miR-20a Promotes the Axon Regeneration of DRG Neurons by Targeting Nr4a3

Dear Editor,

The peripheral nervous system (PNS) regenerates more easily after injury than the central nervous system (CNS) [1]. Sensory neurons in the L4–L6 dorsal root ganglia (DRGs) extend axons to form the sciatic nerve along with motor axons. The DRG neuron is one of the exceptional mature neurons whose axons can regenerate after injury. Following sciatic nerve crush or transection, these sensory neurons rapidly re-extend axons into the lesion site [2]. The switch of DRG neurons from a transmitter to a regenerative state following injury requires the induction of widespread changes through transcription factors, epigenetic modifiers, and microRNAs. Understanding the molecular mechanisms of PNS regeneration may provide clues for the development of new therapeutic strategies for CNS regeneration.

Several studies have indicated that miRNAs play potentially vital roles in different cellular processes. miRNAs are endogenous non-coding small RNAs consisting of 21–23 nucleotides, which bind to the 3′-untranslated region (3′-UTR) of their target mRNAs, leading to the translational inhibition or degradation of target genes [3]. In the PNS, a number of miRNAs participate in neuronal activity; for example, decreased miR-325-5p contributes to chronic visceral pain through CCL2 in DRGs [4], miR-222 promotes neurite outgrowth by targeting PTEN [5], miR-146b increases the proliferation and migration of Schwann cells, and promotes axonal outgrowth via the inhibition of transcription factor Krüppel-like factor 7 [6]. There is a puzzling phenomenon that has not been fully deciphered in axon regeneration. Injuring the peripheral axons triggers axon regeneration in DRG neurons, but injuring the central axons does not. Recently, the expression of miR-20 was found to be down-regulated after injuring the central branch of DRG neurons in the spinal cord dorsal column lesion model [7]. But in our study, we found that the expression of miR-20a was up-regulated after sciatic nerve injury in vivo, and was similarly up-regulated in DRG neurons in vitro. These contradictory expression patterns of miR-20a presumably explain the differential activation of axon regeneration and further elucidate the molecular mechanisms of PNS regeneration.

Up-Regulation of miR-20a

Previous studies have shown that miRNA expression profiles change dramatically in DRGs after sciatic nerve injury [8]. To rule out the influence of other cell types, we purified DRG neurons with 15% BSA in vitro. Agilent microarrays were used to profile the miRNAs and mRNAs in cultured DRG neurons at 0, 3, 6, 12, 18, and 30 h after the cells were plated, which was intended to mimic the process of axon regeneration after injury [9]. Random variation model screening demonstrated that the expression profiles of 45 miRNAs were changed at 3, 6, 12, 18, and 30 h compared with 0 h (Fig. 1A) (GSE173839 microarray). Meanwhile, we obtained 3,085 changed genes during this process from the mRNA microarray (GSE173839). To further explore the function of these altered miRNAs, 1,824 potential downstream targets of these 45 miRNAs were predicted by the TargetScan database. Across the results of mRNA array and database prediction, 335 candidate targets were obtained. The relationships between these 335 candidates and the 45 miRNAs are shown as a network in Fig. S1 and KEGG pathways of the 335 candidate targets are listed in Table S1. Some significant KEGG pathways, such as axon guidance and neuroactive ligand-receptor interaction, were strongly associated with neuronal activities.

Fig. 1.

miR-20a promotes axon regeneration in DRG neurons. A miRNA microarray results showing the expression profiles of 45 miRNAs change at 3, 6, 12, 18, and 30 h compared to 0 h. B qRT-PCR showing the expression of miR-20a at 0, 3, 6, 12, 18, and 30 h during DRG neuron culture (RNU6B (U6) is control; n = 3). C In situ hybridization with miR-20a probe (miR-20a) or blank (control) on the DRG sections at 0 and 4 days after sciatic nerve injury (purple, miR-20a expression; red, nuclei; scale bar, 100 μm). D Electroporation of cultured DRG neurons with the miR-20a mimic or inhibitor, together with GFP plasmids. Left panels: red, Tuj1-labelled neurons; green, GFP-labelled successfully-transfected neurons; blue, nuclei counterstained with Hoechst 33342; scale bar, 100 μm; middle panels: black-and-white images converted from color images to their left; right panels: the length distributions of new neurites. E Maximum length of neurites of each neuron in control (n = 45), miR-20a (n = 45), anti-control (n = 41), and anti-miR-20a (n = 50) groups. F Overview of the time points of the experimental procedure. G Immunohistochemistry showing SCG10 (red) and Tuj1 (green) 3 days after crush of the sciatic nerve (dotted lines, crush site; scale bar, 1000 μm). H Histogram of the lengths of new axons (n = 3). I Relative grey values for the density of new axons at the same distance in the miR-20a and control groups (n = 3). Data are presented as mean ± SEM; *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001.

Among the altered miRNAs, we found that miR-20a, which has been reported to promote neurite extension in the cerebral cortex, was up-regulated [10]. After DRG neurons were plated, the expression of miR-20a was persistently up-regulated from 3 h to 30 h as assessed by the quantitative real-time polymerase chain reaction (qRT-PCR; Fig. 1B), consistent with our microarray results (Fig. S2A). Moreover, we verified the expression of miR-20a at 0, 3, and 9 h, and 1, 4, and 7 days after sciatic nerve injury in vivo (Fig. S2B), consistent with the results in vitro. As a supplementary, in situ hybridization of DRG sections revealed that miR-20a was mainly located in neurons rather than Schwann cells, and its expression at 4 days after injury was higher than at 0 days (Fig. 1C). Together, these results indicate that miR-20a is up-regulated during the axon regeneration of DRG neurons.

miR-20a Promotes Axon Growth In Vitro

To test whether the up-regulation of miR-20a is necessary for axon regeneration after injury, we electroporated cultured DRG neurons with either miR-20a mimics or inhibitors designed to mimic or inhibit the function of endogenous miRNA. Electroporation of cultured DRG neurons with mimic control (control), miR-20a mimics (miR-20a), inhibitor control (anti-control), or miR-20a inhibitors (anti-miR-20a), together with GFP plasmids, which served as a reporter of successful transfection. Based on reported methods, we recorded the length of neurites to assess the function of miR-20a [11]. Immunofluorescent staining of Tuj1 revealed that the new neurites were significantly longer than the mimic control when transfected with miR-20a mimics (Fig. 1D left) and the length distribution of these neurites confirmed this result (Fig. 1D right). In the opposite situation, transfection with miR-20a inhibitors impaired neurite outgrowth (Fig. S2C). We also recorded the maximum (Fig. 1E) and total length (Fig. S2D) of neurites of each neuron, both of which indicated that the over-expression of miR-20a promotes neurite outgrowth while the suppression of miR-20a impairs it.

miR-20a Promotes Axon Regeneration In Vivo

Previous observations suggested that miR-20a promotes neurite outgrowth from cultured neurons, but whether it had a similar action in vivo was not known. To address this issue, we established a DRG injection model to investigate the function of miR-20a in vivo. We injected agomir, a stabilized mimic modified with cy5 fluorescence, into rat L4–5 DRGs. Phosphate-buffered saline (PBS) was used as a blank control. To ensure the efficiency of infection, agomir or PBS were injected twice in two days, the sciatic nerve was crushed on day 5, and the regeneration of axons was assessed on day 8 (Fig. 1F). Based on the detection of cy5 fluorescence, we primarily confirmed the successful injection of miR-20a agomir in DRGs with no fluorescence in the PBS group (negative control, NC; Fig. S2E). SCG10 was used to label new axons 3 days after sciatic nerve injury. Compared with the control group, the length of regenerated axons was increased after miR-20a agomir injection (Fig. 1G, H). In addition, we compared the gray value, which indicated the density of new axons at the same distance in the control and miR-20a groups, and this revealed that the miR-20a group had more new axons (Fig. 1I). The above results suggest that the overexpression of miR-20a enhances axon regeneration after sciatic nerve injury in vivo.

Nr4a3 is the Target of miR-20a

It is known that miRNA exerts its regulatory function by limiting the expression of its target. We next searched for potential targets of miR-20a via intersecting the miRNA targets predicted from the TargetScan database with the altered mRNAs from the mRNA array. Eleven candidates were chosen for their high correlation (Table S2). Among these candidates, Nr4a3 (nuclear receptor subfamily 4 group A member 3, also known as NOR-1) has been shown to regulate axon guidance and neurite extension in the brain [12]. In addition, the expression of Nr4a3 mRNA and protein gradually decreased in DRG neurons after the cells were plated (Fig. 2A, B), and this was negatively correlated with miR-20a. Western blot and qRT-RCR analysis further showed that Nr4a3 was suppressed at the mRNA and protein levels when cells were treated with miR-20a mimics (Fig. 2C, D).

Fig. 2.

Nr4a3 is a downstream target of miR-20a. A, B qRT-PCR analysis of mRNA (A) and Western blots of protein (B) of Nr4a3 in cultured DRG neurons (n = 3). C, D Decreased mRNA (C) and protein expression (D) of Nr4a3 after treatment with miR-20a mimic (n = 3). E Predicted target sites of miR-20a on the 3′-UTR of Nr4a3 together with the wild and mutant sequences. F Relative luciferase activity after the dual-luciferase vectors carrying the wild or mutant 3′-UTR of Nr4a3 were co-transfected with miR-20a. G qPCR analysis showing mRNA expression of Nr4a3 when transfected with siRNA-1 and siRNA-2. H Images of cultured DRG neurons were transfected with siRNA-1 and siRNA-2. Left panels: immunofluorescence of Tuj1 (red); scale bar, 100 μm; middle panel: black-and-white images converted from the color images to their left; right panels: length distributions of new neurites. I Histogram showing the length distributions of new neurites after treating cultured DRG neurons with inhibitor control (anti-control), miR-20a inhibitor (anti-miR-20a), and both miR-20a inhibitor and siRNA-2 of Nr4a3 (anti-miR-20a + s2). Data are presented as mean ± SEM. *P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001.

Based on the TargetScan database, we found that the 3′-UTR of Nr4a3 contained two predicted target sites of miR-20a. The full-length 3′-UTR of rat Nr4a3 was inserted into the pmiR-RB-REPORT™ dual-luciferase vector with firefly luciferase (h-Luc) and Renilla luciferase (h-Rluc). The mutant reporter contained mutated Nr4a3 3′-UTR which was produced by changing the matching sequence from GCACTTT to CGTGAAA (Fig. 2E). The miR-20a mimics were co-expressed with either wild-type reporter or mutant reporter into 293FT cells. We found that overexpression of miR-20a impacted the expression of the wild-type reporter luciferase, but not that of the mutant reporter (Fig. 2F). These results demonstrated that miR-20a specifically represses the expression of Nr4a3 by targeting the Nr4a3 3′-UTR directly.

Down-Regulation of Nr4a3 Improves Axon Growth

To determine the role of Nr4a3 in axon regeneration, we used two specific small-interfering RNAs (siRNAs) to suppress Nr4a3 expression. DRG neurons were transfected with two siRNAs, and after 48 h, qRT-PCR showed that both siRNA-1 and siRNA-2 remarkably reduced the expression of Nr4a3 (Fig. 2G). Immunofluorescent staining of Tuj1 and neurite length distribution showed that the new neurites were much longer when treated with siRNA-1 and siRNA-2 (Fig. 2H), and they also promoted the total length of the neurites of single neurons compared with controls (Fig. S2F). Therefore, down-regulated Nr4a3 promoted the neurite regeneration of cultured DRG neurons, as expected from the effect of up-regulated miR-20a. Next, we co-expressed an miR-20a inhibitor and Nr4a3 siRNAs in DRG neurons to determine whether the neurite growth promotion of miR-20a was caused by Nr4a3 repression. The decrease of neurite outgrowth by the miR-20a inhibitor was restored by siRNA-2 of Nr4a3, as reflected in the neurite length distribution (Fig. 2I) and the total length of neurites of each neuron (Fig. S2G). These results revealed that Nr4a3 is not only a structural target of miR-20a, but also delivers the promoting effect of miR-20a on neurite regeneration in DRG neurons.

Nr4a3 is a member of the Nr4a family, which contains NGFI-B (Nr4a1, Nur77 or TR3), Nurr1 (Nr4a2), and NOR-1 (Nr4a3). This family has been reported to play important roles in neuronal diseases and cancers via diverse pathways, such as GSK3/β-catenin, PI3K-mTOR, P53, and HIF1-α [13]. For example, deletion or inactivation of GSK3β promotes axon regeneration in the adult mammalian CNS [14], activation of PI3K-mTOR signaling by deleting the negative regulator PTEN also induces CNS axon regeneration [15]. Nevertheless, the functions of Nr4a3 in the PNS have not yet been completely explored, and pathways which Nr4a3 takes part in are still unknown. In our work, we found a novel role of Nr4a3: reducing its expression with siRNA promoted axon outgrowth in vitro. As an upstream regulator of Nr4a3, miR-20a was up-regulated in cultured DRG neurons and after sciatic nerve injury. Over-expression of miR-20a enhanced neurite outgrowth in DRG neurons in vitro, and axon regeneration after injury in vivo. In addition, suppression of Nr4a3 mimicked the up-regulating effect of miR-20a on the regeneration of axons in DRG neurons. This study enriches our understanding of the functions of microRNA in axon regeneration, and the role of the target gene Nr4a3 in the PNS.

Supplementary Information

Below is the link to the electronic supplementary material.

Conflict of interest

The authors declare that they have no conflict of interest.