MSCs are multi-potent adult stem cells possessing multi-lineage differentiation capacity and immunosuppressive properties. The sources of MSC are very extensive (Jo et al., 2021) and they can be harvested from bone marrow, umbilical cord, adipose tissue, synovial membrane, gingiva, and some unconventional sources including exfoliated deciduous teeth, menstrual blood, and fetal dermis of accidentally aborted fetuses (Figure 2) (Wang et al., 2019a; Dalirfardouei et al., 2019; Narbute et al., 2019). Accordingly, MSCs should express CD105, CD73, and CD90, with lack of expression of CD34, CD45, CD14 or CD19, CD79a or CD11b, and HLA-DR. Other markers including CD146, CD271, Stro-1, and SSEA-4 are reported to relate to the stemness and sources of MSCs (Lv et al., 2014). Criteria for isolation, culture, and identification of MSCs have been established by the International Society for Cellular Therapy (ISCT) (Galderisi and Giordano, 2014). The biological roles of MSCs in skin regeneration and repair have been reported widely (Kosaric et al., 2019; Zhang et al., 2020; Malhotra et al., 2021). Collectively, the biological roles of MSCs include controlling excessive immune response in skin wounds for their immunomodulatory capacity (Uchiyama et al., 2017), secreting paracrine factors such as EVs to contribute to the healing process (Qi et al., 2014; Bian et al., 2020), and differentiating into skin repair cells for their multiple differentiation potential (Zhang et al., 2020).

MSCs have been frequently used and examined in clinic. For skin wound treatment, a series of clinical trials were performed to evaluate the efficacy of MSC application in serious skin burns (Falanga et al., 2007), non-healing ulcers (Dash et al., 2009; Li et al., 2013; Chen et al., 2018a; Lu et al., 2019; Moon et al., 2019; Kerstan et al., 2021; Zhang et al., 2022a; Kerstan et al., 2022) and advanced limb ischemia (Yang et al., 2013; Lu et al., 2019). For example, Sheng et al. first reported successful employment of MSCs to realize the regeneration of functional sweat glands in patients with deep burn injury (Sheng et al., 2009). Besides, Lu and colleagues showed that application of autologous BM-MSCs in forty-one patients with DFUs could lead to improved ulcerative healing rate and blood infusion, decreased ulcer relapse and amputation during a 3-year follow-up (Lu et al., 2019). In the light of registered trials on https://www.clinicaltrials.gov, more than three hundred clinical trials based on MSC therapy have been finished in patients with including but not only limited to autoimmune or degenerative diseases (Zhou et al., 2021). Here, we listed the clinical trials in patients with non-healing wounds in Table 2. On the whole, MSCs have been proven to have tolerable safety profile and effectiveness in certain clinical settings. However, a lack of verification on safety and therapeutic benefit from large-scale clinical trials barricades the transfer from bench to bedside, impelling researchers to find superior substitute such as MSC-sEVs.

sEVs are predominately characterized by microstructures, sizes, and surface makers. Investigators mainly observe the vesicle-like or cup-like microstructure of sEVs with the help of transmission electron microscopy (Gurunathan et al., 2019; Jafari et al., 2020). Meanwhile, the size of sEVs could be examined by nanoparticle tracking analysis (NTA) or tunable resistive pulse sensing (Sokolova et al., 2011). Together with typical structure and size, sEVs have exclusive markers on their surfaces. One type of the surface proteins are ESCRT-associated proteins, such as tumor suppressor genes (TSG101) and ALG-2 interacting protein X (Alix). While others are ESCRT-independent markers (CD9, CD63, CD81), flotillin, chaperones, and ubiquitinated proteins (de Gassart et al., 2003; Monguió-Tortajada et al., 2019). However, specificity of some markers (such as RACGAP1) and typical subtype markers to distinguish the subtypes of EVs are still debated, calling for more studies to unveil these doubts in the future (Xu et al., 2017).

sEVs are mainly composed of lipids, proteins, and nucleic acids including mRNAs, non-coding RNAs, and DNAs (Keerthikumar et al., 2016) and therefore are thought to be carriers of intercellular biological information. Furthermore, the biochemical content of sEVs varies according to their origin, and thus, the biological information they carry also differs. The heterogeneity of sEV composition is possibly reflective of their origins, sizes, functional impacts on acceptor cells. Reportedly, MSC-sEVs are also enriched in various bioactive molecules which have good therapeutic effects on diverse diseases (Simons and Raposo, 2009).

Furthermore, employment of MSCs and their derivatives has gained momentum in the field of regenerative medicine (Malhotra et al., 2021). A line of basic and clinical trials have demonstrated that MSCs can create a beneficial microenvironment to regulate inflammatory response, facilitating the formation of a properly-vascularized granulation matrix, enhancing the proliferation and migration ability of skin cells and diminishing apoptosis, thus accelerating wound healing (Mazini et al., 2020). MSC-sEVs have also been verified to be therapeutic in preclinical studies and exert immune-modulating and regenerative effects (Bian et al., 2019). Of note, the establishment of immortalized MSCs impacts on neither the quality nor the quantity of MSC-sEVs. Besides, the immortalized treatment is beneficial for its sustainable and cloneable manufacture (Yeo et al., 2013). Compared with other stem cells, MSCs are safer and more favorable in allogenic application, which is evidenced by various clinical trials (Xiao et al., 2013; Squillaro et al., 2016; Bartolucci et al., 2017; Galipeau and Sensébé, 2018; Ban et al., 2021). Accordingly, we speculate that sEVs from MSCs are also safer than that of other stem cells for allogenic transplantation.

Ultracentrifugation-based isolation has been most widely employed and considered to be the golden technique for sEVs separation on account of accessibility, simplicity, high yield, and harvest of comparatively homogenous size groups of sEVs, yet it is lengthy and labor-intensive and might lead to co-sedimentation of non-vesicular proteins (Pegtel and Gould, 2019; Tian et al., 2020). To fix the disadvantages of routine sEV isolation, multiple techniques based on different rationales have been used, such as ultrafiltration, polymer-based precipitation, immunoaffinity capture, size exclusion chromatography, and combined employment of the techniques (Tian et al., 2020). Ultrafiltration provides a faster substitute to ultracentrifugation. However, the particle yield and purity might be compromised because of absorption to cellulose and membranes and deformation of vesicles owing to pressure (Lobb et al., 2015; Konoshenko et al., 2018). Thus, ultrafiltration might be more suitable with limited liquid volume for clinical grade sEVs. As another suitable technique under clinical research conditions, polymer-based precipitation is also time-efficient, low-cost and leads to a high yield. The non-specific mechanism of it also brings about shortcomings including low quality of sEVs and co-sedimentation of non-vesicular contamination (Sidhom et al., 2020). Notably, immunoaffinity capture can harvest sEVs of higher purity but lower yield and is extensively applied as an adjuvant step with ultracentrifugation to enhance pureness. Drawbacks of this method include selecting one subtype of sEVs, not capturing all sEVs, being expensive, and confining to small sample size (Sidhom et al., 2020). Accumulating data have suggested that size exclusion chromatography is superior in purification of sEVs, which is exceptional in preservation of functionally and morphologically intact sEVs. But researchers should prevent biological target denaturation and take the sample volume into consideration when employing this method (Stulík et al., 2003; Sidhom et al., 2020).

Regrettably, different extraction and purification methods of sEVs may produce different populations of sEVs carrying various functional cargoes including proteins and nucleic acids. Additionally, some methods can also lead to the contamination of sEVs with non-vesicular “contaminants.” Above technical limitations likely bring about contradictory or unrepeatable conclusions in the treatment of sEVs on diseases including non-healing wounds. Therefore, we believe that combined employment of the techniques is probably the best choice to obtain highly purified and intact sEVs which can perform in a replicable manner.

The wound healing cascade starts with hemostasis. The immediate response is vasoconstriction of the injured blood vessels to stop bleeding, followed by platelet aggregation, platelet plug formation and the coagulation cascade activation to form a fibrin clot which halts the blood flow and offers a scaffold for inflammatory cells (Rodrigues et al., 2019). Notably, data about the application of MSC-sEVs in hemostasis phase of the wound healing process is limited. It has been reported that MSC-sEVs possess procoagulant activity in vitro. Specifically, Tiffani et al. demonstrated that EVs, either from AD-MSCs or BM-MSCs, were functionally thrombogenic and likely to enhance clotting rates by expressing tissue factor and phosphatidylserine on their surfaces (Chance et al., 2019). So, we have reasons to expect potential effects of MSC-sEVs in hemostasis stage of wound healing process in vivo in the future.

Generally, inflammatory phase, which occurs at several minutes to hours after skin injury and lasts for several days, stands at the beginning stage of the healing cascade. A well-regulated inflammatory response is essential to trigger healing process of skin wounds. However, malfunction of immune cells, such as macrophages, T lymphocytes, and B lymphocytes, is regarded as the culprit to generate persistent excessive inflammation, which is the typical characteristic of chronic non-healing wounds (Li et al., 2021). Accumulating evidence has suggested that MSC-sEVs exert anti-inflammatory and immunomodulating effects on those cells thus facilitating wound healing (Ti et al., 2015; Monguió-Tortajada et al., 2017; He et al., 2019; Ha et al., 2020). They are able to convert pro-inflammatory M1 macrophages predominately into anti-inflammatory M2 phenotype (Figure 3i) (Ti et al., 2015). Besides, Khare et al. (2018) found that MSC-sEVs from bone marrow could regulate the activation and differentiation of B lymphocytes and inhibit proliferation of some types of immune cells, which may help pave the way for resolution of prolonged inflammation. Also, they could suppress the sensitization and proliferation of natural killer cells to alleviate excess inflammation (Fan et al., 2019). Meanwhile, over-abundant inflammatory cytokines are also common factors for extensive inflammation. Studies have demonstrated that sEVs harvested from bone marrow MSCs (BM-MSCs) pretreated with nuclear factor erythroid related factor 2 (Nrf-2) reduced wound inflammation in diabetic rats by downregulating pro-inflammation cytokines like tumor necrosis factor-α (TNF-α) and interleukin-1β (IL-1β) as well as upregulating anti-inflammation factors such as IL-4 and IL-10 (Wang et al., 2021a). Mechanically, functional cargos, such as non-coding RNAs (Fatima et al., 2017), are listed in Table 3 (Li et al., 2020). Therefore, MSC-sEVs may provide a clinically applicable method in alleviating overwhelmed inflammation in chronic non-healing wound.

Angiogenesis is crucial in promoting wound healing and skin repair in terms of transportation of oxygen, nutrients, and immune cells to wounded sites, facilitating cell proliferation and ECM deposition (Wang et al., 2019b). Although healing process is influenced by many factors, poor vascularization may be the most principle cause for wound chronicity (Cheng and Fu, 2018). Furthermore, the depth causes for insufficient angiogenesis could be the undue inflammatory, excessive oxidative stress, DNA damage, and cell cycle arrest, leading to cell senescence and impairing pro-angiogenic ability of endothelial cells (Pulido et al., 2021). A line of evidence has suggested that MSC-sEVs can enhance proliferation and migration of vascular endothelial cells, thus promoting angiogenesis and blood vessel maturation in diabetic wounds (Figure 3ii) (Shabbir et al., 2015; Kang et al., 2016; McBride et al., 2017; Li et al., 2018; Shi et al., 2020; Guillamat-Prats, 2021; Wei et al., 2021). As concluded in Table 4, these desirable effects were associated with elevated expression of genes involved in proliferation (PCNA, cyclin D3), migration (Konoshenko et al., 2018), and angiogenesis (VEGF, Ang1, and Flk1) (Zhang et al., 2015a; Liang et al., 2016; Hu et al., 2018). At a molecular level, a group of non-coding RNAs are detected to enriched in MSC-sEVs, such as HOTAIR (HOX transcript antisense RNA), miR-221-3p, −21, −125a, −17, and −126, which have been proved to mediate favorable bioeffects by regulating above target genes (Tao et al., 2017; An et al., 2019; Yu et al., 2020; Born et al., 2021; Wei et al., 2021; Pi et al., 2022). For instance, we previously found that miR-17-5p in umbilical cord MSC-sEVs (hucMSC-sEVs) accelerated angiogenesis in diabetic mice via targeting PTEN/AKT pathway, thus exerting positive effects on wound healing (Wei et al., 2021). Besides, application of MSC-EVs has also been demonstrated to be efficacious in mitigating excessive oxidative stress and senescence-associated secretory phenotype of endothelial cells (Wang et al., 2020). To be specific, Zhang and coworkers reported that adipose sourced MSC-sEVs could alleviate oxidative stress, diminish reactive oxygen species (ROS) production and ameliorate mitochondrial function in endothelial cells under high glucose by regulating Sirtuin 3 (SIRT3)/superoxide dismutase 2 (SOD2) activity, thereby achieving better vascularization and diabetic wound healing (Sidhom et al., 2020). Likewise, Xiao et al. found that the treatment with MSC-EVs effectively improved the senescence of HG-treated endothelial cells in diabetic wounds via regulating miR-146a/Src pathway, leading to a decrease in aging-related proteins p21/p16/p53 and an enhanced angiogenesis in diabetic wounds (Xiao et al., 2021). Not only RNAs but also active proteins in MSC-sEVs can stimulate angiogenesis in wound restoration (Sung et al., 2019; Tutuianu et al., 2021). For instance, Angiopoietin-2 (Ang2) and deleted in malignant brain tumors 1 (DMBT1) have been reported to be abundant in MSC-sEVs to exert pro-angiogenic functions during wound healing (Chen et al., 2018b; Liu et al., 2021). Loaded with angiogenic components, MSC-sEVs are expected to be a prominent option to induce angiogenesis to treat poorly vascularized wounds in future clinical practice.

Wound healing can be severely retarded by dysfunctional re-epithelization caused by impaired proliferative and migratory capacities of keratinocytes (Berlanga-Acosta et al., 2020). To restore efficient re-epithelialization is indispensable for successful wound healing. As summarized in Table 5, some studies have observed the potency of MSC-sEVs on boosting re-epithelialization in chronic wounds (Figure 3iii) (Zhang et al., 2015b; Ferreira et al., 2017; Sung et al., 2019; Cheng et al., 2020). Specifically, Tutuianu et al. validated regenerative abilities of BMMSC-sEVs. Exposure to these sEVs enhanced the proliferation and migratory abilities of keratinocytes in vitro (Tutuianu et al., 2021). Consistently, Wang et al. applied adipose MSC-sEVs encapsuled in an antibacterial polypeptide-based F127/OHA-EPL hydrogel to treat diabetic wounds and observed remarkable acceleration of re-epithelialization and angiogenesis in vivo (Wang et al., 2019c). Besides, Zhao et al. found that hucMSC-sEV administration dramatically increased cell proliferation and inhibited apoptosis via restraining apoptosis-inducing factor nucleus translocation (Zhao et al., 2020). While some studies postulate that the miRNA related cargoes containing in MSC-sEVs may be responsible for the bioeffects. For example, Gondaliya et al. observed faster re-epithelialization and enhanced wound repair in diabetic mice treated with MSC-sEVs loaded with miR-155 inhibitor via accelerating keratinocyte migration and enhancing fibroblast growth factor-7 (FGF-7) level (Gondaliya et al., 2022). Additionally, miR-205 is reported to be involved in cell migration and proliferation. Lack of miR-205 leads to epidermal defects because of impaired cell proliferation (Wang et al., 2013). This microRNA regulates AKT activation and, therefore, promotes migration of keratinocytes and enhances wound healing (Yu et al., 2010). Moreover, the presence of miR-205 was found in sEV samples from adipose tissue MSCs through the Next-Generation Sequencing experiments. However, it was reported that an miR-205-independent activation of AKT was responsible for the migration and proliferation of keratinocytes in skin wound healing after exposure to adipose tissue MSC-derived EVs (Ferreira et al., 2017). Taken together, there is no doubt that beneficial effects of MSC-sEVs are potential and remarkable. But more studies are required to determine the essential functional constituents and to elucidate the complexity of MSC-sEVs from different tissues for further development of application.

During normal wound healing, fibroblasts proliferate, migrate, and differentiate into myofibroblasts to participate in synthesizing the ECM, to secrete cytokines and growth factors (Dong et al., 2017), and to enhance wound contraction, thereby promoting wound closure (Darby and Hewitson, 2007). However, these capacities of fibroblasts are impaired in chronic wound microenvironments (Wertheimer et al., 2001), which brings about insufficient ECM production and collagen deposition principally in proliferation phase (within 1 month or longer). The mechanism for this impairment includes cellular senescence induced by excessive oxidative stress and advanced glycation end products in diabetic wounds (Bian et al., 2020). Encouragingly, MSC-sEVs are reported to rejuvenate senescent fibroblasts. For example, sEVs from human placental MSCs significantly improved the biological functions of senescent fibroblasts such as promoting their proliferation and migration, enhancing ECM synthesis, and decreasing the overexpression of matrix metalloproteinases (MMPs) (Bian et al., 2020; Zhao et al., 2021). Further study revealed the mechanism for these effects involving inhibiting the expression of receptor for AGEs (RAGE) and stimulating the activation of Smad signaling pathway in these cells (Bian et al., 2020; Zhao et al., 2021). Interestingly, Zhu and colleagues reported that hucMSC-sEVs could not only promote proliferation and migration of fibroblasts, but also remarkedly boost cutaneous nerve regeneration by stimulating fibroblasts to produce nerve growth factors (NGFs), which can stimulate nerve regeneration, re-establish local sensory innervation homeostasis, and enhance angiogenesis, thereby achieving an ideally regenerative healing both morphologically and functionally (Zhu et al., 2022). Additionally, ADSC-Exos can transport miRNAs, lncRNA, and functional proteins (Choi et al., 2018; Cooper et al., 2018; Parvanian et al., 2020; Qian et al., 2021) to promote the migration, proliferation, ECM secretion of fibroblasts (Hu et al., 2016; Wang et al., 2021b).

Pathological scars including keloids and hypertrophic scars arise from excessive wound healing after chronic inflammatory stimulation. The pathological mechanism is the aberrant activation of fibroblasts and myofibroblasts, which synthesize more ECMs in scar tissue, mostly via the abnormal activation of transforming growth factor-β (TGF-β)/Smads pathway and Yes-associated protein predominantly during remodeling phase (at least over 14 days to 1 month) (Finnson et al., 2013; Zielins et al., 2014; Song et al., 2018; Clark, 2021). Recently, multiple sourced MSC-sEVs have been proven effective in achieving scarless and high-quality wound regeneration (Figure 3iv) (Wang et al., 2017; Dalirfardouei et al., 2019; Jiang et al., 2020a). Zhang et al. found hucMSC-sEVs treatment worked smartly not only as an activator of the Wnt/β-catenin signaling pathway to heal impaired skin but also an inhibitor of the signal via exosomal 14-3-3ζ mediated YAP regulation to avoid scar formation (Zhang et al., 2016). Similarly, in another study, Zhang et al. reported that miR-21-5p and miR-125b-5p in MSC-sEVs from umbilical cord blood played critical roles in suppressing myofibroblast differentiation from fibroblasts, thereby favoring scarless wound healing (Zhang et al., 2021). In line with the same idea, Fang et al. examined miRNA profiles in hucMSC-sEVs by high-throughput sequencing and verified that a couple of miRNAs (miR-21, −125b, −23a and −145) decreased myofibroblast formation via suppression of transforming growth factor β (TGF-β) Smad2 pathway (Fang et al., 2016). Encouragingly, similar therapeutic effects were also obtained by another team when they applied MSC-sEVs loaded with tumor necrosis factor-α stimulated gene-6 (TSG-6) or miR-138-5p (Jiang et al., 2020b; Zhao et al., 2022). Additionally, the wounds treated with ADSC-Exos also demonstrated faster wound healing and less collagen deposition (Li et al., 2022). So we conclude that MSC-sEVs exert multiple-mechanism therapeutic effects on wound regeneration (as demonstrated in Table 6).

Most human skin wounds heal with scar and without skin appendages including hair follicles and sweat glands. This outcome seriously affects the appearance of patients and leads to the damage of skin physiological function. Recently, MSC-sEVs were reported to entice regeneration of hair follicles (Table 7). ADSC-EVs containing a variety of cytokines stimulated hair follicle growth by activating the Wnt signaling pathway which is essential to hair follicle induction (Zhang et al., 2022b). Wnt3a is enriched in sEVs from hBMMSCs and BMMSC-sEVs can activate the Wnt/β-catenin signaling in recipient dermal papilla (DP) cells (Rajendran et al., 2022). Our research also implied the hair follicle regeneration potential of sEVs from DP cells (Zhang et al., 2022c). Another group of researchers have successfully designed miR-218-5p abundant sEVs derived from dermal papilla cells and observed a notably promoted hair follicle development by up-regulating β-catenin in mice (Hu et al., 2020). Therefore, MSC-sEVs promote the regeneration of hair follicles mainly by the activation of Wnt/β-catenin pathway (Hu et al., 2020). Additionally, Chen et al. also observed that TGF-β1-enriched MSC-sEVs could realize a faster reconstruction of sweat gland function in wounded skin (Chen et al., 2022a).

Short half-time and rapid clearance in vivo have been the challenges for injecting free sEVs in practical applications (Golchin et al., 2022). Due to high designability of biomaterials, researchers have combined sEVs with innovatively designed biomaterials to prolong the retention time to increase their utilization, and achieve on-demand release in vivo (Las Heras et al., 2020). In this section, we discussed and overviewed some ingenious symphonic designs between sEVs and advanced biomaterials like hydrogel pad, microneedle array, microspheres or electrospun nanofibers.

Changing culture environments of parent cells has been reported to be beneficial to increase the yield of sEVs. For example, a bioreactor composed of hollow fibers (Watson et al., 2016) can sustain larger numbers (more than 10⁹ per mL) of cells and produce highly concentrated cell culture supernatants, thus leading to mass production of sEVs. Using this bioreactor, the yield of sEVs can be improved about 40 folds than conventional system. Additionally, oxygen environment like oxidative stress (Atienzar-Aroca et al., 2016) or hypoxic (King et al., 2012) stimulus was also reported to be responsible for the yield of sEVs. Sandra Atienzar-Aroca et al. observed that retinal pigment epithelium cells (RPEs) secreted larger amounts of sEVs when they were exposed under proper oxidative stress (achieved by treating cells with 40 or 80 nM ethanol) (Atienzar-Aroca et al., 2016). The yield of sEVs can be enhanced about 2 folds when the concentration of ethanol in the cell culture medium of PREs was 80 nM, in comparison with that without ethanol. In another study, it is reported that exposing tumor cell lines (MCF7, SKBR3, and MDAMB 231) to the moderate (1% O₂) and severe (.1% O₂) hypoxia environments significantly increased the number of sEVs present in the conditioned media (King et al., 2012).

Applying external chemical or physical stimulus has been reported to regulate cellular metabolic behavior of parent cells to improve the yield of sEVs. Up to now, several chemical agents have been investigated to affect and improve the secretion of sEVs including BG (Wu et al., 2021), metformin (Liao et al., 2021), cytochalasin B (Pick et al., 2005), monesin (Savina et al., 2003), and serotonin (Glebov et al., 2015) by adding them to the culture medium of parent cells. These reagents affect the production of sEVs through different signaling pathways in parent cells. For example, the ion products of BG can upregulate the expression of neutral sphingomyelinase-2 (nSMase2) and Rab27a to enhance nSMases and Rab GTPases pathways which is essential to regulate the vesicle formation and membrane traffic in human-derived MSCs (Wu et al., 2021). However, metformin was reported to activate autophagy-associated pathway in MSCs to enhance the production of sEVs (Liao et al., 2021). Monesin (a Na⁺/H⁺ exchanger) can induce exchanges in intracellular calcium, thus stimulating sEV release from K562 cells. Similarly, the involvement of cAMP- and Ca²⁺- dependent signaling pathways was reported as the underlying mechanism of serotonin in the regulation of EV secretion from microglia cells (Glebov et al., 2015).

In addition to those chemical or bio-stimulation, applying high frequency acoustic (Ambattu et al., 2020), low level electricity (Fukuta et al., 2020), mechanical forces with the combination of 3D scaffold (Guo et al., 2021), lower pH (Parolini et al., 2009; Ban et al., 2015), and ionizing radiation (Jabbari et al., 2019) can also increase the yield of sEVs from various cells. For instance, Ambattu et al. reported that applying an AC electric field with high frequency acoustic of 10 MHz on U87-MG and A549 cells with 7 cycles over 280 min achieved about 8 to 10 folds enhancement in the production of sEVs (Ambattu et al., 2020). By the treatment of relative lower electric intensity (constant current, .34 mA/cm²) for 60 min, the quantity of sEVs from 3T3-Swiss albino cells and B16F1 cells was enhanced by 1.7 and 1.26 folds, respectively (Fukuta et al., 2020). Guo et al. seeded dental pulp stem cells (DPSCs) or MSCs on Fibra-Cel scaffolds to fabricate a 3D bioreactor and explored the effect of mechanical stimulation with flow stimulation on the yield of sEVs. Surprisingly, the yield of sEVs from above 3D bioreactor under .5 mL/min flow stimulation was improved by about 40.7 and 3.4 folds than the 2D and 3D static counterparts, respectively (Guo et al., 2021).

Overexpressing related regulatory proteins involved in sEV biogenesis via gene transfection can also ensure high yields of sEV production. STEAP3, syndecan-4, and L-aspartate oxidase were identified as three sEV production boosters that were involved in sEV biogenesis, the formation of multivesicular bodies, or cellular metabolism to yield more sEVs with high quality (Kojima et al., 2018). In another similar study, activating actomyosin was believed to pull the actin-bound MVB toward the cellular plasma membrane to facilitate the secretion of sEVs (Mittelbrunn et al., 2015). Kai O. Böker et al. reported that overexpression of tetraspanin CD9 significantly augmented the amount of sEVs secreted by different human cell lines, like HEK293, HeLa, SH-SY5Y, as well as B and T lymphocytes (Boker et al., 2018). Other methods such as fusing with synthetic lipid (Jhan et al., 2020), changing cell culture parameters or collection frequency of sEVs (Patel et al., 2017) were also developed to achieve higher yield of sEVs. Although many studies have explored how to improve the production of sEVs, standard operations and unified understanding of underlying mechanisms need to be investigated in depth.

One of the main reasons that sEVs have greater advantages over MSC therapy is that it can be customized to promote wound healing process through on-demand engineered approaches on the parent cells or the isolated sEVs summarized in Table 8. Generally, according to processed objects, developed engineering methods can be classified as engineering on parent cells or engineering on the isolated sEVs.

Up to now, several methods have been developed by researchers to customize the contents of sEVs by operating parent cells. Genetic engineering is recognized as a common strategy to enrich non-coding RNAs and proteins in the lumen or on the surface of sEVs. By virtue of the transfection of hBMSCs by miR-29b-3p lentiviral vector, the miR-29b-3p-enriched sEVs were generated to suppress excessive capillary proliferation and collagen deposition in the late proliferation and remodeling phases of wound healing. By transfecting HEK293 cells with H19-overexpressing (H19-OE) lentiviral vector and followed by a gradient extrusion method, long non-coding RNA (LncRNA)-H19 was encapsuled into sEV-mimetic nanovesicles (Tao et al., 2018). The obtained H19-sEV-mimetic nanovesicles were able to counteract the regeneration-inhibiting effect of hyperglycemia, thus accelerating the healing process of diabatic wounds. Similar method was employed to load SERPINA1, SERPINF2, and SERPING1 into sEVs respectively to promote healthy ECM and facilitate the formation of a beneficial fibrin scaffold and the resolution of the inflammation phase during wound repair (Park et al., 2022). Moderating culture environment of parent cells can also modulate the content or the therapeutic effect of sEVs. The angiogenesis efficacy of MSC-derived sEVs can be improved by applying an ischemic preconditioning on MSCs for 1 h (Li et al., 2015), which might be a useful solution to boost angiogenesis at wound site. Very recently, our colleagues reported TGF-β1-enriched sEVs by pretreating HUMSCs with isoproterenol (ISO) under anoxic environments. In this study, ISO was employed to promote the secretion of TGF-β1, while the hypoxic environment could boost the yield of sEVs. The engineered sEVs enhanced the migratory behavior and stem cell properties of epidermal keratinocytes and accelerated wound re-epithelization in vivo. To mimic inflammatory microenvironment, lipopolysaccharide (LPS) was employed to stimulate hUC-dMSCs to generate sEVs enriched with miRNA let-7b. They could mitigate inflammation and accelerate diabetic wound healing by upregulating the expression of anti-inflammatory factors and inducing macrophages to shift to anti-inflammatory M2 phenotype (Ti et al., 2015). Pretreating adipose MSCs with inflammatory cytokines like IFNγ or TNFα would enhance the anti-inflammatory efficacy of adipose MSCs-derived sEVs, shifting macrophages from M1 to M2 phenotype (Domenis et al., 2018), which provided a clue for the management of inflammatory process at wound site.

Moreover, several strategies were also proposed to augment the targeting efficacy or therapeutic efficiency of isolated sEVs for accelerating wound healing. To improve therapeutic efficiency, different kinds of cargoes like non-coding RNA, proteins, and drugs were loaded into sEVs. Co-incubation may be the simplest solution to encapsule cargoes into sEVs (Pascucci et al., 2014). Very recently, our group loaded VH298 (a stabilizer of HIF-1α) into sEVs successfully by co-incubating them at 37°C for 1 h (Wang et al., 2022). The VH298-loaded sEVs facilitated the migration and tube formation of HUVECs by activating HIF-1α pathway, thus promoting angiogenesis during wound healing process. Unlike the soft phospholipid bilayer membrane of their parent cells, the high level of cholesterol and sphingomyelin on the membrane of sEVs makes them rigid and difficult to open at room temperature. Therefore, enhancing the permeability of phospholipid bilayer membrane of isolated sEVs is the prerequisite to regulate the contents of isolated sEVs. It has been reported that aggressive environments are needed to destroy membrane structures of sEVs and promote drug fusion. In general, the permeability of phospholipid bilayer membranes can be increased by electroporation (Kamerkar et al., 2017; Usman et al., 2018; Lv et al., 2020), freeze-thaw (Gondaliya et al., 2022) and so on. A modified calcium chloride (CaCl₂) transfection was used to load miR-155 inhibitor into the lumen of BM-MSC-sEVs by letting the mixture of miR-155 inhibitor, sEVs, and CaCl₂, as well as PBS go through a freeze-thaw cycle. The synergistic effects between miR-155 inhibitor and sEVs facilitated the migration of keratinocytes and rebalanced FGF-7 and MMPs levels, thereby accelerating wound healing (Gondaliya et al., 2022). In another study, electroporation was employed to encapsule miR-21-5p mimics into adipose MSC-sEVs. The engineered sEVs enhanced the proliferation and migration behaviors of keratinocytes by activating Wnt/β-catenin pathway and then facilitated diabetic wound healing in vivo (Lv et al., 2020). The great progress in increasing the yield of sEVs and in engineering sEVs will further promote the clinical applications of sEVs in chronic non-healing wounds.