Several experimental strategies have been employed to analyze the selectivity of receptor-G protein interactions:
• Transient co-expression of individual GPCRs with different Ga subunits in cultured mammalian cells.
• Reconstitution of purified receptors and G protein subunits in artificial lipid bilayers. This represents a very difficult and laborious task. Moreover, receptor-G protein interactions are studied in a highly artificial environment.
• 'Immuno-neutralization' studies have been performed with membrane preparations, as well as intact (following microinjection of the antibody) and permeabilized cells. A limitation of the use of such C-terminal Ga antibodies is their inability to discriminate between certain pairs of Ga subunits.
• Receptor-G protein interactions can also be studied via co-immunoprecipitation of receptor-G protein complexes using specific receptor or G protein-recognizing antibodies.
• Incorporation/expression of antisense oligonucleotides complementary to the mRNA for the Ga subunit of interest in cultured cells and transgenic animals. Based on the rules of complementary Watson-Crick base-pairing, an antisense oligonucleotide should bind to the mRNA in question, thereby selectively inhibiting the expression of the Ga subunit. Proper interpretation of such experiments requires demonstration that the expression of the Ga subunit of interest is truly abolished (or reduced) and that the antisense oligonucleotides do not exert non-specific effects on the expression and function of other Ga subunits.
• Traditionally, G protein-mediated responses have been classified into pertussis toxin-sensitive and pertussis toxin-insensitive responses, because of the ability of this bacterial toxin to selectively inactivate G proteins of the Gi/o family (Figure 113). This property is widely exploited to check for the involvement of Gi and Go in cellular responses. At the molecular level, pertussis toxin can catalyze the covalent binding of the ADP-ribose moiety of NAD to a conserved cysteine residue located near the C-terminus (position 4) of Ga and Gao (Figure 113). This will only occur when the G proteins are in the inactive, trimeric state. The ADP-ribosylated Gai/o subunits can no longer interact with receptors and, hence, they can no longer become activated. To determine whether a given GPCR can discriminate between individual Gai/o subunits, cultured cells can be transfected with mutant Gai/o subunits lacking the pertussis toxin-sensitive cysteine residues and then treated with pertussis toxin
to inactivate endogenous Ga^o subunits. Only the transfected Ga^o subunits will remain active. However, as the pertussis-toxin-sensitive cysteine is located within a key contact region for GPCRs, the detailed pharmacology of the GPCR response could potentially be altered by this mutation.
In addition, it is not even necessary to monitor receptor-G protein interaction by measuring a response at the level of the concerned effector component (e.g. adenylate cyclase stimulation in the case of Gs) or even downstream to it. Indeed, techniques have been developed to measure the receptor-Ga coupling and the activation of Ga subunits directly:
• Some (but not all) GPCRs acquire high agonist affinity upon coupling to a G protein. In radioligand binding studies on membranes, this results in the appearance of high-affinity sites for radiolabelled agonists and of shallow agonist/radiolabelled antagonist binding curves.
• Two binding assays on membrane preparations are based on the ability of the activated receptor to promote the exchange of tightly bound GDP by GTP. The same exchange takes place when GTP is replaced by radioactively labelled analogs like [35S]GTPyS or [a32P]GTP azidoanilide (Figures 114 and 115). Accordingly, activation of Ga by a receptor will result in an increased binding rate of these GTP analogs. Since [35S]GTPyS is resistant to the GTP-ase activity of Ga, its binding will be essentially irreversible. Upon photoactivation, [a32P]GTP azidoanilide is even capable of covalently tagging receptor-activated Ga subunits. It should be
I 1 10 100
Figure 115 In CHO cells stably expressing the ^-opioid receptor, DAMGO (= agonist) increases the binding of a-azidianiolino[32P]GTP to multiple Ga subunits. Reprinted from Journal of Neurochemistry, 64, Chakrabarti, S., Prather, P. L., Yu, L., Law, P. Y. and Loh, H. H., Expression of the mu-opioid receptor in CHO cells: ability of mu-opioid ligands to promote alpha-azidoanilido[32P]GTP labeling of multiple G protein alpha subunits., 2534-2543. Copyright (1995) Blackwell Publishing.
noted that an agonist-stimulated receptor does not increase the maximal binding of these GTP analogs, but that they only enhance their association rate.
In practice, the [35S]GTPyS binding assay cannot be used to evaluate receptor-driven G protein activation in intact cells because [35S]GTPyS is unable to cross cell membranes. However, the [35S]GTPyS assay has been used successfully in digitonin-permeabilized cells. Digitonin binds to cholesterol in eukaryotic plasma membranes, creating pores that are permeable to ions and proteins. [35S]GTPyS binding has also been often restricted to the pertussis toxin-sensitive Gi/o family. When compared to the other G proteins, these are usually expressed at higher levels and (together with unrelated cell components like tubulin) they provide a substantial contribution to [35S]GTPyS binding under basal conditions (Milligan, 2003). However, as illustrated in Figure 116, a signal can be easily measured even with a small (e.g. fourfold) stimulation of [35S]GTPyS binding to Gai/o (green). By contrast, considerable (e.g. 20-fold) stimulation of [35S]GTPyS binding to other Ga subunits might be expected to result in only a small signal above basal levels. Recent strategies to overcome these limitations consist in:
• Expression of GPCRs and G proteins of interest in various insect cell lines that express low levels of endogenous G proteins.
• Standard [35S]GTPyS binding assays performed on membrane preparations, coupled with a selective immunocapture step to isolate the G protein(s) of interest.
A complex picture may arise when a GPCR is able to recognize two G proteins that mediate opposite effects such as adenylate cyclase stimulation by Gas and adenylate cyclase inhibition by Ga Such a situation arises for, for example, a2-adrenergic receptors. As shown in Figure 117, the full agonist UK-14,304 inhibits the cAMP production in intact CHO cells expressing a2-adrenergic receptors at high (a2A-H) as well as low (a2A-L) concentration. This process is Gi/o-mediated and when this is prevented (pretreatment of the cells with pertussis toxin), UK-14,304 starts to stimulate cAMP production (i.e. a Gs-mediated process). In fact this process is not very efficient, as demonstrated by the very low response in a2A-L cells. A much higher response is seen in a2A-H cells because of the operating 'receptor reserve'. This suggests that the a2-adrenergic receptor can couple to both Gs and Gi under natural conditions but that coupling to the 'traditional' G; prevails. In this respect, data with mutated a2-adrener-gic receptors suggest that quite different intracellular receptor regions are implicated in the recognition of Ga; and Gas (Table 11).
Was this article helpful?