Isoprenaline shows unique kinase dependencies in stimulating β1AR-β-arrestin2 interaction compared to endogenous catecholamines

Abstract

The β1-adrenergic receptor (β₁AR) is an essential G protein-coupled receptor in the heart. Its dysregulation represents a hallmark of cardiac diseases. Studies have identified a unique mode of β-arrestin interaction, where β₁AR briefly engages with β-arrestins before catalytically accumulating them at the plasma membrane (PM) independently of the receptor. Although receptor phosphorylation crucially impacts β-arrestins, the contributions of specific kinases vital in β₁AR regulation remain unclear. Here, we employed G protein-coupled receptor kinase (GRK) GRK2/3/5/6 knockout cells and the protein kinase A inhibitor H89 in bioluminescence resonance energy transfer-based assays to systematically assess GRKs and protein kinase A in direct β-arrestin2 recruitment to β₁AR and β-arrestin2 translocation to the PM. Furthermore, we compared the effects of the synthetic agonist isoprenaline with the endogenous catecholamines: epinephrine and norepinephrine. We observed pronounced differences in their kinase dependencies to mediate β-arrestin2 translocation to the PM. Upon isoprenaline stimulation, GRKs strongly influenced β-arrestin2 translocation to the PM but had no effect on direct β-arrestin2 recruitment to β₁AR. Additionally, in a GRK2-specific context, protein kinase A inhibition primarily reduced the efficacy of isoprenaline for β-arrestin2 translocation, whereas for GRK5, it decreased potency. Strikingly, these kinase-dependent effects were absent for epinephrine and norepinephrine, suggesting distinct underlying molecular mechanisms for β-arrestin2 accumulation at the PM. This observation could be explained by agonist-specific differences in receptor conformational rearrangements, as suggested by distinct changes in the NMR spectra of β₁AR. Our findings highlight that synthetic and endogenous ligands induce distinct molecular mechanisms in β₁AR regulation, emphasizing the need to consider these differences when translating molecular insights into physiological contexts. SIGNIFICANCE STATEMENT: Our findings reveal mechanistic differences in β1-adrenergic receptor-mediated catalytic activation of β-arrestin2 by synthetic and endogenous agonists, driven by distinct G protein-coupled receptor kinases and protein kinase A dependencies. Although β-arrestin2 translocation to the PM occurred to similar extents with isoprenaline, epinephrine, and norepinephrine, kinase involvement was crucial only upon Iso stimulation of β1-adrenergic receptor. By elucidating these ligand-specific pathways, this study advances our understanding of β1-adrenergic receptor signaling and regulation while additionally highlighting the importance of considering these differences when translating molecular insights into pathophysiological contexts.

Keywords: Bioluminescence resonance energy transfer; G protein-coupled receptor kinases; G protein-coupled receptors; Nuclear magnetic resonance spectroscopy; Protein kinase A; β-arrestin2 translocation.

Fig. 1

Iso-mediated β-arrestin2 interaction with the β₁AR. (A) Schematic representation of the direct β-arrestin2 recruitment NanoBRET assay (PDBs: β₁AR, 7BTS; β-arrestin2, 3P2D; β₁AR–β-arrestin complex, 6TKO; NanoLuc, 7MJB; HaloTag, 8SW8). (B) Concentration–response curves of direct β-arrestin2 recruitment to β₁AR, normalized to the maximum signal in Control cells. Control and ΔQ-GRK (GRK2/3/5/6 knockout) cells were transfected with β₁AR-NanoLuc and HaloTag-β-arrestin2 and stimulated with the indicated concentration of Iso. (C) Schematic depiction of the assay measuring β-arrestin2 translocation to the PM (PDBs: β₁AR, 7BTS; β-arrestin2, 3P2D, 5TV1; NanoLuc, 7MJB; and YFP, 4NDK). (D) Concentration–response curves of β-arrestin2 translocation to the PM, upon β₁AR stimulation, normalized to the maximum signal in Control cells. Control and ΔQ-GRK cells were transfected with unlabeled β₁AR, β-arrestin2-NanoLuc, and YFP-CAAX and stimulated with the indicated concentration of Iso. All measurements are shown as the mean of at least n = 3 independent experiments ± SEM. Statistical analysis was performed using an unpaired two-tailed t test (ns, not significant; ∗∗∗∗, P < .0001), detailed results can be accessed in Supplemental Table 1. (E, F) Representative confocal microscopy images of Control and ΔQ-GRK cells, transfected with β₁AR-CFP, β-arrestin2-YFP, and Rab5-mCherry, before and after stimulation with 1 μM Iso for 15 min.

Fig. 2

Effect of PKA on GRK-specific β-arrestin2 interaction with the β₁AR. (A–D) Concentration–response curves of direct β-arrestin2 recruitment to β₁AR under endogenous GRK expression (A), in ΔQ-GRK cells without GRKs present (B) and with overexpression of GRK2 (C) or GRK5 (D) in ΔQ-GRK cells. All assays were performed in the absence and presence of 5 μM PKA inhibitor H89, after 30 minutes of incubation at 37 °C with 10 μM H89. (E–H) Concentration–response curves of β-arrestin2 translocation to the PM, upon β₁AR stimulation in the same conditions as for the direct β-arrestin2 recruitment (A–D). All curves were normalized to the maximum response measured in Control cells upon Iso stimulation in the respective assays (A, E) and data points are displayed as mean of at least n = 3 independent experiments ± SEM. Statistical analysis was performed using a two-way ANOVA, followed by a Sidak test to compare values at maximal stimulation in the absence and presence of H89 (∗, P < .05). Detailed results can be accessed in Supplemental Table 3, additionally for comparisons of each condition with measurements in Control cells in absence of H89 (two-way ANOVA, followed by a Dunnett’s test). Data for Control and ΔQ-GRK cells in the absence of H89 (A, B, E, F) are shown again from Fig. 1 to allow a direct comparison of the effect of PKA inhibition.

Fig. 3

Effect of PKA on GRK-specific Epi- and NE-mediated β-arrestin2 translocation to the PM. (A–H) Concentration–response curves of β-arrestin2 translocation to the PM upon β₁AR stimulation with the endogenous agonist’s Epi (A–D) or NE (E–H). Transfections and assays were performed as described. All curves were normalized to the maximum response measured in Control cells upon Iso stimulation (Fig. 2E) and data are shown as the mean of at least n=3 independent experiments ± SEM. Statistical analysis was performed using a two-way ANOVA, followed by a Sidak test to compare values at maximal stimulation in the absence and presence of H89 or followed by a Dunnett’s test to compare each condition with measurements in Control cells in the absence of H89. All results can be accessed in Supplemental Table 4.

Fig. 4

Direct comparison of ligand-specific cAMP response using the EPAC1-cAMPs sensor. (A) Schematic representation of the used EPAC1-cAMPs sensor [PDBs: YFP, 4NDK; CFP, 3ZTF; part of EPAC1 used for sensor (E157–E316), 6H7E] (Nikolaev et al, 2004). (B) Changes in the baseline- and vehicle-normalized FRET ratio over time upon stimulation of β₁AR. Control cells featuring endogenous kinase expression were transfected with β₁AR and the EPAC-cAMPs sensor and measured over 30 min after the addition of the indicated ligand (1 μM Iso, 30 μM Epi, or 30 μM NE). (C) Values between approximately 45 and 90 s after stimulation were averaged and displayed as a bar graph. All data are shown as n = 3 independent experiments ± SEM. Statistical analysis was performed using a one-way ANOVA, followed by a Tukey test (ns, not significant). Detailed results can be accessed in Supplemental Table 5.

Fig. 5

Ligand-specific NMR spectra of ¹³C^ε-Met-labeled β₁AR. (A) AlphaFold2 (Jumper et al, 2021) model of the used β₁AR construct. Methyl groups of the methionine residues are marked as purple spheres. The orthosteric binding site is indicated via Epi (shown as blue spheres) through an alignment with Epi-bound β₁AR (PDB: 7BTS). Additionally, the chemical structures of the ligands used in this direct comparison are shown. (B–D) 2D [¹³C, ¹H] NMR correlation spectra of ¹³C^ε-Met-labeled β₁AR in the apo state (B, gray) or in the presence of Iso (B–D, orange), Epi (C, blue) or NE (D, green). Distinct chemical shifts are marked with black boxes (C, D).