Proc. Natl. Acad. Sci. USA Vol. 78, No. 10, pp. 6543-6547, October 1981 Neurobiology

Specific for the : Structure activity relationships (myenteric plexus/opiate/endorphin/multiple opiate receptors/) CHARLES CHAVKIN AND AVRAM GOLDSTEIN Research Foundation and Stanford University, Palo Alto, California 94304 Contributed by Avram Goldstein, July 1, 1981

ABSTRACT The structural features responsible for the high MATERIALS AND METHODS potency and opiate receptor specificity of the dy- norphin in the guinea pig ileum myenteric plexus were examined. The following opioid were obtained from Peninsula Successive removal of'COOH-terminal amino acids from dynor- Laboratories (San Carlos, CA): dynorphin-(1-13), dynorphin- phin-(1-13) demonstrated important contributions of lysine-13, (1-9), dynorphin-(6-13), a-N-acetyldynorphin-(6-13), lysine-11, and arginine-7 to the potency. Removal ofthe NH2-ter- [DAla2]dynorphin-(1-11), [DAla2]dynorphin-(1-13)NH2, a-N- minal tyrosine abolished the biologic activity. Several other struc- methyldynorphin-(1-13)NH2, endo-Gly'-dynorphin-(1-13), and tural modifications were shown to affect potency: substitution of [Arg8]a-neo-endorphin-(1-8). For each of these peptides, pu- D-alanine for glycine-2 reduced the potencies ofdynorphin-(1-13) rity was shown to be >98% by thin-layer chromatography, elec- amide, -(1-11), and -(1-10); and methyl esterification ofthe COOH trophoresis, and amino acid analysis, by the criteria described terminus enhanced the potencies of dynorphin-(1-12), 41-10), (1). [Leu], [DAla2, Leu5]enkephalin, and -(1-9), -(1-8), and -(1-7). Within the dynorphin sequence, lysine- [Leu]enkephalin-Arg6 were obtained from Biosearch (San Ra- 11 and arginine-7 were found to be important for selectivity ofin- fael, CA). Dynorphin-(3-13) and dynorphin-(2-13)-Ser-Asp- teraction with the dynorphin receptor, which is distinguishable Asn-Gln were obtained from Peninsula Laboratories as crude from the ja receptor in this tissue. products after cleavage from the synthesizing resin. These two peptides were purified to homogeneity on C18 reverse-phase The-opioid peptide dynorphin was recently purified from por- high-pressure liquid chromatography (HPLC) as described be- cine pituitary extracts and its sequence was partially deter- low, and composition was confirmed by amino acid analysis. mined. The 13-residue peptide at its NH2 terminus, Tyr-Gly- A series of fragments was prepared with COOH-terminal Gly-Phe-Leu-Arg-Arg-Ile-Arg-Pro-Lys-Leu-Lys, was synthe- amino acids successively removed from dynorphin-(1-13) and sized and shown to contain the pharmacologically active portion from analogues with amino acid substitutions. Carboxypepti- ofthe natural material (1). In the opiate receptor bioassay using dase B (Sigma, type I, diisopropyl fluorophosphate-treated) was the guinea pig ileum myenteric plexus-longitudinal muscle used to prepare dynorphin-(1-12) and dynorphin-(1-11) from preparation, the differences between dynorphin-(1-13) and dynorphin-(1-13) and dynorphin-(1-8) from dynorphin-(1-9). [Leu]enkephalin are particularly interesting because the NH2 Trypsin (Sigma, type XI, diphenylcarbamoyl chloride-treated) terminus of dynorphin corresponds to the [Leu]enkephalin was used to prepare dynorphin-(1-7) and dynorphin-(1-6) from pentapeptide sequence. First, dynorphin-(1-13) is about 700 dynorphin-(1-13), [DA1a2]dynorphin-(1-7) and [DAa2]dynorphin- times more potent than [Leu]enkephalin; and second, the ag- (1-6) from [DAla2]dynorphin-(1-13)NH2, and endo-Gly -d onist effects ofdynorphin-(1-13) are 1/'3th as sensitive to nalox- norphin-(1-6) and endo-Gly'-dynorphin-(1-7) from endo-Gly - one antagonism (1). The higher potency of dynorphin-(1-13) dynorphin-(1-13). Prolyl endopeptidase purified from porcine indicates that the COOH-terminal extension ofthe enkephalin brain by M. Orlowski (6) was used to prepare dynorphin-(1-10) sequence [dynorphin-(6-13)] might provide additional points from and of receptor interaction. The differences in sensitivity to nalox- dynorphin-(1-11) [DAla ]dynorphin-(1-10) from one antagonism suggest that the two agonists have different [DAla2]dynorphin-(1-11). opiate receptor specificity (2); thus, the lower sensitivity ofdy- Peptides from reaction mixtures were purified by repeated norphin-(1-13) to antagonism suggested the existence passage through a C18 reverse-phase HPLC column (0.39 x 30 of a separate dynorphin receptor (1). cm, uBondapak, Waters, 15-35% linear gradient ofacetonitrile Evidence for a separate dynorphin receptor distinct from the in 5 mM trifluoroacetic acid, 1.5 ml/min) with monitoring of (A) and enkephalin (8) receptors has been provided eluate at two wavelengths, 220 and 254 nm, until a single sym- by Wfister et al. (3, 4) who used the mouse vas deferens opiate metrical absorbance peak was obtained. Purity was also con- receptor bioassay. We have recently demonstrated (5) that the firmed in two different thin-layer chromatography systems dynorphin receptor is physically distinct from the ,u receptor (chloroform/methanoVacetic acid/water, 15:10:2:3, Brink- that mediates the effects ofenkephalin and normorphine in the mann Polygram 0.25-mm sil G plates; n-butanoVpyridine/ guinea pig ileum myenteric plexus. acetic acid/water, 21:12:2:15, Eastman cellulose). Concentra- In this paper we present an analysis ofpotency and receptor tions of final products were determined at 274.5 nm by using specificity in a series ofdynorphin analogues in the guinea pig Abbreviations: IC50, concentration for 50% inhibition of guinea pig ileum myenteric plexus-longitudinal muscle bioassay. ileum myenteric plexus-longitudinal muscle contraction; endo-Gly'- dynorphin-(1-13) has glycine inserted after -5 in the dynorphin- The publication costs ofthis article were defrayed in part by page charge (1-13) sequence; HPLC, high-pressure liquid chromatography. payment. This article must therefore be hereby marked "advertise- Throughout this paper, unless otherwise indicated, peptide abbrevia- ment" in accordance with 18 U. S. C. §1734 solely to indicate this fact. tions signify free NH2 and COOH termini. 6543 Downloaded by guest on September 26, 2021 6544 Neurobiology: Chavkin and Goldstein Proc. Nad Acad. Sci. USA 78 (1981)

the tyrosine extinction coefficient, 1340, and confirmed by chain length, however, has the additional effect of introducing molar yields in amino acid analysis. Identities offragments were a negatively charged COOH-terminal acidic group into a novel confirmed by amino acid composition analysis (Dionex analyzer position in the binding site. To eliminate this confounding con- with fluorescence detection) after hydrolysis at 110°C for 24 hr tribution, we removed the negative charge from each of the in 6 M HG1 (sealed tube with argon). Mole ratios were computed peptide fragments by converting the acids to their methyl relative to arginine, with a hydrolysate. of dynorphin(1-13) or esters. [DAla2]dynorphin-(1-13)NH2 prepared and analyzed concur- The IC50 values (and confidence intervals) of the methyl-es- rently, used as standard. terified dynorphin fragments are shown at the left of Table 1. Methyl esters were prepared readily for several fragments, Potency was significantly decreased after the removal oflysine- because the terminal a-COOH is the only acidic group in the 13, lysine-il, and arginine-7. The successive reductions in po- dynorphin-(1-13) sequence. Peptide was dissolved in anhy- tency were about 50%, 25%, and nearly 99%, respectively. drous methanol equilibrated with dry HG1 gas. The reaction Because removal of the other amino acids in the sequence did mixture was fractionated on a C18 reverse-phase HPLC column not significantly affect the potency, this analysis suggests these with the gradient conditions described above. In each case the three amino acids are largely responsible for the potency en- HPLC eluate contained two absorbance peaks, one in the po- hancement, with arginine-7 making the greatest contribution. sition of the parent peptide and a second retained 1-5 min The potencies of the nonesterified dynorphin fragments are longer. This second peak, eluting as expected for the more hy- also shown in Table 1. In this series, removal ofthe COOH-ter- drophobic ester, was purified to homogeneity as before by re- minal amino acid affected potency both by removing the amino peated HPLC. The resulting peptides were shown to be methyl acid functional group and by repositioning the COOH-terminal ester derivatives by saponification; with 90% or greater recov- negative charge. As with the methyl ester dynorphin fragments, ery, the parent (free carboxylic acid) peptide was generated. relative potency was significantly decreased by removal of lys- Potencies ofthe peptides were evaluated by using the guinea ine-13, lysine-li, and arginine-7. Here, however, removal of pig ileum myenteric plexus-longitudinal muscle bioassay as leucine-12 and arginine-6 had the opposite effect of increasing described (7). The concentration of each peptide required to the potency, an effect not seen in the methyl ester series. inhibit the muscle twitch by 50% (IC50) was determined by test- The contribution of the COOH-terminal negative charge to ing at three or more concentrations giving 20-80% inhibition these potency changes can be estimated by comparing the po- and then interpolating ICG5 by log-linear regression analysis. tencies of the COOH-terminal acids and methyl esters of each Agonist effects of all the peptides were reversed by naloxone. peptide. As shown in the last column ofTable 1, methyl ester- Naloxone sensitivity was determined by measuring IC50 in the ification did not significantly change the IC50 of dynorphin- presence and absence of 100 nM naloxone. Peptides were di- (1-13), -(1-11), -(1-6), or -(1-5). However, the potencies ofdy- luted in MeOH/0. 1 M HCl, 1:1 (vol/vol), which, in doses less norphin-(1-12), -(1-10), -(1-9), -(1-8), and -(1-7) were enhanced than 20 ,ul in the 5-ml bath, has no effect on the amplitude of 7- to 13-fold, indicating that the COOH-terminal negative contraction. charge has an adverse effect on potency in these positions. Because we had shown (9) significant adsorptive losses of RESULTS dynorphin-(1-13) but not of [Leu]enkephalin to the glass walls Structural Features Responsible for High Potency. We mea- oftissue baths, we determined the extent ofsuch losses for sev- sured the IC50 values of dynorphin fragments from which eral dynorphin fragments under the bioassay conditions. Dy- COOH-terminal amino acids were sequentially removed, start- norphin peptides terminating in proline-10, lysine-li, leucine- ing with dynorphin-(1-13). It is expected that, when an amino 12, lysine-13, and their methyl esters, and also dynorphin-(1-9), acid critical for receptor binding affinity is removed from the were added to four tissue baths containing muscle strips, each sequence, the potency will decrease. Decreasing the peptide in amount just sufficient to yield its IC50 (Table 1). After 2 min,

Table 1. Potencies of dynorphin peptides COOH-Terminal methyl esters COOH-Terminal free acids Relative Relative ICrO, 95% confidence potency, Potency ICso, 95% confidence potency, IC50 ratio Dynorphin peptide nM interval % change nM interval % (COOH/OMe) YGGFLRRIRPKLK13 0.23 0.16-0.35 129 2.0* 0.33 0.28-0.40 100 1.4 L12 0.57 0.27-1.19 65 0* 5.75 4.16-7.96 9.4 10.1t K11 0.45 0.32-0.64 82 t -0.61 0.46-0.81 64 1.4 P10 1.01 0.71-1.45 19 43 13.2 9.1-19.0 2.2 13.3t R9 1.54 0.99-2.38 23 0.9 14.0 8.57-23.0 1.8 9.lt I8 2.61 1.59-4.28 16 1.4 18.7 10.8-32.2 3.1 7.2t R7 1.94 1.40-2.68 16 1.0 23.3 15.3-35.5 2.0 12.0t R__ 220 140-340 0.18 89t 385 203-733 0.09 1.8 L- 120 65-220 0.27 0.7 182 146-227 0.20 1.5 Dynorphinpeptides were successively truncated fromthe COOH terminus. ICso is the geometric mean computedfrom completelogdose-inhibition curves from 7-48 (mean, 16) separate muscle strips. The 95% confidence interval is calculated from the SEM of the logIC50 values. Relative potency corrects for the variance between muscle strips by comparing the IC6 of nonesterified dynorphin-(1-13) (as 100%) with the ICrO of the givenpeptide determined on the same strip. Potency change is the ratio of relative potencies of the successively shorter dynorphin peptides. Relative potencies and potency changes are not necessarily identical to values calculated from mean IC0 values because they are based on a subset of all the IC50 data. Statistical significance was determined by comparing relative potency values with the Wilcoxon two-sample rank test (8);potency changes and IC50 ratios without a footnote symbol were not statistically significant. Single letter amino acid code: Y, tyrosine; G, glycine; F, ; L, leucine; R, arginine; I, isoleucine; P, proline; K, lysine. *P < 0.05. tp < 0.01. Downloaded by guest on September 26, 2021 Neurobiology: Chavkin and Goldstein Proc. Nad Acad. Sci. USA 78 (1981) 6545

Table 2. Sensitivity of effects of dynorphin peptides to naloxone antagonism COOH-Terminal methyl esters COOH-Terminal free acids Dynorphin Potency 95% confidence Naloxone K., Potency 95% confidence Naloxone K., peptide shift interval nM shift interval nM YGGFLRRERPKLK'3 4.4 3.5-5.5 29 5.0 4.0-6.1 25 L12 4.5 3.3-6.1 29 4.8 3.8-6.1 26 K"1 4.4 2.7-7.1 29 4.3 1.4-13 30 P'0 9.0 3.5-23 13 6.8 5.2-9.0 17 R9 7.1 5.1-9.7 16 6.8 5.0-42 17 I8 8.5 4.8-15 13 14 5.5-42 7.7 R7 10 7.7-14 11 7.5 5.5-10 15 R6 19 11-32 5.6 *26 19-35 4.0 L- 34 18-130 3.0 -34 26-45 3.0 Potency shift is the antilogarithm of the mean of the differences in log ICrO in the presence and absence of 100 nM naloxone. Each value is based on 3-22 (mean, 7) independent determinations. The 95% confidence intervals of the means are calculated as in Table 1. The apparent naloxone dissociation constant (K.) is based on the questionable assumption that, in every case, the agonist is acting through a single receptor; these values are presented only for comparison with published data of others. K. is computed from the equation Ke = C/(DR - 1), derived from the mass law for competitive antagonism at a single ho- mogeneous population of receptors, in which concentration (C) of naloxone is 100 nM, and dose ratio (DR) is the ratio of agonist ICs0 concentrations in the presence and absence of the antagonist (11).

samples of bath fluid were removed for radioimmunoassay in tency after removal of arginine-7 and the nearly equal relative triplicate (10). Recoveries were >70% for all the peptides ex- potencies of dynorphin-(1-5)OMe and dynorphin-(1-6)OMe is cept dynorphin-(1-13), mean (±SEM) recovery of which was that arginine-7, but not arginine-6, is critical for potency. An 50 ± 10%. In general, therefore, our potency estimates are not alternative explanation is that the receptor recognizes a pair of much affected by adsorptive or degradative losses in the bath. basic residues in this position. The latter interpretation is un- However, the true potency ofdynorphin-(1-13) may be as much likely because substitution of glycine for arginine-6 did not af- as twice that indicated in Table 1. fect the potency of dynorphin-(1-7) (Table 3). To determine the contribution ofthe NH2-terminal residues Comparison ofthe potencies ofthe three structurally related to potency, four dynorphin analogues were prepared: dynor- octapeptides dynorphin-(1-8), endo-Gly'-dynorphin-(1-7), and phin-(3-13), dynorphin-(2-13)-Ser-Asp-Asn-Gln,* dynorphin- [Arg8]a-neo-endorphin-(1-8) (Table 3) demonstrates the follow- (6-13), and a-N-acetyldynorphin-(6-13). At concentrations up ing: (i) The identity ofthe residue in the eighth position of the to 10 ,uM, none of the four inhibited the muscle twitch; and dynorphin sequence is not critical; substitution of arginine for none was able to antagonize the agonist effects of 0.4 nM dy- isoleucine-8 does not affect potency because endo-Gly5a-dy- norphin-(1-13) in the single-dose method ofKosterlitz and Watt norphin-(1-7) is nearly equipotent with dynorphin-(1-8). (ii) (11). Furthermore, the simultaneous presence ofeither 10 ,uM The basic charge on the residue in the seventh position in the dynorphin-(6-13) or 10 LM a-N-acetyldynorphin-(6-13) did dynorphin sequence is probably the important structural fea- not enhance or antagonize the potency of [Leu]enkephalin or ture because arginine-7 in that sequence, shown to be critical normorphine. for potency, can be replaced by lysine without loss ofpotency. Structural Features Responsible for Low Sensitivity to An- Furthermore, this analysis suggests that lysine-7 is responsible tagonism by Naloxone. Naloxone antagonized [Leu]enkephalin for the 27-fold higher potency of [Arge]a-neo-endorphin-(1-8) in this bioassay at very low concentration (Ke = 3 nM), whereas relative to [Leu]enkephalin reported by Kangawa et aL (12). a much higher concentration (K& = 29 nM) was required to an- One ofthe strategies successfully used to prepare analogues tagonize dynorphin-(1-13) (Table 2). We used this characteristic of opioid peptides resistant to aminopeptidase is the substitu- to determine which amino acids in the dynorphin-(6-13) se- tion ofD-alanine-2 for glycine-2 (13). As shown by the IC50ratios quence are responsible for the difference in receptor specificity in Table 4, D-alanine-2 substitution reduces the potencies of of dynorphin-(1-13) and [Leu]enkephalin. The ratios of IC,% dynorphin-(1-13) and dynorphin-(1-11), leaves the potencies values in the presence and absence of100 nM naloxone (potency of dynorphin-(1-10) and dynorphin-(1-7) essentially un- shift) are presented in Table 2. changed, and dramatically enhances the potencies of dynor- For the methyl esters, naloxone sensitivity was not affected phin-(1-6) and [Leu]enkephalin. D-Alanine-2 substitution also by removal of lysine-13 or leucine-12 but. was increased 2-fold by removal oflysine-li. It was unaffected by removal ofproline- 10, arginine-9, or isoleucine-8 but was increased 3-fold by re- Table 3. Potencies of dynorphin analogues with amino acid moval ofarginine-7. Thus, two residues, lysine-il and arginine- substitutions 7, are important for reducing sensitivity to naloxone antago- 95% nism, and these are also the most critical for enhancement of IC50, confidence potency. The changes in naloxone sensitivity of the nonesteri- Peptide Sequence nM interval fied dynorphin fragments were similar to those seen with the Dynorphin-(1-7) YGGFLRR 23.3 15.3-35.5 methyl ester fragments. Methyl esterification thus seems to Dynorphin-(1-8) YGGFLRRI 18.7 10.8-32.2 affect potency but not receptor specificity. Endo-Gly6a-dynorphin-(1-6) YGGFLGR 33.9 15.4-75 Effects ofAmino Acid Substitutions on Potency and Recep- Endo-Gly5--dynorphin-(1-7) YGGFLGRR 33.0 15.1-72 tor Specificity. One interpretation of the large change in po- [Arg8]a-neo-endorphin-(1-8) YGGFLRKR 10.4 6.3-17.1 The IC50 and 95% confidence intervals are based on 4-12 determi- * This COOH-terminal extension was one of several studied; addition nations as in Table 1. Data for dynorphin-(1-7) and dynorphin-(1-8) of these four residues does not significantly affect potency. are repeated from Table 1 to facilitate comparison. Downloaded by guest on September 26, 2021 6546 Neurobiology: Chavkin and Goldstein Proc. Natl.'Acad. Sci. USA 78 (1981) Table 4. Potencies and naloxone sensitivities of dynorphin analogues designed to be resistant to aminopeptidase IC50, 95% confidence ICW ratio Potency 95% confidence Naloxone K., Peptide nM interval (Gly2/Ala2) shift interval nM a-N-Me-Dynorphin-(1-13)NH2 0.42 0.30-0.57 - 4.3 3.3-6.5 30.5 [DAIa2]Dynorphin'(1-13)NH2 1.94 1.16-3.26 0.12 8.2 4.6-12 12.7 -(1-11) 2.91 2.06-4.09 0.21 9.1 5.7-13 12.5 -(1-10) 14.4 6.79-30.4 0.92 - -- -(1-7) 13.2 7.67-22.6 1.8 23 20-26 4.7 -(1-6) 23.6 5.9-94.3 16.3 - -- -(1-5) 8.6 1.01-30.4 21.2 44 34-55 2.4 IC50 is the geometric mean of 3-18 determinations (mean, 9). IC5Y ratio compares the mean IC50 for the dynorphin peptide with glycine-2 with the IC5o of the peptide with alanine-2 except that the ICs0 of [DAla ldynorphin-(1-13)NH2 is compared with the IC50 of dynorphin-(1-13) methyl ester. Potency shift and naloxone K. are defined in Table 2 and are based on 4-14 determinations (mean, 7). In the series of the first was an amide. [DAla2]peptides, only

affects the naloxone sensitivity: dynorphin-(1-13) methyl ester, group and a phenolic ring for binding to other opiate receptors dynorphin-(1-11), and dynorphin-(1-7) become more sensitive (16). The inefficacy of dynorphin-(6-13) as compared with dy- to naloxone antagonism but [Leu]enkephalin does not. norphin-(1-13) also indicates that 'a major part of the binding A second strategy used to prepare analogues resistant to energy of the latter is attributable to its NH2-terminal portion. aminopeptidase is NH2-terminal methylation. Dutta et aL (14) Our interpretations ofpotency changes are subject to the lim- demonstrated that a-N-methyl-[Leu]enkephalin potency is 4 itation that, ifpeptides have to diffuse to the receptor through times greater than that of [Leu]enkephalin. As shown in Table tissue containing active peptidases, the relative potencies of 4, a-N-methyldynorphin-(1-13)NH2 was found to be equipo- different peptides could be affected by differing susceptibilities tent with dynorphin-(1-13) methyl ester (cf. Table 1) and to degradation within the tissue. equally sensitive to naloxone antagonism. The naloxone sensitivities ofthe dynorphin fragments divide into three groups: low (K, 25-30 nM), intermediate (Kin 11-17 DISCUSSION nM), and high (K, 3-5 nM). Transition between these groups The 8 and u opiate receptors have been characterized by en- within the dynorphin fragment series follows the removal of kephalin structure-activity studies in guinea pig ileum, mouse lysine-li and arginine-7, respectively. Because these residues vas deferens, and brain membrane binding assays (15). We have are also responsible for enhancement of potency, we propose that the previously shown that the dynorphin receptor and the tt re- dynorphin receptor contains complementary binding ceptor in the guinea pig ileum are physically distinct (5) al- sites for these residues, which the pi receptor lacks. though the dynorphin receptor may be identical to the K re- A diagram summarizing the proposed interactions between ceptor (unpublished data). Here we show which amino acids in dynorphin and its receptor is presented in Fig. 1. The portion the dynorphin-(1-13) sequence are responsible for its high po- of the receptor interacting with the dynorphin-(1-4) sequence tency and its receptor specificity. (or with opiate alkaloids) is designated the tetrapeptide ("mes- The potencies ofdynorphin fragments successively truncated sage") pocket. We regard Tyr-Gly-Gly-Phe as the opioid "mes- from the COOH terminus demonstrate that three amino acid sage" [in the sense introduced by Schwyzer (17)] because it is residues in the dynorphin-(6-13) sequence contribute to the the shortest fragment with typical naloxone-reversible opioid high potency: lysine-13, lysine-il, and arginine-7 (in order of activity, Tyr-Gly-Gly being inactive (18, 19). The portion ofthe increasing importance). We postulate that the dynorphin re- receptor interacting with the dynorphin-(5-13) sequence is des- ceptor contains complementary anionic sites able to bind the ignated the potency-enhancing ("address") domain. It is evi- peptide through ionic interactions at these three critical posi- dently responsible for the specificity and high potency of tions. The importance of the precise location of these critical dynorphin. groups within the sequence has been demonstrated by the de- crease in potency accompanied by glycine insertion between leucine-5 and arginine-6 and by deletion of arginine-6; both "register shift" changes decrease potency about 90% (1). Our conclusion that the other amino acids in the dynorphin- (6-13) sequence contribute little to potency is subject to the reservation that our method only examines successively trun- cated fragments. In a sequence A-B-C-D, for example, if the potency ranking is A-B-C-D >> A-B-C = A-B, we would con- clude that D makes a critically important contribution to po- tency, whereas C does not. Yet C nevertheless might be es- sential to the high potency of A-B-C-D-for example, by affecting the conformation of A-B-C-D in the receptor binding site. The true contribution ofan internal residue, therefore, can only be tested by amino acid replacement without shortening the peptide backbone. In our study, such single amino acid sub- FIG. 1. Diagram of the dynorphin receptor, illustrating putative stitutions were only made at positions 2, 6, 7, and 8. sites of peptide-receptor interaction. Complementary groups are pos- The dynorphin receptor also binds the peptide through in- tulated for binding the a-ammonium and phenolic OH of tyrosine-l, teraction with the is guanidinium of arginine-7, and e-ammonium of lysine-11 and lysine- NH2 terminus, where tyrosine-1 absolutely 13. It is not known if leucine-5 is critical for potency at the dynorphin essential, as shown by our results for des-tyrosine analogues. receptor, although it certainly is at the Au and 8 receptors. See text for This is consistent with the known requirement for a basic amino further explanation. Downloaded by guest on September 26, 2021 Neurobiology: Chavkin and Goldstein Proc. NatL Acad. Sci. USA 78 (1981) 6547

The tetrapeptide pocket of the dynorphin receptor seems to ofboth [DAla2,DLeu5]enkephalin and did not reduce be different from that of the u receptor. This is evidenced by the potency ofdynorphin-(1-13). Moreover, rats made tolerant the differing affinities for naloxone and by the effect ofD-Alan- to the cataleptic effect of sufentanil, a jL agonist, remain sen- ine-2 on receptor specificity. In the myenteric plexus- sitive to the cataleptic effect of dynorphin-(1-13) given by in- longitudinal muscle preparation, D-alanine-2 substitution in- tracebroventricular injection (25). creases the potency of short NH2-terminal fragments of dynorphin (e.g., [Leu]enkephalin). However, it decreases that We thank Dr. Marian Orlowski for generously supplying prolyl en- amide], and such dopeptidase, and Drs. Brian M. Cox, Frances Leslie, Vartan Ghaza- of longer fragments [e.g., dynorphin-(1-13) rossian, and Carmelo Romano for helpful comments. The work was sup- decreased potency is associated with a distinct increase in sen- ported by Grants DA-1199 and DA-7063 from the National Institute on sitivity to naloxone antagonism. Drug Abuse and by a generous gift from David Fasken. A change in receptor specificity following D-alanine-2 sub- stitution in dynorphin-(1-13) had been reported in the mouse 1. Goldstein, A., Tachibana, S., Lowney, L. I., Hunkapiller, M. & vas deferens opiate receptor bioassay (4). Moreover, in a rat Hood, L. (1979) Proc. NatL Acad. Sci. USA 76, 6666-6670. behavioral assay, Herman et at (20) have shown that the cata- 2. Lord, J. A. H., Waterfield, A. A., Hughes, J. & Kosterlitz, H. W. (1977) Nature (London) 267, 495-499. leptic effect of [D-Ala2]dynorphin-(1-11) is much more readily 3. Wuster, M., Schulz, R. & Herz, A. (1980) Eur.J. PharmacoL 62, reversed by naloxone than is the cataleptic effect ofdynorphin- 235-236. (1-13). A practical consequence ofthese findings is that D-alan- 4. Wister, M., Schulz, R. & Herz, A. (1980) Neurosci. Lett. 20, ine-2 substitutions can not be used to stabilize dynorphin pep- 79-83. tides without destroying the specificity of interaction with the 5. Chavkin, C. & Goldstein, A. (1981) Nature (London) 291, dynorphin receptor. 591-593. 6. Orlowski, M., Wilk, E., Pearce, S. & Wilk, S. (1979) J. Neuro- In the "address" sequence, it is of interest that arginine-7, chem. 33, 461-470. which makes the greatest contribution to potency, is the second 7. Goldstein, A. & Schulz, R. (1973) Br. J. Pharmacot 48, 655-666. member of a basic pair. Such basic pairs often represent cleav- 8. Goldstein, A. (1968) Biostatistics: An Introductory Text (Mac- age signals for conversion ofa prohormone to an active millan, New York). (21). Instances are known, however, in which a basic pair signals 9. Ho, W. K. K., Cox, B. M., Chavkin, C. & Goldstein, A. (1980) phosphorylation or pyridoxylation rather than cleavage (22, 23), 1, 143-152. 10. Ghazarossian, V. E., Chavkin, C. & Goldstein, A. (1980) Life Sci. and basic pairs may also occur in proteins not destined for por- 27, 75-86. cessing at all (24). It remains to be determined if, in addition 11. Kosterlitz, H. W. & Watt, A. J. (1968) Br. J. Pharmacol 53, to the special role ofarginine-7 in receptor interaction, the basic 371-381. pair Arg6-Arg7 also serves as a cleavage signal, possibly to ter- 12. Kangawa, K., Matsuo, H. & Igarashi, M. (1979) Biochem. Bio- minate the action of dynorphin at its receptor. phys. Res. Commun. 86, 153-160. The following lines of evidence support the idea that there 13. Pert, C., Pert, A., Chang, J.-K. & Fong, B. T. W. (1976) Science 194, 330-332. is a distinct dynorphin-type opiate receptor in the guinea pig 14. Dutta, A. S., Gormley, J. J., Hayward, C. F., Morley, J. S., Sta- myenteric plexus. cey, G. J. & Turnbull, M. T. (1977) Acta Pharm. Suec. SuppL 14, (i) Specific basic amino acid residues, exactly positioned, ac- 14-15. count for the greatly enhanced potency of dynorphin-(1-13) as 15. Morley, J. S. (1980) Annu. Rev. PharmacoL ToxicoL 20, 81-110. compared with [Leu]enkephalin. 16. Goldstein, A., Goldstein, J. S. & Cox, B. M. (1975) Life Sci. 17, (ii) The same features responsible for the very high potency 1643-1654. 17. Schwyzer, R. (1977) Ann. N.Y. Acad. Sci. 297, 3-26. of dynorphin-(1-13) also account for its relative insensitivity, 18. Morgan, B. A., Smith, C. F. C., Waterfield, A. A., Hughes, J. compared with [Leu]enkephalin, to naloxone antagonism. & Kosterlitz, H. W. (1976) J. Pharm. PharmacoL 28, 660-661. (iii) D-Alanine-2 substitution, which enhances the potency 19. Ling, N. & Guillemin, R. (1976) Proc. Nate Acad. Sci. USA 73, of[Leu]enkephalin and of the shorter dynorphin fragments, has 3308-3310. an opposite effect on the longer and more potent dynorphin 20. Herman, B. H., Leslie, F. H. & Goldstein, A. (1980) Life Sci. 27, fragments. D-Alanine-2 in the longer dynorphin fragments also 883-892. 21. Steiner, D. F., Patzelt, C., Chan, S. J., Quinn, P. S., Tager, H. makes them more sensitive to naloxone antagonism. These find- S., Nielsen, D., Lernmark, A., Noyes, B. E., Agarwal, K. L., ings suggest that the binding site for the NH2-terminal se- Gabbay, K. H. & Rubenstein, A. H. (1980) Proc. R. Soc. London quence Tyr-Gly-Gly-Phe-Leu in dynorphin may be different Ser. B 210, 45-59. from the site to which the same pentapeptide ([Leu]enkephalin) 22. Krebs, E. G. & Beavo, J. A. (1979) Annu. Rev. Biochern. 48, binds when it has no COOH-terminal extension. 923-959. (iv) As we showed elsewhere by selective protection exper- 23. Parsons, T. F. & Preiss, J. (1978)J. BioL Chem. 253, 7638-7645. 24. Dayhoff, M. O., Hunt, L. T., Barker, W. C., Schwartz, R. M. iments, the receptor sites for dynorphin-(1-13) and for & Orcutt, B. C. (1978) Protein Segment Dictionary (National [Leu]enkephalin are physically distinct (5). Biomedical Research Foundation, Washington, D.C.). (v) Tolerance to a and A opiate agonists in the mouse vas 25. Herman, B. H. & Goldstein, A. (1981) Soc. Neurosci. Abstr., in deferens produced by Wuster et al (3, 4) by chronic infusion press. Downloaded by guest on September 26, 2021