Principles
Kinetics versus equilibrium: the importance of GTP in GPCR activation

https://doi.org/10.1016/S0165-6147(99)01394-2Get rights and content

Abstract

Agonist-bound G-protein-coupled receptors (GPCRs) facilitate GDP–GTP exchange on their cognate G proteins. The binding properties of GPCRs are adequately described by the ternary complex model. However, in this article a more realistic (steady-state) model, which is necessary to describe the catalytic effect of agonist-bound receptors on G-protein activation, will be discussed. This model predicts that agonist potency and efficacy might vary from tissue to tissue, depending on the G-protein concentration and can be extended to explain why an agonist’s ability to increase the receptor’s affinity for empty G proteins (in the absence of GTP) is related to the agonist’s efficacy.

Section snippets

Descriptive versus mathematical models

To verify the validity of descriptive models of ligand–receptor interactions, it is necessary to translate them into mathematical equations and then test their predictions against experimental observations. The ternary complex model6 shown in the inset of Fig. 1 is adequate for analysis of binding studies; in the absence of GTP, it predicts the existence of two agonist-binding states, HRG (agonist–receptor–G-protein) and HR, with dissociation constants KH = 1/αK and KL = 1/K, respectively

Ternary complex model: prediction of agonist potency

The ability of the ligand to stabilize HRG is often considered to be equivalent to its ability to activate G proteins7, 18, 19. The ternary complex equations6, 7, 16, 17, 18, 19 consider only three G-protein species [G, RG (receptor–G-protein) and HRG]. By using these equations to describe G-protein activation by GTP, it is implicitly assumed that GTP binding does not significantly affect the G-protein population available to the receptor (i.e. the activated G-protein concentration remains very

Alternative equilibrium models: the quaternary and quinternary complex models

The ternary complex model has been extended to include recognition of guanyl nucleotides by the G protein20 (Fig. 1) and the effect of the nucleotides on the interaction of Gα–Gβγ subunits21. Allosteric interactions are always reciprocal at equilibrium; because GTP inhibits agonist binding6, 22, 23, agonists must decrease the affinity of GTP for RG (Fig. 1). G proteins must be associated with a receptor before the agonist can elicit an effect; agonists cannot affect the nucleotide-binding

The Cassel–Selinger G-protein activation cycle

As shown in Fig. 2, the G-protein–GTP interaction is usually described as a one-way cycle: GTP binding is followed by its rapid hydrolysis by the G protein, and the subsequent GDP dissociation is very slow. Let kon and kcat represent GTP association and hydrolysis rate constants, respectively, and koff represent GTP dissociation. GTP binding at steady state:

GGTP*=kon[G](koff+kcat)[GTP]must always be lower than expected from its affinity for the G protein (KGTP = kon/koff). If kcatkoff, even if

A kinetic model of G-protein activation

G-protein activation by GPCRs is a two-step reaction (Fig. 3)1, 24: (1) the recognition of inactive G-protein and GDP release is followed, in the presence of GTP, by (2) GTP binding and activated G-protein release. Several authors developed the equations that describe the rate of G-protein activation (G*GTP accumulation)9, 24, 25, 26, 27, 28. The equations predict that agonists catalytically induce G-protein activation and that the agonists’ EC50 values depend on G-protein concentration9, 24, 25

Are the kinetic equations more complicated than the ternary complex equations?

If X represents the agonist, G-protein or GTP concentration (all other parameters being held constant), the equations24 that describe the G-protein activation rate as a function of X have the same form as the ‘saturation curve’ equation B = (RtotF)/KD + F) (where Rtot is the density of the binding site, F is the free ligand concentration and KD is the equilibrium dissociation constant). However, Rtot must be replaced by (Vmax)app, which is the rate of G-protein activation observed when X is

Definition of Vmax and agonist efficacy

If the G-protein and GTP concentrations are saturating, all transiently uncoupled receptors (i.e. R and HR) will immediately encounter a new inactive G protein (GGDP) and any transiently emptied G proteins (i.e. RG and HRG) will immediately be re-occupied (by GTP). The G-protein activation cycle (Fig. 3) will ‘turn’ faster if the intermediate complexes HRGGDP and HRGGTP do not accumulate; that is, if the GDP and activated G-protein dissociation rate constants k2 and k4 are large24. Agonists and

Comparison of the EC50 values expected in binding and functional studies

It would be very helpful if binding and functional studies yielded the same information (i.e. if the GTP, G-protein and agonist ‘affinities’ expected in binding and functional studies were identical). Unfortunately, as outlined below, the kinetic models of G-protein activation predict that this is not the case.

Free receptors have been assumed to support G-protein activation, with agonists and inverse agonists merely increasing or decreasing their activity. It is therefore necessary to define

A hypothetical thermodynamic model of G-protein activation

Two hypothetical free-energy profiles29, 30, 31, which describe the receptor-catalysed G-protein activation reaction at saturating GTP concentrations are shown in Fig. 4. The inactive G-protein (GGDP) concentration might be either rate limiting (Fig. 4a) or saturating (Fig. 4b), depending on the receptor and tissue studied. RG and HRG probably do not accumulate in the presence of GTP; they are poorly represented in the reaction medium. Merely stabilizing RG relative to RGGDP is sufficient to

Concluding remarks

Non-equilibrium models are necessary to explain the observations that agonists catalyse GTP binding whereas GTP inhibits agonist binding, and that agonists activate G proteins either through ‘high-affinity’ or ‘low-affinity’ receptors, depending on the receptor and tissue studied. These models predict that G-protein activation is fastest if the quaternary complexes HRGGDP and HRG*GTP are unstable so that GDP and the activated G protein (G*GTP) dissociate rapidly from the ternary complex and

Acknowledgements

Supported by grant 3.4504.99 from the Fonds de la Recherche Scientifique Médicale and by an ‘Action de Recherche Concertée’ from the Communauté Française de Belgique.

References (31)

  • E.J.M. Helmreich et al.

    Biochim. Biophys. Acta

    (1996)
  • H.E. Hamm

    J. Biol. Chem.

    (1998)
  • A.R. Burt

    J. Biol. Chem.

    (1998)
  • A. De Lean et al.

    J. Biol. Chem.

    (1980)
  • E. Ciccarelli

    Regul. Pept.

    (1994)
  • C.K. Mak

    Neuropharmacology

    (1996)
  • J.A. Stanton et al.

    Eur. J. Pharmacol.

    (1997)
  • R.J. Lefkowitz et al.

    Trends Pharmacol. Sci.

    (1993)
  • J.M. Weiss et al.

    J. Theor. Biol.

    (1996)
  • J.M. Weiss et al.

    J. Theor. Biol.

    (1996)
  • M. Waelbroeck et al.

    J. Theor. Biol.

    (1997)
  • A.M. Krumins et al.

    Biochem. Pharmacol.

    (1997)
  • D. Mackay

    J. Theor. Biol.

    (1990)
  • W.K. Vogel et al.

    J. Biol. Chem.

    (1995)
  • L. Shea et al.

    Biochem. Pharmacol.

    (1997)
  • Cited by (0)

    View full text