Abstract

   Chronic low back pain (LBP) caused by intervertebral disc (IVD)
   degradation is a serious socioeconomic burden that can cause severe
   disabilities. Addressing the underlying pathogenic mechanisms of IVD
   degeneration may inspire novel therapeutic strategy for LBP. Herein,
   hypoxic preconditioning improves both the biological function of MSCs
   in hostile microenvironments and enhances the production of small
   extracellular vesicles (sEVs) with desirable therapeutic functions. In
   vitro results reveal that hypoxic preconditional engineering sEVs
   (HP‐sEVs) alleviate the inflammatory microenvironments of IVD
   degradation, enhance the proliferation of nucleus pulposus (NP) cells,
   and promote proteoglycan synthesis and collagen formation.
   Transcriptomic sequencing reveales the excellent therapeutic effects of
   HP‐sEVs in promoting extracellular matrix regeneration through the
   delivery of microRNA(miR)‐7‐5p, which further suppresses p65 production
   and thus the inhibition of Cxcl2 production. Moreover, in vivo results
   further confirm the robust therapeutic role of HP‐sEVs in promoting IVD
   regeneration through the same mechanism mediated by miR‐7‐5p delivery.
   In conclusion, this study provides a novel therapeutic strategy for
   treating IVD degradation and is thus valuable for understanding the
   mechanism‐of‐action of HP‐sEVs in IVD regeneration associated with
   chronic lower back pain.

   Keywords: hypoxic preconditioning, intervertebral disc degradation,
   miRNA, small extracellular vesicles
     __________________________________________________________________

   Hypoxic preconditional engineering small extracellular vesicles
   (HP‐sEVs) are injected into the intervertebral space. HP‐sEVs perform
   an anti‐inflammatory effect, promote the proliferation of nucleus
   pulposus (NP) cells, and promote proteoglycan synthesis, resulting in
   extracellular matrix regeneration. Successful intervertebral disc (IVD)
   regeneration is achieved via administering HP‐sEVs and this finding
   provides a promising method for IVD degradation treatment.

   graphic file with name ADVS-10-2304722-g008.jpg

1. Introduction

   Low back pain (LBP) is often caused by intervertebral disc (IVD) injury
   and degradation which can lead to disability and deteriorated quality
   of patients’ life.^[ [34]^1 , [35]^2 ^] IVD is a hydrated
   fibro‐cartilaginous tissue, composed of gel‐like nucleus pulposus (NP),
   annulus fibrosus (AF), and cartilage endplate and plays a crucial role
   in supporting spine stability and mobility. Specifically,
   proteoglycan‐rich gelatinous NP are rich in hydrated sulfated
   glycosaminoglycans and collagen type II, providing mechanical support
   and transmitting the external forces to AF.^[ [36]^3 , [37]^4 ^]
   Alterations in the biochemical and mechanical microenvironment of NP or
   AF are closely related to IVD degeneration as well as LBP.

   Inflammatory responses triggered by mechanical loading, trauma, and
   aging are some of the most critical mechanisms that are responsible for
   the occurrence and progression of IVD degradation.^[ [38]^5 ^] The
   activation of inflammatory cytokines, including tumor necrosis
   factor‐alpha (TNF‐α), and interleukin‐1 (IL‐1), can result in a
   significant reduction of NP cell density and the homeostasis imbalance,
   which ultimately exacerbates extracellular matrix (ECM) degradation and
   hinder collagen synthesis.^[ [39]^5 , [40]^6 ^] Notably, ECM
   degradation may disturb the function of NP by inducing abnormal
   dehydration, compromised force dissipation, and impeded nutrient
   transportation, thereby accelerating the pathogenic process of IVD
   degradation.^[ [41]^7 , [42]^8 ^] Current treatments available for LBP
   have focused on symptomatic relief via nonsteroidal medication or
   invasive procedures, such as spinal stabilization
   and disc replacement.^[ [43]^5 ^] However, these therapeutic strategies
   do not fully address the underlying pathogenic mechanisms of IVD
   degeneration, limiting their efficacy in the restoration
   of the biological and mechanical functions of IVD.^[ [44]^9 ^]
   Therefore, new strategies to inhibit inflammatory microenvironment, and
   simultaneously regenerate proteoglycan and collagen type II are
   urgently required to treat IVD degradation.

   Recent studies have demonstrated that local, regional injection of
   anti‐inflammatory small molecules, biologics, or stem cells can
   mitigate inflammation in IVD degeneration.^[ [45]^10 ^] Among various
   therapeutic approaches, administering mesenchymal stem cells (MSCs) is
   particularly promising due to its diverse secretome that targets
   heterogeneous inflammation pathways.^[ [46]^11 , [47]^12 ^] However,
   survival of injected MSCs remains a major challenge due to the harsh
   microenvironment (e.g., hypertonic, hypoxic, and avascular conditions)
   in IVD degradation.^[ [48]^13 , [49]^14 ^] Recent studies have reported
   that small extracellular vesicles (sEVs) derived from MSCs showed
   similar therapeutic power of MSCs in cartilage defects.^[ [50]^15 ^]
   Moreover, sEVs derived from MSCs possess anti‐inflammatory and trophic
   properties with satisfactory therapeutic potential for certain tissues,
   including bone, heart, skin, and other reproductive tissues.^[ [51]^16
   , [52]^17 , [53]^18 ^]

   Notably, the IVD microenvironment represents low nutrient availability
   and special hypertonic and hypoxic conditions. Hypoxic preconditioning
   (HP), an adaptive response for sublethal hypoxia, is a promising
   strategy to enhance the survival of transplanted MSCs by lowering
   apoptosis after exposure to the inflammatory microenvironment.^[
   [54]^19 , [55]^20 ^] HP enhances MSCs biological properties, such as
   proliferation, differentiation, and self‐renewal ability.^[ [56]^21 ,
   [57]^22 ^] A previous study demonstrated that sEVs derived from the HP
   of dental pulp stem cells (DPSCs) could better suppress the macrophage
   inflammatory response and osteoclast genesis than native DPSCs.
   However, the effect of hypoxic preconditional engineering sEVs
   (HP‐sEVs) on IVD degradation remains unknown and the underlying
   mechanism needs further investigation.

   In this study, human umbilical cord MSCs (hUCMSCs) were cultured in a
   hypoxic atmosphere, and then through hypoxic conditioning, HP‐sEVs were
   extracted from the conditioned supernatant (Scheme [58] 1 ). After
   injection into the intervertebral space, HP‐sEVs were internalized by
   NP cells. HP‐sEVs performed an anti‐inflammatory effect, promoted the
   proliferation of NP cells, and promoted proteoglycan synthesis, and
   collagen formation, resulting in ECM regeneration. Successful IVD
   regeneration was achieved via administering HP‐sEVs and this finding
   provides a promising method for IVD degradation treatment.

Scheme 1.

   Scheme 1
   [59]Open in a new tab

   Graphical summary of the HP‐sEVs‐based IVD degradation theragnostic.

2. Results

2.1. Identification of hUCMSCs and HP‐sEVs

   hUCMSCs were cultured under hypoxic conditions for 6 h. HP‐sEVs were
   isolated from the conditioned medium of hypoxic preconditioned hUCMSCs
   (Figure [60] 1a). Microscopic images revealed the typical spindle‐like
   morphology of hUCMSCs (Figure [61]1b). Moreover, hUCMSCs showed
   multiple differentiation functions including chondrogenic, adipogenic,
   and osteogenic differentiation potential (Figure [62]S1, Supporting
   Information). Surface markers of hUCMSCs were evaluated using flow
   cytometric analysis (Figure [63]1c). The results showed that the
   positive markers, including CD44, CD166, CD29, CD73, and CD105 were
   expressed in 100% of the cells. The negative markers CD14, CD45, CD34,
   and CD11b were expressed on 4.82%, 0.77%, 0.35% and 1.02% of cells,
   respectively. Subsequently, we cultured hUCMSCs in the hypoxic
   atmosphere and collected the conditioned supernatants. To estimate
   whether a hypoxic atmosphere was successfully created by the AnaeroPack
   MicroAero system, western blotting was used to test the
   hypoxia‐inducible factor‐1 alpha (HIF‐1α) protein. HIF‐1α protein
   expression was increased when hUCMSCs were cultured in the hypoxic
   condition for 6 h (Figure [64]S2, Supporting Information). The average
   diameter of HP‐sEVs was 130.4 nm and the concentration was 2.1 × 10^11
   particles per mL(Figure [65]1d). HP‐sEVs were isolated from the
   conditioned supernatants and then characterized by TEM (Figure [66]1e).
   HP‐sEVs and sEVs exhibited typical cup‐like shapes with double‐layer
   membrane structures. HP‐sEVs and sEVs have similar diameter sizes of
   about 100 nm. Western blot analysis showed that both sEVs and HP‐sEVs
   were positive for the exosomal surface marker including CD9, CD63, and
   CD81. In contrast, hUCMSCs lacked these surface markers
   (Figure [67]1f). These results demonstrate the successful extraction of
   HP‐sEVs from the conditioned medium.

Figure 1.

   Figure 1
   [68]Open in a new tab

   Identification of hUCMSCs and HP‐sEVs. a) Scheme of HP‐sEVs isolation.
   b)The spindle‐like morphology of hUCMSCs. c) Surface markers of hUCMSCs
   were analyzed via flow cytometric analysis. Red curves represent the
   control group and blue curves denote the surface of hUCMSCs. d)
   Particle size of HP‐sEVs characterized by nanoparticle tracking
   analysis (NTA). e) TEM was used to visualize the morphology of sEVs and
   HP‐sEVs. Scale bar: 100 nm. f) Surface markers of sEVs and HP‐sEVs were
   characterized using Western Blot.

2.2. HP‐sEVs Suppress the Inflammatory Microenvironment Induced by TNF‐α

   To validate the role of HP‐sEVs in NP cells in vitro, NP cells were
   co‐incubated with PKH‐26 labeled HP‐sEVs for 24 h (Figure [69] 2a). As
   illustrated in Figure [70]2b, the cytoskeleton of NP cells was stained
   in green fluorescence by actin. HP‐sEVs were stained in red
   fluorescence using PKH‐26, which was absorbed and distributed around
   the nuclei. Inducible nitric oxide synthase (iNOS) and interleukin‐6
   (IL‐6) are key mediators of immune activation and inflammation.^[
   [71]^23 , [72]^24 ^] Interleukin‐10
   (IL‐10) is an anti‐inflammatory cytokine.^[ [73]^25 ^] Macrophages and
   monocytes could generate TNF‐α during acute inflammation. As an
   important inflammatory cytokine, TNF‐α is accountable for various
   signaling events within cells, resulting in necrosis.^[ [74]^26 ,
   [75]^27 ^] In this study, TNF‐α was used to stimulate NP cells to mimic
   the inflammatory environment in vitro. Our results showed that HP‐sEVs
   suppressed iNOS and IL‐6 protein expression following TNF‐α
   stimulation, while upregulated the expression of IL‐10 protein induced
   by TNF‐α (Figure [76]2c). Immunofluorescent was performed to observe
   TNF‐α protein expression in NP cells (Figure [77]2d). The images from
   confocal laser microscopy revealed that TNF‐α stimulation significantly
   upregulated the expression of TNF‐α protein in NP cells, while HP‐sEVs
   inhibited TNF‐α expression in NP cells. Consistently, inflammatory
   genes, including matrix metallopeptidase 13 (MMP13) and TNF‐α in the
   HP‐sEVs group, were significantly down‐regulated compared to the levels
   in the TNF‐α group, while HP‐sEVs significantly increased the
   expression level of arginase 1(ARG1). Inhibition of MMP13 is an
   effective strategy to decelerate both ECM loss and osteoarthritis
   progression.^[ [78]^28 ^] Taken together, HP‐sEVs inhibit the
   inflammatory environment induced by TNF‐α and possess anti‐inflammatory
   roles in vitro.

Figure 2.

   Figure 2
   [79]Open in a new tab

   Anti‐inflammatory effect of HP‐sEVs. a) Mode pattern of co‐incubation
   of nucleus pulposus cells (NP cells) and HP‐sEVs. b) Fluorescence image
   of NP cells uptaken PKH‐26 labeled HP‐sEVs. Green actin was used to
   stain the cytoskeleton. Scale bar: 100 µm. c) Inflammatory proteins
   including iNOS, IL‐6, IL‐10 were estimated by western blotting. d)
   Immunofluorescent of TNF‐α in NP cells. In the sEVs and HP‐sEVs group,
   NP cells were treated with sEVs or HP‐sEVs (20 µg mL^−1), respectively,
   following TNF‐α (50 ng mL^−1) stimulation for 24 h. Scale bar: 100 µm.
   e) Quantitative analysis of TNF‐α positive cells.
   Immunofluorescence‐positive cells were analyzed by Image J software (n
   = 3, ^* p < 0.05, ^** p < 0.01). f) qPCR analysis of
   inflammatory‐related genes, including ARG1, TNF‐α, and MMP13 (n = 3,
   ^*** p < 0.001).

2.3. HP‐sEVs Promote the Proliferation of NP Cells and Facilitate
Proteoglycan Synthesis and Collagen Formation

   To investigate whether HP‐sEVs affect the viability of NP cells in an
   inflammatory microenvironment in vitro, NP cells were cultured with
   HP‐sEVs following TNF‐α stimulation. The results showed that sEVs and
   HP‐sEVs significantly promoted the proliferation of NP cells; HP‐sEVs
   show better effect on enhancing the viability of NP cells (Figure [80]
   3a). Subsequently, a live and dead staining kit was used to estimate
   the survival condition of NP cells under the inflammatory environment.
   HP‐sEVs treatment significantly increased the number of live cells,
   indicating that HP‐sEVs enhanced the proliferation potential of NP
   cells (Figure [81]3b,c). NP is mainly constituted by proteoglycans and
   collagen type II.^[ [82]^29 ^] Aggrecan is a member of the proteoglycan
   family of extracellular matrix proteins and loss of aggrecan is a known
   characteristic of IVD degradation. qPCR was carried out to further
   evaluate the role of HP‐sEVs on IVD regeneration (Figure [83]3d). The
   results revealed that sEVs and HP‐sEVs upregulated the mRNA expression
   levels of aggrecan and collagen type II compared to those in the
   control group. Furthermore, immunofluorescent images from confocal
   laser microscopy revealed that the production of collagen type II was
   significantly enhanced in the NP cells treated with HP‐sEVs plus TNF‐α
   (Figure [84]3e,f). Consistent with the immunofluorescent result, the
   protein expression level of collagen type II and aggrecan was
   significantly increased in the HP‐sEVs group compared to those in the
   TNF‐α group (Figure [85]3g).

Figure 3.

   Figure 3
   [86]Open in a new tab

   HP‐sEVs promote the proliferation of NP cells and facilitate
   proteoglycan synthesis and collagen formation. a) CCK8 analysis of NP
   cells treated with sEVs and HP‐sEVs under inflammatory environments
   (stimulated with 50 ng mL^−1 TNF‐α). b) Fluorescence images of NP cells
   stained with calcein AM (live cells, green fluorescence) and PI (dead
   cells, red fluorescence). In the sEVs and HP‐sEVs groups, NP cells were
   treated with sEVs or HP‐sEVs (20 µg mL^−1), respectively, following
   stimulation with 50 ng/mL TNF‐α for 24 h. Scale bar: 250 µm. c)
   Quantitative analysis of live cells (n = 3, ^* p < 0.05, ^***
   p < 0.001). d) RNA expression levels of collagen type II, and aggrecan
   were determined using qPCR (n = 3, ^* p < 0.05, ^*** p < 0.001). e)
   Immunofluorescent of collagen type II in NP cells stimulated with sEVs
   or HP‐sEVs under inflammatory environments. In the sEVs and HP‐sEVs
   groups, NP cells were treated with TNF‐α (50 ng mL^−1) and sEVs or
   HP‐sEVs (20 µg mL^−1) for 24 h. Scale bar: 100 µm. f) Quantitative
   analysis of collagen type II positive NP cells.
   Immunofluorescence‐positive cells were analyzed by Image J software (n
   = 3, ^* p < 0.05, ^*** p < 0.001). g) Aggrecan and collagen type II
   proteins were evaluated using Western Blot.

2.4. HP‐sEVs Are Conducive to Regenerating NP Tissues and Repairing the
Degraded IVD

   We established an SD rat disc puncture model to investigate the
   potential of HP‐sEVs for IVD degradation treatment in vivo. sEVs and
   HP‐sEVs were injected into the IVD after puncturing with a 21G needle.
   PBS was used as a control. X‐ray imaging of the control group revealed
   that IVD space was obviously narrow, and osteophytes and bone spurs
   were generated from the vertebral body. IVD space in the sEVs group
   showed a little decrease compared to that in the sham group. In
   contrast, the HP‐sEVs group showed a similar IVD space to the sham
   group, indicating that HP‐sEVs have a better therapeutic effect on IVD
   degradation than sEVs in vivo (Figure [87] 4a,c). Notably, IVD space
   was closely associated with ECM content, which was mainly composed of
   proteoglycans and collagen type II. The high signal intensity of
   T2‐weighted MRI is commonly used to evaluate proteoglycans in the IVD.
   Proteoglycans are lost, and water retention is impaired during IVD
   degradation, resulting in dehydration of NP tissue. Herein, a low
   signal of NP and the black disc was observed in the control group,
   while T2‐weighted images in the HP‐sEVs group showed a white high
   signal, similar to that in the sham group (Figure [88]4b,d).
   Hematoxylin and Eosin (HE) and safranin O/fast green staining were
   performed to evaluate the regeneration of newly formed NP tissues
   (Figure [89]4e–g). Few regenerated NP tissues were observed in the
   control group. In contrast, the HP‐sEVs group showed more regenerated
   NP tissues than the sEVs group. Moreover, Giemsa‐positive tissues in
   the HP‐sEVs group were decreased compared to those in the sEVs group,
   indicating that HP‐sEVs could inhibit the proliferation of monocytes in
   vivo (Figure [90]S3, Supporting Information). Immunohistochemical
   staining showed that the HP‐sEVs group had more type II collagen than
   the sEVs group, which was similar to the sham group (Figure [91]4h). In
   brief, administering HP‐sEVs revealed superior biological functions in
   promoting NP tissue regeneration and repairing degraded IVD in vivo.

Figure 4.

   Figure 4
   [92]Open in a new tab

   Imageological and histological evaluation of IVD repaired by HP‐sEVs.
   a) X‐ray and b) MRI analyses of rat coccygeal vertebrae. c) Disc height
   index (DHI) of different groups (n = 3, ^* p < 0.05). d) MRI grade at 6
   weeks after surgery (n = 3, ^** p < 0.01). e) HE staining of IVD. Scale
   bar:250 µm. f) Safranin O/fast green staining of IVD. Scale bar:
   250 µm. g) Histological grade of different treatments at 6 weeks after
   surgery(n = 3, ^* p < 0.05, ^** p < 0.01). h) Immunohistochemical
   staining of collagen type II. Scale bar: 250 µm.

2.5. Discovery of HP‐sEVs‐Associated miRNAs by Small RNA Sequencing

   MicroRNA(miRNA) is instrumental in regulating gene expression by
   binding the 3′‐UTR or amino acid coding sequences of the target gene.^[
   [93]^30 ^] To further verify the underlying mechanism of
   anti‐inflammatory, proteoglycans synthesis and collagen formation
   mediated by HP‐sEVs, miRNA expression profiles of sEVs and HP‐sEVs were
   measured by high‐throughput small RNA sequencing. The count number
   distribution of the total small RNA of sEVs and HP‐sEVs was different
   (Figure [94] 5a). Gene ontology (GO) analysis of molecular function and
   biological processes for differentially expressed miRNAs was performed
   to thoroughly investigate the effect of miRNA involved in the IVD
   regeneration (p < 0.05, log[2] fold change >1). The potential target
   genes of the differentially expressed miRNAs in HP‐sEVs were associated
   with biological processes including protein phosphorylation, cell
   differentiation, regulation of transcription, and cell adhesion
   (Figure [95]5b,c). The results from the Kyoto Encyclopedia of Genes and
   Genomes (KEGG) pathway enrichment analysis (p < 0.05, log[2] fold
   change >1) were based on the potential target genes involved in a broad
   range of signaling pathways, such as MAPK signaling pathway, the
   signaling pathway regulating pluripotency of stem cells and RAP1
   signaling pathway (Figure [96]5d). Furthermore, we selected the first
   few differentially expressed miRNAs (high‐ and low‐expression miRNAs).
   Figure [97]5e,f shows a clustered heatmap and the normalized expression
   levels of the first few differentially expressed miRNAs (high‐ and
   low‐expression miRNAs).

Figure 5.

   Figure 5
   [98]Open in a new tab

   Discovery of HP‐sEVs‐associated miRNAs by small RNA sequencing. a)
   Count number distribution of sEVs and HP‐sEVs. b) Gene Ontology (GO)
   analysis for molecular function of all differentially expressed miRNAs.
   c) GO analysis for the biological process of differentially enriched
   miRNAs. d) Enriched pathways for target genes of miRNAs enriched within
   HP‐sEVs in KEGG pathways. e) Heatmap of clustered dysregulated miRNA
   expression profiles with microarray in HP‐sEVs compared to that in sEVs
   control. f) MiRNA expression levels in HP‐sEVs compared to that in sEVs
   control.

2.6. Transcriptome Profiling of Optimized HP‐sEVs‐Regulated NP Cells Function
and Mechanism

   To explore the potential mechanism of HP‐sEVs‐regulated NP cell
   function, mRNA profiles of HP‐sEVs‐stimulated NP cells were examined
   and comparatively analyzed by mRNA sequencing. KEGG pathway enrichment
   analysis (p < 0.05, log[2] fold change >1) revealed that differentially
   expressed genes (DEGs) were mainly associated with TNF‐α/NF‐κB/Cxcl2
   signaling pathway (Figure [99] 6a). A previous study reported that
   methamphetamine triggers a neuroinflammatory response by
   upregulating TNF‐α and then activating NF‐κB in human brain endothelial
   cells.^[ [100]^31 ^] KEGG relation networks and heatmap distribution
   revealed that Cxcl2 was one of the DEGs and showed closely associated
   with the TNF‐α/NF‐κB signaling pathway, indicating Cxcl2 may be the key
   downstream gene of the TNF‐α/NF‐κB signaling pathway
   (Figure [101]6b,c). Cxcl2 was a chemokine associated with
   proinflammatory function and its downregulation inhibited the
   recruitment of polymorphonuclear cells, which were responsible for the
   elimination of monocytogenes during infection.^[ [102]^32 ^] Gene
   ontology (GO) analysis of cellular components based on DEGs showed that
   ECM was different between the HP‐sEVs and control groups, indicating
   that HP‐sEVs may repair IVD by promoting ECM regeneration
   (Figure [103]6d). TNF‐α/NF‐κB signaling pathway and ECM were altered in
   gene set enrichment analysis (GSEA), suggesting that HP‐sEVs may
   contribute to ECM regeneration by regulating the TNF‐α/NF‐κB/Cxcl2
   signaling pathway (Figure [104]6e).

Figure 6.

   Figure 6
   [105]Open in a new tab

   Transcriptome profiling of optimized HP‐sEVs regulates NP cells'
   function and mechanism. a) KEGG pathway enrichment analysis and b) KEGG
   relation networks based on HP‐sEVs regulated significant target genes.
   c) Heatmap distribution of HP‐sEVs regulating significant target genes.
   In the HP‐sEVs group, NP cells were treated with HP‐sEVs (20 µg mL^−1)
   following TNF‐α (50 ng mL^−1) stimulation for 24 h. In the control
   (Ctrl) group, NP cells were treated with TNF‐α (50 ng mL^−1). d) GO
   analysis for the cellular component of differential target genes. e)
   Gene set enrichment analysis (GSEA) analysis of ECM, NF‐κB signaling
   pathway, and TNF‐α signaling pathway.

2.7. HP‐sEVs Activated TNF‐α/NF‐κB/Cxcl2 Signaling Pathway by Delivering
miR‐7‐5p

   To investigate the miRNA HP‐sEVs activating downstream signaling
   pathways, NP cells were cultured with sEVs or HP‐sEVs, and then qPCR
   was performed to measure the expression level of HP‐sEVs enriched
   miRNAs in NP cells (the first three differentially high or low
   expressed miRNAs, marked with red arrow in Figure [106]5e,f). The
   expression of miR‐7‐5p was significantly upregulated than other miRNAs,
   indicating that HP‐sEVs activate downstream signaling pathways by
   delivering miR‐7‐5p (Figure [107] 7a,b). Transcriptome profiling
   revealed that HP‐sEVs may contribute to ECM regeneration by regulating
   the NF‐κB/Cxcl2 signaling pathway. To further validate whether HP‐sEVs
   regulated the NF‐κB/Cxcl2 signaling pathway by transferring miR‐7‐5p,
   the potential target gene of miR‐7‐5p was predicted via bioinformatic
   analysis. Among the predicted target genes, p65 is one of five
   components of the NF‐κB transcription factor family and plays an
   important role in the transactivation of several target genes involved
   in immunity, inflammation, and proliferation.^[ [108]^33 ^] The
   database showed that miR‐7‐5p could bind the 3′‐UTR region of p65
   (Figure [109]7c). To experimentally investigate whether miR‐7‐5p
   directly binds p65, luciferase reporter plasmids, including
   WT‐psicheck2‐p65 and MUT‐psicheck2‐p65 were constructed and transfected
   into 293T cells. The luciferase value was reduced following miR‐7‐5p
   stimulation, while the reaction was terminated through mutating 3′‐UTR
   regions of p65 (Figure [110]7d). The above results demonstrate that
   miR‐7‐5p targets the 3′‐UTR region of the p65 gene and suppresses the
   expression of p65.

Figure 7.

   Figure 7
   [111]Open in a new tab

   HP‐sEVs activate TNF‐α/NF‐κB/Cxcl2 signaling pathway by delivering
   miR‐7‐5p. a) The first three differentially high expressed miRNAs
   expression levels in NP cells (n = 3, ^*** p＜0.001). b) The first three
   differentially low expressed miRNAs expression levels in NP cells
   treated with sEVs or HP‐sEVs . c) Bioinformatic analysis was used to
   predict the potential target sequences of miR‐7‐5p. d) HEK293T cells
   were transfected with luciferase reporter plasmids WT or MUT 3′‐UTR of
   P65 and miR‐7‐5p, and luciferase activity was detected by the
   Dual‐Luciferase Reporter Assay System (n = 3, ^*** p < 0.001). e)
   TNF‐α/NF‐κB/Cxcl2 signaling pathway‐related proteins in NP cells were
   measured by western blot. f) Schematic diagram of the molecular
   mechanism of HP‐sEVs affection for NP cells.

   Furthermore, we estimated the related proteins of the NF‐κB/Cxcl2
   signaling pathway, such as TNF‐α, NF‐κB, phosphorylated NF‐κB
   (p‐NF‐κB), and Cxcl2. Western blot results showed that TNF‐α
   stimulation upregulated the levels of TNF‐α, p‐NF‐κB, and Cxcl2, while
   protein expression was decreased in the presence of HP‐sEVs
   (Figure [112]7e). Hypoxic preconditioned hUCMSCs secrete numerous
   HP‐sEVs. HP‐sEVs were internalized by NP cells, upregulating the
   expression level of miR‐7‐5p in NP cells and suppressing the level of
   TNF‐α. The increased miR‐7‐5p directly targeted 3′‐UTR of p65 via
   RNA‐induced silencing complex and inhibited the level of p65. The
   inhibitory p65 signal inhibited Cxcl2 production, resulting in aggrecan
   formation, collagen II deposition, and inflammatory downregulation
   (Figure [113]7f).

2.8. miR‐7‐5p Silencing Alleviates HP‐sEVs‐Mediated Anti‐Inflammatory Effect,
Proteoglycans Synthesis, and Collagen Formation

   We investigated whether miR‐7‐5p was essential and involved in
   HP‐sEVs‐mediated IVD regeneration in vitro. Either a miR‐7‐5p inhibitor
   or miR‐7‐5p mimic was transfected into NP cells to deplete or enhance,
   respectively, the expression of mature miRNAs. As shown in Figure [114]
   8a,b, miR‐7‐5p inhibitor treatment significantly diminished the
   HP‐sEV‐mediated downregulation of TNF‐α protein in NP cells. Synthetic
   miR‐7‐5p administration (Group mimic) significantly reduced TNF‐α
   expression compared to that in the control group. Consistently,
   miR‐7‐5p inhibitor administration significantly attenuated the
   downregulation of inflammatory‐related genes including MMP3 and TNF‐α
   (Figure [115]8c,d). Moreover, HP‐sEVs‐mediated collagen type II
   formation and proteoglycans synthesis were reversed by miR‐7‐5p
   inhibitor administration, whereas synthetic miR‐7‐5p significantly
   promoted the expression of aggrecan and collagen type II
   (Figure [116]8e–h), indicating that miR‐7‐5p treatment showed a similar
   effect as HP‐sEVs in promoting collagen type II formation and
   proteoglycans synthesis. Additionally, NF‐κB activator administration
   significantly diminished the miR‐7‐5p‐mediated downregulation of TNF‐α
   and IL‐6 genes in NP cells (Figure [117]S4, Supporting Information).
   Taken together, HP‐sEVs possess anti‐inflammatory effects and
   contribute to NP regeneration, mainly via delivering miR‐7‐5p.

Figure 8.

   Figure 8
   [118]Open in a new tab

   miR‐7‐5p silencing alleviates HP‐sEVs‐mediated anti‐inflammatory
   effect, proteoglycans synthesis, and collagen formation. a)
   Immunofluorescence of TNF‐α in NP cells under inflammatory
   environments. In the control group, NP cells were stimulated with TNF‐α
   for 24 h. In the inhibitor + HP‐sEVs group, NP cells were transfected
   with miR‐7‐5p inhibitor and treated with HP‐sEVs. In the mimic group,
   NP cells were transfected with miR‐7‐5p. Scale bar: 100 µm. b)
   Quantitative analysis of TNF‐α positive NP cells (n = 3, ^* p < 0.05,
   ^** p < 0.01). c,d) qPCR analysis of inflammatory‐related genes
   including TNF‐α and MMP3 (n = 3, ^** p < 0.01, ^*** p < 0.001). e)
   Immunofluorescence of collagen type II in NP cells under inflammatory
   environments. Scale bar: 100 µm. f) Quantitative analysis of collagen
   type II positive NP cells. Immunofluorescence‐positive cells were
   analyzed by Image J software (n = 3, ^** p < 0.01, ^*** p < 0.001).
   g,h) qPCR analysis of proteoglycans synthesis‐related genes, including
   aggrecan and collagen type II (n = 3, ^* p < 0.05, ^** p < 0.01).

2.9. Administered Agomir‐7‐5p Contributes to the Repair of IVD Degradation

   To demonstrate that miR‐7‐5p mediates the effect of HP‐sEVs in
   promoting NP regeneration and repair of degraded IVD in vivo, we used
   agomir‐7‐5p in an SD rat disc puncture model. Agomir‐7‐5p with or
   without sEVs was administered into intervertebral space. MRI revealed a
   high white signal in the agomir group, while the NC group showed a low
   signal of NP and black discs. Moreover, X‐ray imaging of the control
   group revealed that IVD space was narrow, and osteophytes and bone
   spurs were generated from the vertebral body. Agomir‐7‐5p injection
   significantly increased the IVD space, reducing friction between the
   vertebrae (Figure [119] 9a–c). Consistently, Safranin O/fast green and
   HE staining showed that few of the regenerated NP tissues were observed
   in the NC group, while agomir‐7‐5p group showed more regenerated NP
   tissues than in the NC and sEVs group, indicating that agomir‐7‐5p
   contributed to repaired IVD degradation (Figure [120]9d,e). In summary,
   the collective results demonstrate that HP‐sEVs promote NP regeneration
   and repair degraded IVD in vivo, mainly via transferring miR‐7‐5p.

Figure 9.

   Figure 9
   [121]Open in a new tab

   Administered agomir‐7‐5p contributes to the repair of degraded IVD. a)
   X‐ray and MRI analyses of rat coccygeal vertebrae. NC denotes the
   invalid sequence of agomir‐7‐5p. b) MRI grade at 6 weeks after surgery
   (n = 3, ^* p < 0.05, ^** p < 0.01). c) Disc height index (DHI) of
   different groups (n = 3, ^* p < 0.05, ^** p < 0.01). d) Safranin O/fast
   green staining of IVD. Scale bar: 250 µm. e) HE staining of IVD. Scale
   bar: 250 µm.

3. Discussion

   IVD structural integrity is vital for spine stability and mobility.^[
   [122]^8 ^] Therefore, IVD degradation may compress the nerve roots,
   resulting in strained mobility of spinal joints and inducing chronic
   lower back pain.^[ [123]^34 ^] Current therapies for IVD degradation
   include conservative treatments and surgical methods.^[ [124]^5 ^]
   Conservative treatments temporarily alleviate the pain via physical
   therapy and nonsteroidal anti‐inflammatory drugs, while surgical
   techniques such as total disc replacement are often invasive.^[ [125]^5
   , [126]^35 ^] However, it has been a challenging task in terms of
   simultaneous regeneration of NP tissue and restoration of the
   biological and mechanical function of IVD. Herein, we adapted a hypoxic
   preconditioning strategy to harvest hUCMSCs‐derived sEVs (HP‐sEVs) for
   enhanced treatment of IVD degeneration and mitigation of LBP. In vitro
   and in vivo assays supported the robust role of HP‐sEVs for the
   inhibition of the inflammatory microenvironment after IVD degeneration,
   which further regenerated the ECM microenvironment such as proteoglycan
   and collagen type II by transferring miR‐7‐5p and inhibiting the
   NF‐κB/Cxcl2 axis. The superior effects of HP‐sEV over conventional EVs
   are further proven in a rat IVD degeneration model with the improvement
   of NP tissue regeneration and thus the IVD regeneration.

   Administering MSCs is considered a promising therapy for IVD
   degradation due to its favorable anti‐inflammatory properties.^[
   [127]^36 ^] However, the hostile microenvironment in IVD degradation
   contains special hypertonic, hypoxic, and avascular conditions,
   limiting the broad clinical applications of MSC therapy.^[ [128]^13 ,
   [129]^14 ^] Notably, sEVs realized similar therapeutic efficacy of
   MSC‐based therapies in terms of the regeneration of cartilaginous
   tissues.^[ [130]^15 ^] Previous studies have suggested that sEVs
   derived from MSCs pre‐conditioned by LPS may have better regulatory
   abilities for macrophage polarization and thus more effective
   suppression of inflammation, as compared to sEVs derived from
   non‐LPS‐treated MSCs.^[ [131]^37 ^] In this study, we aimed to
   establish a novel functional sEV‐based strategy based on hypoxic
   preconditioning for the effective treatment of IVD degradation. HP is a
   powerful, endogenous protective mechanism that improves the biological
   functions of MSCs in harsh microenvironments.^[ [132]^19 , [133]^20 ,
   [134]^21 , [135]^22 ^] Nevertheless, the therapeutic effect and the
   underlying mechanism of action of HP‐sEVs remain to be investigated in
   the context of IVD degeneration.

   The key to achieving IVD regeneration is to design an effective
   therapeutic therapy that targets the underlying pathogenic mechanisms
   of IVD degradation. The critical issues in the progression of IVD
   degradation have included inflammatory responses, progressive cell
   apoptosis, and deterioration in anabolic activities and cellular
   functions.^[ [136]^38 ^] To address these challenges, we compared the
   morphological and functional differences between sEVs and HP‐sEVs.
   Characterization of sEVs using NTA and TEM suggested that there lacks
   significant morphological differences between the sEVs and HP‐ sEVs in
   terms of their size, shape, or electron density. Moreover, we have also
   shown that HP‐sEVs demonstrate improved anti‐inflammatory function as
   compared to conventional sEVs. Additionally, HP‐sEVs also showed
   superior outcomes in terms of promoting NP cell proliferation,
   proteoglycan formation, and collagen type II synthesis. Proteoglycans
   are key biomacromolecules regulating cell proliferation,
   differentiation, and ECM turnover.^[ [137]^39 ^] Collagen is an
   important component of the ECM and provides structural and mechanical
   support to disc cells.^[ [138]^7 ^] The collective findings in this
   study strongly support the view that HP‐sEVs could improve the
   degenerated IVD microenvironment by inhibiting the inflammatory
   response and decreasing catabolic activities. Meanwhile, HP‐sEVs can
   increase anabolic activity by promoting ECM regeneration and
   facilitating NP cell proliferation, resulting in a favorable IVD
   microenvironment and accelerated regeneration.

   MiRNAs are non‐coding RNAs and also key mediators of the therapeutic
   function of sEVs.^[ [139]^40 , [140]^41 ^] Recent studies have reported
   that sEVs promoted vascularized bone regeneration by transferring
   miR‐23a‐3p.^[ [141]^30 ^] However, a differential analysis of the miRNA
   profile between HP‐sEVs and sEVs and the miRNA‐mediated effects of
   promoting NP regeneration remains to be investigated. In this study,
   RNA sequencing indicated the differential miRNA profile of HP‐sEVs.
   Specifically, miR‐423‐5p, miR‐7‐5p, and miR‐708‐5p were significantly
   enriched in HP‐sEVs compared to that in sEVs, whereas miR‐483‐5p,
   miR‐891a‐5p, and miR‐652‐5p levels were lower in HP‐sEVs than in sEVs.
   We further demonstrated that miR‐7‐5p expression levels were
   significantly upregulated compared to those of other miRNAs in NP cells
   treated with HP‐sEVs. Moreover, HP‐sEVs‐mediated collagen type II
   formation and proteoglycans synthesis were reversed by miR‐7‐5p
   inhibitor administration. In contrast, synthetic miR‐7‐5p showed a
   biological function similar to that of HP‐sEVs in promoting the
   expression of aggrecan and collagen type II. These findings indicated
   the strong anti‐inflammatory effects of HP‐sEVs that are crucial for NP
   regeneration which is likely achieved by the delivery of miR‐7‐5p
   existent in HP‐sEVs. Subsequently, our transcriptome profiling results
   revealed the activation of the NF‐κB/Cxcl2 signaling pathway in the
   HP‐sEVs‐treated NP cells, suggesting that HP‐sEVs may contribute to ECM
   by regulating NF‐κB/Cxcl2 signaling pathway. Moreover, we demonstrated
   that miR‐7‐5p can bind to the 3′‐UTR region of p65, which is one of
   five components of the NF‐κB transcription factor family. In addition,
   miR‐7‐5p mimetics inhibited the production of p65, whereas the mutant
   miR‐7‐5p increased p65 production. NF‐κB signaling pathway regulates
   inflammatory and oxidative responses.^[ [142]^42 , [143]^43 ^] Herein,
   we verified that HP‐sEVs transferred miR‐7‐5p into NP cells.
   Upregulated miR‐7‐5p targets the p65 gene and suppresses p65
   production. The inhibitory p65 signal further inhibits Cxcl2
   production, resulting in inflammatory response downregulation, aggrecan
   formation, and collagen II deposition.

   In summary, our study demonstrates the superior functions of HP‐sEVs in
   promoting IVD regeneration compared to that of normal sEVs. In
   addition, we highlighted the underlying mechanism by which HP‐sEVs
   suppressed inflammatory response and promoted IVD regeneration. Hypoxic
   preconditioning upregulated the production of miR‐7‐5p in sEVs, and the
   enriched level of exosomal miR‐7‐5p significantly promoted the
   therapeutic effect of HP‐sEVs on IVD degradation. HP‐sEVs alleviated
   the inflammatory microenvironment involved in IVD degradation, promoted
   NP cell proliferation, and enhanced proteoglycan synthesis and collagen
   formation, thus promoting IVD regeneration by activating the
   miR‐7‐5p/NF‐κB/Cxcl2 axis. Therefore, we envisage that HP‐sEV‐based
   therapy may serve as a promising treatment for IVD degradation as well
   as lower back pain. Moving forward, a more accurate analysis of the
   spatiotemporal distribution of HP‐sEVs after injection is required to
   understand its pharmacokinetic profiles. Rigorous optimization of
   dosimetry of HP‐sEVs for fully mitigating LBP in both small and large
   animal models, and large‐scale HP‐sEVs production would also be
   essential before its clinical translation.

4. Experimental Section

Cell Isolation and Culture

   Nucleus pulposus cells (NP cells) were isolated from NP tissues of 8
   weeks old Sprague–Dawley rats. According to the previous protocol,^[
   [144]^44 ^] the rats were euthanized by injecting excessive
   pentobarbital sodium. Subsequently, NP tissues were obtained and
   transferred to a sterile PBS. NP tissues were minced to a size of
   1 mm^3 roughly and then digested by 0.1% collagenase type II
   (Invitrogen, 17 101 015, USA) which was dissolved in DMEM/F12(Corning,
   10 092 001, USA). After digested for 4 h, the solution was filtrated by
   a 70 µm cell strainer. The cells were obtained after removing the
   collagenase solution at 316 RCF. NP cells were cultured in DMEM/F12
   containing 10% fetal bovine serum (FBS, Gibco, 10099–141, USA). The
   above procedures were completed in a sterile environment.

hUCMSCs Characterization

   hUCMSCs were cultured in basal medium (DAKAWE, 6 114 021, China)
   supplemented serum analog EliteGro‐Adv (Elitecell Biomedical, EPA‐050,
   Canada). hUCMSCs at a density of 10 × 10^4 cells were stimulated via
   the primary antibodies (Cyagen, HUXMX‐09011, China) for 30 min in a
   dark environment. The cells were centrifuged for 5 min at 1000 rpm,
   followed by washing with PBS. The centrifuge tubes were obtained from
   NEST Biotechnology. Finally, flow cytometry was used to analyze the
   cells after co‐incubating with FITC IgG antibody for 10 min.

   The multipotency of hUCMSCs was evaluated by inducing MSCs into
   chondrogenic, adipogenic, and osteogenic cells. hUCMSCs were incubated
   in chondrogenic, adipogenic, and osteogenic differentiated medium
   (Cyagen, China).

Isolation and Identification of HP‐sEVs

   A hypoxic cultural environment was established by the AnaeroPack method
   according to the previous methods.^[ [145]^45 , [146]^46 ^] In brief,
   AnaeroPack (Cat# D07, Mitsubishi Gas Chemical Company, Japan) was
   placed into a sealed air‐tight container to mimic the hypoxic
   atmosphere environment. Then, hUCMSCs were cultured in a 10 cm dish and
   placed into a sealed air‐tight container. HP‐sEVs were isolated from
   the conditional supernatants of hUCMSCs cultured in the hypoxic
   atmosphere. hUCMSCs were cultured in a medium containing exosomes‐free
   FBS (System Biosciences, USA, 50A‐1) at 37 °C, regular (21% O[2]) or
   hypoxic (1% O[2]) condition. As previously described,^[ [147]^47 ^]
   sEVs were obtained from a conditional medium by gradient
   centrifugation. Briefly, the supernatants were centrifuged at 300 g for
   10 min, 2000 g for 10 min, and 10 000 g for 30 min at 4 °C to clear
   away cell debris. Subsequently, the conditional medium was
   re‐centrifuged at 1 00 000 g for twice at 4 °C and then the sediment
   was resuspended with PBS to obtain sEVs or HP‐sEVs. A transmission
   electron microscope (TEM, G2 spirit BioTWIN by Tecnai) was used to
   observe the morphology of sEVs and nanoparticle tracking was used to
   measure the particle concentration and distribution.

PKH‐26 Staining Assay

   According to the manufacturer's instructions, HP‐sEVs and sEVs were
   labeled through PKH‐26 (Sigma, Germany). Briefly, 4 µL PKH‐26 dye was
   added into 500 µL Dilution C solution to obtain the mixture solution.
   Subsequently, HP‐sEVs were incubated with the above mixture solution in
   a dark environment for 30 min. The process of staining was stopped by
   1% bovine serum albumin. Then, the mixture containing sEVs was
   centrifuged at 1 00 000 g at 4 °C. The labeled HP‐sEVs were resuspended
   by sterile PBS. NP cells were cultured with PKH‐26 labeled HP‐sEVs for
   24 h and then fixed by 4% paraformaldehyde (PFA). Subsequently, NP
   cells were incubated with DAPI. The photograph was observed via
   microscopy.

Western Blot Analysis

   Radio‐immunoprecipitation assay buffer (RIPA, Amizona Scientific LLC,
   China) was used to lysis and extract protein. The BCA method was used
   to measure the protein concentration. After separation by SDS‐PAGE gel,
   proteins were transferred to polyvinylidene difluoride (PVDF)
   membranes. Subsequently, PVDF membranes were blocked via 5% bovine
   serum albumin (BSA) and then co‐incubated with primary antibodies
   overnight at 4 °C. Then, PVDF membranes were washed with tris‐buffered
   saline with tween‐20 (TBST, UElandy, China) three times. PVDF membranes
   were incubated with secondary antibodies for 1 h at room temperature.
   The protein bands were observed with ChemiDoc XRS+ Gel Imaging System
   (Bio‐Rad, USA). Primary antibodies included anti‐NF‐κB (CST, 8242,
   USA), anti‐p‐NF‐κB (CST, 3033, USA), Cxcl2(Abcam, ab275879, USA),
   IL‐6(CST, 12 912, USA), IL‐10(Abcam, ab133575, USA), collagen II
   (Abcam, ab34712, USA), Aggrecan (Abcam, ab3778, USA) and β‐actin
   (Forevertech Biotechnologies, 20270, China).

Quantitative RT‐PCR Analysis

   Trizol reagent was used to extract the total RNA of the cells. RNA was
   extracted by SPARKeasy Cellular RNA Extraction Kit (Sparkjade
   Biotechnology, China). Then, cDNA was obtained after reverse
   transcription of RNA. According to the manufacturer's protocols, cDNA
   was amplified with TB Green Premix EX Taq II (Takara, RR820A, Japan).
   The primers used in qPCR are shown in Table [148]S1 (Supporting
   Information). Reverse transcription of miRNA into cDNA was performed
   using miDETECT A Track miRNA qRT‐PCR Starter Kit (RiboBio, China).
   Internal controls for mRNA and miRNA were GAPDH and U6 RNA,
   respectively. The results were analyzed by the 2^−ΔΔCT quantitative
   method. Data were shown as mean ± SD of biological triplicates.

Cell Proliferation Assay

   The proliferation ability of NP cells was evaluated by cell counting
   kit‐8(CCK‐8, KeyGEN, China) and LIVE/DEAD Fixable dead cell stains
   (Invitrogen, USA). Briefly, NP cells cultured with a density of 5000
   cells were cultured in DMEM/F12 medium containing TNF‐α (50 ng mL^−1),
   sEVs, or HP‐sEVs (20 µg mL^−1) in a 96‐well plate for 1 day.
   Subsequently, CCK8 was added in the medium and co‐cultured for 2–4 h.
   Then, the viability of NP cells was measured via absorbance values at
   450 nm.

   NP cells with a density of 2 × 10^4 were implanted in a 24‐well plate.
   sEVs or HP‐sEVs (20 µg mL^−1) were co‐incubated with the cells for 24
   h. Calcein‐AM and PI were diluted in DMEM/F12 at a ratio of 1:2000. NP
   cells were incubated with the mixture medium for 30 min. Subsequently,
   calcein‐AM and PI solution were used to detect live cells (green
   fluorescence) and dead cells (red fluorescence), respectively.

Immunofluorescence Staining

   NP cells were incubated with sEVs or HP‐sEVs for 3 days. Then, the
   cells were fixed using 4% PFA and then washed with PBS. Subsequently,
   NP cells were treated with 0.25% Triton X‐100 (Elabscience
   Biotechnology, China) and then blocked with 5% donkey serum. NP cells
   were treated with primary antibodies including TNF‐α (Abcam, ab1793,
   USA) and collagen II (Abcam, ab34712, USA) overnight at 4 °C. After
   being washed by PBS, NP cells were conjugated with a fluorescent
   secondary antibody (Invitrogen, USA) and stained with DAPI for 5 min at
   room temperature. Finally, NP cells were visualized via the confocal
   microscope (Carl Zeiss, Germany). Immunofluorescence‐positive cells
   were analyzed by Image J software.

miRNA Inhibitor Transient Infection

   MiRNA inhibitors and mimics were purchased from RiboBio company.
   According to the manufacturer's instructions, when NP cells were
   cultured at 60−70% density, miRNA inhibitors or mimics were transfected
   into NP cells via lipofectamine 3000 (Invitrogen, L3000008). After
   being transfected for 1 day, the transfection medium was removed. Then,
   the growth medium contained tumor necrosis factor‐alpha (TNF‐α) and
   HP‐sEVs were added.

Animal Experiment

   An ethical review committee of Tongji Hospital affiliated with Tongji
   University reviewed and approved the details of animal experiments
   (ID:2022‐DW‐(003)). According to the previous protocol,^[ [149]^29 ^]
   Sprague–Dawley rats were anesthetized and then disinfected with iodine.
   IVD between Co5/6 was exposed by slitting the skin. Then, a needle
   (21 g) was used to puncture the center of the discs, rotated for 5 s
   then the position was maintained for 30 s. Ten microliters sEVs or
   HP‐sEVs were injected into the disc. Four groups were fabricated in
   this study. 1) control (10 µL PBS was injected into the disc), 2) sEVs
   group, 3) HP‐sEVs group, and 4) Sham group. The rats in the sham group
   were only incised in the outer layer of skin at the corresponding site
   without needle puncture. Animals were divided into four groups in this
   study to further demonstrate that miR‐7‐5p mediated the effect of
   HP‐sEVs in vivo. 1) NC (injected invalid sequence of agomir‐7‐5p
   (100 nm)), 2) sEVs group, 3) agomir group (injected agomir‐7‐5p), 4)
   agomir+sEVs group (injected agomir‐7‐5p (100 nm) and sEVs), and 5) Sham
   group.

X‐Ray and MRI Examination

   The therapeutic effect of sEVs and HP‐sEVs for IVD degradation was
   evaluated by X‐ray and MRI at 6 weeks post‐surgery. In brief, the rats
   were anesthetized, and then their tail was exposed in a straight line
   on a molybdenum target radiation‐imaging device. The exposure
   parameters of the X‐ray included 32 kV and 10 s duration. X‐ray images
   were measured by Image J software to obtain disc height index (DHI)
   values. DHI values were calculated by the degraded disc divided by the
   normal disc and represented the disc height change of the degraded disc
   compared with the normal disc. MRI was carried out by Vero 1.5T system
   (Siemens, Germany). T2‐weighted images were captured in the coronal
   plane. The parameter of MRI was that visual field, 80 × 80 mm; slice
   thickness:1.4 mm; repetition time, 4000 ms; echo time, 96 ms.
   T2‐weighted images were graded into I–IV according to the modified
   Thomson classification method.^[ [150]^48 ^]

Histology and Immunohistochemical Analysis

   Animal specimens were placed in 10% formalin for 2 days. Subsequently,
   the samples were decalcified and then embedded in paraffin. Six
   micrometers thick sections of tissue samples were obtained by paraffin
   slicing machine. Then, the tissue sections were stained with Safranin‐O
   & fast green, hematoxylin and eosin (H&E). Immunohistochemical staining
   was according to the previous protocol.^[ [151]^49 ^] In brief, the
   sections of samples were repaired with sodium citrate solution to
   achieve epitope retrieval. Subsequently, the samples were incubated
   with primary antibodies including collagen II (Abcam, USA) overnight at
   4 °C. On the next day, the sections were co‐incubated with a second
   antibody and then stained with hematoxylin.

Statistical Analysis

   All results are expressed as mean ± standard deviation. The unpaired
   Student's t‐test was used to evaluate the difference between the two
   groups. One‐way analysis of variance (ANOVA) was used to evaluate the
   difference between multiple groups. Differences were considered
   statistically significant when p < 0.05. Data were expressed as the
   mean ± standard deviation of three replicates. The t‐test was used to
   compare the mean beta values of each group.

Conflict of Interest

   The authors declare no conflict of interest.

Author Contributions

   H.H. and Z.W. contributed equally to this work. H.H. performed
   investigation, data curation, methodology, and wrote‐original draft.
   Z.W. performed data curation, investigation, and methodology. H.Y.
   performed investigation and data curation. Y.B. performed
   investigation. R.Z. performed conceptualization, supervision, and
   project administration. L.C. performed conceptualization, supervision,
   and project administration.

Supporting information

   Supporting Information
   [152]Click here for additional data file.^ (478.5KB, pdf)

Acknowledgements