Abstract

   The balance among different CD4^+ T cell subsets is crucial for
   repairing the injured spinal cord. Dendritic cell (DC)‐derived small
   extracellular vesicles (DsEVs) effectively activate T‐cell immunity.
   Altered peptide ligands (APLs), derived from myelin basic protein
   (MBP), have been shown to affect CD4^+ T cell subsets and reduce
   neuroinflammation levels. However, the application of APLs is
   challenging because of their poor stability and associated side
   effects. Herein, it is demonstrate that DsEVs can act as carriers for
   APL MBP[87‐99]A^91 (A91‐DsEVs) to induce the activation of 2 helper T
   (Th2) and regulatory T (Treg) cells for spinal cord injury (SCI) in
   mice. These stimulated CD4^+ T cells can efficiently “home” to the
   lesion area and establish a beneficial microenvironment through
   inducing the activation of M2 macrophages/microglia, inhibiting the
   expression of inflammatory cytokines, and increasing the release of
   neurotrophic factors. The microenvironment mediated by A91‐DsEVs may
   enhance axon regrowth, protect neurons, and promote remyelination,
   which may support the recovery of motor function in the SCI model mice.
   In conclusion, using A91‐DsEVs as a therapeutic vaccine may help induce
   neuroprotective immunity in the treatment of SCI.

   Keywords: antigen‐specific immune, dendritic cells, neuroprotective
   immunity, small extracellular vesicles, spinal cord injury
     __________________________________________________________________

   This study uses Dendritic cell‐derived small extracellular vesicles
   (DsEVs) loaded with APL MBP[87‐99]A^91 (A91‐DsEVs). A91‐DsEVs promotes
   activation of Th2 and Treg cells, which can home to the injured spinal
   cord and induce neuroprotective immunity. The treatment with A91‐DsEVs
   can enhance axon regrowth, neuronal survival, and remyelination, which
   may support the functional motor recovery in the SCI model mice.

   graphic file with name ADVS-11-2304648-g009.jpg

1. Introduction

   Spinal cord injury (SCI) is a severe form of central nervous system
   (CNS) damage that leads to significant neurological dysfunction due to
   disruption of neural circuitry.^[ [46]^1 ^] The post‐SCI
   pathophysiological progression involves an immune response.^[ [47]^2 ^]
   CD4^+ T cells are important components of the immune response following
   an injury, and the balance between different subtypes of CD4^+ T cells
   is crucial for neuroprotection.^[ [48]^3 ^]

   During the immune response after SCI, most CD4^+ T cells are biased
   towards T helper type 1 (Th1) and T helper type 17 (Th17) cells. They
   induce an increased release of inflammatory cytokines and
   autoantibodies. This process persists for days or weeks, and aggravates
   neuronal loss, demyelination and axonal damage.^[ [49]^4 , [50]^5 ,
   [51]^6 ^] In contrast, 2 helper T (Th2) and regulatory T (Treg) cells
   have been shown to inhibit the over‐activation of inflammatory
   response.^[ [52]^7 ^] Studies have highlighted the significant role of
   Th2 cells in providing protection and promoting recovery after SCI by
   secreting anti‐inflammatory cytokines.^[ [53]^8 , [54]^9 , [55]^10 ^]
   Treg cells have been shown to play an anti‐inflammatory role and to
   contribute to remyelination and axonal survival after SCI.^[ [56]^11 ,
   [57]^12 ^] Therefore, CD4^+ T‐cell regulation is an important target
   for SCI treatment.^[ [58]^7 ^]

   A promising strategy for changing T‐cell subsets is the use of altered
   peptide ligands (APLs) derived from myelin sheath antigens with
   mutations in critical T‐cell receptor (TCR) or major histocompatibility
   complexes (MHC) contact residues. APLs are analogs derived from native
   peptides that commonly carry amino acid substitutions at TCR contact
   residues.^[ [59]^13 ^] One such antigen is APL of myelin basic protein
   (MBP).^[ [60]^14 ^] A specific segment of MBP consisting of amino acids
   87–99 (MBP[87–99]) has been shown to be a major target of T cells in
   neuroinflammation. Previous studies have demonstrated that
   MBP[87‐99]A^91, an APL of MBP[87‐99] in which the lysine residue at
   position 91 is replaced with alanine, can alter the polarity of
   inflammatory CD4^+ T cells and induce neuroprotective effect.^[ [61]^15
   , [62]^16 ^] However, the clinical application of APLs, such as
   MBP[87‐99]A^91, comes with many challenges, including poor stability
   due to enzymatic degradation in blood circulation^[ [63]^17 ^] and
   damaging biological barriers that result in toxicity to tissues and
   organs.^[ [64]^18 ^] APLs may also cause hypersensitivity responses
   because of their commonly used high concentrations.^[ [65]^19 ^]

   Dendritic cell (DC)‐derived small extracellular vesicles (DsEVs) are
   ideal carriers for peptides to induce antigen‐specific immunity, the
   stable lipid bilayers effectively protect their contents and reduce
   antigen cargo loss.^[ [66]^20 , [67]^21 ^] DsEVs contain MHC I and MHC
   II, along with costimulatory molecules that facilitate peptide
   presentation and T‐cell activation.^[ [68]^22 ^] DsEVs have been shown
   to be non‐immunogenic or possess very low immunogenicity.^[ [69]^23 ^]
   The therapeutic potential of antigen‐specific immunity induced by DsEVs
   has mainly been evaluated in treating conditions such as tumors and
   infectious diseases.^[ [70]^24 , [71]^25 , [72]^26 , [73]^27 ^]

   However, the application of DsEVs for the treatment of CNS injuries
   remains underexplored. For SCI treatment, DsEVs as a therapeutic
   vaccine may offer several advantages: 1) Compared with other
   nanomaterials (liposomes or nanoparticles, among others), DsEVs can
   deliver the peptide cargo directly to the lymphoid organs,
   antigen‐presenting cells or T cells, boosting immunotherapeutic
   efficacy;^[ [74]^28 ^] 2) DsEVs can stimulate antigen‐specific T‐cell
   activation, mitigating the risk of side effects from the generalized
   immune response;^[ [75]^29 , [76]^30 ^] 3) DsEVs could deliver their
   therapeutic benefits through peripheral injection, avoiding secondary
   damage to the lesion area caused by local delivery. Hence, we aimed to
   design a DsEVs‐based nanovaccine platform to induce a CD4^+
   T‐cell‐mediated neuroprotective effect at the injury site and to
   demonstrate the potential of this innovative approach for SCI treatment
   (Scheme [77] 1 ).

Scheme 1.

   Scheme 1
   [78]Open in a new tab

   Small extracellular vesicles derived from dendritic cells (DsEVs)
   loaded with MBP[87‐99]A^91 (A91‐DsEVs) was used for spinal cord injury
   treatment. A91‐DsEVs promoted activation of Th2 and Treg cells, which
   can home to the injured spinal cord and induce neuroprotective
   immunity. The treatment with A91‐DsEVs can enhance axon regrowth,
   neuronal survival and remyelination, which may support the functional
   motor recovery in the SCI model mice.

2. Results

2.1. Characterization of A91‐DsEVs

   In this study, we introduced MBP[87‐99]A^91 or ovalbumin (OVA) as
   vaccine cargo into DsEVs. TAT[47‐57], a cell‐penetrating peptide (CPP),
   was added to peptide to enhance the loading efficiency. Figure [79] 1a
   illustrates the procedure used for A91‐DsEVs isolation.
   Immunofluorescence staining showed clear internalization of
   isothiocyanate (FITC)‐labeled TAT[47‐57]‐MBP[87‐99]A^91 in DCs
   (Figure [80]1b). The loading efficiency of MBP[87‐99]A^91 peptide into
   DCs and DsEVs was quantified using flow cytometry. The result showed
   that MBP[87‐99]A^91 peptide (labeled with FITC) was successfully loaded
   into DCs and A91‐DsEVs at loading rates of 95.1% and 67.9% respectively
   (Figure [81]1c). Flow cytometry was used to assess the expression of
   MHC II and costimulatory molecules in DCs after loading with
   MBP[87‐99]A^91 (Figure [82]S1, Supporting Information). The results
   revealed the expression of MHC II, CD86 and CD80 were raised in
   MBP[87‐99]A^91‐treated group and OVA‐treated group compared to that in
   the PBS‐treated group.

Figure 1.

   Figure 1
   [83]Open in a new tab

   Successful isolation of A91‐DsEVs from DCs culture medium. a) Schematic
   illustration of A91‐DsEVs isolation. b) Immunofluorescence images to
   show FITC‐labeled TAT[47‐57]‐MBP[87‐99]A^91 (green) loaded into DCs
   (red) (Scale bar = 100 µm, n = 3). c) Flow cytometry to quantify the
   loading rate of MBP[87‐99]A^91 in DCs (pink) and A91‐DsEVs (green) (n =
   3). d) Western blotting to detect DsEVs positive markers Alix, TSG101,
   CD63 and negative marker Calnexin in DCs and A91‐DsEVs (n = 3). e) TEM
   to visualize ultrastructure of DsEVs (Scale bar = 100 nm). NTA to
   reveal the size distribution of DsEVs (n = 3). Data were expressed as
   mean ± standard deviation (SD).

   Western blotting was used to investigate the characteristic membrane
   protein markers of A91‐DsEVs, including the positive markers, Alix,
   TSG101, and CD63, and the negative marker Calnexin (Figure [84]1d).
   Alix, TSG101 and CD63 expression was detected in A91‐DsEVs, OVA‐DsEVs,
   and DsEVs. Calnexin was only detected in DCs. Next, vesicle morphology
   of different groups of DsEVs was characterized through transmission
   electron microscopy (TEM) (Figure [85]1e). TEM revealed vesicles with
   typical spherical vesicular structures. To investigate the size
   distribution profile of DsEVs, we performed a size detection using
   nanoparticle tracking analysis (NTA) (Figure [86]1e). NTA revealed the
   particles of A91‐DsEVs have a peak diameter of 113.0±1.8 nm, which was
   similar to that of other groups (OVA‐DsEVs: 114.9±4.5 nm; DsEVs:
   111.1±0.8 nm.). These findings indicated the successful isolation of
   A91‐DsEVs from DCs culture medium.

2.2. A91‐DsEVs Could be Internalized by DCs and CD4^+ T Cells and Induce the
Polarization of CD4^+ T Cells toward Th2 and Treg Subtypes In Vitro

   To evaluate the A91‐DsEVs internalization by DCs and CD4^+ T cells, we
   labeled A91‐DsEVs with
   1,1‐dioctadecyl‐3,3,3,3‐tetramethylindocarbocyanine iodide (Dil). We
   then co‐cultured Dil‐labeled A91‐DsEVs with DCs and CD4^+ T cells
   respectively. After 24 h of co‐culture, DCs were collected for further
   experiments. As shown in Figure [87] 2a, A91‐DsEVs (indicated by orange
   fluorescence) were internalized by DCs and CD4^+ T cells. The viability
   of A91‐DsEVs‐treated and OVA‐DsEVs‐treated DCs was much higher than
   those of the DsEVs‐treated and PBS‐treated groups (Figure [88]2b).
   Similarly, the viability of CD4^+ T cells was higher in the
   A91‐DsEVs‐treated and OVA‐DsEVs‐treated groups than in the PBS‐treated
   group (Figure [89]2b). Flow cytometry was used to assess the expression
   of MHC II and costimulatory molecules in the DCs (Figure [90]2c). The
   expression of MHC II and CD80 in DCs was higher in the
   A91‐DsEVs‐treated group than in the DsEVs‐ and PBS‐treated groups. The
   results also revealed that CD86 expression raised significantly in
   A91‐DsEVs‐treated group than that in the other groups. These findings
   demonstrate that A91‐DsEVs were efficiently internalized by DCs and
   CD4^+ T cells, resulting in the upregulated expression of MHC II and
   costimulatory molecules on DCs and enhanced cellular viability.

Figure 2.

   Figure 2
   [91]Open in a new tab

   A91‐DsEVs internalized by DCs and CD4^+ T cells, increased the
   antigen‐presenting capability of DCs and regulated CD4^+ T cells
   polarization toward anti‐inflammatory phenotype. a) Images to depict
   A91‐DsEVs (orange) internalized by DCs (green) and CD4^+ T cells
   (green). b) The CCK‐8 assay to assess the cell viability of DCs and
   CD4^+ T cells derived from distinct treatment groups, which involved
   co‐cultivation of different DsEVs or PBS with DCs and CD4^+ T cells
   respectively (n = 3). c) Flow cytometry of MHC II, CD86, and CD80
   expression in DCs following treatment with A91‐DsEVs, OVA‐DsEVs, DsEVs,
   or PBS (n = 3). d) Flow cytometry to show the impact of A91‐DsEVs on
   the polarization of CD4^+ T cells on day 3 of co‐culture (n = 3). e)
   ELISA to measure IL‐4, IL‐10, IFN‐γ, and IL‐17A levels in the
   supernatant of the MLR on day 3 of co‐culturing (n = 3). f) ELISA to
   measure NT‐3 and BDNF expression in the MLR system supernatant among
   different groups on day 3 (n = 3). g) and h) ELISA to measure IL‐4,
   IL‐10, IFN‐γ, and IL‐17A levels g), nerve growth factors levels (BDNF
   and NT‐3) h), in the supernatant of the MLR on day 5 of co‐culturing (n
   = 3). MFI, mean fluorescence intensity. Data were expressed as mean ±
   SD. (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 as assessed by
   one‐way ANOVA with Dunnett's multiple comparisons).

   We conducted quantitative real‐time PCR (qRT‐PCR) analysis of DCs
   co‐cultured with different DsEVs or PBS for 24 h to investigate how
   A91‐DsEVs regulate MHC‐II expression. The key molecules regulating
   MHC‐II expression were assessed, including MHC class II transactivator
   (CIITA), regulatory factor X 5 (RFX5), RFX‐associated protein (RFXAP)
   and RFX‐associated ankyrin‐containing protein (RFXANK). The results
   showed a significant increase in CIITA gene expression in DCs treated
   with A91‐DsEVs (Figure [92]S2a, Supporting Information). This was
   further confirmed by an Enzyme‐linked immunosorbent assay (ELISA),
   which demonstrated a substantial increase in CIITA protein levels
   following A91‐DsEVs treatment (Figure [93]S2b, Supporting Information).

   Moreover, to simulate the immune response in vivo, we established an in
   vitro mixed lymphocyte reaction (MLR) system by co‐culturing DsEVs, DCs
   and splenic T cells. We first assed the cell viability of CD4^+ T cells
   on day 3 in the MLR using the Cell Counting Kit‐8 (CCK‐8) assay. The
   result demonstrated that the viability of CD4^+ T cells in
   A91‐DsEVs‐treated group was much higher than that in the other groups
   (Figure [94]S3a, Supporting Information). Flow cytometry was used to
   detect cell subtype changes in splenic CD4^+ T cells in the MLR on day
   3 of co‐culturing. The results revealed decreased counts of IFN‐γ^+
   CD4^+ and IL‐17A^+ CD4^+ T cells, but increased counts of Foxp3^+ CD4^+
   and IL‐4^+ CD4^+ T cells in the A91‐DsEVs‐treated group
   (Figure [95]2d). Furthermore, on day 3 of co‐culture, the supernatant
   of the MLR system was collected to assess the levels of signature
   cytokines in different subsets of CD4^+ T cells. The A91‐DsEVs group
   showed elevated levels of cytokines IL‐4 and IL‐10, along with
   decreased expression of cytokines IL‐17A and IFN‐γ (Figure [96]2e). The
   isotype used for flow cytometry is shown in Figure [97]S4 (Supporting
   Information). Overall, these results underscore the potential of
   A91‐DsEVs to increase the antigen‐presenting capability of DCs and to
   regulate the polarization of CD4^+ T cells towards Th2 and Treg
   subtypes in vitro.

   Because CD4^+ T cells are also a major source of nerve growth factors,
   including brain‐derived neurotrophic factor (BDNF) and neurotrophins‐3
   (NT‐3),^[ [98]^31 ^] we assessed the levels of BDNF and NT‐3 in MLR
   using ELISA. The levels of BDNF and NT‐3 were higher in the
   A91‐DsEVs‐treated group than in the PBS‐treated group (Figure [99]2f).

   To evaluate the immunological response, we also assessed T‐cell
   proliferation and cytokine production rates in the experimental groups
   on day 5 post‐MLR. The CCK‐8 assay revealed enhanced viability of CD4^+
   T cells in the A91‐DsEVs‐treated group compared to those in the other
   three groups (Figure [100]S3b, Supporting Information). In addition,
   the A91‐DsEVs group exhibited elevated levels of IL‐4 and IL‐10,
   accompanied by a reduced expression of IL‐17A and IFN‐γ cytokines,
   compared to those observed in the other groups (Figure [101]2g). The
   levels of BDNF and NT‐3 in the MLR were higher in the A91‐DsEVs group
   than in the other groups on day 5 after co‐culturing (Figure [102]2h).

2.3. A91‐DsEVs Regulated the Polarization of CD4^+ T Cells In Vivo and
Promoted CD4^+ T Cells Infiltrating to the Injury Site

   Figure [103] 3a shows a schematic of the in vivo experimental
   procedure. We initially labeled DsEVs from various groups with
   dimethylindole Red (DIR) and administered them to the mice via the tail
   vein, using PBS as a control. After a 6‐h interval post‐injection, we
   collected the spinal cord, heart, liver, spleen, lungs, and kidneys for
   fluorescence tracking analysis. Notably, no detectable signal was
   observed in the spinal cord or other major organs, whereas fluorescence
   was primarily localized in the spleen and liver (Figure [104]S5,
   Supporting Information). This pattern of DsEVs accumulation in the
   liver is consistent with that previously reported.^[ [105]^32 ^]

Figure 3.

   Figure 3
   [106]Open in a new tab

   A91‐DsEVs regulated the polarization of CD4^+ T cells in vivo and
   facilitated their infiltration into the injury site. a) Schematic
   representation of the in vivo experimental design. b) Flow cytometry of
   splenic CD4^+ T cells from mice in different groups to show the
   modulation ability of A91‐DsEVs on CD4^+ T cell polarization at 3 dpi
   (n = 3). c) Fluorescence tracing analysis of Dil‐labeled CD4^+ T cells
   from mice in different groups in major organs 24 h post‐injection (n =
   3). d) Immunofluorescence images to show the distribution of CD4^+ T
   cells at 7 dpi. The numbers of CD4^+ T cells were quantified from 5
   microscopic fields around the injured site for each sample. (Scale Bar
   = 20 µm, n = 5). Data were expressed as mean ± SD. (*p < 0.05, **p <
   0.01, ***p < 0.001, ****p < 0.0001 as assessed by One‐way ANOVA with
   Dunnett's multiple comparisons).

   To further analyze the effect of A91‐DsEVs on the polarization of
   splenic CD4^+ T cells in vivo, we first isolated splenic T cells from
   mice in different groups at 3 dpi and determined the subtype of splenic
   CD4^+ T cells using flow cytometry. The A91‐DsEVs‐treated group
   revealed increased IL‐4^+CD4^+ and CD25^+Foxp3^+CD4^+ T‐cell counts and
   a reduced IFN‐γ^+CD4^+ and IL‐17A^+CD4^+ T‐cell counts
   (Figure [107]3b). The isotype used for flow cytometry is shown in
   Figure [108]S4 (Supporting Information). This result confirmed that
   immunization with A91‐DsEVs in SCI mice could regulate the polarization
   of CD4^+ T cells towards Th2 and Treg ant‐inflammatory phenotypes.

   To demonstrate the recruitment ability of splenic CD4^+ T cells
   stimulated by A91‐DsEVs toward the injury site, we extracted splenic
   CD4^+ T cells from mice following treatment with different groups of
   DsEVs or PBS. These cells were labeled with Dil dye and subsequently
   introduced into SCI mice via tail vein injection. Fluorescence tracing
   analysis performed 24 h post‐injection showed a higher intensity of Dil
   fluorescence within the spinal cord of A91‐DsEVs‐treated mice compared
   to those observed in other groups (Figure [109]3c). The intensity of
   Dil fluorescence was much higher in the spinal cord and lungs than in
   other organs. Immunofluorescence of spinal cord at 7 dpi was performed
   and the result revealed that more CD4^+ T cells were recruited to the
   lesion area in the A91‐DsEVs‐treated group than in other groups
   (Figure [110]3d). These results indicate that CD4^+ T cells stimulated
   by A91‐DsEVs may migrate to the injured site of spinal cord.

   Local spinal cord antigen exposure is the main driver behind the
   recruitment of antigen‐specific CD4^+ T cells to the injury site.^[
   [111]^33 ^] We investigated the homing mechanism by examining the
   expression of activation markers (CD25, CD69) and chemokine receptors
   (CCR3, CCR4, CCR5, CCR6). The highest expression levels of CD25 and
   CD69 were observed in the A91‐DsEVs group (Figure [112]S6a, Supporting
   Information). Within the A91‐DsEVs group, CCR4 gene expression was
   significantly increased. However, no significant differences were
   observed in the expression levels of CCR3, CCR5, or CCR6, among the
   four groups (Figure [113]S6b, Supporting Information).

2.4. A91‐DsEVs Led to the Infiltration of Polarized CD4^+ T Cells and the
Polarization of Macrophages/Microglia in the Lesion Site

   To analyze the local immune microenvironment established by CD4^+ T
   cells, the aggregation of distinct CD4^+ T cells in the spinal cord was
   evaluated at 7 dpi using immunofluorescence (Figure [114] 4a).
   Remarkably, the A91‐DsEVs‐treated group exhibited significantly
   increased aggregation of CD25^+Foxp3^+ CD4^+ T cells and IL‐4^+ CD4^+ T
   cells compared to those seen in the other groups. Conversely, in the
   A91‐DsEVs treated group, the accumulation of IFN‐γ^+ CD4^+ T cells was
   diminished and that of IL‐17A^+ CD4^+ T cells was reduced.

Figure 4.

   Figure 4
   [115]Open in a new tab

   A91‐DsEVs led to the infiltration of polarized CD4^+ T cells and the
   polarization of macrophages/microglia in lesion site. a)
   Immunofluorescence images to show the distribution of distinct subsets
   of CD4^+ T cells in the injury site at 7 dpi. The bar chart depicts the
   proportions of different subtype cells to CD4^+ T cells in different
   groups. (Scale Bar = 100 µm, n = 5) b) ELISA to measure the expression
   levels of major inflammatory cytokines in injured spinal cord from
   different groups at 7 dpi. c) qRT‐PCR results to show the mRNA
   expression level of IL‐4 and IL‐10, IFN‐γ and IL‐17A in the injured
   spinal cord at 7 dpi. d) Immunofluorescence images to show the
   distribution of different subtypes of macrophages/microglia in the
   injury site at 7 dpi. The bar chart showed the ratio of iNOS^+F4/80^+
   cells and CD206^+F4/80^+ cells to F4/80^+ cells in different groups.
   (Scale Bar = 100 µm, n = 5). Data were expressed as mean ± SD. (*p <
   0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 as assessed by One‐way
   ANOVA with Dunnett's multiple comparisons).

   Subsequently, ELISA was utilized to measure the expression levels of
   IL‐4, IL‐10, IFN‐γ, IL‐17A at the injury area at 7 dpi
   (Figure [116]4b). IL‐4 and IL‐10 levels were higher in the
   A91‐DsEVs‐treated group than in the other groups. Conversely, the
   levels of IL‐17A and IFN‐γ were lower in the A91‐DsEVs‐treated group.
   Furthermore, the relative gene expression analysis in the injured
   spinal cord at 7 dpi revealed an upregulation of IL‐4 and IL‐10
   expression in the A91‐DsEVs‐treated group with concurrent
   downregulation of IFN‐γ and IL‐17A expression (Figure [117]4c).
   Collectively, these findings indicate that A91‐DsEVs may contribute to
   the infiltration of Th2 and Treg cells at the lesion site.

   Macrophages/microglia are important responders to SCI and peak at 7
   dpi. The polarization state of macrophages/microglia could be regulated
   by Th2 or Treg cells.^[ [118]^34 ^] Therefore, their phenotype were
   investigated by assessing the expression of iNOS, a marker associated
   with M1‐type macrophages/microglia. In addition, the expression of
   CD206, a marker of M2‐type macrophages/microglia, was evaluated
   (Figure [119]4d). Following A91‐DsEVs treatment, a significant decrease
   in the population of iNOS^+ macrophages/microglia was observed,
   compared to that observed with other treatments. Conversely, the
   population of CD206^+ macrophages/microglia exhibited a notable
   increase in the A91‐DsEVs‐treated group relative to that in the other
   groups. To further elucidate the phenotype of the microglial recruited
   to the injury site, we detected the microglial marker Iba‐1 using
   immunofluorescence. The results demonstrated a marked reduction in the
   population of M1‐type microglia in the A91‐DsEVs treatment group, in
   contrast to the finding in the other three groups. Concurrently, there
   was a notable increase in the proportion of M2‐type microglia in
   A91‐DsEVs‐treated group (Figure [120]S7a and b, Supporting
   Information).

   Furthermore, we performed qRT‐PCR analysis to assess the mRNA
   expression levels of additional markers associated with M1‐type (IL‐1β
   and IL‐6) and M2‐type (Arg‐1 and CD206) microglia/macrophages. The
   results revealed that the A91‐DsEVs treatment group exhibited decreased
   expression of genes related to M1‐type markers, compared to that in the
   other three groups, while displaying elevated expression of genes
   related to M2‐type markers (Figure [121]S7c, Supporting Information),
   suggesting that A91‐DsEVs induced dominant activation of M2
   macrophages/microglia.

2.5. A91‐DsEVs Suppressed the Inflammatory Signals at the Injury Site

   Given the differences in outcomes between treatment with A91‐DsEVs and
   PBS, we conducted protein level analysis using Luminex to assess the
   cytokine and chemokine levels in spinal cord samples from the
   A91‐DsEVs‐treated group and PBS‐treated group at three different time
   points: 3, 7, and 14 dpi (Figure [122] 5a,b). At 3 dpi, the levels of
   CXCL‐16 and IFN‐γ were downregulated in the A91‐DsEVs‐treated group. At
   7 dpi, reduced levels of IL‐1β, IL‐2, IL‐6, and several other cytokines
   were detected in the A91‐DsEVs‐treated group. Furthermore, the
   expression of IL‐4 and IL‐10 increased, while the level of TNF‐α
   decreased in the A91‐DsEVs‐treated group at 14 dpi. In addition, the
   expression of other proteins, such as CCL‐3, CCL‐4, CCL‐20, CCL‐24, and
   CXCL‐10, was diminished in the A91‐DsEVs‐treated group. Notably, the
   expression levels of CCL‐3, CCL‐20, and CCL‐24 were reduced at both 7
   and 14 dpi.

Figure 5.

   Figure 5
   [123]Open in a new tab

   The effect of A91‐DsEVs on the inflammatory signals at the injury site.
   a) Heatmap to show the Luminex results showing the cytokines level in
   spinal cord samples at 3, 7, and 14 dpi. b) Bar graph to present the
   quantitative analysis of the Luminex results, comparing the expression
   levels of these inflammatory markers at 3, 7, and 14 dpi (n = 3). c)
   Heatmap to illustrate the downregulation of genes associated with
   inflammation at 3, 7, and 14 dpi (n = 3, fold change < 0.5, p < 0.05).
   d,e)  Gene set analysis to demonstrate the downregulation of the
   inflammatory response at 7 dpi (d) and 14 dpi (e). f) GO enrichment
   analysis to show the downregulation (top 20 terms displayed) of genes
   associated with immune and inflammatory responses at 7 dpi. Data were
   expressed as mean ± SD. (*p < 0.05, **p < 0.01, ***p < 0.001, ****p <
   0.0001 as assessed by two‐tailed t‐test).

   RNA sequencing analysis performed on the injured spinal cord
   demonstrated that, compared to the PBS treatment, the A91‐DsEVs
   treatment led to the downregulation of several genes associated with
   inflammation. The heatmap visualized the downregulated expression of
   inflammatory‐related genes, including C3, IL‐1α, IL‐6, Clrp1b, and Tnf
   among others, at 3, 7, and 14 dpi (Figure [124]5c). Csf‐1, a crucial
   regulator of macrophage proliferation and migration, showed decreased
   expression at 7 and 14 dpi. Gene set enrichment analysis revealed a
   weakened inflammatory response in the A91‐DsEVs‐treated group at 7
   (Figure [125]5d) and 14 dpi (Figure [126]5e). Furthermore, the Gene
   ontology (GO) analysis demonstrated the downregulation of genes
   associated with the immune response and inflammatory processes in the
   A91‐DsEVs‐treated group at 7 (Figure [127]5f) and 14 dpi
   (Figure [128]S8, Supporting Information). In addition, the Kyoto
   Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis
   exhibited a downregulation of the TNF‐α signaling pathway, JAK‐STAT
   signaling pathway and IL‐17 signaling pathway in the A91‐DsEVs‐treated
   group, compared to those in the PBS‐treated group at 7 dpi
   (Figure [129]S9a, Supporting Information). At 14 dpi, cytokine‐cytokine
   receptor interactions, the toll‐like receptor signaling pathway, and
   MAPK signaling pathway were downregulated (Figure [130]S9b, Supporting
   Information).

   These data suggest that A91‐DsEVs attenuate neuroinflammation after
   SCI, as evidenced by the downregulation of multiple inflammatory
   signals.

2.6. A91‐DsEVs Promoted Axon Regrowth and Neural Survival after SCI

   We evaluated the therapeutic effects of A91‐DsEVs. Our findings
   demonstrated a higher fluorescence density of neurofilament (NF) and a
   lower fluorescence density of glial fibrillary acidic protein (GFAP) in
   the A91‐DsEVs‐treated group at 35 dpi (Figure [131] 6a,b).
   Immunohistochemistry (IHC) staining of neuronal nuclei (NeuN) revealed
   an increased numbers of surviving neurons in the spinal cord of
   A91‐DsEVs‐treated group at 35 dpi (Figure [132]6c). Nissl staining
   images demonstrated an increase in the number of Nissl‐positive cells
   in the A91‐DsEVs‐treated group at 35 dpi (Figure [133]6d). The TUNEL
   staining results at 7 dpi indicated a reduction in nerve cell apoptosis
   rate in the A91‐DsEVs‐treated group (Figure [134]6e). These findings
   suggest that the immune environment induced by A91‐DsEVs promotes axon
   regrowth and contributes to neural survival after SCI.

Figure 6.

   Figure 6
   [135]Open in a new tab

   A91‐DsEVs promoted axon regrowth and neuronal survival after SCI. a)
   Representative immunofluorescence images of NF (green) and GFAP (red)
   in the spinal cord lesions of mice treated with PBS, DsEVs, OVA‐DsEVs
   or A91‐DsEVs at 35 dpi (scale bar = 500 µm in low magnification, scale
   bar = 100 µm in high magnification, n = 5). b) A quantitative analysis
   was performed to determine the mean fluorescence density of NF and GFAP
   at the center of the injury site (n = 5). c) Representative IHC images
   of NeuN in injured spinal cord from different groups at 35 dpi (Scale
   bar = 100 µm, n = 5) Bar graphs showed the mean numbers of survival
   neurons per visual field. d) Images of Nissl staining in injured spinal
   cord from different groups at 35 dpi (Scale bar = 100 µm, n = 5). The
   arrow in (d) points to the Nissl‐positive cells. Bar graphs showed the
   mean numbers of Nissl‐positive cells per visual field. In (c) and (d),
   five stained sections were randomly selected from each experimental
   mouse to evaluate the average number of surviving neurons in the
   anterior horns of spinal cord. e) Representative TUNEL staining images
   at 7 dpi (Scale bar = 100 µm, n = 5) and quantitative analysis of the
   fluorescence density (n = 5). The density of TUNEL was measured from 5
   microscopic fields for each sample. f) The levels of BDNF and NT‐3
   detected by ELISA and their gene expression levels detected by qRT‐PCR
   in the injured spinal cord from different groups at 7 dpi. g) Pathway
   enrichment analysis of genes that were significantly upregulated (top
   20 terms are displayed) in the A91‐DsEVs‐treated group versus the
   PBS‐treated group at 14 dpi. Data were expressed as mean ± SD. (*p <
   0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 as assessed by One‐way
   ANOVA with Dunnett's multiple comparisons).

   Moreover, we evaluated the expression levels of two major nerve growth
   factors, NT‐3 and BDNF, using ELISA and quantified their gene
   expression levels using qRT‐PCR at 7 dpi (Figure [136]6f). The protein
   expression levels of NT‐3, BDNF, and their associated genes were higher
   in the A91‐DsEVs‐treated group than in the comparison groups. Moreover,
   RNA sequencing analysis at 14 dpi (Figure [137]6g) revealed an
   upregulation of genes associated with axon guidance, axonogenesis,
   nervous system development, neuron projection morphogenesis, chemical
   synaptic transmission, synaptic vesicle endocytosis, and a positive
   regulation of synapse assembly in the A91‐DsEVs‐treated group.

2.7. A91‐DsEVs Promoted Remyelination after SCI

   To investigate whether A91‐DsEVs could affect remyelination following
   SCI, Luxol Fast Blue (LFB) staining and TEM evaluations were performed.
   LFB staining revealed the highest density of myelin in the spinal cord
   tissue of the A91‐DsEVs‐treated group at 35 dpi (Figure [138] 7a). TEM
   analysis was performed at 35 dpi (Figure [139]7b), and the results were
   consistent with those of LFB staining. Quantification of the nerve
   fiber‐specific morphological parameters was conducted, including the
   G‐ratio, myelinated axon area and diameter, and myelin sheath thickness
   measurements (Figure [140]7c). A smaller G‐ratio reflects better
   remyelination. The G‐ratio of the A91‐DsEVs‐treated group was smaller
   than those of the other groups. In addition, compared to the
   OVA‐DsEVs‐treated, DsEVs‐treated, and PBS‐treated groups, the
   A91‐DsEVs‐treated group showed increased myelinated axon diameters and
   areas. Furthermore, the mRNA expression levels of MBP and myelin
   oligodendrocyte glycoprotein (MOG) were assessed. The levels of MBP and
   MOG mRNA were higher in the A91‐DsEVs‐treated group than in the other
   groups at 35 dpi (Figure [141]7d).

Figure 7.

   Figure 7
   [142]Open in a new tab

   A91‐DsEVs promoted remyelination after SCI. a) Representative LFB
   staining images of spinal cord tissue from different groups at 35 dpi.
   The transverse slices of rostral spinal cord segments (5 mm in length,
   5 mm far from the injury epicenter) were used for LFB staining. b)
   Representative images of myelinated axonal regrowth captured by TEM at
   35 dpi (Scale bar = 2 µm, n = 5). c) Quantitative analysis of the
   myelination, including G‐ratio, myelin sheath thickness, myelinated
   axon area, and myelinated axon diameter at 35 dpi (n = 5). d) The mRNA
   expression levels of MBP and MOG measured by qRT‐PCR in the injured
   spinal cord from different groups at 35 dpi, n = 5). e) RNA sequencing
   results showed the upregulated genes associated with neuron protection,
   axonal growth and remyelination in the A91‐DsEVs‐treated group versus
   the PBS‐treated group at 7 and 14 dpi. Data were expressed as mean ±
   SD. (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 as assessed by
   One‐way ANOVA with Dunnett's multiple comparisons).

   RNA sequencing analysis revealed upregulation of numerous genes
   associated with neuron protection, axonal regrowth, and remyelination
   in the A91‐DsEVs‐treated group at 7 and 14 dpi (Figure [143]7e). These
   data and the results mentioned above (Figure [144]6) substantiated the
   hypothesis that A91‐DsEVs treatment could promote neuronal survival and
   myelination of regenerated axons after SCI.

2.8. A91‐DsEVs Improved the Locomotor Function of SCI Model Mice

   Recovery of hind limb movement was achieved after A91‐DsEVs treatment
   at 35 dpi (Figure [145]S10, Supporting Information). We used the Basso
   mouse score (BMS) to evaluate motor functional recovery in the SCI
   model mice. All animals exhibited complete hind limb paralysis, with a
   BMS score of 0 points at 1 dpi. A91‐DsEVs‐treated mice achieved
   significant functional recovery, with a mean BMS of 5.00±0.63 pionts at
   35 dpi. In contrast, the BMS was much lower in the other groups
   (OVA‐DsEVs: 3.83±0.75, DsEVs: 3.50±0.55, PBS: 2.50±0.55) (Figure [146]
   8a). The Catwalk footprint analysis revealed that the A91‐DsEVs‐treated
   group achieved better hind limb motor function recovery
   (Figure [147]8b), with a larger maximum contract area (Figure [148]8c)
   and much higher regularity index (improved walking steps)
   (Figure [149]8d), compared to those achieved by the other groups.

Figure 8.

   Figure 8
   [150]Open in a new tab

   A91‐DsEVs contributed to functional recovery after SCI. a) BMS scores
   and weight changes of mice from different groups (*p < 0.05, A91‐DsEVs
   versus PBS treatment; #p < 0.05, A91‐DsEVs versus DsEVs treatment; &p <
   0.05 A91‐DsEVs versus OVA‐DsEVs) (*p < 0.05, **p < 0.01, ***p < 0.001,
   ****p < 0.0001 as assessed by Two‐way RM ANOVA with Dunnett's multiple
   comparisons). b) Representative footprint images and time view of mice
   from different treatment groups at 35 dpi. c) The 3D footprint
   intensity charts and measurement of maximum contact area of the left
   hind limb at 35 dpi. d) The footfall patterns graph and regularity
   index of mice from different groups. Data were expressed as mean ± SD.
   (*p < 0.05, **p <0 .01, ***p < 0.001, ****p < 0.0001 as assessed by
   one‐way ANOVA with Dunnett's multiple comparisons).

   To assess the biotoxicity of A91‐DsEVs, histopathological examination
   of the lung, liver, and other major organs in the mice was performed
   using HE staining at 3 dpi. The results showed normal morphology and
   structure of the major organs in the A91‐DsEVs‐treated group, without
   any apparent signs of injury (Figure [151]S11a, Supporting
   Information). Lactate dehydrogenase (LDH) and alanine aminotransferase
   (ALT) activity levels were evaluated to determine liver function
   (Figure [152]S11b, Supporting Information). These findings indicated
   that the liver function in mice treated with A91‐DsEVs was comparable
   to those in the other groups. These results suggested that A91‐DsEVs
   had no significant biotoxicity in mice.

3. Discussion

   Excessive activated inflammation following CNS injury leads to
   irreversible dysfunction by disrupting neural circuitry and impairing
   tissue repair mechanisms.^[ [153]^35 ^] Establishing a protective
   immune microenvironment to inhibit excessively activated inflammation
   may help promote functional recovery after SCI. Therefore, in this
   study, we developed a vaccine platform using DsEVs loaded with an APL
   peptide cargo to induce a protective immune microenvironment at the
   injury site and facilitate functional recovery after SCI.

   Recent studies have emphasized the use of therapeutic vaccines against
   SCI. DNA‐, RNA‐ and protein‐based vaccines targeting myelin‐associated
   inhibitors and their associated receptors have been used to treat
   SCI.^[ [154]^36 , [155]^37 , [156]^38 ^] Cell‐based vaccination offers
   a novel approach to the treatment of SCI. As therapeutic vaccines, stem
   cells promote recovery by inhibiting glial scar expansion and reducing
   inflammation. However, the immunosuppressive effects of these stem
   cells may increase tumor risk by suppressing the immune response.^[
   [157]^39 ^] Antigen‐presenting cells such as dendritic DCs can induce
   neuroprotection through T cell‐dependent immune responses. This finding
   suggests the potential of using antigen‐specific DCs to regulate
   neuroinflammation after SCI.^[ [158]^40 ^]

   Recent studies have highlighted the advantages of EVs as novel
   therapeutic vaccines. EVs‐based vaccines have several advantages over
   cell‐based vaccines in terms of their preparation, immunogenicity,
   storage, safety, and component control.^[ [159]^41 ^] Nevertheless, the
   use of EVs is associated with dispersed distribution and low
   efficiency, attributed to the following reasons. 1) The non‐specific
   biodistribution of EVs, leading to their accumulation in unintended
   organs such as the liver, spleen, lungs, kidneys, and pancreas, poses a
   significant challenge for their effective utilization. 2) EVs may be
   taken up by non‐targeted cells.^[ [160]^42 ^]

   The application of DsEVs can enhance the efficacy of immunotherapeutic
   interventions.^[ [161]^43 ^] DsEVs offer the unique ability to directly
   deliver peptide cargo to lymphoid organs or antigen‐presenting cells
   where it can activate antigen‐specific T cells, which may help them
   reach their target.^[ [162]^28 ^] Moreover, because DsEVs are rich in
   MHC II molecules, they can directly upregulate MHC‐II expression on the
   surface of DCs. In addition, DsEVs as carriers can avoid enzyme
   degeneration of antigen peptides in the circulation, thereby
   contributing to the use of a lower concentration of peptide and
   promoting the efficacy of delivery to DC and T cells.^[ [163]^44 ^]
   Finally, DsEVs follow the endocytic pathway and reach multivesicular
   endosomes, which are typically targeted for lysosomal degradation after
   exerting their function.^[ [164]^45 ^] This represents a possible fate
   for DsEVs after internalization and completion of their functional
   effects.

   MBP[87‐99]A^91 is an APL derived from a specific segment (amino acids
   87–99) of MBP, in which the lysine residue at position 91 is replaced
   with alanine. This site has pivotal significance in mediating TCR
   interactions, as validated by previous studies.^[ [165]^46 ^] Replacing
   lysine with alanine in the peptide sequence modulates TCR‐binding
   kinetics.^[ [166]^47 , [167]^48 ^] As a partial TCR agonist,
   MBP[87‐99]A^91 downregulates the production of pro‐inflammatory
   cytokines, increases the levels of Th2 cytokines, and generates a
   microenvironment with anti‐inflammatory features.^[ [168]^49 ^] In our
   experiment, DsEVs were loaded with MBP[87‐99]A^91 peptide. These
   A91‐DsEVs, containing MHC II and costimulatory molecules, presented the
   MBP[87‐99]A^91 peptide‐MHC II complex directly to CD4^+ T cells,
   thereby inducing polarization of CD4^+ T cells toward the Th2
   phenotype.

   After SCI, the local tissue structures are disrupted, leading to
   antigen exposure at the site.^[ [169]^50 ^] MBP, a crucial constituent
   of the myelin sheath, is frequently exposed after SCI and triggers
   immune cell activity.^[ [170]^51 ^] A91‐DsEVs stimulated the conversion
   of CD4^+ T cells into antigen‐specific Th2 and Treg cells, enabling
   their recruitment to injury sites. CCR4 is a pivotal chemokine receptor
   that plays a significant role in the recruitment of both Th2 and Treg
   cells.^[ [171]^52 ^] Studies have demonstrated that increased CCR4
   expression in Treg cells promotes their migration to inflammatory sites
   and exerts immunosuppressive effects.^[ [172]^53 , [173]^54 ^]
   Moreover, a recent study indicated that CCR4 antagonists can
   selectively inhibit the migration of Th2 cells to inflamed tissues by
   blocking highly expressed CCR4 receptors on these cells.^[ [174]^55 ^]
   In our study, the upregulation of CCR4 expression in antigen‐specific T
   cells, stimulated by A91‐DsEVs, suggests that CCR4 is involved in
   regulating the recruitment of Th2 and Treg cells to the injury site.

   DsEVs are non‐immunogenic or possess very low immunogenicity and tend
   to be well‐tolerated by the immune system.^[ [175]^41 ^] In fact,
   MBP[87‐99]A^91 used herein has also been shown to be a non‐immunogenic
   peptide that can mediate the immune response under pathological
   conditions, and change the polarity of inflammatory T cells, without
   inducing immunogenicity.^[ [176]^56 , [177]^57 , [178]^58 , [179]^59 ^]

   To load the peptide onto DsEVs, we used an indirect method involving
   antigen feeding into DCs and subsequent collection of the supernatant
   for DsEVs isolation. This loading strategy ensures a favorable
   interaction between DsEVs and T cells, as it guarantees the association
   of the peptide with major MHC molecules and, as a result, fosters
   peptide presentation and T‐cell activation.^[ [180]^60 , [181]^61 ^] To
   enhance the loading efficiency of the MBP[87‐99]A^91 peptide in DsEVs,
   we fused it to CPP. CPPs are peptides that facilitate the delivery of
   cargo molecules into the cytosol. The CPP‐peptide cargo is delivered to
   the cytosol by direct interactions with the components of the cell
   surface or the plasma membrane. TAT[47‐57], a CPP derived from HIV TAT
   protein, is one of the most commonly used CPPs for intracellular
   peptide delivery.^[ [182]^62 ^‐ [183]^6‐4 ^] Previous studies have
   demonstrated that TAT[47‐57]‐linked peptide cargos possess the
   increased ability to translocate intracellularly into DCs and extend
   the presentation of MHC‐peptide complexes to antigen‐specific T cells,
   facilitating enhanced immune responses without causing biotoxicity.^[
   [184]^65 , [185]^66 ^] Therefore, in this study, we used the
   TAT[47‐57]‐linked peptide to upregulate the loading amount of
   MBP[87‐99]A^91 in DC, which contributed to promoting the loading
   efficiency of the MBP[87‐99]A^91 peptide in DsEVs.

   The balance between different subtypes of CD4^+ T cells is crucial for
   neuroprotection.^[ [186]^67 ^] In our study, we demonstrated that
   immunization with A91‐DsEVs significantly upregulated the levels of Th2
   and Treg cells and downregulated the level of Th1 and Th17 cells in
   vivo. The Th1 and Th17 cells primarily induce spinal cord inflammation,
   causing tissue damage. IFN‐γ is a prototypical Th1 cytokine. The
   continuous expression of Th1 cytokines at the lesion site may lead to
   neuronal damage and demyelination.^[ [187]^7 ^] They are also capable
   of activating M1 macrophages and aggravating lesions in SCI.^[ [188]^4
   ^] Th17 produces cytokines (e.g., IL‐17A, IL‐6, and IFN‐γ), leading to
   the abnormal neuroinflammatory response including the excessive
   activation of microglia and the recruitment of other immune cell types,
   which are involved in neuron or axon injury and demyelination.^[
   [189]^68 , [190]^69 ^]

   IL‐4 is a Th2‐cell‐type cytokine that protects against infection and
   promotes tissue regeneration. The transition from the Th1 to Th2
   response is dominated by anti‐inflammatory factors, which create a
   beneficial microenvironment, induce the polarization of microglia and
   macrophages, and facilitate the protection of neurons.^[ [191]^70 ^]
   However, the presence of Th2 cells in the CNS may be associated with
   certain risks. Therefore, the protective role of IL‐4 may be
   detrimental to the developing brain.^[ [192]^71 ^] Previous studies
   have demonstrated that neuroimmune responses in fetuses and newborns
   markedly differ from those observed in adults.^[ [193]^71 , [194]^72 ^]
   Moreover, a pathogenic Th2 cell subset revealed by single‐cell RNA ‐seq
   may be correlated with disease severity.^[ [195]^73 ^] In general, Th2
   cells play a positive role.^[ [196]^74 ^] In fact, Th2 cells have been
   shown to promote recovery in SCI and other neurological disorders by
   secreting protective cytokines.^[ [197]^8 , [198]^9 , [199]^10 ^]

   Treg cells can inhibit inflammatory response and are thought to be
   targets for balancing the immune environment and improving functional
   outcomes in CNS trauma.^[ [200]^75 ^] It has been shown that Treg
   cell‐derived cytokines such as TGF‐β and osteopontin may help maintain
   immune homeostasis by suppressing excessive immune responses.^[
   [201]^76 , [202]^77 ^] IL‐10 is a key immunesuppressive cytokine
   produced by regulatory T‐cells.^[ [203]^78 ^] Accumulating evidence has
   demonstrated that Tregs alleviate neural injury, promote neurogenesis
   and protect the blood‐brain barrier via IL‐10.^[ [204]^79 ^] Studies
   have also shown that an increased level of IL‐10 can induce an M2
   subtype bias in macrophages/microglia. Tregs mediate the
   differentiation of macrophages toward an M2‐like state by increasing
   the levels of the inflammation‐dampening cytokines under inflammatory
   conditions.^[ [205]^80 , [206]^81 ^]

   The polarization state of macrophages/microglia are regulated by Th2 or
   Treg cells.^[ [207]^81 ^] Therefore, the increased number of
   M2‐polarized macrophages/microglia induced by A91‐DsEVs may be
   attributed to polarized T cells. A series of studies has shown that the
   microenvironment established by M2‐polarized macrophages/microglia may
   release nerve growth factors and promote immune protection and axonal
   growth in SCI.^[ [208]^34 ^] In addition, the multiple downregulated
   pro‐inflammatory cytokines, such as IL‐6, TNF‐α, IL‐1β, suggested that
   A91‐DsEVs effectively established a neuroprotective immune environment
   and inhibited excessive activated inflammatory response after SCI.

   The expression of IL‐4 and IL‐10 at the injury site was elevated at
   both 7 and 14 dpi following the A91‐DsEVs injection. These findings
   suggested a sustained effect of A91‐DsEVs treatment. The expression
   levels of CXCL‐16 and IFN‐γ were downregulated in A91‐DsEVs‐treated
   group at 3 dpi, suggesting that A91‐DsEVs may inhibit neuroinflammation
   in the early phase. This finding may be attributed to the activation
   and infiltration of peripheral DCs and CD4^+ T cells into the injured
   area, following injection of A91‐DsEVs, thereby regulating the immune
   response. CXCL‐16 is an inflammatory cytokine associated with early
   stage neutrophil activation; its levels could be increased by the
   release of pro‐inflammatory cytokines such as IFN‐γ.^[ [209]^82 ^]

   Our results also showed that CCL‐3, CCL‐20, and CCL‐24 levels were
   decreased in the A91‐DsEVs‐treated group. CCL‐3 has been shown to be
   upregulated by the MAPK and NF‐κB signal pathways.^[ [210]^83 ^] CCL‐20
   may aggravate neuroinflammation via Th17 cell recruitment.^[ [211]^84
   ^] CCL‐24 is a chemokine that promotes immune cell trafficking and
   regulates inflammatory responses. Treatment with an anti‐CCL‐24
   monoclonal antibody could exert an immunosuppressive effect.^[ [212]^85
   ^] The levels of these three cytokines were decreased in the
   A91‐DsEVs‐treated group at 7 and 14 dpi. These results provide further
   evidence that A91‐DsEVs may induce a protective immune response after
   CNS trauma.

4. Conclusion

   This study demonstrated that A91‐DsEVs induce antigen‐specific CD4^+ T
   cells homing to the injury site, establish a protective immune
   microenvironment, and promoted functional motor recovery in the SCI
   mice. Our results suggest that DsEVs vaccination strategy may help
   treat CNS truama or inflammatory diseases.

5. Experimental Section

APLs

   Peptide synthesis grade reagents were purchased from QYAOBIO (China).
   The peptide sequence for MBP[87‐99]A^91 was VHFFANIVTPRTP. The peptide
   sequence for a specific segment of OVA, consisting of amino acids
   323–339, was ISQAVHAAHAEINEAGR. The peptide sequence for TAT[47‐57] was
   YGRKKRRQRRR. Prior to final deprotection and cleavage, the
   MBP[87‐99]A^91 peptide segment was labeled with fluorescein FITC
   (5‐FITC‐(Acp)‐VHFFANIVTPRTP‐YGRKKRRQRRR). The peptides were purified
   using high‐performance liquid chromatography, resulting in a purity
   exceeding 95%.

Animals

   Six‐eight‐week‐old C57BL/6 female mice were kept in a 12‐h light/dark
   cycle at 21±2 ˚C and 50% humidity, and they were allowed free access to
   food and water. Female mice were chosen for the whole experiments due
   to fewer post‐operative complications and easier manual bladder
   expressions after experimental paralysis compared to male mice. In
   addition, no sex differences were observed in lesion size or locomotor
   recovery after SCI in mice.^[ [213]^86 ^] This study received approval
   from the Ethics Committee (no. SYDW2019–111) of the Second Affiliated
   Hospital of Harbin Medical University (Harbin, China) in accordance
   with Chinese experimental animal legislation. All the authors
   compliance with all relevant ethical regulations.

Cell culture

   DCs were generated through the culture of bone marrow cells isolated
   from femurs of C57BL/6 mice (six‐eight‐week‐old) in complete medium
   (Gibco, USA) as previously described.^[ [214]^87 ^] Cells were cultured
   in complete medium added with murine GM‐CSF and IL‐4 (20 ng mL^−1,
   Pepro Tech, USA) and refreshed every two days. On day 6, the cells were
   harvested for further analysis. Next, the medium was supplemented with
   TAT[47‐57]‐MBP[87‐99]A^91 or TAT[47‐57]‐ OVA antigen (1 mg mL^−1).
   After 24 h, exosome‐free complete medium was used. After 24 h of
   incubation, the supernatants were collected for further analysis.

DsEVs Isolation

   DsEVs isolation was performed as previously described.^[ [215]^88 ,
   [216]^90 ^] As briefly, the supernatant was purified by sequential
   centrifugation (4˚C, 500 g for 5 min; 2000 g for 20 min; 10,000 g for
   30 min) to remove cells and debris. Subsequently, the purified
   supernatant underwent 0.2 µm filtration and ultra‐centrifugation for
   further refinement (4˚C, 120,000 g for 90 min; Himac CP100NX
   ultracentrifuge, Hitachi Koki, Japan). The pellet was then washed in
   PBS and ultra‐centrifugated again (4˚C, 120,000 g for 90 min). Finally,
   The pellet was resuspended in PBS and stored at −80˚C until use.^[
   [217]^91 ^]

DsEVs Characterization

   Western blotting: Primary antibodies: anti‐CD63 (ab217345, Abcam, USA),
   anti‐Alix (ab275377, Abcam), anti‐TSG101 (ab125011, Abcam),
   anti‐Calnexin (ab313243, Abcam). The membranes were developed using ECL
   kit (Yeasen, China).

   TEM: DsEVs sample was examined by TEM (Hitachi TEM system, Japan).

   NTA: 10 µL DsEVs suspension was loaded into the sample chamber of
   ZetaVIEW S/N 252 (Particle Metrix GmbH, Germany). Data analysis was
   performed with software (ZetaView 8.04.02, Germany).

   Detection of the loading rate of TAT[47‐57]‐MBP[87‐99]A^91 into DsEVs:
   After co‐culturing the FITC‐labeled peptide with DCs for 24 h, DCs were
   collected and stained with phalloidin (Actin‐Tracker Red‐594, C2205S,
   Beyotime) and 4′,6‐diamidino‐2‐phenylindole (DAPI, C1005, Beyotime,
   China). After the isolation of DsEVs.

   DsEVs internalized by DCs and CD4^+ T cells in vitro: DsEVs were
   labelled with the Dil (Solarbio life sciences, China) and then
   co‐cultured with DCs or CD4^+ T cells for 24 h. Cells were fixed and
   followed by staining with phalloidin (Actin‐Tracker Green‐488, C2201S,
   Beyotime) and DAPI (C1005, Beyotime). The images were obtained by
   scanning confocal fluorescence microscopy.

MLR

   DCs were generated as mentioned above. A total of 1 × 10^4 DCs were
   co‐cultured with 1 × 10^5 splenic T cells. A91‐DsEVs, OVA‐DsEVs, DsEVs
   or the same amount of PBS were added in the MLR in different groups.

CCK‐8 Assay

   The cell concentration to 5000 mL was first adjusted. Cell suspension
   was treated by different DsEVs or PBS and 100 µL of cell suspension was
   added to each well of a 96‐well plate. On day 3 and day 5 of MLR
   culture, the culture medium was replaced and added with 10 µL of CCK‐8
   solution. Following a 2‐h inubation, the absorbance at 450 nm was
   measured using a microplate reader (Bio Rad iMak, Japan).

SCI Model

   The surgery procedures were the same as described elsewhere.^[ [218]^92
   ^] Briefly, the mice were divided randomly into four groups. Under
   general anesthesia (1.25% 2,2,2‐tribromoethanol, 0.02 Ml g^−1, Nanjing
   Aibei Biotechnology, China) and sterile conditions, T10 spinal cord
   segment was exposed by laminectomy. Then a weight‐drop device was
   utilized to induce spinal cord contusion injury (RWD Life Science Corp,
   China). A 5.0 g rod was dropped from a height of 5.0 cm, striking the
   exposed spinal cord. Following the impact, the muscle and skin layers
   were meticulously sutured in sequential order. Two hours after SCI,
   mice were injected with 100 µg A91‐DsEVs, OVA‐DsEVs, DsEVs dissolved in
   100 µL PBS or the same amount of PBS via the tail vein.

In Vivo Tracing Analysis

   First, A91‐DsEVs, OVA‐DsEVs, and DsEVs using DiR dye was labeled
   according to the manufacturer's instructions. The specific procedure
   was as follows: 10 µL of DiR working solution was added to 100 µL of
   each vesicle group, and they were incubated at 37 °C for 30 min.
   Subsequently, excess dye was removed by washing. These DiR‐labeled
   vesicles were then injected into the spinal cord‐injured mice via tail
   vein injection (100 µL mouse^−1). After 6 h, various organs including
   the heart, liver, spleen, lungs, kidneys, and spinal cord were
   collected from each group of mice, and the distribution of DiR
   fluorescence was observed using the Living Image software (version 4.4,
   USA).

   The splenic CD4^+ T cells were isolated using the mouse CD4^+ T cell
   isolation kit (Miltenyi Biotec, Germany) at 3 dpi and then stained with
   Dil. Dil‐labeled CD4^+ T cells were then injected into SCI mice through
   peripheral injection (tail vein, 3×10^6 in 100 µL PBS). After 24 h, the
   biodistribution of Dil‐labeled CD4^+ T cells was photographed using an
   in vivo imaging system (IVIS, Night OWL II LB983, Germany), including
   fluorescence images of the liver, heart, spleen, lung, kidney, and
   spinal cord. Bruker MI SE software was used for image analysis.

ELISA

   The CIITA expression level of DCs was detected by ELISA kit (JingKang
   Bio, China) after co‐culturing with different groups of DsEVs or PBS
   for 24 h. The expression level of IL‐4, IL‐10, IFN‐γ, IL‐17A, NT‐3,
   BDNF in supernatant of MLR and spinal cord tissue were detected by
   ELISA kit (JingKang Bio, China). After sacrificing SCI mice at 7 dpi,
   their spinal cords were rapidly collected and stored at −80 °C for
   protein isolation. Segments of injured spinal cord (10 mm in length)
   were homogenized in RIPA buffer. The samples were maintained on ice for
   30 min with intermittent vortexing every 5 min, followed by sonication
   in three pulse bursts for 10 s and repeat vortexing. Then the samples
   were then centrifuged at 14,000 g at 4 °C for 15 min and the
   supernatant for the ELISA was collected. Protein concentration in the
   supernatant was quantified using BCA assay. The absorbance was measured
   with a microplate reader at 450 nm (Bio Rad iMak, Japan).

Tissue Processing

   The whole segments (10 mm in length, centered on the injury epicenter)
   and the rostral spinal cord segments (5 mm in length, 5 mm far from the
   injury epicenter) were separated and post‐fixed in paraformaldehyde
   overnight at 4 ˚C. The whole segments were washed dehydrated in a
   gradient sucrose solution (10%, 20%, and 30%). Samples were frozen and
   sectioned into 5 µm‐thick transverse and longitudinal sections. The
   rostral spinal cord segments were washed, embedded in paraffin wax, and
   subsequently sectioned into transverse slices with a thickness of 5 µm.

Immunofluorescence

   Sample sections or cells were incubated with primary antibodies against
   IFN‐γ (ab280353, Abcam), IL‐4 (bs‐0581R, Bioss antibodies), IL‐17A
   (1:200, ab302922, Abcam), Foxp3 (12653S, Cell Signaling Technology,
   USA), CD4 ( 67786‐1‐Ig, Proteintech, China), iNOS (22226‐1‐AP,
   Proteintech, China), CD206 (24595S, Cell Signaling Technology), F4/80
   (71299S, Cell Signaling Technology, USA), Iba‐1 (ab283346, Abcam), NF
   (2837S, Cell Signaling Technology), GFAP (60190‐1‐Ig, Proteintech)
   overnight at 4 ˚C. After incubation with the secondary antibodies
   including Goat anti‐rat IgG(H+L) Alexa Fluor 647 (ab150159, Abcam),
   Goat anti‐mouse IgG(H+L) Coralite 594 (SA00013‐3, Proteintech) and Goat
   anti‐rabbit IgG(H+L) FITC (SA00003‐2, Proteintech) for 2 h in dark,
   sections were washed with PBS, and the nuclei were counterstained with
   DAPI (Beyotime). Fluorescent images were captured using a confocal
   laser scanning microscope (Zeiss LSM800, Germany). The numbers of CD4^+
   T cells and different subsets of T cell were achieved by were
   quantified from 5 microscopic fields around the injured site for each
   sample. The numbers of iNOS^+F4/80^+ cells, CD206^+F4/80^+ cells,
   F4/80^+ cells, iNOS^+Iba‐1^+ cells, CD206^+Iba‐1^+ cells and Iba‐1^+
   cells were randomly counted from 5 microscopic fields in the injured
   site for each sample.

TUNEL Staining

   The transverse sections of rostral spinal cord segments (5 mm in
   length, 5 mm far from the injury epicenter, 5 µm in thickness) were
   used for TUNEL staining according to the manufacturer's protocol. The
   fluorescent images were captured using a confocal laser scanning
   microcopy (Zeiss LSM800).

IHC

   The transverse slices of rostral spinal cord segments (5 mm in length,
   5 mm far from the injury epicenter) were used for IHC. After dewaxed,
   rehydrated and antigen retrieval, the tissues were blocked, then
   incubated with a rabbit anti‐NeuN antibody (24307S, Cell Signaling
   Technology) at 37 ˚C for 2 h and goat anti‐rabbit IgG secondary
   antibody (1:100; Zhongshan Jinqiao, China), followed by DAB staining
   (Zhongshan Jinqiao). Five stained sections were randomly selected from
   each experimental mouse to evaluate the average number of surviving
   neurons in spinal cord.

LFB Staining

   The transverse slices of rostral spinal cord segments (5 mm in length,
   5 mm far from the injury epicenter) were used for LFB staining. A
   series of sections were dewaxed to 95% ethanol and then soaked in LFB
   (Solarbio Life Science, China) and then washed with 95% ethanol. Then
   the sections were differentiated by LFB differentiation solution
   (Solarbio Life Science) for 15s and in 70% ethanol for 30s. After
   rinsed in distilled water, sections were observed under the microscope
   until the gray and white matter was clear. Then crystal violet staining
   solution (Solarbio Life Science) were used for re‐dyeing the section.
   After washed with water, sections were dehydrated and then sealed with
   neutral gum.

Nissl Staining

   The transverse slices of rostral spinal cord segments (5 mm in length,
   5 mm far from the injury epicenter) were used for Nissl staining. A
   series of sections were dewaxed to 80% ethanol and then treated with
   Nissl dye (Solarbio Life Science) (56 ˚C, 1.5 h), then washed,
   differentiated in 95% alcohol, dehydrated and then sealed. Five stained
   sections were randomly selected from each experimental mouse to
   evaluate the average number of Nissl positive cells.

TEM Experiments

   The whole segments (5 mm in length, centered on the injury epicenter)
   of spinal cords were fixed in 2.5% glutaraldehyde for 24 h at 4 ˚C,
   followed by transferred into 2% osmic acid for 2 h. Then the tissues
   were embedded after dehydration. After solidification, the spinal cord
   segments were cross sectioned (50 nm thickness). The thick sections
   were stained with 0.5% toluidine blue and imaged using a transmission
   electron microscopy (Hitachi TEM System, Japan).

Flow Cytometry

   For cell surface marker staining, cells were stained with antibody on
   ice for 30 min. For intracellular staining, cells were fixed and
   permeabilized by Intracellular Fixation & Permeabilization Buffer Set
   (Thermofisher Scientific, USA) for 30 min, followed by antibody
   incubation recommended concentration for 30 min. Finally, data was
   acquired using Apogee flow system (A60‐Universal, UK) and
   FlowJo_v10.8.1 software. The following fluorochrome‐labeled antibodies
   and staining kits were used for flow cytometry according to the
   manufacturers’ protocols: anti‐mouse‐CD3 FITC (45‐0031‐82, Thermofisher
   Scientific, USA), anti‐mouse‐CD3 PerCP‐Cyanine5.5 (11‐0031‐82,
   Thermofisher Scientific), anti‐mouse‐CD4 PE (12‐0041‐82, Thermofisher
   Scientific), anti‐mouse‐ IFN‐γ FITC (Cat# 11‐7311‐82, Thermofisher
   scientific, USA), anti‐mouse‐ IL‐4 APC (Cat# 17‐7041‐82, Thermofisher
   Scientific), anti‐mouse‐ IL‐17A PerCP‐Cyanine5.5 (12‐0041‐82,
   Thermofisher Scientific), mouse regulatory T cell staining kit
   (88–8111, Thermofisher Scientific), Rat IgG2a kappa Isotype Control
   (eBR2a) PerCP‐Cyanine5.5 (45‐4321‐80, Thermofisher Scientific), Rat
   IgG1 kappa Isotype Control (eBRG1) APC (17‐4301‐81, Thermofisher
   Scientific), Rat IgG1 kappa Isotype Control (eBRG1) FITC (11‐4301‐81,
   Thermofisher Scientific), Rat IgG2a Anti‐Mouse control PE (88‐8111‐40,
   Thermofisher Scientific).

RNA Isolation and qRT‐PCR

   DCs were collected for RNA isolation and qRT‐PCR after stimulated by
   different DsEVs or PBS for 24 h to assess the gene expression levels of
   CIITA, RFX5, RFXAP, and RFXANK. Additionally, spleen samples from SCI
   mice were collected after 3 days of injecting different DsEVs or PBS.
   These splenic CD4^+ T cells were then utilized for RNA isolation and
   subsequent qRT‐PCR analysis. As for RNA isolation and qRT‐PCR of spinal
   cord issue, 10 mm segments of spinal cord (5 mm on each side of the
   injury epicenter) were isolated. After the extraction of total RNA,
   cDNA was synthesized. A Stratagene Mx3005p real‐time PCR system
   (Agilent Technologies, USA) was used for detection. The primer
   sequences used were provided in Table [219]S1 (Supporting Information).

Library Preparation and RNA Sequencing Analysis

   The libraries construction and transcriptome analysis were conducted by
   OE Biotech Co., Ltd. (Shanghai, China). Differential expression
   analysis was performed using the DESeq25. P‐value < 0.05 and foldchange
   > 2 or foldchange < 0.5 was set as the threshold for significantly
   differential expression gene (DEGs). GO analysis and KEGG pathway
   enrichment of DEGs were performed respectively.

BMS

   The BMS primary scoring system assesses hind limb motor function on a
   scale of 0 (complete paralysis) to 9 (completely normal). Mice were
   observed for 5 min on a flat surface, and hind limb motor function was
   assessed using the single‐blind method by two independent blind
   observers. The average score given by both observers for each hind limb
   was recorded as the BMS score for the sample.

Catwalk‐Assisted gait analysis

   Catwalk‐assisted gait analysis (Catwalk XT, Noldus, Netherlands) was
   used to further detect the motor function of SCI mice.

Drug Toxicity Test of DsEVs to Major Organs

   After 3 days of DsEVs injection, heart, liver, spleen, lung, and
   kidneys were taken out for hematoxylin‐eosin staining (HE staining).
   The series of tissue sections were immersed in xylene, rehydrated
   through decreasing concentration of ethanol solution, rinsed with
   distilled water, and then stained with hematoxylin for 2 min. After
   stained with eosin for 1 min, the sections were dehydrated and sealed.

   Simultaneously, the activities of LDH and ALT were measured by the
   enzyme labeling method using the enzyme assay kits (Wanlei Bio, China).
   The absorbance OD value was measured with a microplate reader (Bio Rad
   iMak, Japan).

Statistical Analysis

   SPSS 24.0 and GraphPad Prism 9.0 software were used for statistical
   analysis and the plotting of statistical graphs. All experimental
   results were expressed as mean ± SD. The t‐test was used for two‐group
   comparisons. One‐way ANOVA was used to compare more than two groups.
   Two‐way RM ANOVA was used to compare BMS and body weight change in
   different groups. Statistical significance was accepted for values of p
   < 0.05 (*p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001).

Ethics Approval Statement

   This study received approval from the Ethics Committee (no.
   SYDW2019–111) of the Second Affiliated Hospital of Harbin Medical
   University (Harbin, China) in accordance with Chinese experimental
   animal legislation. All the authors compliance with all relevant
   ethical regulations.

Conflict of Interest

   The authors declare no conflict of interest.

Supporting information

   Supporting Information
   [220]Click here for additional data file.^ (1.5MB, pdf)

   Supplementary Table S1
   [221]Click here for additional data file.^ (11.5KB, xlsx)

Acknowledgements