Fosl1 Targets AMPK to Inhibit Autophagy-Mediated Anti-Apoptotic and Anti-Inflammatory Effects in Spinal Cord Injury


 Background: The objective of this study was to delineate the role of Fosl1 in regulating inflammation and apoptosis following spinal cord injury.Methods: GSE45006 datasets from Gene Expression Omnibus (GEO) were explored to analyze Fosl1 gene expression. Next, we established an animal model to assess Fosl1 and AMPK by western blotting, real-time PCR, and immunohistochemical staining and used immunofluorescence staining to check Fosl1 expression in neurons. Fosl1 silencing was used to assess the effect on AMPK, cell viability, autophagy, inflammation and apoptosis. Subsequently, an AMPK activator and inhibitor were added to PC-12 cells with H2O2-induced injury subjected to si-Fosl1 treatment to examine the change in the above indexes and to determine whether the benefits from Fosl1 silencing occurred via AMPK. Moreover, we employed chloroquine (CQ) and rapamycin (RAP) to activate and inhibit autophagy, respectively, and revealed the effects of the upregulation and downregulation of autophagy following AMPK interference. Finally, an animal model was used to identify the effect of si-Fosl1 in vivo.Results: Based on the analysis of the GSE45006 datasets, Fosl1 was found to be highly expressed and was also found to be significantly enhanced in our animal model. Fosl1 knockdown upregulated AMPK at the protein and mRNA levels, promoted autophagic proteins (LC3 II/I, Beclin1) and inhibited inflammatory factors (IL-1β, IL-6, TNF-α) and apoptosis markers (caspase3, Bax). However, Fosl1 decreased the negatively related autophagic protein p62, the anti-inflammatory factor IL-10 and the anti-apoptotic marker Bcl-2. By utilizing compound C (com, an AMPK inhibitor), we learned that AMPK inhibition exhibited unfavorable effects on autophagy but promoted inflammation and apoptosis following Fosl1 silencing. AMPK activation showed contrasting effects. Moreover, we used CQ (an autophagic inhibitor), which indicated that CQ reversed the benefits of AMPK activation on inflammation and apoptosis. The autophagic activator RAP attenuated the negative effects after com treatment. In vivo, si-Fosl1 increased BBB scores at 7 d and 14 d and motor neurons, meanwhile, it decreased the number of apoptotic cells, and inflammatory cytokine expression at 14 d postoperation. Conclusion: Fosl1 can suppress AMPK to promote inflammation and apoptosis through autophagy in SCI.


Background
Spinal cord injury (SCI) is a disastrous nerve trauma in the central nervous system (CNS) currently without su cient treatment options, which affects several millions people worldwide and leads to catastrophic complications [1,2], such as neurological de cits [3], paraplegia [4], osteoporosis [5], and others. These complications place huge psychological and physical burdens on patients and threaten their health and life quality [6]. Despite strenuous efforts that have been spent on a treatment strategy for SCI [7,8] and that considerable progress has been exhibited in trials, satisfactory clinical bene ts still remain unavailable [9]. In order to explore more effective therapies, an understanding of the mechanism and pathology involved in SCI is critical [10]. Due to abundant harmful factors at the injury site following SCI and the inability of neuron axon regeneration, SCI recovery is almost impossible. Since effective treatments against SCI are not possible, a way to decrease the degree of SCI has become a research focus. The pathological process of SCI involves primary injury and secondary injury, and evidence has indicated that the secondary injury is the essential molecular event [11]. Because the primary injury is unpredictable and irreversible, therapeutic interventions for the secondary injury to reduce damage, inhibit apoptosis, and alleviate the in ammatory response during the acute period of SCI represents a novel research strategy for SCI [12][13][14].
Fos like antigen 1 (Fosl1) is a transcription factor of the fos gene family that has been researched extensively in many diseases [15,16]. It can combine with proteins of the Jun family to mediate tumor occurrence, progression, invasion, metastasis, and other processes [17][18][19][20]. In addition, it is particularly related to cell differentiation, proliferation and apoptosis [21] and plays an important role in cholangitis [22], arthritis [23], and diseases of damage, such as neuron injury [24][25][26][27]. One study has indicated that Fosl1 de ciency protects mouse embryo broblasts from apoptosis induced by oxidative stress [28]. Nevertheless, Fosl1 overexpression enhances the apoptosis of C6 glioma cells [29]. The role of Fosl1 has been investigated in a model of lung injury, where it regulates LPS-induced in ammatory factors such as IL-10 and IL-1β [26,30,31]. Fosl1 is obviously enhanced following injury and increases cell in ammation and apoptosis. However, whether it also has a vital role in the secondary injury of SCI and what that may be are still unclear.
Autophagy is a unique lysosomal-mediated catabolic process in eukaryotic organisms that phagocytizes and degrades abnormal proteins or dysfunctional organelles for maintaining homeostasis in response to different stresses, including nutrient deprivation, injury, and hypoxia [32][33][34]. It can be activated by internal or external stimuli and reduce the harm of detrimental factors or some metabolic wastes, which become isolated within a bilayer membrane of the autophagosome, accompanied by LC3 II formation from LC3 I [35]. Subsequently, these autophagosomes combine with lysosomes to form autolysosomes, which degrade these substances to recycle them. A number of studies have reported that autophagy is upregulated in many diseases, such as cancers, brain injury, and kidney injury [32]. Multiple studies have shown that autophagy has an important protective role in the secondary injury and recovery of SCI [36][37][38]. Autophagy, which usually maintains a low activity, can be induced following SCI and exert antiapoptotic and anti-in ammatory effects [9,39,40], but this autophagy-induced effect in SCI is very limited. Therefore, how to enhance autophagy to reduce apoptosis and in ammation is a research focus, but the underlying molecular mediated-mechanism has still not been elucidated clearly. Thus, we assume that the upregulation of Fosl1 promotes in ammation and apoptosis to aggravate the secondary injury via constraining autophagic activity in SCI. AMP-activated protein kinase (AMPK) is a classic signaling pathway that participates in mediating autophagy and is expressed at a basal level in the body [41]. It can be activated in SCI and plays an important protective role in SCI improvement [37,42]. Hence, we further hypothesized that Fosl1 ampli es in ammation and apoptosis by inhibiting autophagy via the AMPK signaling pathway following spinal cord injury.

Bioinformatic analysis
To investigate the hypothesis that Fosl1 was upregulated in SCI, we explored the Gene Expression Omnibus database (GSE 45006) and used the GPL1355 platforms to download the information of the sham group and the SCI 1 d, 3 d, 7 d, and 14 d groups. After analysis with R language, we ltered out the 150 top differential genes in each time point and generated heatmaps. Subsequently, we analyzed the Fosl1 data to de ne our time point for the SCI animal model.

Animal model, groups and treatments
Eighteen speci c pathogen-free adult Sprague-Dawley (SD) rats weighting 220-250 g were provided by the Anhui Experimental Animal Center of China. Subsequently, the animals were bred in an airconditioned room with a 12 h day and night cycle under a stable temperature (24 ± 2 °C) and humidity (60 ± 5%).
In part 1, the rats were randomly divided into three groups: normal group (n = 6), sham group (n = 6), and SCI group (n = 6). The SCI model was established based on a previously published report using Allen's method [43]. Brie y, following anesthetization, a T10 laminectomy was executed from T8 to T12. Next, rats from the SCI group were subjected to a force (10 g × 5 cm) to produce a T10 contusive SCI. The animals in the sham group received a laminectomy only, and those in the normal group received no treatment. The bladder was emptied twice per day until its function was restored.
In part 5, 20 rats were divided into four groups: sham + saline group (n = 5), SCI + saline group (n = 5), SCI + si-NC group (n = 5), and SCI + si-Fosl1 group. The animals were intrathecally injected with sterile saline, si-NC, or si-Fosl1 immediately and were intramuscularly injected with penicillin for the rst 3 consecutive days after the operation. The BBB scale was applied to evaluate the hindlimb movement at the designed time point, and the rats were sacri ced to remove the target tissues after 14 days.

Cell culture and treatments
PC-12 is a classical neuronal cell line that was purchased from the Chinese Academy of Sciences. To further identify the mechanism involved in SCI, we utilized this cell line treated with H 2 O 2 (100 µmol for 24 h) to mimic neuron injury. PC-12 cells were cultured in high-glucose DMEM with a GlutaMAX™ supplement (10564011, Gibco) and 10% fetal bovine serum (10100147C, Gibco). Our in vitro study involved three parts (part 2, part 3, and part 4). In part 2, we used si-RNA to verify whether there were bene ts following Fosl1 knockdown. In this part, the cells were divided into four groups: PBS, H 2 [44], and H2O2 + si-RNA + compound C (com, an inhibitor, 2 µmol for 24 h) [33]. In part 4, rapamycin (RAP, an activator of autophagy, 5 µmol for 24 h) and chloroquine (CQ, an inhibitor of autophagy, 20 µmol for 24 h) [45] were used to assess whether the effect of the Fosl1 knockdown was generated ultimately via autophagy. The cells were grouped as follows: H 2 O 2 + si-RNA + met VS. H 2 O 2 + si-RNA + met + CQ, H 2 O 2 + si-RNA + com VS. and H 2 O 2 + si-RNA + com + RAP.

Hindlimb locomotor assessment
The Basso, Beattie, and Bresnahan (BBB) scale [46] was applied to evaluate hindlimb locomotion. After surgery, 0 score represented complete hindlimb disability, which indicated that the SCI model was successful. The test was performed by three researchers who were blinded to the speci cs of the three groups. Finally, the scores were averaged.

Nissl staining and TUNEL staining
The rats were sacri ced for removing the spinal cord. The horizontal tissue sections were harvested.
To count the motor neurons, Nissl staining (Beyotime, Shanghai, China) was used to evaluate the Nissl bodies. After staining with crystal violet for 10 min, the tissue sections were observed to count the Nissl bodies in 5 randomly chosen elds with a microscope at a magni cation of 200× (Olympus, Japan).
The TUNEL Apoptosis Assay Kit (Beyotime, Shanghai, China) was used to assess cell apoptosis in the specimens according to the manufacturer's protocol. After staining, the samples were observed with an Olympus immuno uorescence microscope to calculate the number of TUNEL + cells.
To assess the in ammatory level of the spinal cord after si-Fosl1 injection, immuno uorescence staining was applied to examine the in ammatory marker protein. The sections were incubated with primary antibody in a refrigerator for 24 h and then with secondary antibody and DAPI (Beyotime, Shanghai, China) at room temperature. Next, the sections were assessed as described above.

Immunohistochemistry (IHC) staining
To detect histological changes in Fosl1 (1:100, sc-28310, Santa Cruz Biotechnology) and AMPK (1:100, sc74461, Santa Cruz Biotechnology), the slides were stained by IHC. After treatment with hydrogen peroxide, QuickBlock™ Blocking Buffer (Beyotime, Shanghai, China) was used to block non-speci c proteins. The primary antibodies were applied to cover the tissue at 4 °C overnight, and then the tissues were incubated with secondary antibody (ZSGB-BIO, Beijing, China). Following staining with DAB (ZSGB-BIO, Beijing, China) and hematoxylin (Beyotime, Shanghai, China), the slides were visualized using an Olympus microscope.
The spinal cord was extracted and dissolved in RIPA lysis buffer (Beyotime, Shanghai, China) to which protease and phosphatase inhibitors were added. The mixture was incubated on ice for 30 min and centrifuged at 12000 rpm for 10 min, after which the supernatant was collected. Total protein was measured using bicinchoninic acid assays in accordance with the manufacturer's instructions and the concentration was uniform. Subsequently, 5 × SDS loading buffer was added to the supernatant for denaturing the sample. The samples were separated by SDS-PAGE using the same volume per sample. Next, the proteins were transferred to a PVDF membrane (Millipore, MA, USA). After blocking with 5% skimmed milk, the membrane was incubated overnight at 4 °C with the following primary antibodies: (1:20000, AB2839429/AB2839430, A nity Biosciences), and then incubated with ECL (Thermo Scienti c, USA). The bands were checked using an imaging system (Bio-Rad, USA), and the IOD value was measured by Image-Pro Plus software.

Quantitative real-time PCR (qRT-PCR)
Total RNA was extracted from the spinal cord or cells using TRIzol reagent (Invitrogen, USA). The concentration of mRNA was examined using a NanoDrop One device (Thermo, MA, USA) and diluted to 0.5 µg/µl. The purity was assessed using 260/280 nm. Following RNase-free DNase digestion, a reverse transcription kit was used for cDNA synthesis (TaKaRa, Tokyo, Japan). The primers were synthesized by Sangon Biotech Co., Ltd. according to the Pubmed Gene Bank (Table 1). Reverse: ATGAAGGGGTCGTTGATGGC A 20 µl reaction system was used as follows: 1 µl cDNA, 1 µl forward primer, 1 µl reverse primer, 10 µl 2 × mix and 7 µl RNase-free water. The PCR protocol included 95 °C for 2 min, followed by 40 cycles of 95 °C for 10 s and 60 °C for 30 s. GAPDH was used as the standard reference. All experiments were performed in triplicate, and the experimental data were analyzed by the 2-ΔΔCT method.

Enzyme-Linked Immunosorbent Assay (ELISA)
After the cell samples were homogenized with 0.1 M PBS on ice, the mixture was centrifuged at 5000 g for 10 min. The supernatant was removed to another tube for a subsequent assessment. Enzyme-Linked Immunosorbent Assay Kits (IL-1β, IL-6, IL-10, and TNF-α) were provided by Abcam (Cambridge, UK). The optical density values at 450 nm were recorded. All procedures were performed in accordance with the manufacturer's protocol.

Small interfering RNA (si-RNA) transfection
To knockdown Fosl1 expression, PC-12 cells were cultured in a 6-well plate at 1 × 10 5 cells/well in highglucose DMEM without FBS and transiently transfected with si-Fosl1 with a working concentration of 50 nM using a riboFECTTM CP Transfection Kit (RiboBio, Guangzhou, China) according to the manufacturer's instructions. The sequence used was 5′-CATCGAAAGAGTAGCAGCA-3′.
Immuno uorescence microscopy (Olympus, Japan) and qRT-PCR were employed to monitor the e ciency 48 h after transfection.

Cell Counting Kit-8 (CCK-8)
The Enhanced Cell Counting Kit-8 (Beyotime, Shanghai, China) was used to monitor the cell viability. After the cells (100 µl of a 5 × 10 4 cells/ml suspension) were seeded in 96-well plates for 24 h, reagents or PBS were added to the plates for the cells' treatments. Subsequently, 10 µl CCK-8 solution was added to the 96-well plates. After incubation, the 96-well plate was examined at 450 nm in an instrument to detect the enzyme label at each time point.

Flow cytometry (FCM)
Flow cytometry was used to assess the cell apoptosis rate after the different treatments. After digestion with 0.25% trypsin without EDTA, the cells were harvested in an EP tube and washed three times with precooled PBS. The precipitates were resuspended into a single cell suspension in 100 µl 1 × buffer, after which 5 µl annexin V-FITC and 5 µl PI staining solution (Vazyme Biotech, Nanjing, China) were added. Subsequently, another 400 µl 1 × buffer was added to these mixtures, after which the cell apoptosis rate was examined using a ow cytometer.

CO-immunoprecipitation (COIP)
To further investigate the relationship between Fosl1 and AMPK, we employed COIP to examine whether there was a direct interaction between them. Cells were collected and treated by IP-grade Cell Lysis Buffer (Beyotime, Shanghai, China) for 10 min on ice. This mixture was added into a 1.5 ml EP tube for centrifugation (12000 rpm, 4 ℃, 10 min). Subsequently, 10 µl of the mixture was retained as the input.
The remainder was transferred to another EP tube to which a pre-treating solution including 1 µg IgG and 20 ul Protein A/G PLUS-Agarose (Santa Cruz, CA, USA) was added for removal of non-speci c proteins. After centrifugation, the supernatant was transferred to three other EP tubes, to which was added IgG, an anti-Fosl1 antibody, or an anti-AMPK antibody, respectively, in addition to 20 ul Protein A/G PLUS-Agarose, followed by incubation at 4 °C overnight. Following centrifugation, the pellets were rinsed three times with lysis buffer and boiled for 10 min. The samples were detected using western blotting.

Statistical analysis
All results were presented as the mean ± standard error of the mean (M ± SEM). The difference between two groups was analyzed using an unpaired 2-tailed t-test. One-way analysis of variance (ANOVA) with the least signi cant difference (LSD) and Student-Newman-Keuls (S-NK) tests were used to make comparisons among multiple groups with SPSS 21.0 (SPSS, Inc., Chicago, IL, USA). The graphs were generated with GraphPad Prism 6.02 (GraphPad Software, Inc., California, USA). A p < 0.05 was considered to be statistically signi cant.

Fosl1 was signi cantly upregulated at 1 d and 3 d postoperation, as assessed by a bioinformatic analysis of the Gene Expression Omnibus database.
Based on extraction and analysis of data from the GSE 45006 datasets, which included 20 SCI rats and 4 sham rats, we collected the information of the sham group and the SCI groups (at 1 d, 3 d, 7 d, and 14 d postoperation) (Fig. 1A) and used the R language tool to analyze these data. One hundred fty differentially expressed genes were found, which are shown in the four heatmaps (Fig. 1B). Fosl1 was highly expressed at 1 d and 3 d postoperation compared to the sham group, but there was no signi cance between the groups at 7 d and 14 d postoperation (Fig. 1C).
3.2 Fosl1 and AMPK were obviously upregulated in the spinal cord at 1 d postoperation, and Fosl1 was also expressed signi cantly in neurons.
According to the bioinformatic analysis results, 1 d was chosen as the time point for sacri cing animals and harvesting the target tissues. After investigating Fosl1 protein expression in the three groups, we found that Fosl1 were upregulated signi cantly in the injured spinal cord vs. the other groups, and the comparison between the normal group and the sham group was not signi cantly different (Fig. 2B). To examine the expression more speci cally, qRT-PCR was used to examine the Fosl1 mRNA level. Similar to what was seen with the western blotting, it was observed that Fosl1 mRNA in the SCI group was expressed highly vs. the other groups (Fig. 2C). There was also no signi cant difference between the normal group and the sham group at the mRNA level (Fig. 2C). To assess the histological change, we used immunohistochemical staining, which showed that the average optical density of Fosl1 was signi cantly greater compared to the other groups (Fig. 3A).
To further examine Fosl1 expression in neurons, we employed immuno uorescence double-label staining to detect Fosl1 and the neuronal biomarker protein β-III tubulin, which demonstrated that Fosl1 was highly expressed in β-III tubulin-positive cells (Fig. 2F).
AMPK was examined using western blotting, qRT-PCR and immunohistochemical staining as described above. We determined that AMPK was increased obviously at the mRNA and protein levels ( Fig. 2D, E), and there was a signi cant difference in the histological behavior (Fig. 3B).
3.3 Fosl1 and AMPK were enhanced signi cantly in PC-12 cells with H 2 O 2 -induced injury.
To further examine Fosl1 and AMPK expression, the classical H 2 O 2 -induced PC-12 injury model was employed to mimic neuronal cell injury. We found that the Fosl1 expressions at mRNA and protein level in the H 2 O 2 group was increased apparently compared to the PBS group ( Fig. 4B, C). Its expression in the PBS group was negligible. The same methods were used to investigate AMPK expression at the protein and mRNA levels. We discovered a similar result compared to Fosl1, with a signi cant difference between both groups (Fig. 4D, E).

Knockdown of Fosl1 enhanced AMPK expression and autophagic activity in PC-12 cells with H 2 O 2 -induced injury.
To further detect Fosl1 function, si-RNA technology was used to knockdown Fosl1 expression. We found that Fosl1 was downregulated in the si-Fosl1 group compared to the H 2 O 2 group and the H 2 O 2 + si-NC group ( Fig. 4B, C), but the AMPK and p-AMPK expression levels were enhanced vs. both groups (Fig. 4D, E). COIP was employed to identify whether there is a connection between Fosl1 and AMPK. The result indicated Fosl1 directly bound to AMPK in injured PC-12 cells (Fig. 5C). Autophagy is a normal response to external or internal stimuli. To assess autophagic activity, marker proteins such as LC3 II/I, Beclin1, and p62 were tested. We found that the positively related proteins LC3 II/I and Beclin1 were highly upregulated in the si-Fosl1 group, and the negatively related protein p62 was decreased vs. the H 2 O 2 group and the H 2 O 2 + si-NC group (Fig. 5A).

Knockdown of Fosl1 improved the survival rate and decreased in ammation and apoptosis in PC-12 cells with H 2 O 2 -induced injury.
To check the effect of Fosl1 knockdown on injured PC-12 cells, CCK-8 and FCM was applied to assess the cell viability and the apoptotic rate. We learned that the cell viability in the si-Fosl1 group was higher than in the other two groups (Fig. 4A), but the apoptotic rate was lower (Fig. 5B). In addition, in ammation was assessed using ELISA technology. The pro-in ammatory proteins IL-1β, IL-6, and TNF-α were attenuated signi cantly vs. the H 2 O 2 group and the H 2 O 2 + si-NC group, but the protective protein IL-10 was increased ( Fig. 5D). To investigate apoptosis, we performed western blotting to assess the Bcl-2, Bax, and caspase3 proteins. The results showed that the apoptotic-inducing proteins Bax and caspase3 were decreased in the si-Fosl1 group and the apoptotic-inhibiting protein Bcl-2 was enhanced (Fig. 5A).
3.6 Upregulation of AMPK could increase cell viability and autophagic activity, and alleviate in ammation and apoptosis in injured PC-12 cells with Fosl1 knockdown. Inactivated AMPK could neutralize the bene cial effects of si-Fosl1 in injured PC-12 cells.
To investigate the detailed role of AMPK in injured PC-12 cells after si-Fosl1 treatment, an AMPK activator (met) was used to enhance the AMPK activity (Fig. 6B). We found that LC3 II/I, and Beclin1 were increased and that p62 was downregulated (Fig. 7A), which indicated that the autophagic level was enhanced signi cantly in the met group following AMPK activation. The proteins positively related to in ammation and apoptosis, such as IL-1β, IL-6, TNF-α, Bax, and caspase3, were decreased (Fig. 6C, 7A).
Nevertheless, the anti-in ammatory cytokine IL-10 and the anti-apoptotic protein Bcl-2 were increased (Fig. 6C, 7A). The cell viability of met group was better than control group (Fig. 6A), but the apoptotic rate was lower (Fig. 7B).
The same method was applied to inhibit AMPK activity using com. By inactivating AMPK, we observed that the cell viability and autophagic activity were obviously downregulated (Fig. 6A, 7B). However, the levels of the in ammatory and apoptotic levels were enhanced (Fig. 6C, 7A).
3.7 Autophagic activation could attenuate in ammation and apoptosis in injured PC-12 cells with Fosl1 knockdown after AMPK inhibition, but autophagic suppression exhibited the opposite effects.
To identify whether AMPK-mediated in ammation and apoptosis in injured PC-12 were regulated via autophagy, we divided the injured PC-12 cells with si-Fosl1 treatment into two groups. One group had cells treated with AMPK activator (met) and autophagic inhibitor (CQ), and the other group was treated with AMPK inhibitor (com) and autophagic activator (RAP). We found that CQ alleviated the AMPK activator bene ts on in ammation and apoptosis (Fig. 8B, C). Meanwhile, CQ decreased the cell viability and raised apoptotic cells rate (Fig. 8A, D).
The autophagic activator prevented in ammation and apoptosis after com treatment (Fig. 9B, C). moreover, RAP increased the cell viability and reduced apoptotic cells rate (Fig. 9A, D). These outcomes further illustrated that AMPK-mediated in ammation and apoptosis occurred via autophagy.
3.8 si-Fosl1 improved neurological function in rats with SCI.
To verify the si-Fosl1 bene ts on rats with spinal cord injury, we designed four groups (sham + saline, SCI + saline, SCI + si-NC, and SCI + si-Fosl1). The BBB scale was employed to evaluate the hindlimb function at 30 min pre-operation and at 0 d, 1 d, 3 d, 7 d, and 14 d postoperation. The preoperative BBB score of all animals and at each time point in the sham group was 21, but the score in the other groups was 0 at 0 d, which showed that the rat SCI model was successful. Additionally, there was no signi cant difference when comparing the SCI + si-Fosl1 group with the SCI + saline group and the SCI + si-NC group at 1 and 3 d (Fig. 10A). The results were signi cantly different, which indicated that si-Fosl1 could evidently improve the hindlimb locomotion at 7 and 14 d postoperation (Fig. 10A). A comparison between the SCI + saline group and the SCI + si-NC group showed that there was no signi cant difference at each time point (Fig. 10A). This result further demonstrated that Fosl1 knockdown could promote the recovery of neurological function and improve the hindlimb locomotion.
3.9 si-Fosl1 decreased apoptosis and in ammation but increased the survival rate of motor neurons.
To further check the difference at the cellular level, TUNEL and Nissl staining were used to examine the apoptotic cells and the surviving motor neurons, respectively. By analyzing the TUNEL + cells, we found that there were signi cant differences between the SCI + si-Fosl1 group and the other two groups (Fig. 10B, E). To further assess the neurological function of the hindlimb at histologic level, we used Nissl staining at 14 d following SCI. The Nissl bodies were markedly increased in the SCI + si-Fosl1 group ( Fig. 10C, F).
The in ammatory change in the four groups was assessed using immuno uorescence staining at 14 d postoperation. By measuring the reactive astrocytes marker GFAP expression which re ects in ammatory activity, we found that the SCI + si-Fosl1 group fared lower than the SCI + saline group and the SCI + si-NC group (Fig. 10D, G). Moreover, the comparison between the other two groups showed that there was no signi cant difference (Fig. 10D, G). Subsequently, the in ammatory marker proteins, IL-1β, IL-6, IL-10, TNF-α were checked. The results indicated the in ammation level was lower in the SCI + si-Fosl1 group than the other two groups (Fig. 11A-H).

Discussion
Spinal cord injury remains an unsolved neurologic disease with no effective clinical treatment to date, although considerable progress has been made in the past few decades [47]. The main challenge in the eld of SCI is the massive loss of neurons during secondary injury and the extremely limited neuronal regeneration [48,49]. Innovations in stem cell transplantation once presented an effective strategy to provide extrinsic nerve cells to build new networks and rescue neurological function, but satisfactory outcomes have yet to be been achieved [50]. The SCI secondary injury plays a more important role in the neuronal loss, which is often accompanied by in ammation, apoptosis, autophagy, and necrosis [51,52].
However, the precise mechanism is still unclear.
Although Fosl1 is well-known as an essential transcription factor that constitutes the AP-1 compound with c-fos and c-jun in the development of many diseases of injury [24,[53][54][55], its role in the occurrence and progression of SCI has not yet been elucidated. By exploring the GSE45006 and analyzing the datasets, we learned that Fosl1 expression was signi cantly upregulated following SCI at 1 d and 3 d postoperation. In our study, we veri ed that Fosl1 was highly expressed at the mRNA and protein levels at 1 d postoperation. The outcome was consistent with previous studies from the GEO database. Due to neuronal injury playing an essential role in SCI, we used double-label immuno uorescence staining to detect Fosl1 in neurons and discovered that Fosl1 expression was signi cantly enhanced in neurons. In addition, we used 100 nmol/L H 2 O 2 to induce injury in PC-12 cells for 24 h to mimic neuronal injury in vitro. Fosl1 was also observed to be highly expressed at the mRNA and protein levels. These results indicated that the upregulation of Fosl1 in neurons may play an essential role during secondary injury in vivo and vitro.
Small interfering RNA is a double-stranded RNA with a speci c length and sequence that completely complements a speci c mRNA and results in degradation of the target mRNA and blocking of the translation process [56,57]. We used si-RNA technology to precisely assess the Fosl1 role in SCI. After Fosl1 knockdown in the injured neurons, the cells' viability was increased signi cantly, but the cells' apoptotic rate was decreased. Additionally, we found that the pro-in ammatory cytokines IL-1β, IL-6 and TNF-α were downregulated and that the anti-in ammatory cytokine IL-10 was upregulated following Fosl1 knockdown. These outcomes indicated that Fosl1 silencing is advantageous for inhibiting neuronal injury. In addition, the proteins positively related to autophagy and apoptosis, such as Beclin1, LC3 II/I and Bcl-2, were enhanced, but negatively related proteins, such as p62, Bax, and caspase3, were reduced.
Moreover, the signaling pathway proteins AMPK and p-AMPK were increased. These results clearly clari ed that si-Fosl1 can activate AMPK and autophagy and can alleviate neuronal in ammation and apoptosis.
The AMPK signaling pathway is a classical pathway that is involved in modulating autophagy via threonine (Thr-172) phosphorylation [58]. A previous study has shown that AMPK was increased and improved function in a rat model with SCI [59]. This result was shown at the protein and mRNA levels in our study. To determine the detailed role of AMPK in injured neurons, we used its activator (metformin, met) and inhibitor (compound C, com) to treat the injured cells after si-Fosl1 treatment [60][61][62]. After met utilization for 24 h, we found that the apoptotic rate was lower and that the cell viability was also improved. The in ammatory cytokines and apoptotic proteins were diminished signi cantly, although IL-10 and Bcl-2 were upregulated. These interesting phenomena were completely opposite after com treatment, and more importantly, the autophagy activity was decreased. These results suggested that si-Fosl1 exerts anti-in ammatory and anti-apoptotic effects via the AMPK signaling pathway.
Autophagy is regarded as an innate immune defense mechanism against a more serious external or internal injury [63], which performs an anti-in ammatory and anti-apoptotic Author Contributions observe that RAP could reverse the in ammation and apoptosis after AMPK inactivity in injured cells with si-Fosl1 treatment and bene t the cells' viability and apoptotic rate. Combining these results, we were able to de nitely conclude that Fosl1 suppressed AMPK activity to activate in ammation and apoptosis through autophagy in SCI. Our ndings are the rst to demonstrate the novel axis of Fosl1/AMPK/autophagy playing a vital role in SCI.
Next, we utilized an animal model to verify that si-Fosl1 also exerts bene cial effects in vivo. The BBB scores showed that si-Fosl1 improved the hindlimb locomotor function at 7 d and 14 d postoperation, but there was no signi cance at 1 d and 3 d postoperation. The apoptotic cells change was evaluated using TUNEL staining at 14 d, which certi ed that TUNEL + cells was obviously lower than that in the control groups. In addition, Nissl staining was used to count the motor neurons and we found that the number of Nissl bodies was signi cantly better in si-Fosl1 group. Moreover, immuno uorescence staining was used to examine in ammatory marker cytokines (IL-1β, IL-6, IL-10, TNF-α), and in ammatory cytokines expression in the both control groups were worse than that in the si-Fosl1 group, however, the antiin ammatory cytokine IL-10 was higher. Meanwhile, the reactive astrocytes marker (GFAP) was lowly expressed in the si-Fosl1 group. These results revealed that si-Fosl1 exerted anti-in ammatory and antiapoptotic effects on SCI in vitro.

Conclusion
In conclusion, our study is the rst to show that Fosl1 plays a vital role in aggravating SCI secondary injury and that Fosl1 knockdown can promote autophagy-mediated anti-apoptotic and anti-in ammatory effects to improve neurological function via the AMPK signaling pathway. Availability of data and materials

Abbreviations
The datasets used and/or analyzed during the current study are available from the corresponding author on reasonable request.  Figure 1 The mRNA of Fosl1 was highly expressed in rats following SCI at 1 d and 3 d postoperation. A: Schematic of the experimental strategy used to screen the potential differential genes in the GEO database. B: The heatmaps generated by R language using the mRNA expression data of the SCI group and the sham group in GEO at 1 d, 3                  vs. H2O2 group, &p<0.05 vs. H2O2 + si-NC group, &&p<0.01 vs. H2O2 + si-NC group by ANOVA. Figure 5 si-Fosl1 enhanced the autophagic activity and decreased the apoptosis and in ammation. A: The autophagic marker proteins (Beclin1, p62, LC3) and the apoptotic proteins (Bax, Bcl-2, caspase3) were detected using western blotting. B: The apoptotic cell rate was monitored by ow cytometry. C: The COIP was used to examine the interaction of Fosl1 and AMPK in injured PC-12 cells. D: The in ammatory proteins (IL-1β, IL-6, TNF-α, IL-10) were examined using ELISA. All data are presented as the M ± SEM (n=3 in each group). *p<0.05 vs. PBS group, **p<0.01 vs. PBS group, #p<0.05 vs. H2O2 group, ##p<0.01

Figures
vs. H2O2 group, &p<0.05 vs. H2O2 + si-NC group, &&p<0.01 vs. H2O2 + si-NC group by ANOVA. Figure 5 si-Fosl1 enhanced the autophagic activity and decreased the apoptosis and in ammation. A: The autophagic marker proteins (Beclin1, p62, LC3) and the apoptotic proteins (Bax, Bcl-2, caspase3) were detected using western blotting. B: The apoptotic cell rate was monitored by ow cytometry. C: The COIP was used to examine the interaction of Fosl1 and AMPK in injured PC-12 cells. D: The in ammatory proteins (IL-1β, IL-6, TNF-α, IL-10) were examined using ELISA. All data are presented as the M ± SEM (n=3 in each group        The AMPK activator increased the autophagic activity and decreased the apoptotic proteins and apoptotic rate in the injured PC-12 cells with si-Fosl1 treatment. An AMPK inhibitor suppressed autophagy and elevated apoptotic protein expression and the apoptotic rate. A: The autophagic marker proteins (Beclin1, p62, LC3) and the apoptotic proteins (Bax, Bcl-2, caspase3) were measured using western blotting. B: The apoptotic cell rate was examined by ow cytometry. All data are presented as the M ± SEM (n=3 in each group). *p<0.05 vs. H2O2 + si-Fosl1 group, **p<0.01 vs. H2O2 + si-Fosl1 group (n=3) by ANOVA.

Figure 7
The AMPK activator increased the autophagic activity and decreased the apoptotic proteins and apoptotic rate in the injured PC-12 cells with si-Fosl1 treatment. An AMPK inhibitor suppressed autophagy and elevated apoptotic protein expression and the apoptotic rate. A: The autophagic marker proteins (Beclin1, p62, LC3) and the apoptotic proteins (Bax, Bcl-2, caspase3) were measured using western blotting. B: The apoptotic cell rate was examined by ow cytometry. All data are presented as the M ± SEM (n=3 in each group). *p<0.05 vs. H2O2 + si-Fosl1 group, **p<0.01 vs. H2O2 + si-Fosl1 group (n=3) by ANOVA.

Figure 7
The AMPK activator increased the autophagic activity and decreased the apoptotic proteins and apoptotic rate in the injured PC-12 cells with si-Fosl1 treatment. An AMPK inhibitor suppressed autophagy and elevated apoptotic protein expression and the apoptotic rate. A: The autophagic marker proteins (Beclin1, p62, LC3) and the apoptotic proteins (Bax, Bcl-2, caspase3) were measured using western blotting. B: The apoptotic cell rate was examined by ow cytometry. All data are presented as the M ± SEM (n=3 in each group). *p<0.05 vs. H2O2 + si-Fosl1 group, **p<0.01 vs. H2O2 + si-Fosl1 group (n=3) by ANOVA.

Figure 8
Inhibition of autophagy neutralized the bene ts on the autophagic activity and cell viability after AMPK activation in the injured PC-12 cells with si-Fosl1 treatment but promoted in ammation and cell apoptosis. A: Cell viability was detected using CCK-8 kits in the control group and the CQ group. B: The in ammatory level was checked by ELISA kits. C: The autophagic and apoptotic proteins were visualized by western blotting. D: The cell apoptotic rate was inspected using ow cytometry. All data are presented as the M ± SEM (n=3 in each group). *p<0.05 vs. si-Fosl1 + met group, **p<0.01 si-Fosl1 + met group (n=3) by T test.

Figure 8
Inhibition of autophagy neutralized the bene ts on the autophagic activity and cell viability after AMPK activation in the injured PC-12 cells with si-Fosl1 treatment but promoted in ammation and cell apoptosis. A: Cell viability was detected using CCK-8 kits in the control group and the CQ group. B: The in ammatory level was checked by ELISA kits. C: The autophagic and apoptotic proteins were visualized by western blotting. D: The cell apoptotic rate was inspected using ow cytometry. All data are presented as the M ± SEM (n=3 in each group). *p<0.05 vs. si-Fosl1 + met group, **p<0.01 si-Fosl1 + met group (n=3) by T test.

Figure 8
Inhibition of autophagy neutralized the bene ts on the autophagic activity and cell viability after AMPK activation in the injured PC-12 cells with si-Fosl1 treatment but promoted in ammation and cell apoptosis. A: Cell viability was detected using CCK-8 kits in the control group and the CQ group. B: The in ammatory level was checked by ELISA kits. C: The autophagic and apoptotic proteins were visualized by western blotting. D: The cell apoptotic rate was inspected using ow cytometry. All data are presented as the M ± SEM (n=3 in each group). *p<0.05 vs. si-Fosl1 + met group, **p<0.01 si-Fosl1 + met group (n=3) by T test.

Figure 8
Inhibition of autophagy neutralized the bene ts on the autophagic activity and cell viability after AMPK activation in the injured PC-12 cells with si-Fosl1 treatment but promoted in ammation and cell apoptosis. A: Cell viability was detected using CCK-8 kits in the control group and the CQ group. B: The in ammatory level was checked by ELISA kits. C: The autophagic and apoptotic proteins were visualized by western blotting. D: The cell apoptotic rate was inspected using ow cytometry. All data are presented as the M ± SEM (n=3 in each group). *p<0.05 vs. si-Fosl1 + met group, **p<0.01 si-Fosl1 + met group (n=3) by T test.

Figure 9
Activation of autophagy reversed the negative effects on the autophagic activity and cell viability after AMPK inhibition in the injured PC-12 cells with si-Fosl1 treatment and attenuated in ammation and cells apoptosis. A: Cell viability was detected using CCK-8 kits in the control group and the RAP group. B: The in ammatory level was checked by ELISA kits. C: The autophagic and apoptotic proteins were examined by western blotting. D: The cell apoptotic rate was inspected using ow cytometry. All data are presented in ammatory level was checked by ELISA kits. C: The autophagic and apoptotic proteins were examined by western blotting. D: The cell apoptotic rate was inspected using ow cytometry. All data are presented as the M ± SEM (n=3 in each group). *p<0.05 vs. si-Fosl1 + met group, **p<0.01 si-Fosl1 + met group (n=3) by T test.  Activation of autophagy reversed the negative effects on the autophagic activity and cell viability after AMPK inhibition in the injured PC-12 cells with si-Fosl1 treatment and attenuated in ammation and cells apoptosis. A: Cell viability was detected using CCK-8 kits in the control group and the RAP group. B: The in ammatory level was checked by ELISA kits. C: The autophagic and apoptotic proteins were examined by western blotting. D: The cell apoptotic rate was inspected using ow cytometry. All data are presented as the M ± SEM (n=3 in each group). *p<0.05 vs. si-Fosl1 + met group, **p<0.01 si-Fosl1 + met group (n=3) by T test.  (magni cation × 400). All data are presented as the M ± SEM (n=5 in each group). *p<0.05 vs. SCI + saline group, **p<0.01 vs. SCI + saline group (n=5). #p<0.05 vs. SCI + si-NC group, ##p<0.01 vs. SCI + si-NC group (n=5) by ANOVA.