GENERAL CONCLUSIONS and FUTURE DIRECTIONS
In this review, we highlighted the involvement of the different GPCRs in the aetiology of ASD and as potential targets. We analysed the effect of GPCR variants on their level of expression, ligand binding, receptor folding or activation of downstream signalling pathways (Table S2 ). Variants located in introns, untranslated regions or coding regions of mGlu7, 5-HT2A, CB1, GPR37 and GPR85 receptor genes are consistent with the decreased levels of transcript expression observed in patients (Monfared et al., 2021). Based on the evidences in ASD patients, their function and localisation (Figures 2-3 ), and their behavioural predictive validity in animal models (Table S3 ), we classified these GPCRs. OTR, V1A, mGlu5, D2, 5-HT2A, CB1 and GPR37 receptors fulfil most, if not all, criteria. V1Adisplay a clear potential for social interaction, D2 and GPR37 for stereotyped behaviours, and mGlu5, OTR, CB1 and 5-HT2A eventually for both core symptoms. Overall, it is surprising that out of 800 GPCRs, only 23 GPCRs are included in the SFARI list, and that all these genes are classified in the second category, namely ‘strong candidate gene’. We propose to move OTR, CB1 and V1A receptors to the first category ‘high confidence genes’, and add 5-HT2Ato the list. In this review, we also suggest GABAB1, 5-HT6, 5-HT7, D4 and D5 has potential candidates for ASD. Increasing pieces of evidence showed the functional crosstalk between GPCRs within a cell, to control a specific function, such as D2, A2A and mGlu5 for the control of motor activity (Ciruela et al., 2011). In fact, independently of their physical interaction, GPCRs are not individual entities but should rather be considered as a set of GPCRs and isoforms working together in a cell to orchestrate the different signals and regulate downstream signalling network and related cellular processes. Up to hundreds of GPCRs are expressed in the same brain structures or cell types, with the highest diversity in the striatum, cortex and hypothalamus (Vassilatis et al., 2003; Marti-Solano et al., 2020). In 2018, Babu and colleagues identified hundreds of missense and CNV variants in the genes coding for V1B, D1, D2, and D3, 5-HT2A, β2-adrenoceptors and GABAB (Hauser et al., 2018) that might influence their ligand binding or transducer recruitment (Table S4 ). This area of research based on receptor bias is known as pharmacogenomics. Considering the major impact of GPCR signalling (Table S5 ) that are altered in ASD (De Rubeis et al., 2014; Hormozdiari et al., 2015; Gazestani et al., 2019), any slight modification in a GPCR or in a combination of GPCRs would lead to drastic signalling defects and neuronal pathogenicity. Here, we found that at least 15% of the genes listed in the SFARI database are in the signalling networks and cellular downstream processes of GPCRs. Therefore, application of pharmacogenomics to hundreds of GPCRs expressed in the CNS remains an outstanding hypothesis to fully decipher the global effect of GPCRs on pathological processes underlying ASD. Therefore, the impact of GPCRs for autism research has only begun to be highlighted, and rather than a single entity, GPCRs should be considered as one global functional unit of GPCRs expressed in a cell that control signalling networks, in order to understand their contribution to ASD aetiology.
GPCRs meet all the criteria of therapeutic targets for ASD to bypass the placebo effect observed in clinical trials. They contribute to the polygenic ASD aetiology, pathogenic variants are recessive, they are therapeutically rescuable, and they are per essence membrane receptors that display a large pharmacopeia of safe and efficient drugs. We analysed the therapeutic potential of these 26 GPCRs and categorized them as ‘high’, ‘moderate’ and ‘low’ candidates, based on 1) drugs that are already approved or tested for a related disorder, 2) their beneficial effects in animal models of ASD or in clinical trials, 3) their pharmacogenomic profile and 4) their safety (Table S4 ,S6 ). We classified mGlu5, GABAB, D2, 5-HT2A, 5-HT7, CB1 as ‘high’ candidates and OTR, V1A, mGlu7, D1, 5-HT1B and M3 as ‘moderate’ candidates due to the lack of selective ligands. However, the high number of variants of mGlu5 and 5-HT2A might compromise their responsiveness to drugs (Hauser et al., 2018) and might explain why mGlu5-targeted clinical trials have failed. We excluded β2-adrenoceptors, A2A, and AT2 as toxicity or severe side effects have been reported in clinical trials. Finally, despite their strong potential in the future, orphan and olfactory receptors belong to the ‘low’ category as there are no natural ligands or drugs clearly identified. We also ranked CX3CR1, V1B, D3and A3 as ‘low’ candidates, as no selective ligands have been developed and the adequate pharmacological profile remains to be investigated. Finally, considering their involvement in ASD aetiology, their therapeutic potential and results of clinical trials (Figure 4 ), we conclude that D2, 5-HT2A, CB1, OTR, V1Aand GPR37 are the most promising targets for clinical development. D2, 5-HT2A and CB1 are already ongoing for irritability, repetitive behaviours, aggressive and self-injury behaviours, but could be tested on other core symptoms, especially CB1 on social scales. In the near future, when specific ligands will be developed, OTR, V1A and GPR37 should be tested as well. The recent development of antibody fragments targeting GPCRs (Mujić-Delić et al., 2014) and the emergence of high throughput screening by DNA-based bar-coded chemical libraries (Madsen et al., 2020) should boost the identification of new drugs to target GPCRs, including orphan and olfactory receptors. Interestingly, antibody fragments display all types of pharmacological profiles, and can also be used as chaperones, or target oligomers of GPCRs. Considering that GPCRs function as a global GPCR module in a cell, a similar approach might also be applied for future drug development. Finally, the use of several drugs or of a drug targeting multiple GPCRs might be relevant for ASD. Such drugs already exist. For example, aripiprazole, risperidone or cariprazine targets multiple dopamine, serotonin, histamine and/or adrenoceptors (Table S4 ) or arylpiperazine derivatives target multiple 5-HT receptors (Lacivita et al., 2021). Based on their fine-tune pharmacology (biased ligands, oligomerization, global GPCR entities) and their diversity, GPCRs represent the greatest therapeutic options for ASD and hold the promise to successful clinical trials.