Neurohypophyseal hormone-responsive renal adenylate cyclase. III. Relationship between affinity and intrinsic activity in neurohypophyseal hormones and structural analogs.

Binding and adenylate cyclase activation with neurohypophyseal hormones (NHH) and synthetic analogs were examined in two bovine renal medullary membrane preparations. The relationship between binding affinity of NHH and 10 analogs (as determined by competitive binding studies with tritiated I Lys”]vasopressin, [“HILVP and adenylate cyclase activation was studied in one membrane preparation under identical conditions where hormonal stimulation of cyclase was at steady state and binding was at (or near) equilibrium. A linear relationship was demonstrated between the negative logs of peptide concentration required for half-maximal binding (K,,) and for half-maximal enzyme activation (K,,). This finding extends previous evidence indicating that: (a) the bovine membrane binding sites detected by I “HILVP have specificity characteristics expected for functional receptors coupled to adenylate cyclase. and (6) these analogs all occupy the same receptor “pocket.” Study of adenylate cyclase activation by NHH analogs, with and without a preliminary preincubation, showed that the magnitude of enzyme stimulation achieved (percentage of I',,,, relative to LVP) with certain low affinity analogs was increased when these peptides were assayed without preincubation (in the 0to IO-min period) without significant influence on K,L. To maximize the magnitude of adenylate cyclase stimulation achieved with low affinity analogs, the relationship between the apparent affinity and intrinsic activity of 30 NHH analogs was studied in a second membrane, where adenylate cyclase was assayed without preincubation. It

Binding and adenylate cyclase activation with neurohypophyseal hormones (NHH) and synthetic analogs were examined in two bovine renal medullary membrane preparations. The relationship between binding affinity of NHH and 10 analogs (as determined by competitive binding studies with tritiated I Lys"]vasopressin, ["HILVP and adenylate cyclase activation was studied in one membrane preparation under identical conditions where hormonal stimulation of cyclase was at steady state and binding was at (or near) equilibrium.
A linear relationship was demonstrated between the negative logs of peptide concentration required for half-maximal binding (K,,) and for half-maximal enzyme activation (K,,). This finding extends previous evidence indicating that: (a) the bovine membrane binding sites detected by I "HILVP have specificity characteristics expected for functional receptors coupled to adenylate cyclase. and (6) these analogs all occupy the same receptor "pocket." Study of adenylate cyclase activation by NHH analogs, with and without a preliminary preincubation, showed that the magnitude of enzyme stimulation achieved (percentage of I',,,, relative to LVP) with certain low affinity analogs was increased when these peptides were assayed without preincubation (in the 0-to IO-min period) without significant influence on K,L.
To maximize the magnitude of adenylate cyclase stimulation achieved with low affinity analogs, the relationship between the apparent affinity and intrinsic activity of 30 NHH analogs was studied in a second membrane, where adenylate cyclase was assayed without preincubation. It * This work was supported in part by grants from the National Institutes of Health (HD-062371, the Rockefeller Foundation, and the Schweppe Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "adwrtisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. $ To whom requests for reprints should be addressed. Ej Present address, Laboratory of Biochemistry, Faculty of Sci-was found that the apparent affinity of NHH analogs may be decreased about 5 orders of magnitude with only minimal (20%) reduction in intrinsic activity. With further decrease in affinity to a critical K,l value, there is an abrupt decrease in intrinsic activity so that peptides with K,z -2 x 10 4 M are either weak partial agonists or competitive antagonists. The structure-activity relationships established in this study are consistent with previous studies which have shown that: (i) the tripeptide tails of NHH are not necessary for adenylate cyclase activation; (ii) no single amino acid in the ring moieties is essential for biological activity; (iii) receptor-mediated enzyme activation involves the cooperative interaction of multiple elements of the ring with complementary elements of the receptor, with Asn: and Tyr2 of the peptide ring playing "critical" roles. Studies of [il"-acylated Lys%asopressin analogs have revealed it is possible to introduce carboxyl functions in the Lys' side chain with unexpectedly high retention of affinity; these analogs may therefore be useful intermediates for affinity chromatography of NHH receptors. Retro-11 analogs of I Gly71deaminooxytocin and deaminotocinamide, with a free NH, group, were shown to be weak partial agonists; Nformylation, resulting in charge removal, led to loss of agonistic activity with development of antagonistic activity. Structural differences between the bovine and porcine renal receptor have previously been established with respect to position 8 of the vasopressin molecule. The present studies indicate a second species difference, involving the NH, terminus of the peptide. The NH, terminus increases binding affinity in the porcine receptor but not in the bovine receptor where deaminopressinamide has higher affinity than pressinamide, and deaminooxytocin has higher affinity than oxytocin.  [~f-succinyl-Lys*lvasopressin [Ar-glycyl-Lys*lvasopressin [Nf-biotinyl-Lys*lvasopressin lw-fluoresceinyl-thiocarbamyl-Lys81vasopressin ID-aIle"]-retro-D-tocinamide N-Formyl-ID-aIle"]retro-D-tocinamide ln-aIle3, Gly'lretro-n-deaminooxytocin [n-aIle3, Gly7, L-Leu"]retro-n-deaminooxytocin iV-Formylln-aIle3, Gly'lretro-n-deaminooxytocin In-aIle3, Gly', malonamideYlretro-n-deaminooxytocin studies and describe the effects of 30 synthetic analogs of AVP and OT. These studies with NHH' analogs have served to define the relationship between: (a) the apparent affinity for peptide binding to receptor sites and for adenylate cyclase activation; (b) affinity and intrinsic activity (percentage of V max relative to AVP (or LVP), where V,,, is the magnitude of adenylate cyclase activation achieved by maximally stimulating concentrations of peptide); and (c) modification of peptide structure and activity in the NHH-responsive adenylate cyclase system of bovine renal medullary membranes.    with Membrane 5 are shown in Table III and Fig. 1. Fig. 1 shows the relationship between the negative logarithms of K,, and K,(H) with NHH and a set of analogs, which differ in apparent affinity by more than 5 orders of magnitude.
The linear relationship observed provides further evidence that the specific sites in bovine membranes are receptors coupled to adenylate cyclase. There is no straightforward relationship between affinity (l/K,,) and intrinsic activity, in the set of Affinity and Intrinsic ActiuitylNeurohypophyseal Hormones 3233 of DeT varied from 85 to 100% (relative to OT ration 6 (40 to 50 fig of membrane protein/50 ~1 of standard medium) = lOO%), whereas with 12l/z-min preincubation DeT intrinsic activduring the 0-to lo-min period of incubation.
DeOT consistently had ity (V,,,;,, (11)) was only about 50%.  Table I Table III shows that with several partial agonists (DeP, DeT, and I Gly7]retro-n-DeOT) intrinsic activity is increased when membranes are assayed without preincubation. Table  IV illustrates this point with DeT; when assayed in six experiments without preincubation, the V,,,;,,(I) of DeT ranged from 85 to 100% (relative to AVP or OT), whereas with preincubation, the mean V,,,;,,(H) value for DeT was about 50% (range 45 to 56% in five experiments).

Experi
This change in percentage of V,,,,, with low affinity peptide analogs, results from a secondary decrease in enzyme velocity which occurs after about 10 min incubation with the very high concentrations of these peptides required to achieve maximal enzyme stimulation (Fig. 2). Whereas the enzyme activation produced  iTable IV) and P, DeP, and DeT (Table VIII) was also determined.
In view of the linear relationships between K,, and K,l(II) (Fig. 1) and K,, (II) uersus K,,(I) (Table III), the values obtained for K,,(I) with low affinity analogs may be assumed to be approximately equivalent to K,, . In the case of certain analogs, which are weak partial agonists, or competitive inhibitors, or both, the K,, value was considered equivalent to the K, value as calculated by the Schild method (6). Fig. 3 shows the

Various
Nf-acylated Lys* derivatives of LVP were compared in membrane Preparation 6, without preliminary preincubation. The values shown for V,,,,, (I) and K,(I) are the mean of at least two experiments, where the analog and LVP were compared in the same experiment.  relationship between apparent affinity (l/K,,(D) or l/K,) and intrinsic activity (percentage of V,,,, relative to AVP) in this series of 30 analogs and 4 NHH.

DISCUSSION
Bovine renal medullary membranes were previously shown to have specific hormone binding sites which behave as a homogeneous population of molecules with a single affinity constant for tritiated LVP or AVP (1). The present studies with NHH analogs provide strong additional evidence that the specific 1 "HILVP binding sites in bovine membranes have peptide specificity characteristics expected for functional NHH-receptors coupled to adenylate cyclase. Thus, a linear relationship was demonstrated between the negative logarithms of K,, and K,, with four naturally occurring NHH, and a large number of analogs which vary over more than 5 orders of magnitude in apparent affinity (Fig. 1). This finding which extends previous studies, provides strong evidence that all of these peptides occupy the same "pocket" in a single class of receptor molecules, as discussed previously (1).
The peptide concentration-adenylate cyclase activation curves with NHH and analogs in bovine membranes had n,,(A) values which statistically were not significantly different than 1.0. In porcine membranes the LVP concentration-enzyme activation curves in porcine membranes has n,,(A) values about 0.3, indicative of significant negative cooperativity; with other NHH and analogs there was a tendency for apparent negative cooperativity of activation to decrease as the apparent affinity for adenylate cyclase activation decreased, but there was no strict correlation between these parameters. The K,/K,, ratios obtained with NHH and analogs in pig membranes appeared to be correlated with n,, (A), the greater the degree of apparent negative cooperativity the higher the K,,/K,, ratio (10). Thus in porcine membranes, the reported K,,IK, ratio was about 5O:l with LVP, while the ratios with AVP and OT were about 9:l and 1.5:1, respectively; analogs of lower affinity than OT had K,,/K,, ratios approaching 1:l. As previously discussed (1) the basis for the differences in apparent negative cooperativity of hormonal activation observed between porcine and bovine membranes remains to be elucidated.
In the present study, we have been able to define the relationship between the apparent affinity of a peptide analog for receptor binding and its intrinsic activity to stimulate adenylate cyclase in bovine membranes. The K,, values of "very low" affinity analogs (wherein K,, could not be measured from competitive binding studies with 1"HJLVP) was approximated by taking advantage of the fact that: (i) the negative log of K,, is linearly related to the negative log K,, (II) (Fig. 1) and that (ii) with low affinity analogs the K,, values for adenylate cyclase activation assayed with or without preincubation are very similar (Table III). Using l/K,,(I) or l/K, as a measure of relative affinity, it was demonstrated that a progressive decrease in apparent affinity over about 5 orders of magnitude results in only a minor reduction in intrinsic activity, V,,,,, being 80% or greater, of that achieved with AVP or LVP (Fig. 3). However, further decrease in peptide affinity to a critical region (to K,, (I) or Ki values about 2 to 2.5 x 10m4 M) results in a dramatic decrease in intrinsic activity. so that peptides in this affinity range are either weak partial agonists (with I',,,;,, 10 to 50% relative to AVP) or competitive antagonists. Previous studies (1) have shown that the decrease in affinity of LVP relative to AVP is due to two factors: (a) a decrease in the kinetic association constant (k,). which is associated with (b) an equivalent increase in the kinetic dissociation constant (h,.). While we cannot be certain that h, and h, are equivalently modified in all peptide analogs, as a first approximation we may assume that for 2 orders of magnitude of change in K,, (equivalent to K,, = h,.lh,) k, is decreased and h,. increased by about 1 order or magnitude. Since h, for a given peptide is inversely related to the mean time that a peptide molecule occupies a receptor site, Fig. 3 indicates that peptides with very short occupancy times relative to AVP (about 2 orders of magnitude less than AVP) are able to activate most of the cyclase units in the membrane. Further decrease in affinity appears to lead to shortened mean occupancy time, insufficient to permit full activation of enzyme units; thus low affinity analogs of this type (r.g. [Ala"]OT, lVaP]OT, 1 AlaZ]OT, and some retro-n analogs) are weak partial agonists. In the case of other analogs (I n-Leu4 IOT and formylated retro-n analogs) h, may be increased to the point where effective occupancy time is insufficient to permit enzyme activation. so that such analogs are pure competitive antagonists. In examining Fig. 3. three peptides, IAla'lVP, 1 Gly']OT, and 1 Des-Leu-Gly-NH,lOT, appear to have higher intrinsic activity than is expected for their relatively low affinity, suggesting that the modification of structure in these peptides may influence kinetic parameters differentially. Thus, the low affinity of IGly'lOT and IAla'lVP may be primarily related to low h, values for association, rather than to the L-peptide; it was not possible to establish similarity of other elements of the molecule because in solution the conformation of the L or D molecule is not fixed unambiguously.