Experimentalevidence for agonist trafficking potential pitfalls

Much of the difficulty in correctly interpreting agonist dose-response curves arises from the lack of knowledge of the underlying molecular events. Indeed, interpretations 198

-9 -8 -7 -6 -5 -4 -3 -9 -8 -7 -6 -5 -4 -3 [Agonist] (Log M)

-9 -8 -7 -6 -5 -4 -3 -9 -8 -7 -6 -5 -4 -3 [Agonist] (Log M)

Figure 200 Separated two-state models describing the completely independent formation of AR* and AR** (reprinted from Trends in Pharmacological Science, 18, Leff, P., Scaramellini, C., Law, C. and McKechnie, K., A three-state receptor model of agonist action, 355-362. Copyright (1997), with permission from Elsevier). Simulations according to this model with L = 5, M = 3. For ligand 1: KA = 2 jiM, KA* = 0.08 jiM, KA** = 0.1 jiM. For ligand 2: KA = 10 jiM, KA* = 0.01 jiM, KA** = 2 jiM.

could be biased by the fact that, although generated by a single type of receptor, the different response pathways are likely to display differences in 'receptor reserve'.

In addition, each agonist dose-response curve also depends on the strength of its stimulus. When defined as the concentration of activated receptor-G protein complexes, this parameter can be affected by the affinity of the activated receptors for their cognate G proteins as well as by the concentration ratios between each of these participants (see Section 4.12). In this respect, a single activated state of the receptor can still explain why one ligand is able to activate two pathways while another ligand is only able to activate one pathway. Indeed, when the activated receptor has a high affinity for G1 and a low affinity for G2, an agonist of high efficacy may produce enough activated receptors to produce significant coupling to both G proteins while an agonist of low efficacy may only produce enough activated receptors to produce significant coupling to G1. Hence, there is no strict necessity to invoke 'agonist trafficking' to explain the observation that oxymetazoline only stimulates Gi in a2-adrenergic receptor-transfected CHO cells while adrenaline stimulates both Gi and Gs in these cells (Figure 201).

Also, before evoking 'agonist trafficking', great care must be taken that no other receptor or receptor subtype than the receptor of interest is involved. This potential source of artefact is clearly illustrated in a recent study by on porcine a2A-receptor-expressing CHO cell lines (Figure 202). Even when care was taken to avoid differences in 'receptor reserves' between Gs and Gi-mediated effects (see coinciding curves for noradrenaline and adrenaline); oxymetazoline showed selectivity for inducing signalling through the Gi pathway. However, this effect is an artefact since it is mediated

Oxymetazoline Adrenaline Noradrenaline

—>—l—i—l—i—l—i—l—>—l—'—l—1—l—

—>—l—i—l—i—l—i—l—>—l—'—l—1—l—

Figure 201 Gi- (in presence of cholera toxin) and Gs- (in presence of pertussis toxin) mediated effects of two agonists on a2-adrenergic receptor transfected CHO cells. Reproduced from Eason, M., Jacinto, M. and Liggett, S. (1994) Molecular Pharmacology, 45, 696-702, with permission from the American Society for Pharmacology and Experimental Therapeutics.

Figure 201 Gi- (in presence of cholera toxin) and Gs- (in presence of pertussis toxin) mediated effects of two agonists on a2-adrenergic receptor transfected CHO cells. Reproduced from Eason, M., Jacinto, M. and Liggett, S. (1994) Molecular Pharmacology, 45, 696-702, with permission from the American Society for Pharmacology and Experimental Therapeutics.

-i—i—i—i—i—<— -1—i—i—i—,—i—r-

-10-9 -B -7 -6 -5 -10-9 -8 -7 -6 -5 -4 log [oxymetazoline] (M) log [agonist] (M)

-i—i—i—i—i—<— -1—i—i—i—,—i—r-

-10-9 -B -7 -6 -5 -10-9 -8 -7 -6 -5 -4 log [oxymetazoline] (M) log [agonist] (M)

Figure 202 Effect of 'a2-adrenergic agonists' in CHO cells with porcine a2A-receptors at high (a2A-H cells) and low (a2A-L cells) concentrations, respectively. Adenylate cyclase stimulation is measured in the presence of pertussis toxin and is therefore Gs-mediated. Adenylate cyclase inhibition is Gi-mediated. Reproduced from Brink, C. B., Wade, S. M. and Neubig, R. R. (2000) Journal of Pharmacology and Experimental Therapeutics, 294, 539-547, with permission from the American Society for Pharmacology and Experimental Therapeutics.

by endogenous 5-HT1B receptors (i.e. it can be blocked by the 5-HTj antagonist (—)-cyanopindolol). This result strongly emphasises the importance of non-transfected control cells when studying the pharmacological properties of recombinant systems.

Multistate receptors: ligand-mediated sequential changes in receptor conformation

Real-time fluorescence spectroscopy of purified ^-adrenergic receptors labelled with a conformationally sensitive fluorophore (i.e. with a cysteine-reactive, fluorescent probe whose fluorescence is highly sensitive to the polarity of its environment) revealed agonist-induced conformational changes with a t1/2 of 2-3 min. This suggests that the rapid association of agonists is followed by a slower conformational change of the receptor. These findings led to the 'sequential binding and conformational selection' model by Gether and Kobilka (1998). It is assumed that the receptor spontaneously alternates between different active and inactive conformations (Figure 203) and that receptor activation by an agonist occurs sequentially, resulting in a series of intermediate con-formational states (Ra' and Ra") between R and Ra*:

• Agonist binding may involve an initial interaction between receptor and one structural group of the agonist. After this initial event, binding of the remaining groups occurs in a sequential manner. Each interaction between the receptor and the agonist stabilizes one or more transmembrane domain until the agonist finally stabilizes the receptor in the active Ra* state.

• Ri' can be stabilized by inverse agonists in a similar way to Ra* stabilization by agonists.

Figure 203 Sequential binding and conformational stabilization model for the molecular mechanisms of ligand action in GPCRs. Reprinted from Fundamental and Clinical Pharmacology, 19, Vauquelin G. and Van Liefde I. G protein-coupled receptors: a count of 1001 conformations, 45-46, Copyright (2005) Blackwell Publishing.

Figure 203 Sequential binding and conformational stabilization model for the molecular mechanisms of ligand action in GPCRs. Reprinted from Fundamental and Clinical Pharmacology, 19, Vauquelin G. and Van Liefde I. G protein-coupled receptors: a count of 1001 conformations, 45-46, Copyright (2005) Blackwell Publishing.

• Partial agonists may stabilize one of the intermediate states, thereby increasing the chance of spontaneous isomerization to Ra*. Alternatively, they may stabilize unique conformational states having lower affinities for the G protein.

The model is strongly supported by recent studies with angiotensin II (Ang II) analogues. These studies suggest that at least two steps take place to obtain full receptor activation (Figure 204):

Figure 204 Activation of the wild-type ATj receptors (top) and of the AsnJ11Glu ATt receptor CAM by Ang II, Ang III and Ang IV (Le et ai, 2002, reproduced by permission of the American Society for Biochemistry and Molecular Biology).

• A pre-activation step, in which intramolecular interactions constrained within the receptor are broken by Arg2 of angiotensin II. This explains why the potency of Ang IV (i.e. an Ang II fragment without the N-terminal Asp1 and Arg2) is 1000 times lower than the potency of Ang II and Ang III (i.e. an Ang II without Asp1).

• A subsequent activation step in which the C-terminal side of Ang II plays an essential role. Here, Arg2 is no longer needed and this explains why both Ang II and Ang IV have high potency for the Asn111Glu constitutively active AT; receptor mutant (which is assumed to mimic the pre-activated wild-type receptor).

A particularly striking observation with AT1 receptors and many other GPCRs is that the apparent affinity and efficacy of agonists is better for CAMs than for the wild-type receptor. When the WT receptor has very low basal activity, certain CAMs may show hardly detectable constitutive activity in the absence of agonist. Yet, such CAMs could still be detected based on their increased affinity and efficacy for certain agonists. For example, only one AT1 receptor CAM was identified by site-directed mutagenesis (N111A). This CAM showed increased apparent affinity and efficacy for CGP42112A and, based on the same criterion, several other AT1 receptor CAMs were identified following random mutagenesis of this receptor (Figure 205).

Figure 205 Amino acid substitutions resulting in a large (large captions) or small (small captions) increase in sensitivity of ATj receptors to CGP42112A. Reprinted from Proceedings of the National Academy of Science USA, 97, Parnot, C., Bardin, S., Miserey-Lenkei, S., Guedin, D., Corvol, P. and Clauser, E., Systematic identification of mutations that constitutively activate the angiotensin II type 1A receptor by screening a randomly mutated cDNA library with an original pharmacological bioassay, 7615-7620. Copyright (2000) National Academy of Sciences, USA.

Figure 205 Amino acid substitutions resulting in a large (large captions) or small (small captions) increase in sensitivity of ATj receptors to CGP42112A. Reprinted from Proceedings of the National Academy of Science USA, 97, Parnot, C., Bardin, S., Miserey-Lenkei, S., Guedin, D., Corvol, P. and Clauser, E., Systematic identification of mutations that constitutively activate the angiotensin II type 1A receptor by screening a randomly mutated cDNA library with an original pharmacological bioassay, 7615-7620. Copyright (2000) National Academy of Sciences, USA.

Figure 206 Biphenyltetrazole AT1 receptor antagonists. Left: Pro-drugs with masked carboxyl group only undergo fast reversible binding. Right: the active drugs possess an exposed carboxyl group, which is likely to be involved in tight antagonist binding.

Antagonist-ATj receptor complexes have also been shown to adopt at least two distinct states (Figure 208). One state is formed swiftly and binding of the antagonist is fast and reversible, while the other state is formed more slowly (presumably by isomer-ization of the quickly reversible complex) and antagonist binding is much tighter. At equilibrium, both states co-exist and their ratio depends on chemical properties of the bound antagonist:

• Antagonist structure-activity relationship studies indicate that those with a carboxyl group in addition to their acidic tetrazole group (like candesartan and EXP3174) are most prone to forming tight binding complexes (Figure 206).

• Receptor mutation studies in which basic amino acids are replaced by neutral ones indicate that Lys199 at TM5 of the receptor is important for the recognition of the carboxyl group of these antagonists (i.e. replacing Lys199 by the polar, but non-charged Gln produces a much larger drop in affinity for candesartan and EXP3174 as compared to losartan) (Figure 207).

0 0

Post a comment