- Email: [email protected]
Synthesis and biological properties of ovine corticotropin-releasing factor (CRF)
Life Sciences, Vol. 31, pp. 429-435 Printed in the U.S.A.
Pergamon Press
SYNTHESIS AND BIOLOGICAL PROPERTIES OF OVINE CORTICOTROPIN-RELEASING FACTOR (CRF) Javier Sueiras-Diaz, David H. Coy*, Sandor Vigh, Tommie W. Redding, Wei-Yong Huang, Ignacio Torres-Aleman and Andrew V. Schally Endocrine and Polypeptide Laboratory, Veterans Administration Medical Center, and Department of Medicine, Tulane University School of Medicine, New Orleans, Louisiana 70146. (Received in final form May 17, 1982) Summary The 41-resldue sequence of recently identified ovine corticotropin-releasing factor (CRF) was assembled on a benzhydrylamine resin support. Deprotection and cleavage from the resin were accomplished by HF treatment. The crude peptlde was purified by gel filtration and reverse-phase, medium pressure, followed by high-performance liquid chromatography (HPLC). In addition to the usual criteria, the homogeneity of the final material, obtained in 7% yield, was assessed by the isolation and examination of cyanogen bromide cleavage and tryptlc digestion fragments by HPLC and amino acid analysis. The synthetic 41 amino acid CRF stimulated the release of corticotropin (ACTH) in three in vitro systems: isolated rat pituitary quarters, monolayer cultures of dispersed pituitary cells, and superfu~ed pituitary cells on a column, the responses being related to th~ log-dose of CRF in the range of 0.05-125 ng/ml. The synthetic peptide also augmented in vivo release of ACTH in rats pretreated with chlorpromazine, morphine, and Nembutal, as assessed by the measurement of serum corticosterone. The data indicates chemical purity and high biological activity of synthetic material. Corticotropin-releasing factor (CRF) was the first hypothalamic hypophysiotropic factor to be demonstrated (1,2). However, the chemical identification of a peptide which would meet many of the criteria of a physiological CRF was delayed for many years. Recently Vale et al. (3) isolated and structurally identified a 41-resldue peptide from ovine hypothalamus which stimulates the secretion of corticotropin and B-endorphin in vitro and in vlvo (3,4). The successful solid-phase synthesis of several peptides of comparable size (5,6) and the potential of high performance liquid chromatography for purification and for demonstrating homogeneity encouraged us to undertake the synthesis of this peptide, having in mind that our studies on the biological effects of CRF would be helped by the availability of pure peptide. Here we report the results of a rapid solid-phase synthesis of ovlne CRF in good yield and the biological evaluation of the synthetic material.
To whom reprint requests should be addressed.
0024-3205/82/050429-07503.00/0
430
Properties of Ovine CRF
Vol.
31, No. 5, 1982
Materials and Methods Peptide Synthesis. All amino acids were coupled as their Na-t-Boc derivatives (Bachem Inc., Torrance, CA). Reactive side chains were protected as follows: Ser and Thr, benzyl; Lys, 2-chlorobenzyloxycarbonyl; His and Arg, tosyl; Asp and Glu, 4-chlorobenzyl. The amino acid derivatives were coupled to a 1% crosslinked benzhdrylamine resin (Bachem) (2.1 g; 0.49 mmol amino groups/g) in the presence of diisopropylcarbodiimide using a Beckman 990 automatic synthesizer. Boc-Asn and Boc-Gln were coupled in the presence of an equimolar amount of lhydroxybenzotriazole. Boc protection was removed at each stage by two treatments with 33% TFA in CH2CI 2 for 1 minute and 25 minutes. Coupling reactions were monitored at each step by the ninhydrin test (7) and repeated, if incomplete after 60 minutes, using the preformed symmetric anhydride procedure (8) in DMF. If free amino groups persisted they were acetylated with acetylimidazole (5% in CH2C12, 50 minutes). The completed, protected peptide-resin, with its N-terminal Boc group removed to avoid alkylation of Met during HF cleavage steps (9), was deprotected and liberated from the resin support (I raM) by treatment with 60 ml anhydrous HF containing 10% anisole for 1 hr at 0°C. After removal of the HF under nitrogen, the peptide was precipitated with ether, the mixture filtered, and the peptide extracted with 50% acetic acid. Purification and Homogeneity. After reduction in volume, the solution was applied directly onto a column (2.5 x 95 cm) of Sephadex G-50 and eluted with 2 M AcOH. Fractions were collected and aliquots were examined by thin layer chromatography on silica gel plates in the solvent system 1-butanol-pyridine-acetic acid-water (15:10:3:12). The peptide was visualized with ninhydrln and pooled and lyophilized to a constant weight of 3.92 g. This material was examined by high performance liquid chromatography (HPLC) on a column (0.4 x 25 cm) of Synchropack RP-18 (lO ~m, 300 A ° pore size) using a gradient of 25% to 35% isopropanol in 0.1% TFA over 30 minutes. Absorption at 215 nm revealed a major peak (Fig. l) containing contaminants at its leading and trailing edges.
FIG. i HPLC of i00 ug of crude CRF after Sephadex chromatography on C18 Synchropack (i0 ~, 300 A°)(0.4 x 25 cm) using a gradient of 25% to 35% isopropanol in 0.1% TFA over 30 min. Flow rate 1 ml/mln., absorption at 215 nm.
\j 0
I\\ 30 MINUTES
Vol. 31, No. 5, 1982
Properties of 0vine CRF
431
Part of this material (340 mg) was further purified by reversed-phase medium pressure liquid chromatography (RP-MPLC) procedure using a column (2.5 x 45 cm) of Whatman LRP-I (Cl8-bonded silica gel, 13-24 ~m), eluted with a gradient from 15% to 60% acetonitrile in 0.1% TFA (Fig. 2). Aliquots from the main peak were removed and checked by analytical HPLC using the conditions described. Fractions 161-169 were pooled and lyophillzed to give 75 mg of almost pure peptide, 18.5% yield based on resin incorporation. 03
¸
T ~ E h l ~ 6 ER
FIG. 2 Preparative reversed-phase medium pressure liquid chromatography of 340 mg of crude CRF on a column (2.5 x 45 cm) of Whatman LRP-I (C18 silica, 13-26 ~m) using a gradient of 15% to 60% CH3CN in 0.1% TFA. Four ml fractions were collected and fractions 161-169 were pooled. A final purification was achieved by a preparative RP-HPLC procedure using a C18, i0 ~m, 300A = Synchropack column (i x 25 cm), loaded with 11.7 mg and eluted with a gradient of 23% to 36% isopropanol in 0.1% TFA developed over i hr. Fractions were collected manually and aliquots were checked by HPLC. Those fractions Judged pure, favoring purity rather than quantity, were then pooled, concentrated in vacuo and lyophillzed to a constant weight of 4.55 mg, 7% yield based on mM of amino groups substituted on to the starting resin. Amino acid analysis of a sample hydrollzed in 6M HCI at Ii0 ° C for 18 hrs in a sealed evacuated tube gave: Asp, 4.1 (4); Thr, 1.79 (2); Ser, 2.30 (3); Glu, 7.26 (7); Pro, 1.9 (2); Ala, 4.05 (4); Val, 0.95 (I); Met, 0.94 (i); lie, 1.87 (2); Leu, 8.28 (8); Phe, 1.00 (I); His, 1.95 (2); Lys, 2.00 (2); Arg, 1.93 (2). HPLC of the peptlde under the conditions described in the corresponding legend, revealed a high degree of purity (Fig. 3). The same homogeneity was observed when HPLC was run using different columns (Partlsil ODS-2 and ODS-3, Cl8-reversed phase) and conditions. N-Terminus Determination. dimensional chromatography systems described by Woods under an ultraviolet lamp. allow detection of greater
The peptide was dansylated (i0) and subjected to two on polyamide thin-layer sheets in the two solvent and Wang (ii). Dansyl amino acids were visualized Only dansyl serine was visible at loads which would than 1% of contaminating dansyl amino acid.
Tryptlc Digestion. Synthetic peptide (50 ~) was digested with TPCK-Trypsln (i ~g) in buffer (0.2N, N-ethylmorpholine acetate, pH 8.0, i00 ~i) for 16 hrs at 37 ° C. The reaction was terminated by lyophilizatlon. The tryptic fragments were isolated by RP-HPLC using an analytical C18 column (for detailed conditions, see the legend of Fig. 4) and were subjected to amino acid analysis. Results were as follows: CRF (17-23) analyzed as Glu, 2.10;
432
Properties of Ovine CRF
Vol. 31, No. 5, 1982
FIG. 3 HPLC of 80 ug of purified CRF from preparative RP-HPLC using the conditions described in Fig. I.
0
1'5 70 MINUTES
Val, 1.00; Leu, 0.98; Met, 1.00; Thr, 1.00; Lys, 0.96; CRF (1-16) analyzed as Ser, 1.90; Glu, 2.08; Pro, 1.76; lle, 0.97; Leu, 3.67; Asp, 1.03; Thr, 0.97; Phe, 0.97; His, 1.03; Arg, 1.00; CRF (24-35) yielded Ala, 2.51; Asp, 1.97; Glu, 3.50; Leu, 1.03; His, 0.99; Ser, 0.95 ; Arg, 0.88; CRF (37-41) gave Leu, 2.00; Asp, 1.05; lie, 0.85; Ala, 1.06. CRF (17-23) gave the theoretical value for Met after enzymatic hydrolysis indicating that no sulfoxide was present. CNBr Cleavase of Synthetic CRF. CRF (400 ug) was dissolved in i00 ul of 70% TFA and CNBr (400 pg) was added and the mixture stirred at room temperature for 16 hrs. It was then diluted with water and lyophilized. Reversed-phase HPLC of the mixture using a Partisil ODS-3 column and isopropanol gradient from 5% to 50% in 0.1% TRA separated the fragments into two components as expected. The peaks were individually collected, lyophilized and hydrolyzed for amino acid analysis. Results were as follows: CRF (22-41) analyzed as Thr, 0.8; Lys, 2.2; Ala, 4.0; Asp, 2.8; Glu, 3.0; Leu, 3.1; His, 1.0; Ser, 0.8; Arg, 1.2; lie, 1.0. Homoserine-CRF (1-21) analyzed as Ser, 1.8; Glu, 4.0; Pro, 2.0; lie, 0.9; Leu, 5.3; Asp, i.I; Thr, 1.0; Phe, i.i; His, 0.8; Arg, 1.2; Val, 1.2; Hser, 1.0. Measurement of Biological Activity. CRF Assays. The stimulation of the release of ACTH in vitro was measured in three different systems: isolated rat pituitary quarters (i, 12), monolayer cultures of rat pituitary cells (13), and superfusion of rat anterior pituitary cells on a column (14). The ACTH released from pituitary tissue into the medium was measured by a specific radioimmunoassay (RIA) for ACTH. The details of the first two methods and the RIA for ACTH have been reported previously (12). The CRF activity was also tested in vivo in rats pretreated with chlorpromazine, morphine and Nembutal (15), the increase of serum corticosterone being used as the index of ACTH release. A rabbit antiserum against 3-carboxymethyloxime-cortisol conjugated with bovine serum albumin crossreacted 30% with corticosterone and was used in the radioimmunoassay for corticosterone in rat serum (16) (Cambridge Nuclear Corp.). 1251-Labeled cortisol was used as the tracer. The significance of differences between groups was calculated by Duncan's multiple range test (17).
Vol. 31, No. 5, 1982
Properties of Ovine CRF
433
Results and Discussion The solld-phase synthesis of CRF proceeded smoothly with few difficulties. The acylation was usually found to be complete after one or two couplings. The exceptions were Glu 16, which was coupled three times and Va124, which had to be coupled four times followed by an acetylation step. After HF cleavage and passage down Sephadex, the products already contained a major component (monitored by RP-HPLC), showing a fairly good stability to HF treatment. Further purification of this peptide was accompllshed by MPLC and, for the final step, by a preparative RP-HPLC methodology employing a 300 A ° - reverse-phase column, which was chosen because it gave far better resolution than regular 60 A ° reversed-phase packings. The homogeneity and the primary structure of synthetic CRF were investigated by HPLC, amino acid analysis and peptlde mapping. Short individual peptlde fragments split from a large parent polypeptide can provide more reliable amino acid data than the parent molecule, and contaminants resulting from deletions or small modifications to the parent chain will exhibit greater physical differences. The final product was, therefore, digested with CNBr to cleave CRF at the methionyl residue in position 21. The two expected fragments were separated on HPLC and individually collected and characterized by amino acid analysis and direct comparison with the synthetic twenty peptide fragment (22-41) of CRF previously prepared by us. Four fragments were isolated and characterized by amino acid analysis from the tryptic digest of CRF. Their amino acid ratios were consistent with the expected compositions of the tryptic peptides from CRF. The minor side fractions (Fig. 4) were also isolated and analyzed. Their amino acid compositions were consistent with chymotryptic fragments of CRF, probably produced by residual chymotryptlc activity of the TPCK-trypsin preparation used. T1
FIG. 4 T4
T3
%J MINUTES
Analytical RP-HPLC of tryptic fragments of synthetic CRF. TPCK-trypsin digested CRF (50 ~g) dissolved in 0.1% TFA (i00 pl) was applied on Li-Chrosorb RP-18 (I0 ~) (0.4 x 25 cm) and eluted wlth a gradient of from 10% to 50% isopropanol in 0.1% TFA over 30 minutes. Flow-rate was 1.5 ml/min., absorption at 215 nm.
434
Properties of Ovine CRF
Vol. 31, No. 5, 1982
The purified synthetic 41-residue CRF, was tested for its ability to stimulate ACTH release in vitro and in vivo. In two different experiments the synthetic material caused a highly significant stimulation of ACTH release from isolated rat pituitary quarters when added to the incubation medium in doses 1-125 ng/ml (Table i). Doses of 0.05, 0.2 and 0.25 ng/ml did not have a significant effect. The pituitary ACTH response to synthetic CRF in the pituitary quarter system was related to the logarithm of the dose fn the range 1-125 ng/ml. This relationship was linear.
TABLE I Effect of Synthetic CRF on ACTB Release from Rat Pituitary Quarters In Vitro
Material
Control
Dose, ng/ml
ACTH in medium, % of control *
Duncan's P
-
27.2 ± 1.6
CRF
0.2
39.0 ± 5.7
N.S.
CRF
1.0
65.7 ± 3.1
<0.01
CRF
5
94.8 ± 9.5
<0.01
CRF
25
133.4 ± 5.3
<0.01
CRF
125
204.5 ± 15.7
<0.01
*(2nd hr/ist hr) x i00.
-
Shown as mean ± SEM.
Synthetic CRF also stimulated the rate of ACTH secretion from rat pituitary cells on a superfusion column. The minimal effective dose of CRF in this system was 0.05 ng/ml. We found an approximately linear relationship between the logarithm of the dose of CRF in the range of 0.05 ng to 5 ng/ml and the total amount of ACTH released. The 5 ng dose induced a maximal response, that is a 5-fold stimulation of ACTH release as compared to 0.05 ng/ml. The release of ACTH from isolated rat pituitaries or from pituitary cell superfuslon system in response to synthetic CRF has not been reported previously. Our data thus extends the biological evaluation of synthetic CRF previously described by Rivier et al. (4). Synthetic CRF was also tested in monolayer cultures of dispersed pituitary cells in doses of 0.05-204.8 ng/ml. In this system doses of 0.05 ng CRF/ml augmented ACTH secretion, but the stimulation (31.30 ± 2.2 ng ACTH/ml over a 4 hr. period) was not statistically significant compared to controls (ACTH = 22.2 ± 2.9 ng/ml) Doses of 0.2 ng/ml of CRF significantly stimulated ACTH release (33.8 ± 2.2 ng/ ml; p <0.05) and maximal responses were reached at a concentration of 12.8 ng CRF/ml, ACTH rising to 62.0 ± 6.2 ng/ml ( p <0.01). Higher doses of CRF did not result in an additional increase in the ACTH release, the ACTH response to CRF presenting a typical slgmoid dose-response curve. Synthetic CRF was also assayed in vivo for its ability to release ACTH.
The
Vol. 31, No. 5, 1982
Properties of Ovine CRF
435
ACTH stimulation was assessed by the measurement of serum corticosterone by RIA. Twenty min. after intravenous administration of CRF in a dose of 25 ng/ i00 g body weight serum corticosterone levels rose from 82.3 ± 19 ng/ml to 399.3 ± 48 ng/ml (p <0.01). Doses of 5 ng/100 g body weight did not result in a significant elevation of serum corticosterone. Turkelson et al. (18) have recently reported biological data of a CRF synthesized by Peninsula Lab. In conclusion, a rapid purification scheme involving gel permeation followed by two high resolution steps (MPLC and HPLC) has been applied to the isolation of CRF from a complex synthetic mixture. Biological tests carried out in vitro and in vivo indicated that the synthetic CRF possessed potent ACTH releasing activity. Acknowledgments We thank NIAMDD-rat-pltultary Hormone Program for the gifts of materials used in radioimmunoassay. This work was supported by Natlonal Institutes of Health Grants AM18370 (to D.H.C.), AM07467 (to A.V.S.) and by the Veterans Administration. References i. 2. 3. 4. 5. 6. 7. 8. 9. I0. Ii. 12.
13. 14. 15. 16. 17. 18.
N. SAFFRAN and A.V. SCHALLY, Can. J. Biochem. Physiol. 33 408-415 (1955). R. GUILLEMIN and B. ROSENBERG, Endocrinology 57 599-607 (1955). W. VALE, J. SPIES, C. RIVIER and J. RIVIER, Science 213 1394-1397 (1981). C. RIVIER, M. BROWNSTEIN, J. SPIES, J. RIVIER, and W. VALE, Endocrinology II0 272-278 (1982). D.H. COY and J. GARDNER, Int. J. Peptlde Protein Res. 15 73-78 (1980). C.A. MEYERS and D.H. COY, Int. J. Peptlde Protein Res. 16 248-253 (1980). E. KAISER, R. COLESCOTT, C.D. BOSSINGER and P.I. COOK, Anal. Biochem. 34 595-598 (1970). B. HEMMASI and F. BAYER, Int. J. Peptlde Protein Res. 9 63-70 (1977). R.E. NOBLE, D. YAMASHIRO and C.H. LI, J. Am. Chem. Soc. 98 2324-2328 (1926). W.R. GRAY, Methods in Enzymology 25 121-138 (1972). K.R. WOODS and K.T. WANG, Biochem. Biophys. Acta 133 369-370 (1967). A.V. SCHALLY, W.Y. HUANG, T.W. REDDING, A. ARIMURA, D.H. COY, K. CHIHARA, R.C.C. CHANG, V. RAYMOND and F. LABRIE, Biochem. Biophys. Res. Commun. 82 582-588 (1978). W. VALE, G. GRANT, M. AMOSS, R. BLACKWELL and R. GUILLEMIN, Endocrinology 91 562-572 (1972). G.H. MULDER and P.G. SMELIK, Endocrinology I00 1143-1152 (1977). A. ARIMURA, L. SAITO and A.V. SCHALLY, Endocrinology 81 235-245 (1967). C. GOMEZ-SANCHEZ, B.A. MURRY, D.C. KERN and N.M. KAPLAN, Endocrinology 96 796-798 (1975). R.G.D. STEEL and J. TORRIE, Prlnciples and Procedures of Statistics, p. 2, McGraw-Hill, New York (1960). C.M. TURKELSON, A. ARIMURA, M.D. CULLER, J.B. FISHBACK, K. GROOT, M. KANDA, M. LUCIANO, C.R. THOMAS, D. CHANG, J.K. CHANG and M. SHIMIZU, Peptldes 425-429 (1981).
Pergamon Press
SYNTHESIS AND BIOLOGICAL PROPERTIES OF OVINE CORTICOTROPIN-RELEASING FACTOR (CRF) Javier Sueiras-Diaz, David H. Coy*, Sandor Vigh, Tommie W. Redding, Wei-Yong Huang, Ignacio Torres-Aleman and Andrew V. Schally Endocrine and Polypeptide Laboratory, Veterans Administration Medical Center, and Department of Medicine, Tulane University School of Medicine, New Orleans, Louisiana 70146. (Received in final form May 17, 1982) Summary The 41-resldue sequence of recently identified ovine corticotropin-releasing factor (CRF) was assembled on a benzhydrylamine resin support. Deprotection and cleavage from the resin were accomplished by HF treatment. The crude peptlde was purified by gel filtration and reverse-phase, medium pressure, followed by high-performance liquid chromatography (HPLC). In addition to the usual criteria, the homogeneity of the final material, obtained in 7% yield, was assessed by the isolation and examination of cyanogen bromide cleavage and tryptlc digestion fragments by HPLC and amino acid analysis. The synthetic 41 amino acid CRF stimulated the release of corticotropin (ACTH) in three in vitro systems: isolated rat pituitary quarters, monolayer cultures of dispersed pituitary cells, and superfu~ed pituitary cells on a column, the responses being related to th~ log-dose of CRF in the range of 0.05-125 ng/ml. The synthetic peptide also augmented in vivo release of ACTH in rats pretreated with chlorpromazine, morphine, and Nembutal, as assessed by the measurement of serum corticosterone. The data indicates chemical purity and high biological activity of synthetic material. Corticotropin-releasing factor (CRF) was the first hypothalamic hypophysiotropic factor to be demonstrated (1,2). However, the chemical identification of a peptide which would meet many of the criteria of a physiological CRF was delayed for many years. Recently Vale et al. (3) isolated and structurally identified a 41-resldue peptide from ovine hypothalamus which stimulates the secretion of corticotropin and B-endorphin in vitro and in vlvo (3,4). The successful solid-phase synthesis of several peptides of comparable size (5,6) and the potential of high performance liquid chromatography for purification and for demonstrating homogeneity encouraged us to undertake the synthesis of this peptide, having in mind that our studies on the biological effects of CRF would be helped by the availability of pure peptide. Here we report the results of a rapid solid-phase synthesis of ovlne CRF in good yield and the biological evaluation of the synthetic material.
To whom reprint requests should be addressed.
0024-3205/82/050429-07503.00/0
430
Properties of Ovine CRF
Vol.
31, No. 5, 1982
Materials and Methods Peptide Synthesis. All amino acids were coupled as their Na-t-Boc derivatives (Bachem Inc., Torrance, CA). Reactive side chains were protected as follows: Ser and Thr, benzyl; Lys, 2-chlorobenzyloxycarbonyl; His and Arg, tosyl; Asp and Glu, 4-chlorobenzyl. The amino acid derivatives were coupled to a 1% crosslinked benzhdrylamine resin (Bachem) (2.1 g; 0.49 mmol amino groups/g) in the presence of diisopropylcarbodiimide using a Beckman 990 automatic synthesizer. Boc-Asn and Boc-Gln were coupled in the presence of an equimolar amount of lhydroxybenzotriazole. Boc protection was removed at each stage by two treatments with 33% TFA in CH2CI 2 for 1 minute and 25 minutes. Coupling reactions were monitored at each step by the ninhydrin test (7) and repeated, if incomplete after 60 minutes, using the preformed symmetric anhydride procedure (8) in DMF. If free amino groups persisted they were acetylated with acetylimidazole (5% in CH2C12, 50 minutes). The completed, protected peptide-resin, with its N-terminal Boc group removed to avoid alkylation of Met during HF cleavage steps (9), was deprotected and liberated from the resin support (I raM) by treatment with 60 ml anhydrous HF containing 10% anisole for 1 hr at 0°C. After removal of the HF under nitrogen, the peptide was precipitated with ether, the mixture filtered, and the peptide extracted with 50% acetic acid. Purification and Homogeneity. After reduction in volume, the solution was applied directly onto a column (2.5 x 95 cm) of Sephadex G-50 and eluted with 2 M AcOH. Fractions were collected and aliquots were examined by thin layer chromatography on silica gel plates in the solvent system 1-butanol-pyridine-acetic acid-water (15:10:3:12). The peptide was visualized with ninhydrln and pooled and lyophilized to a constant weight of 3.92 g. This material was examined by high performance liquid chromatography (HPLC) on a column (0.4 x 25 cm) of Synchropack RP-18 (lO ~m, 300 A ° pore size) using a gradient of 25% to 35% isopropanol in 0.1% TFA over 30 minutes. Absorption at 215 nm revealed a major peak (Fig. l) containing contaminants at its leading and trailing edges.
FIG. i HPLC of i00 ug of crude CRF after Sephadex chromatography on C18 Synchropack (i0 ~, 300 A°)(0.4 x 25 cm) using a gradient of 25% to 35% isopropanol in 0.1% TFA over 30 min. Flow rate 1 ml/mln., absorption at 215 nm.
\j 0
I\\ 30 MINUTES
Vol. 31, No. 5, 1982
Properties of 0vine CRF
431
Part of this material (340 mg) was further purified by reversed-phase medium pressure liquid chromatography (RP-MPLC) procedure using a column (2.5 x 45 cm) of Whatman LRP-I (Cl8-bonded silica gel, 13-24 ~m), eluted with a gradient from 15% to 60% acetonitrile in 0.1% TFA (Fig. 2). Aliquots from the main peak were removed and checked by analytical HPLC using the conditions described. Fractions 161-169 were pooled and lyophillzed to give 75 mg of almost pure peptide, 18.5% yield based on resin incorporation. 03
¸
T ~ E h l ~ 6 ER
FIG. 2 Preparative reversed-phase medium pressure liquid chromatography of 340 mg of crude CRF on a column (2.5 x 45 cm) of Whatman LRP-I (C18 silica, 13-26 ~m) using a gradient of 15% to 60% CH3CN in 0.1% TFA. Four ml fractions were collected and fractions 161-169 were pooled. A final purification was achieved by a preparative RP-HPLC procedure using a C18, i0 ~m, 300A = Synchropack column (i x 25 cm), loaded with 11.7 mg and eluted with a gradient of 23% to 36% isopropanol in 0.1% TFA developed over i hr. Fractions were collected manually and aliquots were checked by HPLC. Those fractions Judged pure, favoring purity rather than quantity, were then pooled, concentrated in vacuo and lyophillzed to a constant weight of 4.55 mg, 7% yield based on mM of amino groups substituted on to the starting resin. Amino acid analysis of a sample hydrollzed in 6M HCI at Ii0 ° C for 18 hrs in a sealed evacuated tube gave: Asp, 4.1 (4); Thr, 1.79 (2); Ser, 2.30 (3); Glu, 7.26 (7); Pro, 1.9 (2); Ala, 4.05 (4); Val, 0.95 (I); Met, 0.94 (i); lie, 1.87 (2); Leu, 8.28 (8); Phe, 1.00 (I); His, 1.95 (2); Lys, 2.00 (2); Arg, 1.93 (2). HPLC of the peptlde under the conditions described in the corresponding legend, revealed a high degree of purity (Fig. 3). The same homogeneity was observed when HPLC was run using different columns (Partlsil ODS-2 and ODS-3, Cl8-reversed phase) and conditions. N-Terminus Determination. dimensional chromatography systems described by Woods under an ultraviolet lamp. allow detection of greater
The peptide was dansylated (i0) and subjected to two on polyamide thin-layer sheets in the two solvent and Wang (ii). Dansyl amino acids were visualized Only dansyl serine was visible at loads which would than 1% of contaminating dansyl amino acid.
Tryptlc Digestion. Synthetic peptide (50 ~) was digested with TPCK-Trypsln (i ~g) in buffer (0.2N, N-ethylmorpholine acetate, pH 8.0, i00 ~i) for 16 hrs at 37 ° C. The reaction was terminated by lyophilizatlon. The tryptic fragments were isolated by RP-HPLC using an analytical C18 column (for detailed conditions, see the legend of Fig. 4) and were subjected to amino acid analysis. Results were as follows: CRF (17-23) analyzed as Glu, 2.10;
432
Properties of Ovine CRF
Vol. 31, No. 5, 1982
FIG. 3 HPLC of 80 ug of purified CRF from preparative RP-HPLC using the conditions described in Fig. I.
0
1'5 70 MINUTES
Val, 1.00; Leu, 0.98; Met, 1.00; Thr, 1.00; Lys, 0.96; CRF (1-16) analyzed as Ser, 1.90; Glu, 2.08; Pro, 1.76; lle, 0.97; Leu, 3.67; Asp, 1.03; Thr, 0.97; Phe, 0.97; His, 1.03; Arg, 1.00; CRF (24-35) yielded Ala, 2.51; Asp, 1.97; Glu, 3.50; Leu, 1.03; His, 0.99; Ser, 0.95 ; Arg, 0.88; CRF (37-41) gave Leu, 2.00; Asp, 1.05; lie, 0.85; Ala, 1.06. CRF (17-23) gave the theoretical value for Met after enzymatic hydrolysis indicating that no sulfoxide was present. CNBr Cleavase of Synthetic CRF. CRF (400 ug) was dissolved in i00 ul of 70% TFA and CNBr (400 pg) was added and the mixture stirred at room temperature for 16 hrs. It was then diluted with water and lyophilized. Reversed-phase HPLC of the mixture using a Partisil ODS-3 column and isopropanol gradient from 5% to 50% in 0.1% TRA separated the fragments into two components as expected. The peaks were individually collected, lyophilized and hydrolyzed for amino acid analysis. Results were as follows: CRF (22-41) analyzed as Thr, 0.8; Lys, 2.2; Ala, 4.0; Asp, 2.8; Glu, 3.0; Leu, 3.1; His, 1.0; Ser, 0.8; Arg, 1.2; lie, 1.0. Homoserine-CRF (1-21) analyzed as Ser, 1.8; Glu, 4.0; Pro, 2.0; lie, 0.9; Leu, 5.3; Asp, i.I; Thr, 1.0; Phe, i.i; His, 0.8; Arg, 1.2; Val, 1.2; Hser, 1.0. Measurement of Biological Activity. CRF Assays. The stimulation of the release of ACTH in vitro was measured in three different systems: isolated rat pituitary quarters (i, 12), monolayer cultures of rat pituitary cells (13), and superfusion of rat anterior pituitary cells on a column (14). The ACTH released from pituitary tissue into the medium was measured by a specific radioimmunoassay (RIA) for ACTH. The details of the first two methods and the RIA for ACTH have been reported previously (12). The CRF activity was also tested in vivo in rats pretreated with chlorpromazine, morphine and Nembutal (15), the increase of serum corticosterone being used as the index of ACTH release. A rabbit antiserum against 3-carboxymethyloxime-cortisol conjugated with bovine serum albumin crossreacted 30% with corticosterone and was used in the radioimmunoassay for corticosterone in rat serum (16) (Cambridge Nuclear Corp.). 1251-Labeled cortisol was used as the tracer. The significance of differences between groups was calculated by Duncan's multiple range test (17).
Vol. 31, No. 5, 1982
Properties of Ovine CRF
433
Results and Discussion The solld-phase synthesis of CRF proceeded smoothly with few difficulties. The acylation was usually found to be complete after one or two couplings. The exceptions were Glu 16, which was coupled three times and Va124, which had to be coupled four times followed by an acetylation step. After HF cleavage and passage down Sephadex, the products already contained a major component (monitored by RP-HPLC), showing a fairly good stability to HF treatment. Further purification of this peptide was accompllshed by MPLC and, for the final step, by a preparative RP-HPLC methodology employing a 300 A ° - reverse-phase column, which was chosen because it gave far better resolution than regular 60 A ° reversed-phase packings. The homogeneity and the primary structure of synthetic CRF were investigated by HPLC, amino acid analysis and peptlde mapping. Short individual peptlde fragments split from a large parent polypeptide can provide more reliable amino acid data than the parent molecule, and contaminants resulting from deletions or small modifications to the parent chain will exhibit greater physical differences. The final product was, therefore, digested with CNBr to cleave CRF at the methionyl residue in position 21. The two expected fragments were separated on HPLC and individually collected and characterized by amino acid analysis and direct comparison with the synthetic twenty peptide fragment (22-41) of CRF previously prepared by us. Four fragments were isolated and characterized by amino acid analysis from the tryptic digest of CRF. Their amino acid ratios were consistent with the expected compositions of the tryptic peptides from CRF. The minor side fractions (Fig. 4) were also isolated and analyzed. Their amino acid compositions were consistent with chymotryptic fragments of CRF, probably produced by residual chymotryptlc activity of the TPCK-trypsin preparation used. T1
FIG. 4 T4
T3
%J MINUTES
Analytical RP-HPLC of tryptic fragments of synthetic CRF. TPCK-trypsin digested CRF (50 ~g) dissolved in 0.1% TFA (i00 pl) was applied on Li-Chrosorb RP-18 (I0 ~) (0.4 x 25 cm) and eluted wlth a gradient of from 10% to 50% isopropanol in 0.1% TFA over 30 minutes. Flow-rate was 1.5 ml/min., absorption at 215 nm.
434
Properties of Ovine CRF
Vol. 31, No. 5, 1982
The purified synthetic 41-residue CRF, was tested for its ability to stimulate ACTH release in vitro and in vivo. In two different experiments the synthetic material caused a highly significant stimulation of ACTH release from isolated rat pituitary quarters when added to the incubation medium in doses 1-125 ng/ml (Table i). Doses of 0.05, 0.2 and 0.25 ng/ml did not have a significant effect. The pituitary ACTH response to synthetic CRF in the pituitary quarter system was related to the logarithm of the dose fn the range 1-125 ng/ml. This relationship was linear.
TABLE I Effect of Synthetic CRF on ACTB Release from Rat Pituitary Quarters In Vitro
Material
Control
Dose, ng/ml
ACTH in medium, % of control *
Duncan's P
-
27.2 ± 1.6
CRF
0.2
39.0 ± 5.7
N.S.
CRF
1.0
65.7 ± 3.1
<0.01
CRF
5
94.8 ± 9.5
<0.01
CRF
25
133.4 ± 5.3
<0.01
CRF
125
204.5 ± 15.7
<0.01
*(2nd hr/ist hr) x i00.
-
Shown as mean ± SEM.
Synthetic CRF also stimulated the rate of ACTH secretion from rat pituitary cells on a superfusion column. The minimal effective dose of CRF in this system was 0.05 ng/ml. We found an approximately linear relationship between the logarithm of the dose of CRF in the range of 0.05 ng to 5 ng/ml and the total amount of ACTH released. The 5 ng dose induced a maximal response, that is a 5-fold stimulation of ACTH release as compared to 0.05 ng/ml. The release of ACTH from isolated rat pituitaries or from pituitary cell superfuslon system in response to synthetic CRF has not been reported previously. Our data thus extends the biological evaluation of synthetic CRF previously described by Rivier et al. (4). Synthetic CRF was also tested in monolayer cultures of dispersed pituitary cells in doses of 0.05-204.8 ng/ml. In this system doses of 0.05 ng CRF/ml augmented ACTH secretion, but the stimulation (31.30 ± 2.2 ng ACTH/ml over a 4 hr. period) was not statistically significant compared to controls (ACTH = 22.2 ± 2.9 ng/ml) Doses of 0.2 ng/ml of CRF significantly stimulated ACTH release (33.8 ± 2.2 ng/ ml; p <0.05) and maximal responses were reached at a concentration of 12.8 ng CRF/ml, ACTH rising to 62.0 ± 6.2 ng/ml ( p <0.01). Higher doses of CRF did not result in an additional increase in the ACTH release, the ACTH response to CRF presenting a typical slgmoid dose-response curve. Synthetic CRF was also assayed in vivo for its ability to release ACTH.
The
Vol. 31, No. 5, 1982
Properties of Ovine CRF
435
ACTH stimulation was assessed by the measurement of serum corticosterone by RIA. Twenty min. after intravenous administration of CRF in a dose of 25 ng/ i00 g body weight serum corticosterone levels rose from 82.3 ± 19 ng/ml to 399.3 ± 48 ng/ml (p <0.01). Doses of 5 ng/100 g body weight did not result in a significant elevation of serum corticosterone. Turkelson et al. (18) have recently reported biological data of a CRF synthesized by Peninsula Lab. In conclusion, a rapid purification scheme involving gel permeation followed by two high resolution steps (MPLC and HPLC) has been applied to the isolation of CRF from a complex synthetic mixture. Biological tests carried out in vitro and in vivo indicated that the synthetic CRF possessed potent ACTH releasing activity. Acknowledgments We thank NIAMDD-rat-pltultary Hormone Program for the gifts of materials used in radioimmunoassay. This work was supported by Natlonal Institutes of Health Grants AM18370 (to D.H.C.), AM07467 (to A.V.S.) and by the Veterans Administration. References i. 2. 3. 4. 5. 6. 7. 8. 9. I0. Ii. 12.
13. 14. 15. 16. 17. 18.
N. SAFFRAN and A.V. SCHALLY, Can. J. Biochem. Physiol. 33 408-415 (1955). R. GUILLEMIN and B. ROSENBERG, Endocrinology 57 599-607 (1955). W. VALE, J. SPIES, C. RIVIER and J. RIVIER, Science 213 1394-1397 (1981). C. RIVIER, M. BROWNSTEIN, J. SPIES, J. RIVIER, and W. VALE, Endocrinology II0 272-278 (1982). D.H. COY and J. GARDNER, Int. J. Peptlde Protein Res. 15 73-78 (1980). C.A. MEYERS and D.H. COY, Int. J. Peptlde Protein Res. 16 248-253 (1980). E. KAISER, R. COLESCOTT, C.D. BOSSINGER and P.I. COOK, Anal. Biochem. 34 595-598 (1970). B. HEMMASI and F. BAYER, Int. J. Peptlde Protein Res. 9 63-70 (1977). R.E. NOBLE, D. YAMASHIRO and C.H. LI, J. Am. Chem. Soc. 98 2324-2328 (1926). W.R. GRAY, Methods in Enzymology 25 121-138 (1972). K.R. WOODS and K.T. WANG, Biochem. Biophys. Acta 133 369-370 (1967). A.V. SCHALLY, W.Y. HUANG, T.W. REDDING, A. ARIMURA, D.H. COY, K. CHIHARA, R.C.C. CHANG, V. RAYMOND and F. LABRIE, Biochem. Biophys. Res. Commun. 82 582-588 (1978). W. VALE, G. GRANT, M. AMOSS, R. BLACKWELL and R. GUILLEMIN, Endocrinology 91 562-572 (1972). G.H. MULDER and P.G. SMELIK, Endocrinology I00 1143-1152 (1977). A. ARIMURA, L. SAITO and A.V. SCHALLY, Endocrinology 81 235-245 (1967). C. GOMEZ-SANCHEZ, B.A. MURRY, D.C. KERN and N.M. KAPLAN, Endocrinology 96 796-798 (1975). R.G.D. STEEL and J. TORRIE, Prlnciples and Procedures of Statistics, p. 2, McGraw-Hill, New York (1960). C.M. TURKELSON, A. ARIMURA, M.D. CULLER, J.B. FISHBACK, K. GROOT, M. KANDA, M. LUCIANO, C.R. THOMAS, D. CHANG, J.K. CHANG and M. SHIMIZU, Peptldes 425-429 (1981).