Pathway-Focused Profiling of Oligodendrocytes Over-Expressing miR-125a-3p Reveals Alteration of Wnt and Cell-to-Cell Signaling

Abstract

MicroRNAs are small post-transcriptional regulators that modulate gene expression by directly interacting with their target transcripts. Since the interaction between miRNAs and target mRNAs does not require a perfect match, one single miRNA can influence the expression of several genes and lead to a very broad array of functional consequences. Recently, we identified miR-125a-3p as a new regulator of oligodendrocyte development, showing that its over-expression is associated to impaired oligodendrocyte maturation. However, whether and how miR-125a-3p over-expression is causally related to oligodendrocyte maturation is still obscure, as well as the pathways responsible for this effect. To shed light on this issue and to identify the underlying molecular mechanisms, we determined the transcriptomic profile of miR-125a-3p over-expressing oligodendrocytes and, by means of two complementary bioinformatic approaches, we have identified pathways and biological processes consistently modulated by miR-125a-3p alteration. This analysis showed that miR-125a-3p is involved in the regulation of cell–cell interactions and Wnt signaling. By means of pathway-focused PCR arrays, we confirmed that miR-125a-3p induces changes in the expression of several genes encoding for adhesion molecules and gap junctions, which play key roles in oligodendrocytes after exposure to pathological demyelinating stimuli. Moreover, the expression changes of different Wnt targets suggest an over-activation of this pathway. Globally, our studies show that miR-125a-3p over-expression can alter signaling pathways and biological processes essential for myelin formation in oligodendrocytes, suggesting that alteration of miR-125a-3p levels may contribute to impairing oligodendrocyte maturation in demyelinating diseases.

Electronic supplementary material

The online version of this article (10.1007/s10571-020-00836-z) contains supplementary material, which is available to authorized users.

Introduction

In the central nervous system, oligodendrocytes are specialized glial cells that produce myelin, a lipid structure that surrounds axons ensuring their insulation and the saltatory conduction of nerve impulses. Axon-to-oligodendrocyte signals are essential for physiological myelination and contribute to orchestrate the terminal stages of oligodendrocyte differentiation by regulating the expression of both cell–cell and cell-extracellular matrix adhesion molecules, such as integrins, laminins, contactins, and connexins (Laursen et al. 2009). In myelinating glia, gap junctions (GJs) are involved in many physiological processes, including the regulation of cell growth, cell permeability, and calcium signaling, and they also play a fundamental role in myelin maintenance (Nualart-Marti et al. 2013).

Oligodendrocyte development is tightly regulated by intrinsic time-dependent mechanisms that control cell proliferation and maturation via signaling cascades, ending with the activation of several factors that can directly modulate gene expression. Among these regulators, microRNAs emerged as pivotal players in the regulation of oligodendrocyte development, both under physiological and pathological conditions (Marangon et al. 2019). miRNAs control gene expression by binding to complementary sequences in the 3′ untranslated regions of their target messenger RNAs. This interaction does not require perfect complementarity between miRNAs and mRNA sequences; indeed, a 6-base match may be enough to induce mRNA degradation or translational inhibition. This apparently reduces miRNA target specificity, but, on the other hand, it also enables one single miRNA to modulate several biological processes by targeting multiple genes in a shared pathway (Gurtan and Sharp 2013).

In this respect, we recently identified miR-125a-3p as a new regulator of oligodendrocyte maturation, showing that its over-expression decreases the production of the myelin basic protein (MBP), one of the major myelin components (Lecca et al. 2016). Although this former study highlighted the deleterious effect of miR-125a-3p over-expression on OPC maturation, the underlying molecular mechanisms are still unknown. In the present study, we aimed at analyzing the transcriptome of oligodendrocytes over-expressing miR-125a-3p taking advantage of two complementary approaches (pathway- and ontology-based) to identify the over-represented signaling pathways and biological processes and, eventually, by means of pathway-focused PCR arrays, to determine the role of miR-125a-3p in their regulation.

Results

miR-125a-3p Over-Expression in OPCs Inhibits TCF7L2 Signaling

To explore the mechanisms underlying the effects of miR-125a-3p on oligodendrocyte maturation, we performed a transcriptomic analysis of differentiating OPC cultures after miR-125a-3p over-expression. We considered a threshold of ± 1.5-fold change and an FDR-adjusted p-values < 0.05 to obtain the list of genes that are differentially expressed in the two experimental conditions (miR-125a-3p transfected OPCs vs. scramble RNA transfected OPCs). This analysis allowed the identification of 1060 genes that significantly changed after miR-125a-3p over-expression (Fig. 1).

Fig. 1.

Gene expression profiling of OPCs after miR-125a-3p over-expression. Primary-cultured OPCs were grown for 3 days in the presence of growth factors; then the OPCs were transfected with the negative control or miR-125a-3p mimic in the presence of the T3 hormone to promote cell maturation and were lysed 48 h after transfection, for the subsequent RNA extraction and microarray analysis. The microarray revealed 1060 genes significantly modulated after miR-125a-3p over-expression

The list of the differentially expressed genes was first analyzed by means of a pathway-oriented approach. The Ingenuity® Pathway Analysis software (IPA®, QIAGEN) was used to perform an upstream regulator analysis that allows to identify master regulators of gene expression that may be responsible for the observed changes in the experimental dataset, and to understand whether they are likely activated or inhibited. The analysis showed that miR-125a-3p over-expression can influence the activity of several regulators in OPCs, such as cytokines, growth factors, and transcription factors (Table 1). Among the potential candidates, we focused on TCF7L2, an essential effector of the Wnt pathway that can also promote oligodendrocyte maturation in a Wnt-independent manner (Hammond et al. 2015). As suggested by the very low Z score (− 5.186), the TCF7L2 signaling was strongly inhibited after miR-125a-3p over-expression. This observation was based on the downregulation of 27 TCF7L2-related genes in our experimental data set (Fig. 2).

Table 1.

Upstream regulator analysis

Master regulator	Molecule type	Prediction	Z Score	p-value	Number of targets
TCF7L2	Transcription factor	Inhibited	− 5.186	5.21E−08	27
BDNF	Growth factor	Inhibited	− 2.381	4.23E−06	26
CXCL8	Cytokine	Inhibited	− 1.219	8.54E−06	8
1L1B	Cytokine	Activated	1.349	3.26E−05	30
IFNG	Cytokine	Activated	2.541	1.26E−04	36
TGFB1	Growth factor	Activated	1.451	1.69E−04	55
LIF	Cytokine	Activated	1.362	3.29E−04	14
STAT1	Transcription factor	Activated	2.592	3.95E−04	14
IL6R	Receptor	Activated	2.236	6.06E−04	7
MKNK1	Kinase	Activated	2.714	6.47E−04	11

The over-expression of miR-125a-3p in OPCs caused a statistically significant change in the expression of 1060 transcripts. These differentially expressed genes were analyzed with the Ingenuity Pathway Analysis tool (IPA) by performing an upstream regulator analysis to identify possible alterations in the activity of gene expression master regulators that may be responsible for the changes observed in the experimental dataset. The table shows the most promising master regulators resulting from the analysis, with the relative p-value and number of targets modulated. Cut-off: Z =  ± 1. In bold, the transcription factor TCF7L2, selected for subsequent studies is reported.

Fig. 2.

miR-125a-3p over-expression represses TCF7L2 signaling in OPCs. The scheme shows that the alteration of multiple genes after miR-125a-3p over-expression could be related to TCF7L2 signaling inhibition. The arrows represent negative interactions (inhibition). All the transcripts were found downregulated in the experimental dataset. Shape intensity directly correlates to the entity of the downregulation

Based on these results, we hypothesized that the over-expression of miR-125a-3p leads to Wnt signaling over-activation, preventing TCF7L2 pro-myelinating Wnt-independent effect. To assess this hypothesis, we utilized the pathway-focused “Wnt signaling targets” array that simultaneously allowed to profile the expression of 84 key genes responsive to WNT signal transduction (see Table S1 for the full gene list and fold changes). After miR-125a-3p over-expression, changes were detected in 24 out of 84 genes (Fig. 3a). The eight downregulated genes (fold regulation < − 1.5) were the following: Fgf7, Nrp1, Mmp2, Sox2, Met, Plaur, Tcf7l1 ,and Fst. The 16 upregulated genes (fold regulation > 1.5) were the following: Wisp1, Ctgf, Dlk1, Egr1, Gja1, Cdh1, Wisp2, Klf5, Wnt3a, Igf1, Ptgs2, Abcb1a, Cubn, Fn1, Tcf7, and Twist1. The change in the expression levels of the Wnt ligand Wnt3a and Wnt effectors Tcf7, Tcf7l1, and Tcf7l2 was validated by qPCR (Fig. 3b). The upregulation of Wnt3a and TCF7, and the downregulation of Tcf7l1 and Tcf7l2 confirmed an over-activation of the Wnt pathway.

Fig. 3.

Evaluation of the expression of Wnt targets after miR-125a-3p over-expression in OPCs. a Scatter plot comparing the normalized expression of each gene of the “Wnt targets” PCR array between the two groups (miR-125a-3p over-expression vs. negative control). Upper dots represent genes upregulated, lower dots represent genes downregulated, and black dots indicate unchanged genes. Cut-off = ± 1.5. b Histogram shows Log2FC of Sox2, Wnt3a, Tcf7, Tcf7L1, and Tcf7L2 after mir-125a-3p over-expression in OPCs (control set to 0; n = 5 for each group). One sample t-test; *p < 0.05; **p < 0.01; ***p < 0.001

miR-125a-3p Over-Expression in OPCs Alters the Expression of Adhesion Molecules and Gap Junctions

The list of differentially expressed genes after miR-125a-3p over-expression was also evaluated by means of a Gene Ontology-based analysis, which allows to identify common biological processes and functions for differentially expressed genes. The gene ontology-based analysis showed that miR-125a-3p over-expression alters several biological processes related to oligodendrocyte maturation and myelination (Table 2). The most significant biological processes, “Role of cell–cell and ECM-cell interactions in oligodendrocyte differentiation and myelination” and “Cell adhesion–Gap junctions”, were studied more in detail by means of specific PCR arrays (i.e., “Extracellular Matrix & Adhesion Molecules” and “Gap Junctions), that allowed to simultaneously profile the expression of 84 genes important for these processes (see Table S2, S3 for the full gene list, fold changes and p-values). In the “Extracellular Matrix & Adhesion Molecules” array, after miR-125a-3p over-expression, statistically significant expression changes were detected in 23 out of 84 genes (Fig. 4a, b). The seven downregulated genes were the following: Catna1, Cdh2, Col2a1, Mmp2, Tgfbi, Syt1, and Timp3. The 16 upregulated genes were the following: Cd44, Fn1, Entpd1, Fbln1, Hapln1, Itgad, Itgam, Mmp14, Ncam1, Sell, Sparc, Spp1, Timp1, Timp2, Thbs2, and Lamb2. In the “Gap Junctions” array, after miR-125a-3p over-expression, statistical significance expression changes were detected in 19 of 84 genes (Fig. 5a, b). The 13 downregulated genes were the following: Lpar1, Cx32, Cx36, Grb2, Itpr2, Mapk3, Nras, Prkacb, Raf1, Sos2, Tjp2, Tubb4a, and Tubb2b. The 6 upregulated genes were the following: Cx43, Gucy1a2, Gucy1b3, Ma2k2, Map3k2, Tubg1. Globally, these results suggest that over-expression of miR-125a-3p alters the expression of several genes associated to cell-ECM and cell–cell communication, likely contributing to impair oligodendrocyte maturation and myelination.

Table 2.

Ontology-based clusterization with Metacore

Biological process	Total	FDR p-value	No of targets	Targets in the dataset
Role of cell–cell and ECM-cell interactions in oligodendrocyte differentiation and myelination	34	9.1E−05	9	Claudin-11, PLP1, Connexin 43, Connexin 26, Reticulon 4, MAG, GJC3, Myelin basic protein, Connexin 32
Cell adhesion_Gap junctions	30	1.6E-02	6	PKC, Connexin 43, Connexin 31, Connexin 26, Connexin 32, Tubulin beta
Role of Thyroid hormone in regulation of oligodendrocyte differentiation	48	2.6E−02	7	TR-beta, PLP1, p73, MOG, MAG, Myelin basic protein, OATP-A
ATM/ATR regulation of G2/M checkpoint	26	3.1E−02	5	Chk1, Chk2, Cyclin B, Claspin, GADD45 beta
Substance P-stimulated expression of proinflammatory cytokines via MAPKs	43	3.9E−02	6	PKC-delta, PLC-beta, Substance P extracellular region, CCL13, c-Jun, GRO-2
Oxidative stress_activation of NADPH oxidase	59	3.9E−02	7	PKC, PKC-delta, p47-phox, PLC-beta, p22-phox, TRIO, Rac2
HDL-mediated reverse cholesterol transport	44	3.9E−02	6	Pre beta-1 HDL, Nascent HDL, Large apoE-rich HDL, APOA1, PLTP, APOE
Cytoskeleton remodeling	102	5.3E−02	9	Fibronectin, MyHC, MYLK1, MLCK, Collagen I, TGF-beta receptor type I, TRIO, c-Jun, LIMK2

The software Metacore was used to perform an ontology-based analysis of the 1060 differentially expressed genes, in order to identify common biological processes and functions. In table the resulting more significant biological processes and the relative associated genes from the experimental dataset are reported. In bold, the biological processes selected for subsequent studies are reported.

Fig. 4.

Evaluation of the expression of Extracellular Matrix & Adhesion Molecules after miR-125a-3p over-expression in OPCs. a Volcano plot of the “Extracellular Matrix & Adhesion Molecules” PCR array. The relative fold change of each gene was calculated by comparing the expression in miR-125a-3p over-expressing and scramble RNA-treated OPCs (N = 3 for each group). Log2(fold change) are plotted against -Log10(p-value). Green and red indicators show downregulated and upregulated genes, respectively. The blue line indicates a p-value of 0.1. b Histogram shows Log2(FC) of selected adhesion molecules and ECM genes (control set to 0). One sample t-test; ^§p < 0.1; *p < 0.05; **p < 0.01

Fig. 5.

PCR array expression profile of Gap junctions-related genes after miR-125a-3p over-expression in OPCs. a Volcano plot of the “Gap Junctions” PCR array. The relative fold change of each gene was calculated by comparing the expression in miR-125a-3p over-expressing and scramble RNA-treated OPCs (N = 3 for each group). Log2(fold change) are plotted against -Log10(p-value). Green and red indicators show downregulated and upregulated genes, respectively. The blue line indicates a p-value of 0.1. b Histogram shows Log2(FC) of selected components, interactors, and regulators of gap junction (control set to 0). One sample t-test; ^§p < 0.1; *p < 0.05; **p < 0.01

Discussion

The identification of miRNA targets is an important means to elucidate their mode of action and eventually to develop new miRNA-based drugs (Marangon et al. 2019). However, it is worth mentioning that the action of a miRNA on a single specific gene transcript can be rather modest and totally insufficient to explain its pleiotropic biological effects. miRNAs can indeed target hundreds of transcripts serially, simultaneously, and in concert with other transcriptional and epigenetic factors acting on the same pathways. On one hand, this is the reason why a single miRNA can lead to a very powerful effect on cell survival, proliferation, and differentiation (Ebert and Sharp 2012). On the other hand, due to this very peculiar mode of action, when studying miRNAs, it is necessary to adopt a global comprehensive approach focusing on entire biological pathways rather than single direct targets.

Several transcripts involved in pathways connected to myelination have been so far identified as direct targets of miR-125a-3p, such as Fyn kinase (Ninio-Many et al. 2013), which play a key role in regulating oligodendrocyte myelination during development (Peckham et al. 2016) and NRG1 (Yin et al. 2015), a potent chemoattractant that selectively regulates OPC migration and the extent of myelination during early CNS development (Ortega et al. 2012). In a previous study, we demonstrated that the inhibitory effect of miR-125a-3p on oligodendrocyte maturation in terms of expression of myelinating genes was significantly higher than the effect of the same miRNA on its already validated direct targets, including Fyn and Nrg1 (Lecca et al. 2016), suggesting an effect on multiple signaling cascades leading to terminal maturation. For this reason, the aim of the present study was to elucidate the mechanisms regulated by miR-125a-3p during OPC maturation by focusing on entire biological pathways rather than single direct targets. To this purpose, we performed a transcriptomic analysis after its over-expression and then we analyzed differentially expressed genes by means of two complementary approaches to identify common pathways and biological processes (Marangon 2018).

A first pathway-based analysis suggested that the expression changes of several genes in our dataset may be related to TCF7L2 signaling inactivation. TCF7L2 is an important effector of the Wnt/β-catenin pathway, that, as recently shown, can also act in a Wnt-independent manner by interacting with other co-factors (i.e., Kaiso and Sox10), to promote oligodendrocyte maturation (Zhao et al. 2016). Since it is widely known that constitutive activation of Wnt/β-catenin inhibits oligodendrocyte maturation (Hammond et al. 2015), we hypothesize that miR-125a-3p over-expression induces an over-activation of the Wnt pathway, which, in turn, prevents TCF7L2-mediated pro-myelinating effects. Our qPCR data on Wnt signaling after miR-125a-3p over-expression in OPCs revealed an upregulation of the Wnt ligand Wnt3a, in parallel to upregulation of Tcf7 and downregulation of Tcf7l1 and Tcf7l2 downstream effectors, expression changes that have been previously associated to Wnt signaling over-stimulation (Kuwahara et al. 2014), in line with our hypothesis.

A second ontology-based approach allowed us to identify common biological processes for differentially expressed genes, revealing that miR-125a-3p can modulate several processes related to oligodendrocyte maturation and myelination, such as “adhesion molecules and ECM proteins,” “gap junctions,” and “thyroid hormone signaling”. The PCR arrays “Extracellular Matrix & Adhesion Molecules” and “Gap Junctions” were used to simultaneously profile the expression of 84 genes involved in these processes. Our data show that miR-125a-3p over-expression in OPCs alters the expression of several classes of ECM and adhesion molecules, such as catenin (Catn), collagens (Col), integrins (Itg), laminins (Lam), and metalloproteinases (Mmp). Despite some of these classes of molecules may have a dispensable role during OPC in vitro maturation, several reports in the literature show that such alterations are usually associated to dysmyelinating conditions, suggesting that our data could have a high relevance in vivo. For example, it has been reported that disruption of integrin-ECM connection leads to aberrant process and myelin sheath formation (Olsen and Ffrench-Constant 2005). Moreover, antibodies blocking β1-integrin reduce the ability of OPCs to extend their processes in vitro (Buttery and ffrench-Constant 1999). Interestingly, one of the downstream mechanisms that mediate integrin effects on OPC morphological differentiation is the activation of the Fyn kinase, which, in turn, regulates several downstream signaling, such as Rac1 and RhoA (O'Meara et al. 2011), suggesting that, also in this case, miR-125a-3p can influence several actors involved in a common pathway. Other reports have shown that β1-integrin can also form a functional signaling unit by associating to contactin 1 (Cntn1) and regulate Fyn activation by controlling its phosphorylation state (Laursen et al. 2009). These data suggest that miR-125a-3p over-expression not only directly influences Fyn expression, but can also indirectly control its activity by modulating upstream signaling molecules.

The “ECM & Adhesion Molecules” expression profile highlighted an upregulation of genes that are typically expressed by other glial cells, such as CD44 and Spp1, which have been reported to be induced also in oligodendrocytes following various insults to the nervous system, including demyelinating conditions (Tuohy et al. 2004). Interestingly, the oligodendroglial over-expression of CD44, a transmembrane glycoprotein expressed by astrocytes and microglia in the CNS, causes a strong reduction in the number of myelinated fibers, leading to a dysmyelinating phenotype (Tuohy et al. 2004). Moreover, a recent study demonstrated that CD44 is a positive regulator of canonical Wnt signaling (Schmitt et al. 2015). The gene Spp1 encodes for Osteopontin (OPN), a secreted glycoprotein with cytokine-like, chemotactic and anti-apoptotic properties that activates CD44 itself (Selvaraju et al. 2004). In EAE, the administration of recombinant OPN induces relapses, whereas treatment with anti-OPN antibodies ameliorates the disease (Hur et al. 2007). Several studies have demonstrated that MS patients present a strong increase of OPN levels in CSF and blood, in particular in the active phase (Agah et al. 2018). Moreover, in patients with active MS, who underwent disease-modifying treatments, the levels of anti-OPN antibodies were higher than in untreated patients and were associated with low MS severity score (Clemente et al. 2017).

Furthermore, downregulation of the tissue inhibitor of metalloproteinases Timp3, as well as upregulation of Mmp14, Col1a1, Col3a1, Fn1, Lamb2, and Thbs2 were instead found in human MS active lesions (Haddock et al. 2006; Mohan et al. 2010), suggesting that miR-125a-3p over-expression can recapitulate some typical pathological features that OPCs acquire in demyelinating environment. In line with this hypothesis, we have previously found upregulated levels of miR-125a-3p in the CSF of MS patients in relapsing phase (Lecca et al. 2016).

We have also shown that the over-expression of miR-125a-3p alters genes encoding components, interactors, and regulators of GJs, such as pannexin 2 (Panx2), connexins (Cx29, Cx32, Cx43), tubulins (Tubb4a, Tubb2b), surface receptors (Lpar1, Egfr), and protein kinase (Pkc, Nras, Raf1, Map2k2, Map3k2, Mapk3). In the CNS, glial cells express different sets of connexins, which, by forming heterodimers, allow the direct passage of ions and small molecules between different cell types (i.e., oligodendrocyte-to-astrocyte coupling, O–A). Oligodendrocytes mainly express Cx47, Cx32, and Cx29 that may also participate in the formation of oligodendrocyte-to-oligodendrocyte (O–O) gap junctional coupling (Wasseff and Scherer 2011). The GJs expression profile highlighted a global alteration in mitogen-activated protein kinase (MAPK) pathway, as suggested by the change in the expression of Nras, Raf1, MEK2 (Map2k2), MEKK2 (Map3k2), and ERK-1 (Mapk3). Interestingly, it has been previously established that Cx subcellular trafficking, GJ gating, function, and turnover are phosphorylation-dependent and that the mechanism involves the recruitment of MAPK family members (Chen et al. 2013).

We also observed a strong downregulation of Lpar1 (lysophosphatidic acid receptor 1), a G-protein-coupled receptor known to modulate the formation of processes in differentiating oligodendrocytes, allowing their terminal maturation (Garcia-Diaz et al. 2015). Mice lacking the LPA1 receptor exhibit a reduction in cortical oligodendrocytes and defective quantity, quality, and organization of myelinated fibers (Garcia-Diaz et al. 2015). Interestingly, LPA1 receptor stimulation activates MAPK pathway in oligodendrocytes by coupling to Gq subunits and activating PLC and PKC pathways (Yu et al. 2004). Our data show that miR-125a-3p indeed acts at different levels of the GJs signaling, by simultaneously affecting the expression of several key players in the pathway, such as surface receptors (Lpar1), intracellular kinases (MAPKs, PKC), and connexins (Cx47, Cx32 and Cx29).

Of note, the importance of oligodendroglial GJs in myelin formation and maintenance is also demonstrated by the finding that their genetic mutation is related to human disorders characterized by a dysmyelinating phenotype (Papaneophytou et al. 2018). In line with these observations, we hypothesize that the alteration in the expression levels of GJ genes observed after miR-125a-3p over-expression may prevent oligodendrocytes from reaching the myelinating state. Considering that miRNAs do not necessarily lead to the degradation of target mRNAs, a future proteomic analysis after miR-125a-3p over-expression will help to interpret and strengthen these transcriptomic data.

In conclusion, these data suggest that miR-125a-3p modulates different aspects of OPC development, playing an essential role in cell differentiation, myelination, and dysmyelination. The identification of the pathways altered by miR-125a-3p will allow not only to define the networks orchestrating these mechanisms, but also to unveil new pharmacological targets to foster re-myelination in diseases characterized by myelination defects.

Methods

OPC Isolation and Transfection

Primary oligodendrocyte precursor cells were obtained from post-natal day 2 Sprague–Dawley rat cerebral cortices (Chen et al. 2007). Cortical tissues were incubated with 10 ml trypsin–EDTA solution containing 1% DNAse I (final concentration 0.01 mg/ml) (Sigma-Aldrich) for 30 min in a water bath at 37 °C for tissue disaggregation. After the incubation, trypsin was inactivated with HBSS containing 10% of fetal bovine serum (FBS, Euroclone) and tissues were further triturated mechanically with a Pasteur pipet. The cellular suspension was passed through a 100 μm cell strainer (BD) in order to eliminate undissociated tissue residues. Cells were plated in T75 poly-D-lysine (final concentration 10 μg/ml, Sigma-Aldrich)-coated flasks containing DMEM high glucose (Euroclone), 2 mM L-glutamine (Sigma-Aldrich), 1 mM Sodium pyruvate (Sigma-Aldrich), Penicillin 100 U/ml-Streptomycin 100 μg/ml (Euroclone), 2.5 μg/ml Fungizone (Euroclone), and 20% FBS.

After 8 days in culture, flasks were shaken for 3–4 h to promote OPC detaching. The OPC cell suspension was then transferred into a 50 ml tube, centrifuged at 1200 rpm for 15 min, and resuspended in a small amount of Neurobasal (Life Technologies) containing 2 mM L-Glutamine, 1% of Penicillin, 100 U/ml-Streptomycin, 100 μg/ml, and 2% of B27 (Life Technologies). OPCs were plated onto poly-d,l-ornithine-coated (final concentration 50 µg/ml; Sigma-Aldrich) 6-cm dishes (1.5 × 10⁵ cells) for PCR array and qRT-PCR experiments. Cells were plated in Neurobasal medium supplemented with 2% B27 (Life Technologies), 2 mM l-glutamine, 10 ng/ml human platelet-derived growth factor BB (Sigma-Aldrich), and 10 ng/ml human basic fibroblast growth factor (Life Technologies) to promote proliferation for 3 days. OPCs were transfected immediately after switching from proliferating to differentiating medium (in the presence of the T3 hormone). MiR-125a-3p mimic (Dharmacon) was transfected at the final concentration of 50 nM with Lipofectamine RNAiMAX reagent (Life Technologies) following the manufacturer’s protocol. A scrambled miRNA transfection was included as negative control. Cells were lysed with RLT buffer (Qiagen) 48 h after transfection. For each independent experiment, four post-natal day 2 rats were sacrificed to obtain the necessary number of OPCs.

Microarray

Total RNA was extracted by means of RNeasy Micro kit (Qiagen) following the manufacturer’s protocol. RNA quality was assessed with Agilent 2100 Bioanalyzer (Agilent Technologies). RNA with RNA integrity number (RIN) > 7 was used for microarray analysis. Labeled cRNA was synthesized from 100 ng of total RNA using the Low Input Quick-Amp Labeling Kit, one color (Agilent Technologies) in the presence of cyanine 3-CTP. The microarray hybridization was performed by the Microarray Facility, Laboratorio per le Tecnologie delle Terapie Avanzate (LTTA), Ferrara, Italy. Total RNA was hybridized on SurePrint G3 Rat Gene Expression Microarrays (#G4858A-074036, Agilent Technologies). This microarray consists of 60-mer DNA probes synthesized in situ, which represent 30,584 rat transcripts. Hybridization was performed at 65 °C for 17 h in a rotating oven. One-color gene expression analysis was performed according to the manufacturer’s procedure. Feature Extraction 10.7.3 software (Agilent Technologies) was used to obtain microarray raw data. A fold change of ± 1.5 and a FDR-adjusted p-value < 0.05 were considered to obtain the list of genes differentially expressed between the two experimental conditions. Datasets and raw data are publicly-available in GEO Profile (GEO ID: 200143876).

Bioinformatic Analysis

QIAGEN’s Ingenuity® Pathway Analysis (IPA®, QIAGEN Redwood City, www.qiagen.com/ingenuity) was used to perform the upstream regulator analysis on differentially expressed genes after miR-125a-3p over-expression. Z score > 2 indicates that the signaling guided by a transcriptional regulator is likely activated, whereas Z score < 2 indicates that it is likely inhibited in the experimental condition.

The software Metacore™ was used to perform an ontology-based clusterization on differentially expressed genes after miR-125a-3p over-expression in OPCs, to identify common biological processes.

Pathway-Focused PCR Arrays

Total RNA was extracted by means of RNeasy Micro kit (Qiagen) following the manufacturer’s protocol. The PCR arrays “Wnt signaling targets” (PARN-243ZD), “Extracellular Matrix & Adhesion Molecules” (PARN-013ZD), and “Gap Junction” (PARN-144Z) were used to identify genes differentially expressed in OPCs after miR-125a-3p over-expression compared to negative control (See Supplementary material for the full gene list). For each PCR array, cDNA synthesis was performed starting from 500 ng of DNase pre-treated total RNA using RT2 First Strand Kit (Qiagen), following the manufacturer’s protocol. RT2 Profiler PCR Array (SABiosciences) and RT2 SYBRgreen Mastermix (Qiagen) were used to measure gene expression levels. Each array includes five housekeeping genes, that enable normalization of data, a genomic DNA control, that specifically detects genomic DNA contamination, a reverse transcription control, that tests the efficacy of the reverse transcription reaction, and a positive PCR control, that tests the efficacy of the polymerase chain reaction itself. Data were analyzed by RT2 Profiler PCR Array data analysis center v. 3.5 (Qiagen).

Total RNA Extraction, Retrotranscription, and qPCR Analysis

Total RNA was extracted by using RNeasy plus micro kit (Qiagen) by following the manufacturer’s protocol. For gene expression analysis, cDNA synthesis was performed starting from 400 ng of total RNA using SensiFAST™ cDNA synthesis kit (Bioline). The expression of Sox2, Wnt3a, Tcf7, Tcf7l1, and Tcf7l2 was assessed by means of pre-validated PrimePCR assay (Biorad) and SensiFAST™ SYBR Supermix (Bioline) using CFX96 real time PCR system (Biorad) following the manufacturer’s protocol. Relative gene expression was calculated by the ΔCt method normalizing to GAPDH expression.

Statistical Analysis

Data are presented as Log2(fold change) mean ± SEM and were analyzed with the GraphPad Prism 7.04 software. Shapiro–Wilk normality test was performed to assess normal distribution of data. One sample two-tailed t-test was performed to assess statistical significance of data. p < 0.05 was considered as statistically significant. Data with 0.05 < p-value < 0.1 were indicated with § symbol.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Author Contributions

DM designed and performed in vitro experiments, in silico analysis, and qPCR, analyzed the data, prepared the figures, and wrote the manuscript; MPA contributed to data interpretation and discussion and revised the manuscript. DL supervised the project, designed the experiments, and wrote the manuscript. All authors read and approved the final manuscript.

Compliance with Ethical Standards

Conflict of interest

The authors declare that they have no conflict of interest.

Ethical Approval

International (European law Dir. 2010/63/UE) and national (Italian law DL n. 26, 4th March 2014) guidelines for the care and use of animals were followed. All the procedures were approved by the Italian Ministry of Health (735-2015PR to DL).

Research Involving Human Participants

This article does not contain any studies with human participants performed by any of the authors.