Peripheral whole blood microRNA alterations in major depression and bipolar disorder

Read Also

In this study, by assessing the whole miRNome expression in the blood of MD and BD patients (all drug-naïve or drug-free from antidepressants and mood stabilizers), we observed a dysregulation in a number of miRNA transcripts, some specific for MD or BD, whereas others common to both the diseases. Moreover, the RT-PCR results also evidenced significant differences, in terms of miRNA levels, between MD and BD patients.

Regarding MD, RT-PCR validation assays confirmed a significant increase of miR-24-3p and miR-425-3p levels and a decrease for let-7a-5p, let-7d-5p and let-7f-5p expression. These miRNAs were previously implicated in psychiatric diseases, as well as in neuronal molecular mechanisms and behavioural functions. In particular, miR-24-3p was suggested to be a main hypothalamic regulator of oxytocin (Choi et al., 2013), the neuropeptide that regulates several social behaviours such as stress modulation, aggressive behaviour and social recognition (Chini et al., 2014). Interestingly, miR-24-3p was found to be down-regulated in rat hippocampus following chronic treatment with two mood stabilizers, lithium and valproate (Zhou et al., 2009). Moreover, miR-425-3p, here up-regulated in MD patients, was found increased also in a similar study conducted in the peripheral blood of MD patients (Belzeaux et al., 2012). Particularly noteworthy are the data on let-7 family miRNAs. Indeed, in this study we showed a down-regulation of let-7d-5p and let-7f-5p in MD patients and interestingly, in our previous work (Bocchio-Chiavetto et al., 2013), we found them increased in the peripheral blood of MD patients after a 12-weeks treatment with escitalopram, suggesting that these miRNAs may be involved in both the pathogenesis of MD and in the effects of antidepressant drugs. On the other hand, other components of the let-7 family (namely, let-7b and let-7c) were found to be common targets of mood stabilizers in rat hippocampus (Zhou et al., 2009). Let-7 miRNAs belong to the most highly expressed miRNAs in the human brain (Anacker and Beery, 2013) and are supposed to exert a powerful influence on gene expression in the CNS. In particular, let-7 miRNAs exert a pivotal action on neuronal differentiation and maturation during neurodevelopment (Shao et al., 2010) and also on neurogenesis and neuronal plasticity functions in the adult brain.

Our data are consistent, since they showed a down-regulation in MD patients of 3 miRNAs belonging to the let-7 family that are able to regulate almost the same target genes (Fig. 3). Moreover, these 3 miRNAs are coded in the same genetic cluster on chromosome 9 (hsa-let-7a-5p, chr9 94175957-94176036; hsa-let-7d-5p, chr9 94176347-94176433; hsa-let-7f-5p, chr9 94178834-94178920), suggesting a possible impaired transcriptional co-regulation.

Concerning the BD-specific miRNAs, we found an increased blood content of miR-30d-5p, miR-140-3p, miR-330-5p, miR-21-3p and miR-378a-5p. The blood expression of miR-30d-5p and miR-140-3p was increased also in MD patients after AD treatment in our previous study (Bocchio-Chiavetto et al., 2013). With regard to miR-330–5p, it was predicted to regulate many genes involved in neuronal plasticity and neurodevelopment (Cohen et al., 2014). However, in contrast with our data, a decrease of miR-330-5p miRNA was observed in post-mortem brains of BD patients (Moreau et al., 2011) and miR-21-3p levels were found reduced in MD fibroblast cultures (Garbett et al., 2014). Concerning this discrepancy, a tissue-specific miRNA regulation may likely occur, also considering a possible inverse relationship between intracellular and extracellular miRNA content. Moreover, the pharmacological long-term treatment could affect miRNA expression in post-mortem brain samples. Finally, the findings on miR-378a-5p might be of interest, considering that this miRNA is mainly involved in lipid and metabolism homeostasis, that are probably compromised in BD patients, which indeed show an increased vulnerability to develop metabolic syndrome (McElroy and Keck, 2014).

The 2 miRNAs found significantly altered in both the diagnostic groups, miR-330-3p and miR-345-5p, are predicted to regulate several target genes with a putative role in the shared pathogenetic mechanisms between MD and BD, for example the 5-hydroxytryptamine receptor 2C (HTR2C), monoamine oxidase A (MAOA), dopamine receptor D1 (DRD1), calcium/calmodulin-dependent protein kinase 2 (CAMKK2), neurotrophic tyrosine kinase receptor, type 3 (NTRK3), clock homolog (CLOCK), cAMP responsive element binding protein 1 (CREB1), gamma-aminobutyric acid A receptor, alpha 2 (GABRA2), cannabinoid receptor 1 (CNR1), 5,10-methylenetetrahydrofolate reductase NADPH (MTHFR). Furthermore, the parallel dysregulation of these miRNAs in both the disorders suggests their involvement in depressive symptoms manifestation, since both MD and BD patients enrolled for this study are in a depressive state.

Finally, considering RT-PCR results, all the analyzed miRNAs (including MD-specific, BD-specific and commonly altered ones) showed a differential expression when directly comparing MD vs. BD patients. In particular, the levels of the 2 commonly altered miRNAs, higher both in MD and BD patients compared to healthy controls, were also significantly higher in BD vs. MD patients, with MD showing intermediate levels between controls and BD patients.

Overall, the bioinformatic analysis indicated that most of the genes potentially affected by the altered miRNAs are involved in mechanisms associated with neuroplasticity regulation and intracellular signal transduction, further supporting a role for these miRNAs in mood disorders etiology.

However, the reported miRNA alterations have been observed in peripheral blood and it is currently not clear to what extent peripheral miRNA modifications could reflect alterations occurring in the CNS. The alterations observed in the periphery might directly reflect brain modifications, since miRNAs can pass through membranes in free form or in microvesicles (Laterza et al., 2009 and Skog et al., 2008), but it is also possible that changes in blood miRNA expression are due to the alteration/normalization of systems that cause molecular and cellular changes within the brain and peripheral organs as a result of neuroendocrine or neuroimmune responses (Anacker et al., 2011 and Janssen et al., 2010). Future studies investigating miRNA levels in exosomes, which act as cell-to-cell communicators and can derive from the CNS (Sheinerman and Umansky, 2013), may be useful to clarify the observed modifications. We are also aware that the sample size of this study is small and further confirmation in larger samples is needed. Finally, most of the enrolled patients were drug-free, but not drug-naïve from psychotropic drugs, so we cannot exclude an influence of previous therapies on the observed results.

In conclusion, here we report a peripheral blood dysregulation in the expression levels of a panel of miRNAs specific for MD or BD patients, together with common alterations, which could potentially influence several pathways relevant for brain functions. The identification of the genes and biological pathways controlled by these miRNAs could provide new information for clarifying the pathogenesis of these diseases. Moreover, the described miRNA alterations may provide potential biomarkers, which could be integrated with clinical and other biological information to enhance the diagnosis and treatment of mood disorders.