Structure-activity studies with fragments and analogues of salmonid melanin-concentrating hormone

Peptides, Vol. 11, pp. 1103-1108. ©Pergamon Press plc, 1990. Printed in the U.S.A.

0196-9781/90 $3.00 + .00

Structure-Activity Studies With Fragments and Analogues of Salmonid Melanin-Concentrating Hormone B. I. B A K E R , * R. G. K I N S M A N , t C. A. M O S S , t P. D. W H I T E , t P. K. C. P A U L , $ D. W. B R O W N , t M. M. C A M P B E L L t A N D D. J. O S G U T H O R P E : ~

*School of Biological Sciences, ?School of Chemistry, .~Molecular Graphics Unit Bath University, Claverton Down, Bath BA2 7AY, U.K. R e c e i v e d 19 M a r c h 1990

BAKER, B. I., R. G. KINSMAN, C. A. MOSS, P. D. WHITE, P. K. C. PAUL, D. W. BROWN, M. M. CAMPBELL AND D. J. OSGUTHORPE. Structure-activity studies withfragments and analogues of salmonid melanin-concentratinghormone. PEPTIDES 11(6) 1103-1108, 1990.--A number of cyclic and linear fragments and analogues of MCH were synthesized and their biological potencies tested using the isolated carp scale melanophore assay. In this system, the cyclic portion MCH(5-14) exhibited only 0.1% bioactivity, which was markedly enhanced by the addition of the exocyclic sequences MCH(15-17) and MCH(I-4). The exocyclic sequence itself, MCH(1-4,15-17), had minimal activity, however. Substitution of Tyr II with phenylalanine reduced the potency of the ring structure MCH(5-14) by about 4-fold. Substitution of Gly8 with D-alanine reduced the potency of MCH(5-14) 16-fold, while both substitutions together caused a still more marked reduction (200-fold) in bioactivity. Linearized fragments of MCH, extending from MCH(15-17) to [Cys(Acm)S'Ia]MCH(1-17), showed a progressive increase in potency. The linearized forms of MCH, MCH(5-17) and MCH(5-14), were approximately 100-fold or less potent than their cyclic forms. The significant increases in bioactivity produced by the addition of the C- and N-terminal exocyclic sequence even to these linearized forms further emphasizes the importance of these regions for interaction at the receptor site. Neuropeptide

Melanin-concentrating hormone (MCH)

THE melanin-concentrating hormone (MCH) is a neuropeptide first identified in bony fish (3,10), but since found in the brains of all vertebrates examined, including the rat and human (15, 18, 19, 21). In teleost fishes, many axons project from the hypothalamic cell bodies to the posterior pituitary gland (4,16) and the peptide serves as a neurohypophysial hormone (12), causing aggregation of melanin granules in the skin pigment cells and thus promoting skin pallor when the fish moves into a pale-colored environment. Teleost skin melanophores may be very sensitive to the hormone and have provided a useful in vitro bioassay for the peptide. The biological role of MCH in mammals is not yet known but some clue to its neuronal function might be gained from a knowledge of its binding sites in the brain. To obtain this information, it is desirable to have a biologically active, radiolabeled MCH ligand but the construction of such a molecule requires an appreciation of molecular shape and a knowledge of which sequences are crucial for receptor activation. Thus, studies with the fish bioassay show that iodination of the intrinsic tyrosine residue, located in the ring structure of the molecule, significantly diminishes the bioactivity of the peptide (2). Synthesis of a tritiated MCH also requires modification of the parent MCH molecule such as, for instance, the substitution of norvaline in place of methionine, which also affects biological potency (2).

MCH analogues

MCH fragments

Melanophore bioassay

The present work examines the relative potencies of synthetic cyclic fragments and analogues of salmonid MCH, using a teleost melanophore bioassay. The biological potencies of linear fragments, produced during the synthesis of MCH, were also examined to see whether this might highlight the contribution of any one residue to biological activity. METHOD

Peptide Synthesis Peptides were synthesized using a continuous flow variant of the 9-fluorenylmethoxycarbonyl (Fmoc)/polyamide solid phase method, on a semiautomatic Pepsynthesizer apparatus (Cambridge Research Biochemicals, Harston, Cambridge) with continuous spectrophotometric monitoring of the various steps at 330 nm. Chemical synthesis of the cyclic analogues, of which full details have already been published (5), involved the initial synthesis of linear cysteine acetamidomethyl-protected molecule before cyclization. Fragments were purified by rpHPLC, using a preparative 5 Ixm C18 Hypersil column (10 x 250 mm) at a flow rate of 4 ml per min, and on an analytical 5 I~m C18 Spherisorb column (4.6 x 250 mm) at a flow rate of 1 ml per min as previously described (5). Amino acid analysis and fast atom bombardment/ mass spectrometry (FAB/MS) were routinely applied to confirm

1103

BAKER ET AL.

1104

TABLE 1 THE EFFECT OF REMOVING THE EXOCYCLIC SEQUENCES AND OF LINEARIZATION ON THE POTENCY OF MCH-LIKE PEPTIDES

% Potency 1 3 5 7 9 11 13 15 17 Asp-Thr-Met-Arg-Cys-Met-Val-Gly-Arg-Val-Tyr-Arg-Pro-Cy~-Trp-Glu-Val-OH

1. MCH 2. [Cys(Acm)5"14]MCH(1-17)

100.0

Asp-Thr-Met-Arg-Cys-Met-Val-Gly-Arg-Val-Tyr-Arg-Pro-Cys-Trp-Glu-Val-OH Acm Acm

0.9

3. MCH(5-17)

Cys-Met-Val-Gly-Arg-Val-Tyr-Arg-Pro-Cy~-Trp-Glu-Val-OH

1.0

4. [Cys(Acm)5:4]MCH(5-17)

Cys-Met-Val-Gly-Arg-Val-Tyr-Arg-Pro-Cys-Trp-Glu-Val-OH Acm Acm

0.02

5. MCH(5-14)

Cys-Met-Val-Gly-Arg-Val-Tyr-Arg-Pro-Cy~

0.08

6. [Cys(Acm)5'14]MCH(5-14)

Cys-Met-Val-Gly-Arg-Val-Tyr-Arg-Pro-Cys Acm Acm

0.0002

7. Cyclo[Cys(Acm)5'14amide]MCH(5-14)

Cyclo(Cys-Met-Val-Gly-Arg-Val-Tyr-Arg-Pro-Cys "~ ~kAcm Acm ]

the structure of each peptide.

Test Solutions Peptides were dissolved in 1 mM HC1 containing 0.1% bovine serum albumin (BSA, Sigma Chemical Co., Poole, Dorset), and stored at - 4 0 ° C . In early assays, peptides were weighed out on an Oertling microbalance. In later experiments, peptides were dissolved in acetonitrile:water (1:1), and an aliquot was taken for amino acid analysis; the remaining sample was then dried under vacuum and dissolved in acid, as before, to an appropriate concentration based on the valine content of the analyzed sample.

0.0001

tated, the scales from the dorsal surface were scraped off and placed in Hank's balanced salt solution containing 10 - 4 M phentolamine (Ciba Laboratories, Horsham, West Sussex) and 1% BSA (HPA solution), adjusted to pH 7.5 with sodium bicarbonate. The melanophores on the scales became fully dispersed within 30 min. Serial dilutions of the test peptides were prepared in HPA solution in a tissue-culture 24-well plate, and 2-3 scales were placed in each well. The melanophore index was determined after 1 hr. Synthetic salmonid MCH (Peninsula Laboratories Europe Ltd., St Helens, Merseyside) was used as a standard in each assay and the percentage potency was calculated from the ECso relative to the standard. The potency of each peptide was determined in 2-6 separate assays.

Bioassays All peptides were tested in bioassays using the scale melanophores of Chinese grass carp Ctenopharyngodon idellus. Carp were obtained from the Wessex Water Authority and weighed between 25--40 g. They were kept in fresh water at 25°C and fed with commercial fish pellets. For the assay, fish were decapi5

m

j

3

g I:

2

10 . . . . .

:10

10-9

10'-8

10.7

10.6

Peptide Concentration ( mol/litre )

FIG. 1. Dose-response lines of: (a) salmonid MCH, (b) [Cys(Acm)5'zn]MCH(2-17) (L2-17), (c) MCH(5-17) (C5-17) and (d) MCH(5-14) (C5-14) in the carp scale bioassay. Each point is the mean of 3-10 separate assays, with n=3 in each assay. Bars show SEMs.

Statistics Statistical comparisons employed analysis of variance, followed by Scheffe's LSD test. RESULTS

The primary structure of salmonid MCH is shown in Table 1. The importance of a cyclic configuration for MCH bioactivity on carp melanophores was examined by comparing the potencies of cyclic MCH(1-17), MCH(5-17) and MCH(5-14) with that of the linear molecules in which cyclization was prevented by the presence of acetamidomethyl (Acm) groups on the cysteine residues. Linear MCH and MCH(5-17) had 1-2% bioactivity compared with their cyclic derivatives (Figs. 1, 2, Table 1). Linear [Cys(Acm)5'I4]MCH(5-14) was only 0.2% as potent as its cyclic form and, for this particular linear compound, the dose-response plot was clearly nonparallel with MCH (not shown), suggesting abnormal interaction with the receptor. Cyclization of the linear [Cys(Acm)5'I4]MCH(5-14) structure via its terminal groups to form the homodet cyclic (5-14) peptide, leaving the two Acm groups in place, did not improve the potency of the linear molecule (Table 1). This is not surprising as the ring structure of the homodet cyclic peptide is smaller than that of the heterodet cyclic peptide which is cyclized via the Cys side chains, thus imposing a different geometry on the cyclic structure. These bioassays also show that removal of the N- and C-ter-

BIOACTIVITY OF MCH ANALOGUES

1105

TABLE 2 BIOLOGICAL POTENCY OF THE EXOCYCLIC SEQUENCES COMPAREDWITH THAT OF THE RING SEQUENCE OF MCH % Potency

1. MCH

1 3 5 7 9 11 13 15 17 Asp-Thr-Met-Arg-Cys-Met-Val-Gly-Arg-Val-Tyr-Arg-Pro-Cys-Trp-Glu-Val-OH

2. 3. 4. 5.

Asp-Thr-Met-Arg........................................................ Trp-GIu-Val-OH Asp-Thr-Met-Arg.................. NH(CH2)6CO ..................... Trp-Glu-VaI-OH Trp-GIu-Val-OH Cys-Met-Val-Gly-Arg-Val-Tyr-Arg-Pro-Cys-Trp-Glu-Val-OH

MCH(1-4,15-17) MCH(1-4,Aha,15-17) MCH(15-17) MCH(5-17)

6. MCH(5-14)

Cys-Met-Val-Gly-Arg-Val-Tyr-Arg-Pro-Cy]

100.0 0.0004 0.0004 0.0002 1.0 0.08

TABLE 3 THE EFFECT OF PROGRESSIVE INCREASE IN PEPTIDE LENGTH. GROWING FROM THE C-TERMINUS, ON THE BIOLOGICAL POTENCY OF LINEAR MCH % Potency

1. MCH 2. [Cys(Acm)14]MCH(10-14) 3. [Cys(Acm)14]MCH(9-14) 4. MCH(15-17) 5. [Cys(Acm)14]MCH(13-17) 6. [Cys(Acm)14]MCH(12-17) 7. [Cys(Acm)14]MCH(11-17) 8. [Cys(Acm)14]MCH(10-17) 9. [Cys(Acm)~4]MCH(9-17) 10. [Cys(Acm)la]MCH(8-17) 11. [Cys(Acm)14]MCH(7-17) 12. [Cys(Acm)I4]MCH(6-17) 13. [Cys(Acm) ta]MCH(5-17) 14. [Cys(Acm)14]MCH(4-17) 15. [Cys(Acm) 14]MCH(3-17) 16. [Cys(Acm) 14]MCH(2-17) 17. [Cys(Acm)t4]MCH(l- 17)

1 3 5 7 9 11 13 15 17 Asp-Thr-Met-Arg-Cys-Met-Val-Gly-Arg-Val-Tyr-Arg-Pro-Cys-Trp-Glu-Val-OH Val-Tyr-Arg-Pro-Cys Acm Arg-Val-Tyr-Arg-Pro-Cys Acm Trp-Glu-VaI-OH Pro-Cys-Trp-Glu-Val-OH Acm Arg-Pro-Cys-Trp-Glu-Val-OH Acm Tyr-Arg-Pro-Cys-Trp-Glu-Val-OH Acm Val-Tyr-Arg-Pro-Cys-Trp-Glu-Val-OH Acm Arg-Val-Tyr-Arg-Pro-Cys-Trp-Glu-Val-OH Acm Gly-Arg-Val-Tyr-Arg-Pro-Cys-Trp-Glu-Val-OH Acm Val-Gly-Arg-Val-Tyr-Arg-Pro-Cys-Trp-Glu-Val-OH Acm Met-Val-Gly-Arg-Val-Tyr-Arg-Pro-Cys-Trp-Glu-Val-OH Acm Cys-Met-Val-Gly-Arg-Val-Tyr-Arg-Pro-Cys-Trp-Glu-Val-OH Acm Arg-Cys-Met-Val-Gly-Arg-Val-Tyr-Arg-Pro-Cys-Trp-Glu-Val-OH Acre Acm Met-Arg-Cys-Met-Val-Gly-Arg-Val-Tyr-Arg-Pro-Cys-Trp-Glu-Val-OH Acm Acm Thr-Met-Arg-Cys-Met-Val-Gly-Arg-Val-Tyr-Arg-Pro-Cys-Trp-Glu-Val-OH Acm Acm Asp-Thr-Met-Arg-Cys-Met-Val-Gly-Arg-Val-Tyr-Arg-Pro-Cys-Trp-Glu-Val-OH Acm Acm

100.0 <0.00001 <0.00001 0.0002 0.0002 N.D. 0.0005 0.003 0.03 0.01 0.006 0.009 0.02 0.2 0.3 0.9 0.9

1106

BAKER ET AL. (a) 100-0

10-O

1.0. (9 e"

o IX

//A

0.1.

O"01.

0.001. 0"0001_

c

k

c L 5-17

1-17

c

Phe" Ala I Phe" Ala~

L

5-14

C 5-14

(b) 10I

2

1.0_

O C

0.1. 5

9

o

o. O-Ol.

6

7

8

10

0"001-

'._3

15 - 17

0"00011

2

3

4

5

6

7

8

9

10

11

12

13 14

15

I I I I I I I I I I I I I I I Asp Thr Met Arg Cys Met Val Gly Arg Val Tyr Arg Pro Cys Trp Glu .Val OH Acm Asm -=

I Increasing length of linear peptide

FIG. 2. The biological potency of different cyclic and linear analogues of MCH, tested on carp melanophores. The ECs0 value for each peptide was determined from dose-response curves in 2-6 separate assays, with n = 3 for each assay. Potency is the ECso value expressed as a % of salmonid MCH, tested in the same assay. (a) Potency of cyclic MCH(1-17), (517) and (5-14) compared to the equivalent linear [Cys(Acm)5'la]MCH peptides and cyclic MCH(5-14) with substituted residues, including Phe H, D-Alas and Phell,D-Alas together. (b) The influence on potency of progressively increasing the length of linear peptides from MCH(1517) to [Cys(Acm)5'ln]MCH(1-17). The residue number is shown by each value. Bars show SEMs; where none are shown, they are contained within the area of the symbol. minal exocyclic sequences has a drastic effect on bioactivity of both the cyclic and linear molecules. Absence of the N-terminal sequence from MCH, giving MCH(5-17), reduced potency 100fold [Table 1, Fig. 1; F(4,8)= 155, p<0.01]. Absence of both Nand C-terminal sequences to give MCH(5-14) reduced potency to 0.1% of the native molecule for both species of fish (Table 1, F = 84.9, p<0.001). A similar, equally highly significant influence of the terminal sequences is seen with the linear molecules (Fig. 1). Another example of the importance of the exocyclic sequences is provided by the fragments [Cys(Acm)~a]MCH(10-14) and -(9-14), which had no detectable bioactivity (both less than

0.00001%) when tested on carp melanophores at 10 4 and 10 3 M, respectively, whereas the C-terminally extended molecules [Cys(Acm)~a]MCH(10-17) and -(9-17) exhibited 0.003% and 0.03% bioactive potency (Table 3). The bioactivity of three exocyclic fragments was tested on carp scales: the C-terminal sequence MCH(15-17), the peptide formed by the joined N- and C-terminal sequences MCH(1--4,1517), and MCH(1-4,Aha, 15-17 ) in which the inserted amino-heptinoic acid (Aha) unit [-NH(CH2)6CO ] separates the two terminal sequences by the same distance as in native MCH. All three molecules showed very weak MCH-like activity (Table 2), although their dose-response plots appeared parallel to MCH itself. These peptides were purified on an HPLC column which had not previously been used for MCH-like peptides, their identity was checked by FAB/MS, and they were prepared as solutions independently of larger MCH fragments; their intrinsic bioactivity, although weak, is therefore considered to be real and not due to contamination. The side-arm assembly MCH(1-4,15-17), used at a concentration of 10 - 6 M, which is just below its minimum effective dose, did not enhance the potency of cyclic MCH(5-14) when these were tested together on carp scales (results not shown), indicating that the bioactivities of the two fragments were not synergistic; neither did it antagonize the activity of MCH. The main role of the exocyclic regions is probably to improve interaction between the cyclic region and the receptor molecule. Three cyclic peptides with substituted amino acids were tested: [Phe I t]MCH(5-14), [D-AlaS]MCH(5-14), and [D-Alas, Phe 11] MCH(5-14). These changes were selected because the residues Tyr ~j and Gly 8 show interesting properties in molecular modeling studies. The substitutions depressed bioactivity in all cases (Fig. 2a). Replacement of Tyr ~ with Phe reduced potency about 4-fold, F(4,8)= 155, p<0.05; replacement of Gly 8 with D-AIa depressed potency about 16-fold (p<0.01), while substitutions in both positions reduced potency nearly 200-fold (p<0.01). MCH was synthesized by the progressive N-terminal extension of the linearized molecule, which was prevented from premature cyclization by the presence of acetamidomethyl groups on the cysteine residues. Bioassays of the linear fragments showed a progressive increase in bioactivity with increasing length (Fig. 2b). When a regression line was drawn through values for linear MCH(1-17) to linear MCH(11-17) the relationship of most points to the line was highly significant, F(1,27)= 112, p<0.001, but the values for linear MCH(9-17) fell outside the 95% confidence limits, F(9,18) = 5.05, p = 0.002. Scheffe's LSD test showed linear MCH(9-17) was significantly (p<0.05) more potent than linear MCH(I 1-17). DISCUSSION

These studies address three main questions: whether linear fragments of MCH exhibit bioactivity and if so how it relates to that of the cyclic compounds; the contribution of the exocyclic regions (arm structures) towards the potency of the molecule; and the effect on biological potency of changing residues Tyr II and Gly 8. It is clear that the linearized molecules can exert biological activity, although this is markedly less than that of their cyclic equivalents. Molecular dynamic simulations, to investigate the conformation of MCH, MCH(5-14) and linear MCH(5-14) (17), suggest that the formation of the disulphide bridge is important, since without it the linear molecule can adopt additional conformations different from those of the cyclic structures. Nevertheless, the overall conformation adopted by the linear and cyclic sequences was found to be similar: in both cyclic and linear MCH(5-14), the region of maximum flexibility lies between CysS-Arg 9, while the region between Val ~° to C y s 14 is relatively constrained and rigid. This similarity is clearly sufficient to per-

BIOACTIVITY OF MCH ANALOGUES

mit even the linear sequence [Acm(CysS'ln]MCH(5-14) to express some bioactivity. It is not known to what extent the relatively large Acm groups interfere with binding, and it is possible that the bioactivity of linear molecules lacking these added side groups would be greater. The addition of the exocyclic sequences (15-17) and (1-4) markedly enhances the bioactivity of both cyclic and linear molecules when tested on carp melanophores. Thus the potency of cyclic MCH(5-14) is enhanced 10-fold by the addition of the Cterminal side arm to give MCH(5-17), and is increased a further 100-fold by the addition of the N-terminal side arm. The potency of linear MCH(5-17) is similarly increased about 100-fold by the addition of the N-terminal sequence. Interestingly, the linked exocyclic sequences (1-4,15-17) and (1-4,Aha,15-17) themselves exhibit very slight intrinsic activity, but this is trivial compared with the effect they exert in the whole molecule. Molecular dynamic calculations (17) show that the addition of the N- and C-terminal sequences will slightly change the secondary conformation of the ring structure, both by increasing the local conformational flexibility at the two cysteine residues and by affecting certain interactions which involve side chains of residues near the cysteine groups. Examples are the restriction of the side-chain mobility of the Met 6 near the Cys 5, and the breaking of a hydrogen bond between the -NH- of the Arg ~2 and the Pro 13 C = O , near the Cys 14. These changes in secondary structure might improve interaction of the ring structure with the receptor molecule. Additionally, the side arms may also enhance potency by helping to position and bind the peptide to its receptor site, an interpretation that is strengthened by the studies with linear molecules. Thus addition of the C-terminal exocyclic sequence to linear MCH(9-14) enhanced its bioactivity from <0.00001% to 0.03---0.01%; as already noted, N-terminal extension of linear MCH(5-17) to linear MCH(1-17) enhanced its activity 100-fold, as it did for the cyclic molecule. In these cases it is less probable that the exocyclic sequences enhance bioactivity by improving the molecular conformation of the linearized molecule. Our results do not establish whether any one residue in the Nor C-terminal sequences has particular importance. Tests with fragments of MCH(1-14) on the melanophores of Synbranchus have suggested that removal of the terminal Asp 1 causes a major decline in the potency of MCH(1-14) (7), the remaining residues 2-4 having a smaller influence. This was not the case in our tests with the linear MCH sequences, but whether this difference is attributable to the fact that the molecules were linear rather than cyclic, or to the species used for the bioassay, is unknown. Our results do not examine specific residues in MCH(15-17), but it is worth noting that Matsunaga et al. (13) and Castrucci et al. (7) found that Trp j5 plays a crucial role for Synbranchus melanophores, on which MCH(5-15) exhibited 100% potency but MCH(514) only 1% potency. The importance of Tyr ~ in the potency of MCH(5-14) was examined because molecular modeling studies suggested this residue might be important in assisting an appropriate conformation. Paul et al. (17) found that the sequence between Val 1° and Cys 14 assumes a relatively rigid conformation in both cyclic and linear molecules; Pro 13 imposes an important conformational constraint but it was proposed that another constraint could be the hydrogen bonding between Tyr ~1OH and CysSC = O or Met6N-H. Substitution of Tyr ~1 with Phe, which would delete such bonding, depressed the potency of cyclic MCH(5-14), but only by about 4fold. Since nuclear magnetic resonance studies indicate structural proximity between Tyr ~, Pro ~3 and Val m (1), these might suggest that the packing of the aromatic ring of Phe or Tyr into the space above the cyclic sequence is an energetically favorable arrangement and can be attained even without the cross-ring hydro-

1107

gen bonding to maintain an appropriate conformation. The effect of this substitution in MCH itself was not tested but could be expected to be similar. Other changes to the Tyr ~1 residue, examined in previous bioassay studies, have included the introduction of an iodine atom (2) or NO 2 group (11) onto the residue; such modifications reduced molecular potency to a much greater extent, 400- and 1000-fold, respectively, than was apparent from the Phe 11 substitution. It seems probable that the introduction of the iodine atom or the NO2 group would disrupt all types of hydrogen bonding with Tyr ~ and that their large size would further distort the overall shape of the molecule. The importance of Gly s, which occurs within the flexible region of the ring sequence, was also examined. It is well known that the glycine residue has considerable conformational freedom owing to the absence of a constraining side group. Dynamic simulations (17) also suggest that Gly s can adopt alternately an orhelical and gamma-turn conformation. Of these two alternatives, the substitution of glycine with a D-amino acid (e.g., D-Ala) would favor the hydrogen-bonded gamma-turn, and since this substitution was found to significantly diminish biological activity, it is possible that the a-helical conformation is preferred when MCH interacts with its receptor. Interestingly, molecular modeling studies suggest that when Gly is in this conformation, this part of the peptide is characterized by near [3-turus which are secondary structural features often implicated in structure-activity relationships of peptides. The influence of D-Ala8 on the potency of the linear molecule was not examined but the very low potency of linear MCH(5-14) is already near the limits of detection. One surprising and interesting observation was that N-terminal extension of the linear MCH fragments caused a progressive increase in bioactivity. When a regression line was fitted to the potency values for linear peptides MCH(13-17) up to linear MCH( 117), the relationship of most points to the line was highly significant but the values for MCH(9-17) fell outside the 95% confidence region. The potency of linear MCH(9-17) was significantly higher than linear MCH(11-17); addition of residues 8, 7, 6 and 5 did not enhance potency further but the potency of the linear MCH(117) was significantly greater than that of linear MCH(5-17). This, as we suggested earlier, could reflect the ability of the exocyclic sequence to help bind the molecule to the receptor site. These observations encourage the view that, at least in the linear molecule, the region of least flexibility within the cyclic sequence, i.e., between residues Cys ~4 to Arg 9, is particularly important for interaction with the receptor. Some of the MCH fragments investigated here using carp scales, e.g., MCH(5-17) and MCH(5-14), have been tested previously, with somewhat different results, on melanophores of the amazonian eel Synbranchus and on Tilapia (6, 7, 9-11, 13, 14). Thus, in our experiments, MCH(5-17) exhibited only 1% bioactivity relative to native MCH but it is found to have 100% potency when tested on Synbranchus (6, 7, 9) and Tilapia (10,11), and 180% potency on Poecilia (20). One reason for these discrepancies might be that the influence of the C- and N-terminals differs, depending on the species, i.e., there is species variation in the structure of the MCH receptors. Apart from their apparent contribution to binding to the receptor, the exocyclic sequences have also been shown capable of exerting a melanin-dispersing effect on tetrapod and fish melanophores (6, 7, 9, 14); variation in these properties might markedly affect the potency of the fragments in different species. ACKNOWLEDGEMENTS We thank Dr. Eberle for critical reading of the manuscript, Mr. A. Collins of the Statistics section of the Mathematics department for statistical advice, and the SERC for financial support.

1108

BAKER E T AL.

REFERENCES 1. Baker, B. I.; Brown, D. W.; Campbell, M. M.; Kinsman, R. G.; Moss, C. A.; Osguthorpe, D. J.; Paul, P. K. C.; White, P. D. Melanin concentrating hormone; molecular modelling and experimental analysis of conformation. J. Chem. Soc. Chem. Commun. 15431545; 1988. 2. Baker, B. I.; Eberle, A. N.; Baumann, J. B.; Seigrist, W.; Girard, J. Effect of melanin concentrating hormone on pigment and adrenal cells in vitro. Peptides 6:1125-1130; 1985. 3. Baker, B. I.; Rance, T. A. Further observation on the distribution and properties of melanin concentrating hormone. Gen. Comp. Endocrinol. 50:423-431 ; 1983. 4. Bird, D. J.; Baker, B. I.; Kawauchi, H. Immunocytochemical demonstration of melanin-concentrating hormone and proopiomelanocortin-like products in the brain of the trout and carp. Gen. Comp. Endocrinol. 74:442--450; 1989. 5. Brown, D. W.; Campbell, M. M.; Kinsman, R. G.; Moss, C.. A.; Osguthorpe, D. J.; Paul, P. K. C.; White, P. D.; Baker, B. I. Melanin concentrating hormone (MCH)--a structural and conformational study based on synthesis, biological activity, high field NMR and molecular modelling techniques. Biopolymers 29:609-622; 1990. 6. Castrucci, A. M. de L.; Hadley, M. E.; Wilkes, B. C.; Zechel, C.; Hruby, V. J. Melanin concentrating hormone exhibits both MSH and MCH activities on individual melanophores. Life Sci. 40:1845-1851; 1987. 7. Castrucci, A. M. de L.; Lebl, M.; Hruby, V. J.; Matsunaga, T. O.; Hadley, M. E. Melanin concentrating hormone (MCH): The message sequence. Life Sci. 45:1141-1148; 1989. 8. Eberle, A. N. The melanotropins. Chemistry, physiology and mechanisms of action. Basel: S. Karger; 1988. 9. Hadley, M. E.; Zechel, C.; Wilkes, B. C.; Castrucci, A. N. de L.; Visconti, M. A.; Pozo-Alonso, M.; Hruby, V. J. Differential structural requirements for the MSH and MCH activities of melanin concentrating hormone. Life Sci. 40:1139-1145; 1987. 10. Kawauchi, H.; Kawazoe, M.; Tsubokawa, M.; Kishida, M.; Baker, B. I. Characterisation of melanin concentrating hormone in chum salmon pituitaries. Nature 305:321-323; 1983. 11. Kawazoe, I.; Kawauchi, H.; Hirano, T.; Naito, N. Structure-activity relationships of melanin-concentrating hormone. Int. J. Prot. Pept. Res. 29:714-721; 1987.

12. Kishida, M.; Baker, B. I.; Eberle, A. N. The measurement of melanin-concentrating hormone in trout blood. Gen. Comp. Endocrinol. 74:221-229; 1989. 13. Matsunaga, T. O.; Castrucci, A. M. de L.; Hadley, M. E.; Hruby, V. J. Melanin concentrating hormone (MCH): Synthesis and bioactivity studies of MCH fragment analogues. Peptides 10:349-354; 1989. 14. Matsunaga, T. O.; Hruby, V. J.; Lebl, M.; Castrucci, A. M. de L.; Hadley, M. E. Melanin concentrating hormone (MCH): Structurefunction aspects of its melanocyte stimulating hormone-like (MSHlike) activity. Peptides 10:773-778; 1989. 15. Naito, N.; Kawazoe, I.; Nakai, Y.; Kawauchi, H.; Hirano, T. Coexistence of immunoreactivity for melanin-concentrating hormone and ~-melanocyte stimulating hormone in the hypothalamus of the rat. Neurosci. Lett. 70:75-80; 1986. 16. Naito, N.; Nakai, Y.; Kawauchi, H.; Hayashi, Y. Immunocytochemical identification of melanin-concentrating hormone in the brain and pituitary gland of the teleost fishes Onchorhynchus keta and Salmo gairdneri. Cell Tissue Res. 242:41-48; 1985. 17. Paul, P. K. C.; Dauber-Osguthorpe, P.; Campbell, M. M.; Brown, D. W.; Kinsman, R. G.; Moss, C.; Osguthorpe, D. J. Accessible conformations of melanin concentrating hormone: A molecular dynamics approach. Biopolymers 29:623-637; 1990. 18. Pelletier, G.; Guy, J.; Desy, L.; Li, S.; Eberle, A. N.; Vaudry, H. Melanin-concentrating hormone (MCH) is colocalized with ¢x-melanocyte stimulating hormone (a-MSH) in the rat but not in the human hypothalamus. Brain Res. 423:247-253; 1987. 19. Vanghan, J. M.; Fischer, W. H.; Hoeger, C.; Rivier, J.; Vale, W. Characterization of melanin-concentrating hormone from rat hypothalamus. Endocrinology 125:1660-1665; 1989. 20. Visconti, M. A.; Castrucci, A. M. de L.; Hadley, M. E.; Hruby, V. J. Ionic requirements for melanin concentrating hormone (MCH) actions on teleost Poecilia reticulata melanophores. Pigment Cell Res. 2:213-217; 1989. 21. Zamir, N.; Skofitsch, G.; Bannon, M. J.; Jacobowitz, D. M. Melanin-concentrating hormone: Unique peptide neuronal system in the rat brain and pituitary gland. Proc. Natl. Acad. Sci. USA 83:15281531; 1986.