Rare coding variants of the adenosine A3 receptor are increased in autism: on the trail of the serotonin transporter regulome

Abstract

Background: Rare genetic variation is an important class of autism spectrum disorder (ASD) risk factors and can implicate biological networks for investigation. Altered serotonin (5-HT) signaling has been implicated in ASD, and we and others have discovered multiple, rare, ASD-associated variants in the 5-HT transporter (SERT) gene leading to elevated 5-HT re-uptake and perturbed regulation. We hypothesized that loci encoding SERT regulators harbor variants that impact SERT function and/or regulation and therefore could contribute to ASD risk. The adenosine A3 receptor (A3AR) regulates SERT via protein kinase G (PKG) and other signaling pathways leading to enhanced SERT surface expression and catalytic activity.

Methods: To test our hypothesis, we asked whether rare variants in the A3AR gene (ADORA3) were increased in ASD cases vs. controls. Discovery sequencing in a case-control sample and subsequent analysis of comparison exome sequence data were conducted. We evaluated the functional impact of two variants from the discovery sample on A3AR signaling and SERT activity.

Results: Sequencing discovery showed an increase of rare coding variants in cases vs. controls (P=0.013). While comparison exome sequence data did not show a significant enrichment (P=0.071), combined analysis strengthened evidence for association (P=0.0025). Two variants discovered in ASD cases (Leu90Val and Val171Ile) lie in or near the ligand-binding pocket, and Leu90Val was enriched individually in cases (P=0.040). In vitro analysis of cells expressing Val90-A3AR revealed elevated basal cGMP levels compared with the wildtype receptor. Additionally, a specific A3AR agonist increased cGMP levels across the full time course studied in Val90-A3AR cells, compared to wildtype receptor. In Val90-A3AR/SERT co-transfections, agonist stimulation elevated SERT activity over the wildtype receptor with delayed 5-HT uptake activity recovery. In contrast, Ile171-A3AR was unable to support agonist stimulation of SERT. Although both Val90 and Ile171 were present in greater numbers in these ASD cases, segregation analysis in families showed incomplete penetrance, consistent with other rare ASD risk alleles.

Conclusions: Our results validate the hypothesis that the SERT regulatory network harbors rare, functional variants that impact SERT activity and regulation in ASD, and encourages further investigation of this network for other variation that may impact ASD risk.

Figure 1

Pedigree structure of ASD families harboring the Leu90Val or Val171Ile variants. Subjects harboring either variant are indicated by an asterisk. Individuals for whom DNA was not available are labeled NA. Individuals with ASD are indicated by filled circles or squares, while unfilled elements reflect individuals without an ASD diagnosis.

Figure 2

Structure of the A2a adenosine receptor identifying corresponding positions of Leu90 and Val171 residues in the A3 receptor. The structure of the thermostabilized A2a receptor, A2aAR-GL31 (PDB code 2YDO), is depicted as a cartoon in rainbow coloration (N-terminus in blue, C-terminus in red), viewed either (a) parallel to the membrane plane or (b) from the extracellular surface perpendicular to the membrane plane. The endogenous ligand adenosine is shown as a stick model (C, yellow, N, blue, O, red) as is the side chain for Phe168 (C, green). The transmembrane helices are labeled H1-H7, the extracellular loop 2 (EL2) is indicated and H8 is the C-terminal amphipathic helix lying parallel to the membrane plane. A3AR residues Leu90 and Val171 correspond to Val84 and Val171 in A2aAR, and they are both shown as space-filling models (grey).

Figure 3

Cross-species conservation at ADORA3 variant sites detected by Sanger sequencing. Amino acid sequences of the A3AR protein encoded by ADORA3 are aligned for the three variants and their flanking residues: Leu90Val, Val171Ile (cases), and Ala195Thr (controls).

Figure 4

Functional consequences of the A3AR specific agonist IB-MECA on WT-A3AR and Leu90Val-A3AR or Val171Ile-A3AR expressing cells. CHO cells co-transfected with either WT-A3AR or Leu90Val-A3AR or Val171Ile-A3AR and SERT and were stimulated using IB-MECA (1 μM). (A) Levels of cGMP production from WT-A3AR/SERT and Leu90Val-A3AR/SERT co-transfections were measured prior to and after IB-MECA stimulation. Leu90Val-A3AR/SERT cells display elevated basal cGMP production that persists over the entire 60 min time course experiment, revealing enhanced overall production of cGMP (1 tailed t-test P=0.049; n=3). Each of the three independent experiments contained a minimum of three internal replicates. (B) 60 min time-course of IB-MECA induced 5-HT uptake as measured as percent of control (wildtype; time = 0 min). CHO cells were co-transfected with either wildtype-A3AR and SERT or Leu90Val-A3AR and SERT, and 5-HT uptake was stimulated using IB-MECA (1 μM). Cells were incubated with [³H]5-HT (20 nM) for the indicated period. Over the time course (0-60 min) in Leu90Val-A3AR/SERT expressing cells, 5-HT uptake is elevated compared with the wildtype A3AR counterpart (1 tailed t-test, P=0.039; n=4). Each independent experiment contained a minimum of three internal replicates. (C) [³H]5-HT accumulation in cells co-transfected with A3AR and human SERT was measured, and indicate a failure of IB-MECA treatment of Val171Ile-A3AR/SERT to induce SERT dependent 5-HT re-uptake across all tested concentrations. Compared to the peak concentrations (0.8 μM and 1.0 μM) of IB-MECA induced 5-HT uptake in WT-A3AR/SERT, IBMECA induced 5-HT uptake was significantly reduced (for example, 1 μM IB-MECA Val171Ile: 101.8% ± 12.5 vs. WT: 136.3% ± 7.2; two-way ANOVA P=0.005; n=3-4). Significant (P <0.05) findings are indicated by an asterisk (*).