GPCRs expressions are perturbed in four major psychiatric disorders
In this study, we present a comprehensive analysis of the transcriptomic signatures of G-protein-linked signaling across four major psychiatric disorders. The results reveal that the majority (three quarters) of GPCR genes were dysregulated in ASD, SCZ, BP, and MDD. In addition, when compared with non-GPCR genes, the GPCR superfamily as a gene cluster was overrepresented in the four disorders. The analysis also identifies larger GPCR gene expression changes (measured as the average absolute fold change value) in the four psychiatric disorders. Moreover, the results reveal that GPCRs differentially expressed with age tend to be more dysregulated in ASD than random GPCRs (i.e., ASD-dysregulated GPCRs are more likely to follow age-related linear regression than non-dysregulated GPCRs).
Not surprisingly, most GPCR DEs in the four psychiatric disorders belonged to the A-family, which comprises most GPCRs in humans. Noteworthy, GPCR dysregulation was most remarkable in SCZ, with around half of GPCR genes exhibiting upregulation or downregulation. However, unexpectedly, two of the only three known adhesion GPCRs were dysregulated in ASD and SCZ, and over 50% of frizzled GPCRs were dysregulated in SCZ.
At the G protein level, we found, not surprisingly, that the majority of GPCR DEs in the four psychiatric disorders are coupled to Gαi/o isoform, the most common Gα in biological systems. In BP, however, we found a lesser trend of dysregulated GPCRs to couple to Gαq compared with the non-dysregulated GPCRs. Similar to GPCRs, SCZ exhibited the highest number of dysregulated G-protein isoforms.
Disease-specific GPCR transcriptomic signatures
ASD is associated with perturbed Wnt-SMO pathways and neuropeptide transmission
GPCRs dysregulated in ASD included primarily frizzled GPCRs such as FZD1, which transduce SMO signals from the Wnt lipoglycoprotein growth factors [33]. The Wnt pathway plays a critical role in regulating neuronal precursor proliferation, neuronal migration, and fate specification during neurodevelopment. It is also involved in synaptic differentiation and mature synapse modulation [34]. Thus, perturbation of GPCR components of Wnt pathway may cause synaptic and circuit dysfunctions during neurodevelopment, which are considered at the center of the etiology of ASD. Other dysregulated GPCRs include neuropeptides' GPCRs, particularly GPCRs binding melanocortin, orexin, oxytocin, neurotensin, somatostatin, CGRP, and FPR1/2. Of these neuropeptides, oxytocin has been extensively studied in the context of ASD [35–37]. PTH2R, which encodes parathyroid hormone-2 receptor, a class B GPCR, was the second most significant GPCR DEs in ASD after FZD1. PTH2R binds in the brain to the neuropeptide TIP39 (Tuberoinfundibular Peptide of 39 Residues) and is primarily expressed in the hypothalamus and amygdala, brain regions involved in social behavior and fear [38–40]. Anxiety is a typical overlapping symptom in ASD patients [41, 42], suggesting a role for the downregulated PTHR2R gene. In support of this speculation, we found that the ADRA1D gene, which encodes the Alpha-1D adrenergic receptor (α1D-AR), was the top downregulated GPCR DE in ASD. α1D-AR, primarily found in the hippocampus and frontal cortex, exhibits strong desensitization and downregulation in response to its over-activation by noradrenaline [43, 44]. Given the known role of adrenergic receptors in mediating the response to stress [45–47], the profound downregulation of α1B-AR in ASD may reflect a chronic over-activation by noradrenaline in response to stress exposure during developmental stages.
GPCRs differentially expressed as a function of chronological age are at higher risk of dysregulation in ASD than random GPCRs
The onset of the four disorders varies from early life for ASD to late adolescence for SCZ and early adulthood and midlife for BP and MDD. Accordingly, we speculate that GPCRs' dysregulations in these disorders relate to the high dynamics and changeability of GPCR transcriptome as a function of age or in response to stimuli at critical age windows. Our speculation implies that high GPCRs' age-dependent dynamics may lower their threshold for dysregulation. Interestingly, we found that both FZD1 and PTH2R, which exhibited the most significant expression changes in ASD, were also differentially expressed with chronological age. However, their age-dependent expression patterns were in opposite directions from the patterns of their expression changes in ASD. For instance, while FZD1 expression increased in ASD, it was negatively associated with chronological age (decreased across the lifespan), whereas PTH2R expression was decreased in ASD and positively correlated with chronological age (increased across the lifespan).
We asked whether the apparent correlation between the age-dependent expression patterns of FZD1 and PTH2R and their dysregulation in ASD is a random observation or can be generalized to other GPCR DEs in ASD or even to other psychiatric disorders. In each of the four disorders, we identified DEs and non-DEs differentially expressed with chronological age and the direction of their expression across the lifespan. Our findings support a correlation between the age-related GPCR gene expression profiles and their tendency for dysregulation in ASD, but not other disorders. Accordingly, GPCRs differentially expressed with age have a higher propensity for dysregulation in ASD than random GPCRs, and ASD-dysregulated GPCRs are more likely to follow linear regression with age than non-dysregulated GPCRs.
Based on our results, we propose a GPCRs transcriptomic instability theory in ASD, wherein GPCRs transcriptomic instability (malleability) at critical developmental stages increases their tendency for dysregulation in ASD. The generalizability of this hypothesis to non-GPCR gens and other mental disorders is worth further investigation.
SCZ: A dopamine GPCR disorder?
SCZ shows the largest number of GPCR DEs. Compared with ASD DEs, we found that SCZ DEs were more enriched with biogenic amines and lipids transmissions, whereas peptides exhibited less enrichment in SCZ GPCR DEs than in ASD. In addition, a few GPCRs revealed SCZ-specific DEs, particularly the neuropeptides' receptors BDKRB2, MCHR1, KISS1R, and QRFPR.
Along with embryonic linked signaling, FZDs(1–10) also regulate adult neurogenesis [48, 49] and have a crucial role in regulating hippocampal neurons' differentiation and migration [50]. Thus, the dysregulation of nine of the Wnt-SMO pathways in SCZ indicates that early developmental stages and adolescence present a critical window of vulnerability to perturbation of these pathways SCZ [51, 52].
Although biogenic amines neurotransmitters were more enriched in adolescence- and adulthood-onset disorders (SCZ, BP, and MDD) compared to the early life disorder ASD, SCZ exhibited the most significant number of biogenic amines' dysregulated GPCRs.
Notably, the dopamine GPCRs were exclusively perturbed in SCZ, with three dopamine GPCRs (DRD2, DRD4, and DRD5) exhibiting differential expression in SCZ. The adrenergic, histaminergic, serotonergic systems were also enriched in SCZ with seven serotonergic receptors and four adrenergic receptors, two cholinergic receptors, and two histaminergic receptors being DEs in this disorder. These results provide a robust mechanistic basis for the early hypothesis of SCZ, which was centered, for many years, on the hyperactivity of the dopamine system [53, 54]. This hypothesis originated from the observations that antipsychotic medications antagonize dopamine receptors and that certain drugs that enhance dopamine activity, such as amphetamine, cause psychosis in healthy individuals or exacerbate schizophrenic symptoms [55]. Due to the high potency of second-generation antipsychotics for serotonergic, histaminergic, and α-adrenergic receptors, the SCZ hypothesis evolved to implicate other GPCRs [56–58]. However, the dopamine system remains at the center of this theory. The distinctive dysregulation of three dopamine receptors exclusively in SCZ provides a piece of mechanistic evidence that credits dopamine theory and highlights dopamine GPCRs' roles in the neuropathology of schizophrenia. Notably, other non-biogenic amine neurotransmitters GPCRs were also disrupted in SCZ, including the glutamate transmission with four metabotropic glutamate receptors exhibiting abnormal expressions in SCZ.
BP as a disorder of lipid transmission
The landscape of BP GPCR DEs is distinguished from the three other disorders in presenting extensive perturbed lipid transmission and a low tendency of dysregulation of GPCRs coupled to Gq/G11. The disrupted lipid pathways in BP presented those of arachidonic acid derivatives (prostaglandins, leukotrienes, and oxoeicosanoid) and lysophosphatidic acid and sphingosine. Arachidonic acid (AA) is an essential constituent of the cell membrane released from membrane phospholipids through GPCR-Gq-initiated activation of phospholipase A2 [59–61]. AA controls cell membrane fluidity and plays a crucial role in maintaining cell integrity and the functions of specific membrane proteins involved in cellular signaling and brain synaptic plasticity. Further, the AA-derivatives prostaglandins and Leukotrienes are released in the brain and function as neuromodulators via activation of GPCRs. Arachidonic acid turnover and signaling have been implicated in BP [62, 63], and the therapeutic effects of mood stabilizers such as lithium and valproic acid are partially related to their effects on arachidonic acid turnover [62, 64, 65].
On the other hand, the pharmacological actions of BP primary treatments are known to modulate Gαq signaling cascades, namely DAG and IP3 second messengers' pathways. Therefore, the low prevalence of dysregulated GPCRs coupling to Gq/G11 is surprising. Further, the OPRL1 gene, which encodes nociceptin/orphanin FQ (N/OFQ) receptor, is exclusively upregulated in BP. N/OFQ levels were reported to be significantly elevated in the plasma of patients with BP [66], and OFQ receptor antagonists have been proposed as a potential treatment for BP [67].
MDD is associated with upregulated GPCRs and disrupted biogenic amines transmissions
Unlike ASD, SCZ, and BP, GPCR DEs in the MDD exhibited more upregulation than downregulation. The amine theory of MDD suggests that the depletions of amine neurotransmitters may underlie the pathophysiology of MDD, and the therapeutic actions of current antidepressants rely on increasing monoamine levels via inhibiting their metabolism or uptake. Furthermore, the depletion of monoamine transmitters is known to cause a compensatory upregulation or supersensitivity of their receptors [68–71]. Therefore, our finding that most amine GPCRs were upregulated in MDD substantiates the contribution of monoamine signaling to the MDD pathophysiology and may reveal compensatory mechanisms to counterbalance the reduced levels of the monoamine transmitters [72].
One important GPCR, exclusively dysregulated in MDD (downregulated), was CRHR1. CRHR1 mediates the action of corticotropin-releasing hormone (CRH), one of the most extensively studied systems concerning its potential role in stress and depression [73, 74]. Recent data on CRHR1 antagonists suggest that this receptor might be a promising target for the treatment of MDD. Compelling evidence indicates that early stress enhances adult depressive mood through perturbed GPCRs. According to the early stress theory of depression, CRH-CRHR1 signaling initiates the hormonal stress-response pathway involving perturbed glucocorticoid (GC) signaling pathways [75, 76]. The role of early-life stress in adulthood development of depression is also mediated through other neurotransmitters GPCRs, including glutamate, GABA, endocannabinoids, and neuropeptides, several of which are dysregulated in MDD [16].
Dysregulated GPCRs overlap across psychiatric disorders
Not only did SCZ show the largest number of dysregulated GPCR genes, but its DEs showed the highest levels of GPCR transcriptomic overlap with ASD, BP, and MDD.
Interestingly, GPCR DEs shared between ASD and SCZ exhibited more extensive fold changes in ASD than SCZ. Initially, SCZ and BP were viewed as different phases of the same disorder, with ASD manifesting itself as an earlier phase of SCZ. Our study provides further evidence for the overlapping nature of these disorders and possible common etiological components and argues for a theory that the degree of the expression changes of GPCRs, and probably other genes, may determine whether the disorder manifests itself as ASD in early childhood or SCZ later in adolescence.
The largest proportion of dysregulated GPCRs were shared between SCZ and BP, with approximately 75% of BP GPCR DEs overlapping with SCZ. This high degree of overlap is in accordance with the abundant genetic and transcriptomic overlaps and the highly shared clinical features between SCZ and BD. Among the top shared GPCRs between BP and SCZ are SUCNR1, LGR6, and CX3CR1, which mediate the succinate, Wnt, and chemokine neurotactin signaling, respectively. These three receptor pathways are essential for microglia migration to their synaptic targets, pointing at a critical role of microglia functional disruptions in the pathophysiology of both ASD and SCZ.
Unlike the high overlap observed among ASD, BD, and SCZ, MDD exhibited lesser commonalities with the three other disorders suggesting that the etiology of MDD is less related to those of the three other disorders. The substantial overlap of GPCR DEs among SCZ, ASD, and BP and their age-related shared clinical manifestations provide a compelling theory that ASD, SCZ, and BP can be viewed as a single "spectrum disorder," wherein clinical manifestations vary in their explicit phenotype expression across genetically predisposed individuals.
Dysregulated GPCRs are involved in the regulation of sleep-wake and feeding
An intriguing observation was that many dysregulated GPCRs and their signaling systems were involved in sleep-wake and feeding behavior. Sleep disturbances are primary comorbid conditions across the four disorders ASD, SCZ, BP, and MDD. The majority of ASD patients suffer from sleep disturbance [77, 78], often characterized by insomnia. Similarly, SCZ and the manic phase of BP share the reduced need for sleep [79–82]. In MDD and during the depressive phase of BP, patients present both hypersomnia and severe insomnia [83–86].
The sleep-wake cycle is modulated by a complex interaction between the different transmitter and peptide systems located throughout the brain. For example, neurotransmission systems in the brainstem and hypothalamus, including serotonin, norepinephrine, dopamine, acetylcholine, glutamate, and histamine, act together to promote the generation and maintenance of wakefulness (for review, [87, 88]. On the other hand, GABA and Adenosine are crucial transmitters for sleep promotion.
Neuropeptides involved in sleep-wake regulation are synthesized and released from the hypothalamic nuclei, including orexin, galanin, melanocortin, neuropeptide Y, and melanin-concentrating hormone corticotropin-releasing hormone (CRH), and somatostatin (for review, [89]). Sleep-wake cycles are highly regulated processes under the tight control of the circadian rhythm, which is synchronized to the 24h light-dark cycle. Thus, light is the most persistent and powerful circadian entrainers across all ages. Interestingly, light/photon GPCRs (opsins) were among the dysregulated GPCRs across the four psychiatric disorders. Whether light can activate brain opsins and the functional significance of such activation are not known; however, the dysregulations of the brain light receptors, in accordance with the dysregulation of transmitters' GPCRs involved in sleep-wake regulation, provide compelling evidence for the link between GPCRs' dysregulation and sleep disturbances in the psychiatric disorders.
Strikingly, almost all the dysregulated GPCRs that mediate the signaling of brain transmitter systems in sleep are also known for their role in feeding behavior and body weight. Examples illustrating this view are GPCRs for the neurotransmitters noradrenaline, serotonin, and dopamine. Orexin, ghrelin, neuropeptide-Y, melanin-concentrating hormone, GHRH, endorphins, cannabinoid, galanin, nociceptin (OFQ) are orexigenic peptides, whereas melanocortin, GLP-1, corticotropic releasing hormone, oxytocin, neuromedin U, and neurotensin act as anorexic peptides (for review, [90, 91]). Appetite changes, bodyweight loss, and obesity have frequently been associated with psychiatric disorders. In addition, there is considerable evidence for comorbidities in psychiatric disorders with eating and appetite disorders. For example, there is a profound shift in the appetitive characteristics in depression from weight loss in the depression early stage to weight gain at later stages [92, 93]. Further, eating behavior disorders such as binge eating, anorexia nervosa, bulimia nervosa, and night eating have been considered characteristic features of schizophrenia (for review, [94]). Similarly, several studies have illustrated the prevalence of certain eating disorders such as anorexia nervosa and bulimia nervosa in ASD, and low body weight has often been seen in male subjects with ASD (for review, [95]). Fast weight fluctuation is a common feature in BP, with more than 5% loss or gain of body weight can be observed in one month, and overweight and disordered eating behaviors differentially impact BD patients [96, 97]. Since sleep and feeding behaviors are regulated by internal circadian rhythm, our results raise an intriguing hypothesis that perturbed brain GPCR signaling may play a key role in mechanistically mediating the consequences of circadian rhythm disruption in psychiatric disorders.
In conclusion, the results of our study support the growing anticipation that perturbed brain GPCR signaling pathways might be at the center of the etiological and pathophysiological mechanisms of psychiatric disorders. Furthermore, by highlighting shared disrupted GPCR pathways across ASD, SCZ, BP, and MDD, and identifying GPCRs' signaling signature, responsible for particular clinical phenotypes across psychiatric disorders, our study suggests that targeting specific GPCRs could serve as a common therapeutic strategy to treat distinct clinical phenotypes, such as social deficits or sleep disorders, across psychiatric disorders.