miR‐181d‐5p promotes neurite outgrowth in PC12 Cells via PI3K/Akt pathway

Summary

Introduction

miRNAs dysregulate in spinal cord injury (SCI) and have been demonstrated to play a crucial role in neurite outgrowth. However, the underlying mechanism remains elusive. In this study, we constructed a mouse model of SCI, extracted RNA from injured spinal cord tissue for the use of microarray assay. miR‐181d‐5p which is one of the most significantly expressed miRNAs in miRNA‐mRNA network, abundantly expressed in center system and highly conserved across different spices, was chosen for our further study.

Aims

To demonstrate whether miR‐181d‐5p can promote neurite outgrowth in PC12 cells via PI3K/Akt signaling pathway, we performed function analysis of miR‐181d‐5p with LV‐miR‐181d‐5p and LV‐sh‐GFP to infect PC12 cells.

Results

Through microarray assay analysis, we totally found 262 significantly expressed miRNAs and 2973 target genes in SCI and observed that their expression dynamically changed postinjury. Here, we provided enough evidences that the overexpression of miR181d‐5p significantly decreased the expression of PTEN, upregulated p‐Akt expression, increased neurite outgrowth‐related proteins (GAP‐43 and NF‐200) and synaptic vesicle‐related proteins (Synapsin and PSD95), and then promoted neurite outgrowth in PC12 cells. Furthermore, we confirmed that miR‐181d‐5p could directly target to the 3′‐UTR of PTEN mRNA through dual‐luciferase report assay.

Conclusions

Our study supports that aberrant expression of miRNAs is involved in the pathogenesis of SCI, miR‐181d‐5p plays an important role in neurite growth in PC12 cells via PI3K/Akt signaling pathway and may be a candidate target for the treatment of SCI in the future.

1. INTRODUCTION

Spinal cord injury (SCI) still remains a worldwide problem, which includes primary mechanical injury and secondary injury caused by various biological factors. Abnormal expression of genes after SCI is one of the most important causes of secondary injury of SCI. miRNAs are a type of noncoding RNA, which are short in length (about 17‐25nt) and highly conserved in evolution. miRNAs can regulate gene expression by preventing mRNA translation or inducing degradation through complementary pairing of oligonucleotides to target mRNAs.

As is universally known, approximately 25 000‐30 000 genes exist in human body, a third of which are estimated to be regulated by miRNAs. Each miRNA can regulate more than 150 target genes, which plays an important role in various biological processes, including cell development, necrosis, proliferation, and differentiation.1 Moreover, accumulating evidences support that some miRNAs together with their target genes participate in neuronal proliferation and development.2 For example, miRNA26a promotes neurite outgrowth through suppressing PTEN expression.3 Overexpression of miR‐124 makes great contribution to the control of neurite outgrowth during neuronal differentiation by regulation of the cytoskeleton.4

PTEN is the main inhibitor of PI3K/Akt pathway which is a classical cell survivor signaling pathway.5 PTEN inhibition promotes the intrinsic regenerative outgrowth of axons.6 miRNA‐29c participates in mice brain development and boots neurite outgrowth in PC12 cells via PTEN pathway.7 MicroRNA‐222 targeting PTEN promotes neurite outgrowth from adult dorsal root ganglion neurons following sciatic nerve transection.8 These studies strengthen the importance of the role of PTEN in neurite outgrowth.

Two freeware programs available on the internet including TargetScan and microT were applied to search for potential miRNA‐mRNA binding region, and these two freeware programs both supported that PTEN is the most probable target gene of miR‐181d‐5p. In immune system, miR‐181 regulates PI3K signaling by modulating PTEN which is required for normal lymphopenia‐induced T‐cell proliferation and for normal B‐cell development.9 In addition, it places miR‐181 as a central in vivo regulator of cellular metabolic fitness during homeostasis and development.9 In our study, through infecting PC12 cells with lentiviral LV‐miR‐181d‐5p and its control, we observed that the neurite of PC12 cells infected with LV‐miR‐181d‐5P was significantly longer than those in control group. Based on RT‐PCR or Western blot, we found LV‐miR‐181d‐5P increased levels of miR‐181d‐5p, decreased the PTEN expression, and markedly upregulated p‐Akt level. Results indicated that miRNA‐181d‐5p could modulate PTEN expression and promote neurite outgrowth in PC12 cells.

2. EXPRERIMENTAL PROCEDURES

2.1. Mouse model of spinal cord contusion injury and extraction of total RNA

Male ICR mice (20‐25 g) aged 6‐8 weeks were purchased from Experimental Center of Suzhou University, which were divided into three mice/cages and maintained in SPF environment. The mice were allowed environmental adaptation before operation for a week with normal circadian rhythms (one 12‐hour light‐dark cycle). After the mice were anesthetized, we made a cut in the middle line of the eighth to twelfth thoracic vertebrae (T8‐T12), stripped the paraspinal muscles and removed laminectomy at T10 exposing the spinal cords. All mice underwent a contusive spinal cord injury at T10 with an NYU impactor dropping from 25 mm high. Comfortable and sterile environment was provided for injured mice after the wound was sutured with absorbent suture, ensuring that they were urinated twice at every morning and evening. Five groups of mice were sacrificed at 4 hours, 1 day, 3 days, 1 week, or 3 weeks post‐SCI respectively, and spinal cord segment containing the injury epicenter (T8‐T12) was quickly removed for RNA extraction. All animal studies were finished according to the guidelines of animal research from science and technology department of China and approved by the Ethics Committee of animal research in Suzhou University.

Total RNA of the obtained spinal cord tissues and different lentiviral‐infected PC12 cells were extracted with Trizol (Life Technologies, Carlsbad, CA, USA) in accordance with the manufacturer's instructions. The RNA obtained from spinal cord tissues was used for miRNA microarray assay.

2.2. Cell culture

PC12 cells were obtained from the Type Culture Collection of the Chinese Academy of Sciences (Shanghai, China.) and grown in DMEM/F12 (Hyclone, Logan, UT, USA) containing 25 nmol/L high‐glucose, 10% fetal bovine serum (FBS) (Hyclone), and 50 μ/mL penicillin‐streptomycin solution (Hyclone).

2.3. Microarray research objective and concrete research method

Total RNA of the injured spinal cord tissues was extracted at 4 hours, 1 day, 3 days, 1 week, and 3 weeks postinjury respectively and processed into a miRNA microarray assay. Firstly, through comparing different expressed miRNAs in 1 day, 3 days, 1 week, 3 weeks group with the 4 hours group, significantly expressed miRNAs were searched and miRNAs with at least 2‐fold significantly changed were collected. Secondly, series test of cluster (STC) was applied to analyze selected miRNA expression trends, screen out the most significant and most mainstream miRNA population with the change of sample order, and find the mechanism of common expression of miRNA from the functional point of view.10, 11

Thirdly, the target genes of miRNA were predicted via miRanda and TargetScan, and their intersection was taken. Then Gene Ontology analysis (GO‐analysis), which is a cross‐species, comprehensive, and descriptive database constructed by the Gene Ontology Consortium, was used for significant functional analysis of target genes.12 Pathways analysis of target genes was also performed based on Kyoto Encyclopedia of Genes and Genomes (KEGG).13 KEGG is a database of systematic analysis of genes (and their coding products), gene function, genome information, which helps researchers to study gene and expression information as a whole network. Fisher exact test and chi‐squared test were used for significant analysis of pathways that the target genes participate in, and meaningful pathways were filtered according to P value <0.05. The data of experiment are presented as mean ± SD which was considered statistically significant when P value was <0.05.

Finally, a dynamic miRNA‐mRNA network was constructed based on the correlation analysis between the differentially expressed miRNA and mRNAs as described previously.14 In the network analysis, the degree was defined as the sum of the links that one node has to all of the other nodes. The K‐core in the graph theory was defined as a method of simplifying the analysis of graph topologies. The clustering coefficient was defined as a measure of the relationship between a gene and its neighboring genes. Then, we calculated Pearson's correlation coefficient for each pair of genes and selected significant correlation pairs to build the network.

2.4. Real‐time PCR (validation microarray results, miRNA, and mRNA)

To validate microarray results and assess miRNA expression level, we used Trizol (Life Technologies) to extract RNA from cells and tissue specimens in accordance with the manufacturer's protocols. Poly (A) tailing kit (NEW ENGLAND Biolabs, Beverly, MA, USA) was used to detect miRNAs. The ingredients contained the following: 0.5‐μg total RNA was mixed with 1 μL 10× poly (A) buffer, 1 μL ATP, and 0.5 μL E coli poly (A).We adjusted the total volume to 10 μL with RNase‐free ddH2O. RNA was converted into single‐stranded cDNA by reverse transcriptase (TAKARA, Dalian, China). The involved ingredients: 0.5‐μg total poly(A)‐tailing RNA was mixed with 4 μL 5× reaction buffer, 1 μL common dNTP, 1 μL Ribolock RNA inhibitor (2 U), 1 μL Revert Aid, and 1 μL adaptor (dt). The total volume was adjusted to 20 μL with RNase‐free ddH2O. Reverse transcription process was performed at 42°C for 60 minutes, followed by an inactivation reaction at 70°C for 10 minutes. The PCR mixture was as follows: 10 μL SYBR Green, 8 μL ddH2O, 1 μL cDNA, and 1 μL primer. Each well was totally added with 20 μL PCR mixture. A Real‐Time Instrument (Bio‐Rad, Hercules, CA, USA) of CFX96™ was utilized to perform RT‐PCR with the following primers (Table 1). Thermal cycle parameters were as below: 72°C for 20 seconds, 95°C for 300 seconds, 60°C for 15 seconds, 40 cycles of 95°C for 15 seconds, and 65–95°C to generate the curve of dissociation. An activating reaction for 900s at 70°C was utilized to define every gene's expression. The expression of miRNAs was normalized to that of U6. The related changes in gene expression via 2^−ΔΔCt method were studied, and every test was repeated for three times (n = 3). The melting temperatures (Tm) were documented, and every gene's expression was defined from the threshold cycle (Ct).

Table 1.

Primers for RT‐PCR

Primer	Sequence (5′‐3′)	Base (bp)
mmu‐miR‐182‐5p	CGGTATGGCAATGGTAGAAC	20
mmu‐miR‐181 d‐5p	GGCGGAACATCATTGTTGTC	20
mmu‐miR‐383‐3p	GGTCACAGTGAACCGGTC	18
mmu‐miR‐138‐5p	GGCCACCGCACTGCCTG	17
Universal primer	CCGAGCTGGTGTTGTGAATC	20
NF200 forward	CGGAGGGAAAGGAAG AAATAAA	22
NF200 reverse	AGAGGGACACCCAGAATAGCC	21
GAP43 forward	GCACATCGGCTTGTTTAGGCT	21
GAP43 reverse	GGAGGGAGATGGCTCTGCTACT	22
PTEN forward primer	AAGGACGGACTGGTGTAA	18
PTEN reverse primer	CCTGAGTTGGAGGAGTAGAT	20
GAPDH forward primer	GCAAGTTCAACGGCACAG	18
GAPDH reverse primer	ACGCCAGTAGACTCCACGAC	20
u6 forward	CTCGCTTCGGCAGCACA	17
u6 reverse	GCGAGCACAGAATTAATACGAC	22

GAPHD, glyceraldehyde‐3‐phosphate dehydrogenase.

To detect mRNA expression level, we converted total RNA (1 μg aliquots) to cDNA with reverse transcriptase kit (TAKARA) and diluted resultant cDNA by 1:5. The Real‐Time Instrument of CFX96™ was utilized for detection and amplification. The related expression of the mRNA levels was normalized to the glyceraldehyde‐3‐phosphate dehydrogenase (GAPDH).The software of SDS was applied to obtain CT value from the amplification plot.

2.5. miRNA‐181d‐5p target gene prediction

Among all the significantly expressed miRNAs, we chose miR181d‐5p as our research object, which is a key miRNA in building of miRNA‐mRNA‐network. Then TargetScan (http://www.targetscan.org/) and microT (http://www.targetscan.org/) were commonly used to search for potential miRNA‐mRNA binding region. These two freeware programs both support that PTEN is the most probable target gene of miRNA‐181d‐5p.

2.6. Cell transfection and lentivirus infection

To perform analysis of miR‐181d‐5p on neurite outgrowth in PC12 cells, PC12 cells were infected by LV‐miR‐181d‐5p and LV‐sh‐GFP using polybrene (5 μg/mL), respectively. Then the PC12‐miR‐181d‐5p cells and PC12‐sh‐GFP cells were cultured with puromycin(5 μg/mL). LV‐miR‐181d‐5p and LV‐sh‐GFP were obtained from Genechem.

2.7. Immunofluorescence

The expression of GAP43 and NF200 in PC12 cells was examined by immunofluorescence assay. PC12 cells were infected with LV‐miR‐181d‐5p and LV‐sh‐GFP using polybrene (5 μg/mL) respectively and were seeded into 6‐well plates. After 72 hours of cell culturing, we removed the medium, washed with phosphate buffered saline (PBS) three times, and fixed the cells with ice‐cold 3% paraformaldehyde for 30 minutes at room temperature. After three times washes with PBS, the cells were permeabilized with 0.2% Triton X‐100 (Sigma, St. Louis, MO, USA) in PBS for 15 minutes, blocked in 5% skim milk for 1 hours, and washed three times with PBS containing 0.05% Tween 20 (PBST).We incubated cells with 1:200 dilution of anti‐GAP43 (1:200, no: BA0878; BOSTER, Wuhan, China) and anti‐NF200 (NF200 KD 1:200, no: ab77745; Abcam, Cambridge, UK) antibodies for 1 hours, respectively. After incubation with the primary antibody, we washed the cells three times with PBST and incubated the cells with a corresponding secondary antibody for 1 hours. The secondary antibodies were Cy3‐labeled goat anti‐rabbit‐IgG antibody and Cy3‐labeled goat anti‐mouse‐IgG antibody (1:200; Boster). After further washing three times with PBST, then cells nuclei were stained with DAPI (4,6,‐diamidino‐2‐phenylindole) for 5 minutes at room temperature. After three times of washing cells with PBST, we observed cells and acquired images with an inverted fluorescence microscope (Leica, Wetzlar, Germany).

2.8. Western blotting

Western blot was performed to detect PTEN, p‐Akt, Synapsin, PSD95, NF‐200, and GAP‐43 protein expression in PC12 cell lines. The culture medium was discarded, and the cultured cells were washed three times with PBS before treatment with radio‐immunoprecipitation assay (RIPA) lysis buffer at 4°C for 10 minutes. The mixture was transferred to an EP tube and centrifuged 12 000 g at 4°C for 15 minutes. BCA method was used to determine the protein concentration in the supernatant. 40 μg cell lysates were denatured and separated by sodium dodecyl sulfate‐polyacrylamide gel electrophoresis (SDS–PAGE) and transferred onto the polyvinylidene fluoride (PVDF) membrane. After blocking with 5% skim milk for 1 hours at room temperature, we incubated PVDF with 1:1000 dilution of anti‐PTEN, anti‐Akt, anti‐p‐Akt, anti‐GAP‐43, anti‐NF‐200(Cell Signal Technology, Beverly, MA, USA), anti‐Synapsin, and anti‐PSD95 (Abcam) primary antibodies, followed with incubation of horseradish peroxidase (HRP)‐conjugated antibodies (Thermo Pierce, Waltham, MA, USA). The thickness of GAPDH (Thermo Pierce) was used to determine the amounts of protein from the infected cells.

2.9. Luciferase reporter assay

The appropriate PTEN gene 3′UTR or its mutant (TGAATGT was mutated to GTCCGTG) was cloned into luciferase vector (GV272) and transfected with miR‐181d‐5p mimic or a NT control into HEK‐293T cells in a 24‐well plate using x‐tremeGENE HP (Invitrogen). Forty‐eight hours after transfection, we obtained the cell extract which was used to detect the firefly and Renilla luciferase activities via the dual‐luciferase assay system (Promega, Madison, WI, USA) according to the manufacturer's protocols. To ensure a stable luciferase signal before measuring, all the reagents and cells were placed at room temperature.

2.10. Neurite outgrowth analysis

Quantification of PC12‐miR‐181d‐5p cells and PC12‐sh‐GFP cells with neurite and morphological analysis was performed through fluorescent microscope. We labeled PC12 cells by adding CM‐Dil (Invitrogen) to PC12 cell culture medium. At least ten fields selected randomly with more than 100 cells were picked out for analysis of neurite length of PC12 cells. PC12 cells with at least one neurite equal to or greater than the length of their cell body were defined as positive for neurite outgrowth. The positive PC12 cells were counted and expressed as a percentage of total cells in the field. The neurite length was identified and defined as the maximum neurite length per cell per field. ImageJ software was applied to measure the neurite length of PC12 cells, and data were recorded correspondingly. Each experiment was repeated three times and analyzed, respectively.

2.11. Statistical analysis

Data were acquired from RT‐PCR, Western blot, and dual‐luciferase assay system and presented as mean ± SD. A Student's t test and one‐way analysis of variance (ANOVA) were applied to determine the significance of differences between the treated group and the control group. P < 0.05 was considered statistically significant.

3. RESULTS

3.1. Microarray data

The microarray data (Table 2) showed that the miRNAs and mRNAs expression dynamically changed from the initial injury to 1 day, 3 days, 1 week, and 3 weeks postinjury. The number of upregulation miRNAs did not change much after injury while that of downregulation miRNAs increased from the initial injury to 1 week and declined at 3 weeks after SCI. The number of differentially expressed mRNAs increased at 1 day, 3 days, and 1 week and appeared to decline at 3 weeks after injury. The miRNAs and miRNAs with a fold change ≥2 and an FDR ≤ 0.05 were collected. The differentially expressed miRNAs and mRNAs at five time points were integrated, and a union set was created. This union set contained 63 significantly altered expressed miRNAs and 89 important mRNAs. Based on this set, the heat maps, and hierarchical clustering visually demonstrated the differential expression of the miRNAs and mRNAs (Figure 1A,B). The microarray data have been approved and assigned in the National Center for Biotechnology Information Gene Expression Omnibus(GEO) [GSE67515: GEO].

Table 2.

The number of this table showed the quantity of differential expression of miRNAs and mRNAs by using normal spinal cord as the control

Groups	mRNA		miRNA
	Upregulation	Downregulation	Upregulation	Downregulation
1d vs 0d	309	149	62	10
3d vs 0d	541	532	39	28
1w vs 0d	1045	1132	69	119
3w vs 0d	746	104	57	66

Figure 1.

A remarkable differential expression of miRNAs (A) and mRNA (B) at five continuous time points after spinal cord injury (SCI) was showed in Heat maps and hierarchical clustering. Red and green indicated the high and low expression levels. Black represented the mean expression levels. Each row represented a single group, and each column represented a miRNA or mRNA

3.2. STC for analysis of miRNA expression trend

To measure the expression trend of miRNAs and identify the cluster of diverse genes which have similar expression patterns, series tests of cluster were performed. This method explicitly took into account the dynamic nature of the temporal gene expression profiles during clustering and identified the number of distinct clusters.

As was shown in Figure 2, the differential expression profile was composed of seven differentially expressed miRNA clusters. Each cluster stood for a model trend profile and contained genes with similar expression trends.

Figure 2.

STC analysis was used to analyze miRNAs differential expression. A, We analyzed the expression patterns of these miRNAs and chose seven model profiles for epitomization. Each coordinate axis was representative of an expression pattern which corresponded to the trend of different miRNAs transient expression. The red line in coordinate axis has statistical significance (adj‐P‐value <0.05) while the blue line does not. The upper number in the coordinate axis is the number of expression profiles and value. The horizontal axis represents the time point, and the vertical axis stands for the miRNAs expression level. The number of vertical axes does not represent the actual value of miRNAs expression but the marker of gene expression level (B and C). Two typical significant profiles were picked out to exhibit the specific pattern of gene expression. The curve showed individual gene expression profiles. The upper number in the coordinate axis is the number of profile; genes assigned represents the number of actual differential expression genes in this trend; genes expected stands for the theoretical number of miRNAs in this trend according to the random distribution; adj‐P value represents the significant level of the ratio of the actual miRNAs number to the theoretical number of genes in this trend. We defined the significance at adj‐P‐value <0.05

3.3. Prediction of miRNA target genes

To accurately identify target genes for miRNAs, miRanda and TargetScan were used to predict the target genes of miRNA. The number of target genes that MiRanda and TargetScan predicted was 22 512 and 6780 respectively, in which 2973 were same. From the results, we also got 212 mRNAs negatively correlated with the expression of miRNA.

3.4. Target gene annotation and pathway analysis

To find out the significant function of target genes and the related signal pathways, function analysis, and pathway analysis were performed via GO‐analysis and KEGG, respectively. We obtained 332 significant functions and 41 related signal pathways of target genes. The results indicated that the main functions of the target genes of miRNAs included cell adhesion, induction of apoptosis, angiogenesis, regulation of growth, and embryonic cranial skeleton morphogenesis (Figure 3A). The main signaling pathways contained focal adhesion, PI3K‐Akt signaling pathway, pathways in cancer and endocytosis, and many more (Figure 3B).

Figure 3.

Function analysis and pathway analysis related to target genes after spinal cord injury (SCI). A, GO‐analysis was used to analyze the main functions of target genes having obvious trend according to the Gene Ontology which is the key functional classification of NCBI. Part of the GO‐analysis result includes cell adhesion, induction of apoptosis, angiogenesis, and so on. B, Pathway analysis was applied to find out the significant pathway of the differential genes according to Kyoto Encyclopedia of Genes and Genomes (KEGG) which is an important tool used to analyze relationship among genes, function of differential genes. Part of the results of pathway analysis includes focal adhesion, Pi3k‐Akt signaling pathway, and pathway in cancer

3.5. Building of miRNA‐mRNA‐network

To determine the mRNA‐regulated core miRNA, a miRNA‐mRNA network was built. The miRNA‐mRNA network including the degree and model number of miRNAs was shown in Figure 4. The degree of mmu‐miR‐182‐5P and mmu‐miR‐181d‐5P was the highest two, and mmu‐miR‐128‐3P and mmu‐miR‐338‐3P ranked third and fourth respectively, which indicated that these four miRNAs were key factors in this miRNA‐mRNA network.

Figure 4.

Dynamic miRNA‐mRNA network was constructed. Dynamic miRNA‐mRNA network associated with pathology of spinal cord injury (SCI) was constructed which contained 63 miRNAs and 89 mRNAs. In this network, the size of the circle was on behalf of the ability of the gene to interact, according to the quantification of the degree. Hence, we identified various key miRNAs with high degrees and K‐cores, such as miR‐138‐3p, miR‐181d‐5p, miR‐182‐3p, and miR‐383‐3p which might play crucial roles in SCI pathology

3.6. RT‐PCR validation microarray data

To perform bioinformatics analysis in the subsequent steps, the reliability of the microarray data should be validated with RT‐PCR firstly. Thus four the most important miRNAs including miR‐138‐3p, miR‐181d‐5p, miR‐182‐3p, and miR‐383‐3p were picked out for validation of the microarray results. As was shown in Figure 5, the variation from the RT‐PCR was similar to that of the microarray. The results of RT‐PCR indicated that our microarray data were reliable and could be used for further research.

Figure 5.

The reliability of microarray results was validated via RT‐PCR. A, miR‐138‐5p , B, miR‐181d‐5p, C, miR‐182‐5p, and D, miR‐383‐3p were picked out for validation. Red curve represents the result of RT‐PCR, while the green curve represents the data of the miRNAs microarray. Y‐axis indicates the relative expression level of miRNAs. X‐axis represents different time points postinjury. RT‐PCR exhibited similar result compared with microarray data

3.7. miR181d‐5p for further study

We searched for the most probable binding site of miRNA‐181d‐5p against 3′UTR of PTEN in databases of rodent transcription via TargetScan. Collectively, we obtained three probable binding sites including two poorly conserved, one highly conserved. Not all the prediction results are shown. Only the highly conserved one was shown in Figure 6A and selected for further study. Firstly, we predicted downstream genes for miR‐181d‐5p through TargetScan which was also applied to search the most probable genes against the sequences of miR‐181d‐5p in databases of rodent transcription.

Figure 6.

The forecast of miRNA‐181d‐5p for PTEN was shown in Figure 6. A, Three potential‐binding sites of miRNA‐181d‐5p against PTEN via TargetScan were obtained, but only the highly conserved was shown in Figure 6A, which was located at nt2299‐2305 of PTEN 3′UTR. The concrete miRNA's parameters for PTEN are as below: site type represents an exact match to positions 2‐8 of the mature miRNA (the seed + position 8).The context++ score (CS) for a specific site is the sum of the contribution of 14 features. The context++ score percentile rank is the percentage of sites for this miRNA with a less favorable context++ score. The weighted context++ score (WCS) accounts for the abundance of each 3′‐UTR tandem isoform in which the site exists, as estimated from the UTR profile (a compendium of 3P‐seq datasets from the same species). The WCS for a specific site is the sum of the contribution of 14 features. The conserved branch length score is the sum of phylogenetic branch lengths between species that contain a site. To help control for individual UTR conservation, for each stop site of a gene, the median BLS of the most prevalent 3′UTR and its orthologs were calculated. PCT, the probability of conserved targeting has been calculated for all highly conserved miRNA families. B, 3′UTR of the PTEN gene contains one separate miR‐181d‐5p‐binding seed sequences. C, Dual‐luciferase reporter assay was applied to detect the effect of miR‐181d‐5p and its NC on the luciferase intensity controlled by 3′UTR of PTEN. **P < 0.01.*P < 0.05 Mean ± SD (n = 4)

3.8. Luciferase assay validated the relationship between miR‐181d‐5p and PTEN

Through sequence analysis, it could be found that the potential miR‐181d‐5p targeting regions are located at nt2299‐2305 (Figure 6A) and nt4914‐4920 of the PTEN 3′‐UTR, but only nt2299‐2305 of nucleotide sequences is highly conserved across different species which was chosen for subsequent study.

To demonstrate the direct interaction between miR‐181d‐5p and PTEN, 3′UTR sequences of PTEN containing the potential target site (wild type, WT) or mutated sequences (mutant type, MUT) (Figure 6B) were cloned into the GV272 control vector, respectively. We transfected this vector into 293T cells with miR‐181d‐5p mimic or negative control. In the PTEN 3′‐UTR NC group, the luciferase activity did not change much (Figure 6C). In WT group, miR‐181d‐5p decreased luciferase activities markedly compared with negative controls (Figure 6C). But miR‐181d‐5p mimic did not affect the luciferase activity in the MUT group (Figure 6C). These results were consistent with the bioinformatic prediction that miR‐181d‐5p could directly target the 3′UTR of PTEN.

3.9. miR‐181d‐5P‐activated PI3K/Akt pathway to promote neurite outgrowth in PC12 cells

Previous study has demonstrated that PTEN is the main inhibitor of PI3K/Akt pathway which exerts effect on the neurite outgrowth.5, 15, 16 PTEN inhibitor facilitates neurite outgrowth.17, 18, 19 In this study, we assessed the role of LV‐miR‐181d‐5p on neurite outgrowth in PC12 cells.

To determine the biological effects of miR‐181d‐5P on neurite outgrowth, we performed functional analyses with LV‐miR‐181d‐5p and LV‐sh‐GFP. Firstly, we infected PC12 cells with LV‐miR‐181d‐5p and LV‐sh‐GFP. Two stable cell lines were generated: PC12‐miR‐181d‐5p cells and PC12‐sh‐GFP cells. We then assessed changes in PTEN mRNA and protein levels via RT‐PCR and Western blotting.

Compared with PC12‐sh‐GFP group, LV‐miR‐181d‐5p increased miR‐181d‐5p levels by 6.85‐fold (Figure 7A) and reduced PTEN mRNA and protein expression levels (Figure 7B) by 48% and 56% (Figure 7C,D), respectively. We also observed that miR‐181d‐5p upregulate p‐AKT expression level by 2.74‐fold over controls following PTEN downregulation (Figure 7C,E,F).

Figure 7.

As shown in Figure 7, miR‐181d‐5p has effect on the PC12 cells. A, After being infected with the lentivirus LV‐sh‐GFP and LV‐miR‐181d‐5p, RT‐PCR confirmed the expression degrees of miR‐181d‐5p in the PC12 cells. Mean ± SD (n = 4). The t test of student was utilized ***P < 0.001. B, At mRNA level, miR‐181d‐5P was inversely correlated with PTEN expression. PC12‐miR‐181d‐5p cells inhibited PTEN expression compared with PC12‐sh‐GFP cells. Mean ± SD (n = 4). Student's t test was utilized here ***P < 0.001. C, Similar results were proved by Western blot analysis. At the protein level, PC12‐miR‐181d‐5p cells enhanced the protein levels of p‐Akt and reduced PTEN without changing Akt expression. Glyceraldehyde‐3‐phosphate dehydrogenase GAPDH was utilized as a loading control. Mean ± SD (n = 4). Herein, the t test of student was utilized ***P < 0.001. D‐F, Quantification of densitometric degrees of Akt, p‐Akt, and PTEN. Mean ± SD (n = 4). Herein, student's t test was utilized ***P < 0.001

Both GAP‐43 and NF‐200 are neurite outgrowth‐related proteins which have been proved in previous study.20, 21 Then, we next detected GAP‐43 and NF‐200 expression levels in PC12‐miR‐181d‐5p cells and PC12‐sh‐GFP cells.Cell immunofluorescence results showed that GAP43 and NF200 were expressed in both PC12‐sh‐GFP cells and PC12‐miR‐181d‐5p cells. As was showed in (Figure 8A,B), GAP43 and NF200 mRNA significantly increased in PC12‐miR‐181d‐5p cells compared with PC12‐sh‐GFP cells. Western blotting analysis was performed to exhibit similar results. PC12‐miR‐181d‐5p cells expressed GAP‐43 and NF‐200 proteins at least 2.35‐fold and 2.23‐fold as much as PC12‐sh‐GFP cells (Figure 8C,D,E).

Figure 8.

NF200 and GAP43 expression in PC‐miR‐181d‐5p cells and PC‐sh‐GFP cells. A, B, After 3 d of culturing, PC12‐miR‐181d‐5p cells expressed NF‐200, GAP‐43 higher than PC12‐ sh‐GFP as presented in RT‐PCR. Mean ± SD (n = 4). Herein, the t test of student was adopted ***P < 0.001, **P < 0.01. C, The analysis of Western blot represents the bands of protein on the gel for PC12‐ sh‐GFP and PC12‐miR‐181d‐5p cells, severally. In the expression of neural protein, they had considerable differences. D, E, The amount of densitometry degrees of NF‐200 and GAP‐43 is shown. Mean ± SD (n = 4). Herein, the t test of student was utilized ***P < 0.001

Both Synapsin and postsynaptic density protein (PSD95) are synaptic vesicle‐related proteins clustering at most synapses.22, 23 Western blot analysis was performed to confirm that PC12‐miR‐181d‐5p cells expressed Synapsin and PSD95 protein at 2.68‐fold and 2.87‐fold as much as PC12‐sh‐GFP cells, respectively (Figure 9A,B,C).The neurite of PC12 cells transfected with LV‐miR‐181d‐5P (Figure 9B,D) was significantly longer than those in the PC12‐sh‐GFP group (Figure 9A,C). The proportion of positive neurite‐bearing cells in PC12‐miR‐181d‐5p cells (73.5 ± 5.5%) (P < 0.01) is significantly greater than that in PC12‐sh‐GFP cells (17.3 ± 3.5%) (Figure 9E). Likewise, compared with PC12‐sh‐GFP cells (16.1 ± 2.3 μm), the length of the longest neurite significantly increased to 203.2 ± 9.6μm (P < 0.001) in PC12‐miR‐181d‐5p cells (Figure 9F). From above all, we came to a conclusion that miR‐181d‐5P downregulated PTEN expression and activated PI3k/Akt pathway to promote neurite outgrowth of PC12 cells.

Figure 9.

PC12‐miR‐181d‐5p cells upregulate synaptic vesicle‐related protein and promote neurite outgrowth. A, Compared with PC12‐sh‐GFP cells, PC12‐miR‐181d‐5p cells expressed more Synapsin and PSD95 protein. B, C, Quantification of densitometric degrees of Synapsin and PSD95. Mean ± SD (n = 4). Herein, the t test of student was utilized. **P < 0.01, ***P < 0.001. D, After 3 d of culturing, the images of anti‐NF‐200 and anti‐GAP43 stand for the cells of PC12‐ sh‐GFP (A, C) and PC12‐miR‐181d‐5p (B, D). Scale bar = 100 μm. Mean ± SD (n = 4). E‐F, The number of positive PC12 cells and the length of neurites are shown in the bar graphs E and F, severally. Data are presented as mean ± SEM from three independent experiments. Herein, the t‐test of student was utilized. **P < 0.01

4. DISCUSSION

Microarray assay plays a crucial role in performing analysis of miRNA.24 We produced a reliable and steady animal model and extracted RNA from injured spinal cords for the use of microarray assay. Microarray assay results were also validated with several significantly expressed miRNAs selected from miRNA‐miRNA network by RT‐PCR. Through the analysis of microarray assay results, we observed that the miRNAs and mRNAs expression dynamically changed postinjury and totally collected 262 significantly altered expressed miRNAs and 2973 target genes in SCI. The key to study the biological function and mechanism of miRNAs is to accurately identify the target genes of miRNAs and performed analysis of their target genes.25 Via function analysis and pathway analysis of target genes, we found 332 significant functions of target genes (mainly related to cell adhesion, induction of apoptosis, and angiogenesis) and obtained 41 related signaling pathways, including focal adhesion, PI3K/Akt signaling pathway, and cancer pathway A miRNA‐mRNA network containing 63 significantly expressed miRNAs, and 89 important target genes was created which revealed that miRNA regulation of target genes is so complex: a miRNA can regulate two or more genes (such as mmu‐miR‐138‐3p can simultaneously regulate Igfbp5 and Phldb2), and vice versa, some genes can be simultaneously regulated by two or more miRNAs (Col5a1 can be regulated by miR‐181d‐5p or miR‐182‐5p,Cald1 can be regulated by miR‐182‐5p or miR‐383‐3p). As a result, microarray assay is a very important bioinformatics research tool and can offer a great deal of useful data for analysis of miRNAs.

Accumulating evidences support that miRNAs are important mediators of axonal regeneration and neurodevelopment.2, 26 MicroRNA‐34a regulates neurite outgrowth, spinal morphology, and function.27 miR‐133b regulate neurite outgrowth via ERK1/2 and PI3K/Akt signaling pathway by RhoA suppression.28

Both miR‐181d‐5p and PI3K/Akt signaling pathway selected as our subsequent research target were from microarray assay result. The whole miRNA181 family members share the same 5′seed and consist of six mature miRNA‐miRNA181a1, miRNA181a2, miRNA181b1, miRNA181b2, miRNA181c, and miRNA181d, which are encoded in three independent paralog precursor transcripts on three separate chromosomes.29 The bioinformatics analysis indicates that PTEN may be the potential target for miR‐181d‐5p. Thus, luciferase activity assay was applied to identify the effect of miR‐181d‐5p on PTEN expression. The luciferase activity data showed that miR‐181d‐5p was capable of directly targeting the 3′UTR of PTEN. RT‐PCR and Western blotting results also showed that the overexpression of miR‐181d‐5p downregulated PTEN expression in PC12 cells and inversely correlated with miR‐181d5p expression.

PTEN is the principal and nonredundant negative regulator of the PI3K pathway and it might act as an intrinsic brake on the regenerative outgrowth.18, 30 miR‐29a modulates neurite outgrowth in PC12 cells via targeting PTEN.31MicroRNA‐182 which is another significantly expressed miRNA in microarray assay result, and neuron‐enriched microRNA has been demonstrated that plays a significant role in regulating neuronal axon outgrowth and dendrite tree formation through PTEN/AKT pathway.32

In immune system, miR‐181 regulates cellular metabolic fitness required for robust proliferation, determines organism size, and is considered to be a PTEN rheostat and a central regulator of the PI3K pathway in vivo.9 The importance of miR‐181 in immune system is clear. However, its underlying signaling mechanisms in the nervous system are not clearly clarified. In this study, we observed that the neurite in PC12‐miR‐181d‐5p cells was significantly longer than that of PC12‐sh‐GFP cells. Through RT‐PCR and Western blotting, we demonstrated that overexpression of miR181d‐5p could decrease the expression of PTEN, upregulate p‐Akt, and then promote neurite outgrowth‐related proteins (GP‐43 and NF‐200) and synaptic vesicle‐related proteins (Synapsin and PSD95) expression in PC12‐miR‐181d‐5p. These findings were the first time to provide the evidence that miRNA‐181d‐5p could promote neurite outgrowth via PI3K/Akt pathway through downregulating PTEN expression in PC12 cells.

Although it is too early to speculate the clinical use of miRNA, it does provide a promising research hint. In conclusion, miRNA‐181d‐5p is a crucial regulator of neurite outgrowth through PI3K/Akt pathway and might be a suitable therapeutic target for neural disease.

CONFLICT OF INTEREST

The authors declare no conflict of interest.

Shen LM, Song ZW, Hua Y, Chao X, Liu JB. miR‐181d‐5p promotes neurite outgrowth in PC12 Cells via PI3K/Akt pathway. CNS Neurosci Ther. 2017;23:894–906. 10.1111/cns.12761