Fig 2: STARPA reveals differential GRK expression in commonly used cell lines. (a) STARPA was used to determine relative GRK expression in HEK293, HeLa, HepG2, Jurkat, K562, MCF-7, Molm-13, U2OS, and U-251 MG cells. Lysates of different cell lines together with the validated STARPA standards (Figure 4c,d) were loaded onto polyacrylamide gels. Western blot analysis using anti-HA antibody confirmed the STARPA standards. Immunoblotting using anti-GRK antibodies for each isoform allowed the determination of endogenously expressed GRK levels in the cell lines. Shown are representative images of three independent blots from identical lysates. (b) Close ups of the STARPA standards from the immunoblot presented in (a) displaying the cross-reactivity of the anti-GRK2 and anti-GRK6 antibody towards GRK3 and GRK5, respectively. Below, the quantification of three independent blots ± SD. The means of the obtained signals are indicated and were used to determine the respective cross-reactivity coefficient of the antibodies. (c) STARPA blots were quantified according to each GRK isoform-specific antibody, and the relative signal compared to the respective signal in the standards is shown as mean + SD of three independent blots from identical lysates. The calculated relative expression levels of GRK2 and GRK6 before (striped) and after (solid) cross-reactivity correction are presented.

Fig 3: GRK isoform specificity of several commercially available GRK antibodies. HEK293 cells were transiently transfected with pcDNA3 plasmids encoding either GRK2, GRK3-1, GRK3-2, GRK5, GRK6-1, GRK6-2, GRK6-3, or GRK6-4 as indicated. GRK isoform specificity of indicated GRK antibodies was investigated in Western blot analysis. Representative blots are shown. Two different antibodies raised against GRK2 (a,b), GRK3 (c,d), GRK5 (e,f), and GRK6 (g,h) were used for the analysis. Approximate molecular weight of each antigen is stated. Detailed information on each antibody examined is listed in Table 1. Actin served as loading control.

Fig 4: Genetic deletion of each GRK differentially regulates ?-secretase activity and Aß generation(A–D) Aß40 and Aß42 generation and representative western blot analysis of HEK293 cells following expression of APP-C99 in CRISPR control cells (white bars) and each GRK KO cell line (gray bars): (A) Adrbk1 (GRK2), (B) Adrbk2 (GRK3), (C) Grk5 (GRK5), and (D) Grk6 (GRK6). ****p < 0.0001 by ANOVA with Tukey’s post-test. Bars denote mean ± SEM of 3 independent experiments (n = 3) performed in triplicate. The means from each experiment are indicated (white circles).

Fig 5: GRK2, 3, 5 or 6 can individually mediate a ligand-independent interaction of the V2R and ß-arrestin1.a–d ?Q-GRK or Control cells were transfected with V2R-Halo-Tag and one of the following ß-arrestin1 (ßarr1)-NanoLuc fusion constructs: wild type (a, b) or ßarr1 lacking the finger loop region (dFLR; c, d). Additionally, either GRK2, 3, 5, 6 or the empty vector (EV) were transfected as indicated. The dynamic BRET changes are shown as ligand concentration-response curves normalised to baseline values and vehicle control. All data points are calculated as ? net BRET fold change, mean of n = 3 independent experiments ± SEM. The same dataset is presented as bar graphs, displaying the mean BRET values + SEM before (baseline) and after stimulation with 10 µM [Arg8]-vasopressin (AVP; stimulated), normalised to the basal BRET ratio derived from the ?Q-GRK + EV condition (dashed line). To test whether the baseline BRET ratios were significantly elevated compared to the respective ?Q-GRK + EV baseline, an ANOVA and one-sided Dunnett’s test was performed (*p < 0.05; **p < 0.01; ns not significant; a, b ?Q-GRK + GRK2 p = 0.0113, ?Q-GRK + GRK3 p = 0.0142, ?Q-GRK + GRK5 p = 0.0188, ?Q-GRK + GRK6 p = 0.0036, Control + EV p = 0.6186; c, d: ?Q-GRK + GRK2 p = 0.0073, ?Q-GRK + GRK3 p = 0.0033, ?Q-GRK + GRK5 p = 0.0024, ?Q-GRK + GRK6 p = 0.0013, Control + EV p = 0.1054). e, f ?Q-GRK or Control cells were transfected with V2R-CFP (green), Rab5-mCherry (magenta), ßarr1-YFP (not shown) and either EV, GRK2 or GRK6 as indicated. Images were taken before (basal) and after 15 min of 100 nM AVP stimulation. Representative images are shown in (e) and Supplementary Fig. 8e–h. The co-localisation of V2R and ßarr1 or Rab5 was quantified using Squassh and SquasshAnalyst (number of images per respective condition; ?Q-GRK + EV (35), ?Q-GRK + GRK2 (35), ?Q-GRK + GRK6 (33), Control + EV (30)). f Data are presented as mean fold change in co-localisation signal + SEM normalised to unstimulated (baseline) ?Q-GRK + EV condition. Co-localisation prior to stimulation was compared using ANOVA and two-sided Dunnett’s test (*p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001; ns not significant). All exact p values, test statistics, effect sizes, confidence intervals and degrees of freedom are provided in the Source Data files. g Schematic depiction of ßarr1 interactions with the V2R in the absence and presence of ligand, facilitated by high expression levels of GRKs.