Domain swapping in G-protein coupled receptor dimers
 

Christopher A. Reynolds

Department of Biological Sciences, Central Campus, Wivenhoe Park, Colchester, Essex, CO4 3SQ, United Kingdom

with assistance from Christopher R. Snell, Robert P. Bywater and Paul R. Gouldson
(other affiliations given below)


 





KEYWORDS: agonist, antagonist, inverse agonist, GPCR, G-protein, constitutive acctivity, auto activation, R*, high affinity binding, domains, domain swapping, dimerisation, chimeric receptors, co-expression, bell-shaped dose-response curves, functional rescue, site-directed mutagenesis, molecular modelling, molecular dynamics, brownian dynamics, correlated mutation analysis.

Introductory animation


An animation of the domain swapping dimerisation, viewed from the extracellular side of the membrane. Each transmembrane alpha helix is represented as a coloured circle. Each monomer has one domain coloured in red and one in blue, with the hinge loop (intracellular loop 3) coloured in green. Notice that in the domain swapped dimer each 7 TM bundle is completely red or completely blue because the monomers have swapped domains.  Notice that the rearrangement can occur essentially without the A domain (helices 1-5) changing its structure.

return to Chris Reynolds' home page


Contents
 

First known reference to domain swapping in GPCRs.

What is domain swapping?

Evidence for domains in GPCRs

Possible reasons for domain swapping in GPCRs (as opposed to other modes of dimerisation)

Receptors which show some evidence of dimerisation.

Evidence for domain swapping in GPCRs

Evidence for dimerisation in GPCRs

The oyster club

Other research groups interested in GPCR dimerisation

Acknowledgments


First known reference to domain swapping in GPCRs.

As far as we are aware, these are the first references to domain swapping in GPCRs.

Prior to this, domain swapping in GPCRs was present by C.A. Reynolds at the following meetings. The main article on domain swapping in GPCRs at the initiation of this web page is This article should be cited for information taken from this web page.

top

What is domain swapping?
 

Domain swapping is a very efficient method of forming oligomers since the interactions within the monomer are reused in the dimer. There is thus no need to evolve a new site on the surface which in one monomer mutually recognises the corresponding site on the other momomer, since in the domain swapped dimer the recognition requirement has already largely been accounted for.A schematic diagram is shown below, along with a diagram to show the proposed domain swapping mechanism in GPCRs.

General references on domain swapping in other systems.

top

Evidence for domains in GPCRs
 


GPCRs may be cut at the points marked by the blue arrows. If the fragments are expressed separately, it is possible, depending on the receptor, that a functional receptor may be produced. On the whole however, it appears that GPCRs are most tolerant to this splitting only if split in intracellular loop three which lies between helices 5 and 6. Active (in terms of binding and/or coupling) combinations include:
 
 
1-2 + 3-7 
1 + 2 + 3-7 
1-3 + 4-7 
1-5 + 6-7 
1-3 + 4-5 + 6-7 
1-5 + 6-7 
1-5 + 6-7 
1-3 + 4-7 
1-4 + 5-7 
1-5 + 6-7 
1-5 + 6-7 
1-5 + 6-7		 Bacteriorhodopsin 

Rhodopsin 
 
 

beta2-adrenergic 
M2 muscarininc 
 

V2 vasopressin 
GnRH
Selected references are given below, the above information was extracted from Guderman et al., (1997)

top

Possible reasons for domain swapping in GPCRs (as opposed to other modes of dimerisation)


The left hand monomer contains a fatal mutation on the A domain (helices 1-5), the second monomer contains a fatal mutation on the B domain (helices 6-7). When these are coexpressed, the fatal mutations are confined to one 7-helix bundle, leaving the other intact. This combination of receptors should be fully functional. This mechanism will not protect from constitutively active mutations or mutations which prevent expression on the cell surface. The best example of where this mechanism will not work is in the X-linked vasopressin receptor.
top

Receptors which show some evidence of dimerisation.

top

Evidence for domain swapping in GPCRs

The primary evidence for domain swapping comes from Maggio's excellent work on the coexpression of chimeric receptors, as illustrated below. Some of the other evidence for GPCR dimers given below also points towards domain swapping. These articles are denoted DS. Much of the evidence involves functional rescue experiments which may or may not apply to the normal receptors, these are denoted FR. However, if domain swapping can occur in functional rescue of defective receptors, there is no reason why it should not also occur in the normal pathways. Evidence for domain swapping/dimerisation can also be inferred from the work of Monnot.


 
 

top

Evidence for dimerisation in GPCRs

The evidence listed below has not been sorted in a way to reflect its importance. For a number of these articles, interpretations other than dimerisation may offer equally valid explanations. There is a user submission form below - if readers wish to submit additional data in the form given below, it will be entered on the web.

Molecular Dynamics and Brownian Dynamics simulations (Gouldson, Kamiya; 1997/8)

The simulations on the dimeric peptides is related to the work of Wade (see below). The remaining molecular dynamics simulations explore the energies of various dimers, including the domain swapped dimer, in the presence and absence of agonist and antagonist. The simulations are consistent with, but do not prove, the idea that the active high affinity form of the receptor is the domain swapped dimer, and that the role of agonist is to shift the equilibrium in favour of the domain swapped dimer.
    The Brownian dynamics simulations explore the gating mechanism of the extracellular loops and show that the loop gating may not be sufficient in itself to permit entry of the ligand. The inference is that some kind of movement similar to domain swapping may be required to create a sufficiently large opening for the ligand to bind.

Fifty percent of receptors in the high affinity form (Potter 1991)

Potter observed that only 50 % of muscarinic M2 receptors were in their high affinity forms at any one time. It is the 50 % that points towards dimerisation.

Autoactivation of GPCRs when overexpressed in transgenic mice

This may be interpreted by assuming that the basal activity of G-protein coupled receptors is related to the rate of dimer formation and that dimer formation is greatly assisted by agonist binding or a higher concentration of receptors.

Activation by bivalent antibodies and by antagonists coupled back to back

Given the view that agonism follows dimer formation, it is relatively straightforward to explain the activation of receptors by both bivalent antibodies and by antagonists coupled back to back since they will assist in bringing two monomers together.

Dimers of peptides from intracellular loop three shown strong activation

This is consistent with this view of receptor activation as it suggests that the G-protein requires two copies of this loop. Moreover, this latter observation supports the idea that the active dimer is the 5,6-domain swapped dimer rather than say the 1,2-dimer as observed in the rhodopsin two dimensional projection maps. The two copies of intracellular loop three are about 70 Å apart in the projection maps while the linker is merely a disulphide bridge. However, the two copies of intracellular loop three move much closer together in the 5,6-domain swapped dimer (by about 40 Å  - see Gouldson, P.R., Reynolds, C.A. (1997) Biochem. Soc. Trans., 25, 1066-1071).

The two-dimensional electron microscopy projection maps contain dimers

The article by Corless explicitly mentions the occurrence of dimers, but dimers are visible in other projection maps.

Other receptors, e.g. receptor tyrosine kinases, dimerise

This is very circumstantial evidence but it would be very interesting if GPCRs were unusual in functioning as monomers!

Bell-shaped dose response curves

Bell-shaped dose response curves may have several origins, but dimerisation is one of them, as described in: De Meyts, P., Urso, B., Christoffersen, C.T. & Shymko, R.M. (1995). Mechanism of insulin and IGF-I receptor activation and signal transduction specificity - receptor dimer cross-linking, bell-shaped curves, and sustained versus transient signalling, Ann. N.Y. Acad. Sci. 766, 388-401.

The immunological studies of Ciruela

Ciruela raised antibodies against the second extracellular loop and the third intracellular loop of the A1-adenosine receptor and provided further evidence for the 5,6-domain swapped dimer. He described identification of both monomers and dimers and their interconversion by agonist or antagonist binding. However, the antibody against the third intracellular loop did not bind to the dimer. If the physiological dimer is indeed a 5,6-domain swapped dimer, then the reasons for this are apparent since the two adjacent copies of intracellular loop three would sterically hinder each other and prevent access by the antibody.

SDS-resistant dimers

It was observed that peptides derived from transmembrane helix six can inhibit G-protein dimerisation and activation; additional observations: inverse agonists favour reversal of dimer formation, other helix 6 peptides do not inhibit.

Peptides derived from helix seven have been shown to inhibit dimerisation

(This is similar to Hebert's work but not as extensive - the experiments on activation were not done).

Genetic information

Highly correlated patterns of change (Correlated mutations) exist amongst the external (lipid-facing) residues. If these external residues had no function, then such patterns would not be anticipated. Evidence that correlated mutations do occur at protein/domain interfaces has been provided by Pazos on 20 proteins (Pazos, F., Helmer-Citterich, M., Ausiello, G., Valencia, A. (1997) J. Mol. Biol., 271, 511-523.) and by ourselves for the MHC class II dimer of heterodimers (Nilsson, A., Wijayawardene, M., Gkoutos, G., Wilson, K.M., Fernandez, F., Reynolds, C.A., submitted.)

Site-directed mutagenesis:

Mutation of Tyr 205 in the neurokinin NK-1 receptor results in loss of (this is thought to be an external residue though the authors do not identify it as such). Similar results were obtained for the corresponding Tyr 206 in the neurokinin NK-2 Neither of these mutations affected antagonist binding, and this is consistent with, but not proof of, the idea that the mutation affects helix-helix packing at the dimer interface. Similarly, mutation of Gly 276, Gly 280 and Leu 284 to alanine in Hebert's helix six peptide (see above) significantly reduced its ability to inhibit dimerisation; again, these correspond to external residues.

Evidence for heterodimerisation

A number of researchers have discussed this topic, e.g. at the UPpsala heptahelical receptors meeting in 1996. The following are probably the most prominent articles to date:
 

GABABR1 and GABABR2 were found not to function unless they were coexpressed.
 
 

Dimers and receptor deactivation

Here it is suggested that agonist decreases the concentration of dimers. This is in contrast to the work of Hebert (1996) on the b 2-adrenegric receptor which showed that agonist increased the dimer concentration. There are a number of possible explanations for these differing conclusions. For example, the opioid receptor couples to a different G-protein.

Since Cvejic and Devi did not carry out any dose response experiments, we cannot rule out the possibility that they are observing the tail end of a bell-shaped dose response curve (see their figure 2). Either way, both groups observe that dimerisation is important.

Disulphide-linked GPCR dimers

top

Agonist induced increase in receptor size

Agonist, but not antagonist causes an increase in the receptor size; other explanations besides receptor aggregation are given.

top

Activity of bivalent drugs

Here is yet another example of bivalent ligands which may exhibit greater affinity and selectivity for a receptor subtype over their monomer counterparts.

top

Reader supplied evidence for dimerisation

To submit your own evidence for dimerisation, please email C.A.Reynolds@essex.ac.uk, with:

top

The oyster club

The oyster club is a goup of scientists who meet informally to discuss domain swapping in GPCRs. The founding members are

<Picture of an oyster smack rather than an oyster catcher?>
top

Other research groups interested in GPCR dimerisation

under construction

top
 

Acknowledgments

We wish to acknowledge the following for supporting our domain swapping research.

Novo Nordisk, the BBSRC (B/06081) and the EPSRC (94309861).

This page was last modified in December 1998

top
return to Chris Reynolds' home page