GPCR dimerization

GPCRs are traditionally regarded to exist and to be fully functional as monomers. Yet, recent findings suggest that they also exist as homo- as well as heterodimers (i.e. dimers between, respectively, the same and different receptor molecules). As dimerization is often observed in recombinant cell systems, there is a major concern that this process could be due to receptor overexpression and, hence, about the physiological relevance of this process. However, dimerization has also been observed when the receptors are expressed at endogenous levels and many other proteins exist and function as dimers (e.g. tyrosine kinase receptors, transcription factors and steroid receptor). GPCR dimerization also shows some specificity: i.e. homodimers and heterodimers containing closely related GPCRs appear to be more easily formed than heterodimers between distantly related GPCRs. The importance of heterodimerization in ligand recognition and signalling is further exemplified by the y-aminobutyric acid-binding GABAb receptor (Figure 139). It is only functional as a heterodimer between GABABR1 and GABABR2 (two naturally occurring

External interactions • [igand bindinq

• localization associations

C3 loop • signalling ■ cytoskeletal associations

Figure 141 Interactions involving distinct GPCR domains that affect localization, signalling and pharmacological properties of the receptor. Reprinted from Trends in Pharmacological Sciences, 22, Milligan, G. and White, J. H., Protein-protein interactions at G protein-coupled receptors, 513-518, © (2001), with permission from Elsevier.

• localization

Intracellular interactions

* signalling

* trafficking

• localization

• cytoskeletal

C3 loop • signalling ■ cytoskeletal associations associations

Figure 141 Interactions involving distinct GPCR domains that affect localization, signalling and pharmacological properties of the receptor. Reprinted from Trends in Pharmacological Sciences, 22, Milligan, G. and White, J. H., Protein-protein interactions at G protein-coupled receptors, 513-518, © (2001), with permission from Elsevier.

non-functional GPCRs). GABABR1 alone displays low agonist affinity and does not traffic to the cell surface because it possesses an endoplasmic reticulum retention motif. This motif gets masked upon heterodimerization with GABABR2 (upon co-expression). The heterodimer now reaches the cell surface and is functionally active. In the same line, native tissue metabotropic glutamate mGlu5 receptors exist almost exclusively as homodimers.

Biochemical and biophysical techniques have been used to identify GPCR dimers:

• Radiation inactivation (target size analysis) was used in the eighties. This technique is based on the inverse relationship between the size of a macromolecule and the dose-dependent inactivation of that macromolecule by ionizing radiation. These studies suggested that, for example, P2-adrenergic receptors had a higher molecular mass than expected and hence, could be dimers.

• The use of cross-linking agents and/or photoaffinity labelling reagents followed by purification of the radiolabelled complex using gel exclusion chromatography.

• Differential epitope tagging and selective immunoprecipitation have been an invaluable tool to provide biochemical evidence for the presence of GPCR dimers. Using this technique, the two GPCRs under investigation are each tagged with a distinct epitope and expressed in heterologous cells (that do not normally express these receptors). Antibodies to one epitope are used to immuno-isolate

Figure 142 Cells were co-transfected with p2-adrenergic and FLAG-tagged K-opioid receptors. Samples were immunoprecipitated with anti-(32-adrenergic receptor antibody, resolved by SDS/ PAGE and immunoblotted with anti-FLAG antibody. Reproduced with permission, from Ramsay, D., Kellett, E., McVey, M., Rees, S. and Milligan, G., 2002, Biochemical Journal, 365, 429-440. © The Biochemical Society.

Figure 142 Cells were co-transfected with p2-adrenergic and FLAG-tagged K-opioid receptors. Samples were immunoprecipitated with anti-(32-adrenergic receptor antibody, resolved by SDS/ PAGE and immunoblotted with anti-FLAG antibody. Reproduced with permission, from Ramsay, D., Kellett, E., McVey, M., Rees, S. and Milligan, G., 2002, Biochemical Journal, 365, 429-440. © The Biochemical Society.

the receptor-containing complex, and the associating receptor in the complex is visualized using antibodies to the second epitope (Figure 142). A major concern with this technique is the possibility of artifactual aggregation (due to the inherent hydrophobic nature of GPCRs) under the solubilization/immunoprecipitation conditions. To ensure that dimers are not induced, cells individually expressing differently tagged receptors can be mixed prior to solubilization and immunoprecipitation. Under these conditions, dimers should be observed only in cells co-expressing the two receptors and not in the mixture of cells.

• The bioluminescence resonance energy transfer (BRET) (Figure 143) technique involves the emission of light upon the catalytic degradation of a suitable substrate by a (receptor-coupled) enzyme. This light can activate an energy acceptor (such as green fluorescent protein (GFP) coupled to a second receptor), resulting in its fluorescence. Fluorescence can only be measured if the energy donor and acceptor are located within relatively small distances (i.e. 50 A = one receptor radius) of each other. However, this does not necessarily imply a physical interaction. Whereas BRET uses the oxidation of coelenterazine by Renilla luciferase to excite the fluorescence of yellow fluorescent protein (YFP), a recent variant (BRET2) (Figure 143) has been introduced in which Renilla luciferase oxidizes a modified form of coelenterazine to excite the fluorescence of a GFP mutant (designated GFP2). The bioluminiscence and the fluorescence spectra are more effectively resolved in this newer assay so that the fluorescence signal becomes more distinct.

• The fluorescence resonance energy transfer (FRET) (Figure 144) technique is similar to BRET, with the exception that the energy donor molecule, a variant of GFP (generally, cyan fluorescent protein (CFP)) is excited by an external

Principle Luciferase Gfp Renilla

Figure 143 Top: Principle of the BRET2 technique. Bottom: Luminescence and fluorescence spectra for the traditional BRET and the more recent BRET2 approach. Reproduced with permission, from Ramsay, D., Kellett, E., McVey, M., Rees, S. and Milligan, G., 2002, Biochemical Journal, 365, 429-440. © The Biochemical Society.

Figure 143 Top: Principle of the BRET2 technique. Bottom: Luminescence and fluorescence spectra for the traditional BRET and the more recent BRET2 approach. Reproduced with permission, from Ramsay, D., Kellett, E., McVey, M., Rees, S. and Milligan, G., 2002, Biochemical Journal, 365, 429-440. © The Biochemical Society.

light source. The energy emitted by CFP is used to excite an acceptor molecule, another variant of GFP (generally, yellow fluorescent protein (YFP)). Here again, fluorescence can only be measured if the energy donor and acceptor are located within relatively small distances (i.e. 100 A = 2 X receptor radius) of each other.

From the evidence gathered thus far, it appears that some GPCRs are assembled as dimers in the endoplasmic reticulum whereas others assemble to dimers (or even oligomers) at the cell surface. Also the effect of agonists on receptor di-/oligomers is quite variable. There are at least two possible scenarios for the assembly and maturation of the GPCR dimer (Figure 145):

Gpcr Forms Dimer
Figure 144 Principle of the FRET technique.

Figure 145 Subcellular locations where GPCR dimers are thought to form. (A) Receptor dimerization in the endoplasmic reticulum and transport as dimers to the cell surface. (B) Receptor di- or multimerization at the cell surface under the influence of agonists. Reprinted from Pharmacology and Theraputics, 92, Rios, C. D., Jordan, B. A., Gomes, I. and Devi, L. A., G protein-coupled receptor dimerization: modulation of receptor function, 71-87. Copyright (2001), with permission from Elsevier.

Figure 145 Subcellular locations where GPCR dimers are thought to form. (A) Receptor dimerization in the endoplasmic reticulum and transport as dimers to the cell surface. (B) Receptor di- or multimerization at the cell surface under the influence of agonists. Reprinted from Pharmacology and Theraputics, 92, Rios, C. D., Jordan, B. A., Gomes, I. and Devi, L. A., G protein-coupled receptor dimerization: modulation of receptor function, 71-87. Copyright (2001), with permission from Elsevier.

Figure 146 Top: Co-expression of HA-tagged D3 dopamine receptor mutants (lacking TM6 and TM7) and FLAG-tagged wild-type D3 receptors in the same cells. Both form heterodimers (co-immunoprecipitation experiments) and show a similar cytosolic localization (confocal laser microscopy shown here). Bottom: Control experiment. All receptors reside at the cell surface when HA- and FLAG-tagged wild-type D3 receptors are introduced into the same cell line. Reproduced from Karpa, K. D., Lin, R., Kabbani, N. and Levenson, R. (2000) The dopamine D3 receptor interacts with itself and the truncated D3 splice variant D3nf: D3-D3nf interaction causes mislocalization of D3 receptors. Molecular Pharmacology, 58, 677-683, with permission from the American Society for Pharmacology and Experimental Theraputics.

Figure 146 Top: Co-expression of HA-tagged D3 dopamine receptor mutants (lacking TM6 and TM7) and FLAG-tagged wild-type D3 receptors in the same cells. Both form heterodimers (co-immunoprecipitation experiments) and show a similar cytosolic localization (confocal laser microscopy shown here). Bottom: Control experiment. All receptors reside at the cell surface when HA- and FLAG-tagged wild-type D3 receptors are introduced into the same cell line. Reproduced from Karpa, K. D., Lin, R., Kabbani, N. and Levenson, R. (2000) The dopamine D3 receptor interacts with itself and the truncated D3 splice variant D3nf: D3-D3nf interaction causes mislocalization of D3 receptors. Molecular Pharmacology, 58, 677-683, with permission from the American Society for Pharmacology and Experimental Theraputics.

• (A) GPCRs assemble in an intracellular compartment and are shuttled to the cell surface as dimers (e.g. GABABR1 and GABABR2). In this respect, it has also been found for several GPCRs that co-expression of a mutant truncated at the N-and C-termini with wild-type receptors may inhibit the trafficking of the latter to the cell surface (Figure 146).

• (B) GPCRs are synthesized in the endoplasmic reticulum (ER) and shuttled to the cell surface as monomers, where they may assemble as dimers in response to agonists.

Also, the interactions that hold a dimer together seem to differ considerably among the receptors. It can be mediated either by covalent (disulfide) and/or non-covalent (hydrophobic) interactions, and can involve associations of different receptor domains. e.g.:

• Dimerization of the metabotropic glutamate receptors was found to be dependent on the formation of disulfide bonds between cysteines in their large amino-terminal domains.

• The involvement of TM6 and TM7 has been implicated for D2 dopamine receptor dimerization since peptides encoding these regions inhibit dimer formation.

• For the 8-opioid receptor, dimerization was eliminated by deletion of 15 amino acids at the C-terminus, indicating the involvement of this part of the receptor in dimerization. C-terminal regions also participate in GABAbR1-GABAbR2 receptor heterodimerization.

Some receptors appear to swap some of their TM domains during the heterodimeri-zation process. Evidence of this 'domain swapping' is provided by functional rescue experiments:

• When TM6 and TM7 are exchanged between a2C adrenergic and muscarinic receptors, the receptor mutants are not able to bind their initial radioligands when expressed separately. However, both a2C adrenergic and muscarinic radioligands are able to exhibit significant binding to cells where both mutants are co-expressed.

• Co-expression of truncated P-adrenergic receptors (one with TM 1 to TM5 and the other with TM 6 to TM7) restores its functional (i.e. adenylate cyclase stimulation) and binding activities.

To explain how binding pockets are recovered in such experiments, a model was proposed (Gouldson et al., 1997) in which, upon receptor dimerization, the original binding pockets of the two subunit monomers are replaced by two binding pockets with similar structures except that they are formed from regions donated by both monomers (Figure 147).

Dimerization may alter how a receptor binds and functionally responds to a ligand but there is no evidence supporting a universal role of dimerization for GPCR activation (Table 15):

• In some cases the physical interaction between GPCRs leads to functional activation or enhanced functional activity. In other cases such an interaction appears to lead to functional receptor inactivation.

TM1 —TM5 with TM6-TM7 TM1-TM4 with TM5-TM7

TM1 —TM5 with TM6-TM7 TM1-TM4 with TM5-TM7

Figure 147 Possible formation of new binding pockets by GPCR transmembrane domains upon dimerization. Reproduced from Lee, S. P., O'dowd, B. F., Ng, G. Y. K., Varghese, G., Akil, H., Mansour, A., Nguyen, T. and George, S. R. (2000) Inhibition of cell surface expression by mutant receptors demonstrates that D2 dopamine receptors exist as oligomers in the cell. Molecular Pharmacology, 58, 120-128, with permission from the American Society for Pharmacology and Experimental Theraputics.

Figure 147 Possible formation of new binding pockets by GPCR transmembrane domains upon dimerization. Reproduced from Lee, S. P., O'dowd, B. F., Ng, G. Y. K., Varghese, G., Akil, H., Mansour, A., Nguyen, T. and George, S. R. (2000) Inhibition of cell surface expression by mutant receptors demonstrates that D2 dopamine receptors exist as oligomers in the cell. Molecular Pharmacology, 58, 120-128, with permission from the American Society for Pharmacology and Experimental Theraputics.

Table 15 Interacting GPCRs. Reprinted from Cellular Signalling, 14, Brady, A. E. and Limbird, L. E., G protein-coupled receptor interacting proteins: Emerging roles in localization and signal transduction, 297-309. Copyright (2002), with permission from Elsevier.

GPCR heterodimerization

Functional role(s) implicated

CCR2

CCR5

Required for activation of receptor

associated JAK kinase, presumably

due to transphosphorylation synergistic

activation of Ca2+ responses by MCP-1

(CCR2) and RANTES (CCR5) ligands

Heterodimerization also recruits dissimilar signalling pathways

GABAB R1

GABAB R2

Cell surface expression; pharmacologically

appropriate ligand binding; coupling to

GIRK/Kir3 currents

K-opioid receptor

8-opioid receptor

Modified pharmacological profile

^-opioid receptor

8-opioid receptor

Potentiation of m agonist signals by low

concentrations of 8 -selective agonist

A1 adenosine receptor

D1 receptor

Uncoupling of D1 receptor from

activation of adenylate cyclase

when A1 and D1 receptors are

simultaneously activated

SSTR1

SSTR5

Dimerization induced by agonist;

required for receptor activation;

promotes internalization

D2 receptor

SSTR5

Creates novel receptor with enhanced

affinity for dopamine and SST

agonists; enhanced coupling to

inhibition of adenylate cyclase

AT1 receptor

Bradykinin B2

Increases action of Gq and Gi receptor

receptor

activation in heterologous and smooth

muscle cells; modulates endocytic

trafficking

GPCR Homodimerization

Functional role(s) implicated

K-opioid receptor

Function unknown

8-opioid receptor

Increasing levels of agonist causes dimer

to monomer transition, which appears

to induce receptor internalization

P-adrenergic receptor

BRET signal is constitutive and not

regulated by agonist Agonist-

facilitated; wild-type receptors have

dominant positive effect on some

mutant P2-adrenegic receptors

mGlu receptor

Occurs in ER and does not require

glycosylation

M3 muscarinic receptor

Function unknown

Table 15 (Continued)

GPCR Homodimerization

Functional role(s) implicated

Ca2 +-sensing receptor

Increases affinity for extracellular Ca2+;

accelerates rate of agonist-elicited response

D2 receptor

Dimers are targets for nemonapride;

monomers are targets for spiperone;

disrupted by peptides derived from D2 R

transmembrane regions

STE2 (yeast a -mating

Exist as dimers constitutively; dimers also

factor receptor)

exist during endocytosis TRH receptor

Agonist stimulates BRET signal;

constitutive oligomerization

LH receptor

Receptor oligomerization increases to a

greater extent with hCG occupancy than

with LH occupancy

• In the same line, agonists can stabilize the dimeric form (e.g. for p2-adrenergic receptors) or decrease the level of dimer formation (e.g. for 8-opioid receptors).

• Finally, such interaction may result either in an increased or decreased (e.g. for the p2-adrenergic-K-opioid receptor combination) desensitization and internali-zation of the receptors.

Of special pharmacological interest is that the heterodimerization of some GPCRs may result in the formation of novel recognition sites with a completely distinct pharmacological profile. For example, heterodimers with k- and 8-opioid receptors have no significant affinity for any of the subtype-selective ligands, but possess high affinity for ligands that are less subtype-selective. Moreover, a specific ligand for one receptor in a heterodimer can alter the binding of a specific ligand for the other receptor. For example, the affinity of an SSTR5-selective agonist for its receptor is affected by dopaminergic ligands when D2 dopamine receptors are co-expressed (Figure 148).

0 0

Post a comment