Vol. 181, No. 2, 1991 December 16, 1991
BIOCHEMICAL
AND BIOPHYSICAL RESEARCH COMMUNICATIONS
Pages 664-670
Topochemical Analysis of Morphiceptin and Dermorphin Bioactivities Toshimasa YAMAZAKI,
Seonggu RO, and Murray GOODMAN*
Department of Chemistry, 0343, University of California, San Diego, La Jolla, CA 92093 Received
October
11,
1991
Summary To elucidate the topochemical requirements for bioactivities
of morphiceptin (TyrPro-Phe-Pro-NH2) and dermorphin (Tyr-D-Ala-Phe-Gly-Tyr-Pro-Ser-NH*), we have designed and synthesized two diastereomers Tyr-(L and D)-(NMe)Ala-Phe-D-Pro-NI$. Both the analogs display high activities in the guinea pig ileum assay. The only difference in the composition of these two diastereomers arises from the chirality at residue 2. The high p-receptor activities are attributed to structures where the Tyr*-L-(NMe)Ala2 amide bond assumes a cis configuration while the Tyrt-D-(NMe)Ala2 amide bond assumes a trans configuration. Accessible space studied for the second residues of these molecules continns the fact that the L(NMe)Ala’ analog belongs to the morphiceptin family of opioids while the D-(NMe)Ala’ analog belongs to the dermorphin class of opioids. The similarity in the spatial array of the analogs explains their high p-receptor activities and indicates that they are likely recognized at the same opioid receptor. 0 1991
Academic
Press,
Inc.
Morphiceptin (Tyr’-Pro2-Phe3-Pro4-NH2) and dermorphii (Tyr*-D-Ala2-Phe3-Gly4-TyrsPro6-Ser7-NH2) are highly p-receptor selective peptide opioids. It is well known that the receptor recognition of these peptides requires specific arrangements of the amine and phenolic groups of Tyr’ and the aromatic group of Phe3. Since the biologically important Tyr’ and Phe3 are joined by a single amino acid, the second residue must play a significant role in orienting these residues in the correct array necessary for bioactivity. The morphiceptin and dermorphin classes of opioids exhibit opposite chiral requirements at residue 2. Morphiceptin requires an L-chirality for Pro2 as indicated by the fact that Tyr-D-Pro-Phe-Pro-NH2 is inactive (1). On the contrary, incorporation of L-amino acids at position 2 of dermorphin results in a remarkable reduction in bioactivity (2). Unlike dermorphin which adopts only the truns amide bond linking residues 1 and 2, morphiceptin exhibits cis and trans isomers about~the Tyr’-Pro’ amide bond in a ratio of 30 : 70 (3). The cis/trms configuration about the Tyr’-Pro2 amide bond is particularly significant for the relative orientation of the biologically important Tyr’ and Phe3 residues. The activities of the morphiceptin analogs incorporating D-Pro and (L and D)-Val at position 4 show that the structure of residue 4 does not affect recognition at the receptor site (4,5). In previous papers (6,7), * To
whom
correspondence
should
CKM-291x/91 $1.50 Copyright 0 1991 by Academic Press, Inc. All rights of reproduction in any form reserved.
be
addressed.
664
Vol.
181,
No.
2, 1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
we reported on morphiceptin analogs containing 2-aminocyclopentanecarboxylic acid (2-A&) as the second residue in place of proline Tyr-2-A&-Phe-X4-NH2 [X4 = Pro and (L and D)-Val] to eliminate cislfruns isomerization about the amide bond between residues 1 and 2. Among the four stereoisomers, only the analogs containing cis-(lS,2R)-2-A& show bioactivity. Although the ~-A&Z analogs can only adopt a tram amide bond structure about Tyr-2-A&, the bioactive analogs Tyr-cis(lS,2R)-2-A&-Phe-X4-NH2 are topochemically similar to morphiceptin with the Tyr-Pro amide bond in a cis configuration (7). We therefore proposed that a cis configuration about the Tyr’-Pro’ amide bond is required for bioactivity of the morphiceptin related analogs containing a proline as the second residue. To elucidate the configurational requirements of the Tyr’-X2 amide bond (where X2 represents residue 2) and the conformations of the second residue for morphiceptin and dermorphin bioactivities, we synthesized Tyr’-X2-Phe3-D-Pro4-NH2 [X2 = L-(NMe)Ala and D(NMe)Ala]. Since an N-methylalanine is an acyclic mimetic for a proline, the L-(NMe)Ala* and the D-(NMe)Ala* analogs are closely related to morphiceptins; the active Tyr-Pro-Phe-D-ProNH2 and the inactive Tyr-D-Pro-Phe-D-Pro-NH*, respectively. ln addition, Tyr-D-(NMe)AlaPhe-D-Pro-NH2 is also an N-methylated analog of the bioactive D-Ala* dermorphin-like molecule : Tyr-D-Ala-Phe-D-Pro-NH2 (8). In order specifically to explain the topochemical effects of the second residue on bioactivity, we also synthesized Tyr-Ala-Phe-D-Pro-NH2 as a control molecule although we expected that this analog would be inactive. After due consideration of accessible space for the second residues of all these opioids synthesized, we predicted that both the L- and D-(NMe)Ala* analogs would be active at the u-receptor. However, we postulated that the activity of the former molecule is attributed to structures with the Tyr’(NMe)Ala* amide bond in a cis configuration (morphiceptin-like) while a trans structure about the Tyr’-D-(NMe)Ala* amide bond is responsible to the activity of the latter molecule (dermorphin-like). Below we present our experimental evidence to support these structural proposals to explain bioactivity. MATERIALS
AND METHODS
Synthesis The compound Tyr-D-(NMe)Ala-Phe-D-Pro-NH* was synthesized using solid phase techniques. The protected tetrapeptide was assembled on a methylbenzhydrylamine resin employing Boc strategy and the DCC/HOBt coupling method. A low-high HF cleavage was used to cleave the peptide from the resin and to remove protecting groups. The syntheses of the other compounds were carried out in solution using Boc chemistry and EDC/HOBt coupling method. Acidolytic deprotection of protected tetrapeptides with TFA and thioanisole provided crude products. The crude products were purified by affinity chromatography and gel filtration. The purities of the final products were examined by HPLC (> 99 %). The charaterizations of the product are summarized in Table I. Spectroscopy Measurements The ‘H-NMR spectra were recorded at 500 MHz on a General Efectric GN-500 spectrometer. The samples were prepared in DMSO-& at a concentration of 10 mM. Tetramethylsilane was used as an internal reference for the determination of chemical shifts. All of the proton resonances were assigned using two-dimensional HOHAHA (9) and ROESY (10) experiments. The HOHAHA experiments were carried out using the MLEV17 (11, 12) with a mixing time of 100 ms and TPPI (13). The ROESY spectra were acquired using a 200 ms mixing time. Computer Simulations Molecular mechanics calculations were carried out employing the DISCOVER flexible geometry program (14). The (jr, v) energy contour maps.of N-acetyl-N’665
Vol.
181, No. 2, 1991
BIOCHEMICAL
AND BIOPHYSICAL RESEARCH COMMUNICATIONS
TableI. Characteristicsof SynthesizedAnalogs FAB-MS NJ+)
Amino Acid Analysis Ratio
Tyr-(NMe)Ala-Phe-D-Pro-NH2 Tyr-(NMe)Ala-Phe-Val-NH2 Tyr-Ala-Phe-D-Pro-NH2
TLC? Rf 0.30 0.68 0.45
510 512 496
Tyr-D-(NMe)Ala-Phe-D-Pro-~z
0.40
510
Y(1.00) F(0.98)P(1.03) Y(1.00) F(0.96)V(l.00) Y(1.00) F(0.98)A(l.O1) P(1.02) Y(1.00) F(1.00)P(1.01)
Compound
?‘aluesweredetermined in a mixtureof butyl alchol,water,andaceticacid with a ratio of 4:l:l.
methylamide derivatives of amino acids (Ac-X-NHMe) were produced by constraining the angles $ and v to a particular value with a force constant of 100 kcal mol-’ and minimizing the energy with respectto all the other coordinates in the molecule. RESULTS
AND DISCUSSION
The in vitro bioactivities of the tetrapeptides Tyr-X2-Phe-D-Pro-NH2 determined using the guinea pig ileum (GPI) and mouse vas deferens (MVD) assays are summarized in Table II. The GPI and MVD assays were used to assessthe bioactivities at the p- and &receptors, respectively. The receptor affinities were also determined by displacement of relatively selective radioligands from rat brain membrane binding sites (Table III). The compound r3HIDAMG0 served as a selective p-receptor label, and either [3HlDPDPE or [3H]C1-DPDPE was used to determe 6-
Table II. GuineaPig Ileum (GPI) and MouseVas Deferens(MVD) Assays of TetrapeptidesRelatedto Morphiceptin and Dermorphin GPI Compound Tyr-(NMe)Ala-Phe-D-Pro-NH2
1C50/ nM
32.4+ 8.05’ 30.6f 6.1b 213.0f 45.58 Tyr-(NMe)Ala-Phe-Val-NH2 458f 72b Tyr-Ala-Phe-D-Pro-NH2 28100f 7500’ 26900f 7200b Tyr-D-(NMe)Ala-Phe-D-Pro-NH2 73.43f 15.98 85.4f 5.gb Tyr-Pro-Phe-D-Pro-NH* 28.7f 2.4b 20.7f 2.4b Tyr-Pro-(NMe)Phe-D-Pro-NH* Tyr-D-Ala-Phe-D-Pro-NH2 74f 13d
MVD ICsc/ nM
MVD/GPI ICs@ltio
1390Lt14308 202f 2gb 7178f 850’ 4470f 670b Ca
> 1OOOOOb 2046* 391a 603f 147b 1508f 180b 1250f 220b 1490f 18od
27.ga
7.06b 52.5b 60.4b 20.Id
“Valuesweremeasured by theNIDA OpiateCompound TestingProgram. bvdues weremeasured by Dr. P.W. Schillerof theClinicalResearch Instituteof Montreal. ‘A maximumvalueof inhibitionwas26.3% at theconcentrationof 2 x lOaM. dMatthieset al. (1984)Peptides 5,463-470.
Vol.
181, No. 2, 1991
BIOCHEMICAL
AND BIOPHYSICAL
RESEARCH COMMUNICATIONS
Table III. ReceptorBinding Assaysof Morphiceptin and DermorphinTetrapeptides
[3H]DAG0 Kj’/nM
Compound Tyr-(NMe)Ala-Phe-D-Pro-NH2 Tyr-(NMe)Ala-Phe-Val-NH* Tyr-Ala-Phe-D-Pro-NH2 Tyr-D-(NMe)Ala-Phe-D-Pro-NH* Tyr-Pro-(NMe)Phe-D-Pro-NH.r
7
5.5 21.8 loo000 18.1 6.5
[3H]Cl-DPDPE lq/nM 1000OOa 2042.3 > 1ooooo 1308.2 7 1oooooa
‘A value wasmeasuredusing [3H]DPDPE.
receptor affinities. In general, the bioassy data and the receptor binding results agree with each other. As expected, the analog Tyr-Ala-Phe-D-Pro-N& is inactive at both the GPI and MVD tests. The N-methylation of L-Ala* of the inactive analog results in a highly potent analog Tyr(NMe)Ala-Phe-D-Pro-N*, displaying approximately the same activity profile as the parent morphiceptin analog Tyr-Pro-Phe-D-Pro-NH2. The ‘H-NMR studies indicate that the bioactive (NMe)Ala* analog exhibits cis and rrans forms about the amide bond between residues 1 and 2 (29 : 71) similar to the Pro* analog (30 : 70) whereas the inactive Ala* analog adopts only the trim form, The D-(NMe)Ala’ analog is also highly active, displaying the same potency as TyrD-Ala-Phe-D-Pro-NH2 (8). A ratio of 19 : 8 1 was observed for cis and trans configurational isomers about the amide bond between residues 1 and 2 in Tyr-D-(NMe)Ala-Phe-D-Pro-NH*. Since the analogs Tyr-(L and D)-X2-Phe-D-Pro-NH2 [x2 = Ala, (NMe)Ala, and Pro] differ only the second residue, the differences in bioactivities could be interpreted in terms of the conformations and configurations of residue 2. In order to assessaccessible space for the second residues of the above analogs, energy calculations were carried out for six N-acetyl-N’methylamide derivatives [AC-X2-NHMe; X2 = Ala, D-Ala, (NMe)Ala, D-(NMe)Ala, Pro, or DPro]. The derivatives Ac-X2-NHMe display the largest accessible space because they do not have neighboring residues which might restrict conformations for the X2 residues in a long peptide sequence. The backbone conformation of such model compounds is characterized by a set of angles $ and v, respectively representing the rotational states about the skeletal single bonds N-Ca and p-C(O) of the X2 residues, in addition to a cisltrans configuration about the amide bond preceding the X2 residues. The allowed ($, v) space representing 5 kcal mol-’ or less in energy calculated for AcAla-NHMe and the tram isomers of Ac-(NMe)Ala-NHMe and Ac-Pro-NHMe are shown in Figure 1A. All the space accessible to the tram isomers of Ac-(NMe)Ala-NHMe and Ac-ProNHMe are also allowed for Ac-Ala-NHMe. Thus, the inactive analog Tyr-Ala-Phe-D-Pro-NH2 can assume the same overall topochemistry as Tyr-(NMe)Ala-Phe-D-Pro-NH2 and Tyr-Pro-PheD-Pro-NH2 with the amide bond between residues 1 and 2 in a tram configuration. If the bioactive forms of the (NMe)Ala* and Pro* analogs were to have the trans amide bond, the Ala* 667
Vol.
181, No. 2, 1991
BIOCHEMICAL
AND BiOPHYSlCAL
RESEARCH COMMUNICATIONS
120
PO 60 8B
0 -60
-LB0 -120
Weg
-60
0
Weg
60
120
l&l
-iSO -160 -120
-60
0
60
120
Weg
Figure 1. Allowed (I$, v) regions for Ac-X-NHMe; A: Ac-Ala-NHMe (solid), Ac-(NMe)Ala-Name (dashed), andAc-Pro-NHMe (dotted) with tram amide bond I?: Ac-(NMe)Ala-NHMe (solid) and AC-FroNI-IMe (dashed) with cis amide bond C: Ac-D-Ala-NHMe (solid), Ac-D-(NMe)Ala-NHMe (dashed),and Ac-D-Pro-NHMe (dotted) with tram amidebond.
analog would also be active. The biological studies show that the Ala’ analog is totally inactive (Tables II and III). These results indicate that structures for Tyr-(NMe)Ala-Phe-D-Pro-NH* and Tyr-Pro-Phe-D-Pro-NH2 in which a truns amide bond links residues 1 and 2 cannot be responsible for the bioactivities of these molecules. We, therefore, conclude that high p-receptor activities of the (NMe)Ala2 and Pro2 analogs arise from structures in which the amide bonds between residues 1 and 2 are in a cis configuration. To confirm the above postulate the allowed ($, v) regions were calculated for the cis configurational isomers of Ac-(NMe)Ala-NHMe and Ac-Pro-NHMe (Figure 1B). Because both the (NMe)Ala2 and Pro2 opioid analogs display almost the same bioactivities, residue 2 in each compound most likely assumes similar conformations shown in the shaded region. The shaded region, of course, designates conformations common to both analogs and therefore defines common structures which can be recognized at the receptor for both the opioids. We conclude that high p-receptor activities of the morphiceptin analogs are attributed to structures where the second residue assumes conformations in the shaded region in Figure 1B. In addition, a cis configuration is required about the amide bond linking residues 1 and 2. The analogs Tyr-D-Ala-Phe-D-Pro-NH2 and Tyr-D-(NMe)Ala-Phe-D-Pro-NH2 are biologically active while Tyr-D-Pro-Phe-D-Pro-NH2 is inactive. To estimate topochemical requirements of the second residues of the D-Ala2 and D-(NMe)Ala’ analogs for bioactivity, we studied the allowed (I$, v) space for Ac-X2-NHMe [X2 = D-Ala, D-(NMe)Ala, and D-Pro]. Although the D-(NMe)Ala2 analog displays cis and truns isomers about the Tyr-D:(NMe)Ala amide bond, only the trans isomer can assume similar conformations to Tyr-D-Ala-Phe-D-Pro-NH2. The DAla2 analog adopts only a truns amide bond between residues 1 and 2. The shaded (I$, u/) regions in Figure 1C show the conformations accessible to Ac-D-Ala-NHMe and the truns isomer of Ac-D-(NMe)Ala-NHMe but prohibited for the truns isomer of Ac-D-Pro-NHMe. The overlapping region of I$ - -70” and w - 60” cannot explain the “bioactive conformations” since the region is also accessible for the Ac-Ala-NHMe. (Recall that the opioid Tyr-Ala-Phe-D-Pro668
180
Vol.
181, No. 2, 1991
BIOCHEMICAL
AND BIOPHYSICAL
RESEARCH COMMUNICATIONS
Figure 2. The topochemical array for bioactivity of A: Tyr-(NMe)Ala-Phe-D-Pro-NH* (NMe)Ala-Phe-D-Pro-NH% while the Tyr-D-(NMe)Ala
and B: Tyr-DIn the structure A, the Tyr-(NMe)Ala amide bond assumes a cis configuration amide bond assumes a truns configuration in the structure B.
NH2 is inactive.) We must look elsewhere for the structures required for bioactivity. The “bioactive conformations” for Tyr-D-Ala-Phe-D-Pro-NH* and Tyr-D-(NMe)Ala-Phe-D-Pro-NH* arise from the orientation of the second residues in the shaded region where Q - 130’. In addition, these analogs require a truns configuration about the amide bond linking residues 1 and 2 for bioactivity. Thus, in terms of the truns amide linking residues 1 and 2 and the chirality of the second residue, Tyr-D-Ala-Phe-D-Pro-NH* and Tyr-D-(NMe)Ala-Phe-D-Pro-NH2 are defined as dermorphin-like opioids. Structures of the morphiceptin analog Tyr-(NMe)Ala-Phe-D-Pro+@ and the dermorphin analog Tyr-D-(NMe)Ala-Phe-D-Pro-NHz, considered to be closely related to bioactive forms at the p-receptors, are shown in Figures 2A and 2B, respectively. These structures represent the preferred conformations for each molecule estimated by ‘H-NMR and molecular modeling studies (15). The two structures are topochemically almost equivalent. The relative spatial arrangements of the functional groups, i.e. the amine and phenolic groups of Tyr’ and the aromatic ring of Phe3, are almost the same in both the structures. However, the structures of the second residues are different in the two molecules. These results indicate that the chirality change of the (NMe)Ala2 residue can be compensated by 180” rotation about the amide bond between residues 1 and 2, i.e. 0’ about Ty@VMe)Ala (cis amide) and 180’ about Tyr-D(NMe)Ala (truns amide). We conclude that the topochemical similarities of the preferred conformations of the morphiceptin and dermorphin analogs indicate that these two classes of peptide opioids likely interact with the same p-receptor subtype. ACKNOWLEDGMENTS : The authors wish to thank support of this research through grants @A-05539, DA-06254) from the National Institute for Drug Abuse (NIDA). We also gratefully acknowledge Dr. P.W. Schiller of the Clinical ResearchInstitute of Montreal and The NIDA Opiate Compound Testing Program for biological assaysof our analogs. 669
Vol.
181, No. 2, 1991
BIOCHEMICAL
AND BIOPHYSICAL
RESEARCH COMMUNICATIONS
REFERENCES 1. 2. 3. 4.
5. 6. 7. 8.
9. 10. 11. 12. 13. 14. 15.
Chang, K.-J., Killian, A., Hazum, E., Cuatrecasas, P., and Chang, J.-K. (1981) Science 212, 75-77. Rossi, A. C., de Castiglione, R., and Perseo, G. (1986) Peptides 7,755-759. Goodman, M., and Mierke, D. F. (1989) J. Am. Chem. Sot. 111.3489-3496. Chang, K.-J., Wei, E. T., Kiilian, A., and Chang, J.-K. (1983) J Pharmacol. Exp. Tber. 227, 403408. Yoshikawa, M., Tani, F., Yoshimura, T., and Chiba, H. (1986) Agric. Biol. Chem. 50, 2419-2421. Mierke, D. F., Ni$ier, G., Schiller, P. W., and Goodman, M. (1990) Int. J. Peptide Protein Res. 35,35-45. Yamazaki, T., Prijbstl, A., Schiller, P. W., and Goodman, M. (1991) Int. J. Peptide Protein Res. 37,364-381. Matthies, H., Stark, H., Hartrodt, B., Ruethrick, H.-L., Spieler, H.-T., Barth, A., and Neubert, K. (1984) Peptides 5,463-470. Bax, A., and Davis, D. G. (1985) J. Am. Chem. Sot. 107,2820-2821. Bothner-By, A. A., Steppens, R. L., Lee, J., Warren, C. D., and Jeanloz, R. W. (1984) J. Am. Chem. Sot. 106,811-813. Bax, A., and Davis, D. G. (1985) J. Magn. Resonance 65,355-360. Levitt, M. H., Freeman, R., and Frenkiel, T. (1982) J. Magn. Resonance 47,328-330. Bodenhausen, G., Vold, R. L., and Vold, R. R. (1980) J. Magn. Resonance 37,93-106. Hagler, A. T. (1985) The Peptides (Udenfriends, S., Meienhofer, J., and Hruby, V. J., Eds.), Vol. 7, pp. 214-296, Academic Press, Orlando. The details of our conformational studies will be dicussed in a subsequent paper.
670
BIOCHEMICAL
AND BIOPHYSICAL RESEARCH COMMUNICATIONS
Pages 664-670
Topochemical Analysis of Morphiceptin and Dermorphin Bioactivities Toshimasa YAMAZAKI,
Seonggu RO, and Murray GOODMAN*
Department of Chemistry, 0343, University of California, San Diego, La Jolla, CA 92093 Received
October
11,
1991
Summary To elucidate the topochemical requirements for bioactivities
of morphiceptin (TyrPro-Phe-Pro-NH2) and dermorphin (Tyr-D-Ala-Phe-Gly-Tyr-Pro-Ser-NH*), we have designed and synthesized two diastereomers Tyr-(L and D)-(NMe)Ala-Phe-D-Pro-NI$. Both the analogs display high activities in the guinea pig ileum assay. The only difference in the composition of these two diastereomers arises from the chirality at residue 2. The high p-receptor activities are attributed to structures where the Tyr*-L-(NMe)Ala2 amide bond assumes a cis configuration while the Tyrt-D-(NMe)Ala2 amide bond assumes a trans configuration. Accessible space studied for the second residues of these molecules continns the fact that the L(NMe)Ala’ analog belongs to the morphiceptin family of opioids while the D-(NMe)Ala’ analog belongs to the dermorphin class of opioids. The similarity in the spatial array of the analogs explains their high p-receptor activities and indicates that they are likely recognized at the same opioid receptor. 0 1991
Academic
Press,
Inc.
Morphiceptin (Tyr’-Pro2-Phe3-Pro4-NH2) and dermorphii (Tyr*-D-Ala2-Phe3-Gly4-TyrsPro6-Ser7-NH2) are highly p-receptor selective peptide opioids. It is well known that the receptor recognition of these peptides requires specific arrangements of the amine and phenolic groups of Tyr’ and the aromatic group of Phe3. Since the biologically important Tyr’ and Phe3 are joined by a single amino acid, the second residue must play a significant role in orienting these residues in the correct array necessary for bioactivity. The morphiceptin and dermorphin classes of opioids exhibit opposite chiral requirements at residue 2. Morphiceptin requires an L-chirality for Pro2 as indicated by the fact that Tyr-D-Pro-Phe-Pro-NH2 is inactive (1). On the contrary, incorporation of L-amino acids at position 2 of dermorphin results in a remarkable reduction in bioactivity (2). Unlike dermorphin which adopts only the truns amide bond linking residues 1 and 2, morphiceptin exhibits cis and trans isomers about~the Tyr’-Pro’ amide bond in a ratio of 30 : 70 (3). The cis/trms configuration about the Tyr’-Pro2 amide bond is particularly significant for the relative orientation of the biologically important Tyr’ and Phe3 residues. The activities of the morphiceptin analogs incorporating D-Pro and (L and D)-Val at position 4 show that the structure of residue 4 does not affect recognition at the receptor site (4,5). In previous papers (6,7), * To
whom
correspondence
should
CKM-291x/91 $1.50 Copyright 0 1991 by Academic Press, Inc. All rights of reproduction in any form reserved.
be
addressed.
664
Vol.
181,
No.
2, 1991
BIOCHEMICAL
AND
BIOPHYSICAL
RESEARCH
COMMUNICATIONS
we reported on morphiceptin analogs containing 2-aminocyclopentanecarboxylic acid (2-A&) as the second residue in place of proline Tyr-2-A&-Phe-X4-NH2 [X4 = Pro and (L and D)-Val] to eliminate cislfruns isomerization about the amide bond between residues 1 and 2. Among the four stereoisomers, only the analogs containing cis-(lS,2R)-2-A& show bioactivity. Although the ~-A&Z analogs can only adopt a tram amide bond structure about Tyr-2-A&, the bioactive analogs Tyr-cis(lS,2R)-2-A&-Phe-X4-NH2 are topochemically similar to morphiceptin with the Tyr-Pro amide bond in a cis configuration (7). We therefore proposed that a cis configuration about the Tyr’-Pro’ amide bond is required for bioactivity of the morphiceptin related analogs containing a proline as the second residue. To elucidate the configurational requirements of the Tyr’-X2 amide bond (where X2 represents residue 2) and the conformations of the second residue for morphiceptin and dermorphin bioactivities, we synthesized Tyr’-X2-Phe3-D-Pro4-NH2 [X2 = L-(NMe)Ala and D(NMe)Ala]. Since an N-methylalanine is an acyclic mimetic for a proline, the L-(NMe)Ala* and the D-(NMe)Ala* analogs are closely related to morphiceptins; the active Tyr-Pro-Phe-D-ProNH2 and the inactive Tyr-D-Pro-Phe-D-Pro-NH*, respectively. ln addition, Tyr-D-(NMe)AlaPhe-D-Pro-NH2 is also an N-methylated analog of the bioactive D-Ala* dermorphin-like molecule : Tyr-D-Ala-Phe-D-Pro-NH2 (8). In order specifically to explain the topochemical effects of the second residue on bioactivity, we also synthesized Tyr-Ala-Phe-D-Pro-NH2 as a control molecule although we expected that this analog would be inactive. After due consideration of accessible space for the second residues of all these opioids synthesized, we predicted that both the L- and D-(NMe)Ala* analogs would be active at the u-receptor. However, we postulated that the activity of the former molecule is attributed to structures with the Tyr’(NMe)Ala* amide bond in a cis configuration (morphiceptin-like) while a trans structure about the Tyr’-D-(NMe)Ala* amide bond is responsible to the activity of the latter molecule (dermorphin-like). Below we present our experimental evidence to support these structural proposals to explain bioactivity. MATERIALS
AND METHODS
Synthesis The compound Tyr-D-(NMe)Ala-Phe-D-Pro-NH* was synthesized using solid phase techniques. The protected tetrapeptide was assembled on a methylbenzhydrylamine resin employing Boc strategy and the DCC/HOBt coupling method. A low-high HF cleavage was used to cleave the peptide from the resin and to remove protecting groups. The syntheses of the other compounds were carried out in solution using Boc chemistry and EDC/HOBt coupling method. Acidolytic deprotection of protected tetrapeptides with TFA and thioanisole provided crude products. The crude products were purified by affinity chromatography and gel filtration. The purities of the final products were examined by HPLC (> 99 %). The charaterizations of the product are summarized in Table I. Spectroscopy Measurements The ‘H-NMR spectra were recorded at 500 MHz on a General Efectric GN-500 spectrometer. The samples were prepared in DMSO-& at a concentration of 10 mM. Tetramethylsilane was used as an internal reference for the determination of chemical shifts. All of the proton resonances were assigned using two-dimensional HOHAHA (9) and ROESY (10) experiments. The HOHAHA experiments were carried out using the MLEV17 (11, 12) with a mixing time of 100 ms and TPPI (13). The ROESY spectra were acquired using a 200 ms mixing time. Computer Simulations Molecular mechanics calculations were carried out employing the DISCOVER flexible geometry program (14). The (jr, v) energy contour maps.of N-acetyl-N’665
Vol.
181, No. 2, 1991
BIOCHEMICAL
AND BIOPHYSICAL RESEARCH COMMUNICATIONS
TableI. Characteristicsof SynthesizedAnalogs FAB-MS NJ+)
Amino Acid Analysis Ratio
Tyr-(NMe)Ala-Phe-D-Pro-NH2 Tyr-(NMe)Ala-Phe-Val-NH2 Tyr-Ala-Phe-D-Pro-NH2
TLC? Rf 0.30 0.68 0.45
510 512 496
Tyr-D-(NMe)Ala-Phe-D-Pro-~z
0.40
510
Y(1.00) F(0.98)P(1.03) Y(1.00) F(0.96)V(l.00) Y(1.00) F(0.98)A(l.O1) P(1.02) Y(1.00) F(1.00)P(1.01)
Compound
?‘aluesweredetermined in a mixtureof butyl alchol,water,andaceticacid with a ratio of 4:l:l.
methylamide derivatives of amino acids (Ac-X-NHMe) were produced by constraining the angles $ and v to a particular value with a force constant of 100 kcal mol-’ and minimizing the energy with respectto all the other coordinates in the molecule. RESULTS
AND DISCUSSION
The in vitro bioactivities of the tetrapeptides Tyr-X2-Phe-D-Pro-NH2 determined using the guinea pig ileum (GPI) and mouse vas deferens (MVD) assays are summarized in Table II. The GPI and MVD assays were used to assessthe bioactivities at the p- and &receptors, respectively. The receptor affinities were also determined by displacement of relatively selective radioligands from rat brain membrane binding sites (Table III). The compound r3HIDAMG0 served as a selective p-receptor label, and either [3HlDPDPE or [3H]C1-DPDPE was used to determe 6-
Table II. GuineaPig Ileum (GPI) and MouseVas Deferens(MVD) Assays of TetrapeptidesRelatedto Morphiceptin and Dermorphin GPI Compound Tyr-(NMe)Ala-Phe-D-Pro-NH2
1C50/ nM
32.4+ 8.05’ 30.6f 6.1b 213.0f 45.58 Tyr-(NMe)Ala-Phe-Val-NH2 458f 72b Tyr-Ala-Phe-D-Pro-NH2 28100f 7500’ 26900f 7200b Tyr-D-(NMe)Ala-Phe-D-Pro-NH2 73.43f 15.98 85.4f 5.gb Tyr-Pro-Phe-D-Pro-NH* 28.7f 2.4b 20.7f 2.4b Tyr-Pro-(NMe)Phe-D-Pro-NH* Tyr-D-Ala-Phe-D-Pro-NH2 74f 13d
MVD ICsc/ nM
MVD/GPI ICs@ltio
1390Lt14308 202f 2gb 7178f 850’ 4470f 670b Ca
> 1OOOOOb 2046* 391a 603f 147b 1508f 180b 1250f 220b 1490f 18od
27.ga
7.06b 52.5b 60.4b 20.Id
“Valuesweremeasured by theNIDA OpiateCompound TestingProgram. bvdues weremeasured by Dr. P.W. Schillerof theClinicalResearch Instituteof Montreal. ‘A maximumvalueof inhibitionwas26.3% at theconcentrationof 2 x lOaM. dMatthieset al. (1984)Peptides 5,463-470.
Vol.
181, No. 2, 1991
BIOCHEMICAL
AND BIOPHYSICAL
RESEARCH COMMUNICATIONS
Table III. ReceptorBinding Assaysof Morphiceptin and DermorphinTetrapeptides
[3H]DAG0 Kj’/nM
Compound Tyr-(NMe)Ala-Phe-D-Pro-NH2 Tyr-(NMe)Ala-Phe-Val-NH* Tyr-Ala-Phe-D-Pro-NH2 Tyr-D-(NMe)Ala-Phe-D-Pro-NH* Tyr-Pro-(NMe)Phe-D-Pro-NH.r
7
5.5 21.8 loo000 18.1 6.5
[3H]Cl-DPDPE lq/nM 1000OOa 2042.3 > 1ooooo 1308.2 7 1oooooa
‘A value wasmeasuredusing [3H]DPDPE.
receptor affinities. In general, the bioassy data and the receptor binding results agree with each other. As expected, the analog Tyr-Ala-Phe-D-Pro-N& is inactive at both the GPI and MVD tests. The N-methylation of L-Ala* of the inactive analog results in a highly potent analog Tyr(NMe)Ala-Phe-D-Pro-N*, displaying approximately the same activity profile as the parent morphiceptin analog Tyr-Pro-Phe-D-Pro-NH2. The ‘H-NMR studies indicate that the bioactive (NMe)Ala* analog exhibits cis and rrans forms about the amide bond between residues 1 and 2 (29 : 71) similar to the Pro* analog (30 : 70) whereas the inactive Ala* analog adopts only the trim form, The D-(NMe)Ala’ analog is also highly active, displaying the same potency as TyrD-Ala-Phe-D-Pro-NH2 (8). A ratio of 19 : 8 1 was observed for cis and trans configurational isomers about the amide bond between residues 1 and 2 in Tyr-D-(NMe)Ala-Phe-D-Pro-NH*. Since the analogs Tyr-(L and D)-X2-Phe-D-Pro-NH2 [x2 = Ala, (NMe)Ala, and Pro] differ only the second residue, the differences in bioactivities could be interpreted in terms of the conformations and configurations of residue 2. In order to assessaccessible space for the second residues of the above analogs, energy calculations were carried out for six N-acetyl-N’methylamide derivatives [AC-X2-NHMe; X2 = Ala, D-Ala, (NMe)Ala, D-(NMe)Ala, Pro, or DPro]. The derivatives Ac-X2-NHMe display the largest accessible space because they do not have neighboring residues which might restrict conformations for the X2 residues in a long peptide sequence. The backbone conformation of such model compounds is characterized by a set of angles $ and v, respectively representing the rotational states about the skeletal single bonds N-Ca and p-C(O) of the X2 residues, in addition to a cisltrans configuration about the amide bond preceding the X2 residues. The allowed ($, v) space representing 5 kcal mol-’ or less in energy calculated for AcAla-NHMe and the tram isomers of Ac-(NMe)Ala-NHMe and Ac-Pro-NHMe are shown in Figure 1A. All the space accessible to the tram isomers of Ac-(NMe)Ala-NHMe and Ac-ProNHMe are also allowed for Ac-Ala-NHMe. Thus, the inactive analog Tyr-Ala-Phe-D-Pro-NH2 can assume the same overall topochemistry as Tyr-(NMe)Ala-Phe-D-Pro-NH2 and Tyr-Pro-PheD-Pro-NH2 with the amide bond between residues 1 and 2 in a tram configuration. If the bioactive forms of the (NMe)Ala* and Pro* analogs were to have the trans amide bond, the Ala* 667
Vol.
181, No. 2, 1991
BIOCHEMICAL
AND BiOPHYSlCAL
RESEARCH COMMUNICATIONS
120
PO 60 8B
0 -60
-LB0 -120
Weg
-60
0
Weg
60
120
l&l
-iSO -160 -120
-60
0
60
120
Weg
Figure 1. Allowed (I$, v) regions for Ac-X-NHMe; A: Ac-Ala-NHMe (solid), Ac-(NMe)Ala-Name (dashed), andAc-Pro-NHMe (dotted) with tram amide bond I?: Ac-(NMe)Ala-NHMe (solid) and AC-FroNI-IMe (dashed) with cis amide bond C: Ac-D-Ala-NHMe (solid), Ac-D-(NMe)Ala-NHMe (dashed),and Ac-D-Pro-NHMe (dotted) with tram amidebond.
analog would also be active. The biological studies show that the Ala’ analog is totally inactive (Tables II and III). These results indicate that structures for Tyr-(NMe)Ala-Phe-D-Pro-NH* and Tyr-Pro-Phe-D-Pro-NH2 in which a truns amide bond links residues 1 and 2 cannot be responsible for the bioactivities of these molecules. We, therefore, conclude that high p-receptor activities of the (NMe)Ala2 and Pro2 analogs arise from structures in which the amide bonds between residues 1 and 2 are in a cis configuration. To confirm the above postulate the allowed ($, v) regions were calculated for the cis configurational isomers of Ac-(NMe)Ala-NHMe and Ac-Pro-NHMe (Figure 1B). Because both the (NMe)Ala2 and Pro2 opioid analogs display almost the same bioactivities, residue 2 in each compound most likely assumes similar conformations shown in the shaded region. The shaded region, of course, designates conformations common to both analogs and therefore defines common structures which can be recognized at the receptor for both the opioids. We conclude that high p-receptor activities of the morphiceptin analogs are attributed to structures where the second residue assumes conformations in the shaded region in Figure 1B. In addition, a cis configuration is required about the amide bond linking residues 1 and 2. The analogs Tyr-D-Ala-Phe-D-Pro-NH2 and Tyr-D-(NMe)Ala-Phe-D-Pro-NH2 are biologically active while Tyr-D-Pro-Phe-D-Pro-NH2 is inactive. To estimate topochemical requirements of the second residues of the D-Ala2 and D-(NMe)Ala’ analogs for bioactivity, we studied the allowed (I$, v) space for Ac-X2-NHMe [X2 = D-Ala, D-(NMe)Ala, and D-Pro]. Although the D-(NMe)Ala2 analog displays cis and truns isomers about the Tyr-D:(NMe)Ala amide bond, only the trans isomer can assume similar conformations to Tyr-D-Ala-Phe-D-Pro-NH2. The DAla2 analog adopts only a truns amide bond between residues 1 and 2. The shaded (I$, u/) regions in Figure 1C show the conformations accessible to Ac-D-Ala-NHMe and the truns isomer of Ac-D-(NMe)Ala-NHMe but prohibited for the truns isomer of Ac-D-Pro-NHMe. The overlapping region of I$ - -70” and w - 60” cannot explain the “bioactive conformations” since the region is also accessible for the Ac-Ala-NHMe. (Recall that the opioid Tyr-Ala-Phe-D-Pro668
180
Vol.
181, No. 2, 1991
BIOCHEMICAL
AND BIOPHYSICAL
RESEARCH COMMUNICATIONS
Figure 2. The topochemical array for bioactivity of A: Tyr-(NMe)Ala-Phe-D-Pro-NH* (NMe)Ala-Phe-D-Pro-NH% while the Tyr-D-(NMe)Ala
and B: Tyr-DIn the structure A, the Tyr-(NMe)Ala amide bond assumes a cis configuration amide bond assumes a truns configuration in the structure B.
NH2 is inactive.) We must look elsewhere for the structures required for bioactivity. The “bioactive conformations” for Tyr-D-Ala-Phe-D-Pro-NH* and Tyr-D-(NMe)Ala-Phe-D-Pro-NH* arise from the orientation of the second residues in the shaded region where Q - 130’. In addition, these analogs require a truns configuration about the amide bond linking residues 1 and 2 for bioactivity. Thus, in terms of the truns amide linking residues 1 and 2 and the chirality of the second residue, Tyr-D-Ala-Phe-D-Pro-NH* and Tyr-D-(NMe)Ala-Phe-D-Pro-NH2 are defined as dermorphin-like opioids. Structures of the morphiceptin analog Tyr-(NMe)Ala-Phe-D-Pro+@ and the dermorphin analog Tyr-D-(NMe)Ala-Phe-D-Pro-NHz, considered to be closely related to bioactive forms at the p-receptors, are shown in Figures 2A and 2B, respectively. These structures represent the preferred conformations for each molecule estimated by ‘H-NMR and molecular modeling studies (15). The two structures are topochemically almost equivalent. The relative spatial arrangements of the functional groups, i.e. the amine and phenolic groups of Tyr’ and the aromatic ring of Phe3, are almost the same in both the structures. However, the structures of the second residues are different in the two molecules. These results indicate that the chirality change of the (NMe)Ala2 residue can be compensated by 180” rotation about the amide bond between residues 1 and 2, i.e. 0’ about Ty@VMe)Ala (cis amide) and 180’ about Tyr-D(NMe)Ala (truns amide). We conclude that the topochemical similarities of the preferred conformations of the morphiceptin and dermorphin analogs indicate that these two classes of peptide opioids likely interact with the same p-receptor subtype. ACKNOWLEDGMENTS : The authors wish to thank support of this research through grants @A-05539, DA-06254) from the National Institute for Drug Abuse (NIDA). We also gratefully acknowledge Dr. P.W. Schiller of the Clinical ResearchInstitute of Montreal and The NIDA Opiate Compound Testing Program for biological assaysof our analogs. 669
Vol.
181, No. 2, 1991
BIOCHEMICAL
AND BIOPHYSICAL
RESEARCH COMMUNICATIONS
REFERENCES 1. 2. 3. 4.
5. 6. 7. 8.
9. 10. 11. 12. 13. 14. 15.
Chang, K.-J., Killian, A., Hazum, E., Cuatrecasas, P., and Chang, J.-K. (1981) Science 212, 75-77. Rossi, A. C., de Castiglione, R., and Perseo, G. (1986) Peptides 7,755-759. Goodman, M., and Mierke, D. F. (1989) J. Am. Chem. Sot. 111.3489-3496. Chang, K.-J., Wei, E. T., Kiilian, A., and Chang, J.-K. (1983) J Pharmacol. Exp. Tber. 227, 403408. Yoshikawa, M., Tani, F., Yoshimura, T., and Chiba, H. (1986) Agric. Biol. Chem. 50, 2419-2421. Mierke, D. F., Ni$ier, G., Schiller, P. W., and Goodman, M. (1990) Int. J. Peptide Protein Res. 35,35-45. Yamazaki, T., Prijbstl, A., Schiller, P. W., and Goodman, M. (1991) Int. J. Peptide Protein Res. 37,364-381. Matthies, H., Stark, H., Hartrodt, B., Ruethrick, H.-L., Spieler, H.-T., Barth, A., and Neubert, K. (1984) Peptides 5,463-470. Bax, A., and Davis, D. G. (1985) J. Am. Chem. Sot. 107,2820-2821. Bothner-By, A. A., Steppens, R. L., Lee, J., Warren, C. D., and Jeanloz, R. W. (1984) J. Am. Chem. Sot. 106,811-813. Bax, A., and Davis, D. G. (1985) J. Magn. Resonance 65,355-360. Levitt, M. H., Freeman, R., and Frenkiel, T. (1982) J. Magn. Resonance 47,328-330. Bodenhausen, G., Vold, R. L., and Vold, R. R. (1980) J. Magn. Resonance 37,93-106. Hagler, A. T. (1985) The Peptides (Udenfriends, S., Meienhofer, J., and Hruby, V. J., Eds.), Vol. 7, pp. 214-296, Academic Press, Orlando. The details of our conformational studies will be dicussed in a subsequent paper.
670
