Cloning and construct origin
The transfected GRK constructs and receptors were of human origin and β-arrestin2 was of bovine origin. The NanoLuciferase (NLuc) and Halo-Tag genes were obtained from Promega and the plasmids expressing b2AR-, M2R- and M5R-NLuc and Halo-β-arrestin2 have been described before16. The bARK-CT construct was provided by Professor Silvio Gutkind. Plasmids expressing human full-length GRK2 and GRK3 in pcDNA3 have been described in Drube et al., 202216 and plasmids expressing GRK2/3-D110A and GRK2/3-R587Q have been described and characterized in Jaiswal, 202334. The mutant constructs were assessed for similar levels via Western blot analysis34. The GRK2-D110A/R587Q double mutant was created by using GRK2-D110A as the template and 5’GGAGATCTTCGCCTCATACATCATGAAGGAGCTGCTGG as forward primer and 5’CCAGCAGCTCCTTCATGATGTATGAGGCGAAGATCTCC as reverse primer. GRK3 D110A/R587Q double mutant was created using GRK3 R587Q as the template and 5’TTCCCCAACCAGCTCGAGTGGC as the forward primer and 5’GCCACTCGAGCTGGTTGGGGAA as the reverse primer to introduce D110A. GRK2/3-CAAX constructs were generated by inserting the CAAX overhangs using Gibson assembly to GRK2/3 WT, D110A, R587Q and D110A/R587Q. The generated constructs were validated by sequencing at Eurofins Genomics GmbH. The pcDNA3 backbone was used as a control and is referred to as the empty vector (EV). The mNeonGreen-GRK constructs were generated through the insertion of dsDNA strings (GeneArt) encoding GRK2/3 WT, -D110A, -R587Q and -D110A, R587Q with or without a following sequence encoding the CAAX motif (NPPDESGPCCMSCKCVLS), downstream the mNeonGreen coding sequence, using In-Fusion Cloning technology (Takara Bio)51,52.
Cell culture
CRISPR/Cas9-generated HEK293 knockout cells of GRK2/3/5/6 (ΔQ-GRK), GRK2/3 (ΔGRK2/3) or GRK2 (ΔGRK2) and CRISPR/Cas9 HEK293 control cells (Control) with unaltered GRK expression16 were cultured at 37 °C with 5% CO2 in Dulbecco’s modified Eagle’s medium (DMEM; Sigma-Aldrich, D6429), complemented with 10% fetal calf serum (Sigma-Aldrich, F7524) and 1% penicillin and streptomycin mixture (Sigma-Aldrich P0781). Cells were passaged every 3–4 days and regularly checked for infections with mycoplasma using the LONZA MycoAlert mycoplasma detection kit (LT07-318).
Western blot
The following antibodies were used for the detection of GRK2/3 and their mutants: mouse anti-GRK2 (Santa Cruz, sc-13143), rabbit anti-GRK3 (Cell signaling technology, 80362), mouse anti-actin (Sigma-Aldrich, A5441), HRP-conjugated goat anti-mouse (SeraCare, 5220-0341) and HRP-conjugated goat anti-rabbit (SeraCare, 5220-0336).
GRK2 knockout cells (GRK2 Western blot) or GRK2/3 knockout cells (GRK3 Western blot) were washed once with PBS and subsequently lysed with RIPA buffer (1% NP-40, 1 mM EDTA, 50 mM Tris pH 7.4, 150 mM NaCl, 0.25% sodium deoxycholate) at room temperature for 15 min. The lysates were centrifuged at 10,000 g for 15 min at 4 °C and the protein amount was estimated using the BCA protein estimation kit (Thermo Fisher Scientific, 23225). Sample loading buffer was added to each cell lysate followed by denaturation at 95 °C for 5 min. 10 μg of total protein was loaded onto each lane of 10% polyacrylamide gels and following electrophoretic separation, was transferred to nitrocellulose membranes. After the membranes were blocked with TBST (supplemented with 5% non-fat dry milk) and washed, the protein was detected by incubating overnight at 4 °C with specific primary antibodies, as listed above. As secondary antibodies, we used HRP-conjugated goat anti-mouse or anti-rabbit antibodies (diluted 1:10000), incubated at room temperature for one hour. Chemiluminescence signal was detected using LAS4000 Image Reader (Fujifilm 2.11, Life Science) and quantified using the Fujifilm Multi Gauge software (V3.0). Signal for each sample of GRK2/3 was background-corrected and then normalized to its respective background-corrected actin signal.
Localization of GRK constructs using confocal microscopy
For the assessment of employed GRK construct localization, we utilized N-terminally tagged NeonGreen-GRK fusion protein plasmids. Corresponding to the respective construct, a C-terminal H-Ras CAAX motif tethers the GRK to the plasma membrane. The cells were seeded in 6 cm dishes (ΔQ-GRK 1.6 × 106 cells per dish) and transfected the following day with 1 µg of the indicated NeonGreen-GRK construct or empty vector (EV), according to the Effectene transfection reagent manual (Qiagen, #301427). After 24 h 1.0 × 106 cells were re-seeded on poly-D-Lysine-coated 24 mm round glass coverslips and maintained for 24 h in complete culture medium. Images of living cells were acquired using an inverted laser scanning confocal microscope (DMi8 TCS SP8, Leica microsystems) equipped with a HC PL APO CS2 63x/1.40 oil objective (Leica).The NeonGreen fluorophore was excited at a wavelength of 496 nm and emission was detected in a bandwidth of 512–540 nm. 1024 × 1024 pixels format images were acquired and later processed with ImageJ (https://imagej.nih.gov/ij/ NIH, Bethesda).
Fluorometric assessment of GRK2 and GRK3 construct expression
ΔQ-GRK cells were seeded and transfected as described in above in the “Localization of GRK constructs using confocal microscopy” section. The following day, 40,000 cells per well were seeded into black poly-D-lysine-coated 96-well plates (Brand, 781968). For each transfection, technical replicates were seeded as quadruplicates. Before measuring the next day, the cells were washed twice using measuring buffer (140 mM NaCl, 10 mM HEPES, 5.4 mM KCl, 2 mM CaCl2, 1 mM MgCl2; pH 7.3). The measurements were performed in measuring buffer using a Synergy Neo2 plate reader (Biotek), the Gen5 software (version 2.09) and a corresponding 485/20 excitation filter (BioTek, 1035014) and 516/540 emission filter cube (BioTek, 1035047). The measured intensity was normalized to background (empty vector-transfected control).
Bioluminescence resonance energy transfer (BRET) measurements
The intermolecular BRET measurements to investigate β-arrestin recruitment were conducted as described before16. In short, cells were seeded into 6 cm dishes (ΔQ-GRK 1.6 × 106 cells per dish) and transfected the following day with 0.5 µg of the indicated GPCR-NLuc, 1 µg of Halo-tag-β-arrestin2 and 0.25 µg of one GRK construct or empty vector (EV), according to the Effectene transfection reagent manual (Qiagen, #301427). Each transfection was adjusted with EV to contain 2.5 µg total DNA. The following day, 40,000 cells per well were seeded into poly-D-lysine-coated 96-well plates (Brand, 781965) and the Halo-ligand was added (1:2,000; Promega, G980A). For each transfection, technical replicates were seeded as triplicates and a mock labeling condition was included without the Halo-ligand. Before measuring the next day, the cells were washed twice using measuring buffer (140 mM NaCl, 10 mM HEPES, 5.4 mM KCl, 2 mM CaCl2, 1 mM MgCl2; pH 7.3). After aspiration, the NLuc-substrate furimazine (Promega, N157B) in measuring buffer (1:3,500) was added. The measurements were performed in a Synergy Neo2 plate reader (Biotek), the Gen5 software (version 2.09) and a customized filter cube (fitted with a 555 nm dichroic mirror and a 620/15 bandpass filter). First, basal values were measured for 3 min, followed by addition of the indicated agonist and measurement of the stimulated values for 5 min. After the first measurement upon stimulation, four data points (43 sec intervals) were averaged for the concentration-response curves. The human b2AR was stimulated with isoproterenol (Iso; Sigma-Aldrich, I5627, dissolved in water). The human M2R and M5R were stimulated with Acetylcholine (ACh; Sigma-Aldrich, A6625, dissolved in measuring buffer).
In case of the bARK-CT-mediated inhibition of the GRK2-Gβγ interaction32,36 and the measured effects on β-arrestin2 recruitment, 1.2 × 106 CRISPR/Cas9 HEK293 control cells were seeded into 6 cm dishes. The cells were transfected as described above with 0.5 µg of M5R-NLuc and 1 µg of Halo-tag-β-arrestin2, in addition to 0.5 µg or 1 µg of bARK-CT or 1 µg EV as a control. The total amount of transfected DNA was adjusted to 2.5 µg with EV, when necessary.
For the measurement of β-arrestin2 recruitment to M5R in presence of the guanine nucleotide dissociation inhibitor FR90035938, ΔQ-GRK and CRISPR/Cas9 HEK293 Control cells (Control) were transfected and re-seeded as described above. On the day of the measurement, cells were pre-incubated in cell culture medium (DMEM, 10% fetal calf serum, 1% penicillin and streptomycin mixture) supplemented with 300 nM FR900359 (10 mM stock solution, solved in DMSO) for 1 h at 37 °C. Following pre-incubation, cells were washed twice with measuring buffer supplemented with 300 nM FR900359. Addition of furimazine substrate and measurements were then carried out in measuring buffer supplemented with 300 nM FR900359. The data points at 2–4 min after stimulation were averaged for the concentration-response curves.
To measure the effect of FR900359 on receptor internalization, ΔQ-GRK and CRISPR/Cas9 HEK293 Control cells were seeded as described above and transfected with 0.1 µg M5R-NLuc and 1 µg early endosome-tethered mNeonGreen-FYVE39,40,41. The total amounts of DNA were adjusted to 2 µg and the transfections were carried out according to the Effectene transfection reagent manual (Qiagen, #301427). The pre-incubation with 300 nM FR900359 was performed as described above. The measurement was performed using a 410/515 filter (BioTek, 1035072) for 30 min after stimulation. The data points of the last 10 min were averaged for the concentration-response curves.
Analysis, statistics and reproducibility
The measured BRET ratios were labeling corrected by subtraction of the respective mock-labelled condition and subsequently the averaged technical replicates of the stimulated values were divided by the respective averaged baseline values. To get the final dynamic Δ net BRET change, the ligand-dependent labeling-corrected BRET change was divided by the vehicle control and calculated as percent changes. These corrected BRET changes were normalized to the maximum value of GRK2- or GRK-CAAX-mediated recruitment, as indicated in the respective figure legends. In the bARK-CT experiment, the Δ net BRET changes were normalized to the β-arrestin2 recruitment at the highest ligand concentration in absence of bARK-CT (EV-transfected condition). For the FR900359 inhibition experiments, the Δ net BRET changes were normalized to the maximal signal in CRISPR/Cas9 HEK293 control cells without the inhibitor present. In case of the M5R translocation to early endosomes, no labeling procedure was necessary due to utilization of the fluorophore NeonGreen as acceptor, hence there was no labeling correction in the analysis. All data are shown as mean of at least three independent experiments ± SEM as indicated. Statistical comparisons were made in GraphPad Prism 7.03 using one-way ANOVA and subsequent Turkey’s test. The supplementary bar graphs of the Halo labeling- and vehicle-corrected mean Δ net BRET changes + SEM before (basal) and after stimulation with the indicated ligand were normalized to the basal BRET ratio derived from the EV-transfected condition (Δ net BRET fold change). Here, statistical differences within one condition between basal and stimulated or between differently transfected conditions were tested using two-way ANOVA followed by a Sidak’s or Tukey’s test respectively. In all cases, a type I error probability of 0.05 was considered significant.
Plate reader experiments were performed in technical replicates of three or four from the same transfection, stimulated with identical ligand solution, as indicated in the respective methods sections. The technical replicates were averaged for each n. All shown experiments represent at least n = 3 of independent experiments with independent transfections. Exact n numbers are provided in the respective figure legends.