Bone marrow mesenchymal stem cell exosomes-derived microRNA-216a-5p on locomotor performance, neuronal injury, and microglia inflammation in spinal cord injury

Background: MicroRNA-216a-5p (miR-216a-5p) mediates inflammatory responses and neuronal injury to participate in the pathology of spinal cord injury (SCI). This study intended to explore the engagement of bone marrow mesenchymal stem cell exosomes (BMSC-Exo)-derived miR-216a-5p in locomotor performance, neuronal injury, and microglia-mediated inflammation in SCI rats. Methods: Rat BMSC or BMSC-Exo was injected into SCI rats. GW4869 treatment was adopted to suppress the exosome secretion from BMSC. Subsequently, miR-216a-5p-overexpressed BMSC-Exo (BMSC-miR-Exo) or negative-control-overexpressed BMSC-Exo (BMSC-NC-Exo) were injected into SCI rats. Results: The injection of BMSC or BMSC-Exo enhanced locomotor performance reflected by Basso, Beattie & Bresnahan score (p < 0.001), and neuronal viability reflected by NeuN+ cells (p < 0.01), but attenuated neuronal apoptosis reflected by TUNEL positive rate, cleaved-caspase-3 expression, and B-cell leukemia/lymphoma-2 expression (p < 0.05). Additionally, the injection of BMSC or BMSC-Exo suppressed microglia M1 polarization-mediated inflammation reflected by IBA1+iNOS+ cells, tumor necrosis factor-α, interleukin (IL)-1β, and IL-6 (p < 0.01). Notably, the effect of BMSC on the above functions was retarded by the GW4869 treatment (most p < 0.05). Subsequently, the injection of BMSC-miR-Exo further improved locomotor performance (p < 0.05), while inhibiting neuronal apoptosis (p < 0.05) and microglia M1 polarization-mediated inflammation (p < 0.05) compared to BMSC-NC-Exo. Interestingly, the injection of BMSC-miR-Exo reduced toll-like receptor 4 (TLR4) (p < 0.01), myeloid differentiation factor 88 (p < 0.05), and nuclear factor kappa B (NF-κB) (p < 0.05) expressions versus BMSC-NC-Exo. Conclusion: BMSC-Exo-derived miR-216a-5p enhances functional recovery by attenuating neuronal injury and microglia-mediated inflammation in SCI, which may be attributable to its inhibition of the TLR4/NF-κB pathway.


Introduction
Spinal cord injury (SCI) is one of the most disabling and destructive neurological disorders, which severely affects patients' quality of life (Ahuja et al., 2017b).According to the Global Burden of Disease Study 2019, the incident and prevalent cases of SCI are 0.9 million and 20.6 million in 2019 (Ding et al., 2022).In general, SCI patients often have a partial or complete loss of sensory and locomotor performances below the plane of injury, showing a range of clinical symptoms such as abnormal muscle tone, urinary and defecation disorders, and limb paralysis (McDonald and Sadowsky, 2002;Eli et al., 2021).In terms of treatment, the main neuroprotective therapies for acute or early SCI include surgery, medication, and vasopressor therapy (Ahuja et al., 2017a).Unfortunately, there is no way to reverse damage to the spinal cord completely, and effective strategies that promote functional recovery from chronic SCI are still lacking (Venkatesh et al., 2019).Under these situations, it is necessary to explore potential treatment that improves the neurological function and locomotor performance of SCI.
Bone marrow mesenchymal stem cell exosomes (BMSC-Exo) have exhibited certain therapeutic effects on SCI (Liu W. Z. et al., 2021).For example, one previous study illustrates that BMSC-Exo attenuates apoptosis and promotes autophagy in neuronal cells of SCI rats (Gu et al., 2020).In addition, another study discovers that BMSC-Exo suppresses neuronal apoptosis and accelerates locomotor performance recovery by activating the Wnt/β-catenin pathway in SCI rats (Li et al., 2019).Furthermore, BMSC-Exo facilitates the phagocytosis of macrophages to clean myelin debris, thereby improving functional recovery after SCI (Sheng et al., 2021).These studies indicate that BMSC-Exo could act as a promising regimen for SCI.Notably, BMSC-Exo possesses abundant RNA, lipids, and proteins; these contents play an essential role in the process of intercellular material information transfer (Kalluri and LeBleu, 2020).Therefore, the underlying mechanism of BMSC-Exo in treating SCI deserves further investigation.
MicroRNA-216a-5p (miR-216a-5p) possesses the ability to regulate neuroinflammatory responses, neuronal injury, and neurogenesis in several neurological diseases, such as traumatic brain injury and Alzheimer's disease (Xu et al., 2020;Xin Y. et al., 2021;Shao, 2021).For instance, one study indicates that miR-216a-5p attenuates neuroinflammatory response by inhibiting the high-mobility group box 1/nuclear factor kappa B (NF-κB) pathway (Shao, 2021).In addition, another study elucidates that increased miR-216a-5p in brainderived neurotrophic factor-induced MSCs-Exo promotes neurogenesis after traumatic brain injury in PC12 cells (Xu et al., 2020).In terms of SCI, one study illustrates that miR-216a-5p is enriched in hypoxic BMSC-Exo and may potentially participate in the hypoxic BMSC-Exo-mediated microglial polarization (Liu et al., 2020).However, it should be mentioned that the previous study only discloses a possibility that BMSC-Exo may result in the increase of miR-216a-5p under hypoxic conditions, thus further helping to strengthen functional recovery after SCI (Liu et al., 2020).While the direct engagement of BMSC-Exo-derived miR-216a-5p in SCI is unknown and should be further investigated.
Accordingly, this research intended to study the direct influence of BMSC-Exo-derived miR-216a-5p on locomotor performance, neuronal injury, and inflammation in SCI rats.

Animals and ethics
The adult Sprague-Dawley (SD) rats (8 weeks old) from Shanghai SLAC laboratory animal (Shanghai, China) were housed in standard conditions (light-shade alternating cycle for 12 h, 24 °C ± 2 °C, and humidity of 50%-60%).All experimental and surgical procedures were permitted by the Animal Care and Use Committee and carried out according to the National Institutes of Health Guidelines.

BMSCs culture, GW4869 treatment, and exosomes separation
Rats were euthanized with cervical dislocation after being anaesthetized by isoflurane inhalation, then femurs and tibias were isolated.The muscle attached to the bone was removed to expose the marrow cavity.Bone marrow was flushed by low-glucose Dulbecco's modified eagle medium (DMEM; Gibco, United States).The cell suspension was then filtered, centrifuged, and plated in lowglucose DMEM containing 10% fetal bovine serum (FBS; Gibco, United States).The BMSCs were harvested after the adherent cells reached 80-90% confluence, and the BMSCs at 3rd passage were used for further experiments.The immunophenotypes of BMSCs were identified by flow cytometry after being stained with antibodies against CD29, CD44, CD90, CD105, CD34, and CD45 (BD, United States).The GW4869 (10 μM; Yeason, China) was adopted to suppress the exosomes secretion of BMSCs, and the isolation of BMSC exosomes (BMSC-Exo) was carried out in the presence of Exosome Isolation Kit (Yeasen, China) according to the kit's protocol.

Establishment and grouping of SCI model
The SCI model was constructed as previously described (Li X. et al., 2020).Briefly, a laminectomy was performed to expose the spinal cord.Then, the spinal cord was contused with Infinite Horizon Impactor (PSI, United States).The injection of BMSCs or BMSC exosomes was carried out via tail intravenous after the surgery.
Firstly, 30 rats were separated into 5 groups (n = 6 per group): the Sham group, which received laminectomy only and injection of PBS; the SCI group, which received SCI surgery and injection of PBS; the BMSC group, which received SCI surgery and injection of BMSCs (5×10 6 cells in 500 μL PBS); the BMSC-GW group, which received SCI surgery and injection of GW4869-treated BMSCs (5×10 6 cells in 500 μL PBS); and the BMSC-Exo group, which received SCI surgery and injection of BMSC-Exo (200 μg exosomes in 500 μL PBS).
Then, additional 24 SCI rats were adopted to determine the engagement of miR-216a-5p.Briefly, rats were separated into 4 groups (n = 6 per group) after SCI surgery.The rats in the SCI group were given injection of PBS, and the rats in the BMSC-Exo, the BMSC-NC-Exo, and the BMSC-miR-Exo groups were given injection of BMSC-Exo, BMSC-NC-Exo, or BMSC-miR-Exo (200 μg exosomes in 500 μL PBS), respectively.
The Basso, Beattie & Bresnahan (BBB) locomotor score (Wang et al., 2020) were applied to quantify the locomotion recovery at 1, 3, 7, 14, 21, and 28 days after SCI surgery.All rats were euthanized at Day 28, and samples from spinal cord lesions were obtained for further assays.

Hematoxylin-eosin (HE), TUNEL, and immunofluorescence (IF) staining
The tissues from the spinal cord lesions were fixed in 4% paraformaldehyde (Sangon, China), embedded in paraffin, and sliced into 4 μm thick sections.The HE and TUNEL staining were carried out in the presence of HE or TUNEL Staining kits (Yeason, China).For IF staining, sections were blocked with 10% BSA (Beyotime, China) and incubated with anti-NeuN, anti-IBA1, or anti-iNOS antibodies (1:100, Abcam, United States), followed by incubation of fluorescently labeled secondary antibodies (1:250, Abcam, United States) for 1 h at 37 °C.The DAPI (Servicebio, China) staining was used to mark the nuclei.The antibody staining was specific.

RT-qPCR
RT-qPCR was performed to assess miR-216a-5p level in BMSCs, BMSC exosomes, or spinal cord lesion tissues.In brief, the total RNA was isolated with TRIzol reagent (Invitrogen, United States).

Statistical analysis
GraphPad software 7.0 (GraphPad, United States) was used for statistical analysis.Among the sham, SCI, BMSC, BMSC-GW, and BMSC-Exo groups, 6 biological replicative samples were tested in each group, and 30 rats were tested in total.Among the SCI, BMSC-Exo, BMSC-NC-Exo, and BMSC-miR-Exo groups, 6 biological replicative samples were tested in each group, and 24 rats were tested in total.Regarding BMSC transfection and exosome isolation, 3 biological replicative samples were tested.Comparisons were accomplished with one-way ANOVA followed by Tukey's post hoc test.p < 0.05 was considered statistically significant.

BMSC-exo enhanced locomotor performance in SCI rats
Initially, the SCI model was established, then BMSC or BMSC-Exo was injected through the caudal vein.BBB score was evaluated on days 1, 3, 7, 14, 21, and 28.All rats were euthanized on day 28 (Figure 1A).Meanwhile, the BBB score on day 28 was higher in the BMSC group than in the SCI group (p < 0.001) and the BMSC-GW group (p < 0.01); in addition, the BBB score on day 28 was also higher in the BMSC-Exo group than in the SCI group (p < 0.001) (Figure 1B).Results of HE staining were shown in Figure 1C, which revealed that tissue injury and inflammation infiltration were reduced, while neuron numbers were increased in the BMSC group compared to the SCI group and BMSC-GW group.The same trend was found in the BMSC-Exo group versus the SCI group.The rat bone mesenchymal markers, including CD29 + , CD44 + , CD90 + , CD105 + , CD34 + , and CD45 + were detected to ensure that the cell source came from rat bone mesenchymal.It turned out that the levels of CD29 + , CD44 + , CD90 + , and CD105 + were 96.9%, 97.8%, 96.0%, and 98.1%, which were all over 95.0% and indicated that the cell source did come from rat bone mesenchymal.Meanwhile, the levels of CD34 + and CD45 + were 1.26% and 1.16%, which were less than 2.0% and indicated that the cell source did not come from other positions of rats (Supplementary Figure S1).

BMSC-exo attenuated neuronal injury in SCI rats
The TUNEL staining was conducted to detect cell apoptosis in SCI rats on day 28 (Figure 2A).TUNEL positive rate was diminished in the BMSC group versus the SCI group (p < 0.001) and the BMSC-GW group (p < 0.05).At the same time, TUNEL positive rate was also reduced in the BMSC-Exo group versus the SCI group (p < 0.05) (Figure 2B).Subsequently, the apoptotic markers, cleaved-caspase3 and BCL2 were determined by Western blot on day 28 to verify the above finding (Figure 2C).It was found that cleaved-caspase3 expression was decreased in the BMSC group by comparison with the SCI group (p < 0.001) and BMSC-GW group (p < 0.01), as well as reduced in the BMSC-Exo group versus the SCI group (p < 0.001) (Figure 2D).Nevertheless, BCL2 expression exhibited an opposite trend among groups (most p < 0.05), except for no change between the BMSC group and the BMSC-GW group (Figure 2E).
The IF staining was performed to detect neuronal viability in SCI rats on day 28 (Figure 2F).It was discovered that NeuN + cells were raised in the BMSC group versus the SCI group (p < 0.001) and the BMSC-GW group (p < 0.01).Meanwhile, NeuN + cells were also enlarged in the BMSC-Exo group versus the SCI group (p < 0.01) (Figure 2G).

BMSC-exo inhibited microglia M1 polarization-mediated inflammation in SCI rats
The IF staining was performed to detect microglia M1 polarization in SCI rats on day 28 (Figure 3A).It was found that IBA1 + iNOS + cells were diminished in the BMSC group versus the SCI group (p < 0.001) and the BMSC-GW group (p < 0.05).Additionally, IBA1 + iNOS + cells were also reduced in the BMSC-Exo group versus the SCI group (p < 0.001) (Figure 3B).
Regarding the inflammatory cytokines, TNF-α (Figure 3C), IL-1β (Figure 3D), and IL-6 (Figure 3E) in the lesions were all decreased on day 28 in the BMSC group by comparison with the SCI group (all p < 0.01) and the BMSC-GW group (all p < 0.05), as well as diminished in the BMSC-Exo group versus the SCI group (all p < 0.01).

BMSC-exo-derived miR-216a-5p strengthened locomotor performance in SCI rats
The relative miR-216a-5p expression in BMSCs was increased following the transfection with miR-216a-5p mimics versus the NC mimics (p < 0.001) (Figure 4A), which indicated the transfection was successful.After that, the exosomes of transfected BMSCs were isolated.The relative miR-216a-5p expression in BMSC-Exo was upregulated following the transfection with miR-216a-5p mimics versus the NC mimics (p < 0.001) (Figure 4B).The relative miR-216a-5p expression in the spinal cord was increased in the BMSC-miR-Exo group versus the BMSC-NC-Exo group (p < 0.001) (Figure 4C).In terms of locomotor performance, the BBB score on day 28 was higher in the BMSC-miR-Exo group than in the BMSC-NC-Exo group (p < 0.05) (Figure 4D).Results of HE staining were shown in Figure 4E, which suggested that neuronal injury and inflammation infiltration were attenuated in the BMSC-miR-Exo group versus the BMSC-NC-Exo group.

BMSC-exo-derived miR-216a-5p suppressed neuronal injury in SCI rats
By TUNEL staining (Figure 5A), it was found that the TUNEL positive rate on day 28 was reduced in the BMSC-miR-Exo group versus the BMSC-NC-Exo group (p < 0.01) (Figure 5B).Then Western blot was conducted on day 28 to confirm the above finding (Figure 5C), which displayed that cleaved-caspase3 expression was decreased (p < 0.05) (Figure 5D) but BCL2 expression was raised (p < 0.05) (Figure 5E) in the BMSC-miR-Exo group versus the BMSC-NC-Exo group.By IF staining (Figure 5F), it was figured out that NeuN + cells on day 28 only showed an increasing trend in the BMSC-miR-Exo group versus the BMSC-NC-Exo group but did not achieve statistical significance (p > 0.05) (Figure 5G).

Discussion
The therapeutic effect of BMSC-Exo for SCI and the potential underlying mechanisms have been revealed by previous studies (Xin W. et al., 2021;Jia et al., 2021;Sheng et al., 2021).One study finds out that BMSC-Exo improves the function of macrophages by enhancing phagocytosis of myelin debris internalization, which assists in functional recovery after SCI (Sheng et al., 2021).At the same time, a previous study reports that BMSC-Exo preserves the integrity of the blood-spinal cord barrier and enhances the functional recovery after SCI through the tissue inhibitors of matrix metalloproteinase 2/matrix metalloproteinase pathway (Xin W. et al., 2021).In the current study, several interesting findings were discovered.Firstly, BMSC-Exo improved locomotor performance in SCI rats.A possible reason would be that BMSC-Exo might enhance locomotor performance by inhibiting neuronal injury and the microglia M1 polarization-mediated inflammation (Liu et al., 2020;Xu et al., 2020).Secondly, it was found that BMSC-Exo hindered neuronal injury in SCI rats.A reason behind this would be that BMSC-Exo might activate the c-Jun N-terminal kinase 1/c-Jun pathway, endoplasmic reticulum to nucleus signaling 1 (Ern1), etc. to reduce cell apoptosis and improve neuronal cell viability in SCI rats (Li R. et al., 2020;Jiang et al., 2023).Thirdly, BMSC-Exo inhibited microglia M1 polarizationmediated inflammation in SCI rats.This finding was partly in line with a previous study, in which BMSC-Exo ameliorates injury by regulating microglial polarization to inhibit NLRP3 inflammasomemediated inflammation (Liu X. et al., 2021).Fourthly, it was also discovered that BMSC-Exo inhibited the TLR4/NF-κB pathway in SCI rats.This finding was partly in line with previous studies (Liu et al., 2020;Jiang and Zhang, 2021).
MiR-216a-5p/TLR4/NF-κB pathway participates in cell apoptosis and inflammation in various neurological disorders (Ouyang et al., 2022;Huang et al., 2023).For instance, one previous study reports that miR-216a-5p targets TLR4/NF-κB pathway to inhibit inflammation and the apoptosis of chondrocytes (Huang et al., 2023).Meanwhile, another study indicates that miR-216a-5p directly hinders the TLR4/ NF-κB pathway to prevent intestinal epithelial barrier dysfunction (Ouyang et al., 2022).Partly in line with these previous studies (Ouyang et al., 2022;Huang et al., 2023), the present study also figured out that BMSC-Exo-derived miR-216a-5p inhibited the TLR4/ NF-κB pathway in SCI rats.According to previous studies, miR-216a-5p could directly target TLR4, thereby hindering the subsequent activation of NF-κB (Ouyang et al., 2022;Huang et al., 2023).Therefore, it was speculated that the inhibition of miR-216a-5p on neuronal injury and inflammation was mediated by preventing the TLR4/NF-κB pathway in SCI rats.However, this speculation needed further experiments to validate.It should be clarified that we applied p-p65/p65 to reflect the activation or inhibition of the NF-κB pathway, which was in line with previous studies (Wang et al., 2019;Yang et al., 2020).The reason for detecting p-p65 was as follows: p65 was a member of the NF-κB pathway, which exerted biological function by being phosphorylation, and p-p65 was the phosphorylated form of p65; thus, the change of p-p65/p65 could reflect the activation or inactivation of the NF-κB pathway.In the current study, it was found that p-p65/p65 was decreased by BMSC-Exo-derived miR-216a-5p, which indicated that BMSC-Exo-derived miR-216a-5p inactivated the NF-κB pathway.In terms of the detection of TLR4/β-ACTIN and myD88/β-ACTIN, we used β-ACTIN as the internal reference to reflect the expression of TLR4 and myD88 proteins, which was in line with previous studies (Liu et al., 2020;Rong et al., 2021;Gu et al., 2023).Overall, this study discovered that BMSC-Exo-derived miR-216a-5p reduced the TLR4/β-ACTIN, myD88/β-ACTIN, and p-p65/p65 in SCI rats, which indicated that BMSC-Exo-derived miR-216a-5p inactivated the TLR4/NF-κB pathway in SCI rats.
In summary, BMSC-Exo-derived miR-216a-5p improves functional recovery by attenuating neuronal injury and microgliamediated inflammation in SCI, which may benefit from its inactivation of the TLR4/NF-κB pathway.However, more shreds of evidence are required to validate our findings.

FIGURE 2
FIGURE 2 Effect of BMSC-Exo on neuronal injury in SCI rats.Representative images of cell apoptosis by TUNEL staining (A), comparison of TUNEL positive rate (B), detection of cleaved-caspase3 and BCL2 by Western blot (C), comparison of cleaved-caspase3 (D) and BCL2 (E), representative images of NeuN + cells by IF staining (F), and comparison of NeuN + cells (G), among the sham, SCI, BMSC, BMSC-GW, and BMSC-Exo groups.Six biological replicative samples were tested in each group, and 30 rats were tested in total among the sham, SCI, BMSC, BMSC-GW, and BMSC-Exo groups.

FIGURE 5
FIGURE 5 Effect of BMSC-Exo-derived miR-216a-5p on neuronal injury in SCI rats.Representative images of cell apoptosis by TUNEL staining (A), comparison of TUNEL positive rate (B), detection of cleaved-caspase3 and BCL2 by Western blot (C), comparison of cleaved-caspase3 (D) and BCL2 (E), representative images of NeuN + cells by IF staining (F), and comparison of NeuN + cells (G), among the SCI, BMSC-Exo, BMSC-NC-Exo, and BMSC-miR-Exo groups.Six biological replicative samples were tested in each group, and 24 rats were tested in total among the SCI, BMSC-Exo, BMSC-NC-Exo, and BMSC-miR-Exo groups.