Using literature-based discovery to identify candidate genes for the interaction between myocardial infarction and depression

Myocardial infarction (MI) is a highly prevalent cardiovascular disease. The American Heart Association released a scientific statement in 2014 and recommended that depression should be considered a risk factor for adverse medical outcomes in patients with acute coronary syndrome [1]. Depression may cause many adverse outcomes, including autonomic dysfunction [2], inflammation [3], endothelial dysfunction [4, 5], hyperactivity of the hypothalamic-pituitary-adrenal axis [6], and poor compliance [7], which subsequently lead to an increased risk of MI. Both the severity and cumulative duration of depressive symptoms have a negative impact on the MI prognosis [8]. On the other hand, patients with MI may have a higher prevalence of depression [9]. In an assessment of 10,785 patients with MI performed using a structured clinical interview, depression was common and persistent in MI survivors. Major depression was identified in approximately 1 of 5 (19.8%) patients hospitalized with MI [10]. Thus, understanding the interaction between MI and depression is very important for the development of therapeutic interventions and determining patients’ needs.

The biomedical support discovery system (BITOLA) is a sophisticated bioinformatics tool that enables new discoveries, such as mining new information from the literature without using patient tissue samples, especially for identification of key candidates, and finding potentially new relationships among various biomedical concepts [11, 12]. Some researchers have used the text mining tools to identify candidate genes for diseases [13], such as multiple sclerosis and bilateral polymicrogyria [12, 14, 15]. In addition, using the BITOLA system, genes neural cell adhesion molecule 1 (NCAM1) and CD4 were identified as potential candidate genes in the interaction between depression and oral lichen planus [16].

Because the molecular mechanisms underlying the interaction between MI and depression remain unclear, the aim of the study is to identify new potential candidate genes linking these two diseases.

In this study, we present for the first time a preliminary literature mining work exploring candidate genes related to MI and depression. By integrating data from the literature, we revealed 4 genes of interest (GNB3, CNR1, MTHFR, and NCAM1) that were likely to be associated with the aetiology of both MI and depression.

G proteins play an important role in intracellular signal transduction from the cell surface [24]. A C3T polymorphism at nucleotide 825 in exon 10 of the G protein β3 subunit gene (GNB3/C825T) was demonstrated to be associated with enhanced intracellular signal transduction [25] and a variety of cardiovascular risk factors, including hypertension [25], obesity [26], dyslipidaemia [27], diabetes, and atherosclerosis [28]. An association between GNB3/C825T and MI has also been reported [29]. In addition to the roles mentioned above, studies have implicated a role for GNB3/C825T in depressive disorder [30‐32] and the efficacy of antidepressants for the treatment of major depression disorders [33]. In the present study, we found the highest GNB3 expression in the heart ventricle and cingulate cortex of the brain (Fig. 2), which was in accordance with the aetiology of depression [34] . Thus, further study of GNB3 is essential for assessment of the interaction between MI and depression.

Cannabinoid receptor 1 (CNR1) is one member of the seven transmembrane G-protein coupled receptor family and can regulate the levels of second messenger mainly through coupling with G proteins after activation by endocannabinoids [35, 36]. The CNR1 receptor may play a protective role through a wide variety of mechanisms, including inhibition of excessive noradrenaline release from the sympathetic nerve fibres [37], lowering inflammation, oxidative stress, fibrosis, and excitotoxicity, and enhancing blood flow [38]. Therefore, cannabinoid receptor agonists can be considered as a prospective group of compounds for creation of drugs that are able to protect the heart against ischaemia-reperfusion injury in the clinical setting [39]. Over the past few years, numerous studies have suggested that depression directly results in the hyperactivity of the hypothalamic-pituitary-adrenal axis [6]. Studies have also suggested that CNR1 negatively regulates the hypothalamic-pituitary-adrenal axis function [40, 41]. In addition, mice lacking CNR1 can develop depressive-like behaviours or disorders [42]. Specifically, in our study, high CNR1 expression in the brain areas was observed at the nucleus accumbens (Fig. 3), which has been suggested to be related to a lack of interest and other symptoms of depression [43]. The evidence above suggests that targeting the endocannabinoid system may evolve as a novel therapeutic concept to limit the devastating consequences of MI and depression.

Methylenetetrahydrofolate reductase (MTHFR) is a key enzyme involved in homocysteine metabolism. An elevated total plasma homocysteine level has been demonstrated to be associated with both cardiovascular disease and depression [44, 45]. Because the C-to-T transition can cause reduced enzyme activity and elevated total plasma homocysteine levels, a positive relationship may exist between the MTHFR 677 C → T polymorphism and these two diseases, which has also been demonstrated [46, 47]. This polymorphism was also associated with a risk of MI [48, 49]. Moreover, the results confirmed those of very recent meta-analyses of genome-wide association studies, suggesting that MTHFR was a genetic overlap candidate gene that likely was shared between mood disorders and cardiovascular diseases [50]. These findings provide some concrete directions for further research.

NCAM1, which is also known as CD56, is a member of the immunoglobulin superfamily [51]. NCAM1 was first identified in brain tissue and is the best surface antigen for identification of human NK cells [52]. Numerous studies have suggested that NCAM1 is a gene of interest associated with the pathogenesis of depressive disorder [52‐54]. Experimental evidence showed that NCAM deficiency in mice resulted in a depression-like phenotype that could be reversed by an NCAM-derived peptide [55]. In the present study, the NCAM1 gene was mainly expressed in the cerebral cortex and amygdala in the brain (Fig. 5), which are involved in the pathogenesis of depression [56]. In addition to its role in depression, studies have also suggested its correlations with MI [57]. One study demonstrated that NCAM1 was upregulated under metabolic stress in cardiomyocytes and suggested that NCAM1 was a cardioprotective factor [58]. Hence, this evidence may have implications for the role of NCAM1 in communication between MI and depression that warrants further exploration.

In conclusion, using literature mining methods, the GNB3, CNR1, MTHFR, and NCAM1 genes were identified and directly or indirectly implicated in the regulation of MI and depression. Although additional research is needed to confirm these findings, our study reduced the candidate causal genes to a manageable number and might present potential new clues for future research.