On the degradation of dermorphin and D-Arg2-dermorphin analogs by a soluble rat brain extract

Degradation of dermorphin, [D-Arg2]dermorphin and [D-Arg2, Gly3, Phe4]dermorphin in a soluble rat brain extract was examined. The former two heptapeptides were degraded in a similar fashion to produce corresponding N-terminal tetrapeptide as the main degradation product along with the parallel release of Tyr5, Pro6 and Ser7-NH2. Tyr-D-Arg-Phe-Gly showed a good enzymatic stability. When captopril, an angiotensin-converting enzyme inhibitor, was present in the incubation mixture, hydrolysis of the Gly4-Tyr5 bond was markedly suppressed and resulted in release of the corresponding N-terminal hexapeptide as the main degradation product. Combined use of captopril and amastatin, an aminopeptidase inhibitor, markedly suppressed the hydrolysis of these peptides. On the other hand, [D-Arg2, Gly3, Phe4]dermorphin was hydrolyzed easier than the other two heptapeptides and considerable amounts of Tyr1 and Phe4 were released after 20 hr incubation while the N-terminal tetrapeptide, Tyr-D-Arg-Gly-Phe, showed a good enzymatic stability. On the basis of these results, possible degradation pathways of these heptapeptides were discussed.     

Dermorphin-based potential affinity labels for mu-opioid receptors.

Dermorphin and [Lys7]dermorphin, selective micro -opioid receptor ligands originating from amphibian skin, have been modified with various electrophiles in either the 'message' or 'address' sequences as potential peptide-based affinity labels for micro -receptors. Introduction of the electrophilic isothiocyanate and bromoacetamide groups on the para position of Phe3 and Phe5 was accomplished by incorporating Fmoc-Phe(p-NHAlloc) into the peptide followed by selective deprotection and modification. The corresponding amine-containing peptides were also prepared. The pure peptides were evaluated in radioligand binding experiments using Chinese hamster ovary (CHO) cells expressing micro - and delta-opioid receptors. In dermorphin, introduction of the electrophilic groups in the 'message' domain lowered the binding affinity by > 1000-fold; only [Phe(p-NH2)3]dermorphin retained nanomolar affinity for micro -receptors. Modifications in the 'address' region of both dermorphin and [Lys7]dermorphin were relatively well tolerated. In particular, [Phe(p-NH2)5,Lys7]dermorphin showed similar affinity to dermorphin, with almost 2-fold higher selectivity for micro -receptors. [Phe(p-NHCOCH2Br)5]- and [Phe(p-NHCOCH2Br)5,Lys7]dermorphin exhibited relatively high affinity (IC50 = 27.7 and 15.1 nm, respectively) for micro -receptors. However, neither of these peptides inhibited [3H]DAMGO binding in a wash-resistant manner.     

Molecular mechanics studies of dermorphin

Molecular mechanical simulations have been carried out on dermorphin. Presence of D-Ala2 at the N-terminus and L-Pro6 residue at the C-terminus indicated the probability of beta-turns. From the stereochemical considerations, three types- II', III' and V' - for the beta-turn at the N-terminus of the peptide and two types-I and III- for the C-terminus side of the peptide are possible. In our molecular mechanics calculations, we considered six folded and one extended conformations for dermorphin to asses the relative stabilities. Three of the six folded conformations are lower in energy and have the following general feature-similar in energy, three hydrogen bonds, semirigid beta-sheet segment and favorable Tyr1-Tyr5 interaction. The presence of beta-sheet structure might play a role in mu-receptor selective interaction of dermorphin.     

Dehydro-dermorphins. I. Influence of the stereo-orientation of aromatic groups in dermorphin-(1-5)-peptides to opioid activity

Dehydrophenylalanine having the Z-configuration (delta Phe) and D-Phe were incorporated in positions 3 and/or 5 into dermorphin-(1-5)-peptide amide (H-Tyr-D-Ala-Phe-Gly-Tyr-NH2) in order to study the effect of structure or configurational changes. On GPI preparation, whereas the activity of [L-Phe5]-pentapeptide was fourfold higher than parent peptide and comparable to that of dermorphin, the substitution of Phe3 by its D-enantiomer was barely tolerated. The pentapeptides containing delta Phe in positions 3 and/or 5 displayed even lower potency: particularly the unsaturation at position 3 alone or at position 3 and 5 was very detrimental to mu activity.     

Purification of an acid proteinase from Aspergillus saitoi and determination of peptide bond specificity.

The specificity and mode of action of an acid proteinase (EC 3.4.23.6) from Aspergillus saitoi were investigated with oxidized B-chain of insulin, angiotensin II and bradykinin. Further purification of acid proteinase was performed with N,O-dibenzyloxycarbonyl-tyrosine hexamethylene-diamino-Sepharose 4B affinity chromatography and isoelectric focusing. The purified enzyme was free of any other proteolytic activity demonstrated in Asp. saitoi. Acid proteinase from Asp. saitoi hydrolyzed primarily two peptide bonds in the oxidized B-chain of insulin, the Leu(15)-Tyr(16) bond and the Phe(24)-Phe(25) bond. Additional cleavages of the bonds His(10)-Leu(11), Ala(14)-Leu(15) and Tyr(16)-Leu(17) were also noted. Primary splitting sites at Leu(15)-Tyr(16) and Phe(24-)-Phe(25) with acid proteinase from Asp. saitoi were identical with those reported in the work of cathepsin D (EC 3.4.23.5) from human erythrocyte. Hydrolysis of angiotensin II was observed at the Tyr(4)-Ile(5) bond. In conclusion, peptide bonds which have a hydrophobic amino acid such as phenylalanine, tyrosine, leucine and isoleucine in the P'1 position (as defined by Berger and Schechter, [29]) are preferentially cleaved by the trypsinogenactivating acid proteinase from Asp. saitoi.     

Action of human pepsins 1,2,3 and 5 on the oxidized B-chain of insulin.

Human pepsins 1 and 2 attack the B-chain of oxidized insulin at pH 1.7 at the same bonds as does human pepsin 3. At pH 3.5, pepsins 1 and 2 attack insulin B-chain at essentially the same bonds as at pH 1.7, but more slowly. For all three enzymes, the first bond to be hydrolysed is Phe(25)-Tyr(26), followed simultaneously by Glu(13)-Ala(14), Leu(15)-Tyr(16) and Tyr(16)-Leu(17). Human pepsin 5, however, attacks Phe(24)-Phe(25) first of all, followed by Leu(15)-Tyr(16) and Tyr(16)-Leu(17). The results suggest that each pepsin has only one active site. Acid hydrolysis indicates that the sites of enzymic cleavage are not bonds with an inherent instability at low pH.     

14-membered cyclic opioids related to dermorphin and their partially retro-inverso modified analogues. I. Synthesis and biological activity.

As a continuation of our program to study structure-activity relationships of opiate peptides, we report the syntheses and biological activities of a series of 14-membered cyclic dermorphin analogues closely related to enkephalin analogue Tyr-c[D-A2bu-Gly-Phe-Leu] incorporating a phenylalanine at the third position in place of glycine. In addition to two parent dermorphin analogues Tyr-c[D-A2bu-Phe-Phe-(L and D)-Leu], four stereoisomeric retro-inverso modified analogues Tyr-c[D-A2bu-Phe-gPhe-(S and R)-mLeu] with a reversed amide bond between residues four and five, and Tyr-c[D-Glu-Phe-gPhe-(L and D)-rLeu] with two reversed amide bonds between residues four and five, and between residue five and the side chain of residue two have been synthesized. The results from the guinea pig ileum (GPI) and mouse vas deferens (MVD) assays show that all analogues are superactive at either one or both opiate receptors and in general display higher activities as compared to the corresponding enkephalin analogues with a glycine at the third position. Results from the in vitro biological assays and conformational analysis using 1H-NMR spectroscopy (adjoining paper) will provide useful information to understand the role of the Phe3 aromatic side chain in dermorphin, and that of the Phe4 aromatic side chain in enkephalin, on opiate activity since these cyclic dermorphin analogues contain two Phe residues at both the third and fourth positions.     

The specificity of bovine spleen cathepsin S. A comparison with rat liver cathepsins L and B.

The peptide-bond-specificity of bovine spleen cathepsin S in the cleavage of the oxidized insulin B-chain and peptide methylcoumarylamide substrates was investigated and the results are compared with those obtained with rat liver cathepsins L and B. Major cleavage sites in the oxidized insulin B-chain generated by cathepsin S are the bonds Glu13-Ala14, Leu17-Val18 and Phe23-Tyr26; minor cleavage sites are the bonds Asn3-Gln4, Ser9-His10 and Leu15-Tyr16. The bond-specificity of this proteinase is in part similar to the specificities of cathepsin L and cathepsin N. Larger differences are discernible in the reaction with synthetic peptide substrates. Cathepsin S prefers smaller neutral amino acid residues in the subsites S2 and S3, whereas cathepsin L efficiently hydrolyses substrates with bulky hydrophobic residues in the P2 and P3 positions. The results obtained from inhibitor studies differ somewhat from those based on substrates. Z-Phe-Ala-CH2F (where Z- represents benzyloxycarbonyl-) is a very potent time-dependent inhibitor for cathepsin S, and inhibits this proteinase 30 times more efficiently than it does cathepsin L and about 300 times better than it does cathepsin B. By contrast, the peptidylmethanes Z-Val-Phe-CH3 and Z-Phe-Lys(Z)-CH3 inhibit competitively both cathepsin S and cathepsin L in the micromolar range.     

The delta-selective opioid peptide dermenkephalin and the mu-selective hybrid peptide dermenkephalin-[1-4]-dermophin-[5-7] display strikingly different conformations despite identical tetrapeptide N-termini. A quantitative 2-D NMR and molecular modeling analysis

The selective recognition of the aminoterminal binding pharmacophore Tyr-D-Xaa-Phe of the opioid heptapeptide dermorphin, Tyr-D-Ala-Phe-Gly-Tyr-Pro-Ser-NH2 (DRM)1, and of dermenkephalin, Tyr-D-Met-Phe-His-Leu-Met-Asp-NH2 (DREK), by the mu-opioid receptor and delta-opioid receptor, respectively, depends upon the constitution / conformation of the C-terminal tripeptide. The hybrid peptide DREK-[1-4]-DRM-[5-7] is very potent at, and exquisitely selective for the mu-opioid receptor, and differs only from dermenkephalin by its C-terminal tripeptide. Comparison of the structural features of DREK-[1-4]-DRM-[5-7] and dermenkephalin by nmr analysis and molecular modeling revealed striking differences, as well in the trans (Tyr5 - Pro6) isomer (population 75%) than in the cis isomer.. Whereas the folded C-terminal tail of dermenkephalin influenced the tertiary structure of the N-terminal tetrapeptide and placed the Tyr1 and Phe3 aromatic rings in definite orientations that are best suited for the delta-receptor, there were only weak contacts, as shown by NOE data, between the aminoterminal and carboxyterminal parts of the hybrid peptide. This promoted increased flexibility of the whole backbone and relaxed orientations for the side-chains of Tyr1 and Phe3 that are compatible with the mu-receptor but unsuitable for the delta-receptor. The steric hindrance introduced by Pro6 in DREK-[1-4]-DRM-[5-7], plus the absence of large hydrophobic side-chains in positions 5 and 6 may prevent close contacts between the N-terminal and C-terminal domains and reorientation of the main pharmacophoric elements Tyr1 and Phe3.     

Dimeric dermorphin analogues as mu-receptor probes on rat brain membranes. Correlation between central mu-receptor potency and suppression of gastric acid secretion

The opioid receptor preference for dermorphin and several dimerized structural analogues was investigated using rat brain synaptosomes and correlated with the potencies of intracerebroventricularly administered dimeric dermorphin peptides to inhibit gastric acid secretion. The carboxyl terminus of dermorphin or amino-terminal dermorphin analogues was bridged by dihydrazide or (poly)ethylenediamine structures. Synaptosomal membranes were prepared for radioligand binding assay in the presence of soybean trypsin inhibitor and preincubated to remove endogenously bound opioid peptides before storage at -70 degrees C. Specific radiolabeled agonists used in the radioligand binding assays were [D-Ala2,N-methyl-Phe4,Gly-ol5] [3H] enkephalin for mu-receptors and [D-Ala2,D-Leu5] [3H]enkephalin for delta-receptors. delta-Receptor binding assays were conducted in the presence of 2.6 microM [N-Me-Phe3,D-Pro4]morphiceptin to suppress peptide binding to mu-receptors. [D-Ala2,N-methyl-Phe4,Gly-ol5]enkephalin and dermorphin had affinities of 1.39 and 1.22 nM for mu-receptors and 355.8 and 178.6 nM for delta-receptors, respectively. Affinities of dimeric-dermorphin0 for mu- and delta-receptors, and the mu-selectivity ratio, exceeded values characteristic of dermorphin. The dimerized amino-terminal dermorphin analogues are peptides whose receptor binding differed from the parent molecule; e.g. the affinity of dimeric tetrapeptides toward mu-receptors was reduced but was increased for delta-receptors relative to monomeric dermorphin-(1-4)-amide. Dimeric tetradermorphin linked by a bridge containing 12 methylene units (di-tetra-dermorphin12), exhibited a dramatic loss in the mu-selectivity ratio as a result of diminished mu-affinity. On the other hand, substitution of Gly4 by Sar in di-tetra-dermorphin2 enhanced binding to mu-receptors: substitution of D-Arg2 for D-Ala resulted in an increased binding to mu-receptors while decreasing binding to delta-receptors, yielding a peptide with the highest mu-selectivity ratio. These substitutions of D-Arg2 and Sar4 in dimeric amino-terminal dermorphin pentapeptides enhanced binding to both mu- and delta-receptors relative to dermorphin-(1-5)-amide, but led to a decrease in its mu-selectivity ratio. Several dimeric dermorphin analogues exhibited an enhanced mu-selectivity ratio relative to their monomeric analogues. Dimeric peptides, which had a relatively high affinity for mu-receptors, were effective in the suppression of gastric acid secretion.     

Synthesis and biological activities of linear and cyclic enkephalin analogues containing a psi (E,CH = CH) or psi (CH2CH2) isosteric replacement.

The peptide CO-NH function was replaced by a trans carbon-carbon double bond or by a CH2-CH2 isostere in enkephalin analogues of DADLE, DCDCE-NH2 or DPDPE. In DADLE the 2-3 and the 3-4 peptide bond was modified, whereas in the cyclic analogues the Gly3-Phe4 bond was replaced by the isosteres Gly psi (E,CH = CH)Phe [5-amino-2-(phenylmethyl)-3(E)-pentenoic acid] or Gly psi (CH2CH2)Phe [5-amino-2-(phenylmethyl)pentanoic acid]. In general, the modification results in a drop in potency which is the largest for the flexible CH2-CH2 replacement. The Gly3 psi (E,CH = CH)Phe4 DCDCE-NH2 analogue retains considerable potency. These results confirm the importance of the peptide function at the 2-3 and 3-4 position in enkephalin analogues for biological potency.     

Conformational features responsible for the binding of cyclic analogues of enkephalin to opioid receptors. III. Probable binding conformations of mu-agonists with phenylalanine in position 3.

Theoretical conformational analysis was carried out for several tetrapeptide analogues of beta-casomorphin and dermorphin containing a Phe residue in position 3. Sets of low-energy backbone structures of the mu-selective peptides [N-Me-Phe3, D-Pro4]-morphiceptin and Tyr-D-Orn-Phe-Asp-NH2 were obtained. These sets of structures were compared for geometrical similarity between themselves and with the low-energy conformations found for the delta-selective peptide Tyr-D-Cys-Phe-D-Pen-OH and nonactive peptide Tyr-Orn-Phe-Asp-NH2. Two pairs of geometrically similar conformations of mu-selective peptides, sharing no similarity with the conformations of peptides showing low affinity to the mu-receptor, were selected as two alternative models of probable mu-receptor-bound backbone conformations. Both models share geometrical similarity with the low-energy structures of the linear mu-selective peptide Tyr-D-Ala-Phe-Phe-NH2. Putative binding conformations of Tyr1 and Phe3 side chains are also discussed.     

Morphiceptin and beta-casomorphin-5 analogues containing a reduced peptide bond: selective mu-receptor agonists and a novel mu antagonist, H-Tyr-Pro psi (CH2-NH)Phe-Pro-Gly-OH.

In order to prevent enzymatic degradation of beta-casomorphin-5 (1) and morphiceptin, reduced peptide bonds were incorporated at the 2-3 and 3-4 bonds, respectively. The analogues were synthesized by a combination of solid phase methodology and reductive alkylation of resin-bound peptide amines with Boc-amino acid aldehydes (Boc: tert-butyloxycarbonyl) in the presence of NaBH3CN. During reversed phase high pressure liquid chromatography purification, peak shape distortions could be observed. Epimerization was excluded, based on gas chromatography/mass spectroscopy analysis, which indicated acceptable levels of racemization (less than 3%) in the crude product. Instead, the phenomena could be attributed to slow cis/trans isomerizations originating from the Xxx-Pro bonds in the sequence. The presence of different conformational isomers was also established by 1H-nmr spectroscopy in DMSO-d6. All analogues showed high stability in blood plasma, enhanced binding affinity for the mu receptor, and very low binding to the delta receptor. While the Phe 3 psi(CH2-N)Pro4 analogues (3) and (5) displayed agonist activity, the Pro 2 psi(CH2-NH)Phe3 modified analogue (2) showed antagonist activity comparable to D-Phe-Cys-Tyr-D-Trp-Arg-Thr-Pen-Thr-NH2.     

Derivatives of Clostridium acidi-urici ferredoxin containing altered amino acid sequences. Semisynthetic synthesis, biological activity, and stability.

The semisynthetic syntheses and some properties of derivatives of Clostridium acidi-urici ferredoxin that contain amino acid deletions or replacements in the peptide chain are described. All 16 stable derivatives prepared, with the exception of [Trp2]ferredoxin, were fully active as electron carriers in the two enzymatic assay systems tested: the phosphoroclastic system and the ferrodoxin-dependent reduction of cytochrome c. E1Trp1]Ferredoxin had 70% of the activity of native ferredoxin in both assay systems. The stability in aerobic solution of [Ala1]ferredoxin, which had had its natural alanyl NH2-terminal residue removed and then replaced chemically, is the same as that of the native ferrodoxin (half-life of approximately 54 days). The relative stabilities of derivatives with a replacement or deletion of the NH2-terminal residue are as follows: [Ala1]- greater than or equal to [Phe1]-, [Lys1]-, [ Pro1]-, [Leu1]- greater than [Met1]- greater than [Gly1]- greater than [Glu1]- greater than des-Ala1-ferrodoxin. The data indicate that a large bulky residue, but not a negatively charged residue, is tolerated in position 1 of the peptide chain and the greatly decreased stability (half-life = 1 day) of des-Ala1-ferredoxin confirms the importance of the NH2-terminal residue for the stability of the protein. The relative stabilities of derivatives containing Ala1, but including a replacement for the normal Tyr2, are as follows: Native greater than [Trp2]- greater than or equal to [Phe2]- greater than [His2]- greater than [Leu2]- greater than [Pro2]ferredoxin. [Gly2]- and des-Ala1-Tyr2-apoferredoxin did not form stable derivatives upon reconstitution with iron and sulfide, nor did [3-NO2-Tyr2, 30]- and [Leu2,3-NO2-Tyr30]apoferredoxins. Other relatively stable and fully active derivatives prepared included: [3-NH2-Tyr30]-, [3-F-Phe2]-, and [2-F-Phe2]ferredoxin. The behavior of these various derivatives demonstrates the importance of the peptide chain for the stability of C. acidi-urici ferredoxin and shows that the activity of ferredoxin can be altered by a single amino acid substitution in the peptide chain.     

Exogenous adenosine 3': 5'-monophosphate can release yeast from catabolite repression.

A new simple fast and reproducible purification procedure for the proteinase from rat liver mitochondria has been worked out. The specificity of cleavage of peptide bonds in glucagon, oxidized A and B chains of insulin and yeast proteinase B inhibitor by the proteinase of the inner mitochondrial membrane has been studied. The proteinase hydrolyzed three peptide bonds in glucagon, Tyr (13) - Leu (14), Trp (25) - Leu (26) and Phe (22) - Val (23) (minor cleavage site); none in the insulin A chain; one in the B chain of insulin, Tyr (16) - Leu (17); and three in the yeast proteinase B inhibitor, Phe (4) - Ile (5), Phe (20) - Leu (21) and Tyr (41) - Thr (42) (minor cleavage site).Thus, the mitochondrial proteinase cleaves peptide bonds at the carboxyl site of an aromatic amino acid and the amino site of a leucine, isoleucine, threonine or valine. The comparison with chymotrypsin A shows that the mitochondrial proteinase cleaves peptide bonds in a more restricted manner.     

The vegetable rennet of Cynara cardunculus L. contains two proteinases with chymosin and pepsin-like specificities

The flowers of cardoon (genus Cynara) are traditionally used in Portugal for cheese making. In this work the vegetable rennet of the species Cynara cardunculus L. was characterized in terms of enzymic composition and proteolytic specificity of its proteinases (cardosin A and cardosin B). Cardosin A was found to cleave insulin B chain at the bonds Leu15-Tyr16, Leu17-Val18 and Phe25-Tyr26. In addition to the bonds mentioned cardosin B cleaves also Glu13-Ala14, Ala14-Leu15 and Phe24-Phe25 indicating that it has a broader specificity. The kinetic parameters for the hydrolysis of the synthetic peptide Leu-Ser-Phe(NO₂)-Nle-Ala-Leu-oMe were also determined and compared to those of chymosin and pepsin. The results obtained indicate that in terms of specificity and kinetic parameters cardosin A is similar to chymosin whereas cardosin B is similar to pepsin. It appears therefore that the enzyme composition of cardoon rennet closely resembles that of calf rennet.     

Structural requirements for dermorphin opioid receptor binding

Structural features influencing binding activity of dermorphin to opioid receptors have been investigated in the rat brain through the synthesis and evaluation of binding affinity of a series of synthetic dermorphin analogs. Tritiated dermorphin was used as primary ligand. The single population of high affinity dermorphin binding sites present in the rat brain is clearly of an opioid nature since bound radiolabeled dermorphin was fully displaced with high affinity either by morphine or naloxone. Displacement of tritiated dermorphin by all alkaloid opiates or dermorphin related peptides tested was monophasic, consistent with simple competitive inhibition at a single population of binding sites. Dermorphin (Tyr-D-Ala-Phe-Gly-Tyr-Pro-Ser-NH2) was the most potent competitor in all experiments. The D-configuration of the amino acid residue in position 2 was found to be of crucial importance for binding. Replacement of D-Ala2 with L-Ala led to a deleterious effect, this analog being 1/5000th as potent as dermorphin in displacing bound tritiated dermorphin from its receptor. Shorter dermorphin homologs, dermorphin-(1-4)-NH2 and dermorphin-(1-3)-NH2, were found to be 20 and 40-fold less potent, respectively, than dermorphin. The C-terminal carboxamide function is of significant importance for manifestation of the full intrinsic binding potency of dermorphin. Deamidated dermorphin had 1/5th the potency of the parent peptide. This suggests that while the whole dermorphin sequence is required for the expression of the full intrinsic binding activity of the molecule, the N-terminal tripeptide is a key structure as it contains the features which allow receptor recognition.     

Dehydro-dermorphins. II. Synthesis and biological activity of unsaturated dermorphin hexa and heptapeptides

The Phe3 and/or Tyr5 residues in dermorphin (H-Tyr-D-Ala-Phe-Gly-Tyr-Pro-Ser-NH2) and its N-terminal hexapeptide-amide were replaced by delta-Phe or by Phe5 in order to examine the effect on opioid activity. On GPI preparation, the substitution of Phe5 for Tyr5 was well tolerated, whereas the hexa and heptapeptides containing delta Phe in position 3 and/or 5 displayed low potency. The unsaturation at position 3 alone or at positions 3 and 5 was particularly detrimental to mu activity. In the tail flick test, the influence of unsaturation or substitution at positions 3 and 5 generally matched the results of the in vitro assay. Dehydropeptides showed comparatively low antinociceptive effects and [Phe5] analogues displayed about 50% of the analgesic potency of the original peptides.     

N-terminal tetrapeptide of dermorphin and D-Arg-substituted tetrapeptides: inactivation process of the antinociceptive activity by peptidase

Degradation products of the N-terminal tetrapeptide of dermorphin, H-Tyr-D-Ala-Phe-Gly-OH (ALPG) and D-Arg2-substituted tetrapeptide analogs of dermorphin, H-Tyr-D-Arg-Phe-Gly-OH (ARPG), H-Tyr-D-Arg-Phe-Gly-NH2 (TDAPG-NH2) and H-Tyr-D-Arg-Phe-beta-Ala-OH (TDAPA) by enkephalin degrading enzymes were studied by using reversed-phase high-performance liquid chromatography. After 5 and 25 hr incubations of the peptides with solubilized enzymes of mouse brain or spinal cord, liberation of the appreciable Tyr1 residue was observed in ALPG but not in ARPG, TDAPG-NH2 and TDAPA. When ARPG and TDAPG-NH2 were incubated with enzymes for 25 hr, a main degradation product was the N-terminal tripeptide produced from the hydrolysis of Phe3-Gly4 bond. Conversely, TDAPA did not produce the N-terminal tripeptide after 25 hr incubation with enzymes. In the enzyme assay, Tyr1-D-Arg2 bond of ARPG, TDAPG-NH2 and TDAPA was more stable than that of ALPG to the cleavage by aminopeptidase M (AP-M). Phe3-Gly4 bond of ALPG, ARPG and TDAPG-NH2 were easily hydrolyzed by carboxypeptidase Y (CP-Y) within 3 hr incubation, whereas the hydrolysis of Phe3-beta-Ala4 bond of TDAPA by CP-Y was not observed after 3 hr incubation. The present results and previous behavioural data suggest that a potent and prolonged antinociceptive activity of the D-Arg-substituted tetrapeptides is mainly attributed to the stability of Tyr1-D-Arg2 bond against aminopeptidase of peptidases.