Pretreatment of cells coexpressing GHSR1a and DRD2 for 30 min with increasing concentrations of the GHSR1a agonists Z-VAD-FMK MK-677 (Patchett et al., 1995) or ghrelin reduces dopamine-induced Ca2+ mobilization by 60%–75% of the control response (Figure 5A). MK-677 with a longer half-life than ghrelin is significantly more efficient than ghrelin in attenuating DRD2-induced Ca2+ signaling (MK-677 EC50 = 0.064 ± 0.0005 nM, ghrelin EC50 = 0.87 ± 0.019 nM; p < 0.05; Figure 5A). Similarly, preincubation with dopamine or quinpirole reduces ghrelin-induced Ca2+ release by 60% and 50%, respectively (Figure 5B), but preincubation with the D1R-selective agonist SKF81297 fails to inhibit

the ghrelin-induced response (Figure 5B). Cross-desensitization observed with a GHSR1a agonist or DRD2 agonist is consistent with a mechanism involving formation of GHSR1a:DRD2. We employed time-resolved (Tr)-FRET to test for heteromer formation because this technology is ideal for monitoring cell surface protein-protein interactions at physiological concentrations of receptors (Maurel et al., 2008). We introduced

a SNAP-tag at the GHSR1a N terminus and showed its appropriate expression on the cell surface and its functional activity (Figures S3A and S3B). Specific labeling of SNAP-GHSR1a was demonstrated by SDS-PAGE in-gel fluorescence, fluorescent confocal microscopy, and dose-dependent cell surface labeling with BG-488 (Figures S3C–S3E). To optimize the Tr-FRET signal,

cells expressing SNAP-GHSR1a were incubated with a fixed concentration of energy donor (terbium cryptate, buy PFI-2 100 nM) and increasing concentrations of acceptor (Figure S3F) and a linear relationship between receptor concentration and Tr-FRET signal was established (Figure S3G). When GHSR1a is expressed alone, it forms homomers and, consistent with formation of GHSR1a homomers, the Tr-FRET signal is reduced according to the ratio SNAP-GHSR1a to GHSR1a such that at a ratio of 1:1 Tr-FRET is reduced to 59% ± 6% and to 17% ± 3.7% at a 1:5 ratio (Figure 6A). When DRD2 is substituted for GHSR1a, the Tr-FRET signal generated by GHSR1a:GHSR1a homomers is Carnitine palmitoyltransferase II reduced to 62% ± 10% by a 1:1 ratio of GHSR1a to DRD2 (p < 0.01), and 36.6% ± 6.5% by a 1:5 ratio, consistent with formation of GHSR1a:DRD2 heteromers (Figure 6A). When a control GPCR, RXFP1, is coexpressed with SNAP-GHSR1a, the Tr-FRET is not attenuated (Figure 6A). To confirm GHSR1a:DRD2 formation, we prepared CLIP-tagged GHSR1a and SNAP-tagged DRD2 and examined expression of these receptors by confocal microscopy. Both the CLIP- and SNAP-tagged receptors are colocalized on the cell surface (Figure 6B). We then conducted saturation assays observing robust saturable Tr-FRET signals indicative of specific heteromerization rather than random collisions (Figure 6C). As a further test of heteromerization of GHSR1a and DRD2 we utilized a SNAP-tagged DRD2 variant.