Activation of Gαs and Gαq at plasma membrane and early endosomes upon AVP stimulation.

a, Left: Illustration of EbBRET-based biosensors used to monitor G protein activation by the V2R at the plasma membrane. Right: Kinetics of the recruitment of mG proteins at the plasma membrane upon stimulation of V2R-expressing HEK293 cells with 1 μM AVP. See also Supplementary Fig. 1a for expression of each mG construct. b, Left: Illustration of EbBRET-based biosensors used to monitor G protein activation by the V2R from early endosomes. Right: Kinetics of the recruitment of mG proteins to early endosomes upon stimulation of V2R-expressing HEK293 cells with 1 μM AVP. See also Supplementary Fig. 1b for the expression of each mG construct. n = 3 biological replicates for mGs, mGsi, and mGsq, and n = 4 for mG12 for a and b. c, Confocal microscopy of HEK293 cells expressing RFP-Lck, V2R, and Halo-mGs (left panels), or Halo-mGsq (right panels) stimulated for 10 minutes with vehicle (upper panels) or 1 μM AVP (bottom panels). d, Scatter plots showing Lck/mG colocalization performed on 5 representative images. e, Confocal microscopy of HEK293 cells expressing RFP-EEA1, V2R, and Halo-mGs (left panels), or Halo-mGsq (right panels) stimulated for 45 minutes with vehicle (upper panels) or 1 μM AVP (bottom panels). f, Scatter plots showing EEA1/mG colocalization performed on 6 representative images. Asterisks mark statistically significant differences between vehicle and AVP treatments as assessed by two-way ANOVA and Sidak’s post hoc test for multiple comparisons (***P ≤ 0.001, ****P ≤ 0.0001). All data are presented as mean ± s.e.m.

Formation of megaplexes with Gαs or Gαq upon stimulation of V2R with AVP.

a, Kinetics of mG protein recruitment to the plasma membrane upon stimulation of V2β2AR with 1 μM AVP. See Supplementary Fig. 2 for expression of each mG construct. n=3 for mGs, mGsi, and mGsq, and n=4 for mG12. b, Left: Illustration of EbBRET biosensors used to monitor βarr recruitment to the plasma membrane and endosomes. Right: Kinetics of the recruitment of βarr1 and βarr2 to the plasma membrane (upper panel) and to early endosomes (bottom panel) upon stimulation of V2R or V2β2AR with 1 μM AVP. See Supplementary Fig. 3 for the relative expression of V2R or V2β2AR at plasma membrane and of βarrs. n=3 for all conditions. c, Left panel: Illustration of nanoBiT biosensors used to monitor simultaneous coupling of Gα proteins and βarr1 to GPCRs. Right panels: Kinetics of the proximity between SmBiT-βarr1 and LgBiT-mGs (upper panel) or LgBiT-mGsq (bottom panel) upon stimulation of the V2R or V2β2AR with 1 μM AVP. n=3 for all conditions. d,e, Confocal microscopy of HEK293 cells expressing Halo-mGs or Halo-mGsq, strawberry-βarr2, and V2R (left panels), or V2β2AR (right panels). The cells were stimulated for 45 minutes with vehicle (upper panels) or 1 μM AVP (bottom panels). f, Scatter plots of the percentage of βarr2 colocalization with mGs or mGsq upon stimulation with 1μM AVP (6 representative images). Asterisks mark significant differences between V2R and V2β2AR assessed by two-way ANOVA and Sidak’s post hoc test for multiple comparisons (***≤0.001, ****≤0.0001). No statistical difference (ns) was detected between mGs and mGsq for V2R (P=0.6640) and V2β2AR (P=8446). Data are presented as mean ± s.e.m.

Contribution of megaplex to endosomal Gαs and Gαq signaling.

a,b, AVP dose-response curves of the recruitment of mGs (left panel) and mGsq (right panel) to the plasma membrane and early endosomes using parental or Δβarr1/2 cells. The cells were stimulated for 10 minutes (plasma membrane) or 45 minutes (early endosomes) with AVP. See Supplementary Fig. 4a,b for relative V2R expression levels at the plasma membrane in parental and Δβarr1/2 cells. n=4 biological replicates for each condition. See also Supplementary Table 1 for parameters related to dose response curves. c, Left: Illustration of EbBRET biosensors used to monitor AVP-mediated internalization of the V2R. Right: Kinetics of V2R internalization upon stimulation with AVP 0.1 μM. See Supplementary Fig. 4c for relative expression of V2R in parental and Δβarr1/2 cells. n=4 and asterisks mark significant differences from zero as assessed by one sample t test (*≤0.05, **≤0.01, ***≤0.001). d,e, Transduction coefficients of mGs and mGsq recruitment to the plasma membrane and early endosomes in V2R- or V2β2AR-expressing HEK293 cells. The cells were stimulated 10 minutes (plasma membrane) or 45 minutes (early endosomes) with 1 μM AVP. See Supplementary Fig. 5 and Supplementary Table 2 for the dose-response curves and associated parameters, respectively. See also Supplementary Fig. 4d,e for the relative expressions of V2R and V2β2AR at the plasma membrane. Asterisks mark significant differences between the V2R and V2β2AR as assessed by two-way ANOVA and Sidak’s post hoc test for multiple comparisons (*≤0.05). No statistical difference (ns) was detected between the V2R and V2β2AR for mGs (P=0.7692) and mGsq (P=0.7445) at the plasma membrane. n=4 for each condition. All data are presented as mean ± s.e.m.

Updated model of V2R signaling. At the plasma membrane, AVP binding to V2R results in receptor-mediated Gαs or Gαq activation. This initial G protein activation at the plasma membrane is followed by V2R internalization into early endosomes. This internalization occurs primarily in a βarr-dependent manner, leading to the formation of a megaplex with Gαs or Gαq/11 and robust activation of these G proteins from endosomes. Additionally, a minor population of V2R internalize in a βarr-independent fashion, which also leads to minor but significant Gαs or Gαq/11 activation from endosomes.

Scatter plots showing equivalent expression of Rluc-mG constructs for the kinetics experiments represented in Fig. 1a (a) and Fig. 1b (b).

n=3 biological replicates for mGs, mGsi, and mGsq, and n=4 for mG12. Statistical differences between mGs and the other mG proteins were assessed by one-way ANOVA and Dunnett’s post hoc test for multiple comparisons. No statistical differences were detected as compared to mGs (ns). In a, P=0.9985, 0.9130, 0.5902 for mGsi, mGsq, and mG12 respectively. In b, P=0.9999, 0.9072, 0.9019 for mGsi, mGsq, and mG12 respectively. The mean ± s.e.m are represented.

Scatter plots showing the relative expression of each Rluc-mG construct for the kinetics experiments represented in Fig. 2a.

n=3 biological replicates for mGs, mGsi, and mGsq, and n=4 for mG12. Statistical differences between mGs and the other mG proteins were assessed by one-way ANOVA and Dunnett’s post hoc test for multiple comparisons. No statistical differences were detected as compared to mGs (ns). P=0.9686, 0.8091, 0.1528 for mGsi, mGsq, and mG12 respectively. The mean ± s.e.m are represented.

Relative expression of V2R and V2β2AR at the plasma membrane and of βarrs for the experiments represented in Fig. 2b.

a,b, Scatter plots of relative V2R and V2β2AR expression at the plasma membrane (upper panels) and of βarr1 and βarr2 (bottom panels). n=3 biological replicates for each condition. Statistical differences between V2R- and V2β2AR-expressing cells were assessed by two-way ANOVA and Sidak’s post hoc test for multiple comparisons. No statistical differences were detected (ns). In a, upper panel, P=0.4485 for βarr1, and P=0.7995 for βarr2. In a, bottom panel, P=0.9730 for βarr1, and P=0.3188 for βarr2. In b, upper panel, P=0.6893 for βarr1, and P=0.6316 for βarr2. In b, bottom panel, P=0.0954 for βarr1, and P=0.8251 for βarr2. The mean ± s.e.m are represented.

Relative receptor expression related to the experiments in Fig. 3.

a,b Scatter plots of relative expression of V2R in parental versus Δβarr1/2 cells used to measure mGs and mGsq recruitment to the plasma membrane and early endosomes in Fig. 3a,b. Statistical differences between parental and Δβarr1/2 cells were assessed by two-way ANOVA and Sidak’s post hoc test for multiple comparisons. No statistical differences were detected (ns). In a, P=0.5785 for mGs, and P=0.4970 for mGsq. In b, P=0.7238 for mGs and P=0.4882 for mGsq. c, Scatter plots showing the relative expression of V2R-Rluc in parental versus Δβarr1/2 cells used to monitor V2R internalization in Fig. 3c. Relative V2R-Rluc expression was assessed by monitoring the relative luminescence units (RLU) emitted by the luciferase. No statistical differences were detected (ns) using a paired t test (P=0.9748). d,e, Scatter plots showing the relative expression of V2R and V2β2AR at the plasma membrane in the cells used to determine the transduction coefficients represented in Fig. 3d,e. Statistical differences between the V2R and V2β2AR were assessed by two-way ANOVA and Sidak’s post hoc test for multiple comparisons. No statistical differences were detected (ns). In d, P=0.0996 for the plasma membrane, and P=0.1242 for early endosomes. In e, P=0.1615 for the plasma membrane, and P=0.2944 for early endosomes. n=4 biological replicates for all conditions. The mean ± s.e.m are represented.

AVP dose-response curves for the recruitment of mGs and mGsq to the plasma membrane and early endosomes by the V2R and V2β2AR.

a,b, Dose-dependent recruitment of mGs and mGsq in HEK293 cells upon 10 minutes (plasma membrane) or 45 minutes (early endosomes) of AVP treatment. See Supplementary Table 2 for dose-response parameters and Supplementary Fig. 4d,e for relative expression of the V2R and V2β2AR at the plasma membrane. n=4 biological replicates for each condition and the mean ± s.e.m are represented.

Parameters related to AVP dose-response curves of the mGs and mGsq recruitment to the plasma membrane or early endosomes in parental and Δβarr1/2 cells.

ΔEbBRET values from Fig. 3a,b were fitted using four parameters equation with the bottom fixed at zero. n=4 biological replicates for each condition. Statistical differences for AVP-induced maximal efficacy (ΔEbBRET) and potency (LogEC50) between parental and Δβarr1/2 cells were assessed by comparing independent fits with a global fit that shares the selected parameter using extra sum-of-squares F test (**≤ 0.01, ****≤ 0.0001). The mean ± s.e.m are represented.

Parameters related to AVP dose-response curves of the mGs and mGsq recruitment to the plasma membrane or early endosomes in cells expressing the V2R or V2β2AR.

ΔEbBRET values from Supplementary Fig. 5 were fitted using four parameters equation with the bottom fixed at zero. n=4 biological replicates for each condition. Statistical differences between AVP-induced maximal efficacy and potency of mGs or mGsq recruitment by V2R and V2β2AR were assessed by comparing independent fits with a global fit that shares the selected parameter using extra sum-of-squares F test (**≤0.01, ***≤ 0.001). Statistical significance for Log(τ/Ka) values from V2R-expressing cells compared to cells expressing the V2β2AR were assessed by two-way ANOVA and Sidak’s post hoc test for multiple comparisons (*≤0.05). The mean ± s.e.m are represented.