The ADORA1 mutation linked to early-onset Parkinson's disease alters adenosine A1-A2A receptor heteromer formation and function

Abstract

Adenosine modulates neurotransmission through inhibitory adenosine A₁ receptors (A₁Rs) and stimulatory A_2A receptors (A_2ARs). These G protein-coupled receptors are involved in motor function and related to neurodegenerative diseases such as Parkinson's disease (PD). An autosomal-recessive mutation (G279^7.44S) within the transmembrane helix (TM) 7 of A₁R (A₁R^G279S) has been associated with the development of early onset PD (EOPD). Here, we aimed at investigating the impact of this mutation on the structure and function of the A₁R and the A₁R-A_2AR heteromer. Our results revealed that the G279^7.44S mutation does not alter A₁R expression, ligand binding, constitutive activity or coupling to transducer proteins (G_αi, G_αq, G_α12/13, G_αs, β-arrestin2 and GRK2) in transfected HEK-293 T cells. However, A₁R^G279S weakened the ability of A₁R to heteromerize with A_2AR, as shown in a NanoBiT assay, which led to the disappearance of the heteromerization-dependent negative allosteric modulation that A₁R imposes on the constitutive activity and agonist-induced activation of the A_2AR. Molecular dynamic simulations allowed to propose an indirect mechanism by which the G279^7.44S mutation in TM 7 of A₁R weakens the TM 5/6 interface of the A₁R-A_2AR heteromer. Therefore, it is demonstrated that a PD linked ADORA1 mutation is associated with dysfunction of adenosine receptor heteromerization. We postulate that a hyperglutamatergic state secondary to increased constitutive activity and sensitivity to adenosine of A_2AR not forming heteromers with A₁R could represent a main pathogenetic mechanism of the EOPD associated with the G279^7.44S ADORA1 mutation.

Keywords: A(1)R-A(2A)R heteromer; Adenosine A(1) receptor; Constitutive activity; Early-onset Parkinson’s disease.

Figure 1.. Expression of A₁R^wt and A₁R^G279S in HEK-293T cells.

(A) Cryo-EM structure of A₁R (PDB id 6D9H) [66] showing the positions of G279^7.44 (orange circle) within TM 7, the highly conserved NP^7.50xxY motif (purple sticks), the adenosine orthosteric binding site (yellow surface), and the Gi binding site (green surface). (B) Immunofluorescence detection of A₁R^wt-NL and A₁R^G279S-NL in stable transfected HEK-293T cells. Cells stably expressing A₁R^wt-NL and A₁R^G279S-NL were processed for immunocytofluorescence detection of A₁R (red) using a rabbit anti-A₁R (1 µg/ml) antibody. Nuclei were stained with DAPI (blue) (see Methods). Scale bar: 10 μm. (C) Representative immunoblot showing total and cell surface density of A₁R^wt-NL (lane 1) and A₁R^G279S-NL (lane 2) in stable transfected HEK-293T cells. Total and cell surface extracts were analysed by SDS-PAGE and immunoblotted using rabbit anti-A₁R (1 µg/ml) antibody (see Methods). The asterisk denotes the possible identification of the A₁R homodimer, as previously described [67]. (D) Relative quantification of A₁R^wt-NL and A₁R^G279S-NL cell surface density. The immunoblot protein bands corresponding to A₁R^wt-NL and A₁R^G279S-NL were quantified by densitometric scanning; values were normalized to the respective amount of α-actinin in each lane to correct for protein loading in the total extract. Then, the densitometric scanning of cell surface A₁R^wt-NL and A₁R^G279S-NL was related to the normalized total amount of receptor (Cell surface/Total) and expressed as mean ± SEM of four independent experiments. No significant difference was observed in cell surface expression between A₁R^wt-NL and A₁R^G279S-NL (p = 0.7516, Student t test).

Figure 2.. A₁R cell surface ligand binding determinations.

(A) Schematic representation of the NanoBRET assay. Nanoluciferase (NL) fused to the N-terminal domain of A₁R (i.e., A₁R^wt-NL and A₁R^G279S-NL) acts as a donor in a bioluminescence resonance energy transfer (BRET) process emitting light at 490–10 nm in presence of coelenterazine 400a (blue circles). The fluorescent ligand (i.e., CA200634) acts as acceptor of the BRET process, thus the light excites the BODIPY attached to the ligan, which subsequently emits fluorescence at 650–80 nm. HEK-293T cells expressing A₁R^wt-NL (B) and A₁R^G279S-NL (C) were incubated with increasing concentrations of CA200634 in the absence (Total) or presence (Non-specific) of 1 μM DPCPX and the fluorescent ligand binding was determined by NanoBRET (see Materials and Methods). Results are expressed as mean ± SEM of three independent experiments performed in triplicate.

Figure 3.. Agonist-induced receptor/transducer proteins coupling does not differ between A₁R^wt and A₁R^G279S.

(A) Schematic representation of the NanoBiT assay. Nanoluciferase (NL) SmBiT and LgBiT complementary fragments are fused to A₁R^wt or A₁R^wt/G279S and the different transduction proteins, respectively. As the agonist induce receptor`s coupling to transduction proteins, the NL subunits reconstitute an active NL, which becomes a real-time light reported of receptors’ coupling. The A₁R structure (in light blue; adapted from PDB: 5N2S) coupled to the mini-G_αo protein (in grey; adapted from [68]) is shown. Also, the structures of the SmBiT and LgBiT are shown in dark blue and purple, respectively (adapted from PDB: 5IBO). (B) Representative real-time profiles of agonist-induced receptor coupling to G_αi and G_αq. HEK-293T cells expressing A₁R^wt-SmBiT (blue circles) or A₁R^G279S-SmBiT (red squares) plus G_αi-LgBiT (left panel) and G_αq-LgBiT (right panel) were challenged with CPA (200 nM) and the receptor/G protein coupling was determined by NanoBiT (see Materials and Methods). Data show the NL bioluminescent signal after subtracting the vehicle signal and normalizing by basal signal (before CPA addition). Results are expressed as mean ± SD of one representative experiment. (C) Quantification of the overall coupling of A₁R^wt-SmBiT (blue circles) or A₁R^G279S-SmBiT (red squares) to G_αi/o-LgBiT, G_αq-LgBiT, G_α12/13-LgBiT, G_αs-LgBiT, β-arrestin2-LgBiT and GRK2-LgBiT as indicated in panel B. Results are shown as percentage of the area under the curve (AUC) extracted from the real-time NanoBiT profiles (shown in panel B).

Figure 4.. Ligand-induced concentration-response curves of A₁R/G_αi protein coupling.

HEK-293T cells expressing A₁R^wt-SmBiT (blue circles) or A₁R^G279S-SmBiT (red squares) plus G_αi-LgBiT were challenged with increasing concentrations of CPA (A) or DPCPX (B) and the receptor/G_αi protein coupling was determined by NanoBiT (see Materials and Methods). Results are shown as percentage of the area under the curve (AUC) for each CPA and DPCPX concentration induced real-time NanoBiT profile (see Fig. 3B) and expressed as mean ± SEM of four and three independent experiments each performed in triplicate, respectively.

Figure 5.. Computational simulations of A₁R^wt and A₁R^G279S.

(A) Average root-mean-square fluctuation (RMSF) during 1 μs unrestrained MD simulations of A₁R^wt (3 replicas in blue) and A₁R^G279S (4 replicas in red). Detailed view of average RMSF and associated B-factor values in the I239^6.40-L253^6.54 (left) and I272^7.37-I286^7.51 (right) stretch of amino acids (two helical turns before and after S246^6.47 or G/S279^7.44, respectively). (B) Average unit twist angles (see Methods) along these stretches of amino acids in which position i was assigned to either S246^6.47 (left) of G/S279^7.44 (right) in A₁R^wt and A₁R^G279S. Boxplots of unit twist angles at the (i, i-3) turn of TMs 6 and 7 (broken line). Statistical significance was calculated by non-parametric Mann−Whitney test. (C) Contour plots of the evolution of the center of mass of amino acids T257^6.58-C260^6.61 in TM 6 during the MD simulations of A₁R^wt and A₁R^G279S. The xy plane is as defined by the Orientations of Proteins in Membranes (OPM) [69]. Distributions of the x and y values are shown on the right x-axis and top y-axis, respectively. The inward movement of the top of TM 6 is illustrated in the right panel. (D) Sequence logo of class A GPCRs showing the highly conserved CWxP^6.50 motif. A₁R replaces Cys by Ser to form the receptor specific SWxP^6.50 motif. (E) Detailed views of a water-mediated hydrogen bond between S246^6.47 and S279^7.44 in A₁R^G279S (bottom panel) and a highly conserved water molecule that energetically stabilizes the strong Pro^6.50-kink distortion in TM 6 of class A GPCRs by forming hydrogen bonds with the backbone carbonyl at position 6.47 and the backbone N-H amide at position 6.51 present in both A₁R^wt (top) and A₁R^G279S (bottom). The percentage of presence of these water molecules are shown.

Figure 6.. Impact of G279^7.44S mutation on A₁R/A_2AR heteromerization.

(A) A₁R-A_2AR heteromer formation in living cells. HEK-293T cells transiently transfected with A_2AR-LgBiT and A₁R^wt-SmBiT or A₁R^G279S-SmBiT were processed for NanoBiT assay. The NanoBiT signal (RLU) was normalized by co-transfected CFP (fluorescence). Results are expressed as mean ± SEM of six independent experiments performed in quadruplicates. **P < 0.01 Student t-test. (B) Constitutive activation of adenylyl cyclase by A_2AR. HEK-293T cells permanently expressing A_2AR-NL were transiently transfected with A1R^wt, A1R^G279S or empty vector (mock) and the basal cAMP accumulation determined as described in Materials and Methods. Results were normalized by the basal cAMP accumulation in mock transfected cells and expressed as means ± SEM of the of four independent experiments each performed in triplicate. *P < 0.05, one-way ANOVA with Tukey’s post-hoc test. (C) A_2AR agonist-induced concentration-response curves of cAMP accumulation. HEK-293T cells transfected as descried in panel B were challenged with increasing concentrations of CGS21680 and the cAMP accumulation determined as described in Materials and Methods. Results were normalized by the basal cAMP accumulation in mock transfected cells and expressed as means ± SEM of the of four independent experiments each performed in triplicate. ^#(F_(6,56) = 8.368) P < 0.0001, extra sum-of-squares F test.

Figure 7.. Representative structures obtained in MD simulations of the A₁R^wt-A_2AR and A₁R^G279S-A_2AR heteromers.

(A) The movement of TMs 5 & 6 forming the TM 5/6 interface (blue for A₁R^wt/A_2AR and red for A₁R^G279S/A_2AR) is shown, whereas the other TM helices correspond to the initial model (in gray). Evolution of the center of mass of amino acids M177^5.35-Y182^5.40 in TM 5 and T257^6.58-C260^6.61 in TM 6 of either A₁R^wt or A₁R^G279S and P173^5.34-F180^5.41 in TM 5 and N253^6.55-F258^6.60 in TM 6 of A_2AR during three replicas of unbiased 1 μs MD simulations (100 structures collected every 10 ns in each replica) and their distributions. The xy plane is defined as in Figure 6. (B) The movement of A_2AR relative to A₁R^wt or A₁R^G279S within the heteromer was monitored by the center of mass of A_2AR in the MD simulations of the A₁R^wt/A_2AR and A₁R^G279S/A_2AR heteromers. Black arrows represent the movement of TM helices in the presence of the single point G279^7.44S mutation (A₁R^G279S, in red) relative to wild type sequence (A₁R^wt, in blue).