Research on the TSPAN6 regulating the secretion of ADSCs-Exos through syntenin-1 and promoting wound healing

Abstract

Background

Exosomes (Exos) from adipose-derived stem cells (ADSCs) have a high inclusion content and low immunogenicity, which helps to control inflammation and accelerate the healing of wounds. Unfortunately, the yield of exosomes is poor, which raises the expense and lengthens the treatment period in addition to impairing exosomes’ therapeutic impact. Thus, one of the key problems that needs to be resolved in the current exosome study is increasing the exosome yield.

Methods

Tetraspanin-6 (TSPAN6) overexpression and knockdown models of ADSCs were constructed to determine the number of exosomes secreted by each group of cells as well as the number of multivesicular bodies (MVBs) and intraluminal vesicles (ILVs) within the cells. Subsequently, the binding region of the interaction between TSPAN6 and syntenin-1 was identified using the yeast two-hybrid assay, and the interaction itself was identified by immunoprecipitation. Finally, cellular and animal studies were conducted to investigate the role of each class of exosomes.

Results

When compared to the control group, the number of intracellular MVBs and ILVs was significantly larger, and the number of ADSCs^TSPAN6+-Exos was more than three times higher. However, TSPAN6’s ability to stimulate exosome secretion was reduced as a result of syntenin-1 knockdown. Additional yeast two-hybrid assay demonstrated that the critical structures for their interaction were the N-terminal, Postsynaptic density protein 95/Discs large protein/Zonula occludens 1 (PDZ1), and PDZ2 domains of syntenin-1, and the C-terminal of TSPAN6. In animal trials, the wound healing rate was best in the ADSCs^TSPAN6+-Exos group, while cellular experiments demonstrated that ADSCs^TSPAN6+-Exos better enhanced the proliferation and migration of human skin fibroblasts (HSFs) and human umbilical vein endothelial cells (HUVECs).

Conclusion

TSPAN6 stimulates exosome secretion and formation, as well as the creation of MVBs and ILVs in ADSCs. Syntenin-1 is essential for TSPAN6’s stimulation of ADSCs-Exos secretion. Furthermore, ADSCs^TSPAN6+-Exos has a greater ability to support wound healing, angiogenesis, and the proliferation and migration of a variety of cells.

Background

ADSCs are mesenchymal stem cells derived from adipose tissue with a high capacity for self-renewal, multidirectional differentiation, and the ability to produce hundreds of cytokines [1, 2]. They are generally regarded as one of the most promising adult stem cell types for therapeutic applications due to their large supply and simplicity of access. According to reports, exosomes and paracrine cytokines are the primary mechanisms by which ADSCs exercise their biological effects [3]. The study of ADSCs-Exos has advanced quickly during the past five years, particularly in the area of wound healing [4]. According to studies, ADSCs-Exos may accelerate the healing process by decreasing the expression of inflammatory factors and stimulating neovascularization, fibroblast proliferation, and collagen production through a variety of signaling pathways, including Wnt/ꞵ-catenin, PI3K/Akt, etc [5].

A twofold invagination of the plasma membrane and the development of MVBs containing ILVs are the primary steps in the synthesis of the exosome [6]. Early-sorting endosomes (ESEs), which are formed by the first invagination of the plasma membrane, are cup-shaped structures that hold cell membrane proteins and other related soluble proteins. ESEs are also produced by the endoplasmic reticulum and intracellular Golgi apparatus, and they can eventually develop into late-sorting endosomes (LSEs). Re-invagination of the LSEs’ membrane results in the formation of MVBs, which in turn causes the MVBs to produce a number of ILVs [7]. MVBs can either merge with the cell membrane and release ILVs as exosomes, or they can fuse with autophagosomes or lysosomes and be degraded, allowing their contents to enter the cell cycle.

Insufficient exosome production, however, not only reduces the therapeutic effect of exosomes but also lengthens the duration of treatment and raises the expense of treatment [8]. Researchers have used the regulatory mechanisms behind exosome biogenesis by manipulating specific intracellular components to increase exosome production to investigate new paths to better address the issue of low exosome production [9, 10]. It has now been demonstrated that various tetraspanins regulate the classification of proteins in exosomes. The tetraspanin protein family members are a class of low-molecular-weight glycoproteins that are extensively expressed on the membranes of eukaryotic cells. They are involved in lymphocyte activation, tissue differentiation, tumor invasion and metastasis, and cell adhesion and proliferation [11]. The majority of tetraspanins (TSPANs) are proteins found on the cell surface that play a key role in mediating signal transduction across cell membranes and intercellular communication. TSPAN6 functions as an oncogenic gene in the development of glioblastoma [12]. By modulating exosome release, TSPAN6 overexpression in glioblastoma cells stimulates angiogenesis. Meanwhile, TSPAN6 plays a crucial role in the MVBs, mediating the exosome release and the lysosomal degradation of amyloid precursor protein fragments [13]. According to research, TSPAN6 is more abundant in MVBs and ILVs and can recruit syntenin-1 via the PDZ1 domain of the cytoplasmic adaptor (syntenin-1) [13]. With its two contiguous tandem PDZ domains, syntenin-1 resembles a molecule similar to an adaptor. Numerous publications have shown that syntenin-1 interacts with a wide range of proteins [14]. Multiple peptide patterns are recognized by syntenin-1’s PDZ domains with low to moderate affinity. Syntenin-1 controls the structure of the cell membrane by interacting with other proteins. Consequently, syntenin-1 can induce neural cells to protrude along their neurites, tumor cells to metastasize, and different cell types to produce exosomes. Besides, syntenin-1 can interact with proteins involved in exosome formation and control how internal membrane structure and the cargo of exosomes are rearranged [15]. Several tetraspanin proteins are present in exosomes, and one of these, CD63, has been shown to interact directly with syntenin-1. Previous research has also demonstrated that the CD63/syntenin-1 protein complex regulates the conversion of early endosomes into multivesicular bodies and their transport to the plasma membrane [16]. TSPAN6 may have an impact on exosome formation; however, its exact function in exosome secretion is still unclear [13, 17].

To investigate whether TSPAN6 can specifically regulate the exosome secretion of ADSCs-derived exosomes and thereby increase exosome yield, we present a new research idea for the exosome secretion mechanism of ADSCs-derived exosomes from the perspective of TSPAN6 in this study.

Methods and materials

Isolation and identification of ADSCs

With approval from the Ethics Committee of the Second Xiangya Hospital of Central South University, adipose tissue was collected from the abdomen of ten Sprague-Dawley (SD) male rats (12 weeks old, 160–180 g). All rats were given isoflurane (0.41 mL/min at 4 L/min fresh gas flow, 2%) inhalation anesthesia before being put to death with a 1.5% sodium pentobarbital (dose: 200 mg/kg) intraperitoneal injection. Rats’ abdomens were dissected after a thorough sterilization, and the abdominal adipose tissue was removed. Blood vessels and fascia were adequately removed from rat adipose tissue. Adipose tissue that had been sheared was placed in a sterile centrifuge tube and then digested using the same volume of collagenase type I (0.25%, Solarbio, C8140). Once the digestion was finished, the adipose tissue switched the form of chyme, to which an equal volume of ADSCs complete medium was added to stop the process. ADSCs complete medium consisted of 90% dulbecco’s modified eagle medium (DMEM) (high glucose, Procell, PM150210), 9% special grade fetal bovine serum (FBS) (Procell, 164210-50), and 1% penicillin-streptomycin solution (Gibico, 15140122). Next, the adipose tissue was filtered through a sterile cell strainer, centrifuged (1000 rpm/min for 5 min), the supernatant was discarded to preserve the precipitates, and the cellular precipitates were resuspended using ADSCs complete medium. The cells were moved to a CO₂ incubator for culture (37 °C, 5% CO₂, saturated humidity) after digestion and centrifugation. Third-generation cells’ surface marker expression levels (CD29 (FITC anti-rat CD29 Antibody, BioLegend, 102205), CD44 (FITC anti-rat CD44 Antibody, BioLegend, 203906), CD73 (Anti-CD73/FITC antibody, Bioss, bs-4834R-FITC), CD90 (FITC anti-rat CD90 Antibody, BioLegend, 202503), CD105 (CD105 Monoclonal Antibody, FITC, Thermo Fisher Scientific, MA1-19594), CD14 (Anti-CD14/FITC antibody, Bioss, bs-1192R-FITC), CD34 (CD34 Monoclonal Antibody, FITC, Thermo Fisher Scientific, 11-0341-82), CD45 (CD45 Monoclonal Antibody, FITC, Thermo Fisher Scientific, 11-0451-82), and CD106 (CD106 antibody/FITC, Biorbyt, orb434364)) were detected by flow cytometry (BECKMAN COULTER, B53000), and then the Rat Adipose MSC Lipogenic (OriCell, RAXMD-90031), Osteogenic (OriCell, RAXMD-90021), and Chondrogenic (OriCell, RAXMD-90041) Induced Differentiation Kit for trilineage-induced differentiation was used to identify the cells.

Isolation and identification of ADSCs-Exos

The ADSCs were cultured in exosome-depleted medium for 48–72 h to collect the medium and were then removed by centrifuging at 4 ℃ and 300 g for 10 min. Death cells was removed by centrifuging at 4 ℃ and 2000 g for 10 min. Then, the cell debris was removed by centrifuging at 4 ℃ and 10,000 g for 30 min. Next, the medium was twice ultracentrifuged at 4 ℃ and 100,000 g for 70 min to purify the exosomes. Transmission electron microscopy (Japan Electronics Co., Ltd, JEM-F200) was used to examine the shape of exosomes; western blotting was used to identify the expression of CD9, CD63, and TSG101 in exosomes; and nanosight particle size analysis (ParticleMetrix, ZetaView PMX 110) was used to identify the dispersion of exosomes. The exosome-depleted medium consisted of 90% DMEM (high glucose, Procell, PM150210), 9% exosome-depleted FBS (Systembio, EXO-FBS-250 A-1), and 1% penicillin-streptomycin solution (Gibico, 15140122).

Reverse transcription-quantitative polymerase chain reaction (RT-qPCR)

Total ribonucleic acid (RNA) was extracted with TRIzol reagent (Absin, abs9331), and a nanodrop spectrophotometer (NanoDrop Lite Plus Microvolume Spectrophotometer, Thermo Scientific, NDL-PLUS-CN) was used to measure the purity and quantity of the isolated RNA. The isolated RNA could be kept at -80 °C if it wasn’t going to be used right away. Genomic deoxyribonucleic acid (DNA) was first eliminated, and then, using the SYBR^® Green Premix Pro Taq HS qPCR Kit (Accurate Biology, AG11702), it was amplified. By using β-actin as the endogenous loading for mRNA, the 2^−ΔΔCt method was used to assess the value of gene expression. The following primer pairs were applied to detect the rat cDNAs: TSPAN6-F: ACACTTTCATCTTTTGGATCACTGG, TSPAN6-R: ACAAAAGGCACATTGGT-GGC, β-actin-F: CCCATCTATGAGGGTT-ACGC, and β-actin-R: TTTAATGTCACGCACGATTTC.

Western blotting

From tissues or cell lysates, total protein was extracted using radioimmunoprecipitation assay (RIPA) buffer (Beyotime, P0013B). The protein sample was conducted on sodium dodecyl sulfate - polyacrylamide gel electrophoresis (SDS-PAGE) (NCM Biotech, P2012) and then transferred onto polyvinylidene fluoride (PVDF) membranes following concentration analysis using the BCA Kit (Bioss, C05-02001). The membranes were blocked by 5% non-fat milk at 25 °C for 1 h and reacted with TSPAN6 (Proteintech, 1:1000, 12293-1-AP), CD9 (Abcam, 1:1000, ab307085), CD63 (Santa Cruz, 1:1000, sc-5275), TSG101 (Abcam, 1:10000, ab125011), syntenin-1 (Santa Cruz, 1:1000, sc-515538), and β-actin (Abcam, 1:5000, ab6276) primary antibodies overnight at 4 °C. The use of horse radish peroxidase (HRP)-conjugated secondary antibodies (Thermo Fisher Scientific, 32460) was followed by the detection of the protein bands using the enhanced chemiluminescence (ECL) reagent (Zenbio, 17046) and then captured with the Bio-Rad system (Bio-Rad, ChemiDoc XRS+). Image J was used to analyze the level of protein expression using endogenous loading of β-actin.

Cells proliferation assay

Cell proliferation was measured using a Cell Counting Kit-8 assay (CCK-8, Solarbio, CA1210). ADSCs, HUVECs, and HSFs (1.5 × 10³ cells/well) were seeded onto 96-well plates and cultured in exosome-depleted medium with 100 µg/mL of ADSCs-Exos or an equivalent volume of exosome diluent (phosphate buffer saline, PBS). The blank group was a collection without any cells. CCK-8 solution (10 µL/well) was added to the medium on days 1, 2, 3, 4, and 5. Cells were incubated at 37 ℃ for 2 h. A microplate spectrophotometer (Multiskan SkyHigh Microplate Spectrophotometer, Thermo Fisher Scientific, A51119600DPC) was used to quantify the absorbance at 450 nm, and the optical density data represented the rate of cell proliferation.

Immunocytochemistry

Polyethyleneimine was applied to round coverslips to promote cell growth, and 4% paraformaldehyde in PBS was used to fix the cells. After immersing in antigen retrieval buffer for 10 min, the coverslips were washed with PBS. Samples were incubated in PBS with either 0.1–0.25% Triton X-100 (Biosharp, BL934A) for 10 min. To prevent unspecific antibody binding, cells were incubated for 30 min in PBST (PBS + 0.1% Tween 20) containing 1% bovine serum albumin (BSA) (Biosharp, BS114) and 22.52 mg/mL glycine. In a humidified chamber, cells were incubated in diluted primary antibody in 1% BSA in PBST for 1 h at room temperature or for the entire night at 4°C. Then, cells were incubated for 1 h at room temperature in the dark with the secondary antibody in 1% BSA. Next, cells were provided with one-minute 4’,6-diamidino-2-phenylindole (DAPI) (Thermo Fisher Scientific, D1306) incubation. Finally, the images were gathered using a fluorescent microscope (Carl Zeiss, Axi Vert A1).

Co-immunoprecipitation (Co-IP)

The Pierce Co-IP kit (Thermo Fisher Scientific, 26149) was used as a guide during this process. The IP lysate was pre-cooled on ice before being put into the cell culture dishes, where it was lysed for 5 min on ice after the cells had been twice cleaned with PBS buffer. A control agarose resin was used to pretreat cell lysates. Furthermore, three IP lysate washes were conducted after adding protein lysates to the resin that had been cross-linked with antibodies and incubating at 4 ℃ for the entire night. Western blotting was used to detect bound proteins after they were eluted.

Yeast two-hybrid assay

Subcloning vectors for TSPAN6 and syntenin-1 were designed to amplify and guarantee plasmid yields greater than 200 ng. Yeast extract Peptone Dextrose Adenine (YPDA) medium (Coolaber, PM2011) was used to cultivate monoclonal yeast species, and the resulting organisms were collected. The bacteria were resuspended using 1× TE/LiAc solution (TaKaRa, 630439), and each tube of the reaction system received two plasmids (1 µg each). Subsequently, 5 µL of pre-denatured Carrier DNA (TaKaRa, 630440) was added first, followed by 300 µL of 1×PEG/LiAc, and mixed thoroughly and left at 30 ℃ for 30 min. Each tube was filled with 20 µL of dimethyl sulfoxide (DMSO) (Solarbio, D8371), mixed, and water-bathed for 15 min at 42 °C. The reaction tubes were centrifuged and resuspended in YPDA Plus liquid medium for 60 min. The organisms were then resuspended with saline and spread out on SD/-Trp/-Leu and SD/-Trp/-Leu/-His/-Ade plates (Coolaber, PM2221, PM2111), respectively, and incubated in a biochemical incubator.

Wound healing assay

After inoculating HUVECs/HSFs (iCell, iCell-h110, iCell-0051a) into 6-well plates, care was taken to maintain a constant cell density in each well, ideally under control so that the cells would grow into a complete monolayer by the second day. The midline of the wells was scraped through using a 10 µL pipette head. After removing the original medium, the cells were cleaned three times using a sterile PBS buffer, and the suitable medium for culture was chosen based on the needs of the experiment. Under an inverted microscope (Olympus, CKX53), the scratch width was measured, and after 12 h of culture, the difference in width was recorded. Image J was used to assess the HUVECs’ and HSFs’ capacity for migration.

Transwell assay

The Transwell plate (Corning, 3422) was divided into two chambers: the upper chamber and the lower chamber. Resuspended in 100 µL of serum-free medium, 1 × 10⁴ cells/well of HUVECs/HSFs were seeded in the upper chamber. After applying approximately 600 µL of different treatments into the lower chamber for 24 h, the transwell chambers were thoroughly cleaned three times using PBS. The cells were then fixed for 30 min with 4% paraformaldehyde and stained for 30 min with 0.1% (w/v) crystal violet (Abiowell, AWC0142a). After staining, the average numbers of cells that invaded the membrane were determined via microscopy (Olympus, CKX53).

Tube formation assay

Following an overnight thawing period in a 4 °C refrigerator, 200 µL of Matrigel (Corning, 354234) was added to each pre-cooling 24-well plate, and the plate was equally swirled on ice to prevent the Matrigel from gelling too soon. The 24-well plate was then incubated for 1 h in the culture incubator to facilitate Matrigel gelation. Following the seeding of HUVECs (1 × 10⁵ cells/well) onto Matrigel-coated 24-well plates, the cells were treated with either PBS or exosomes (100 µg/mL) from different groups. An inverted light microscope (Olympus, CKX53) was used to view tube formation nine hours after seeding, and ImageJ was used to count the tube formation.

Animal operation

Twenty-seven 6-week-old male SD rats, weighting 160–180 g, were obtained from Hunan SJA Laboratory Animal Co., Ltd and randomly divided into three groups (9 rats/group): the ADSCs^TSPAN6+-Exos group, the ADSCs-Exos group, and the control group. A computer-based random order generator was used to produce random numbers. Every animal received therapy at the same time each treatment day, with randomization used to establish the trauma area counts and the order of exosome injections. The rats were anesthetized by inhaling isoflurane (0.41 mL/min at 4 L/min fresh gas flow, 2%) and then receiving 1.5% sodium pentobarbital (60 mg/kg) intraperitoneally. The rats were given circular, full-thickness wounds of 2 cm in diameter on their backs. PBS, ADSCs-Exos, or ADSCs^TSPAN6+-Exos were subsequently applied to the wounds. Starting on day 0, the exosome injection regimen was carried out every three days. Eight spots along the periphery of every circular incision received subcutaneous injections. 200 µL of exosomes at a concentration of 1 µg/µL were given to one rat, while 200 µL of PBS buffer was given to the control group. Following the procedure, a skin patch was applied to the wounds, and all of the rats were returned to the biosafety facility. The animals were excluded if they died prematurely, preventing the collection of behavioral and histological data. However, there were no exclusions during this study. Nine rats were divided into each group for the purposes of evaluating the healing of wounds (n = 9); three rats for each group underwent hematoxylin-eosin (HE) staining (n = 3); three rats underwent Masson staining (n = 3); and three rats underwent immunofluorescence staining (n = 3). We collected a minimum of three samples for each outcome index. Four researchers worked on each rat: the first was in charge of randomly assigning each animal to a group and was the only one who knew which animal was in which treatment group. Although the second researcher was in charge of performing the experiments, he was unaware of the treatment that each animal received. The third researcher was in charge of gathering experimental data, but he was also blind to the treatments that each animal had undergone. In charge of data analysis, the fourth researcher was aware of which animals belonged to the same group but not the precise treatments they had all received. At days 0, 3, 6, 9, 12, and 15 postoperatively, the skin patch was removed, and a digital camera was then used to take pictures of the wounds’ healing results. The wound area was measured and examined using ImageJ software. On day 15, all rats were euthanized by intraperitoneal injection of pentobarbital sodium at an injection dose of 180 mg/Kg. The formula [(C0-Ct)/C0] ×100% was used to calculate the wound healing rates. The wound dimension at day 0 was denoted by C0, and the wound dimension at each time point was represented by Ct. (The work has been reported in line with the ARRIVE guidelines 2.0, Additional file 5)

Histological and immunofluorescence analysis

The excised tissues, which included the wound bed and the surrounding healthy skin on day 15 after the wound, were preserved in 10% formalin, dried using a series of increasing alcohol concentrations, embedded in paraffin, and sectioned into 4-µm-thick pieces perpendicular to the wound surface for histological examination. Histological observations of wound scar width (the distance between the ends of scars during wound healing) were made using HE staining. Masson staining was utilized to determine the collagen maturity level (collagen content per unit area of tissue) in the wound beds.

To ascertain the angiogenesis of wound beds, immunofluorescence (IF) staining for CD31 (Abcam, 1:200, ab222783) and α-smooth muscle actin (α-SMA) (Abcam, 1:50, ab7817) was performed. Briefly, the sections were rehydrated and treated with citric acid (pH 6.0) to retrieve the antigen, blocked with BSA for 20 min, and finally incubated with antibody at 4 °C overnight. After three times PBS washes, the sections were incubated for 60 min at 37 °C with a secondary antibody labeled with fluorescein isothiocyanate (Abcam, ab150079, ab150113). The nucleus was stained with 5 µg/mL DAPI. Then, an upright fluorescence microscope (Carl Zeiss, Axi Vert A1) was applied to observe the stained section in a dark chamber. ImageJ software was employed to count the number of new blood vessels per unit area.

Statistics

All data were shown as mean ± standard error of the mean (SEM). A one-way analysis of variance (ANOVA) was employed to evaluate the significance of the difference. A P < 0.05 was considered significant. All of the data were statistically compared using GraphPad Prism 9 software. One-way ANOVA was performed when the data were normally distributed with homogeneous variance. The Kruskal-Wallis H test was employed when there was heterogeneity in variances.

Results

Preparation and characterization of ADSCs-Exos and ADSCs^TSPAN6+-Exos

We isolated primary ADSCs (See Additional file 1) from rat inguinal fat pads, collected supernatants from ADSCs (no treatment) and TSPAN6 overexpression-treated ADSCs to separate ADSCs-Exos and ADSCs^TSPAN6+-Exos by ultracentrifugation, respectively. Identification of the extracted exosomes was performed by Transmission electron microscopy (TEM), Nanoparticle Tracking Analysis (NTA), and WB. All exosome groups were found to be spherical, with a cup-like or teato-like structure that had a brighter outer ring and a slightly darker center, as seen in TEM images presented in Fig. 1. Exosomes obtained from ADSCs and ADSCs^TSPAN6+ had average sizes of 139.7 nm and 147.1 nm, respectively, according to NTA analysis, and the majority of exosomes were found to be between 40 and 160 nm in diameter. According to the WB analysis, isolated exosomes expressed the protein markers CD9, CD63, and TSG101 positively, but not Calnexin. These findings validate the effective extraction of exosomes from both groups by demonstrating that ADSCs-Exos and ADSCs^TSPAN6+-Exos had comparable particle size, shape, and protein markers.

Fig. 1

Characteristics of exosomes. Characteristics of exosomes. A. Observation of ADSCs-Exos morphology and structure by transmission electron microscopy (indicated by the black arrow, bar: 200 nm); B. Detection of the diameter and distribution of ADSCs-Exos by Nanosight; C. Western Blot was used to identify ADSCs-Exos surface protein markers; D. Observation of ADSCs^TSPAN6+-Exos morphology and structure by transmission electron microscopy (indicated by the black arrow, bar: 200 nm); E. Detection of the diameter and distribution of ADSCs^TSPAN6+-Exos by Nanosight; F. Western Blot was used to identify ADSCs^TSPAN6+-Exos surface protein markers. (Full-length blots are presented in Additional file 4, Figure S3)

TSPAN6 favors the generation of exosomes

We created a model of ADSCs with TSPAN6 overexpression and knockdown (See Additional file 2) in order to more thoroughly investigate the impact of TSPAN6 on exosome secretion from ADSCs. As seen in Fig. 2, the proliferative capability of the cells in the various groups was initially assessed to exclude the impact of the quantity of source cells on the quantity of exosomes released. The number of cells that were the source of exosomes was nearly equal in each group. Following the extraction of the exosomes from each of the three cell groups, NTA and WB performed quantitative and semi-quantitative analyses. The findings demonstrated that whilst the number of exosomal particles obtained from ADSCs increased 3.385-fold following TSPAN6 overexpression, the number of exosomal particles obtained from ADSCs dropped to 67.96% of the origin number following TSPAN6 knockdown. Similar outcomes were shown by WB semi-quantification (Fig. 2). These findings demonstrated that TSPAN6 was able to favorably control the release of ADSCs-Exos, leading to an increase in the production of exosomes, but it had no discernible influence on the proliferative ability of ADSCs.

Fig. 2

Analysis of ADSCs derived exosomes in each group. A. Effect of TSPAN6 overexpression lentivirus and knockdown lentivirus on proliferation of ADSCs; B. NTA quantitative analysis of exosomes; C. Exosome-associated protein expression levels in each group; D. Semi-quantitative statistical results of Western Blot (*: p < 0.05; **: p < 0.01; ***: p < 0.001). (Full-length blots are presented in Additional file 4, Figure S4)

TSPAN6 regulates the formation of MVBs and ILVs in ADSCs

This study used TEM to examine the morphology and quantity of MVBs as well as ILVs in ADSCs^TSPAN6+, ADSCs^TSPAN6−, and control ADSCs in order to investigate the mechanism underlying TSPAN6’s effect on exosome secretion. The results demonstrated that, in comparison to the control group, ADSCs^TSPAN6+ had a considerably higher number of MVBs and ILVs in MVBs (Fig. 3), but not a statistically different morphology for ILVs. In contrast, following TSPAN6 knockdown, ADSCs showed a large drop in the quantity of MVBs and ILVs inside MVBs (Fig. 3), although there was no significant change in the shape of ILVs. The aforementioned findings indicate that TSPAN6 stimulates the synthesis of MVBs and ILVs in ADSCs but has no effect on the size or shape of ILVs, indicating that exosome generation is the primary mechanism by which TSPAN6 stimulates exosome secretion.

Fig. 3

MVBs and ILVs in ADSCs. A-C. The morphology and number of MVBs and ILVs in ADSCs^TSPAN6+ (bar: 2 μm; 600 nm; 250 nm); D-F. The morphology and number of MVBs and ILVs in ADSCs (bar: 2 μm; 800 nm; 250 nm); G-I. The morphology and number of MVBs and ILVs in ADSCs^TSPAN6− (bar: 2 μm; 800 nm; 250 nm); J. 20 different electron microscopic images of cells were randomly selected from each group, and the number of MVBs in each cell was counted; K. 60 different electron microscopic images of MVBs were randomly selected from each group, and the number of ILVs in each MVB was counted; L. Comprehensive statistics were used to predict the number of ILVs in each group of cells. (*: p < 0.05; **: p < 0.01; ***: p < 0.001)

Furthermore, we used TSG101 (red) and CD63 (green) as surface markers for MVBs to do immunofluorescence labeling, which allowed us to see the number and subcellular location of MVBs in each group of ADSCs. The findings indicated (Fig. 4) that while there was no difference in the distribution location of the positive particles, there were significantly more TSG101⁺ and CD63⁺ particles in the ADSCs^TSPAN6+ group than in the control group; on the other hand, there were significantly fewer TSG101⁺ and CD63⁺ particles in the ADSCs^TSPAN6− group than in the control group, and the distribution of the particles was more sparsely distributed.

Fig. 4

Immunofluorescence observation of MVBs in ADSCs. A-D. The expression levels of TSG101 and CD63 in ADSCs^TSPAN6+ (bar: 10 μm); E-H: The expression levels of TSG101 and CD63 in ADSCs (bar: 10 μm); I-L: The expression levels of TSG101 and CD63 in ADSCs^TSPAN6− (bar: 10 μm); M: The number of TSG101 positive particles in each group; N: The number of CD63 positive particles in each group. (*: p < 0.05; **: p < 0.01; ***: p < 0.001)

Subsequently, we isolated whole-cell lysates from three cell groups and used Western blot analysis to measure the expression levels of TSG101, CD9, and CD63 in the cells. The findings indicated that none of these marker proteins significantly changed in the cells of any group (Fig. 5), indicating that TSPAN6 had no discernible impact on the exosome marker proteins in whole cells. Consequently, it can be demonstrated that variations in the quantity of exosomes released by the cells account for the majority of the variation in the expression levels of these marker proteins in exosomes, as opposed to variations in the expression levels of these proteins in whole cells.

Fig. 5

The expression levels of exosome related proteins. A. The expression levels of exosomal marker proteins in each group; B. Statistical results of expression levels of exosomal marker proteins in each group (*: p < 0.05; **: p < 0.01; ***: p < 0.001). (Full-length blots are presented in Additional file 4, Figure S5 and S6)

Role of Sytenin-1 in the regulation of exosome secretion by TSPAN6

The first steps in the detection of syntenin-1 expression in ADSCs, ADSCs^TSPAN6+, and ADSCs^TSPAN6− groups were Western Blot and RT-qPCR. The results demonstrated (Fig. 6) that, in comparison to the control group, the expression of syntenin-1 was down-regulated in the ADSCs^TSPAN6− group and up-regulated in the ADSCs^TSPAN6+ group. Given that TSPAN6 may influence syntenin-1 expression in adipose stem cells and that there appears to be a positive correlation between the two expressions, it is plausible that they may work in concert to control the release of exosomes originating from adipose stem cells. By applying immunoprecipitation assays, we examined the possibility of TSPAN6 and syntenin-1 interacting with one another in order to learn more about their interaction. The findings are displayed in Fig. 6, TSPAN6 can be detected in the samples by immunoprecipitation using the syntenin-1 antibody, and syntenin-1 can be detected in the samples by immunoprecipitation using the TSPAN6 antibody simultaneously, suggesting that TSPAN6 and syntenin-1 have a mutual binding effect.

Fig. 6

Role of Syntenin-1 in the regulation of exosome secretion by TSPAN6. A. The expression levels of syntenin-1 in each group of ADSCs (Full-length blots are presented in Additional file 4, Figure S7); B. Statistical results of expression levels of syntenin-1 in each group; C. The expression levels of syntenin-1 mRNA in each group of ADSCs; D. Syntenin-1 antibody Co-IP detection (Full-length blots are presented in Additional file 4, Figure S8); E. TSPAN6 antibody Co-IP detection (Full-length blots are presented in Additional file 4, Figure S9); F. NTA was performed to detect exosomes extracted from ADSCs(syn-) and ADSCs^TSPAN6+(syn-); G. The expression levels of exosome-related proteins in the ADSCs(syn-) and ADSCs^TSPAN6+(syn-) (Full-length blots are presented in Additional file 4, Figure S10); H. Statistical analysis of the expression levels of exosome-related proteins in the ADSCs(syn-) and ADSCs^TSPAN6+(syn-). (*: p < 0.05; **: p < 0.01; ***: p < 0.001)

We created a stable syntenin-1 knockdown adipose stem cell model [ADSCs(syn-)] in this study (See Additional file 3), and we then transfected TSPAN6 overexpressing lentivirus and empty lentivirus to investigate if TSPAN6 could continue to stimulate the secretion of exosomes derived from adipose stem cells after syntenin-1 knockdown. Figure 6 displays the experimental results: NTA revealed no statistically significant difference in the particle counts of the two exosome groups, and the WB results similarly revealed no statistically significant variation in the expression of related proteins between the two exosome groups. The findings imply that following syntenin-1 knockdown, TSPAN6’s role in stimulating exosome secretion is blocked.

Binding region for the interaction between TSPAN6 and syntenin-1

In order to confirm that the yeast two-hybrid system is accurate, this experiment used BK-53/AD-T as a positive control and BK-Lam/AD-T as a negative control. As shown in Fig. 7, all of the self-activated groups were able to grow on SD/-Trp/-Leu (DDO) plates, demonstrating that the bait yeast was constructed successfully and that the expressed relevant proteins were not toxic to yeast cells. On SD/-Trp/-Leu/-His/-Ade (QDO) plates, however, none of the self-activated groups was able to grow, indicating that the aforementioned proteins are not self-activating and can be used to validate interactions.

Fig. 7

Yeast two-hybrid self-activated group. A. Growth of self-activated group on SD/-Trp/-Leu (DDO); B. Growth of self-activated group on SD/-Trp/-Leu/-His/-Ade (QDO)

Formal yeast two-hybrid tests were carried out following the determination that all expressed proteins lacked self-activating function and were non-toxic. The outcomes of yeast two-hybrid system of TSPAN6 wild-type (TSPAN6^WT) and TSPAN6 C-terminal truncated type (TSPAN6^ΔC) with syntenin-1, respectively, showed that the TSPAN6^WT and syntenin-1^WT groups were able to grow on both DDO and QDO plates (Fig. 8), suggesting that co-expression of the two protein molecules was non-toxic and that there was an interaction between the two protein molecules. On DDO plates, but not on QDO plates, the TSPAN6^ΔC with syntenin-1^WT were able to grow, suggesting that the two protein molecules co-expressed are non-toxic but do not interact. These results imply that the essential structure for their interaction is TSPAN6’s C-terminal.

Fig. 8

Yeast two-hybrid experimental group. A. Growth of experimental group of TSPAN6 on SD/-Trp/-Leu; B. Growth of experimental group of TSPAN6 on SD/-Trp/-Leu/-His/-Ade; C. Growth of experimental group of syntenin-1 on SD/-Trp/-Leu; D. Growth of experimental group of syntenin-1 on SD/-Trp/-Leu/-His/-Ade

The results of yeast two-hybrid system between TSPAN6^WT with syntenin-1 N-terminal truncated type (syntenin-1^ΔN), syntenin-1 PDZ1 domain truncated type (syntenin-1^ΔPDZ1), syntenin-1 PDZ2 domain truncated type (syntenin-1^ΔPDZ2), and syntenin-1 C-terminal truncated type (syntenin-1^ΔC), respectively, are shown in Fig. 8. On DDO and QDO plates, TSPAN6^WT with syntenin-1^ΔC were able to grow, indicating that the co-expression of the two protein molecules is interactive and non-toxic. On DDO plates, but not on QDO plates, the TSPAN6^ΔC with syntenin-1^ΔN, syntenin-1^ΔPDZ1, and syntenin-1^ΔPDZ2 were able to grow, suggesting that the two protein molecules co-expressed are non-toxic but do not interact. The findings imply that the essential structures for their interaction are the syntenin-1 N-terminal, PDZ1, and PDZ2 domains.

Functional analysis of ADSCs^TSPAN6+-Exos

We used PBS, ADSCs-Exos, and ADSCs^TSPAN6+-Exos to co-culture HUVECs and HSFs, respectively, and used the CCK-8 assay to measure each group’s cell proliferation capacity. The results, as illustrated in Fig. 9, demonstrated that when compared to PBS and ADSCs-Exos, the cells in the ADSCs^TSPAN6+-Exos group had the strongest proliferative ability in both HUVECs and HSFs. According to the aforementioned findings, ADSCs^TSPAN6+-Exos more effectively supports HUVEC and HSF proliferation. Subsequently, we used the Transwell migration assay and the wound healing assay to evaluate each group’s cell migration capacity. The results are displayed in Fig. 10, in both HUVECs and HSFs, the migration rates of cells in the ADSCs-Exos and ADSCs^TSPAN6+-Exos groups were significantly greater than those of the PBS group, with the ADSCs^TSPAN6+-Exos group exhibiting the highest migration rate. The results mentioned above indicate that HUVECs and HSFs migration may be more effectively facilitated by the exosomes isolated from ADSCs following TSPAN6 overexpression. We also use the tube formation assay to evaluate exosomes’ capacity to stimulate angiogenesis in each group. Figure 9 presents the results, it can be shown that the ADSCs^TSPAN6+-Exos and ADSCs-Exos groups exhibited significantly higher values than the PBS group in terms of total tube length, total branching point, and total loops. Additionally, the ADSCs^TSPAN6+-Exos group outperformed the ADSCs-Exos group. These results suggested that ADSCs^TSPAN6+-Exos was more effective in stimulating HUVECs tube formation, which in turn suggested that it was more capable of stimulating angiogenesis. Therefore, Western blot of hypoxia-inducible factor-1α (HIF-1α) and vascular endothelial growth factor (VEGF) was conducted to assess the underlying mechanisms of ADSCs^TSPAN6+-Exos. The ADSCs^TSPAN6+-Exos group had the greatest levels of HIF-1α and VEGF expression (Fig. 11). The outcomes also confirmed that ADSCs^TSPAN6+-Exos could efficiently promote HUVECs proliferation and migration through the HIF-1α/VEGF pathway. Furthermore, Wnt/β-catenin signaling was identified in HSFs. The findings indicated that the addition of ADSCs-Exos or ADSCs^TSPAN6+-Exos triggered Wnt/β-catenin signaling, and that the ADSCs^TSPAN6+-Exos might augment the Wnt/β-catenin pathway’s activation in HSFs (Fig. 11).

Fig. 9

Cell proliferation and tube formation assays. A. The proliferation capacity of HSFs; B. The proliferation capacity of HUVECs; C. The tube formation assay of HUVECs in each group (100×); D. Statistics of total tube length in each group; E. Statistics of total branching points in each group; F. Statistics of total loops in each group. (*: p < 0.05; **: p < 0.01; ***: p < 0.001)

Fig. 10

Cell migration assays. A. Wound healing assay for the migration of HUVECs (bar: 100 μm); B. Statistics of the wound healing assay of HUVECs; C. Wound healing assay for the migration of HSFs (bar: 100 μm); D. Statistics of the wound healing assay of HSFs; E. Transwell assay for the migration of HUVECs (×100); F. Statistics of the transwell assay of HUVECs; G. Transwell assay for the migration of HSFs (×100); H. Statistics of the transwell assay of HSFs. (*: p < 0.05; **: p < 0.01; ***: p < 0.001)

Fig. 11

Mechanisms of ADSCs^TSPAN6+-Exos promoting proliferation and migration of the HUVECs and HSFs. A. Western blot analysis of HIF-1α and VEGF in HUVECs (Full-length blots are presented in Additional file 4, Figure S11); B. Statistical results of expression levels of HIF-1α in each group; C. Statistical results of expression levels of VEGF in each group; D. Western blot analysis of Wnt and β-catenin in HSFs (Full-length blots are presented in Additional file 4, FigureS12); E. Statistical results of expression levels of Wnt in each group; C. Statistical results of expression levels of β-catenin in each group. (*: p < 0.05; **: p < 0.01; ***: p < 0.001)

Following the successful construction of a 20 mm-diameter rat wound model in animal studies, ADSCs^TSPAN6+-Exos, ADSCs-Exos, and PBS were injected near the wound’s edge, respectively. Figure 12 displays the results, the wounds in all three groups of rats healed gradually over time, with the wound healing rates of the ADSCs-Exos and ADSCs^TSPAN6+-Exos groups being significantly higher than those of the control group starting on day 6; the wound healing rates of the ADSCs^TSPAN6+-Exos group starting on day 9 being significantly higher than those of the ADSCs-Exos group and the control group, and reaching 97% on day 15 (Additional file 6, Table S1). This finding implies that exosomes produced by ADSCs overexpressing TSPAN6 may be more effective in accelerating rat wound healing. On day 15, the wound tissue was stained with HE to determine the width of the scar. The PBS group had the broadest scar at day 15, as illustrated in Fig. 12, whereas the wound treated with ADSCs^TSPAN6+-Exos had the thinnest scar, almost fully healed (Additional file 6, Table S2). Masson staining was used to assess the level of collagen deposited in the wound tissue. As can be seen in Fig. 12, at day 15, the PBS group had the least amount of collagen deposition roughly 20%, while the ADSCs^TSPAN6+-Exos group had more than 70% (Additional file 6, Table S3). This difference in the degree of collagen deposition showed that the ADSCs^TSPAN6+-Exos was able to enhance and speed up wound healing. Next, the tissue around the wound was stained with CD31 and αSMA in order to measure the amount of neovascularization present. As demonstrated in Fig. 13, CD31 staining revealed that the ADSCs^TSPAN6+-Exos treatment group had the highest neovascular density, while the PBS group had the lowest. The vascular density of the ADSCs^TSPAN6+-Exos and the ADSCs-Exos did not differ significantly, according to αSMA staining, but both groups’ vascular densities were much larger than those of the PBS group (Additional file 6, Table S4 and S5). According to the above findings, exosomes generated from ADSCs following TSPAN6 overexpression are more favorable for wound healing and can enhance neovascularization at the location of the lesion.

Fig. 12

Wound repair assay in rats. A. Representative images of exosomes promoting skin wound repair in each group during 0 ~ 15 days; B. Superimposed wound surface of each group at different time points; C. Statistics of wound healing rate in each group; D. HE staining (bar: 200 μm) and Masson (bar: 25 μm) staining were performed on wounds of each group on the 15th day; E. Statistics of wound scar width; F. Statistics of collagen deposition. (*: p < 0.05; **: p < 0.01; ***: p < 0.001)

Fig. 13

Rat wound with immunofluorescence staining. A. CD31and αSMA staining were performed on wounds of each group on the 15th day (bar: 50 μm); B. Statistics of CD31 staining in each group; C. Statistics of αSMA staining in each group. (*: p < 0.05; **: p < 0.01; ***: p < 0.001)

Discussion

Here, we show that TSPAN6 positively controls the secretion of ADSCs-Exos. We also find that its exosome secretion was boosted by more than three times, whereas the exosome secretion of ADSCs^TSPAN6− exhibited a downward trend. The ADSCs^TSPAN6+ group exhibited a considerable increase in the number of ILVs within them, as well as an increased quantity and bigger morphology of MVBs in comparison to control ADSCs. Furthermore, we discovered that the cytoplasmic adapter molecule syntenin-1, in conjunction with TSPAN6, regulates exosome secretion. The crucial structures for this interaction are the C-terminal of TSPAN6, the N-terminal, the PDZ1domian, and the PDZ2 domain of syntenin-1. Furthermore, compared to ADSCs-Exos, ADSCs^TSPAN6+-Exos can more effectively stimulate angiogenesis, the migration and proliferation of HUVECs and HSFs, and wound healing.

Researchers used to consider exosomes-a type of double-layered membrane vesicles with a diameter of roughly 40–160 nm-to be “trash” produced by cells. But as more and more data are published, it becomes clear that exosomes may transmit materials and convey information between cells [18]. They can also prevent immunological rejection, which is why clinical researchers are paying close attention to them as a kind of cell-free therapy [19]. The exosomes from ADSCs in particular perform the same function as the original cells, but they are safer and offer significant benefits when it comes to usage and storage [20,21,22]. However, exosome yields have always been poor, and this low yield has limited the clinical application of exosomes in addition to affecting their therapeutic impact and increasing treatment time and expense. Thus, one of the most important issues that need to be resolved in the current exosome-related research is the investigation of easy and effective ways to boost the exosome yield.

In this work, it was discovered that TSPAN6 positively regulates exosome secretion, which is enhanced by over three times when ADSCs^TSPAN6+ are created. Additionally, it was found that TSPAN6 could encourage the development of MVBs and ILVs in ADSCs, indicating that TSPAN6 is directly responsible for the secretion of ADSC-Exos because it is involved in the biogenesis of exosomes. TSPANs are a large class of evolutionarily very conserved cell membrane proteins, with the N-terminal and C-terminal ends located on the cytoplasmic side, interspersed with four highly hydrophobic transmembrane domains (TM1 to TM4) [23]. Exosomes have been shown to be substantially enriched for TSPANs in earlier research. It has been demonstrated that TSPANs, which have 33 members in mammals, control protein transport via separating membrane structures(24). Although some common tetraspanins-like CD9, CD63, and CD81-are important exosome constituents and classical exosome indicators, there hasn’t been much research on the function of tetraspanins in exosome formation. It has been observed that CD63 and syndecan are similar in their ability to bind with syntenin-1, the cytoplasmic adapter; however, CD63 does not have the same effect on exosome secretion as syndecan [25]. The present investigation revealed that there was no significant alteration in the expression of CD9, CD63, or TSG101 proteins in any group’s cells. This finding implies that TSPAN6 has no effect on the expression of these proteins in whole cells. The differential in the quantity of exosomes released by the cells in each group, however, was the primary reason for the significant increase in the expression level of the relevant proteins in the ADSCs^TSPAN6+-Exos group. The protein content increases with the quantity of exosomes. Furthermore, the protein content of exosomes made up a very small percentage of the total protein content of the cells; hence, variations in the protein content of exosomes were unlikely have an impact on the overall protein content of the cells. Consequently, the gray value of the bands of exosome-associated marker proteins can, when suitable, reflect the number of exosomes released by a cell.

It has been demonstrated that syntenin-1 interacts with phospholipids, phosphatidylinositol [4, 5] bisphosphate, and a range of proteins [26]. It is a versatile intracellular adaptor protein that can take part in many different physiopathological processes that occur inside the cell. It controls, among other things, the shape of neural membranes, the trafficking of receptor proteins, the generation of exosomes, the formation of synapses, and the growth of axons [27, 28]. It also plays a role in the pathophysiology of several tumor cell types, particularly when it comes to the control of tumor invasion and stemness [29]. Therefore, the function of syntenin-1’s binding proteins is primarily responsible for its functional variety. In addition to regulating membrane shape through interactions with membrane-bound proteins and the formation of functional complexes on cell membranes, syntenin-1 also plays a role in the subcellular transport of proteins involved in the processes of endocytosis and exocytosis [30]. It has been documented that syntenin-1 regulates the reorganization of internal membrane structures and exosome cargo in addition to interacting with proteins involved in exosome formation [31]. Previous research has demonstrated that exosomes contain a large number of tetraspanins, among which CD63 can interact directly with syntenin-1 [32]. Additionally, the CD63/syntenin-1 protein complex regulates the transport of multivesicular bodies from early endosomes to the plasma membrane. Several researchers have proposed that syntenin-1 could be a new exosome identifier because studies have demonstrated that syntenin-1 is present in exosomes from a range of cell lines and in exosomes acquired through a variety of different isolation methods [33]. Researchers collected exosomes, apoptotic vesicles, and cell microvesicles from a range of cell types to further confirm its universality. They discovered that syntenin-1 was exclusively found in exosomes.

We found that ADSCs^TSPAN6+ had up-regulated syntenin-1 expression, indicating a positive correlation between the two. And transfection of TSPAN6 overexpressing lentivirus could no longer enhance exosome secretion after knocking down syntenin-1 in adipose stem cells, indicating that TSPAN6’s function in boosting exosome secretion was blocked at the time. This finding unequivocally highlights syntenin-1’s critical function in exosome secretion. Our research also shows that TSPAN6 can directly interact with syntenin-1. The C-terminal of CD63, the PDZ1 domain, and the C-terminal of syntenin-1 are the actual sites of interaction between CD63 and syntenin-1 [32]; however, the precise binding location of TSPAN6 to syntenin-1 is still unknown. Since both TSPAN6 and CD63 are members of the tetraspanins family, we hypothesize that their interactions with syntenin-1 may be comparable, meaning that TSPAN6’s C-terminal serves as a critical location for its function.

To confirm the binding location of the interactions, we performed a yeast two-hybrid assay. The critical structures for the interaction, according to the experimental results, are the C-terminal of TSPAN6, the N-terminal, the PDZ1 domain, and the PDZ2 domain of syntenin-1. The fact that TSPAN6 and CD63 are members of the same tetraspanin family and have certain structural similarities supports our previous speculation that TSPAN6’s C-terminal is the critical structure for its function. According to reports, syntenin-1 has an autoinhibitory domain at its N-terminal, and tyrosine phosphorylation at the N-terminal may control this structure [34]. The function of the C-terminal of syntenin-1 has not been clarified, but some have proposed that it is necessary for homodimerization, a crucial step in syndecan-related signaling [35]. The structural domains of syntenin-1, PDZ1 and PDZ2, are known to bind proteins with varying affinities [14]. They can be divided into three groups according to their binding peptide sequences: the first group is called (-S/T-X-Φ), the second is called (-Φ-Χ-Φ), and the third is called (-D/E-Χ-Φ). The letters Φ and X stand for hydrophobic and any amino acid residue, respectively. The majority of peptide ligands preferentially bind to the PDZ2 domain of syntenin-1 due to differences in the binding properties of the PDZ1 and PDZ2 domains, with the PDZ1 domain having a relatively weak affinity for the target protein and the PDZ2 domain having a relatively strong affinity. The globular structure of the PDZs domain is primarily composed of two α-helical chains and six β-folded chains, with an overall length of 80–90 amino acid residues. The binding proteins’ C-terminal peptide can interact with these chains by inserting into the grooves between the β2 chains and the α2 helices. The two PDZs’ structural domains are similar, according to crystal structure studies, but PDZ1’s peptide binding groove is narrower than PDZ2’s, which accounts for PDZ1’s poorer target protein binding capabilities.

Exosomal inclusions, diameter size, and cellular origin all contribute to exosome heterogeneity, which in turn influences how recipient cells utilize exosomes [36]. Since we increased TSPAN6 expression in ADSCs to induce exosome secretion in our study, there may be differences between the exosomes produced in this way and those secreted by regular ADSCs. As a result, we must do an experimental analysis to determine the role of these exosomes. We examined the role of ADSCs^TSPAN6+-Exos based on wound healing because research on ADSCs-Exos has been growing quickly in recent years, particularly in the field of tissue and wound repair [37]. In cellular experiments, we discovered that ADSCs^TSPAN6+-Exos could more effectively support HUEVCs and HSFs proliferation and migration. Upon investigating the pertinent signaling pathways, we found that the HIF-1α/VEGF signaling pathway was significantly activated in HUVECs. HIF-1α is a transcription factor that controls the transcription of a series of downstream genes, such as VEGF, Glut1, and Lactate Dehydrogenase A, among others [38]. Furthermore, it increases the microvessels’ permeability, which permits plasma macromolecules to leak out and deposit in the extravascular matrix. This replenishes the matrix with nutrients, promoting tissue cell proliferation and the formation of new capillaries. In HSFs, we found that the Wnt/β-catenin signaling pathway was significantly activated by ADSCs^TSPAN6+-Exos. The Wnt/β-catenin signaling pathway possesses more than 20 target genes, including c-myc, metalloproteinases, survivin, cyclinD1, and others [39]. It has been demonstrated that the activated Wnt/β-catenin signaling pathway accelerates the cell cycle in G0 phase by up-regulating cyclinD1 [40] and its associated protein kinases (e.g., cyclin-dependent protein kinase 4) and/or down-regulating its inhibitory proteins (e.g., p21, p27). The Wnt/β-catenin signaling pathway plays a multifaceted role in dermal fibroblasts. It has been discovered that the pathway stimulates the proliferation, migration, and differentiation of dermal fibroblasts [41], resulting in an increase in dermal thickness and the promotion of wound healing. In the animal experiments section, by using CD31 and αSMA immunofluorescence, vessel density was double-stained; endothelial cells were stained with CD31 and smooth muscle cells and pericytes were stained with αSMA. In this study, there was no difference in the number of αSMA⁺ vessels; however, more CD31⁺ vessels were observed in the ADSCs^TSPAN6+-Exos group than in the ADSCs-exos group. This is due to the fact that smooth muscle, pericytes, and endothelial cells, which stabilize the vascular structure and regulate its permeability, make up mature blood vessels [42]. Whereas immature vasculature displays αSMA⁻ and CD31⁺ vessels, mature vessels display both of these characteristics. Given that the majority of the vessels formed in the ADSCs^TSPAN6+-Exos group were immature compared to those in the standard exosome group, this could be one explanation for the results mentioned above.

In this work, we discovered a novel gene that can stimulate exosome secretion. TSPAN6 was shown to do this by encouraging the creation of MVBs and ILVs. To increase the range of uses for exosomes in medical applications, we can co-express multiple genes involved in exosome biogenesis (such as Rab GTPase and endosomal sorting complex required for transport proteins) to generate stem cells with an exosome-productive genotype. However, the genes that optimize exosome production without compromising cell viability and function still need to be thoroughly screened. More significantly, there is an unavoidable chance that the exosomes could become contaminated with undesirable virus genes due to the similarity between viruses and exosomes on multiple levels [43]. While genetic modification approaches can enhance our comprehension of exosome secretion and synthesis, the pursuit of augmenting exosome production through these methodologies is nascent.

There are a few other limitations to this study as well. Firstly, it only employed a rat animal model; no experimental validation in human or other animal-derived cells was carried out. Consequently, further validation of TSPAN6 in a variety of cell types is required to confirm that TSPAN6 has a broad function in promoting exosome secretion. Secondly, there has been insufficient analysis of the molecular mechanism by which ADSCs^TSPAN6+-Exos contributes to the promotion of wound repair in animal investigations.

Conclusion

TSPAN6 stimulates exosome secretion and synthesis, as well as the creation of MVBs and ILVs in ADSCs. The C-terminal of TSPAN6, together with the N-terminal, the PDZ1 domain, and the PDZ2 domain of syntenin-1, all play critical roles in controlling the release of exosomes produced by ADSCs. Furthermore, ADSCs^TSPAN6+-Exos has a greater ability to stimulate angiogenesis, enhance the migration and proliferation of HUVECs and HSFs, and support wound healing. From experimental animal models to clinical human applications, there is still a long and difficult path ahead, but we are confident that these obstacles will be addressed in further research.