Disclosing common biological signatures and predicting new therapeutic targets in schizophrenia and obsessive–compulsive disorder by integrated bioinformatics analysis

Schizophrenia (SCZ) is a complex and severe mental illness mainly characterized by a combination of positive symptoms (delusions and hallucinations), negative symptoms (social withdrawal, paucity of spontaneous speech and amotivation) and debilitating cognitive deficits [1]. Overall, the prevalence of SCZ in both sexes is approximately equal, but with earlier onset and greater severity in males than in females [2]. Environmental and social factors such as childhood trauma and social isolation predispose individuals to SCZ [3, 4]. However, it is highly heritable (~ 80%) with complex polygenic architecture which both common and rare genetic variants contribute to its etiology [5]. Emerging evidence has suggested that SCZ has a common molecular etiology with other psychiatric disorders such as obsessive–compulsive disorder (OCD), autism spectrum disorders (ASD) and bipolar disorder (BD), despite their unique clinical characteristics [6].

OCD, with the prevalence of about 1–3% in the general population, is a long-lasting and devastating mental disorder characterized by recurrent, persistent and unwanted thoughts, images or impulses called obsessions and repetitive behaviors called compulsions [7]. The World Health Organization has ranked it as one of the most debilitating disorders worldwide as it results in poor quality of life and can substantially impair the patient’s occupational, marital, emotional and social functioning [8, 9]. Moreover, plenty of epidemiological studies demonstrated that co-occurrence rates in OCD are generally higher than rates of other comorbid disorders [10‐12].

In the past years, the categorizing of psychiatric disorders has been rearranged through the increasing a lot of studies that emphasize co-occurring and/or comorbid disorders [13]. Relating to SCZ, the specific co-occurrence of obsessive–compulsive symptoms (OCS) and SCZ has been revealed for more than a century (in 1878) [14]. A meta-analysis study indicated that nearly 38.3% of SCZ patients experience anxiety disorders during their illnesses. Also, the prevalence of OCD in these patients was reported at 12.1% [15]. In the same way, another meta-analysis reported that the prevalence of OCS and OCD in SCZ patients is 30.3% and 12.3%, respectively [16]. These results along with abundant research that focus on the presence of OCD and OCS among the SCZ patients emerged the concept of “schizo-obsessive disorder” as a new specific clinical entity [17‐19]. Meanwhile, the accurate examination of these patients manifested that they have more severe psychotic and depressive symptoms, lower social functioning and higher suicidality [13]. Although the major psychiatric disorders are very debilitating, early diagnosis and treatment can substantially reduce the unfavorable outcomes of them [7]. Therefore, identifying more possible biomarkers and effective drugs can be essential for reducing the severity of these disorders.

There are two hypotheses for the co-occurrence of OCS during the course of SCZ. One assumes that second-generation antipsychotics, especially clozapine, might exacerbate or generate second-onset OCS. The second is an important role of genetic risk factors that dispose of patients with SCZ to develop OCS [20]. In the past decade, studies using both gene data and genome-wide association study (GWAS) have debated that some genes may be responsible for the co-occurrence of SCZ and OCS/OCD. [20‐22]. However, as far as we know, there has been no bioinformatic study performed with a special focus on the common genes between these disorders. In a recent study, O’Connell et al. displayed common genetic etiology for SCZ, BD, ASD and OCD. They proposed that more research on shared components of these disorders is needed to obtain actionable and translatable results in order to combat psychiatric disorders [6].

Although there have been many studies done on SCZ and OCD, the common pathogenesis of them is not been well identified at the molecular level until now. Based on this point, we assume that the co-occurrence of SCZ and OCD is rooted in the genetic similarities and it may reveal the shared genetic basis, potential biomarkers and therapeutic targets between the two disorders. In this regard, we conducted a comprehensive bioinformatics analysis to identify common genes, molecular functions, cellular components and biological pathways along with predicting transcription factors (TFs) and posttranscriptional regulator microRNAs (miRNAs) as well as to repurpose candidate drugs for both SCZ and OCD.

Previous studies have shown that SCZ is closely related to OCD [13, 15‐19]. In the present study, we used globally accessible databases (Geneweaver and Harmonizome) for exploiting related genes of SCZ and OCD to survey (by conducting bioinformatic analyses) the hypothesis that these two mental disorders display prospective significant shared genetic basis, potential biomarkers and therapeutic targets in terms of common genes. Initially, we found a big common gene set (with 89 genes) between SCZ and OCD which can be evidence for the common pathogenesis and co-occurrence of these two disorders. The results of enrichment analysis on the common genes showed that ten genes (BDNF, SLC6A4, GAD1, HTR2A, GRIN2B, DRD2, SLC6A3, COMT, TH and DLG4) are the most central genes in associated with both disorders.

BDNF was found as the most central shared gene (Fig. 1). This gene encodes BDNF protein which is a member of a large family of neuronal growth factors called neurotrophins and plays a pivotal role in neurogenesis, differentiation and cell survival [33]. Meanwhile, it contributes to the transcription and translation of proteins involved in the synaptogenesis, development and stability of synapses [34]. BDNF has been widely studied in psychiatric disorders and its possible roles in the pathophysiology of them were discussed in the last two decades. It is released at the synapse and affects synaptic plasticity, subsequently induces critical changes in cognitive functions, learning and memory [35, 36]. While, a defect in the regulation of BDNF release can cause abnormalities in the underlying brain processes and cognitive dysfunctions in the psychiatric disorders [37]. BDNF signaling may substantially promote the structure and functioning of several neural circuits involved in the modulation of various neurotransmitter systems, including the dopaminergic [38], serotoninergic [39] and GABAergic [40] systems, all closely related to SCZ. In this regard, the normal development of these systems may disturb by dysfunction of BDNF-TrkB signaling during critical developmental periods, consequently leading to physiological dysregulation and vulnerability to SCZ[41]. Relating to OCD, it has been demonstrated that sequence variants in the BDNF gene are strongly associated with OCD [42]. For example, it was reported that the Val66Met BDNF gene variant contributes to the OCD pathobiology [43, 44]. In addition, more studies have shown that this variant lead to functional differences and consequently decrease the activity of the BDNF system, which can be a risk factor for OCD [45].

Furthermore, the interface between SCZ and OCD can be explained through the serotoninergic neurotransmission as two important common genes are SLC6A4 and HTR2A which have critical functions in the serotoninergic pathway. SLC6A4 by the encoding of serotonin transporter (5-HTT) has a crucial role in the regulation of serotonin via reuptake of it from synaptic clefts [46]. Because this gene is involved in the pathogenesis of SCZ, it is considered as a candidate gene for SCZ [47]. It has been reported that mRNA and protein levels of 5-HTT were changed in SCZ patients compared with healthy subjects [48, 49]. Likewise, there is a lot of evidence reporting that serotonergic system is implicated in the pathophysiology of OCD, in particular well-known anti-obsessional efficacy of selective serotonin reuptake inhibitors (SSRIs) have suggested an important role for 5-HTT in the etiology of OCD [50]. Zitterl et al. showed a significant reduction in 5-HTT availability in the brain regions of OCD patients [51]. Clearly, several studies demonstrated that there is a significant association between the SLC6A4 polymorphisms and OCD susceptibility [52]. The HTR2A gene encodes the serotonin receptor type 2A (5-HTR2A) and is abundantly expressed in the glutamatergic neurons and GABA-ergic interneurons in the prefrontal cortex and hippocampal regions which both neurotransmission systems are well known to be associated with SCZ [53]. It has been shown that 5-HTR2A is involved in the pathogenesis of SCZ as its activation regulates both dopaminergic and glutamatergic transmission in the brain [54]. Relating to OCD, findings from candidate gene studies demonstrated that HTR2A is a most important gene for development of the disorder [55]. Also, recent meta-analytic evidence showed that polymorphic variants within this gene are significantly associated with OCD pathogenesis [56, 57].

GAD1 is one of the other important genes in both SCZ and OCD. It encodes the glutamic acid decarboxylase-67 (GAD67) enzyme in multiple cortical regions (particularly in the prefrontal cortex and hippocampus) which is responsible for most cortical γ-Aminobutyric acid (GABA) synthesis [58]. This gene with a lower expression level in the SCZ subjects and consequently lower levels of GAD67 mRNA and protein is a well-known biomarker for SCZ [59]. In OCD patients, a recent clinical study reported that GABA abnormalities can be found within the anterior cingulate cortex [60]. Besides, Zhang et al. indicated that GABA concentration in the prefrontal cortex contributes to the psychopathology of OCD [61].

According to our results, another candidate gene for co-occurrence of SCZ and OCD is GRIN2B encodes the NR2B subunit of N-methyl D-aspartate (NMDA) glutamate receptors and may play an important role in synaptic plasticity, circuit formation and brain development [62]. Recently, it has been reported that variations in GRIN2B can be associated with SCZ which may be due to abnormalities of the NR2B subunit and consequently altered function of NMDA receptors [63]. In addition, hypofunction of NMDA receptors has been suggested as a mechanism in the pathogenesis of SCZ, based on the point that noncompetitive antagonists of the NMDA receptors, such as ketamine and phencyclidine, induce SCZ-like symptoms [64]. In contrast, it is proposed that neuronal excitotoxicity resulting from hyperactive NMDA receptors has an important role in SCZ [65]. Clinical and preclinical evidence have suggested that dysregulation of glutamatergic system contributes to the etiology of OCD [66]. In imaging studies of OCD, it has been shown that a hyperglutamatergic dysfunction may lead to abnormalities in the cortico–striatal–thalamo–cortical circuits [67]. In addition, it is demonstrated that the GRIN2B gene is implicated in OCD as mutations of it have been associated with the disorder in males [68] and the presence of contamination obsessions and cleaning compulsions [69].

The DRD2 gene, coding for dopamine D2 receptor, is an attractive candidate gene for SCZ due to its role in dopaminergic system [70]. Several polymorphisms of this gene have been identified related to SCZ, hence it is well considered as a causative factor in SCZ [71]. Besides, DRD2 likely contributes to the OCD pathology since it has been demonstrated that the dopaminergic system is implicated in inducing or aggravating the symptoms of OCD [72] and in particular it is reported that the DRD2 A2 allele is significantly higher in male OCD patients compared to controls[73].

Other important shared genes between SCZ and OCD are SLC6A3, COMT, TH and DLG4. The SLC6A3 gene encodes the dopamine transporter and can be determinative in the regulation of dopamine in the synaptic cleft. It has been well indicated that its polymorphisms are risk factors for SCZ [74]. Also, the manifestation of OCD symptoms can be associated with SLC6A3 [75]. Since the COMT enzyme has a crucial role in the metabolism of dopamine, the COMT gene is considered as an important factor in the etiology of SCZ [76]. Higher dopamine levels are likely involved in OCD, hence COMT can be a suitable candidate for OCD [77]. Tyrosine hydroxylase (TH), encoded by TH gene, is a rate-limiting enzyme that produces dopamine in the brain and can be related to SCZ [78]. Dopamine neurotransmission is likely associated with OCD [77]. Therefore, the TH gene may be involved in OCD. Post-synaptic density protein 95 (PSD95), encoded by the DLG4 gene, has an essential rolein regulating NMDA receptor activity and altered expression of DLG4 has been revealed in the post-mortem brain of subjects with SCZ [79]. In OCD pathogenesis, since glutamatergic system can be disrupted [80], DLG4 may be an interference factor in this system.

In our study, the significant molecular functions exhibit receptor activity as the most common of them associated with SCZ and OCD. As recently reviewed, multiple serotonin receptors (5-HTR) are implicated in SCZ, including 5-HT1AR, 5-HT2CR, 5-HT3R, 5-HT6R, 5-HT7R and several studies link 5-HT2AR to the pathogenesis of SCZ as its activation regulates both dopaminergic and glutamatergic transmission [54]. It is worth mentioning in SCZ that the more-focused and well-known receptors are dopamine D2 receptors involved in the mesolimbic dopamine pathway, and their hyperactivity is responsible for the cause of the positive symptoms of SCZ, with all antipsychotics acting to block them [81]. Also, Dopamine D2 receptors likely play a part in the OCD pathology as decreased striatal D2/D3 and D1 receptors were reported in patients with OCD [82]. Positron emission tomography (PET) showed a significant reduction of 5-HT2AR availability in OCD patients compared to healthy subjects [83].

Based on cellular component enrichment analysis, the cellular components involved in the transition of the signals, namely, neuron projection and synapse were the most important cellular components in both SCZ and OCD. As mentioned, serotonergic synapse markers (5-HTT and 5-HT2AR) contributed to the development of OCD. Besides, the neurocircuitry of OCD is located between different brain regions which are densely innervated by neuron projections of the serotonergic and dopaminergic neurotransmitter systems [84]. Also, Rosoklija et al. reported the structural abnormalities of dendrites in SCZ and mood disorders [85]. In addition, Faludi and Mirnics showed that synaptic disturbances are important and integral part of SCZ pathophysiology [86].

KEGG pathway analysis indicated that the identified common genes mostly contributed to serotonergic synapse, cocaine addiction, dopaminergic synapse, amphetamine addiction, alcoholism, taste transduction, glutamatergic synapse, tyrosine metabolism and estrogen signaling pathways. As discussed above, several studies have shown that serotonergic pathway has a substantial role in the pathophysiology of SCZ and OCD. In addition to the dopamine and glutamate theory of psychosis, there is also the serotonin theory states that cortical 5-HT2AR hyperfunction can also result in psychosis [81].

TF analysis predicted that HMGA1, MAPK14, HINFP and TEAD2 are the more relevant TF biomarkers associated with both SCZ and OCD. According to the gene cards database (www.​genecards.​org); HMGA1 is associated with Type 2 Diabetes Mellitus and Multiple Lipomatosis. Likewise, Diseases associated with MAPK14 include Chlamydia and Alexander Disease. HINFP show an association with Cerebro-oculo-facio-skeletal Syndrome 2 and Fetal Alcohol Spectrum Disorder. Diseases associated with TEAD2 include Sveinsson Chorioretinal Atrophy and Multiple Acyl-CoA Dehydrogenase Deficiency [87].

This study predicted 121 significant miRNAs for common genes through miRDB database (Fig. 3). Among them, hsa-miR-3121-3p, hsa-miR-495-3p, hsa-miR-369-3p, hsa-miR-186-5p, hsa-miR-218-5p, hsa-miR-22-3p, hsa-miR-330-3p, hsa-miR-543, hsa-miR-1271-5p, hsa-miR-96-5p, hsa-miR-148b-3p, hsa-miR-152-3p, hsa-miR-148a-3p, hsa-miR-19b-3p and hsa-miR-19a-3p were more relevant identified biomarkers in associated to both SCZ and OCD. In human, almost 70% of miRNAs express in the nervous system [88] and they have a role in the regulation of neural structure and function such as formation of dendrites and dendritic spines, axon growth, neural developments and the maturation process [89]. Also, abnormal expressions of miRNAs could be important as they are involved in the occurrence of neuropsychiatric disorders [90]. Therefore, considering them as potential biomarkers for co-occurrence of SCZ and OCD can be useful for more diagnostic and therapeutic targets.

A class of non-coding RNA molecules as miRNAs act as negative regulators of post-transcriptional processes that mediate gene expression profile and subsequent biological functions. In the present study, we predicted 121 miRNAs with target predicting score of more than 95 concerning SCZ and OCD. Pan et al., showed that hsa-miR-144-3p is associated with schizophrenia through a decrease in the expression of ATPase Na⁺/K⁺ transporting subunit beta 2 (ATP1B2) and PI3K/Akt/mTOR signalling in rat hippocampus [91]. It has been reported that miR-3121-3p suppressed RAP1 GTPase activating protein (Rap1GAP) [92], while this protein is crucial for development of dentate gyrus and prevention of SCZ-like behavior in mice [93]. Importantly next-generation sequencing and real-time quantitative polymerase chain reaction (qRT-PCR) on peripheral blood cells in patients with SCZ, suggested hsa-miR-22-3p as a biomarker for these patients [94]. Interestingly, the level of hsa-miR-22-3p also significantly increased in patients with OCD compared to healthy subjects [95]. A bioinformatics study on microarray dataset from brain post-mortem samples of SCZ patients identified hsa-miR-26a-5p as a differentially expressed miRNA compared to control samples [96]. Also, comparison of expression pattern of several miRNAs in peripheral blood cells by qRT-PCR revealed a significant increase in the expression of miR-26a-5p in OCD patients compared to control subjects [97].

Finally, current research identified drugs (particularly Haloperidol, Fluoxetine, Clozapine and Melatonin) that may have a potential influence on the co-occurrence of SCZ and OCD. Haloperidol, a highly effective first-generation antipsychotic (FGA), is one of the most used antipsychotic drugs, and it has very high anti-dopaminergic activity in the mesolimbic dopamine pathway [98] hence it is very efficient for the treatment of SCZ [99]. However, like other FGAs, it is associated with severe extrapyramidal side effects [100]. Haloperidol is considered as second in line with antipsychotic augmenting agents in OCD patients who showed an inadequate response to SSRIs but better response with antipsychotic augmentation [101]. Fluoxetine is one of the oldest SSRIs and is applied as a first-line drug for the treatment of some mental illnesses such as major depressive disorder, premenstrual dysphoric disorder, panic disorder and bulimia nervosa [102]. Furthermore, it is widely used in OCD patients due to its good therapeutic response, good compliance and low side effects [103, 104]. Also, in treating SCZ, it has been indicated that the use of antidepressants like fluoxetine as an add-on therapy to antipsychotics can improve the negative symptoms in patients with chronic SCZ [105]. Clozapine, a dibenzodiazepine developed in 1961, is an antipsychotic approved in treating resistant SCZ [106]. It has been indicated that clozapine is more effective than any other antipsychotic drug (first or second-generation) in the treatment of resistant SCZ [107]. However, the anti-serotonergic effects of clozapine in cortico-striatal serotoninergic circuits may induce OCD-like behavior in mice and generate OCS in patients treated with it [108, 109]. Therefore, it is hypothesized that decrease plasma concentration of clozapine may alleviate OCD, but it can exacerbate the severity of SCZ [20]. Clinicians should be aware of the exacerbation of OCD and OCS after chronic prescription of clozapine in treatment of SCZ. However, prescription of SSRIs along with antipsychotic medications such as clozapine and olanzapine might be a proposed alternative treatment in co-occurrence of SCZ and OCD. As, Stryjer et al. reported that administration of 20 mg/day escitalopram in patients with SCZ and OCD that were treated with antipsychotic drugs (i.e. clozapine, risperidone and quetiapine), decreased the total Yale Brown Obsessive–Compulsive Scale scores (Y-BOCS, P = 0.001) and Positive and Negative Syndrome Scale scores (PANSS, P = 0.03) [110]. Melatonin, the endogenous hormone that regulates circadian rhythms, is used exogenously for the treatment of sleep disorders [111]. Available evidence has suggested that melatonin is linked to SCZ. Sleep disorder is a usual feature of SCZ [112] and appear to be caused by abnormal melatonin functions and abnormal circadian implicated in the pathophysiology of SCZ [113]. Additionally, decreased melatonin level in SCZ patients was reported [114‐116]. A recent systematic review of melatonin use for SCZ showed that adjunctive melatonin therapy can be useful for sleep, metabolic profile and tardive dyskinesia in SCZ patients [117]. Sleep disorders are also prevalent among OCD patients, as up to 48% of them report these disorders [118]. Furthermore, Monteleone et al. showed that overall plasma concentrations of melatonin are lower in patients with OCD compared with normal controls [119]. Hence, melatonin therapy may be helpful for OCD. Our study utilized a bioinformatic method to identify common molecular and cellular mechanisms and predict new therapeutic drugs for SCZ and OCD. However, the present study has a number of limitations. One limitation was the present study failed to verify main findings with experimental results due to lack of confirmatory experimental animal studies that mimic SCZ and OCD-like behavior. Another limitation is lack of an integrated comprehensive database to support genetic basis of OCD in human subjects. Furthermore, the function of most predicted miRNAs have yet to be determined. Therefore, validation of predicted results requires further in vitro, in vivo, and especially clinical future experimental researches with large sample size.

Obsessive–compulsive behavior is a common comorbid condition with schizophrenia. Herein, we conducted a comprehensive enrichment analysis on the common genetic basis of SCZ and OCD. Regarding comorbid disorders, bioinformatics studies may help us to better understand the common pathophysiology of these disorders by identifying possible biomarkers and underlying mechanisms involved in them. Furthermore, finding more potential therapeutic options for disorders can be another implication of such analyses and future experimental studies.