Peptides,Vol. 13, pp. 1207-1213, 1992

0196-9781/92 $5.00 + .00 1992 PergamonPressLtd.

Printed in the USA.

Interaction of Opioid Peptides and Other Drugs With Multiple 6ncx Binding Sites in Rat Brain: Further Evidence for Heterogeneity H E N G X U , * J O H N S. P A R T I L L A , * B R I A N R. DE COSTA,~" K E N N E R C. R I C E r A N D R I C H A R D B. R O T H M A N .1

*Clinical Psychopharmacology Section, NIDA Addiction Research Center, PO Box 5180, Baltimore, AID 21224 and -~Laboratory of Medicinal Chemistry, NIDDK, NIH, Bethesda, MD 20892 R e c e i v e d 8 J u n e 1992 XU, H., J. S. PARTILLA, B. R. DE COSTA, K. C. RICE AND R. B. ROTHMAN. Interaction ofopioidpeptides and otherdrugs with multiple 6,~ binding sites in rat brain: Further evidencefor heterogeneity. PEPTIDES 13(6) 1207-1213, 1992.--Recent pharmacological data strongly support the hypothesis of 6 receptor subtypes as mediators of both supraspinal and spinal antinociception (rt and 52 receptors). In vitro ligand binding data, which are fully supportive of the in vivo data, are still lacking. A previous study indicated that [3H][D-Ala2,D-LeuS]enkephalinlabels two binding sites in membranes depleted of ~t binding sites by pretreatment with the site-directed acylating agent, 2-(p-ethoxybenzyl)-l-diethylaminoethyl-5-isothiocyanatobenzimidazoleHCI (BIT). The main goal of the present study was to develop a ligand-selectivity profile of the two ~n~ binding sites. The data indicated that naltrindole and oxymorphindole were relatively selective for site 1 (20-fold). [D-Ser2,Thrr]Enkephalinand deltorphinII were only 2.7-fold and 2.2-fold selective for site 1. [O-Pen2,D-PenS]Enkephalin and deltorphin-I were 80-fold and 38-fold selective for site 2.3-1odo-Tyr-l>-Ala-Gly-Phe-o-Leu was 52-fold selective for site 1. Morphine had moderate affinity for site 1 (K~ = 16 nM), and was about I l-fold selective for site 1. Thus, of the 10 drugs studied, only DPDPE and DELT-I were selective for site 2. Viewed collectively with other data, it is likely that the 6~ receptor and the 6,~ binding site are synonymous. Opioid peptides

6,cx binding sites

Drug interaction

SEVERAL independent lines of evidence support the hypothesis of an opioid receptor complex composed of distinct, yet interacting, ~,, 6, and perhaps r binding sites (6,27,28). This model postulates two classes of fi binding sites: a 6 binding site not associated with the opioid receptor complex, termed the fin¢~site (the ncx stands for not in the complex), and a 6 site associated with the receptor complex, termed the 6¢x site (the cx stands for in the complex). The 6.~ site has high affinity for ligands such as [D-PenE,D-PenS]enkephalin (DPDPE) (13), low affinity for morphine, and is synonymous with what is commonly identified as the 6 binding site. In contrast, the ft, site has high affinity for /~ ligands, low affinity for ~ ligands, and is optimally labeled with [3H][D-Ala2,LeuS]enkephalin. Due to its low affinity for the ~¢x site (K~ of about 1000 nM), [3H]DPDPE cannot be used to label the 6¢x site. The ~ncxsite is thought to mediate the direct antinociceptive effects of DPDPE, an effect that is selectively reversed by the 6selective irreversible antagonist, [D-Ala2,LeuS,Cys6]enkephalin (DALCE) (7). In contrast, the 6cx site is thought to mediate the modulatory effects of subantinociceptive doses of enkephalinrelated peptides on morphine-induced antinociception, an effect that is blocked by the irreversible antagonist, 5'-naltrindole-isothiocyanate (5'-NTII) (27). Requests for reprints should be addressed to Richard B. Rothman. 1207

Other data, which support the hypothesis of 6 receptor subtypes, come from work demonstrating that certain enkephalin analogs (2,31) and B-casomorphan analogs (1) are less potent in the mouse vas deferens (MVD) bioassay than would be predicted on the basis of their affinity for the rat brain 6 receptor. Vaughn et al. (32) further demonstrated, using [3H][o-PenEpC1-Phe4,o-PenS]enkephalin ([3H]-pCIDPDPE) to label 5 binding sites of the M V D and rat brain, that [o-Ala2,(2R,3S)-AEPhe4LeuS]enkephalin had 33-fold lower affinity for the ~ binding sites of the MVD than for rat brain ~ binding sites. Porreca and associates have recently provided compelling in vivo evidence for 6 receptor subtypes as mediators of both supraspinal and spinal antinociception (8,11). Their work was greatly facilitated by the development of selective 6 receptor agonists such as deltorphin-II (DELT-II) (5) and DPDPE (13), selective reversible ~ antagonists such as naltrindole (17), as well as the selective irreversible ~ antagonists 5'-NTII (18) and DALCE (3,7). Based on their findings, Porreca and colleagues tentatively classified ~ receptors into two subtypes: the 61 receptor, which is selectively activated by DPDPE and blocked by DALCE, and the t52 receptor, which is selectively activated by DELT-II and blocked by 5'-NTII.

1208

XU ET AL. TABLE 1 EFFECT OF BIT AND DIGIT ON DISPLACEMENTOF [3H][D-Ala2,D-Leu~]ENKEPHALINBINDING TO THE ~¢~ BINDING SITE BY DPDPE AND DELT-I CNT/CNT-P2*

CNT/BIT-P2

CNT/DIGIT-P2

Drug

ICs0

Slope Factor

IC~o

SlopeFactor

IC5o

SlopeFactor

DPDPE DELT-I

8.9+0.3 9.6 + 0.4

0.88 +0.03 0.78 + 0.02

13.1 _+ 1.8~" 13.6 -+ 2.3?

0.71 + 0.06t 0.68 --- 0.07

12.7 -+ 1.1t 9.65 _+ 1.2

0.82 +0.05 0.68 + 0.05

CNT/CNT-P2, CNT/BIT-P2, and CNT/DIGIT-P2 membranes were prepared from a common pool of lysed-P2 membranes as described in the Method section. Nine point DPDPE and DELT-I inhibition curves were generated by displacing 2.9 nM [3H][D-Ala2,D-LeuS]enkephalin.The data of two experiments were pooled, and fit to the two-parameter logistic equation for the best-fit estimates of the IC5o(nM _+SD) and slope factor (N _+SD) reported above. * LY164929 (100 nM) was present to block binding to the ~x site ?p < 0.01 when compared to CNT/ CNT-P2 membranes, as determined with the F-test.

Viewed collectively, it is apparent that two different schema of ~ receptor subtypes have developed: an earlier model, based on studies of the opioid receptor complex, and a more recent model, which was developed with recently available novel agonists and antagonists. A point of some interest is the relationship between the ~n~ and ~cxbinding sites, and the ~ and ~2 receptors. Observations that DALCE selectively inhibits the ~ncx binding site (24), and blocks the effect of DPDPE at the ~cx receptor (7), suggest that the ~ and ~cx binding sites may be synonymous. The ~ receptors are selectively activated by deltorphin-II, and selectively blocked by the irreversible ~ antagonist, 5'-NTII (8). Doses of 5'-NTII that block deltorphin-II effects also block modulation of morphine antinociception by ~ ligands (16,27), an effect thought to be mediated via the bCx binding site (6,27). Thus, these data suggest that the ~x site and the ~2 receptor might also be synonymous. An earlier study reported initial data supporting the existence of subtypes of the 5,~xbinding site distinguished by DPDPE and deltorphin-I (DELT-I) (34). The goal of the present study was to generate a more detailed ligand-selectivity profile of the ~nCx subtypes, and to see if it correlated with the pharmacological profile of the ~ and 62 receptors. The data presented here provide additional evidence that supports the existence of subtypes of the ~c~ binding site. METHOD

Preparation of Membranes Membrane preparations were prepared as described previously (28). Briefly, lysed-P2 membranes were incubated for 60 min at 25°C in l0 m M Tris-HCl, pH 7.4, containing 100 m M NaC1, 3 m M MnCI2, and 2 # M GTP. After the membranes were washed three times by repeated centrifugation and resuspension with ice-cold 10 m M Tris-HCl, pH 7.4, the membrane pellets were resuspended with l0 m M MOPS buffer, pH 7.4, containing 3 m M MnCl2. Following a 60-min incubation at 25°C with either 1 # M of the irreversible # ligand, 2-(p-ethoxybenzyl)-1-diethylaminoethyl-5-isothiocyanatobenzimidazole-HCl (BIT), or no drug (CNT), the membranes were washed three times by repeated centrifugation and resuspension into ice-cold 10 m M Tris-HC1, pH 7.4, and the final membrane pellets were kept at - 7 0 ° C until assayed. In some experiments, the a receptor irreversible ligand, di-o-tolyl-guanidine-isothiocyanate (DIGIT) (9), was used. In keeping with established nomenclatures (28), these types of membranes are termed CNT/BIT-P2, CNT/CNT-P2, and CNT/DIGIT-P2.

Ligand Binding Assay, Data Analysis, Statistics, and Chemicals As described (28), incubations with [3H][D-Ala2,DLeuS]enkephalin were conducted for 4-6 h at 25°C in 10 m M Tris-HCl, pH 7.4, containing 100 m M choline chloride, 3 m M MnC12. and a protease inhibitor cocktail [bacitracin (100 #g/ ml), bestatin (10 ~g/ml), leupeptin (4 ~g/ml), and chymostatin (2 ug/ml)]. In some experiments, 100 n M LY164929 [MeTyrD-Ala-GIy-N(Et)-CH(CH2-Ph)CH2-N(CH3)2] was included at a final concentration of 100 n M to block binding to the 5¢x (#) binding site (26). Single inhibition curves were fit to the twoparameter logistic equation (21) for the best-fit estimates of the IC50 and slope factor using MLAB-PC (Civilized Software, Bethesda, MD). Binding surfaces (12,22,30) were fit to one- and two-site binding models for the best-fit parameter estimates using MLAB-PC, which is a true implementation of MLAB as described by Knott and Reece for the DEC-10 computer system (10). Statistical differences between one- and two-site binding models, and between binding parameters, used the F-test (14), as previously described (25). Peptides were purchased from Peninsula Laboratories (Belmont, CA). [3H][D-AIa2,DLeuS]Enkephalin (sp.act. = 30 Ci/mmol), was purchased from Dupont New England Nuclear Corp. 5'-Guanylyimidodiphosphate (GppNHp) was purchased as the lithium salt from Sigma Chemical Co. (St. Louis, MO). Naltrindole and oxymorphindole were synthesized in the Laboratory of Medicinal Chemistry. 3Iodo-Tyr-D-Ala-Gly-Phe-D-Leu ([127I]DADL) was synthesized and purified at the Brown University Macromolecular Synthetic Facility (Providence, RI). [The sources of the other reagents have been described (28).] RESULTS

Effect of Nonspeciftc Acylation on DPDPE and DELT-I Inhibition Curves Our previous study (34) reported that the slope factor for DPDPE inhibiting [3H][D-Ala2,D-Leu~]enkophalin binding to the ~n~xsite was significantly decreased by BIT. Isothioeyanate-based irreversible ligands are highly reactive, and in addition to acylating the target receptor, they also, by reacting with free malihydryl groups, are capable of acylating membrane proteins. The series of experiments reported in Table l was carried out to assess the degree to which nonspecific acylation by BIT was responsible for the decreased slope factor. In these experiments, a common pool oflysed-P2 membranes was pretreated with either no drug, 1 # M BIT, or 1 a M DIGIT,

OPIOID PEPTIDES AND OTHER DRUG INTERACTION

which irreversibly binds to phencyclidine and ~ binding sites (19). The data (Table 1) demonstrated that pretreatment of membranes with DIGIT caused a statistically significant increase in the ICs0 of the DPDPE inhibition curve, and did not alter the DELT-I curve, relative to the inhibition curves observed with CNT/CNT-P2. Pretreatment of membranes with BIT also increased the IC5o of the DPDPE inhibition curve, but also decreased its slope factor. Additionally, BIT increased the IC50 of the DELT-I inhibition curve without significantly decreasing its slope factor.

Effect of GppNHp on DPDPE and DELT-I Inhibition Curves To test the hypothesis that the low slope factors resulted from the presence of two affinity states of the 6 ~ binding site, DPDPE and DELT-I inhibition curves were generated in the absence and presence of GppNHp, which would be predicted to increase the ICs0 values and increase the slope factors to 1.0 (33). The data (Table 2) showed that GppNHp modestly increased the ICso values of DPDPE and DELT-I, decreased the slope factors of the DPDPE inhibition curves, and had no significant effect on the slope factor of the DELT-I inhibition curves.

Binding Surface Analysis of [3H][D-Alae,o-LeuS]Enkephalin Binding to the ~ x Binding Site We used binding surface analysis to quantitatively test the hypothesis that [3H][D-Ala2,D-LeuS]enkephalin labeled two classes of 6~¢xbinding sites. The data set presented in our first study was extended by the use of oxymorphindole, which had a selectivity for the two sites opposite to that of DPDPE and DELT-I. The experimental design used is outlined in Table 3. Briefly, two concentrations of [3H][D-Ala2,D-LeuS]enkephalin were each displaced by eight concentrations of test drug, in the absence and presence of blocking agents. The blocking agents serve to partially block [3H][D-Ala2,D-LeuS]enkephalin binding to one of the two putative 6~x binding sites, thus providing additional information about the interaction of the ligands with the unblocked site. Additionally, the use of the blocked design, which links the different binding surfaces together by the use of

TABLE 2 EFFECT OF GppNHla ON DISPLACEMENTOF [3H][D-AIa2,D-LeuS]ENKEPHALINBINDING BY DPDPE AND DELT-I USING RAT BRAIN CNT/BIT-P2 MEMBRANES Drug DPDPE 0 ~M GppNHp 10 tzMGppNHp 50/zMGppNHp DELT-I 0 #M GppNHp 50/~M GppNHp

IC5o (nM _+SD)

Slope Factor

11.7 _+0.65 23.3 _+ 1.9" 16.8 -+ 2.1

0.77 _+0.03 0.64 -+ 0.03* 0.61 _-2-0.04*

7.06 _+0.98 14.3 _+ 1.8"

0.61 _+0.05 0.69 -+ 0.06

Nine point DPDPE and DELT-I inhibition curves were generated by displacing2.9 nM [3H][D-Ala2,D-LeuS]enkephalin.The data of three experiments were pooled, and fit to the two-parameter logisticequation for the best-fit estimates of the IC5o(nM _+ SD) and slope factor (n _+SD) reported above. The addition of GppNHp (both 10 uMand 50 #M) decreased specificbinding by 48%. * p < 0.01 when compared to control (F-test).

1209

TABLE 3 EXPERIMENTAL DESIGN OF THE [3H][D-AIa2,D-LeuS]ENKEPHALIN BINDING EXPERIMENTS Surface 1

2 3 4 5 6 7 8 9 l0 Il 12 13

Primary Displacer

BlockingDrug

[D-AIa2,D-Leu 5]Enkephalin [D-Ala2,D-LeuS]Enkephalin [o-Ala2,D-Leu 5]Enkephalin [D-Ala2,D-Leu 5]Enkephalin DPDPE DPDPE DPDPE DELT-| DELT-I DELT-I Oxymorphindole Oxymorphindole Oxymorphindole

None 5 nM DPDPE 5 nM DELT-I l nM Oxymorphindole None 5 nM DELT-I l nM Oxymorphindole None 5 nM DPDPE l nM Oxymorphindole None 5 nM DPDPE 5 nM DELT-I

The experimental design used to characterize 6ncxbinding sites is summarized above. Two concentrations of [3H][D-AIa2,D-Leu5]enkephalin (0.5 and 2.5 nM) were each displaced by eight concentrations of the indicated primary displacing drug, in the absence and presence of the indicated concentrations of blockingdrugs. Initial experiments were carfled out to insure that the concentration range of the displacing drugs inhibited binding by 10% to 90% of control, and that the blocking concentrations of drugs inhibited binding between 30 to 80% of control. Each unblocked binding surface generated 18 data points, and each blocked binding surface generated 20 data points. The data of the 13 binding surfaces were generated over several days. The data were pooled and fit to various binding models as described in the Results section.

ligands that are common to all the binding surfaces, serves to decrease the dependency values and the standard errors of the parameter estimates. The complete data set, which consisted of 392 data points, is shown in Fig. I A-E. As reported in Table 4, fitting the data to a one-site binding model resulted in a sum-of-squares (SS) of 1084. Fitting the data to a two-site binding model resulted in a highly significant decrease (p < 0.001) in the SS to 850. The best-fit parameter estimates (Table 4) indicated that [3H][DAla2,D-LeuS]enkephalin labeled two binding sites with Bm~xvalues of 45 fmol/mg protein and 96 fmol/mg protein, and Kd values of 6.9 n M and 1.2 nM. [D-Pen2,D-PenS]Enkephalin and DELT-I were 95-fold and 37.7-fold selective for site 2. In contrast, oxymorphindole was 17.2-fold selective for site 1.

Ligand-Selectivity Study of [3H ][D-Ala2,o-LeuS]Enkephalin Binding to Subtypes of the 6ncxBinding Site Ligand-selectivity studies were carded out as described in Table 5, and illustrated for DELT-II in Fig. 2. As reported in Table 5, naltrindole, like oxymorphindole, was relatively selective for site 1 (20-fold). [D-Ser2,Thrr]Enkephalin (DSTLE) and DELT-II were only 2.7-fold and 2.2-fold selective for site 1. As noted above, DPDPE and DELT-I were selective for site 2; [~27I]DADL was 52-fold selective for site 1. Morphine had moderate affinity and selectivity (11-fold) for site 1. Thus, of the 10 drugs studied, only DPDPE and DELT-I were selective for site 2. DISCUSSION Recent pharmacological data strongly support the hypothesis of 6 receptor subtypes as mediators of both supraspinal and spinal antinociception (8,1 l). In vitro ligand binding data, which are

1210

XU ET AL.

B.

A. DADL BINDING ~URFACE8

IADL BINDING SURFACES

100

uno,Md ,1~ i 7 mll.[4 o

........

i

oo._ i

7"-"

• mm~ogmw0~0~'0 /#l

o'~,i

1

O

-1~

-tl

-10

LOGADL'0~--IM

-7

-0

.1

-S

-11

-10

-9 -I LOG [DAD(.] M

O. DELTORPHIN4 BINDING SURFACES 100 bl • rod

100

-o"-~

- -

...... f 40

4°

i

2°

J

0 .,,

.,°

.

0

_

_

.,

.°

T

2O

1

0

.°

E. O X Y M O f t ~

| n m tumuli

-111

•

~,:

*7

-°

!

ti

i

i

k i

1

°11

-t0 -II -8 -7 LOG (OILTOIIPHIN-t] M

-°

.It

-O

-6

BINOING Ir4JNFACE8

100

* * 0 , , , hl~.l s '

L___ U , . i m , d

'° 20

0

-'Ill

-11

-I0

-9

-0

-7

kOO [OXYMOIIPtWIDOLE] II

FIG. 1. (A) [3H][D-Ala2,D-LeuS]Enkephalin[0.5 nM (dashed line) and 2.5 nM (solid line)] was displaced by [D-Ala2,D-LeuS]enkephalin,in the absence and presence of 5 nM DPDPE. (B) [3H][~Ala2,D-LeuS]Enkephalin [0.5 nM (dashed line) and 2.5 nM (solid line)] was displaced by [~AIa2,DLeuS]enkephalin, in the presence of 5 nM DELT-I and 1 nM oxymorphindole. (C) [3H][D-Ala2,D-LeuS]Enkephalin(0.5 nMand 2.5 riM) was displaced by DPDPE, in the absence and presence of 5 nM DELT-I and l nM oxymorphindole. (D) [3H][D-Ala2,D-Leu~]Enkephalin(0.5 nM and 2.5 riM) was displaced by DPDPE, in the absence and presence of 5 nM DPDPE and l nM oxymorphindole. (E) [3H][D-Ala2,D-LeuS]Enkephalin(0.5 taM and 2.5 nM) was displaced by oxymorphindole, in the absence and presence of 5 nM DPDPE and 5 nM DELT-I. fully supportive of the in vivo data, are still lacking. Negri et al. (15) first reported preliminary evidence for ~ receptor subtypes. That study was somewhat limited, since there were not sufficient ligand-selectivity data to permit comparison with the in vivo data, and since experiments that could distinguish between the occurrence of two states of a single receptor and two separate sites were not conducted. Moreover, although our first study resolved two ~n= binding sites, the ligand selectivity of the two sites was also too limited for a valid comparison with the in vivo data. The data of the present study considerably extend the work reported in our first paper (34). Previous studies showed that the slope factor for DPDPE inhibiting [3H][D-Ala2,D-Leu~]enkephalin binding to the ~,cx site

is significantly decreased by BIT. Since the ability to resolve a two-site binding model is related to the occurrence of a low slope factor, the issue raised by this observation is that the two sites might have in some manner been created by pretreatment of the membranes with BIT. Although BIT binds irreversibly to the ~ (5c~) binding site (20,28), it is very likely that the highly reactive isothiocyanate group of BIT nonspecitieally aeylates susceptible sulfhydryl groups of membrane proteins. Thus, although BIT does not irreversibly bind to the ~ site, it is possible that it might alter its properties via nonspeeific acylation. To address this question, membranes were pretreated with BIT, as well as an unrelated acylating agent, DIGIT, which acylares PCP and a binding sites, but not opioid sites (19). It seemed

OPIOID PEPTIDES A N D O T H E R D R U G I N T E R A C T I O N

1211

TABLE 4 BEST-FIT ESTIMATESOF THE TWO-S1TE BINDING MODEL: 6~, BINDING SITES Site 1 Bm~ (fmol/mg protein) [D.Ala2,D.Leu5]Enkephalin (Kd, aM) DPDPE (Ki, aM) DELT-I (Ki, aM) Oxymorphindole (Ki, aM)

45 _+ I l

TABLE 5 STRUCTURE-ACTIVITYSTUDY OF SUBTYPES OF THE 6~-~BINDING SITES Site 2

K~(nM +_SD) Site l

Site 2

6.9 _+2.6 190 _+ 132 147 _+ 86 0.06 _+0.05 0.018 _+0.027 0.43 _+0.16 0.69 _+0.51 0.12 _+0.10 16 _+ 15

1.2 _+0.1 2.0 _+0.2 3.9 -+ 0.4 1.03 _+0.10 0.36 _+0.06 1.17 _+0.07 1.53 _+0.19 6.25 _+0.55 176 _+20

96 _+ 6

6.9 _+ 2.6 190 _+ 132 147 _+ 86

1.2 _+0.1 2.0 + 0.2 3.9 + 0.4

0.06 +_ 0.05

1.03 +_ 0.10

The data presented in Fig. 1 (392 points) were fit to the one-site model, with a SS of 1084. When fit to a two-site model, the SS decreased to 850. This was a highly significant change (F > 20, p < 0.0001). Each parameter value reported above is _+SD.

reasonable to hypothesize that alterations in binding parameters produced by DIGIT could be assumed to reflect nonspecific acylation. As reported in Table l, the low slope factor of DELTI, which was observed with control membranes (0.78), was not significantly altered by either BIT or DIGIT. Whereas BIT increased the ICs0 of DELT-I, D I G I T did not. In contrast, BIT significantly decreased the slope factor of DPDPE, whereas DIGIT did not, and both D I G I T and BIT increased the ICs0 of DPDPE. The different effects produced by BIT and D I G I T on the binding parameters of the DPDPE and DELT-I inhibition curves indicate that the results of this experiment are more difficult to interpret than we hoped they would be. As noted above, the most likely explanation for the lower slope factors in CNT/BITP2 membranes is that acylation of sulfhydryl groups by BIT produces a conformationai change in the ~ binding sites such that the selectivity of the peptides, in particular DPDPE, for the two sites is increased. In the absence of acylation, DPDPE would presumably have about the same Ki for both sites, whereas after acylation, it would be more selective between the two sites. These data suggest that nonspecific acylation can affect ligands differently. Similar nonspecific acylation has been observed with the 6-selective acylating agents, FIT and (+)-trans-SUPERFIT (29). Since the slope factor of DELT-I is not altered, it is likely that its selectivity for the two sites is not altered by BIT. The postulated conformational change is presumably sufficiently subtle that it is more readily detected with DPDPE than with DELTI. Most importantly, the fact that low slope factors occur in the absence of acylation is evidence that they are not artifacts of acylation. Although low slope factors are consistent with the occurrence of two separate binding sites, they can also occur if the [3H]ligand labels two states of a single receptor that are interconvertible by guanyl nucleotides (4,33). According to the latter model, the [3H]ligand, which is usually an antagonist, labels high- and lowaffinity states of a single receptor with about equal affinity. The displacing drug, which has high affinity for one state and low affinity for the other state, displaces [3H]ligand binding with a low slope factor. The addition of G T P or its metabolically stable analog, GppNHp, converts the high-affinity state to the lowaffinity state. As a result, the slope factor of the displacing drug increases to 1 and the ICso of the displacing drug also increases, reflecting an exclusive interaction with the low-affinity state. The experiments reported in Table 2 failed to find evidence that the low slope factors observed for DPDPE and DELT-I inhibition curves are due to two states of a single receptor. For example, both 10 # M G p p N H p and 50 # M G p p N H p signifi-

[o-Ala2,o-LeuS]Enkephalin * DPDPE* DELT-I* Oxymorphindole* Naltrindole DSTLE DELT-II []:71]DADL Morphine

The structure-activity studies were generated with CNT/BIT-P2 in three independent experiments as follows. Experiment l: Two concentrations of [3H][t)-Ala2,D-LeuS]enkephalin (0.5 and 2.5 aM) were each displaced by eight concentrations of test drug. Experiment 2: Two concentrations of [3H][D-Ala2,D-LeuS]enkephalin (0.5 and 2.5 aM) were each displaced by eight concentrations of test drug in the presence of 5 nM DPDPE. Experiment 3: Two concentrations of [~H][D-Ala2,DLeuS]enkephalin (0.5 and 2.5 aM) were each displaced by eight concentrations of test drug in the presence of 1 aM oxymorphindole. The data of all three experiments were combined (52 data points), and fit to the two-site model with the Ka values of [3H][D-Ala2,D-LeuS]enkephalin, the Ki values of DPDPE and oxymorphindole, and the Bmaxvalues fixed to those reported in Table 4. * These results are from Table 4.

cantly decreased the slope factor for DPDPE, and did not affect the slope factor of DELT-I. Although 50 ~,M G p p N H p significantly increased the ICs0 of DELT-I, it had no significant effect on the ICso of DPDPE. To characterize the two (~.~ binding sites in greater detail, the experiments outlined in Table 3 were carried out. These data extend the data set reported in our previous study (34), by the addition of binding surfaces utilizing oxymorphindole, to a total of 392 points. The complete data set is shown in Fig. 1. As reported in Table 4, two binding sites were readily resolved. Whereas DPDPE and DELT-I were relatively selective for site 2, oxymorphindole was relatively selective for site I. [D-Ala2,D LeuS]Enkephalin was only 5.7-fold selective for site 2. With this

loo ~1 '~

f __1

I

Blocked ~ith 1 n M

eo _ o . . , o . , e ~ ~

OZI="

u~,*c,d

.

.

.

.

.

.

.

.... ~"--

eL

" "" '-'~.

I"

4 0

I~ m

20

5 nM DRDPE

:

"m ~'.

-Ik t ~

............

D

a. o -1

-11

-10

-9

-8

LOG [DELTORPHIN-II]

-7 M

-6

FIG. 2. [3H][D-Ala2,D-LeuS]Enkephalin(0.5 aM and 2.5 riM) was displaced by DELT-II, in the absence and presence of 5 aM DPDPE and ! aM oxymorphindole.

1212

XU ET AL.

degree of selectivity, [3H][D-Ala2,D-LeuS]enkephalin binding surfaces alone do not distinguish between the two sites, which supports fitting [3H][D-Ala2,D-LeuS]enkephalin binding surfaces generated with CNT/BIT-P2 to one-site binding models (28). The addition of the oxymorphindole data resulted in a significant change in the binding parameters as compared to the data of our previous study (34), especially in regards to the Bm~ values. For example, the Bmax values reported previously were 12.9 and 120 fmol/mg protein (34). In the present study, the Bm~xvalues were 45 and 96 fmol/mg protein. In our experience, this type of change is commonly observed when a data set is made more constrained. The ligand-selectivity study was conducted as illustrated in Fig. 2 for DELT-II. As reported in Table 5, naltrindole, like oxymorphindole, was relatively selective for site 1 (20-fold). [DSer2,Thr6]Enkephalin and DELT-II were only 2.7-fold and 2.2fold selective for site 1. As noted above, DPDPE and DELT-I were selective for site 2. Morphine had moderate affinity for site 1 (g~ = 16 nM), and was about 1 l-fold selective for this site; [~27I]DADL was 52-fold selective for site 1. Thus, of the 10 drugs studied, only DPDPE and DELT-I were selective for site 2. These findings suggest that [3H]DELT-I might be used to selectively label site 2, and that [127I]DADL might be used to selectively label site 1, using DPDPE to block binding to site 2. Recent studies with [3H]naltrindole showed that it labels an apparent single class of binding sites in rat brain membranes with a Ka of 0.037 nM and a Bm~ value of 63 fmol/mg protein (35). These values approximate the values observed here for site 1 (Ki = 0.018 nM, Bma~ = 45 fmol/mg protein), suggesting that [3H]naltrindole might selectively label site 1. Alternatively, it is also possible that under the assay conditions used by Yamamura et al. (35), [3H]naltrindole might have similar affinity for site 1 and site 2, and therefore label both sites. The data of the present study indicate that, in the presence of a blocking concentration of DPDPE, [3H]naltrindole could conceivably be used to selectively label site 1. However, since naltrindole has a Ki for the K2~ binding site in the low nanomolar range (23), it would be important to exclude the labeling of this site by this radioligand.

Porreca et al. have recently provided compelling in vivo evidence for 8 receptor subtypes as mediators of both supraspinal and spinal antinociception (8,11). Based on their findings, 8 receptors were tentatively classified into two subtypes: the ~l receptor, which is selectively activated by DPDPE and blocked by DALCE, and the 82 receptor, which is selectively activated by DELT-II and blocked by 5'-NTII. The ligand selectivity of the two tSncxbinding sites is not entirely concordant with this definition of 81 and 82 receptors, which would predict that DELTII and DPDPE have opposite selectivity between the two sites. However, it is also possible that DELT-II and DPDPE might have similar binding selectivity, but opposite efficacy at the two subtypes of the 8ncx site, which would be consistent with the pharmacological data. The fact that DELT-II is about 2-fold selective for site 1, and DPDPE is selective for site 2, is consistent with the pharmacological data. However, DELT-I is, like DPDPE, selective for site 2. Although data on the selectivity of DELT-I for ~ and 82 receptors are not known, there is no reason to expect that DELT-II and DELT-I would have different biological activities. In addition, the fact that DALCE, a selective 81 antagonist (3,7), irreversibly blocks [3H][D-Ala2,D LeuS]enkephalin binding to the tSncx site suggests that neither subtype of the 8ncxsite is related to the 82 receptor (24). A point of considerable interest is the relationship between the 8n~x and 8¢x binding sites, and the 81 and 82 receptors. Observations that DALCE selectively inhibits the 8nc~binding site (24), and blocks the effect of DPDPE at the 8,cx receptor (7), suggest that the 8~ and (5,~xbinding sites may be synonymous. The 82 receptors are selectively activated by deltorphin-II, and selectively blocked by the irreversible 8 antagonist, 5'-NTII (8). Doses of 5'-NTII that block deltorphin-II effects also block modulation of morphine antinociception by t5 ligands (16,27), an effect thought to be mediated via the 8~ binding site (6,27). Thus, these data suggest that the 8cx site and the 82 receptor might also be synonymous. Future studies will address this important issue, as well as the related question of heterogeneity of the 8~x binding sites. In summary, the present study provides additional data to support the existence of subtypes of the 8,¢x binding site.

REFERENCES

1. Brantl, V.; Pfeiffer, A.; Herz, A.; Henschen, A.; Lottspeich, F. Antinociceptive potencies of beta-casomorphin analogs as compared to their affinities towards mu and delta opiate receptor sites in brain and periphery. Peptides. 3:793-797; 1982. 2. Bryan, W. M.; Callahan, J. F.; Codd, E. E.; Lemieux, C.; Moore, M. L.; Schiller, P. W.; Walker, R. F.; Huffman, W. F. Cyclic enkephalin analogues containing alpha-amino-beta-mercapto-beta,betapentamethylenepropionic acid at positions 2 or 5. J. Med. Chem. 32:302-304; 1989. 3. Calcagnetti, D. J.; Fanselow, M. S.; Helmstetter, F. J.; Bowen, W. D. [D-Ala2,LeuS,Cysr]enkephalin:Short-term agonist effects and long-term antagonism at delta opioid receptors. Peptides 10:319326; 1989. 4. Childers, S. R.; Snyder, S. H. Guanine nucleotides differentiate agonist and antagonist interactions with opiate receptors. Life Sci. 23: 759-761; 1978. 5. Erspamer, V.; Melchiorri, P.; Falconieri Erspamer, G.; Negri, L.; Corsi, R.; Severini, C.; Barra, D.; Simmaco, M.; Kreil, G. Deltorphins: A family of naturally occurring peptides with high affinity and selectivity for delta opioid binding sites. Proc. Natl. Acad. Sci. USA 86:5188-5192; 1989. 6. Holaday, J. W.; Porreca, F.; Rothman, R. B. Functional coupling among opioid receptor types: Implications for anesthesiology. In: Estafanous, F. G., ed. Opioids in anesthesia II. Boston: ButterworthHeineman Publishers; 1990:50-60.

7. Jiang, Q.; Bowen, W. D.; Mosberg, H. 1.; Rothman, R. B.; Porreca, F. Opioid agonist and antagonist antinocieeptive properties of [DAla2,LeuS,Cysr]enkephalin:Selective actions at the delta,,,,~om~x~ site. J. Pharmacol. Exp. Ther. 255:636-641; 1990. 8. Jiang, Q.; Takemori, A. E.; Sultana, M.; Portoghese, P. S.; Bowen, W. D.; Mosberg, H. I.; Porreca, F. Differential antagonism ofopioid delta antinociception by [D-Ala2,LeuS,Cys~]enkephalinand naltrindole 5'-isothiocyanate: Evidence for delta receptor subtypes. J. Pharmacol. Exp. Ther. 257:1069-1075; 1991. 9. Kim, C. H.; Reid, A. A.; Rothman, R. B.; Jacobson, A. E.; Rice, K. C. Synthesis and binding studies of affinity ligands based on ditolylguanidine (DTG). A study of their interactions with the sigma binding site. In: Harris, L. S., ed. Nida research monograph: Problems of drag dependence 1988. Washington, DC: Supt. of Docs., US Govt. Print Off.; 1989. 10. Knott, G. D.; Reece, D. K. MLAB: A civilized curve fitting system. Proceedings of the Online 1972 International Conference 1:497526; 1972. 11. Mattia, A.; Vanderah, T.; Mosberg, H. I.; Porreca, F. Lack ofantinocieeptive cross-tolerance between [D-Pen2,D-Pen~]enkephalin and [D-Ala2]deltorphin II in mice: Evidence for delta receptor subtypes. J. Pharmacol. Exp. Ther. 258:583-587; 1991. 12. McGonigle, P.; Neve, K. A.; Molinoff, P. B. A quantitative method of analyzing the interaction of slightly selective radioligands with multiple receptor subtypes. Mol. Pharmacol. 30:329-337; 1986.

OPIOID PEPTIDES A N D O T H E R D R U G I N T E R A C T I O N 13. Mosberg, H. I.; Hurst, R.; Hruby, V. J.; Gee, K.; Yamamura, H. I.; Galligan, J. J.; Burkes, T. F. Bis-penicillamineenkephalins possess highly improved specificity toward delta opioid receptors. Proc. Nail. Acad. Sci. USA 80:5871-5874; 1983. 14. Munson, P. J.; Rodbard, D. LIGAND: A versatile computerized approach for characterization of ligand-binding systems. Anal. Biochem. 107:220-239; 1980. 15. Negri, L.; Potenza, R. L.; Corsi, R.; Melchiorri, P. Evidence for two subtypes of ~ opioid receptors in rat brain. Eur. J. Pharmacol. 196: 335-336; 1991. 16. Porreca, F.; Takemori, A. E.; Sultana, M.; Portoghese, P. S.; Mosberg, H. I. Modulation of mu-mediated antinociception in the mouse involves opioid delta-2 receptors. J. Pharmacol. Exp. Ther. 263: 147-152; 1992. 17. Portoghese, P. S.; Sultana, M.; Takemori, A. E. Naltrindole, a highly selective and potent non-peptide delta opioid receptor antagonist. Eur. J. Pharmacol. 146:185-186; 1988. 18. Portoghese, P. S.; Sultana, M.; Takemori, A. E. Naltrindole 5'-isothiocyanate: A nonequilibrium, highly selective delta opioid receptor antagonist. J. Med. Chem. 33:1547-1548; 1990. 19. Reid, A. A.; Kim, C.-H.; Thurkauf, A.; Monn, J. A.; De Costa, B. R.; Jacobson, A. E.; Rice, K. C.; Bowen, W. D.; Rothman, R. B. Wash-resistant inhibition of phencyclidine and haloperidolsensitive sigma binding sites in guinea pig brain by putative affinity ligands: Determination of selectivity. Neuropharmacology 29:10471053; 1990. 20. Rice, K. C.; Jacobson, A. E.; Burke, T. R., Jr.; Bajwa, B. S.; Streaty, R. A.; Klee, W. A. Irreversible ligands with high selectivity toward mu or delta opiate receptors. Science 220:314-316; 1983. 21. Rodbard, D.; Lenox, R. H.; Wray, H. L.; Ramseth, D. Statistical characterization of the random errors in the radioimmunoassaydoseresponse variable. Clin. Chem. 22:350-358; 1976. 22. Rothman, R. B. Binding surface analysis: An intuitive yet quantitative method for the design and analysis ofligand binding studies. Alcohol Drug Res. 6:309-325; 1986. 23. Rothman, R. B.; Bykov, V.; De Costa, B. R.; Jacobson, A. E.; Rice, K. C.; Brady, L. S. Interaction of endogenous opioid peptides and other drugs with four kappa opioid binding sites in guinea pig brain. Peptides 1 l:311-331; 1990. 24. Rothman, R. B.; Bykov, V.; Jacobson, A. E.; Rice, K. C.; Long, J. B.; Bowen, W. D. A study of the effect of the irreversible delta receptor antagonist [D-Ala2,LeuS,Cys6]enkephalin on 6cx and tSncx opioid binding sites in vitro and in vivo. Peptides 13:691-694; 1992. 25. Rothman, R. B.; Bykov, V.; Mahboubi, A.; Long, J. B.; Jiang, Q.; Porreca, F.; De Costa, B. R.; Jacobson, A. E.; Rice, K. C.; Holaday,

1213

26. 27.

28.

29.

30.

31.

32.

33.

34. 35.

J. W. Interaction of beta-funaltrexamine with [aH]cycloFOXY binding in rat brain: Further evidence that beta-FNA alkylates the opioid receptor complex. Synapse 8:86-99; 1991. Rothman, R. B.; Bykov, V.; Ofri, D.; Rice, K. C. LY164929: A highly selective ligand for the lower affinity [3H]D-ala2-D-leuS-enkephalin binding site. Neuropeptides 11:13-16; ! 988. Rothman, R. B.; Holaday, J. W.; Porreca, F. Aliosteric coupling among opioid receptors: Evidence for an opioid receptor complex. In: Herz, A., ed. Handbook of experimental pharmacology, vol. 104, "Opioids." Berlin: Springer-Verlag; 1992:217-237. Rothman, R. B.; Long, J. B.; Bykov, V.; Jacobson, A. E.; Rice, K. C.; Holaday, J. W. Beta-FNA binds irreversibly to the opiate receptor complex: In vivo and in vitro evidence. J. Pharmacol. Exp. Ther. 247:405-416; 1988. Rothman, R. B.; Mahboubi, A.; Bykov, V.; Kim, C.-H.; Jacobson, A. E.; Rice, K. C. Probing the opioid receptor complex with (+)trans-SUPERFIT I. Evidence that [D-Pen2,D-PenS]enkephalin interacts with high affinity at the 6c~ binding site. Peptides 12:359364; 1991. Rovati, G. E.; Rodbard, D.; Munson, P. J. DESIGN: Computerized optimization of experimental design for estimating Kd and Bmax in ligand binding experiments. II Simultaneous analysis of homologous and heterologous competition curves and analysis blocking and of "multiligand" dose-response surfaces. Anal. Biochem. 184: 172-183; 1990. Shimohigashi, Y.; Takano, Y.; Kamiya, H.; Costa, T.; Herz, A.; Stammer, C. H. A highly selective ligand for brain delta opiate receptors, a cyclopropyl(E)Phe(4)-enkephalin analog, suppresses mu receptor-mediated thermal analgesia by morphine. FEBS Lett. 233: 289-293; 1988. Vaughn, L. K.; Wire, W. S.; Davis, P.; Shimohigashi, Y.; Toth, G.; Knapp, R. J.; Hruby, V. J.; Burks, T. F.; Yamamura, H. I. Differentiation between rat brain and mouse vas deferens delta opioid receptors. Eur. J. Pharmacol. 177:99-101; 1990. Weding, L. L.; Puttfarcken, P. S.; Cox, B. M. Multiple agonist-affinity states ofopioid receptors: Regulation of bindingby guanyl nucleotides in guinea pig cortical, NGI08-15, and 7315c cell membranes. Mol. Pharmacol. 33:423-431; 1988. Xu, H.; Ni, Q.; Jacobson, A. E.; Rice, K. C.; Rothman, R. B. Preliminary ligand binding data for subtypes oftbe delta opioid receptor in rat brain membranes. Life Sci. 49:PL141-PL146; 1991. Yamamura, M. S.; Horvath, R.; Toth, G.; Otvos, F.; Malatynska, E.; Knapp, R. J.; Porreca, F.; Hruby, V. J.; Yamamura, H. I. Characterization of naltrindole binding to delta opioid receptors in rat brain. Life Sci. 50:PL119-PL124; 1992.