Primate relaxin: Synthesis of gorilla and rhesus ... - Springer Link

Journal of Protein Chemistry, Vol. 13, No. 3, 1994

Primate Relaxin: Synthesis of Gorilla and Rhesus Monkey Relaxins John D . Wade, 1"4 Selena S. Layden, 1 Paul F. Lambert, 2 Hariclia Kakouris, 3 and Geoffrey W. Tregear *

Received January 1, 1994

The synthesis of the hormone relaxin from the species Gorilla gorilla (gorilla) and Macaca mulatta (rhesus monkey) has been achieved. Each of the two chains which constitute the peptide structures was assembled separately, the A-chains (24 amino acids) by the Boc-polystyrene solid-phase procedure and the B-chains (29 and 28 amino acids) by the Fmoc-polyamide (gorilla) and the Boc-polystyrene (rhesus monkey) solid-phase methods. After cleavage from the solid supports, the separate chains were purified to a high degree of homogeneity. Oxidative combination of the respective A- and B-chains in solution at high pH afforded the synthetic relaxins in low overall yield. Chemical and physiochemical characterization of the products confirmed both their purity and their conformational similarity to the human hormone. The synthetic gorilla and rhesus monkey relaxins were both found to possess potent chronotropic and inotropic activity in the isolated rat cardiac atrium assay. KEY WORDS: Solid-phase peptide synthesis; gorilla relaxin; rhesus monkey relaxin; isolated rat

heart chronotropic and inotropic assay.

the brain and heart of the rat (Osheroff et al., 1990; Osheroff and Phillips, 1991; Kakouris et al., 1992), although its function(s) in these systems remain unknown. Like insulin, relaxin is produced from a single prohormone precursor and, after enzymic scission of the C-peptide, the mature hormone comprises two dissimilar peptide chains designated A and B (Hudson et al., 1984). Unlike insulin, there is considerable variation in relaxin primary structure between species, with only 12 of approximately 54 residues being invariant. These include all three cystines which hold the hormone in a threedimensional structure analogous to that of insulin (Schwabe and McDonald, 1977). Analysis of the human relaxin genes has shown that there are two nonallelic relaxin sequences (Hudson et al. 1984). Both of the human gene sequences are consistent with expression and synthesis of a functional relaxin molecule. H o w e v e r only one of these genes, gene

1. I N T R O D U C T I O N

Relaxin is a peptide hormone produced principally by the corpus luteum of the ovary during pregnancy (Sherwood, 1988). It acts on target tissues in the reproductive tract to inhibit uterine contraction during most of pregnancy and to cause cervical ripening and remodeling of pelvic connective tissue in preparation for parturition (Downing and Sherwood, 1985). Relaxin has also been found in the male in seminal plasma. Recently binding sites for the hormone have been located in ~,H0ward Florey Institute of Experimental Physiology and Medicine, University of Melbourne, Parkville, Victoria~3052, Australia. ZPresent address: MacFarlane Burnet Centre for Medical Research, Fairfield Hospital, Fairfield, Victoria 3078, Australia. 3Department of Pharmacology, University of Melbourne, Parkville, Victoria 3052, Australia. 4To whom correspondence should be addressed. 315

I)277-F,(133/94/04(~I-0315507,00/IJ (~) 1994 Plcnum Publishing Corporation

316

Wade et al.

2, is expressed in the ovar3~ during pregnancy. The function of the gene 1 form of human relaxin has yet to be established. Native human gene 2 relaxin isolated from corpora lutea was shown to consist of a 24-amino acid A-chain and a 29-residue B-chain (Stults et al., 1990). As part of a comprehensive study of the structure and function of relaxin, we have synthesized and compared the biological activity of a number of relaxins including those from pig,. rat, and human (gene 1 and gene 2). With respect to the role of primate relaxin, the primary structure of relaxin from both the species Gorilla gorilla (gorilla) and Macaca mulatta (rhesus monkey) was determined by complementary D N A sequencing (Evans et al., 1994; Crawford et al., 1989). As in the human, two genes were detected in the gorilla, with one of these, named gene 2, being similar to the human gene 2. It is highly likely that the gorilla gene 2 relaxin is expressed in vivo and thus it is the focus of our interest. The gorilla gene 1 sequence appears to be nonfunctional due to point mutations in the A- and B-chain coding regions (Evans et al., 1994). In contrast, only a single gene was found to code for relaxin in the rhesus monkey. As shown in Fig. 1, the gorilla gene 2 relaxin consists of a 24-amino acid A-chain with the same primary structure as the human gene 2 peptide. The B-chains of the two hormones are of identical length (28 residues), but the gorilla peptide differs by three residues at the amino terminus. Two of the substitutions, B e and B 7, are homologous, with aspartic acid replacing glutamic acid in both positions. The third change at residue B 5 is

nonhomologous--a lysine residue for methionine. In contrast, the rhesus monkey relaxin has a significantly different primary structure, with five residues being replaced in the A-chain compared to the human/gorilla peptide. The B-chain is predicted to consist of 28 residues (one less at the amino-terminal end) and its sequence has differences of 6 and 7 residues compared to the gorilla and human relaxins, respectively. We undertook the chemical synthesis of both the rhesus monkey and gorilla gene 2 relaxins to enable a comparison of their activity with that of the human gene 2 peptide and thus allow some determination of the influence of the various amino acid substitutions on the biological activity.

2. EXPERIMENTAL 2.1. Materials The reagents for peptide synthesis were purchased from Fluka (Germany). Both N~-Boc and Fmoc-amino acids were obtained from Auspep (Australia), as was the Boc-Cys (MeBzl)-PAMresin used in the Boc-synthesis of the A-chain. Keiselguhr-supported polydimethylacrylamide resin (Pepsyn K, - 0 . 1 0 m m o l / g loading) was from MilliGen and derivatized before use as previously described (Wu et al., 1988). DMF was distilled prior to use. Synthetic human gene 2 relaxin [A(1-24), B(1-29)] was a generous gift from Genentech (South San Francisco, CA). 2.2. Techniques and Instrumentation

A-chaln

1

I0

5

15

20

h u m a n gene 2

QLYSALANKCCHVGCTKRSLARFC

gorilla

QLYSALANKCCHVGCTKRSLARFC

rhesus monkey

QLYMTLSNKCCHIGCTKKSLAKFC

B-~bmn

1

5

10

15

h u l n a n gene 2

DSWMEEVIKLCGRELVRAQ

gorilla

DSWKDDVIKLC

rhesus monkey

KWM DDVI

25

IAICGMSTWS

GRELVRAQ

KACGRE

20

LVRAQ

IAICGMSTWS IAI C G KST

LG

Fig. 1. Primary structure of human gene 2, gorilla gene 2, and rhesus monkey relaxins. Disulfide bond pairings are Cys A l~ Cys A ~s, Cys AU-Cys B u , and Cys A24-Cys B 23 in the human and gorilla relaxins and Cys Am-Cys A 15, Cys aU-Cys B t~ and Cys Ae4-Cys B 22 in rhesus monkey relaxin.

Routine amino acid analyses were carried out on a Beckman (Palo Alto, CA) System 6300 anlayzer after hydrolysis with 6 N hydrochloric acid containing 0.1% phenol, in evacuated, sealed tubes for 24 hr at 100~ Reversed-phase high-performance liquid chromatography (RP-HPLC) was carried out on a Waters (Milford, MA) 600 multisolvent delivery system with a variable wavelength detector. Peptide sequencing was carried out on an Applied Biosystems (Foster City, CA) model 470A protein sequencer connected to an Applied Biosystems model 120A PTH analyzer. Capillary electrophoresis was performed using an Applied Biosystems 270A System. Mass spectroscopy was performed on a Perkin Elmer (Thornhill, Ontario, Canada) SCIEX API 3 ion spray unit

Primate Relaxin: Synthesis

317

using H § as the primary charge agent. Circular dichroism (CD) spectra were obtained on an AVIV (Lakewood, N J) model 62DS unit at the Department of Biochemistry, University of Melbourne. Peptides were dissolved in 0.1% aqueous TFA at a concentration of 100/~g/ml and CD measured over wavelengths ranging from 200 to 250 nm.

-4~ Workup of the peptide was the same as described for the A-chains. The crude freeze-dried peptide was then subjected to Nin-deformylation by 30-min treatment with 10% 2-ethanolamine in 6 M urea, p H 8.0, at room temperature followed by acidification with neat TFA and then purification as described in Section 2.5.

2.3. Peptide Synthesis

2.5. Purification and Characterization

2.3.1. Synthesis of A-Chain

The separate A- and B-chain peptides were purified by preparative RP-HPLC on a Vydac C4 column using a gradient of 0.1% TFA in CH3CN. The resulting products were characterized by analytical RP-HPLC, SDS-polyacrylamide gel electrophoresis, capillary electrophoresis, amino acid analysis, and sequencing.

The A-chain of both gorilla and rhesus monkey relaxins was assembled by Boc-polystyrene chemistry on an ABI 430A peptide synthesizer as previously described (Kelly et al., 1988). The scale of syntheses was 0.25 mmol. 2.3.2. Synthesis of B-Chain The gorilla relaxin B-chain was synthesized by the continuous-flow Fmoc-polyamide method on a MilliGen 9050 synthesizer using Fmoc-amino acid O-pentafluorophenyl esters as described (Shen et al., 1990). The synthesis scale was 0.20 mmol. The B-chain of rhesus monkey relaxin was assembled by the Boc-polystyrene method as described for the A-chains above.

2.4. Peptide-Resin Cleavage and Deprotection 2. 4.1. A-Chains A single-stage treatment of the protected A-chain peptide-resins with 90% HF and 10% p-cresol was carried out at -4~ for lhr. The cleaved peptides were triturated with ice-cold ether containing 1% 2-mercaptoethanol, extracted into 0.1% aqueous trifluoroacetic acid containing 20% acetonitrile and 1% 2-mercaptoethanol and freezedried. 2.4.2. B-Chain Gorilla Relaxin. A 6-hr treatment of the peptide-resin with 85% TFA/10% anisole/5% ethanedithiol was employed after which the peptide was isolated by precipitation from ice-cold ether. After trituration, the peptide was extracted into 5% aqueous acetic acid and freeze-dried. Rhesus Monkey Relaxin. The dry peptide-resin was treated with 90% HF/10% m-cresol for 1 hr at

2.6. Relaxin Chain Combination: General Procedure The S-reduced A-chain (30 mg) and S-reduced B-chain (20 mg) were each dissolved in a degassed buffer of 36 m130 mM DTT, 8 M urea, 20 mM Tris, p H 8 . 0 , and then combined. The solution was stirred at room temperature for 20hr. Aliquots were removed for RP-HPLC analysis. When the B-chain had been consumed completely, the reaction was stopped by the addition of neat TFA. The resulting solution was centrifuged (300 rpm, 10 min) to remove insoluble matter and then the relaxin was isolated by preparative RP-HPLC on a Vydac C4 support. The product was characterized comprehensively as described above for the separate chains and also by ion spray mass spectroscopy.

2.7. Biological Activity The activities of the synthetic gorilla and rhesus monkey relaxins were measured using the in vitro isolated rat atrium assay. Cumulative dose-response curves were generated by the addition of increasing concentrations of relaxins (Kakouris et al., 1992). Chronotropic effects were determined by measurement of beats per minute above baseline using the spontaneously beating right atria. Inotropic effects were determined by measurement of increases in isometric tension using the electrically stimulated left atria.

318 3. RESULTS A N D DISCUSSION

The chemical synthesis of a two-chain peptide remains a considerable challenge particularly if more than two disulfide bonds are present. The principal difficulty relates to the acquisition of the correct cystine pairings. Experimental approaches for the synthesis of such two-chain peptides are varied and have been discussed previously (Du et al., 1965; Seiber et al., 1977; Bullesbach, 1992; Maruyama et al., 1992). For the synthesis of gorilla and rhesus monkey relaxins, we used the simplest approach, that of separate chemical assemblies of the S-reduced A- and B-chains followed by their oxidative combination in solution. This protocol has enabled us to prepare successfully a variety of mammalian relaxins. The A-chains of gorilla and rhesus monkey relaxin were synthesized by the Boc-polystyrene methodology. The use of Fmoc-based methods was not considered, due to the presence of C-terminal cysteine in both peptides. This residue is prone to significant racemization during the N~-Fmoc deprotection step by the base, piperidine. Such a side reaction is not observed during Boc-peptide synthesis (Atherton et al., 1991). The B-chain of rhesus monkey relaxin was also prepared by the Boc-polystyrene method. However, more recent work in our laboratory showed that the B-chain of mammalian relaxin could be assembled in higher purity by the Fmoc solid-phase procedure. This is most likely due to the chemistry being less destructive to sensitive residues in the B-chain, namely tryptophan, cysteine, and methionine. For this reason, the gorilla relaxin B-chain peptide was assembled by the Fmoc method. The A- and B-chains of each species were assembled in good purity and, following purification by conventional RP-HPLC, were obtained in average yield of 10-13% and 5-7% for the A- and B-chains respectively. The latter figure, although low, is acceptable because of the considerable difficulty caused by the sparing solubility of the S-reduced B-chains. Best conditions for dissolution were afforded by addition of 8 M urea, 25 mMTrisHC1, p H 8 . 0 , containing 1% mercaptoethanol and incubation at 37~ for 2hr. Immediately prior to HPLC purification, the mixture was diluted with 0.1% aqueous TFA, centrifuged, and passed through a 15-~m filter prior to RP-HPLC purification. Each of the four purified chains gave

Wade et al.

acceptable amino acid analysis data (results not given). Further confirmation of their high purity was afforded by RP-HPLC in two different buffer systems and by capillary zone electrophoresis at p H 2 . 5 (data not shown). In addition, amino acid sequencing confirmed that the peptides possessed the correct arrangement of amino acids (data not shown) with no significant preview. For both the gorilla and rhesus monkey relaxin, the combination of the two chains in solution was rapid, being essentially complete after just 5hr as indicated by the absence of the S-reduced B-chain on RP-HPLC. It is our experience that the presence of minor peptide impurities can markedly slow the rate of chain combination, with some reactions taking up to 72 hr before being complete. The recovery of gorilla relaxin from the chain combination was low, with just 0.30rag being obtained from 30.0mg of A-chain and 20.0mg of B-chain. The principal source of loss was the precipitation of the S-reduced B-chain during the combination. Despite the use of a variety of denaturants and organic solvents, the degree of solubilization was insufficient to give combination yields comparable to those achieved for other mammalian relaxins. These have been as high as 50% relative to the starting B-chain peptide. This result was particularly curious in the light of the fact that the gorilla B-chain differs from the human gene 2 counterpart by only the three N-terminal residues, two of which are homologous and the third being the replacement of a methionine with lysine, which could be expected to enhance the solubility of the peptide. The human gene 2 relaxin has been prepared in good yield and less difficulty was experienced in its preparation (Lambert, 1988). The synthetic gorilla and rhesus monkey relaxins were purified by conventional RP-HPLC in overall yield of 0.2% and 0.7%, respectively, relative to the starting resin-bound B-chain. The low yield attests to the need to develop alternative methods for the synthesis of two-chain, disulfidebonded peptides. Such investigations are currently being carried out in our laboratory. Amino acid analysis (Table I) and sequencing of the purified synthetic relaxins showed each to comprise one A-chain or one B-chain only. Further, the relaxins could be S-reduced to produce two peptides which, when separated, were identified as being the A- and B-chains respectively. RP-HPLC of the synthetic relaxins in two different

Primate Relaxin: Synthesis

319

Table i. Amino Acid Composition of Synthetic Gorilla and Rhesus Monkey Relaxins" Amino acid Asp Thr Ser Glu Gly Ala Cys b

VaF Met lie c Leu Tyr Phe His Lys Trp b Arg

Gorilla

Rhesus monkey

4.12 (4) 1.97 (2) 4.95 (5) 3.14 (3) 3.18 (3) 5.12 (5) ND (6) 2.62 (3) 0.87 (1) 2.67 (3) 4.87 (5) 1.08 (1) 0.97 (1) 1.14 (1) 3.92 (4) NO (2) 3.91 (4)

3.05 (3) 3.36 (3) 3.69 (3) 3.27 (3) 4.49 (4) 4.29 (4) ND (6) 0.81 (2) 2.17 (2) 2.97 (4) 5.53 (5) 0.81 (1) 1.14 (1) 0.93 (1) 7.82 (7) ND (1) 1.62 (2)

a Theoretical values in parentheses. b Not determined; destroyed during hydrolysis. c Ile-Val bond incompletely acid-hydrolyzed.

c

s o
O ..O

residues near the amino terminus of the B-chain has little effect on activity in this assay system. This region, therefore, may not have a significant role in sequence-specific receptor binding. In contrast, the rhesus monkey relaxin showed weaker chronotropic and ionotropic activity when compared with both human and gorilla gene 2 peptides (Fig. 4). This result reflects the fact that it differs from human relaxin by 13 residues that are fairly evenly distributed throughout the hormone and as a consequence may have a different overall tertiary structure. Further, specific residues which interact with the receptor may be altered even if the peptide secondary structure is little changed. It is not possible as yet to identify the key substitutions which lead to a decline in biological activity in this assay system. Systematic replacement of individual residues is required to shed further light on their importance. The two-chain, three-disulfide-bonded peptides gorilla gene 2 and rhesus monkey relaxin have thus been successfully assembled and demonstrated to have activity in the atrial assay. These peptides are currently being employed in our continuing studies of the role of relaxin in the primate.

p.

.2

o.1

C

ACKNOWLEDGMENTS

0 ffl . I~ 0.0 10-11

-...., 1 0 "~~

1 0 "g

1 0 .8

1 0 .7

log [Rlxl (M) e-

200"

(o

0 .o

._=

This research was supported by an Institute block grant from the National Health and Medical Research Council of Australia. We thank Bill Rule, Mick Petrovski, and Marie John for technical support and Alun Jones, Centre of Drug Design and Development, University of Queensland, Australia, for provision of ion spray mass spectroscopy results. We acknowledge helpful discussions with Larry Eddie (Howard Florey Institute) and Roger Summers (Department of Pharmacology, University of Melbourne).

100'

REFERENCES Q) ~3 0 0

CC

.

0 "11

-

9

. . . .

,

. . . . . . . .

1 0 "~~

,

. . . . . . . .

1 0 .9

,

1 0 "s

. . . . . . . .

,

1 0 .7

log [RIx] (M) Fig. 4. Chronotropic (lower panel) and inotropic (upper panel) effects of synthetic human gene 2 (11), gorilla gene 2 (9 and rhesus monkey (&) relaxins on in vitro rat atria, n = 5 experiments for each peptide, p >0.05 (unpaired t-test) comparing -log relaxin concentration with ECso.

Atherton, E., Hardy, P. M., Harris, D. E., and Matthews, B. H. (1991). In Peptides 1990 (Giralt, E., and Andreu, D., eds.), ESCOM, Leiden, The Netherlands, pp. 243-244. Bullesbach, E. (1992). Kontakte (Darmstadt), 1992, 21-29. Crawford, R. J., Hammond, V. E., Roche, P. J., Johnston, P. D., and Tregear, G. W. (1989). J. Mol. Endocrinol. 3, 169-174. Downing, S. J., and Sherwood, O. D. (1985). Endocrinology 116, 1215-1220. Du, Y. C., Jiang, R.-Q., and Tsou, C.-L. (1965). Sci. Sinica 114, 229-236. Evans, B. A., Fu, P., and Tregear, G. W. (1994). Endocrin. J. 2, 81-86.

Primate Relaxin: Synthesis Hudson, B. A., John, M., Crawford, R., Haralambidis, J., Scanlon, D., Gorman, J., Tregear, G., Shine, J., and NiaU, H. (1984). EMBO J. 3, 2333-2339. Kakouris, H., Eddie, L. W., and Summers, R. J. (1992). Lancet 339, 1076-1078. Kelly, P. J., Lambert, P. F., Tregear, G. W., and Johnston, P. D. (1988). In Peptides 1988 (Jung, G., and Bayer, G., eds.), de Gruyter, Berlin, pp. 178-180. Lambert, P. F. (1988). Ph.D. Thesis, University of Melbourne, Melbourne, Australia. Maruyama, K., Nagata, K., Tanaka, M., Nagasawa, H., Isogai, A., Ishizaki, H., and Suzuki, A. (1992). J. Protein Chem. 11, 1-12. Osheroff, P. L., and Phillips, H. S. (1991). Proc. Natl. Acad. Sci. USA 88, 6413-6417. Osheroff, P. L., Ling, V. T., Vandlen, R. L., Cronin, M. J., and Lofgren, J. A. (1990). J. Biol. Chem. 265, 9396-9401.

321 Schwabe, C., and McDonald, J. K. (1977). Science 197, 914915. Seiber, P., Kamber, B., Hartmann, A., Johl, A., Riniker, B., and Rittel, W. (1977). Heir. Chim. Acta 60, 27-37. Shen, J.-H., Eddie, L. W., Lambert, P. F., Tregear, G. W., and Wade, J. D. (1990). In Innovation and Perspectives in Solid Phase Synthesis (Epton, R., ed.), SPCC, UK, pp. 571-576. Sherwood, O. D. (1988). In The Physiology of Reproduction (Knobil, E., and Neill, J., eds.), Raven Press, New York, pp. 585-673. Stults, J. T., Bourell, J. H., Canova-Davis, E., Ling, V. T., Laramee, G. R., Winslow, J. W., Griffin, P. R., Rinderkneeht, E., and Vandlen, R. L. (1990). Biomed. Environ. Mass Spectrom. 19, 655-664. Wu, C.-R., Wade, J. D., and Tregear, G. W. (1988). Int. J. Pep#de Protein Res. 31, 47-57.