LncRNA MIAT Promotes Spinal Cord Injury Recovery in Rats by Regulating RBFOX2-Mediated Alternative Splicing of MCL-1

LncRNA myocardial infarction–associated transcript (MIAT) alleviates acute spinal cord injury (ASCI)-induced neuronal cell apoptosis, but the specific mechanism of it involved in regulating SCI progression needs further exploration. Here, a SCI rat model was established, followed by administration with adenovirus-mediated MIAT overexpression vector (Ad-MIAT) alone or together with Ad-RBFOX2 (RNA binding fox-1 homolog 2). The data indicated that MIAT overexpression promoted motor function recovery, improved morphology of injured tissues, and restrained neuron loss and cell apoptosis in SCI rats. Then, PC-12 cells were treated with H2O2 to induce cell injury. And highly expressed MIAT suppressed H2O2-caused decrease in cell viability and increase in cell apoptosis. MIAT stabilized RBFOX2 protein expression by binding to RBFOX2, thereby promoting RBFOX2-induced upregulation of anti-apoptotic MCL-1L (myeloid cell leukemia sequence 1) and reduction of pro-apoptotic MCL-1S. And silencing RBFOX2 in vitro blocked the inhibitory effect of MIAT on cell apoptosis. Moreover, MCL-1-specific steric-blocking oligonucleotides (SBOs) were used to transfer the MCL-1 pre-mRNA splicing pattern from MCL-1L to MCL-1S. SBOs reversed the protection effect of RBFOX2 overexpression on H2O2-induced cell injury. Furthermore, overexpression of MCL-1L instead of MCL-1S facilitated autophagy activation in H2O2-stimulated cells. Interestingly, co-overexpression of MIAT and RBFOX2 had a better promoting effect on SCI recovery. In conclusion, MIAT mitigated SCI by promoting RBFOX2-mediated alternative splicing of MCL-1. Our findings might provide a promising therapeutic target for SCI.


Introduction
Spinal cord injury (SCI) mainly consists of primary injury caused by mechanical compression and secondary injury caused by a series of pathophysiological cascades shortly after injury [1]. The latter leads to pathological changes in injured tissues, including edema, apoptosis and necrosis, inflammation, and local autophagy, which seriously affect neurogenesis in spinal cord segments [2,3]. Due to irreversible changes in the structure and function of neurons, SCI can cause long-term physical damage and disability, and permanently affect the life quality of patients [4]. One of the most destructive processes of secondary SCI is neuronal apoptosis, as it is one of the main obstacles to the recovery of motor function [5]. Therefore, the prevention of cell apoptosis after SCI may promote spinal cord tissue repair and motor function improvement.
LncRNAs (longer than 200 nucleotides) regulate gene expression at the epigenetic, transcriptional, and translational levels, and participate in many biological processes such as chromatin remodeling, cell differentiation, proliferation, and apoptosis [6]. LncRNA myocardial infarction-associated transcript (MIAT) was initially reported to increase the risk of myocardial infarction and to be highly expressed in the nervous system and retinal tissues [7]. With the deepening of research, the role of MIAT in various diseases has been reported, including human cancers [7,8], neurological diseases [9], and cardiovascular diseases [10]. A previous study showed that MIAT expression is reduced after SCI, which may be due to neuronal death [11]. In addition, a latest research showed that highly expressed MIAT inhibits nerve cell apoptosis and promotes the recovery of motor function in acute spinal cord injury (ASCI) rats [12]. These evidences suggest the involvement of MIAT in the progression of SCI, but its mechanism still needs to be further explored.
Alternative splicing (AS) is a mechanism by which the pre-mRNA of a single gene generates different forms of mature mRNA through the integration or exclusion of specific exons [13]. RNA-binding proteins (RBPs) regulate AS by binding pre-mRNAs of genes and interacting with splicing machinery [14]. The RNA binding fox-1 homolog 2 (RBFOX2) is widely expressed in many tissues throughout life, unlike its hominins, RBFOX1 and RBFOX3, which are mainly expressed in muscle and neuronal tissues [15]. Rbfox2 is an important regulator of tissue-specific and signal-responsive alternative splicing, and can function by binding to the (U)GCAUG motif [16]. Rbfox2 plays important roles in myogenesis 3, neurodevelopment 2, embryonic stem cell survival, and tumor progression [17][18][19][20], but its role in neuron apoptosis remains to be explored. Moreover, a recent study indicated that RBFOX2 regulates the alternative splicing of KIF1B in ovarian cancer cells, and interference with RBFOX2 results in pro-apoptotic KIF1B-β preferentially splicing and promotes anoikis [21]. Given the important role of RBFOX2 in neural development, we hypothesized that RBFOX2 might participate in the SCI process.
Here, we demonstrated that highly expressed lncRNA MIAT contributed to functional recovery in SCI rats and reduced H 2 O 2 -induced nerve cell apoptosis in vitro by stabilizing RBFOX2 protein expression. Given that RBFOX2 is a regulator of alternative splicing, we further investigated whether MIAT played its role through RBFOX2-mediated gene alternative splicing.
According to a previous study [22], a pair of MCL-1 SBOs was designed to target the splicing sites at the 3′ and 5′ ends of exon 2 of MCL-1L pre-mRNA, so that exon 2 could be spliced off, thus changing the splicing mode of MCL-1L pre-mRNA from MCL-1L to MCL-1S mRNA. All steric-blocking oligonucleotides (SBOs) were synthesized, and the Endo-Porter delivery system was purchased from Gene Tools (Philomath, OR, USA). The sequences of the SBOs at the 3′ acceptor and 5′ donor splice sites of Mcl-1 pre-mRNA exon 2 were 5′-CGA AGC ATG CCT GAG AAA GAA AAG C-3′ and 5′-AAG GCA AAC TTA CCC AGC CTC TTT G-3′, respectively. The non-targeting oligonucleotide sequence was 5′-CCT CTT ACC TCA GTT ACA ATT TAT A-3′ (Con-SBO). The delivery efficiency of the Endo-Porter system was optimized by using a fluorescence microscope to observe the number of fluorescent positive cells in a limited field of view.

Animals and Treatments
Adult male Sprague-Dawley (SD) rats (weighing 200-220 g) were obtained from Center for Animal Experiment of Henan province (Zhengzhou, Henan, China). Rats were housed in standard conditions with controlled temperature (24 ± 2 °C) with 12:12 light/dark cycle and freely fed and watered. Animal experiments performed in our study were approved by the Animal Ethics Committee of Xi'an Jiaotong University. Animals were randomly divided into 5 groups: Sham group, SCI group, SCI plus adenovirus vector group, SCI plus Ad-MIAT group, SCI plus Ad-MIAT, and Ad-RBFOX2 group. The rat SCI model was established by using the improved Allen method. In brief, rats were anesthetized by intraperitoneal injection of 10% chloral hydrate (3 mL/kg, Sino Chemical Reagent Company, Shanghai, China). Rats' backs were shaved to expose the T9-11 spinous process and vertebral segments. Rats underwent laminectomy of vertebral T10, and then, a 10-g rod (2.5 mm in diameter) was dropped from a height of 12.5 mm to the spinal cord to generate a moderate contusive injury in the SCI group. After injury, the spinal cord was washed with normal saline, the incision was sutured, and antibiotics were administered for three consecutive days. Rats in the sham group were only subjected to laminectomy. The empty vector, Ad-MIAT, and Ad-RBFOX2 plasmids (200 µg of plasmids precipitated in 200 µL of PBS) were administered via tail vein injection immediately after SCI. At day 14 after SCI, a 10-mm-long segment of the spinal cord centered at the injury epicenter was harvested for further examination.

Assessment of Locomotor Capacity
The assessment of athletic ability was based on the Basso-Beatie Bresnahan (BBB) exercise scale and the bevel test. All behavioral assessments were scored by three individuals who were blinded to grouping, and performed on days 1, 7, 14, 21, and 28 after SCI. The BBB test scores range from 0 to 21. The total score for severe neurological impairment was 0, while a total score of 21 indicated normal performance. The bevel test was performed on the test equipment. The maximum angle at which the rat remained in its position for more than 5 s without falling was recorded.

Preparation of Spinal Cord Slices
Rats were deeply anesthetized with a lethal dose of sodium pentobarbital (80 mg/kg) and transcardially perfused with cold salt water, followed by 4% paraformaldehyde. Spinal tissues at the lesion site were dissected immediately after perfusion, immobilized overnight, and cryopreserved in graded concentrations of sucrose (12%, 18%, and 24%). A 10-mm segment of the spinal cord including the injury epicenter was sectioned in transverse horizontal plane by using a cryostat (CM3050S, Leica, Wetzlar, Germany) at 10 or 20 μm thickness.

Hematoxylin-Eosin Staining and Nissl Straining
For HE straining, sections were stained in hematoxylin solution for 5 min, washed in tap water, and soaked in 1% acid alcohol for 30 s. After washing again, the sections were stained in eosin for 30 s and dehydrated in graded concentrations of ethanol. For Nissl straining, sections were dewaxed with xylene and rehydrated in graded concentrations of ethanol. Next, sections were stained in Nissl staining solution (Solarbio, Beijing, China). Cells with typical neuronal morphology and nuclei were counted.

Fluorescent Protein-LC3B Dot Assay
Cells were transfected with mRFP-eGFP-LC3B plasmid for 12 h. Then, cells were treated in groups as shown in Fig. 7. After 48 h, cells were fixed in 4% paraformaldehyde and washed in PBS. DAPI was used to stain nuclei. A confocal laser scanning microscope (Leica TCS SP8, Leica Biosystems, Wetzlar, Germany) was used to capture fluorescence images (400 ×), and the data were analyzed with Image J software.

TUNEL Staining
Sections of the spinal cord were fixed, sealed, and incubated with the terminal deoxynucleotidyl transferase-mediated deoxyuridine triphosphate nick end labeling (TUNEL) reaction mixture (Roche, Basel, Switzerland) at 37 °C for 1 h according to the manufacturer's instructions. DAPI was used for counter staining. A fluorescence microscope was used to capture the images of the apoptotic cells. Total number of TUNEL-positive cells was counted from each slide, and the averages were recorded to examine the numbers of apoptotic cells.

Apoptosis Assay
Cell apoptosis was measured with Annexin V-fluorescein isothiocyanate (FITC) Apoptosis Detection Kit (BD Biosciences, San Diego, CA, USA). In brief, the treated cells were collected and resuspended in 100 μL of binding buffer. Next, the cells were cultured with 5 μL of Annexin V-FITC and 5 μL propidium iodide (PI) for 15 min in the dark. Flow cytometry was performed to measure the cell apoptosis, and the obtained data were analyzed using FlowJo software.

Cell Viability Assay
To evaluate the cell viability, the MTT assay was performed according to the manufacturer's instructions. Briefly, cells were seeded into 96-well plates with a density of 1 × 10 4 cells/well. After incubation for 48 h, 20 μL of MTT (Sigma-Aldrich, Shanghai, China) reagent was pipetted into each well, and cells were incubated for another 3 h. Next, 200 μL of dimethyl sulfoxide solution (Sigma-Aldrich) was added, and the absorbance at 490 nm was measured with a microplate reader (Bio-Rad Model 550; Hercules, CA).

RNA Pull-Down Assay
MIAT cDNA was inserted into pSPT18/19 vector, and then transcribed in vitro with biotin-labeled nucleotides by using T7 RNA polymerase (Roche, Indianapolis, IN, USA), followed by purification with RNeasy Mini Kit (Qiagen, Valencia, CA). RNA pull-down assay was performed by using a Pierce Magnetic RNA-Protein Pull-Down Kit (Thermo Fisher Scientific, Waltham, MA, USA) according to the manufacturer's instructions. The lysate of PC-12 cells were incubated with 50 pmol of biotinylated MIAT (MIAT probe) or NC probe for 1 h at room temperature, and then incubated with 50 µL of streptavidin agarose magnetic beads (Life Technologies) at 4 °C for 1 h. The RNA-protein immunocomplex was eluted with Biotin Elution Buffer and then boiled in sodium dodecyl sulfate (SDS) loading buffer for 10 min. Western blotting was used to detect the protein expression of RBFOX2.

RNA Immunoprecipitation Assay
The Magna RNA immunoprecipitation kit (17-700, Millipore, Billerica, MA, USA) was used for RIP assay. Cells were harvested by centrifugation at 4 °C (1500 rpm) for 5 min. The cell pellet was lysed by using RIP lysis buffer containing 1% protease inhibitor and 200 U/mL RNase inhibitor (Life Technologies) on ice for 15 min, followed by centrifugation at 4 °C (14,000 rpm) for 10 min. The cell supernatant was collected and a part of supernatant was used as an input. The magnetic beads was rinsed with RIP lysis buffer, and then incubated with anti-RBFOX2 antibody or IgG (AP112, Sigma, St. Louis, MO, USA) for 2 h at 4 °C. Next, the remaining cell supernatant was incubated with antibody-coated magnetic beads at 4 °C for 6 h. TRIzol reagent (Invitrogen, USA) was used to extract the RNA bound to anti-RBFOX2 antibody, and qRT-PCR was performed to measure the MIAT level in the immunoprecipitation complex.

RT-qPCR
Total RNA was isolated from PC-12 cells and spinal cord tissues by using Trizol reagent following the manufacturer's protocols. The samples were reversely transcribed by using the Bestar qPCR RT kit (DBI Bioscience, Ludwigshafen, Germany). qPCR amplification was carried out by using DBI Bestar ® SybrGreen qPCR MasterMix (DBI Bioscience) under the following conditions: 94 °C for 2 min, followed by 40 cycles of 94 °C for 20 s, 58 °C for 20 s, and 72 °C for 20 s. GAPDH was used as a endogenous control, and the relative transcript levels were calculated using the 2 −ΔΔCT method.

Statistical Analysis
Results in this study were presented as mean ± standard deviation. The comparison of more than two groups was conducted by one-way ANOVA test. The differences between two groups were investigated by Student's t-test. P < 0.05 was considered to be a significant difference.

Overexpression of MIAT Promoted the Recovery of Experimental SCI Rats
In order to explore the role of MIAT in the progression of SCI, a rat SCI model was established, and an adenovirusmediated MIAT overexpression vector was injected into rats through the tail vein. The data displayed that MIAT expression was significantly downregulated after SCI induction, and was upregulated after Ad-MIAT administration (Fig. 1A). Moreover, compared with the sham group, the BBB scores and maximum dip angle were markedly decreased in the SCI group. And the BBB scores and maximum dip angle were notably increased in the Ad-MIAT administration group compared with the SCI + vector group (Fig. 1B, C). Compared with the sham group, SCI group showed obvious structural damage and abnormal tissue arrangement in histopathology. Overexpression of MIAT markedly improved histological morphology and reduced necrotic tissues (Fig. 1D). Furthermore, the results of Nissl staining showed that neuronal loss in the ventral horn was decreased after overexpression of MIAT (Fig. 1E). In addition, TUNEL staining showed extensive apoptotic cell death in SCI rats compared with the sham group. And the high expression of MIAT resulted in a prominent decrease in the number of TUNEL-positive cells (Fig. 1F).

Overexpression of MIAT Alleviated H 2 O 2 -Induced PC-12 Cell Apoptosis
In order to explore the mechanism by which MIAT regulates the progress of SCI, PC-12 cells were treated by H 2 O 2 to simulate SCI-induced cell damage in vitro. MIAT expression was observably reduced after treatment with Fig. 1 Overexpression of MIAT promoted the recovery of experimental SCI rats. A SCI rat model was established, and Ad-MIAT plasmids (200 µg of plasmids precipitated in 200 µL of PBS) were administered via tail vein injection immediately after SCI. A The expression of MIAT in injured spinal cord segments was detected 2 weeks after surgery. B, C The Basso-Beattie-Bresnahan and inclined plane test were used to assess the the recovery of motor func-tion in rats. D HE staining was executed 2 weeks after surgery to evaluate the extent of spinal cord tissue injury. E Nissl staining was conducted 2 weeks after surgery to quantify neuron loss. F TUNEL staining of neuronal apoptosis was performed 2 weeks after surgery. N = 8 in each group. *P < 0.05 and **P < 0.01 compared with the sham group. # P < 0.05 compared with the SCI + vector group. Each test was independently repeated at least three times H 2 O 2 , but was significantly upregulated after subsequent Ad-MIAT infection ( Fig. 2A). Moreover, H 2 O 2 treatment led to a significant decrease in cell viability of PC-12 cells and a significant increase in cell apoptosis, which was partially abolished by MIAT overexpression (Fig. 2B,  C). Furthermore, the enforced MIAT expression reversed the upregulation of cleaved-caspase-3 and cleaved-caspase-9 induced by H 2 O 2 (Fig. 2D).

MIAT Directly Bound to RBFOX2
We used the StarBase online database to predict the RBP motifs in the MIAT sequence, and found that there are two RBFOX2 binding motifs in the MIAT sequence (Fig. 3A,  B). Next, RIP and RNA pull-down assays were performed to verify the binding of MIAT and RBFOX2. MIAT was significantly enriched in complexes immunoprecipitated with anti-RBFOX2 antibody by using MIAT-specific primers instead of control non-specific primers, while almost no MIAT enrichment was measured in the complexes immunoprecipitated with IgG antibody (Fig. 3C). The RNA pull-down assay using the MIAT probe showed a significant increase of RBFOX2 protein in the MIATprotein complex compared with using control probe The mRNA and protein levels of RBFOX2 were detected after overexpressing or silencing MIAT. H PC-12 cells were treated with 5 μM CHX to evaluate the effect of MIAT on the stability of RBFOX2 protein. N = 5 in each group. *P < 0.05, **P < 0.01, # P < 0.05 compared with Mut-both group. Each test was independently repeated at least three times (Fig. 3D). Subsequent RNA pull-down test using WT or MUT biotinylated MIAT transcripts demonstrated that mutations of either motif inhibited the binding between MIAT and RBFOX2, and mutations of the two motifs showed a stronger inhibitory effect (Fig. 3E). In addition, the regulatory effect of MIAT on RBFOX2 expression was analyzed, and no significant changes in the mRNA level of RBFOX2 were observed after MIAT overexpression or interference (Fig. 3F), while overexpression of MIAT significantly increased RBFOX2 protein level, and knockdown of MIAT showed an opposite effect (Fig. 3G). Thus, we speculated that MIAT might regulate RBFOX2 expression by stabilizing its protein expression. Next, we treated cells with a protein synthesis inhibitor cycloheximide (CHX) and analyzed RBFOX2 protein expression at specified time points. The data indicated that RBFOX2 protein was stably expressed within 30 h in cells transfected with NC siRNA, whereas it was gradually downregulated in cells transfected with MIAT siRNA (Fig. 3H).

MIAT Suppressed H 2 O 2 -Induced Cell Apoptosis by Upregulating RBFOX2
Compared with the control group, the expression of MIAT and RBFOX2 showed a noteworthy reduction in H 2 O 2 -treated cells. And infection with Ad-MIAT upregulated the expression of MIAT and RBFOX2 in the presence  (Fig. 4A, B). While compared with the Ad-MIAT treatment group, transfection with RBFOX2 siRNA only suppressed the protein level of RBFOX2 without altering MIAT expression (Fig. 4A, B). Overexpression of MIAT upregulated cell viability and reduced cell apoptosis and cleaved-caspase-3/9 expression compared with the H 2 O 2 treatment group (Fig. 4C-E). Moreover, the low expression of RBFOX2 effectively restrained MIAT overexpressioninduced increase in cell viability and decrease in cell apoptosis and cleaved-caspase-3/9 expression in H 2 O 2 -treated cells (Fig. 4C-E).

RBFOX2 Regulated the Alternative splicing of MCL-1
The two splicing variants of MCL-1 (MCL-1L and MCL-1S) were shown in Fig. 5A. Since the MCL-1L is pro-apoptotic and the MCL-1S is anti-apoptotic, we investigated whether H 2 O 2 treatment led to a conversion between the two splicing variants. The data displayed that treatment with H 2 O 2 notably inhibited the mRNA and protein levels of MCL-1L, but increased MCL-1S, which was reversed by overexpressing RBFOX2 (Fig. 5B, C). And the enforced expression of MIAT played a similar role as RBFOX2 overexpression. Moreover, silencing RBFOX2 effectively impeded the upregulation of MCL-1L and downregulation of MCL-1S caused by MIAT overexpression (Fig. 5D, E). Since RBFOX2 regulates alternative splicing by binding to (U)GCAUG motif, we investigated whether this motif is present in introns 1 and 2 of MCL-1 pre-mRNA. We found that only one (U)GCAUG sequence exists in MCL-1 intron 1, indicating that RBFOX2 was likely to regulate the alternative splicing of MCL-1 by binding this motif (Fig. 5A).

RBFOX2 Restrained H 2 O 2 -Induced Cell Apoptosis by Modulating Alternative Splicing of MCL-1
To investigate the role of MCL-1 in H 2 O 2 -induced cell apoptosis, cells were treated with fluorescein-tagged MCL-1-specific steric-blocking oligonucleotides (SBOs), which transferred the MCL-1 pre-mRNA splicing pattern from MCL-1L to MCL-1S. We found that the protein level of MCL-1L was increased, and MCL-1S was reduced after assay was carried out to measure cell viability. C Cell apoptosis was analyzed with flow cytometry. N = 5 in each group. *P < 0.05. Each test was independently repeated at least three times overexpressing RBFOX2, which was effectively reversed by SBOs (Fig. 6A). Moreover, SBOs observably hindered the increase in cell viability and the reduction in cell apoptosis caused by the overexpression of RBFOX2 in H 2 O 2 -stimulated cells (Fig. 6B, C).

MCL-1L Overexpression-Induced Autophagy Activation in H 2 O 2 Stimulated Cells
In order to explore whether MCL-1 regulates PC-12 cell apoptosis by affecting autophagy activation, H 2 O 2 pretreated cells were infected with Ad-MCL-1L or Ad-MCL-1S. Treatment with H 2 O 2 upregulated LC3-II level and reduced P62 expression, which was reversed by infection with Ad-MCL-1S (Fig. 7A, B), while MCL-1L overexpression group showed higher LC3-II expression and lower P62 expression compared with H 2 O 2 group (Fig. 7A, B). Moreover, the formation of MRFP-EGFP-LC3B spots was examined by confocal laser scanning microscopy. The data showed that treatment with H 2 O 2 increased autophagy flux, and MCL-1L overexpression further enhanced the autophagy flux, while infection with Ad-MCL-1S suppressed the autophagy flux compared with H 2 O 2 treatment group (Fig. 7C). Consistently, H 2 O 2 stimulation observably facilitated the number of autophagosomes in PC12 cells, which was increased by overexpressing MCL-1L and reversed by overexpressing MCL-1S (Fig. 7D).

Co-overexpression of MIAT and RBFOX2 Promoted Motor Function Recovery, and Attenuated Structural Damage and Neuronal Death After SCI
Compared with the SCI + vector group, the BBB scores and the maximum dip angle were significantly increased. And the BBB scores and the maximum dip angle in Ad-MIAT and Ad-RBFOX2 combined administration group were higher than those in Ad-MIAT administration group (Fig. 8A, B). Moreover, overexpression of MIAT alleviated SCI-induced structural injury, neurons loss, and cell Western blotting was used to assess the protein levels of LC3 and P62. C Representative images of mRFP-eGFP-LC3B spots were captured by a confocal laser micro-scope (magnification × 1000). D Transmission electron microscopy is used to observe the accumulation of autophagosomes (magnification × 8000). N = 5 in each group. *P < 0.05. Each test was independently repeated at least three times apoptosis in spinal cord tissues (Fig. 8C-E). Compared with Ad-MIAT treatment alone group, the co-treatment of Ad-MIAT and Ad-RBFOX2 markedly suppressed cavitation and ameliorated the structural damage in the spinal cord, as well as restrained the neurons loss in the ventral horn (Fig. 8C,  D). And co-overexpression of MIAT and RBFOX2 showed better anti-apoptotic effect compared with overexpression of MIAT alone (Fig. 8E). Furthermore, we measured the expression of MIAT and RBFOX2 in vivo, and found that their expression was increased after overexpressing MIAT, while subsequent RBFOX2 overexpression only upregulated the level of RBFOX2 protein without changing MIAT expression (Fig. 8F, G).

Discussion
The growing evidence indicates that lncRNAs are involved in cell apoptosis induced by SCI injury. LncRNA XIST (X-inactive specific transcript) regulates AKT phosphorylation by competitively binding miR-494 to accelerate neuronal apoptosis in rats after SCI [23]. Silencing lncRNA BDNF-AS (Brain-Derived Neurotrophic Factor Antisense) exerts anti-apoptotic effects in both ASCI rats and hypoxic nerve cells by modulating the miR-130b-5p/PRDM5 axis [24]. It was reported that lncRNA MIAT plays a key role in brain development and its dysregulation may cause neurological diseases [9]. The enforced expression of MIAT can Fig. 8 Co-overexpression of MIAT and RBFOX2 promoted motor function recovery, and attenuated structural damage and neuronal death after SCI. A, B The Basso-Beattie-Bresnahan and inclined plane test were performed to evaluate the recovery of motor function. C Representative images of spinal cord tissue sections with HE staining. D Nissl staining was conducted 2 weeks after surgery to quantify neuron loss. E Neuronal apoptosis was analyzed with TUNEL staining 2 weeks after surgery. F, G MIAT and RBFOX2 expression in injured spinal cord segments was measured 2 weeks after surgery. N = 8 in each group. *P < 0.05 and **P < 0.01 compared with the SCI + vector group. # P < 0.05 compared with the SCI + Ad-MIAT group. Each test was independently repeated at least three times suppress neuron apoptosis in the neonatal rat brain injury model induced by hypoxia-ischemia [25]. Moreover, MIAT was reported to be downregulated in a SCI rat model [11]. A recent study showed that MIAT promotes VEGFA transcription by increasing RAD21 protein expression, and contributes to functional recovery in ASCI rats [12]. The results in this study showed that overexpression of MIAT promoted motor function recovery, improved morphology of injured tissues, and inhibited neuron loss and cell apoptosis in SCI rats. In addition, overexpression of MIAT in vitro hindered H 2 O 2 -induced apoptosis of PC-12 cells, suggesting that MIAT promoted functional recovery in SCI rats probably through suppression of apoptosis.
Post-transcriptional regulation is mainly regulated by RBPs, as RBPs dynamically coordinate maturation, transport, and stability of all types of RNAs [26,27]. Therefore, the unbalanced expression or functional changes of RBPs may cause many diseases [28]. It was reported that lncRNAs are important factors in post-transcriptional regulation by co-regulating mRNA stability and translation with micro-RNAs or RBPs [29,30]. Furthermore, lncRNAs can also affect the stability of RBPs. A previous study showed that lncRNA FAM83H-AS1 promoted radiation resistance and cell metastasis of ovarian cancer by stabilizing HuR protein expression [31]. Our results showed that MIAT bound to RBFOX2 protein and stabilized the expression of RBFOX2 protein without affecting its mRNA level. Rbfox2 has been reported to be involved in many neurological diseases and regulate the alternative splicing of many important neuronal transcripts [32]. The lack of Rbfox2 in the central nervous system can cause disruption of cerebellar development [19]. We found that RBFOX2 was significantly lower expressed in H 2 O 2 -stimulated PC-12 cells. Moreover, silencing RBFOX2 reversed the blocking effect of MIAT on H 2 O 2 -induced apoptosis, and overexpressing RBFOX2 strengthened the promoting effect of MIAT on functional recovery in SCI rats, showing that both RBFOX2 and MIAT played protective roles in the SCI progression. LncRNAs regulate pre-mRNA splicing by interacting with splicing factors or with mRNA itself [33][34][35]. The position of (U)GCAUG motif relative to the alternative exon determines the role of RBFOX protein in splicing. Motifs located downstream of alternative exons usually promote RBFOX-dependent exon inclusion, while motifs located upstream usually suppress exon inclusion [19]. In addition, a previous research suggested that silencing MALAT1 in ovarian cancer cells inhibits the expression of RBFOX2, leading to the preferential splicing of the pro-apoptotic KIF1B subtype (KIF1B-β) and increasing anoikis [21]. We found that RBFOX2 regulated alternative splicing of MCL-1 by switching the splicing pattern of MCL-1 pre-mRNA to anti-apoptotic MCL-1L. Moreover, we also found that only one (U)GCAUG sequence exists in MCL-1 intron 1, indicating that RBFOX2 was likely to regulate the alternative splicing of MCL-1 by binding this motif.
Apoptosis is a genetically conserved programmed cell death that is essential for normal cell homeostasis [36]. Many apoptosis-related genes (such as Bcl-x, Bcl-2L and Mcl-1) can undergo alternative splicing, leading to different subtypes with antagonistic (anti-apoptotic or pro-apoptotic) functions [22]. Mcl-1 is an important member of the Bcl-2 gene family and is generally regarded as an anti-apoptotic factor. The full-length transcript of MCL-1 encodes the antiapoptotic MCL-1L subtype. When the second exon of the pre-mRNA of MCL-1 is clipped, the anti-apoptotic MCL-1S variant is encoded [37]. In addition, high expression of MCL-1L and low expression of MCL-1S become key survival and drug resistance molecules involved in the avoidance of apoptosis in some tumors [38][39][40]. Furthermore, studies have shown that changes in the expression of miR-20a and miR-29b promote neuronal cell death in mice with SCI by downregulating Mcl-1 and upregulating BH3-only protein [41], suggesting that Mcl-1 is involved in the process of nerve cell apoptosis induced by SCI. We used SBOs to treat cells, which transferred the MCL-1 pre-mRNA splicing pattern from MCL-1L to MCL-1S. And SBOs effectively abolished the reversal effect of RBFOX2 on H 2 O 2 -induced cell damage. In addition, MCL-1 has been shown to regulate autophagy. The lack of MCL-1 in the mouse heart leads to impaired autophagy activation, leading to rapid cardiomyopathy and death [42]. Moreover, McL-1 regulates apoptosis and starvation-induced autophagy in a developmentally controlled manner [43]. Our results revealed that the promotion effect of H 2 O 2 on autophagic flux and autophagosome formation was promoted by overexpression of MCL-1L and inhibited by overexpression of MCL-1S.
In conclusion, overexpression of MIAT increased RBFOX2 protein level by binding with RBFOX2 protein, resulting in the transformation of RBFOX2-mediated splicing mode of MCL-1 pre-mRNA to anti-apoptotic MCL-1L, reducing the apoptosis of nerve cells, and promoting the recovery of SCI rats. Moreover, our data indicated that the regulatory effect of MCL-1 on cell apoptosis was probably achieved by modulating autophagy. Our research may provide a promising molecular target for the treatment of SCI.
Data availability All data generated or analyzed during this study are included in this published article. Data will be made available from the corresponding author on reasonable request.