorientation and decrease of rotational buffer (pH 8.5). Fractions containing the major component properties of the tyrosine side chain (ring-current effects). were pooled, lyophilized, and rechromatographed on succes- Two types of conformationally related vicinal coupling con- sive columns of Sephadex G-25 by using n-BuOH/pyridine/ stants were measured in [a-MeTyr4]AII and compared with H20 (65:35:35, upper phase) and n-BuOH/AcOH/H20 (4:1:5, those of All: the JNH-CHa value, which is tied to the dihedral upper phase) as the solvent systems. TLC on cellulose (E. angle 45 of the backbone, and the3J,,,l coupling, which yields Merck), RF0.73 in n-BuOH/AcOH/H20 (4:1:5); RF 0.55 in n- the side-chain orientation (rotamer distribution). Ofcourse, the BuOH/pyridine/H20 (10:2:5); RF 0.68 in n-BuOH/pyridine/ a-methylation of the tyrosine residue reduces the amount of H20 (65:35:65); RF 0.71 in n-PrOH/H20 (2:1); RF 0.76 in n- information obtainable, as neither the NH-CHa coupling nor BuOH/AcOH/H20/pyridine (30:6:24:20). lonophoresis was the tyrosine side-chain orientation are available in this analog. carried out on Schleicher & Schuell 2043A filter paper strips The coupling constants observed for the other residues, how- in a model R, series D, Beckman electrophoresis cell at 400 V ever, agree well with those of All. Although the JNH-CHG COU- in HCO2H/AcOH buffer (pH 1.95) prepared by diluting 60 ml pling constants of the valine-3 and isoleucine-5 residues are of HCO2H and 240 ml of AcOH to 2 liters with distilled H20 somewhat smaller than in All (i.e., [Asp', Ile5]AII), the varia- and Beckman barbiturate buffer B-2 (pH 8.6). Histidine was run tion is within the limits of the difference observed between simultaneously as a reference compound; E(His) indicates elec- [Asp',Val5]AII and All. As to the rotamer distribution of the trophoretic mobility relative to His = 1.00: E(His) 0.82 (pH principal side chains of All, a parameter that is sensitive to 1.95); E(His) 1.25 (pH 8.6). The peptide was hydrolyzed in a overall or local conformational changes (4 and 4i angles), no sig- sealed tube in 12 M HCl/propionic acid (1:1) at 155°C for 2 hr nificant difference exists between the two peptides (12). in the presence of 0.1 ml of 90% aqueous phenol. Amino acid 'H NMR spectra ofpeptides in dimethyl sulfoxide also permit analysis, performed on a model MM-100 Glenco amino acid the observation of temperature effects on the NH proton chem- analyzer, gave the ratio: Asp 1.0, Arg 0.99, Val 0.89, Tyr(a-Me) ical shifts that can be related to overall conformational stability 1.14, Ile 0.95, His 0.98, Pro 0.98, Phe 0.98. and internal hydrogen bonding. The slopes ofthe temperature- Pressor Activity. Determination of the pressor activity (9) dependence curves of the NH protons in [a-MeTyr4]AII are was carried out on vagotomized ganglion-blocked rats; [a- generally less pronounced than those in All, which suggests that MeTyr4]AII showed 92.62 ± 5.32% (n = 24, 8 rats) pressor the a-methylation of the pivotal tyrosine residue has slightly activity ofAll. increased the global rigidity in the peptide backbone. Enzymatic Degradation with a-Chymotrypsin. Incubation Circular Dichroism. CD has been shown to be a very sen- of the peptide solution (0.1 M ammonium acetate buffer, pH sitive method for monitoring even small conformational changes 8.5) with chymotrypsin (beef pancreas, triple crystallized) (3) in peptides and proteins (4, 10, 11). With respect to the All for 1 hr followed by TLC indicated the absence of degradation peptides, the CD spectrum in the peptide region of [a- products when the chromatogram was sprayed with ninhydrin. MeTyr4]AII in trifluoroethanol, and a pH titration curve of the Under the same conditions, angiotensin II was completely de- tyrosine 'Lb and at 275 nm in aqueous solution, have been graded into two components (presumably tetrapeptides). shown to respond readily to global or local structural changes, 'H NMR. The peptides All and [a-MeTyr4]AII were studied respectively. The CD spectrum of [a-MeTyr4]AII in TFE, re- under identical conditions-0.01 M in the cationic state in deu- corded at 250-180 nm consists essentially of two bands, a neg- terated dimethyl sulfoxide. Tetramethylsilane was used as in- ative one at 225 nm and a positive one at 195 nm, consistent with ternal reference. 'H NMR spectra were obtained on 250-MHz the antiparallel ,B-sheet conformation and practically identical and 400-MHz instruments. Signal assignment was carried out to that ofAll in the same solvent (13). The intensities ofthe two on the basis of double resonance experiments and by compar- bands are somewhat higher in the a-methylated analog, how- ison with previous work (10). ever, again suggesting greater stability in and around position Circular Dichroism (CD). CD spectra were recorded on a 4 of the backbone (Fig. 2). In aqueous solution, the CD spec- Dichrograph model III (Jobin Yvon) using fused quartz cells of trum of the aromatic region is also identical in shape to that of 1.0-cm and 0.01-cm path length. The results are expressed in All, albeit of slightly higher intensity at 275 nm. Just as in All, molar ellipticity, and pH titrations were carried out as described titration of the carboxyl group (the positive slope has been (11). shown to be due to the influence of the j3-carboxyl Downloaded by guest on September 29, 2021 Biochemistry: Khosla et al Proc. Nati Acad. Sci. USA 78 (1981) 759

All

lw -4 rI T

CaH His Tyr Phe I CaH2 Pro [a-MeTyr4]AII I

4.7 4.3 3.9 3.5

I I 1 i

Shift (relative to tetramethylsilane), ppm

FIG. 1. 1H NMR spectra at 400 MHz of All and [a-MeTyr4]AI.

group) from pH I to pH 4 increases the intensity of the tyrosine does not lead to any major change in the overall solution con- signal, and titration of the histidine imidazole group causes a formation of the peptide. Small localized effects that can be total inversion of the CD signal. The amplitude of the titra- detected by the sensitive spectroscopic methods CD and NMR tion-i. e., the difference of ellipticity, [0]275 between pH 4 and pH 8-is a function of the respective orientations of the ty- rosine and histidine side chains, in other words, of the time- 300 averaged distance between them (4, 10). The increased ampli- I I~~AI tude of titration in [a-MeTyr4]AII with respect to that of AII hints at a slightly reduced distance, perhaps due to a sterically 200 more hindered rotation of the tyrosine side chain (modified rotamer distribution). At higher pH values, the tyrosine chro- mophore titrates, going from the phenol to phenolate form; the 100 e CD spectrum at pH 12 consists of a strong positive band at 293 tot as nm, just in AIl. C.4Mu 4U 0 DISCUSSION The conformational properties of a-methylated amino acids, -100 l a:-MeTyr4]AIj <% 0v_ such as a-aminoisobutyric acid, have been investigated by the- oretical and experimental methods. Whereas early calculations suggested preference for helical-type conformations around this -200 residue (14), IR and NMR spectroscopy studies ofpeptides con- taining this amino acid in inert solvents showed that, in solution, -300 no such preference exists (15, 16). a-Aminoisobutyric acid has been found in the C7 conformation both in solution (15) and in 3 5 ~~~~~79 11 the crystal state (4 = 70, 4i = 64°) (17) and in the corner positions 2 and 3 of f-turns (15, 16): this residue and, by extrapolation, pH any a-methylated amino acid residue, displays nearly the same conformational properties as other usual residues. It is therefore FiG. 2. Ellipticity at 275 nm as a function of pH for All and [a- not surprising that a-methylation ofthe tyrosine residue in AII MeTyr4]AII. Downloaded by guest on September 29, 2021 760 Biochemistry: Khosla et aL Proc. Natl. Acad. Sci. USA 78 (1981)

do not extend far beyond the site of the chemical modification. 5. Khosla, M. C., Stachowiak, K., Khairallah, P. A. & Bumpus, F. This corroborates the results of Marshall et al. (18), who sug- M. (1979) in Peptides: Structure and Biological Functions, Pro- gested that a-methylation introduces only minor NMR ceedings of the Sixth American Peptide Symposium, Washington, DC, June 17-22, eds. Gross, E. & Meienhofer, J. (Pierce Chem- modifications. ical, Rockford, IL), pp. 467-470. Considering the ever stronger correlation between the con- 6. Khosla, M. C., Leese, R. A., Maloy, W. L., Ferreira, A. T., formation and the biological activity ofAII peptides, it is logical Smeby, R. R. & Bumpus, F. M. (1972)J. Med. Chem. 15, 792-795. to predict full biological activity for [a-MeTyr4]AII. As a-meth- 7. Khosla, M. C., Munoz-Ramirez, H., Hall, M. M., Khairallah, P. ylation of the tyrosine residue practically eliminates the sus- A. & Bumpus, F. M. (1977) J. Med. Chem. 20, 1051-1055. of the to 8. Koenig, W. & Geiger, R. (1970) Chem. Ber. 103, 788-798. ceptibility peptide chymotryptic hydrolysis, the syn- 9. Pickens, P. T., Bumpus, F. M., Lloyd, A. M., Smeby, R. R. & thesis of further analogs of this type (which conserve the Page, I. H. (1965) Circ. Res. 17, 438-448. essential conformation while being resistant to enzymatic at- 10. Fermandjian, S., Lintner, K., Haar, W., Fromageot, P., Khosla, tack) can be expected. In addition to a-methylation in position M. C., Smeby, R. R. & Bumpus, F. M. (1976) in Peptides, ed. 4 (or 5), suitable modifications at the NH2 and COOH termini Loffet, A. (Univ. Bruxelles, Bruxelles, Belgium), pp. 339-352. are necessary to stabilize these peptides against enzymatic at- 11. Lintner, K., Fermandjian, S., Regoli, D. & Barabe, J. (1977) Eur. tack by and J. Biochem. 61, 395-401. endopeptidases exopeptidases. 12. Fermandjian, S., Piriou, F., Lintner, K., Toma, F., Lam-Thanh, H., Fromageot, P., Khosla, M. C., Smeby, R. R. & Bumpus, F. We are grateful to Drs. R. Hirschmann and D. F. Veber (Merck Sharp M. (1979) in Peptides: Structure and Biological Functions, Pro- & Dohme) for a generous supply of L-a-methyltyrosine and to Messrs. ceedings of the Sixth American Peptide Symposium, Washington, E. Bachynsky, S. Forgac, and J. Blum and Miss C. Lakios for their ex- DC, June 17-22, eds. Gross, E. & Meienhofer, J. (Pierce Chem- cellent technical assistance. This work was supported in part by National ical, Rockford, IL), pp. 205-208. Institutes of Health Grant HL-6835. 13. Greff, D., Fermandjian, S., Fromageot, P., Khosla, M. C., Smeby, R. R. & Bumpus, F. M. (1976) Eur. J. Biochem. 61, 297-305. 1. Khosla, M. C., Page, I. H. & Bumpus, F. M. (1979) Biochem. 14. Burgess, A. & Leach, S. J. (1973) Biopolymers 12, 2599-2605. PharmacoL 28, 2867-2882. 15. Aubry, A., Protas, J., Boussard, G., Marraud, M. & Neel, J. (1978) 2. Khosla, M. C., Munoz-Ramirez, H., Hall, M. M., Smeby, R. R., Biopolymers 17, 1693-1711. Khairallah, P. A., Bumpus, F. M. & Peach, M. J. (1976) J. Med. 16. Rao, C. P., Nagaraj, R., Rao, C.N.R. & Balaram, P. (1980) Bio- Chem. 19, 244-250. chemistry 19, 425-431. 3. Stachowiak, K., Khosla, M. C., Plucinska, K., Khairallah, P. A. 17. Flippen, J. L. & Karle, I. L. (1976) Biopolymers 15, 1081-1092. & Bumpus, F. M. (1979)J. Med. Chem. 22, 1128-1130. 18. Marshall, G. R., Bosshard, H. E., Kendrick, N.C.E., Turk, J., 4. Piriou, F., Lintner, K. Fermandjian, S., Fromageot, P., Khosla, Balasubramanian, T. M., Cobb, S.M.H., Moore, M., Leduc, L. M. C., Smeby, R. R. & Bumpus, F. M. (1980) Proc. Natl. Acad. & Needleman, P. (1976) in Peptides, ed. Loffet, A. (Univ. Brux- Sci. USA 77, 82-86. elles, Bruxelles, Belgium), pp. 361-368. Downloaded by guest on September 29, 2021