Vol. 122, No. 3, 1984 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

August 16, 1984 Pages 1218-1224

ISOLATION AND CHARACTERIZATION OF CAPRINE CORTICOTROPIN-RELEASING FACTOR

Nicholas Ling, Frederick Esch, Peter BBhlen, Andrew Baird and Roger Guillemin

Laboratories for Neuroendocrinology The Salk Institute for Biological Studies, La Jolla, CA 92037

Received July 9, 1984

SUMMARY. The hypophysiotropic peptide, corticotropin-releasing factor (CRF) mated from caprine hypothalamic median eminence tissue by means of acid extraction, immunoaffinity chromatography, gel filtration and two steps of reverse-phase high-performance liquid chromatography (RPLC). Amino acid sequence determination using a gas-phase sequencer established the primary structure of the first 39 residues from the NH,-terminus. The nature of the COOH-terminal dipeptide was elucidated by Staphylococcus aureus V8 protease digestion of the native peptide, dansylation of the digest and comparative RPLC studies with the dansylated dipeptides Ile-Ala-NH,, Ile-Ala-OH, Ala-Ile- NH, and Ala-Ile-OH. The complete structure of the peptide was established as: H-Ser-Gln-Glu-Pro-Pro-Ile-Ser-Leu-Asp-Leu-Thr-Phe-His-Leu-Leu-Arg-Glu-Val- Leu-Glu-Met-Thr-Lys-Ala-Asp-Gln-Leu-Ala-Gln-Gln-Ala-His-Ser-Asn-Arg-Lys-Leu- Leu-Asp-Ile-Ala-NH2, which is identical to that of ovine CRF.

The control of pituitary adrenocorticotropin (ACTH) secretion is mediated

primarily by a hypothalamic corticotropin-releasing factor (CRF) and to a

lesser extent by vasopressin and norepinephrine (1). In 1981, Vale et al. (2)

reported the characterization of the first hypothalamic CRF, isolated from the

ovine species. When the structure of ovine CRF was established, Vale et al.

(2) noted that it had considerable sequence homology with sauvagine (3), a

peptide isolated from the skin of a South American frog, Phyllomedusa sauvagie

and recognized by Falaschi et al (4) as a putative CRF of amphibians in view

of its in vitro ACTH-releasing activity. Subsequently, two CRF-related --

peptides were characterized from the urophysis of two species of fish, one

from sucker (5) and the other from carp (6). Because most physiological

studies on the regulation of ACTH secretion are done with rat, Rivier et al.

ABBREVIATIONS: CRF = corticotropin-releasing factor; GRF = growth hormone-releasing

factor; CRF-LI = CRF-like immunoreactivity; ACTH = adrenocorticotropin; RPLC = reverse-phase high-performance liquid chromatography; >PhNCS-AA = phenyl- thiohydantoin-amino acid.

Vol. 122, No. 3, 1984 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

(7) isolated and characterized rat CRF. More recently, Shibahara et al. (8)

reported the structure of human CRF deduced from the cDNA encoding the gene

for human CRF from a genomic library. The structure of human CRF deduced by

them was identical to that of rat CRF.

Our laboratories have recently devised a relatively rapid and efficient

method for the isolation of growth hormone-releasing factor (GRF) from porcine

(9), bovine (lo), caprine. ovine and human (11) hypothalamus. The procedure

involves acid extraction of the hypothalamic tissues, concentration of the GRF

activity in the extract on an immunoaffinity column, gel filtration of the

immunoreactive GRF recovered from the affinity column. followed by two steps

of reverse-phase high-performance liquid chromatography (RPLC). Since the GRF

depleted tissue extracts from bovine and caprine hypothalamus were still

available, we initiated a project to isolate and characterize bovine and

caprine CRF using the same methodology applied in the isolation of GRF. The

successful characterization of caprine CRF is reported here and the work on

bovine CRF has been reported in another paper (12).

METHODS AND MATERIALS

Derivatized amino acids used in the synthesis of the dipeptides were of the L- configuration and purchased from Bachem, Inc (Torrance, CA). The Na-amino function was protected exclusively with the t-butyloxycarbonyl (Boc) group. Chloromethyl-substituted polystyrene resin cross-linked with 1% divinylbenzene was obtained from Lab Systems, Inc. (San Mateo, CA). The substitution was 0.9 mmol chlorine per gram according to the manufacturer. p-Methylbenzhydrylamine resin was synthesized in our laboratory using a previously published procedure (13). The substitution was 0.60 mmol amine per gram by Gisin analysis (14).

Peptide isolation. Approximately 2600 caprine hypothalamic median eminence fragments weighing 148 g after lyophilization were boiled in 825 ml water for 3 min to inactivate endogenous proteases. After cooling in an ice-bath, an equal volume of 0.6 M HCl containing phenylmethylsulfonyl fluoride and pepsta- tin A (10 pg/ml each) was added and the mixture homogenized for 5 min using a Polytron (Brinkman Instruments, Los Angeles, CA). The work-up of the rest of the extraction procedure was the same as that used for the isolation of human GRF (11). The GRF activity in the caprine hypothalamic extract was first re- moved by pumping the solution through an anti-GRF immunoaffinity column used in the isolation of porcine (9) and bovine (10) GRF; The caprine hypothalamic extract with the GRF activity removed was then pumped through another immuno- affinity column (3.5 X 7.2 cm, V bed = 70 ml) constructed by coupling to Affi- gel 10 (Bio-Rad Laboratories, Richmond, CA) a purified IgG fraction raised in rabbits (11) immunized with a mixture of [TyrZ1]oCRF(21-41) and methylated bovine serum albumin by a previously described procedure (15). After washing off the unbound material with phosphate-buffered saline (pH 7.4) (0.020 M Na2HPOh/0.150 M NaCl/O.Ol% NaN,), the adsorbed CRF activity was eluted with 1 M CH,COOH. Location of the CRF-LI in the column effluents was performed with radioimmunoassay using the same pool of [TyrZ1]oCRF(21-41) antiserum as that

1219

Vol. 122, No. 3, 1984 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

used to construct the affinity column. The fractions containing CRF-LI were pooled and further purified by gel filtration on Sephadex G-75 (Pharmacia Fine Chemicals, Piscataway, NJ) and two steps of RPLC, all done at room tempera- ture. Details of the chromatographic conditions are given in the legend to figure 1.

Structural characterization. Amino acid analyses (16) and sequence determina- tion 1111 were performed as previously described. To determine the nature of the COOH-terminal dipeptide of caprine CRF, 100 pmol of the native peptide was digested with Staphylococcus aureus V8 protease in 0.020 M Na,HPO, (pH 7.8) at 37" for 15 h with a 1:35 (mole/mole) enzyme to substrate ratio (18). After dansylation (19) the digestion mixture was analyzed by RPLC and compared with the retention time of the four synthetic dipeptides Ile-Ala-NH2, Ile-Ala-OH, Ala-Ile-NH2 and Ala-Ile-OH which were prepared by solid phase methodology (20).

Dipeptide Synthesis. The dipeptides with a free acid at the COOH-terminus were prepared with the chloromethyl-substituted polystyrene resin and the ones with an amidated COOH-terminus were prepared with the p-methylbenzhydrylamine resin using a previously published procedure (11). The synthetic peptides were purified by ion-exchange chromatography and partition chromatography as described (11).

RESULTS AND DISCUSSION

The isolation of caprine CRF was achieved by using the five step purifi-

cation procedure involving acid extraction, immunoaffinity chromatography, gel

filtration and semi-preparative and analytical RPLC. Gel filtration of the

CRF-LI recovered from the immunoaffinity column yielded one zone of CRF activ-

ity (Figure 1A) which upon semi-preparative RPLC, was separated into two main

CRF-LI fractions (Figure 1Bl. The latter eluting fraction containing the

major CRF-LI was further purified on an analytical RPLC system which yielded

one major peak of CRF-LI with distinct 254 nm UV absorption (Figure 1C). The

CRF-LI peak from Figure 1C was subjected to amino acid analysis and results

showed that it contained 1.7 nmol of peptide with a composition in agreement

with that of ovine CRF (Table 11.

Direct Edman degradation of 500 pmol of the peptide in a Model 470A gas-

phase protein sequencer (Applied Biosystems, Foster City, CA) yielded its

primary structure up to 39 amino acids from the NH,-terminus (Figure 2).

Comparison of the established sequence of caprine CRF(l-39) with the amino

acid composition of the intact peptide narrowed the COOH-terminus to one of

four possible dipeptides: Ile-Ala-NH2, Ile-Ala-OH, Ala-Ile-NH2 and Ala-Ile-OH.

These four dipeptides were synthesized and used to compare with RPLC analysis

of the dansylated Staphylococcus aureus V8 protease digestion product from

native caprine CRF. Figure 3 shows that only synthetic dansyl-Ile-Ala-NH,

1220

Vol. 122, No. 3, 1984 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

500 750 1000 1250 1500 1750 VOLUME (ml)

I3 r-

I

1’

I’ II /

/ -120 -ii Ll

-1 -96 <

-72 - s

i -40 I

LA

-24 5 - 0

0 30 60' - 90 120

TIME (minutes)

TIME (minutes)

FIGURE 1: Isolation of caprine CRF.

A) Sephadex G-75 gel filtration of the immunoaffinity purified caprine CRF. The sample was applied to a 4.5 x 117 cm column and eluted with 1 M CHICOOH/O.P% e,-mercaptoethanol (v/v) at 60 ml/h and 15 ml fractions were collected.

B) Semi-preparative RPLC of the pooled CRF-LI from gel filtration. The sample (eluting between 1095 and 1260 ml in Fig. 1Al was loaded by pumping it directly onto an Ultrasphere ODS column (5 urn particle size. 1 X 25 cm, Altex) and eluted with a 0.25 M triethylammonium phosphate (pH 3.0)/aceto- nitrile mobile phase. Fractions of 3 ml were collected at 1 ml/min.

Cl Analytical RPLC of the major CRF-LI from semi-preparative RPLC. The sample was loaded onto a RP-300 Cl8 column (10 Vm particle size, 0.46 x 25 cm, Brownlee) and eluted with a 0.2% heptafluorobutyric acid/acetonitrile mobile phase. Fractions of 2.5 ml were collected at 1 ml/min.

1221

Vol. 122, No. 3, 1984 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

TABLE I: AMINO ACID COMPOSITION OF CAPRINE CRF

Amino Acid Caorine bCRF Nearest Inteaer Number

Asx 3.93 Thr 1.95 Ser 2.60 Glx 7.01

GUY 0

Ala 3.73 Val 1.09 Met 0.78 Ile 2.25 Leu 8.41

Tyr 0 Phe 1.00 His 1.88

Tv 0

LYS 2.18

Arg 2.20 Cya* 0 Pro 1.99

(4) (2) (3) (7)

(4) (1) (1) (2) (8)

(1) (2)

(2) (2)

(2)

* Cysteine was determined as cysteic acid.

elutes with a retention time identical to that of one of the dansylated

digestion products. Thus the complete primary structure of caprine CRF is

established as shown in Figure 4 using

Amount Applied: 500 pmols Initial Yield: 38.0%

6 10 15

approximately 600 pmol of peptide. The

Average Repetitive Yield: 92.2% Average Lag per Cycle: 1.2%

SOEPPlSLDLTFHLLREVLEMTKADOLAOOAHSNRKLLDXX 6001

FIGURE 2: Gas-phase sequence analysis ot 500 pmols caprine CRF.

1222

Vol. 122, No. 3, 1984 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

Oansyl+Chloride

Dansic Acid

\

/ /

/ /

A’ /

/ /

/ /

/ /

/ A-l-OH

I-A-OH

10 20 TIME (min)

100

80

60 f

1

40 g

20

I

FIGURE 3: Reverse-phase high-performance liquid chromatographic identifica- tion of the COOH-terminal dipeptide of caprine CRF. The dansylated samples were loaded onto an Ultrasphere ODS column (5 urn particle size, 0.46 X 25 cm, Altex) and eluted at 1 ml/min with the indicated gradient of solvent B added to solvent A. Solvent A was 0.25 M NaOAclpH 5.9):MeOH:THF (80:19:1, v/v) and Solvent B was 20% solvent A and 80% MeOH. The column effluent was monitored with a Kratos/Schoeffel FS950 Fluoromat fluorescence detector. Elution positions,of the synthetic dipeptides are as indicated by the arrows.

established structure of caprine CRF is identical to that of ovine CRF. The

earlier eluting CRF-LI fraction in Figure 16 was also purified to homogeneity

by analytical RPLC (data not shown). It has the same amino acid composition

as that presented in Table 1 and may thus represent an oxidized form of

caprine CRF.

Characterization of caprine CRF adds the structure of one more hypothala-

mic peptide with high intrinsic CRF activity in addition to the already char-

acterized ovine (21, bovine (121, rat (71 and human (8) hypothalamic CRF (Fig-

6 10 II 20 16 90 36 40

BOVINE CAF SOEPPISLDLTFHLLREVLEMTKADCtLA9~AHNNRKLLDlA-NH2

SQEPPISLDLTFHLLREVLEMTKADQLAC2QAHSNFtKLLDiA-NH2

SEEPPlSLDLTFHLLREVLEMARAEQLAClQA~SNRKLMElI-NH,

BE;; 0 NDDPPlSlDLTFHLLRNMlEMARlENEREaAGLNRKYLDEV 1 L__l 1 r-1 ;1 ijffl- NH,

NDDPPlSlDLTFHLLRNMlEMARNENQREfJA0LNRKYLDEV-NH2

QQPPlSlDLSLELLRKMlElEKOEKEKEKQtJAANNRLLLDTl-NH2

FIGURE 4: Primary structures of native peptides with CRF activity.

1223

Vol. 122, No. 3, 1984 BIOCHEMICAL AND BIOPHYSICAL RESEARCH COMMUNICATIONS

we 4). Including the three CRF-related peptides: sauvagine (3), sucker uro-

tensin I (5) and carp urotensin I (61, isolated from the lower vertebrates, a

total of six different peptides with CRF activity have been characterized.

ACKNOWLEDGEMENTS

We thank M. Mercado, R. Schroeder, F. Castillo, K. Von Dessonneck, T. Durkin, R. Klepper, M. Regno, D. Angeles and T.C. Chiang for their expert technical assistance. This research was supported by program grants from NIH (HD-09690 & AM-188111 and the Robert J. & Helen C. Kleberg Foundation.

REFERENCES

1. 2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

13.

:::

16. 17. 18.

19.

20.

Vale, W. and Rivier, C. (1977) Fed. Proc. 36, 2094-2099. Vale, W., Spiess, J., Rivier, C. and RivG, J. (1981) Science 213 1394-1397.

-*

Montecucchi, P., Henschen, A. and Erspamer, V. Physiol. Chem. 360, 1178.

(1979) Hoppe-Zeyler's Z.

Falaschi, P., mchiorri, P., D'Urso, R., Negri, L., Rocco, A. and Erspamer, V. (1980) Hormone Res. 13, 329. Lederis, K., Letter, A., McMasteT D., Moore, G. and Schlesinger, D. (1982) Science 218, 162-164. Ichikawa, T., mster, D., Lederis, K. and Kobayashi, H. (1982) Peptides 3, 859-867.

T\i'vier, J., SpieSS, J. and Vale, W. (1983) Proc. Natl. Acad. Sci. USA 80 4851-4855.

-'

Shibahara, S., Morimoto, Y., Furutani, Y., Notake, M., Takahashi, H., Shimizu, S., Horikawa, S. and Numa, S. 779.

(1983) The EM80 Journal 1, 775-

Bbhlen, P., Esch, F., Brazeau, P., Ling, N. and Guillemin, R. (1983) Biochem. Biophys. Res. Commun. 116, 726-734 Esch, F., ahlen, P., Ling, N., Brazeau, P. and Guillemin, R. (1983) Biochem. Biophys. Res. Commun. 117, 772-779. Ling, N., Esch, F., Btlhlen, PTing, N., Brazeau, P., Wehrenberg, W.B. and Guillemin, R. (1984) Proc. Natl. Acad. Sci. USA 81 (in press). Esch, F., Ling, N., Bt)hlen, P., Baird, A. and Gllemin, R. (1984) Biochem. Biophys. Res. Commun. (in press). Rivier, J., Vale, W., Burgus, R., Ling, N., Amoss, M., Blackwell, R. and Guillemin, R. (1973) J. Med. Chem. 16, 545-549. Gisin, B.J. (1972) Anal. Chim. Acta-58, 248-249. Benoit, R., Bbhlen, P., Ling, N., -@skin, A., Esch, F., Brazeau, P., Ying, S-Y. and Guillemin, R. (1982) Proc. Natl. Acad. Sci. USA 79 917- 921.

-’

Bb'hlen, P. and Schroeder R. (1982) Anal. Biochem.126, 144-156. Esch, F. (1984) Anal. Biochem. 136, 39-47. Houmard, J. and Drapeau, G. (19m Proc. Natl. Acad. Sci. USA 69 3506- 3509.

-'

Tapuhi, Y., Schmidt, D.E., Lindner, W. and Karger, B.L. (1981) Anal. Biochem. 115, 123-129. Merrifield,.B. (1963) 3. Pm. Chem. Sot. 85 2149-2154. -’

1224