Human Bone Marrow Mesenchymal Stromal Cell-Derived Extracellular Vesicles Promote Proliferation of Degenerated Nucleus Pulposus Cells and the Synthesis of Extracellular Matrix Through the SOX4/Wnt/β-Catenin Axis

Objective: Intervertebral disk degeneration (IDD) is a major cause of pain in the back, neck, and radiculus. Mesenchymal stem cells (MSCs)-derived extracellular vesicles (EVs) are therapeutic in musculoskeletal degenerative diseases such as IDD. This study explored the effect and functional mechanism of human bone MSCs (hBMSCs)-derived EVs in proliferation and apoptosis of degenerated nucleus pulposus cells (DNPCs) and extracellular matrix (ECM) synthesis. Methods: Extracellular vesicles were isolated from hBMSCs and identified. DNPCs were induced by TNF-α. EVs were incubated with DNPCs for 24h. Internalization of EVs by DNPCs, DNPCs proliferation, apoptosis, and expressions of ECM synthetic genes, degrading genes and miR-129-5p were assessed. Downstream target genes of miR-129-5p were predicted. Target relation between miR-129-5p and SRY-box transcription factor 4 (SOX4) was verified. DNPCs proliferation, apoptosis, and ECM synthesis were measured after treatment with EVs and miR-129-5p inhibitor or SOX4 overexpression. Expressions of SOX4 and Wnt/β-catenin pathway-related proteins were determined. Results: hBMSC-EVs promoted DNPCs proliferation, inhibited apoptosis, increased expressions of ECM synthetic genes, and reduced expressions of ECM degrading genes. hBMSC-EVs carried miR-129-5p into DNPCs. Silencing miR-129-5p in EVs partially inverted the effect of EVs on DNPCs proliferation and ECM synthesis. miR-129-5p targeted SOX4. SOX4 overexpression annulled the effect of EVs on DNPCs proliferation and ECM synthesis. Expressions of Wnt1 and β-catenin were decreased in EVs-treated DNPCs, while silencing miR-129-5p in EVs promoted expressions of Wnt1 and β-catenin. Conclusion: hBMSC-EVs promoted DNPCs proliferation and ECM synthesis by carrying miR-129-5p into DNPCs to target SOX4 and deactivating the Wnt/β-catenin axis.


INTRODUCTION
Intervertebral disk degeneration (IDD) is a common chronic disease leading to a series of clinical symptoms including cervical spondylitis and intervertebral disk herniation (Chan and Gantenbein-Ritter, 2012;Kos et al., 2019). A recent study has revealed that IDD will be initiated under various physiological and pathological conditions and can be influenced by a range of factors including genetics, cell senescence, mechanical load, extracellular matrix (ECM) degradation, inflammation, and apoptosis (Stokes and Iatridis, 2004;Galbusera et al., 2014). IDD is characterized by ECM loss and apoptosis of nucleus pulposus cells (NPCs; Kadow et al., 2015). However, the exact mechanism of IDD remains unknown. It is difficult to reverse IDD progression once occurred, and there is no effective method to prevent the degenerative changes in IDD (Sakai and Schol, 2017). Traditional treatment methods of IDD such as conservative treatment and surgical intervention can only mitigate low back pain, worse yet, often cause recurrence on the same or adjacent sites and the degeneration of intervertebral disk is not eliminated (Richardson et al., 2007). Therefore, it is warranted to explore more effective therapeutic strategies for IDD.
Bone marrow mesenchymal stromal cells (BMSCs) are the main source of mesenchymal stromal cells (MSCs) which secrete a variety of cytokines and growth factors (Strioga et al., 2012). The therapeutic effect of BMSCs on the occurrence of IDD has attracted researcher's attention recently due to their ability to suppress NPC apoptosis (Yang et al., 2010;Richardson et al., 2016). Extracellular vesicles are a group of cell-derived membranous structures which comprises exosomes and microvesicles (van Niel et al., 2018). EVs mediate intercellular communication by carrying proteins, RNA species, lipids, and other biomacromolecules from parent cells to target cells (El Andaloussi et al., 2013). BMSCs have been reported to transfer EVs to protect against tissue or organ damage (Eirin et al., 2017;Wang et al., 2018b). BMSC-EVs repress NPC apoptosis induced by inflammatory cytokines (Lu et al., 2017;Cheng et al., 2018) but the effect and mechanism of BMSC-EVs in IDD remain unclear. microRNAs (miRNAs) are a kind of small non-coding RNAs regulating target gene expressions and have been reported to participate in IDD development (Wang et al., 2015;Lan et al., 2021). Currently, the study about BMSC-EVs affecting NPC function via miRNAs is in its infancy. Wen et al. (2021) have reported that BMSC-EVs promote IDD repair by transferring miR-199a. Zhu et al. (2020) have shown that exosomal miR-532-5p released from BMSCs delays IDD by targeting RASSF5. Existing studies have revealed the association of miR-129-5p with IDD and the functionality of miR-129-5p overexpression to inhibit NPC apoptosis and degrade IDD risk by modulating FADD/BMP-2 Li et al., 2020). SRY-box transcription factor 4 (SOX4) has been documented to avert the effect of miR-499a-5p on NPC proliferation and ECM synthesis . The Wnt/βcatenin pathway is associated with NPC senescence and apoptosis and ECM degradation (Chen et al., 2017). However, whether BMSC-EVs regulate NPCs in IDD through the miR-129-5p/ SOX4/Wnt/β-catenin axis has not been reported. This study investigated the functional mechanism of BMSC-EVs on degenerated NPCs (DNPCs) proliferation and apoptosis and ECM synthesis with the purpose to provide new reference for IDD management.

Culture and Identification of Human BMSCs
Human BMSCs (hBMSCs) purchased from Cyagen Biosciences (Guangzhou, China) were cultured in Dulbecco's modified Eagle's medium (DMEM)/F12 medium (Gibco, Grand Island, NY, United States) in an incubator at 37°C with 5% CO 2 . Once reaching 80% confluence, the immunophenotypes of hBMSCs were identified. Specifically, about 2 × 10 5 hBMSCs were collected and centrifuged at 150 g for 5 min with the supernatant removed. Next, the cells were resuspended with phosphate buffer saline (PBS), and 100 μl suspension was transferred to an Eppendorf (EP) tube. Fluorescein isothiocyanate (FITC)-labeled monoclonal antibodies (CD29, CD34, CD44, and CD90; BD Bioscience, Franklin Lakes, NJ, United States) were added into cell suspension and reacted in the dark for 30-60 min (with IgG as an isotype control). After three times of PBS washes, cells were suspended in 450 μl PBS and detected using flow cytometry.
Next, the adipogenic and osteogenic abilities of hBMSCs were determined. For adipogenic differentiation, 2 × 10 4 hBMSCs were seeded into 6-well plates and cultured in adipogenic differentiation basic culture medium A (Cyagen Biosciences) for 3 days and in adipogenic differentiation basic culture medium B (Cyagen Biosciences) for 1 day. The procedures were repeated three times. The hBMSCs were fixed with 4% paraformaldehyde, stained with Oil red O (Sigma-Aldrich, St. Louis, MO, United States) for 30 min, and observed under an inverted microscope (Olympus, Tokyo, Japan). For osteogenic differentiation, 3 × 10 4 hBMSCs were seeded into 6-well plates and cultured in osteogenic differentiation medium (Cyagen Biosciences) for 4 weeks. Subsequently, hBMSCs were fixed with 4% paraformaldehyde and stained with 1% Alizarin red Frontiers in Physiology | www.frontiersin.org 3 October 2021 | Volume 12 | Article 723220 (Sigma-Aldrich) for 5 min. Mineralized nodule formation was observed under the inverted microscope (Olympus).

Isolation and Identification of hBMSC-EVs
The supernatant was removed when the confluence of hBMSCs reached around 80%. Subsequently, 10% EV-free fetal bovine serum (FBS; EVs in FBS were removed by centrifugation at 100,000 g for 10 min) was added and cultured in the CO 2 incubator for 48 h at 37°C. EVs were extracted using ultrahigh-speed centrifugation . When the medium was changed, the supernatant was collected and centrifuged for 10 min at 4°C at 500 g, followed by another centrifugation for 20 min at 12,000 g. The supernatant was centrifuged at ultra-high speed for 2 h at 100,000 g using a 0.22 μM filtration membrane. The precipitation was resuspended with PBS and then centrifuged at ultra-high speed for 2 h, followed by resuspension with PBS and storage at −80°C. The isolated EVs were identified after the morphology of EVs was observed under a transmission electron microscope (TEM), and the size distribution of EVs was analyzed using nanoparticle tracking analysis (NTA), and the expression of CD9 and CD63 was detected on the surface of BMSC-EVs via Western blot. Additionally, hBMSC supernatant added with EV inhibitor GW4869 (20 μg/ml conditioned medium; Sigma-Aldrich) was used as a control (GW). The identified EVs were lysed with the lysis buffer. Total protein content was determined using the bicinchoninic acid (BCA) kit (23225, Thermo Fisher Scientific, Shanghai, China). EVs used in this study were divided into six groups: GW group (hBMSC supernatant added with GW4869), EVs group, EVs + RNase group (EVs added with RNase), EVs + RNase + SDS group (EVs added with RNase and lysis buffer SDS), EVs-NC group (EVs isolated from hBMSCs transfected with negative control (NC) of miR-129-5p inhibitor for 24 h), and EVs + inhi group (EVs isolated from hBMSCs transfected with miR-129-5p inhibitor for 24 h). hBMSC transfection was conducted in strict conformity with the manufacturer's instructions. Briefly, miR-129-5p inhibitor or miR-129-5p inhibitor NC (GenePharma, Shanghai, China) was transfected into hBMSCs using Lipofectamine 2000 (Invitrogen, Carlsbad, CA, United States). Following experiments were performed after 24 h.

Immunocytochemistry
Nucleus pulposus cells and DNPCs climbing sheets were fixed with 4% paraformaldehyde (Beyotime, Shanghai, China) for 30 min, rinsed with PBS and immersed with 20% H 2 O 2 methanol solution for 30 min at room temperature. After blockade with 1:10 horse serum (Beyotime) for 20 min at room temperature, the serum was removed, and cells were added with Collagen II (Col II) primary antibody (1:500, ab34712; Abcam, Cambridge, MA, United States) and incubated at 4°C overnight. Cells were then added with secondary antibody IgG (1:2,000, ab205718; Abcam) and incubated for 1 h at room temperature, followed by 20-min incubation with SABC reagent (Maxim Corp, Fuzhou, China) in the 37°C constant temperature water bath. After color development using 2,4-diaminobutyric acid (DAB), cells were counterstained with hematoxylin (Beyotime) and observed under a microscope (TS100; Nikon, Kanagawa, Japan).

Reverse Transcription-Quantitative Polymerase Chain Reaction (RT-qPCR)
Total RNA was extracted from cells using a TRIzol kit (Invitrogen) as per the instructions. cDNA template was synthesized from total RNA using Transcriptor First Strand cDNA Synthesis kit (TaKaRa Biotechnology, Dalian, China). TaqMan sequence and probe were obtained from Takara (Tokyo, Japan). Quantitative PCR was conducted using SYBR Green II (TaKaRa) with ABI PRISM 7900 Sequential Detection System under the reaction condition of pre-denaturation at 95°C for 10 min and 40 cycles of denaturation at 95°C for 10 s, annealing at 60°C for 20 s and extension at 72°C for 34 s with U6 as an internal control. Data were analyzed using the 2 −ΔΔCt method. Primer sequences (Sangon Biotech, Shanghai, China) are given in Table 1.

CCK-8 Assay
CCK-8 assay was adopted to assess cell proliferation. Cells were seeded into 96-well plates (1,500 cells/well) after transfection, and added with CCK-8 solution (Dojindo Molecular Technologies, Kumamoto, Japan) at baseline (day 0) and on day 1, 2, 3, 4 and incubated with 5% CO 2 for 2 h at 37°C. The optical density value of each well at 450 nm was detected using a microplate reader (Tecan, Mannedorf, Switzerland).

Flow Cytometry
Degenerated nucleus pulposus cells at the third passage under good growth conditions were detached with trypsin. The cell concentration was adjusted to 1 × 10 6 cells/ml. After washing with PBS, 1 ml cells were mixed with 100 ml mouse anti-human CD24-phycoerythrin (PE) monoclonal antibody with the isotype control set. Cells were incubated for 30 min at room temperature under conditions devoid of light, washed with 2 ml PBS and then counted using flow cytometry. The positive rate of CD24 in DNPCs was documented based on the fluorescence intensity with the isotype control as the negative cell group.

Statistical Analysis
SPSS 21.0 (IBM Corp., Armonk, NY, United States) and GraphPad Prism 8.01 (GraphPad Software Inc., San Diego, CA, United States) were used for data analysis and graph plotting. Normality and homogeneity of variance tests were performed first to verify that the data were consistent with normal distribution and homogeneity of variance. Data were presented as mean ± standard error. Pairwise comparisons were analyzed using the independent sample t test, while comparisons among groups were analyzed using one-way analysis of variance (ANOVA), followed by Tukey's multiple comparisons test. p < 0.05 was indicative of statistical significance.

Identification of hBMSCs and EVs
The purchased hBMSCs were initially purified for three generations to obtain relatively uniform and active cells, which were in spindle shapes and arranged in a whirlpool pattern ( Figure 1A). Flow cytometry showed that expressions of CD29, CD44, and CD90 in hBMSCs were positive, while CD34 expression was negative ( Figure 1B). Moreover, Oil red O staining after adipogenic differentiation induction of hBMSCs verified lipid deposition, which reflected the lipid differentiation ability of hBMSCs ( Figure 1C). hBMSCs were stained red by Alizarin red after osteogenic differentiation induction ( Figure 1D). Subsequently, EVs were isolated from hBMSCs with ultrahigh-speed centrifugation. TEM showed that EVs were in round or oval shape with non-uniform granules and membranous structure ( Figure 1E). NTA was conducted to further clarify the size distribution, and the result showed that EVs were about 100 nm in size with the concentration of 5.8 × 10 7 particles/ ml ( Figure 1F). Western blot showed that CD9 and CD63 were enriched in isolated EVs, while Calnexin was not expressed (Figure 1G). These results indicated that hBMSC-EVs were successfully obtained.

hBMSC-EVs Promoted DNPCs Proliferation and ECM Synthesis
To study the effect of EVs on DNPCs, DNPCs were firstly obtained through TNF-α induction. NPCs observed under the microscope exhibited clear morphology of short spindle and polygonal shape, while DNPCs were in spindle and irregular shape (Figure 2A). Col II is the most abundant collagen in nucleus pulposus tissues. Immunocytochemistry showed that the cytoplasm of NPCs was stained dark tan, while the cytoplasm of DNPCs were stained light yellow, which was in consistency with the degeneration of Col II synthesis ( Figure 2B). These results suggested that correct and usable DNPCs were obtained.
Next, Dil-labeled EVs were incubated with DNPCs. After 24 h, immunofluorescence results demonstrated red fluorescence in DNPCs (Figure 2C), suggesting that EVs were internalized by DNPCs. Furthermore, DNPCs were treated with EVs and hBMSC supernatant added with GW4869 (GW group) for 24 h. CCK-8 assay showed that relative to the blank group and GW group, the proliferative ability of DNPCs was enhanced after EV treatment (all p < 0.05, Figure 2D). Flow cytometry showed that DNPCs apoptosis was decreased after EV treatment compared with the blank group and GW group (all p < 0.01, Figure 2E). In addition, it has been reported that excessive ECM destruction, loss of Col II and Aggrecan in particular, contributes significantly to the occurrence and development of IDD (Sivan et al., 2014). Western blot showed that expressions of Aggrecan and Col II were increased, while expressions of MMP3 and ADAMTS5 were decreased in DNPCs after EV treatment (all p < 0.01, Figure 2F), suggesting that EVs promoted ECM synthesis in DNPCs. Above results demonstrated that hBMSC-EVs promoted DNPCs proliferation and ECM synthesis and inhibited apoptosis.
EVs Carried miR-129-5p Into DNPCs miR-129-5p is associated with IDD and upregulation of miR-129-5p inhibits NPC apoptosis in IDD and reduces IDD risk by regulating FADD/BMP-2 Li et al., 2020). RT-qPCR showed that miR-129-5p expression was decreased in DNPCs relative to that in NPCs (p < 0.01), while Frontiers in Physiology | www.frontiersin.org 6 October 2021 | Volume 12 | Article 723220 miR-129-5p expression was elevated in DNPCs after EV treatment (Figures 3A,B), and miR-129-5p expression in hBMSC-EVs was higher than that in hBMSC supernatant treated with GW4869 (p < 0.01, Figure 3C). Herein, we speculated that hBMSC-EVs affect DNPCs proliferation and apoptosis and ECM synthesis by carrying miR-129-5p. After treating EVs with RNase, miR-129-5p expression in EVs showed no significant difference, while miR-129-5p expression was considerably decreased after treatment with lysis buffer SDS and RNase (all p < 0.01, Figure 3C), which elicited that miR-129-5p was encapsulated in EVs. Then, hBMSCs were transfected with miR-129-5p inhibitor and then EVs were isolated. miR-129-5p was downregulated in hBMSCs and EVs after transfection (p < 0.01, Figures 3D,E). Subsequently, DNPCs were treated with EVs-inhi, and the result showed that miR-129-5p expression was reduced in the DNPCs + EVs-inhi group relative to that (F) expressions of ECM synthetic genes Aggrecan and Col II and ECM degrading genes MMP3 and ADAMTS5 detected by Western blot. Cell experiment was repeated three times. Data were presented as mean ± standard error. Comparisons among groups were analyzed using one-way ANOVA, followed by Tukey's multiple comparisons test. *p < 0.05, **p < 0.01.

Silencing miR-129-5p Inverted the Promoting Effect of EVs on DNPCs Proliferation and ECM Synthesis
To explore whether hBMSC-EVs affect DNPCs proliferation and apoptosis and ECM synthesis by transporting miR-129-5p, DNPCs were incubated with EVs in different groups for 24 h. The results showed that DNPCs proliferation was reduced, apoptosis was increased, and ECM synthesis was decreased after silencing miR-129-5p in EVs (all p < 0.05, Figures 4A-C).
These results elicited that silencing miR-129-5p in EVs inverted the promoting effect of EVs on DNPCs proliferation and ECM synthesis, which further clarified that EVs promoted DNPCs proliferation and ECM synthesis by carrying miR-129-5p into DNPCs.

miR-129-5p Targeted SOX4
To investigate the downstream regulatory mechanism of EVs in promoting DNPCs proliferation and ECM synthesis via miR-129-5p, eight downstream target genes of miR-129-5p were identified by searching on the StarBase, TargetScan, RNAInter, and miRDB database ( Figure 5A). Among the eight downstream target genes, SOX4 was reported to be upregulated in degenerated NP tissues relative to healthy nucleus pulposus tissues and have the ability to promote NPC apoptosis and inhibit ECM synthesis . Western blot showed that SOX4 expression in DNPCs was higher than that in NPCs (p < 0.01, Figure 5B). Moreover, the target relationship between miR-129-5p and SOX4 3′UTR sequence was predicted on the StarBase (Figure 5C), and dual-luciferase assay verified the target relationship between miR-129-5p and SOX4 (p < 0.01, Figure 5D). Besides, Western blot showed that SOX4 expression was decreased in DNPCs after EV treatment and increased after were analyzed using independent sample t test, while data in panels (B-F) were analyzed using one-way ANOVA, followed by Tukey's multiple comparisons test. **p < 0.01.

SOX4 Overexpression Partially Inverted the Promoting Effect of EVs on DNPCs Proliferation and ECM Synthesis
Degenerated nucleus pulposus cells were transfected with SOX4 overexpression plasmid along with EV treatment. Western blot showed that SOX4 expression in the DNPCs-oe-SOX4 + EVs group was increased compared with that in the DNPCsoe-NC + EVs group (p < 0.01, Figure 6A), indicating that SOX4 was successfully overexpressed. The promoting effect of EVs on DNPCs proliferation and ECM synthesis was inverted after overexpressing SOX4 (all p < 0.05, Figures 6B-D). These results demonstrated that hBMSC-EVs downregulated SOX4 and thus promoted DNPCs proliferation and ECM synthesis and inhibited DNPCs apoptosis by carrying miR-129-5p into DNPCs.

DISCUSSION
Intervertebral disk degeneration remains a chief cause of morbidity and disability and medical costs around the world (Kos et al., 2019). MSC-EVs are regarded as a novel therapeutic approach for IDD (Liao et al., 2021). The present study elicited that hBMSC-EVs promoted DNPCs proliferation and ECM synthesis by carrying miR-129-5p into DNPCs to target SOX4 and inhibit the activation of the Wnt/β-catenin axis. After identification of hBMSCs and EVs, DNPCs were induced by TNF-α, and Dil-labeled EVs were cocultured with DNPCs. Our result showed that EVs were internalized by DNPCs after 24-h incubation. MSC-EVs are reported to inhibit NPC apoptosis, promote proliferation, and participate in matrix organization and distribution of ECM components (Rilla et al., 2019;Piazza et al., 2020). Our results found that the proliferative A C B FIGURE 4 | Silencing miR-129-5p inverted the promoting effect of EVs on DNPCs proliferation and ECM synthesis. (A) DNPCs proliferation detected by CCK-8 assay; (B) DNPCs apoptosis detected by flow cytometry; (C) expressions of ECM synthetic genes Aggrecan and Col II and ECM degrading genes MMP3 and ADAMTS5 detected by Western blot. Cell experiment was repeated three times. Data were presented as mean ± standard error. Comparisons among groups were analyzed using one-way ANOVA, followed by Tukey's multiple comparisons test. *p < 0.05, **p < 0.01.
Frontiers in Physiology | www.frontiersin.org 9 October 2021 | Volume 12 | Article 723220 ability of EVs-treated DNPCs was enhanced, while DNPCs apoptosis was decreased. Col II is a main component of gelatinous tissues of nucleus pulposus (Chen et al., 2016). Destruction of Col II and Aggrecan is an important feature of IDD . MMP3 plays a central role in IDD and ADAMTS5 is the primary aggrecanase mediating IDD (Ngo et al., 2017;Zhao et al., 2020). Our results showed that expressions of Col II and Aggrecan were increased, while expressions of MMP3 and ADAMTS5 were decreased in DNPCs after EV treatment. MSC-EVs have been identified to modulate NPC proliferation and apoptosis (Cheng et al., 2018). EVs carrying miRNAs impact ECM degradation in osteoarthritis (Withrow et al., 2016). Consistently, our results demonstrated that hBMSC-EVs inhibited apoptosis and promoted proliferation of DNPCs and ECM synthesis. miR-129-5p was reported to be involved in IDD progression and was differentially expressed in NPCs of IDD patients . Our result showed that miR-129-5p expression was decreased in DNPCs and elevated after EV treatment. Additionally, miR-129-5p expression in hBMSC-EVs was higher than that in GW-4869-treated hBMSC supernatant. To confirm whether hBMSC-EVs affect DNPCs proliferation and apoptosis and ECM synthesis via the delivery of miR-129-5p, EVs were treated with RNase and lysis buffer SDS. The results showed predicted on the StarBase database; (D) target relationship between miR-129-5p and SOX4 verified by dual-luciferase assay; (E) SOX4 expression in DNPCs detected by Western blot. Cell experiment was repeated three times. Data were presented as mean ± standard error. Data in panels (B,D) were analyzed using independent t test, while data in panel (E) were analyzed using one-way ANOVA, followed by Tukey's multiple comparisons test. **p < 0.01.
that miR-129-5p expression in EVs showed no significant difference after RNase treatment but was decreased in EVs treated with RNase and lysis buffer SDS, suggesting miR-129-5p was encapsulated by EVs. Subsequently, EVs were isolated from hBMSCs transfected with miR-129-5p inhibitor. The result showed that miR-129-5p expression was decreased significantly in transfected hBMSCs, isolated EVs, and DNPCs treated with EVs-inhi. Human synovial MSCs-derived exosomes relieved IL-1β-induced osteoarthritis by carrying miR-129-5p into chondrocytes (Qiu et al., 2021). Similarly, our results showed that hBMSC-EVs carried miR-129-5p into DNPCs. Furthermore, miR-129-5p knockdown promotes NPC apoptosis and causes collagen I synthesis (Chen et al., 2018;Li et al., 2020). Our results showed that EVs-treated DNPCs proliferation was decreased, and apoptosis and ECM synthesis decreased after silencing miR-129-5p in EVs. Zhao et al. have reported that miR-129-5p regulates cell apoptosis in nucleus pulposus in IDD (He et al., 2015). miR-129-5p serves as a negative regulator (D) expressions of ECM synthetic genes Aggrecan and Col II and ECM degrading genes MMP3 and ADAMTS5 detected by Western blot. Cell experiment was repeated three times. Data were presented as mean ± standard error. Comparisons among groups were analyzed using one-way ANOVA, followed by Tukey's multiple comparisons test. *p < 0.05, **p < 0.01.
of collagen I in systemic sclerosis (Nakashima et al., 2012). Collectively, EVs promoted DNPCs proliferation and ECM synthesis by carrying miR-129-5p into DNPCs. Similar to miR-129-5p, miR-129-3p enhances the viability and restrains the apoptosis of chondrocytes in osteoarthritis . However, the involvement of miR-129-3p in IDD has not been reported. In addition, miR-129-3p and miR-129-5p exert similar effects in diseases via different mechanisms. For example, in the amelioration of diabetes, miR-129-3p decreases the apoptosis and recruitment of neutrophils by regulating the translation of Casp6 and Ccr2 (Umehara et al., 2019), whereas miR-129-5p inhibits ECM degradation by inhibiting the expression of Sp1-mediated MMP9 (Wang et al., 2018a).
To explore the downstream mechanism of EVs in DNPCs proliferation and ECM synthesis via miR-129-5p, downstream target genes of miR-129-5p were investigated. Among the screening results, SOX4 was reported to contribute to ECM and IDD progression (Zhang et al., 2021). Our results showed that SOX4 expression was increased in DNPCs. Target relationship between miR-129-5p and SOX4 was verified by dual-luciferase assay. Furthermore, SOX4 expression was decreased in EVs-treated DNPCs and increased after silencing miR-129-5p in EVs. Taken together, miR-129-5p targeted SOX4. Moreover, SOX4 overexpression plasmids were transfected into EVs-treated DNPCs to explore its effect. Upon SOX4 overexpression, the promoting effects of EVs on DNPCs proliferation and ECM synthesis were averted. According to Sun et al. ( 2019) SOX4 downregulation attenuates TNF-α-induced NPC apoptosis and balances anabolism and catabolism of ECM. SOX4 overexpression reverses the effect of circITCH depletion on enhancing NPC proliferation and expressions of Aggrecan and Col II (Zhang et al., 2021). These results demonstrated that hBMSC-EVs promoted DNPCs proliferation and ECM synthesis by carrying miR-129-5p into DNPCs to target SOX4.
In conclusion, this study elicited that hBMSC-EVs promoted DNPCs proliferation and ECM synthesis by carrying miR-129-5p into DNPCs to target SOX4 and inhibit the activation of the Wnt/β-catenin pathway. However, this study has its limitations. Firstly, the effect of hBMSC-EVs on DNPCs proliferation and ECM synthesis was verified through in vitro experiment only. Animal experiment and clinical validation are absent. Secondly, A B FIGURE 7 | Extracellular vesicles inhibited the activation of Wnt/β-catenin pathway via the miR-129-5p/SOX4 axis. (A) Expressions of Wnt1 and β-catenin in NPCs and DNPCs detected by Western blot; (B) expressions of Wnt1 and β-catenin in DNPCs in different treatment groups. Cell experiment was repeated three times. Data were presented as mean ± standard error. Data in panel (A) were analyzed using independent t test and data in panel (B) were analyzed using one-way ANOVA, followed by Tukey's multiple comparisons test. **p < 0.01. the age factor was not considered. The occurrence of IDD is closely associated with age, and age could affect the functioning of NPCs (Lehmann et al., 2014;Tang et al., 2019;Wang et al., 2020). In addition, the secretion of miRNAs in BMSCs-derived exosomes changes with aging (Davis et al., 2017;Xun et al., 2021). Hence, animal experiment and clinical trials are to be performed and human NPCs and hBMSCs from patients of different ages are to be used in future studies. Besides, upstream long non-coding RNAs of miR-129-5p or other miRNAs in EVs and other downstream target genes and pathways related to miR-129-5p should be investigated in future studies.

DATA AVAILABILITY STATEMENT
The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

AUTHOR CONTRIBUTIONS
HW contributed to the study concepts, study design, and definition of intellectual content. FL and WB contributed to the literature research. HW and FL contributed to the manuscript preparation. JZ contributed to the manuscript editing and review. WB and JZ contributed to the experimental studies and data acquisition. HW and GZ contributed to the data analysis and statistical analysis. All authors contributed to the article and approved the submitted version.

FUNDING
This study was supported by grant from the Shaanxi Provincial Key Research and Development Program (No. 2020SF-085).