Extracellular vesicle microRNA cargoes: candidates for diagnosis and targeted therapy of cutaneous wound healing

Abstract

Background

Non-healing wounds lead to tissue dysfunction, contributing to pathological conditions and compromising overall health. Recently, microRNAs have been shown to regulate diverse biological events, including cell migration, proliferation, angiogenesis, scar formation, and wound closure. Interestingly, cells selectively sort microRNAs into extracellular vesicles, which can be up-taken by cells to regulate the healing process. Consequently, extracellular vesicles and their microRNA content serve as an in-situ reservoir, providing tissue-specific signals on demand.

Aims and methods

This review aims to explore the association between microRNAs and the wound healing process, and the roles of extracellular vesicle-derived miRNAs in different cell populations within the wound microenvironment. We searched PubMed, Web of Science, and Scopus using the keywords: exosomes, mesenchymal stem cell-derived exosomes, wound healing, and microRNA. Relevant data were collected and summarized.

Results

This review provided information about extracellular vesicles, microRNA biogenesis, and their roles in wound healing. Results highlighted the roles of extracellular vesicle-derived microRNAs in anti-inflammatory responses, promoting cell migration, proliferation, and differentiation while inhibiting apoptosis during wound repair. Beyond their natural functions, extracellular vesicles and their microRNA cargoes hold significant potential for therapeutic and diagnostic applications in wound healing. We also outlined the potential use of extracellular vesicle-derived microRNAs as drug delivery systems and miRNA-based monitoring tools.

Conclusion

Extracellular vesicle-derived microRNAs play important roles in wound healing. Despite their promise, challenges such as lack of standardization in isolation, precise delivery, and reliable quantification methods remain obstacles for clinical implementation in disease diagnosis and prognosis.

1. Background

Skin is a vital organ for our body due to its role as a physical barrier to protect against the invasion of external insults. When the skin is damaged, the organism’s survival is threatened, leading to the rapid activation of repair mechanisms to restore skin integrity and keep the organism alive and functioning. The wound-healing process in most multicellular organisms, including humans, involves a well-orchestrated series of overlapping phases. In humans, these are described as haemostasis, inflammation, migration, and tissue remodeling that sequentially re-establish homeostasis, remove pathogens and non-vital tissue, close the wound, and restore the function of tissue [1]. Each phase is regulated strictly by different biological components, encompassing cytokines, various growth factors, and extracellular matrix proteins, which are produced by multiple cell populations.

Normally, the majority of wounds heal in most individuals under normal circumstances. Nevertheless, people who suffer from advanced systemic diseases, such as diabetes, age-related conditions, and coronary vascular disease, often experience wound healing defects [1]. The expression of multiple factors in the wound healing process is a disorder in these patients; however, the exact nature of these factors and their impact on wound healing remains to be fully understood.

Extracellular membrane vesicles (EVs) are lipid membrane-enclosed vesicles and are secreted by almost all cell types. For a long time, EVs were regarded as ‘cell dust’ and thought to be waste products released solely by cells cultured in vitro. In fact, EVs can be extracted from every body fluid and contain diverse bioactive molecules, including RNA, DNA, miRNAs, protein, organelles, and lipids [2–6]. These molecules reflected the pathological and physiological characteristics of parental cells. In addition, the lipid bilayer is adorned with molecular markers that reflect their origin [7], facilitating receptor-mediated recognition and cargo delivery to target cells via membrane fusion or endocytosis [8]. In this way, EVs stemming from different cell populations compose distinct EV cargoes and exert varying effects on target cell responses [9]. Certainly, it is hypothesized that the cargo contents of EVs not only reflect the disease conditions of parental cells but also contribute to the disease phenotypes of recipient cells. Thus, employing EV cargoes is rapidly emerging as an alternative to traditional approaches for diagnostics and therapeutics.

MiRNAs, a subtype of small non-coding RNAs, are included in the repertoire of molecular cargoes linked with EVs. MiRNAs serve as key regulators of transcription and translation, thereby influencing cellular functions and phenotypes. Indeed, the expression of specific miRNAs, including miR-20, miR-199a, miR-429, and miR-34a, is associated with the epidermis and hair follicle maturation and development [10]. Additionally, miR-21 and hypoxia-induced miR-210 expression have been reported to regulate the formation of granulation tissue, keratinocyte proliferation, and wound closure [11]. Moreover, miRNAs also regulate angiogenesis by targeting genes that encode angiogenic factors [12], thereby contributing to healing enhancement or inhibition. For example, miR-126 loaded in exosomes (EXs) derived from bone marrow mesenchymal stem cells (BMMSCs) has shown the potential to profoundly increase tube formation, presenting a promising strategy to accelerate wound healing in vivo [13].

Taken together, miRNAs included in EV cargoes secreted from wound healing-related cell populations, such as fibroblasts, immune cells, and endothelial cells, present a non-invasive and first-of-its-kind screening tool for wound monitoring, potentially with sufficient power to inform wound type and healing trajectory. Accompanying screening tools, miRNAs packed in EVs secreted from other cell types, such as stem cells, also demonstrated the potential for wound repair.

EV classification

Many studies have demonstrated that nearly all cell types release phospholipid-membrane-bound vesicles into the extracellular space through an evolutionarily conserved process. These EVs, overlapping in size and molecular markers, can be categorized into three main types, including apoptotic bodies (ABs), microvesicles (MVs), and EXs, based on their distinct formation and release mechanisms. Among these, EXs have been extensively studied. They demonstrate a cup-shaped morphology with diameters from 30 nm to 250 nm [14]. EXs are formed through an inward budding process inside multivesicular bodies (MVBs), which later fuse with the plasma membrane to release the vesicles. The EX biogenesis is regulated by two pathways: (1) the endosomal sorting complex required for transport (ESCRT)-dependent and (2) ESCRT-independent pathways [15]. In addition, transmembrane proteins, such as CD63, CD9, and CD81, and other plasma membrane-associated proteins, are typically present in EXs and are often more abundant in the vesicles than in the lytic cells [16]. Recently, it has been demonstrated that EXs lack proteins from the nucleus and mitochondria, but may contain low levels of proteins from the Golgi apparatus and endoplasmic reticulum, which are considered ‘non-exosomal markers’ [16].

MVs range from 100 nm to 1000 nm in diameter and exhibit irregular shapes under the microscope [17]. The MV biogenesis involves intricate machinery that enables cargo transport, lipid segregation, and vesicle fission [18]. By contrast with EXs, MV formation occurs through outward budding of the plasma membrane at sites where alterations in protein composition and phospholipid translocation regulate the plasma membrane curvature and rigidity [19]. Although this mechanism is not yet fully understood, phosphatidylserine, normally positioned at the inner leaf of the plasma membrane, has been observed to be translocated to the outer leaf, triggering MV generation [20]. Besides, Ras-related GTPase ADP-ribosylation factor 6 and components of the ESCRT system are also requisite for this process [19]. Additionally, the interaction between arrestin domain-containing protein-1 and TSG101 may lead to the re-localization of TSG101 from endosomal membranes to the plasma membrane, influencing membrane curvature [21]. Like EXs, MVs are expected to contain components involved in their formation and release, such as myristoylation, palmitoylation, ceramide, and cholesterol [22,23].

ABs are apoptosis products and exhibit heterogeneous morphology with diameters ranging from 1 µm to 5 µm [15]. During programmed cell death, cells experience significant morphological changes, nuclear chromatin, and organelle condensation. Subsequently, the plasma membrane blebs and cellular components disintegrate into different membrane-enclosed vesicles [24] known as apoptotic bodies. Therefore, ABs may carry nuclear fragments and cellular organelles. Additionally, higher levels of a protein associated with the nucleus (i.e. histones), mitochondria (i.e. HSP60), Golgi apparatus, and endoplasmic reticulum (i.e. GRP78) would be expected in ABs. In normal physiological conditions, ABs attract macrophages due to the translocation of the phosphatidylserine residue to the external membrane surface. Although many studies still demonstrated the role of ABs in wound healing, this EV population has not yet been utilized in clinical applications.

2. MicroRNA biogenesis and association of microRNAs with EVs and wound healing

2.1. MicroRNA biogenesis

MiRNAs are short and small non-coding RNAs, typically 19–24 nucleotides in length, which have recently been identified as regulators of mRNA cleavage or translational repression [25]. Primary miRNAs stem from the introns of protein-coding host genes, and they need to undergo multiple processing steps to become mature miRNAs [25,26]. Estimatively, miRNAs regulate the translation of more than 60% of protein-coding genes through multiple pathways [27,28]. RNA-induced silencing complex (RISC)- incorporated miRNAs complement partially with their target mRNAs at the 3′ untranslated region (3′ UTR) [29]. Typically, this interaction occurs through the pairing of nucleotides two and seven at the 5′ end of the miRNA to the 3′ end of the target mRNA [30,31]. As a result, miRNAs can suppress their target mRNAs either by degrading them or inhibiting the translation process [32] (Figure 1).

Figure 1.

miRNA biogenesis and function. Initially, miRNAs are transcribed from DNA into primary miRNA (pri-miRNA) by RNA polymerase II (pol II). Drosha then cleaves the hairpin structure from pri-miRNA to form precursor miRNA (pre-miRNA), which is subsequently transported to the cytoplasm by the GTP-dependent exportin-5 complex. In the cytoplasm, Dicer removes the loop domain from the pre-miRNA, resulting in the formation of a double-stranded miRNA. Following the degradation of one miRNA strand, the complementary strand becomes the mature miRNA. The mature miRNA and its complementary sequence of target mRNAs are then complexed with the RISC. Through this complex, individual miRNAs can regulate the stability (via degradation or cleavage) or repress the translation of one or many target mRNAs.

2.2. Detection of miRNAs in extracellular vesicles

During EV biogenesis, various cargoes were enriched in EVs, including a series of nucleic acids, proteins, and lipids [33]. Among these therapeutic molecules, miRNAs have recently emerged as an essential factor in regulating both pre- and post-transcriptional gene expression [33,34]. As a consequence, cell behaviors are altered, affecting the wound healing process, such as promoting fibroblast and endothelial cell proliferation, migration, and angiogenesis while inhibiting apoptosis.

As mentioned, EVs from various sources also contain genetic materials, including miRNAs [35–37]. After 121 miRNAs in mast cell-derived EXs were detected by Valadi et al. [37] and 33 expressed miRNAs were also found in blood-derived EXs by Hunter et al. [38], many other studies have also reported the presence of miRNAs in EVs from saliva, cancer cells, and primary cells [3,38–45]. Interestingly, besides the increasing identification of EV-derived miRNAs, their functions in recipient cells are also being deciphered [38,45,46]. Most miRNAs, which were detected in blood-derived EXs, are responsible for the regulation of blood cell differentiation and metabolic pathways [38]. Additionally, exosomal miRNAs originating from T cells that were transferred to antigen-presenting cells at the immune synapse, influencing gene expression in recipient cells [42]. Moreover, miR-124a in exosomal neurons regulates the expression of astroglial glutamate transporter 1 (GLT1) indirectly involved in extracellular glutamate concentration [47].

It is highlighted that selectively sorted miRNAs into EXs depend on the origin, as well as the physiological or pathological status of secreting cells, making these miRNAs potential diagnostic biomarkers. Kim et al. have discovered that six exosomal miRNAs (miR-3565, miR-3124-5p, miR-200b-3p, miR-6515, miR-3126-3p, and miR-9-5p) were up-regulated and one exosomal miRNA (miR-92b-5p) was down-regulated in small cell lung cancer, compared to normal samples [48]. Moreover, the three-miRNAs panel (miR-200b-3p, miR-3124-5p, and miR-92b-5p) remarkably enhanced the diagnostic accuracy in small cell lung cancer [48]. Similarly, there is a difference between exosomal miRNA concentration in plasma from lung adenocarcinoma patients and the control group [49].

As mentioned, EV miRNAs could be delivered between cells at a distance, contributing to various EV-mediated biological functions [38,40,41]. With the ability to recognize multiple mRNA targets through sequence complementarity and RNA-binding proteins, miRNAs that regulate gene expression in both pre- and post-transcription stages have opened new options for designing therapeutic agents [25,38,42,47].

2.3. Association of microRNAs with wound healing

Concerning skin cell proliferation and migration, miR-210, miR-205, and miR-184 are all expressed throughout the proliferative stage [50–52]. Specifically, miR-205 inhibits the expression of SH2-containing phosphoinositide 5-phosphatase (SHIP2), affecting AKT signaling pathways, which decrease keratinocyte apoptosis and promote keratinocyte migration [51,52]. Conversely, miR-184 antagonized miR-205 to maintain the expression of SHIP2 protein, thereby dampening AKT signaling, which increases keratinocyte apoptosis and cell death [51,52]. Also, miR-210 is hypoxia-sensitive and has been shown to target E2F transcription factor 3, profoundly affecting keratinocyte migration [50,53]. Moreover, the expression of miR-21 and miR-31 is respectively upregulated by transforming growth factor beta 1 and 2 (TGF-β1 and TGF-β2). Thus, these miRNAs promote keratinocyte migration and proliferation by targeting the invasion of T-lymphoma, tissue inhibitor of metalloproteinase 3 (TIMP3), epithelial membrane protein 1 (EMP1), and metastasis-inducing protein 1 (TIAMI1) [54,55]. The influence on keratinocytes has been observed in other miRNAs, although the targeted genes were different. For example, in keratinocytes, the production of inflammatory chemokines, including C-X-C motif chemokine ligand 5 (CXCL5), interleukin 8 (IL8), and CXCL1 were suppressed by miR-132, thus leading to the reduced activity of endothelial cells activating blood vessel formation, as well as decreasing the force of attraction between neutrophils and the sites of wounded skin [56]. Furthermore, miR-132 inhibited the nuclear factor kappa B (NF-kB) pathway, increased the signal transducer and activator of transcription (STAT3), and extracellular signal-regulated kinase (ERK) pathways activation by targeting heparin-binding EGF-like growth factor (HB-EGF), facilitating the transition from the inflammation to the proliferation phase [57]. Other evidence also points to other various miRNAs, such as miR-146a, miR-198, and miR-200b, contributing to the healing process through enhancing cellular growth, namely [58–60]. Nevertheless, there is still uncertainty surrounding the mechanism of this enhancement, which requires a more thorough investigation.

Numerous miRNAs have been documented to modulate angiogenesis, such as miR-1, miR-23a, miR-21, miR-133a/b, miR-29b, miR-939, miR-377, miR-126, miR-146, miR-200, miR-210, miR-218, and miR-4530 [60,61]. miR-200b, for instance, enhanced the expression of vascular endothelial growth factor receptor 2 (VEGFR2) and GATA binding protein 2 (GATA2). Therefore, the downregulation of miR-200b led to VEGFR2 and GATA2 being derepressed, thus enabling wound angiogenesis [60]. Furthermore, basic fibroblast growth factor (bFGF) and VEGF-A enhanced the expression of miR-132 and miR-130a, which specifically target proangiogenic stimuli, including GAX, HOXA5, and p120RasGAP, respectively [62,63].

Regarding the regulation of miRNAs on the healing process’s remodeling phase, miR-29b and miR-29a, which were controlled by TGF-β1, platelet-derived growth factor subunit B (PDGF-B), and IL4, regulated the expression of collagen (COL) (type I, III, IV, and V), SMADs, and β-catenin in normal skin fibroblasts [64–66]. While some miRNAs enhanced cutaneous wound healing, overexpression of miR-21 and miR-130a delayed re-epithelialization and then inhibited wound healing [11]. In addition, miR-21, miR-29, miR-125b, miR-23a, and miR-145 were responsible for the inhibition of SMAD7 and phosphate and tensin homolog (PTEN) expression, causing an enrichment in collagen deposition that resulted in the formation of an abnormal scar in the remodeling phase [61,67,68]. These data suggest numerous therapeutic targets for the regulation of the wound healing process utilizing miRNAs, namely scar formation, angiogenesis, cell migration, or promotion of wound re-epithelialization.

2.4. Mechanism of sorting miRNAs into EVs

Recent studies have proposed four main mechanisms for the process of loading miRNAs into EVs, although their understanding of these mechanisms is still unclear. Firstly, the neutral sphingomyelinase 2 (nSMase2)-dependent pathway, which nSMase2 plays a part in packaging miRNAs into EVs, since overexpressing nSMase2 increased the miRNA number, while blocking nSMase2 produced the opposite result [69]. Secondly, the miRNA motif and sumoylated heterogeneous nuclear ribonucleoproteins (hnRNPs)-dependent pathway: when a sumoylated hnRNPA2B1 binds with GGAG motifs at the miRNA sequences’ 3′ portion, these miRNAs then go through packaging into EXs [70]. Furthermore, the binding of hnRNPC and hnRNPA1 proteins to miRNAs indicates these proteins directly sort miRNAs into EVs [70]. This pathway is similar to a recent study demonstrating that motifs, such as CGGGAG, are recognized by Alyref and FUS proteins, further regulating miRNA export into EVs [71]. Thirdly, the 3′-end of the miRNA sequence-dependent pathway: the uridylated 3′ end is a crucial signal determining the directions of miRNAs sorting to become EVs for this pathway, whereas the adenylated 3′ end is required for the enrichment of miRNAs in cells [72]. Lastly, the specific sequences of miRNAs direct their incorporation into extracellular vesicles (EVs) by preferentially binding to argonaute 2 (AGO2), the central component of the miRISC complex [73]. AGO2 knock-out decreases the number of miRNAs that were transported into EVs, such as miR-150 and miR-451, into HEK293T-derived EXs [74]. In conclusion, specific proteins, enzymes, and sequences in certain miRNAs may contribute to the selective sorting of miRNAs into EVs.

3. Association of EV-derived microRNAs with different cell populations in wound microenvironment

3.1. EV interaction at the wound microenvironment

In the injured microenvironment, EVs can facilitate healing through different mechanisms by interacting with neighbor cells or cells at a distance. EVs can (1) bind to target cells’ surface receptors and then trigger the downstream signalling pathways, (2) be internalized into the cytoplasm and subsequently induce intracellular signalling by releasing the cargo contents, or (3) directly fuse with the cellular membrane [15]. Accordingly, it can promote the wound healing process and result in variations of recipient cells [15].

The interaction between EXs and recipient cells, classified as a ligand-receptor interaction, may result in downstream activation of cellular signalling or transfer of EXs’ contents into the recipient cells. Internalization of EXs into the target cells has been documented in many reports [75–77], but the underlying mechanism of the binding between target cells and EXs, which leads to the activation of internalization, remains unknown. Evidence suggests that proteins mediating EV uptake, such as tetraspanins, lectins, proteoglycans, integrins, immunoglobulins, and lipid rafts, facilitate endocytosis and EV attachment [78,79]. Recently, Joshi et al. proposed that in the cytoplasm, EVs react to pH for conducting fusion with the endosomal membrane and release their contents [80]. Nonetheless, little is known about this mechanism.

An alternative way to facilitate the delivery of EV content into target cells is to fuse the EV membrane with the cellular plasma membrane directly. This mechanism may be controlled by soluble N-ethylmaleimide attachment protein receptor (SNARE) proteins, regulating target specificity and fusion in intracellular vesicle trafficking [81]. In addition, when bringing together two lipid bilayers, the outer leaflets are in direct contact, the result is the emergence of a hemifusion stalk, which can expand to produce the hemifusion diaphragm bilayer, facilitating the fusion [82].

Considering EVs’ innate ability to transport bioactive molecules, developing targetable EVs is of paramount importance in precision medicine. EXs may possess targeting behavior and display an inclination for specific cell types or tissues. The ‘cell specificity’ of EVs is also based on the recipient cells’ phagocytic activity [83]. Their high phagocytic capacity may be the key to explaining their internalization of EVs with such high efficiency. Present chemical and genetic engineering technologies can be applied to target ligands, specifically those on the EV surface, namely lipid insertion, click chemistry, genetic modification, metabolic labeling, affinity binding, and enzymatic ligation [84]. By genetically modifying donor cells, the targeting moieties could be genetically fused to vesicle membrane proteins or lipid-binding proteins/peptides, thus spontaneously expressing on the EV surface [84]. In terms of surface modification, targeting proteins or peptides (1) are directly inserted into EV membrane through a simple mixing and incubation, (2) react with or bind to reactive groups on membrane molecules, (3) react with metabolically altered membrane molecules through Click chemistry, (4) are linked to affinity molecules or conjugated to EV membrane protein by enzyme ligase, or (5) spontaneously fused with EV membrane, displaying functional groups on the surface of the hybrid vesicles [84–87]. These techniques for EV surface modification have shown promising results in tissue engineering, targeted therapy, and cellular immunotherapy. To address potential shortcomings, researchers are striving to enhance their stability, versatility, and safety in preclinical and clinical applications.

3.2. Roles of EV-derived miRNAs in different cell types associated with the wound healing process

Recently, many studies have been conducted to uncover the roles of exosomal miRNAs in cutaneous skin wound healing (Table 1). These involved both in vitro assays on immune cells, endothelial cells, keratinocytes, and fibroblasts and in vivo assays using full-thickness skin wound mice or diabetic mouse models. Most reports showed that exosomal miRNAs have the ability to enhance cell migration, proliferation, tube formation, and anti-apoptosis. In addition, exosomal miRNAs can promote healing processes such as re-epithelialization, tissue remodeling, and wound closure, while also reducing scar width.

Table 1.

Therapeutic functions of exosomal miRNAs in skin wound healing.

Cell sources	miRNAs	Recipient cell types	Target genes of miRNAs	Effects	Reference
hUCB plasma	miR-21-3p	Fibroblasts	Inhibiting PTEN and SPRY1	Promoting the proliferation and migration of fibroblasts, accelerating re-epithelialization, reducing scar widths, and enhancing angiogenesis in vitro and in vivo	[150]
BMSC	miR-21-5p	Fibroblasts	Inhibiting SPRY2, activating PI3K/AKT and ERK1/2 signaling pathways	Promoting proliferation, migration, angiogenesis, and fibroblast function both in vitro and in vivo	[151]
Human AFSC	let-7-5p, miR-22-3p, miR-27a-3p, miR-21-5p, and miR-23a-3p	Fibroblasts	Inhibiting TGF- βR1 and TGF- βR2 of TGF-β signaling pathway	Reducing collagen fiber deposition and in situ myofibroblast formation, promoting the reconstruction of cutaneous cells in vitro	[152]
ADMSC	miR-181a	Fibroblasts	Targeting SIRT1	Inhibiting fibroblast-myofibroblast differentiation in vitro	[153]
Human amnion MSC	miR-135a	Fibroblasts	Downregulating LATS2	Increase cell migration and wound epidermalization, thus promoting wound healing in vitro and in vivo	[154]
Mouse ESC	mmu-miR-291a-3p	Fibroblasts	TGF-βR2 signaling pathway	Inhibiting cellular senescence; accelerating the excisional skin wound healing process in vitro and in aged mice model	[155]
ADMSC	miR-21-3p, miR-126-5p, miR-31-5p (up regulate), miR-99b, miR-146-a (downregulate)	Fibroblasts	Activating PI3K/AKT signaling pathway	Promoting proliferation and migration of fibroblasts, inhibiting inflammation, thus enhancing the secretion of vascular growth factors and extracellular matrix both in vitro and in vivo	[105]
ADMSC	has-miR-4484, has-miR-619-5p, and has-miR-6879-5p	Dermal fibroblasts	Inhibiting NPM1, PDCD4, CCL5, and NUP62 genes	Promoting wound healing by stimulating the proliferation and migration of dermal fibroblasts in vitro	[156]
Young fibroblast	miR-125b	Old fibroblasts	SIRT7	Reducing myofibroblasts and dysfunctional ECM deposition during aged wound closure and promoting proliferation, and migration of fibroblasts both in vitro and in vivo	[157]
ADMSC	miR-146a	Fibroblasts		Promoting the proliferation, and migration of fibroblast, upregulating SERPINH1, and p-ERK2 expressions, and enhancing neovascularization both in vitro and in rat back model	[158]
EPSC	miR-425-5p, miR-142-3p	Fibroblasts	Suppressing TGF-β expression	Inhibiting the differentiation of fibroblasts to myofibroblasts, thus inhibiting scar formation both in vitro and in vivo	[159]
UCMSC	miR-21-5p, miR-125b	Myofibroblasts	Inhibiting TGF- β signaling pathway	Stimulating wound healing by suppressing fibroblast – myofibroblast differentiation both in vitro and in vivo	[68]
ADMSC	miR-29a	Hypertrophic scar-derived fibroblasts	Targeting TGF-β2/SMAD3 signaling pathway	Inhibiting the fibrosis of HSFBs and scar hyperplasia after scalding in mice, promoting wound healing, and reducing pathological scar formation in vitro and in vivo	[119]
MSC	miR-138-5p	Hypertrophic car-derived fibroblasts	Targeting SIRT1 gene	Inhibiting proliferation, migration, and protein (NF-kB, a-SMA, TGF-β1) expressions, thus alleviating scar formation in vitro	[160]
ADMSC	miR-141-3p	Hypertrophic scar-derived fibroblasts	Targeting TGF-β2 to suppress the TGF-β 2/SMAD pathway	Decreasing the thickness of the scar, improving fibroblast distribution, and collagen fiber arrangement, and downregulating the levels of a-SMA, COL I, FN, TGF-β2, and p-SMAD2/3 in the scar both in vitro and in vivo	[161]
hADSC	miR-192-5p	Hypertrophic scar-derived fibroblasts	Targeting IL-17RA and then activating the SMAD pathway	Decreasing expression of COL I, COL III, α-SMA, IL-17RA, and p-SMAD2/p-SMAD3, increasing SIP1 in HSFs; inhibiting fibroblast proliferation and migration; promoting wound healing, attenuating collagen deposition and hypertrophic scar fibrosis in vitro and in mice model	[77]
Torvastatin-pretreated BMMSC	miR-221-3p	Endothelial cells	Upregulating AKT/eNOS signaling pathway	Promoting endothelial cell proliferation and migration, tube formation, and VEGF level; facilitating blood vessel formation and promoting wound regeneration in vitro and in vivo	[141]
Mouse EPC	miR-221-3p	Endothelial cells	Involving in the AGE-RAGE and the p53 signaling pathway	Increasing expression level of VEGF, CD31, and cell proliferation marker Ki67; promoting skin wound healing in vitro and in vivo	[162]
UCMSC	miR-135b-5p, miR-499a-3p	Endothelial cells	MEF2C signaling pathway	Enhancing proliferation and migration of HUVEC, promoting vessel formation and angiogenesis both in vitro and in vivo	[163]
ADMSC	miR-125a-3p	Endothelial cells	Inhibiting PTEN	Promoting angiogenesis both in vitro and in mice wound tissue	[143]
UCMSC	miR-17-92	Endothelial cells		Accelerating cell proliferation, migration, angiogenesis, and enhancing against erastin-induced ferroptosis both in vitro and in vivo	[148]
BMMSC	miR-542-3p	Fibroblasts and endothelial cells		Promoting cellular proliferation, collagen deposition, neovascularization, and accelerating wound closure both in vitro and in mice model	[149]
ADSC	miR-126-3p	Fibroblasts and endothelial cells	Increasing PI3KR2 expression	Increasing wound healing rate, collagen deposition, and newly formed vessels both in vitro and in rat models	[164]
ADMSC	miR-21	Keratinocytes	Changing TGF-β1 expression; improving MMP9 expression through PI3K/AKT pathway	Improving in vitro HaCaT migration and proliferation	[125]
ADMSCs	miR-19b	Keratinocytes	Targeting CCL1 to regulate the TGF-β pathway	Promoting proliferation and migration of keratinocytes, thus accelerating skin wound healing rate both in vitro and in vivo	[126]
ADSC	miR-21-5p	Keratinocytes	Activating WNT/β-catenin signaling pathway	Promoting proliferation and migration of keratinocytes in vitro and in vivo, increasing re-epithelialization, collagen remodeling, angiogenesis, and vessel maturation in vivo	[132]
BMMSC	miR-93-3p	Keratinocytes	Inactivating APAF1	Promoting proliferation and migration, and suppressing apoptosis of keratinocytes in vitro	[165]
UCMSC	miR-150-5p	Keratinocytes	Targeting PTEN and then Activating PI3K/AKT signaling pathway	Promoting growth and migration, while inhibiting apoptosis of keratinocytes in vitro	[131]
BMMSC	miR-223	Macrophages	Targeting PKNOX1	Accelerating wound healing by inducing macrophage polarization both in vitro and in mice model	[89]
EPC	miR-182-5p	Keratinocytes	Targeting PPARG	Promoting proliferation and migration while inhibiting apoptosis of HaCaTs both in vitro and in diabetic mice model	[133]
iPSCs-derived keratinocyte	miR-762	Keratinocytes, endothelial cells	PML (noncoding RNA)	Enhancing migration of keratinocyte and endothelial cells in vitro and in vivo	[166]
ADSC	miR-132	Endothelial cells and M2 macrophages	NF-κB signaling pathway	Accelerating diabetic wound healing and enhancing M2 macrophages polarization, angiogenesis, and increasing collagen remodeling, further reducing the inflammation in wound tissues both in vitro and in vivo	[147]
Mononuclear cell	miR-150-5p			Modulating skin cell proliferation, migration, survival, and angiogenesis; completing re-epithelization and tissue rremodeling in vitro	[131]
Peroxiredoxin II-stimulated dermal MSC	miR-221, miR-21-5p			Accelerating wound healing by protecting dermal MSCs from reactive oxygen species-induced apoptosis both in vitro and mouse model	[167]
Irradiated HaCaT	miR-27a		Targeting MMP2	Leading to slowed cell migration which could affect wound healing both in vitro and in vivo	[168]
Hypoxia ADMSC	miR-21-3p, miR-126-5p, miR-31-5p, miR-99b, miR-146-a		PI3K/AKT signaling pathway	Promoting diabetic wound healing and inhibiting inflammation both in vitro and in vivo	[105]
hFibroblast	miR-130a, miR210	Fibroblast, endothelial cells		Enhancing proliferation, migration in fibroblast and promote tube formation both in vitro and in vivo	[169]
ADMSC	miR-486-5p	Human microvascular endothelial cells, human fibroblast		Improved the proliferation and motility of fibroblast and microvascular endothelial cells, along with enhancing their capacity to support angiogenesis in vivo model.	[170]

UCB: umbilical cord blood; BMSC: bone mesenchymal stem cell; AFSC: amniotic fluid stem cell; MSC: mesenchymal stem cell; ESC: embryonic stem cell; ADMSC: adipose-mesenchymal derived stem cell; FB: fibroblast; ADS: adipose-derived stem cell; EPSC: epidermal stem cell; UCMSC: umbilical cord-derived mesenchymal stem cell; BMMSC: bone marrow-derived mesenchymal stem cell; EPC: endothelial progenitor cell; HaCaT: human epidermal keratinocyte; miR: microRNA.

3.2.1. Immune cells and anti-inflammation effects at wound sites

The inflammatory phase plays a crucial role in achieving hemostasis and activating the immune system, which safeguards us from invading pathogens and aids in eliminating damaged tissues [88]. Thus, regulating immune cells and cytokines related to the inflammatory process during wound healing is vital to facilitate the skin healing process. Exosomal miRNAs have been demonstrated to modulate inflammation-related disorders in the wound healing context due to their immunomodulatory, pro-regenerative, and restoring homeostasis capacities [89,90]. For example, miR-233 and let-7b carried by EXs derived from BMMSC and lipopolysaccharide-primed umbilical cord-derived mesenchymal stem cells (UCMSCs) increased the expression of M2 macrophages via targeting PBX/Knotted 1 Homeobox 1 or inhibiting toll-like receptor 4 (TLR4) and promoting the NF-κB/STAT3/AKT signaling pathway, respectively [89,90]. Suppressing TLR4 signaling has also been achieved by miR-181c presented in UCMSC-derived EXs, which resulted in the reduction of burn-related inflammation [91]. Additionally, miR-146a-5p and miR-548e-5p in EXs secreted by amniotic fluid-derived MSC showed their anti-inflammatory effects through regulating phosphorylation of NF-κB, AKT, and mitogen-activated protein kinase (MAPK) proteins [92]. These EXs are also revealed to modulate the expression of cytokines, such as increasing IL-10 and reducing tumor necrosis factor α (TNF-α) levels [89]. Other miRNAs, including miR-155 and miR-146a, have been shown to control the differentiation and formation of regulatory T cells (T_regs) [93], which are responsible for balancing immune homeostasis and mediating cutaneous wound healing [94]. However, there is currently a lack of research on the impact of these miRNAs from EXs on the differentiation and generation of T_regs.

3.2.2. Exosomal miRNAs modulate fibroblasts’ growth factor production, proliferation, migration, and myofibroblast differentiation

In injured tissues, fibroblasts promote wound healing by producing regulatory molecules and cross-talk with other cell types. As a result, fibroblasts release: (1) metalloproteinases that degrade and lose the nearby extracellular matrix to facilitate the infiltration of inflammatory cells into the wound areas [95], or (2) extracellular matrix components (fibronectin, hyaluronic acid, proteoglycans, and collagen) to support keratinocyte migration for re-epithelialization process [95,96], or (3) pro–angiogenesis molecules (VEGF, FGF, Ang-1, TSP) which are necessary for endothelial cells to engage in angiogenesis [97–101]. Therefore, utilizing miRNAs to target pathways associated with fibroblast activities offers an alternative approach to enhance wound healing efficacy. Indeed, numerous miRNAs have been employed to manipulate fibroblast proliferation, migration, differentiation, and apoptosis, thus promoting wound healing and preventing abnormal scar formation. Interestingly, the majority of these miRNAs targeted two critical signaling pathways that govern fibroblast behaviors, which are phosphatidylinositol 3-kinase (PI3K)/protein kinase B (AKT) and transforming growth factor β (TGF-β)/SMAD pathways.

The PI3K/AKT signaling pathway regulates cell proliferation, migration, differentiation, angiogenesis, and metabolism to maintain skin homeostasis [102]. Dysregulation of the PI3K/AKT pathway has been demonstrated to be associated with several human cutaneous disorders, whereas activating this pathway may improve wound repair and tissue regeneration [103]. In fact, the immunohistochemistry results of Gao et al. indicated that during the healing process, the expression of PI3K, AKT, p-PI3K, and p-AKT reached its peak in fibroblasts and was significantly correlated to wound time [104]. Similarly, Wang et al. reported hypoxia adipose mesenchymal-derived stem cells (ADMSC)- EXs with upregulated levels of miR-21-3p, miR-126-5p, and miR-31-5p, and downregulated levels of miR-99b and miR-146a activated the PI3K/AKT signaling pathway in fibroblasts [105] (Figure 2). As a result, these exosomal miRNAs promoted the healing of diabetic wounds by increasing fibroblast proliferation, migration, vascular growth factors, and extracellular matrix secretion [105]. Additionally, a recent study using in vitro and in vivo models has demonstrated that let-7 promoted cell proliferation [106], and let-7 in EVs identified from saliva and keratinocytes can enhance the wound healing process [3,39]. A study by Jiang et al. found that let-7c directly targeted HSP70, modulating Bax and Bcl-2 expression in the PI3K/AKT pathway, thereby inhibiting fibroblast migration and proliferation during wound healing [107].

Figure 2.

Therapeutic role of miRNA in accelerating wound healing by targeting the PI3K/AKT signaling pathway. Dysregulation of the PI3K/AKT pathway is commonly found at wound sites, with proteins in this pathway being overexpressed (blue) or downregulated (red). Utilizing these miRNAs to silence abnormally upregulated proteins (blue) can indirectly activate wound healing-stimulated proteins (red), thereby triggering the PI3K signaling pathway. This pathway includes numerous Sub-pathways, such as mTOR, eNOS, RAS/RAF, and NF-κB. The mTOR niche is shown to be primarily responsible for supporting keratinocyte migration, proliferation, and invasion, whereas eNOS contributes to enhanced endothelial cell vascularization. Furthermore, the RAS/RAF Sub-pathway promotes cell survival and proliferation, whereas NF-kB induces apoptosis. As a result, miRNA-mediated activation of the PI3K/AKT pathway and its Sub-pathways plays an important role in accelerating wound healing and paves the way for potential therapeutic applications in regenerative medicine and cutaneous wound care. The black line indicates the activation, while the red line indicates the inhibition.

The transition from fibroblasts into myofibroblasts is crucial in the tissue remodeling phase, which is responsible for wound contraction, vascularization, and cellularity decline [100,108,109]. Skin accomplishes this activity by activating the TGF-β/SMAD signaling pathway. When TGF-β activates its downstream effectors SMAD proteins, inflammatory cells are recruited, angiogenesis is stimulated, fibroblasts differentiate into myofibroblasts, and extracellular matrix (ECM) proteins are synthesized [110,111]. To avoid scar formation, myofibroblasts have to undergo apoptosis, and the ECM is removed upon wound closure and tissue regeneration [112,113]. In contrast, scar tissue will develop when fibroblasts continue to proliferate and differentiate into myofibroblasts even after the wound has been successfully repaired [114]. This abnormal process results in an excessive build-up of connective tissue, and one of the primary factors contributing to this phenomenon is the aberrant activation of the TGF-β/SMAD pathway [115–117]. Therefore, downregulating this pathway by miRNAs in fibroblasts or myofibroblasts could attenuate abnormal scar formation [68]. In this regard, Fang et al. have found that specific exosomal miRNAs derived from MSCs, including miR-21, miR-23a, miR-125b, and miR-145, play a key role in suppressing myofibroblast formation by inhibiting excess α-SMA and collagen deposition associated with the activity of the TGF-β/SMAD2 signaling pathway [68]. Another unfavorable outcome that needs to be considered is hypertrophic scars, which are related to the excessive deposition of collagen and changes in the structure of other extracellular proteins [118]. Yuan et al. have found that miR-29a, delivered by ADMSC-EXs, targeted TGF-β2 of the TGF-β2/SMAD3 signaling pathway, leading to an inhibition of HSFB fibrosis and scar hyperplasia after scalding in mice [119]. Similarly, a study by Li et al. demonstrated that ADMSC-EXs miR-192-5p reduced pro-fibrotic protein levels by targeting IL-17RA, thus facilitating wound healing and attenuating collagen deposition [77] (Figure 3). In summary, these data suggest that miRNAs carried by EXs hold promising potential to promote wound healing while suppressing scar formation.

Figure 3.

Modulation of the TGF-β signaling pathway by miRNAs to prevent aberrant scar formation and promote wound closure. Overexpression of the TGF-β/SMAD signaling pathway in fibroblasts causes excessive α-SMA and collagen synthesis and deposition, and abnormal collagen fiber cross-linking, leading to hypertrophic scarring and fibrosis. To attenuate skin scars, miRNA-based therapeutic strategies targeting this dysregulated pathway in fibroblasts are proposed. These miRNAs prevent myofibroblast over-differentiation and excessive ECM deposition, which are key factors of scarring. However, while TGF-β is suggested to promote keratinocyte migration that supports epithelialization, its effects on keratinocyte proliferation are still conflicting. Given the dual role of TGF-β in the re-epithelialization phase, the application of these miRNAs as therapeutic agents at wound sites should be considered carefully to avoid converse effects. The black line indicates the activation, while the red line indicates the inhibition. The blue color indicates proteins being overexpressed, whereas the red color indicates proteins being downregulated at the wound site.

3.2.3. Exosomal miRNAs promote keratinocyte proliferation and migration and inhibit apoptosis for wound healing

Keratinocyte migration plays a crucial role in the healing process by rapidly covering dermal and mucosal wound surfaces to re-establish an epidermal barrier, also called re-epithelialization [120,121]. Re-epithelialization is the result of three overlapping keratinocyte functions: migration, proliferation, and differentiation. Keratinocytes secrete TGF-β1 and TGF-β2 into the wound fluid via paracrine and autocrine to regulate wound healing [122]. Numerous studies have demonstrated the role of TGF-β in promoting keratinocyte migration that supports epithelialization [123,124]; however, the effects of TGF-β on keratinocyte proliferation are still conflicting. In fact, this contradiction is observed in two independent studies utilizing ADMSC exosomal miR-21 and miR-19b to promote skin wound healing. Over-expressed miR-21 inhibited the TGF-β1 expression, thus balancing MMP-2 and TIMP-1 protein expression, which ensures successful re-epithelialization [125]. On the other hand, miR-19 in ADMSC-derived EXs with significant expression levels targeted chemokine (C-C motif) ligand (CCL) to increase TGF-β level, hence improving therapeutic effects on skin damage caused by H₂O₂-treated human epidermal keratinocyte (HaCaTs) [126] (Figure 3). These data suggest that silencing TGF-β by miRNAs can either promote or suppress wound healing associated with keratinocyte proliferation; thus, using miRNAs as therapeutic agents for skin wound healing that target the TGF-β pathway should be carefully considered.

In normal tissue, downstream targets of mammalian target of rapamycin (mTOR) in the PI3K/AKT pathway, particularly phosphorylated forms of the S6 ribosomal protein (pS6) and serine 473 AKT (pAKT473) accumulated in the granular layer [127–129]. However, during the wound healing process, the presence of pAKT473 and pS6 is expanded to the spinous layer of the transitional epithelium and then to all layers in the migrating epithelial tongue, which was always absent from the normal epithelium [130]. This evidence indicates that the PI3K/AKT/mTOR pathway is activated in the healing process, and this function is attributed to the activation of the mTOR complexes. Squarize et al. have shown that inhibiting mTOR by rapamycin delayed wound healing, while activating mTOR by targeting its upstream regulator, phosphatase and tensin homolog (PTEN) accelerated re-epithelialization by increasing migratory and proliferation behavior of keratinocytes [130]. Similarly, the genetic knockdown of PTEN by miRNAs led to promoting growth and migration while inhibiting apoptosis of keratinocytes, as evidenced by the study employing UCMSC-EXs with a high level of miR-150-5p to target PTEN in keratinocytes [131] (Figure 2). Therefore, miRNAs targeting PTEN that activate the PI3K/AKT/mTOR pathway in keratinocytes represent a novel intervention strategy to accelerate the skin healing rate.

Therapeutic approaches targeting the epidermal keratinocytes may open the doors for optimal diabetic wound care. With regard to this, Lv et al. employed ADMSC exosomal miR-21-5p to target the Wnt/β-catenin signaling pathway [132], and Li et al. utilized endothelial progenitor cell (EPC)-derived EXs mediated miR-182-5p to downregulate PPARG in keratinocytes [133]. Both of these efforts successfully promoted the proliferation and migration of keratinocytes in vitro, thus accelerating diabetic wound healing. Collectively, these findings unveil a new role of exosomal miRNAs in the treatment of diabetic skin wounds, but there is a lack of investigations on the healing context through keratinocyte functions modulated by exosomal miRNAs.

3.2.4. Exosomal miRNAs promote endothelial cell proliferation, migration, and angiogenesis

Angiogenesis, referred to as the formation of new blood vessels, is a critical step in wound healing. During the healing process, endothelial cells proliferate and migrate toward the wound sites, following the direction of the tip cell, and then differentiate to form a continuous vessel [134,135]. Reportedly, MSCs-derived EXs carrying different miRNAs could promote endothelial cell proliferation, migration, and angiogenesis by targeting different signaling pathways; for example, the vascular endothelial growth factor (VEGF)/AKT/eNOS pathway. VEGF is a potent angiogenic factor that promotes angiogenesis and contributes significantly to normal wound repair [136]. VEGF binds to the VEGF receptor (VEGFR) on the endothelial cell membrane to phosphorylate AKT, further activating endothelial nitric oxide synthase (eNOS) [137]. eNOS is an enzyme that synthesizes NO and regulates NO expression by mediating the activity of upstream signals in this pathway. NO is required to maintain vascular integrity and blood flow, such as causing vasodilation, controlling vascular tone, and inhibiting platelet aggregation and adhesion to the vascular wall [138,139]. Additionally, NO can promote angiogenesis by enhancing the survival, proliferation, and migration of endothelial cells [140]. Thus, regulating NO production in endothelial cells by upregulating the VEGF/AKT/eNOS signaling pathway is crucial for the vascularization of wound healing. Yu et al. discovered that the AKT/eNOS signaling pathway in diabetic rats was over-expressed when injecting EXs derived from atorvastatin-pretreated MSCs, and blocking this pathway attenuated the pro-angiogenic effects of atorvastatin-EXs [141]. They also highlighted that overexpression of miR-221-3p elevated the AKT/eNOS pathway activity, which promotes endothelial cell proliferation and migration, facilitating blood vessel formation and wound regeneration for diabetic mouse models [141]. Additionally, PTEN, another important component of the PI3K/AKT pathway, can inhibit angiogenesis by downregulating VEGF expression. Meanwhile, PTEN knockdown significantly increases PI3K and p-AKT expression levels, suggesting that PTEN is a key negative regulator of the wound healing process [142]. Overexpression of miR-125a-3p in human ADMSCs–EXs inhibited the expression of PTEN in human umbilical vein endothelial cells (HUVEC) [143]. As a result, PTEN silencing promoted the survival, migration, and angiogenesis of HUVECs, finally promoting wound healing and angiogenesis in mice wound granulation tissues [143] (Figure 2).

Moreover, nuclear factor kappa B (NF-kB) signaling plays a role in maintaining endothelial cell integrity and vascular homeostasis in vivo [144,145]. NF-kB activation in endothelial cells is considered a way to block angiogenesis, yet the specific mechanisms behind this inhibition are still elusive [144]. NF-kB could impair angiogenesis via (1) a decrease in ECM degradation capacity to inhibit endothelial cell migration [144] or (2) inducing vascular endothelial growth inhibitors or angiostatic agents [146]. Therefore, suppressing the NF-κB signaling pathway, specifically in endothelial cells, might be an attractive therapy for wound repair. Besides enhancing M2 macrophage polarization, inhibiting NF-kB by miR-132 presented in ADMSC-derived EXs also resulted in enhanced proliferation and migration of HUVECs, reduced local inflammation, increased angiogenesis, thus promoting diabetic wound healing both in vitro and in vivo [147].

Nevertheless, several studies have reported that miRNAs originating from MSC-EXs, including miR-17-92 and miR-542-3p, increase endothelial cell proliferation, migration, and angiogenesis through unknown mechanisms [148,149]. Thus, these results offer promising potential for further investigation of injured skin treatment by repairing and enhancing blood vessels for both acute and chronic wounds.

3.3. Potential diagnostics and prognostics of EV miRNAs in cutaneous wound healing

As previously noted, specific miRNAs present at a wound site can either promote or hinder the healing process. MiR-124 and miR-378a exhibited elevated expression levels in fibroblasts at wound sites, and sponging them by lncRNA MALAT1 enriched in EXs accelerated the proliferation and migration while inhibiting apoptosis of fibroblasts [171,172]. Additionally, vascular cells, endothelial cells, immune cells, and platelets release EVs containing miRNAs, contributing to the miRNA pool in blood plasma [173–175]. These circulating miRNAs are transported to wound sites, where they influence the healing process. For example, under inflammatory stress, vascular endothelial cells released miR-191 into the bloodstream, which was subsequently taken up by distant dermal endothelial cells or dermal fibroblasts of inflamed skin wounds [175]. The miR-191 acts as a negative regulator, inhibiting the angiogenesis and migration capacity of diabetic dermal fibroblasts or endothelial cells, leading to defective wound healing in diabetic patients [175]. Thus, examining the miRNA composition in wound exudates could reveal clues to the healing status. Interestingly, numerous exosomal miRNAs have been found with considerable concentrations in chronic wounds (Table 2), enabling them to be potential candidates for diagnostic and prognostic markers of the wound healing process. Even though these miRNAs serve as diagnostic and prognostic markers for wound status, further research is needed to uncover their specific biological effects and potential redundancies in the complex wound healing process [176].

Table 2.

EV-derived miRNAs in skin wound healing diagnosis.

Cell sources	Inhibitors	Recipient cell types	miRNAs	Effects	Reference
ADSC	Lnc H19	Fibroblasts	miR-19b	Upregulating SOX9, thus accelerating human fibroblast proliferation, migration, and invasion for wound healing in vitro and in vivo	[177]
ADSC	Lnc MALAT1	Fibroblasts	miR-124	Inducing wound healing by activating the WNT/β-catenin pathway, promoting cell proliferation, migration, while inhibiting apoptosis in vitro	[172]
ADSC	mmu-circ_0000250	Endothelial cells	miR-128-3p	Promoting wound healing in diabetes by upregulating SIRT1 which increased angiogenesis and suppressed apoptosis of endothelial cells both in vitro and in vivo	[178]
ADSC	Lnc MALAT1	Fibroblasts	miR-378a	Promoting proliferation and migration, increasing re-epithelialization, collagen remodeling, angiogenesis, and vessel maturation by downregulating FGF2 expression	[171]
ADSC	mmu_circ_0001052	Endothelial cells	miR-106a-5p	Inhibiting apoptosis while promoting proliferation, migration, angiogenesis, and expression of FGF4, VEGF, and p38	[179]
ADSC	NEAT1	Keratinocytes	miR-17-5p	Promoting wound healing by upregulating pathways related to wound healing, such as autophagy pathway via NEAT1/miR-17-5p/ULK1 axis	[180]
ADSC	LncRNA FOXD2-AS1	Keratinocytes	miR-185-5p	Promoting HaCaT proliferation, migration via downregulating miR-185-5p and upregulating ROCK2	[181]
MSC	miR-155 inhibitor	Keratinocytes	miR-155	Increasing FGF-7 levels, keratinocyte migration, anti-inflammatory action, and angiogenesis, leading to accelerated wound healing in diabetic wounds	[182]
UCMSC	CircHIPK3	Endothelial cells	miR-20b-5p	Promoting cell proliferation, migration, and angiogenesis via upregulating NRF2 and VEGFA under diabetic condition	[183]
BMMSC	Lnc KLF3-AS1	Endothelial cells	miR-383	Increasing blood vessel formation, reducing inflammation by upregulating VEGFA	[184]

ADSC: adipose-derived stem cell; MSC: mesenchymal stem cell; UCMSC: umbilical cord-derived mesenchymal stem cell; BMMSC: bone marrow-derived mesenchymal stem cell.

4. Challenges of using extracellular vesicle-derived miRNAs for monitoring wound healing and wound repair

Previous evidence suggests that miRNAs have potentials to modulate gene expression in skin cell populations, thereby facilitating or preventing normal wound healing. In this context, miRNA-based therapies, utilizing both miRNA mimics and miRNA inhibitors (also known as anti-miRNAs), have emerged as innovative strategies to restore the normal expression level and function of wound-related genes. miRNA mimics are synthetic double-stranded small RNA molecules that are similar to the original miRNA sequences and, hence, functionally aim to restore the loss of these miRNAs in diseased cells [185]. Anti-miRNAs, on the other hand, are single-stranded molecules that are designed to bind and inhibit the functions of target miRNAs [185]. Notwithstanding, it is impossible to directly introduce miRNA mimics or anti-miRNAs into the human body due to their susceptibility to being degraded by nucleases in serum or in the cell’s endocytic compartment, rapid clearance from blood, high toxicity, and low tissue permeability [185,186]. However, EVs possess good biocompatibility and biodegradability; thus, they could address these issues by protecting miRNAs from degradation, prolonging blood circulation time, and crossing the physiological and blood-brain barrier. Additionally, EVs can not only load miRNA drugs but can also enrich more therapeutic miRNAs by the endogenous cargo sorting pathway. In this regard, EV-loaded miRNAs could avoid their limits in clinical application for the treatment of various diseases and make them an ideal candidate for new therapeutic targets [187].

Nevertheless, the lack of an appropriate standardized loading strategy is a challenge in putting EX innovation technology into clinics. Recent research has discovered that transfection reagents might contaminate EVs during transfection, and traditional centrifugation methods cannot effectively separate transfection complexes from EVs [188,189]. As a result, this contamination potentially affected EV’s delivery capacity and cargo transfer to recipient cells [188,189]. However, four innovative and effective strategies for loading miRNAs into EXs, including (1) sonication, (2) hydrogen (pH) gradient loading, (3) exosome-liposome fusion loading, and (4) commercial kit EXO-Fects, have been utilized to improve not only loading efficiency but also minimize immunogenicity and increase stability and half-life of the EXs [190–193]. Remarkably, the current generation of the Exo-Fect kit provides a rapid, easy, and highly efficient workflow for EX loading, achieving 95% delivery efficiency to target cells without specialized equipment, optimization, and non-toxicity [194]. Apart from modifying EXs, miRNA modification by cholesterol to form asymmetric miRNA oligonucleotides with a hydrophobic portion could increase miRNA stability, facilitate cell membrane binding, promote cellular internalization, and induce effective gene silencing [195,196]. In a study by Hade et al., they coupled miRNA with peptide YARA, and the results suggested that this miRNA modification not only significantly increased the loading effect of EXs compared to naked miRNAs but also enhanced proliferation, migration, and invasion of fibroblasts, thus providing an innovative strategy for miRNA delivery in wound healing [197]. It is noteworthy that chemical modification could prevent the recognition and loading of miRNA mimics into AGO and RISC complexes [198].

In addition, one miRNA molecule could target multiple genes, a phenomenon known as ‘too many targets for miRNA effect’ (TMTME) [199]. This typical and inevitable property of miRNAs is due to the incomplete complementarity between miRNAs and target sequences. Unlike FDA-approved drugs, including siRNA drugs, which usually have only a few targets, miRNAs can affect numerous pathways [199]. Thus, introducing or removing miRNAs in human cells can change various unknown and unpredictable pathways and biological processes, possibly leading to physiological dysfunctions or new diseases [199]. This TMTME property of miRNAs might also contribute to serious adverse events that have hampered the success of miRNA-based therapy in clinical trials. While surface engineering of EVs can improve miRNA’s solubility and efficacy, minimize unwanted interactions with healthy tissues, and reduce toxicity and side effects [199,200], these techniques pose challenges for certain cell types that are hard to transfect, such as stem cells and red blood cells [201]. Hence, non-genetic strategies like lipid and protein modifications should be explored further to overcome the limitations of genetic modification and enhance compatibility. Other drawbacks of these drug delivery systems include their high cost, limited productivity in delivering nucleotides, and challenges in adjusting dosages effectively [199]. Furthermore, packing exogenous artificial miRNAs into EVs will promote competition with endogenous miRNAs for the intracellular machinery, thus triggering unexpected gene expression and leading to unwanted side effects [202–204].

In addition, the EV source is also of paramount importance in determining their therapeutic potential in wound healing. For instance, EVs secreted by ADMSCs promote wound healing in diabetic ulcer via improving angiogenesis, and re-epithelialization; meanwhile, EVs from BMMSCs demonstrated lower efficacy under the same condition [205]. Besides, although the same source of parental cells, cargoes of EVs have also been influenced by different types of culture medium [206]. More importantly, EVs from HUVECs, enriched with miR-106b, actually delayed wound closure by inhibiting JMJD3 and RIPK3, whichleads to a reduction in angiogenesis and a decrease in viability of fibroblasts and keratinocytes [207]. Therefore, therapeutic strategies involving EVs should carefully consider the biological cargoes and intended outcome, as certain EVs may hinder rather than promote tissue regeneration.

Last but not least, recently, there has been no ‘gold standard’ for EV isolation, and the method used to isolate EVs from body fluids is critical for the characterization of their cargoes since the different isolation methods cause differences in EV concentration, purity, and size. In fact, RNA sequencing analysis showed that mRNA profile differed in EVs isolated by ultracentrifugation, ExoQuick, or Total Exosome Isolation Reagent [208]. Although 588 common miRNAs were detected in all three EV populations, there are approximately 200 miRNAs that are unique to each EV sample isolated by these methods [208]. Similar results were reported by Ding et al. and Rekker et al., stating that the exosomal miRNA profiles depend on the isolation methods used [209,210]. Additionally, the isolation methods would also be carefully considered when using exosomal miRNAs as biomarkers. The choice of an appropriate and easy-to-handle protocol is crucial, especially in clinical routines, since it might influence the results of high-throughput differential expression analysis.

5. Future direction

Although a considerable number of miRNAs have been demonstrated to have therapeutic effects on skin wounds, most of these studies lack a temporal view to investigate the functional role of miRNAs during the healing process. As the regulation of miRNAs on their target gene expression alters dynamically to adapt to wound repair’s requirement, a miRNA could be defined as a ‘pro-healing’ when its continuous expression is beneficial for one healing phase but could be considered an ‘anti-healing’ element for other healing phases [211]. Therefore, future studies should focus on deciphering the temporal expression patterns of miRNAs with strong potential for skin wound healing to identify precise treatment phases.

As miRNAs are unstable and susceptible to degradation, miRNA-based therapies need advanced transport systems to deliver therapeutic miRNAs to target wound areas precisely, safely, and efficiently [212]. Nanocarriers and delivery strategies are two crucial elements for managing the spatiotemporal release of therapeutic miRNAs during the dynamic wound healing process, which changes in both time and space [212,213]. Therefore, future research needs to concentrate on developing alternative transfer methods to avoid triggering unwanted immune responses and then identifying the right transporter for the right miRNA to target the right tissue in the wound area, a concept also known as precision medicine. Additionally, current methods of miRNA quantification are mostly relative rather than absolute, thus limiting the safety and effectiveness of their clinical translation [214]. More accurate quantification and detection methods of miRNAs are required in the future to overcome this drawback.

Recently, EVs encapsulated in a smart hydrogel that can sense and respond to parameters in the wound environment, such as O₂, pH, ROS level, and glucose concentration, are considered promising techniques for delivering EVs to the wound site and ensuring prolonged and sustained EV release [215,216]. Meanwhile, EVs are ideally suited as biological nanocarriers for miRNAs due to their high biocompatibility, low immunogenicity, and ease of modification. Thus, future direction should improve the properties of hydrogel in order to better control the release and response to specific stimuli of this nanovesicle in transporting therapeutic miRNA cargoes to the wound environment [217,218]. Moreover, multi-functional biomaterials are expected to play a significant role in future wound treatment by providing a balanced environment (e.g. alleviating wound hypoxia, promoting angiogenesis, reducing oxidative stress, and enhancing wound closure rates) [214,215,218]. Recently, a study by Oh et al. discovered that EVs derived from Streptococcus mutants, an oral commensal bacterium, have therapeutic potential as a carrier for tRNA to promote chronic wound healing through a TLR3-dependent mechanism [219]. However, there is still limited data on their capability to carry miRNAs.

In summary, EXs’ advantages render them appealing as promising tools for miRNA delivery in wound repair or miRNA-based monitoring of wound conditions. However, several obstacles still remain, including the lack of standardized loading and isolation techniques, complexity in miRNA sorting mechanisms, and low exosome purity. These hurdles hinder the progress of current research on exosomal miRNAs in wound healing; hence, further research on their therapeutic and diagnostic potential should be explored before application. These attempts to advance EV-derived miRNA-based strategies would not only enhance the speed and quality of chronic wound healing without affecting other normal physiological processes but also address the obstacles of translating miRNA-based therapies into clinical trials, providing excellent targeting solutions for skin wound healing.

6. Conclusion

Research has highlighted the regulatory role of EV-derived miRNAs in wound healing, with a particular emphasis on exosomal miRNAs. The selective enrichment of specific miRNAs in EVs is linked to the physiological and pathological states of the secreting or parental cells, underscoring their potential for disease diagnosis and prognosis. While EVs can influence the extracellular matrix (ECM) and facilitate cellular processes, certain miRNAs are preferentially packaged into EVs, affecting specific biological events in distinct cell types during wound healing. However, the current body of research remains insufficient to establish EV miRNAs as definitive biomarkers for wound healing diagnosis and prognosis. Therefore, further studies are required to assess the diagnostic and prognostic potential of EV miRNAs, particularly those isolated from body fluids such as wound exudates. Despite existing challenges, continued research may pave the way for EVs to revolutionize wound treatment and diagnosis.

Glossary

Abbreviations

MiRNA

microRNA

EV

extracellular vesicle

BMMSC

bone marrow mesenchymal stem cell

EX

exosome

AB

apoptotic body

MV

microvesicle

MVB

multivesicular bodies

ESCRT

endosomal sorting complex required for transport

TSG

tumor suppressor gene

HSP

heat shock protein

CD

cluster of differentiation

RISC

RNA-induced silencing complex

UTR

untranslated region

GLT

glutamate transporter 1

SHIP

SH2-containing phosphoinositide 5-phosphatase

TGF-β

transforming growth factor beta

TIMP

targeting tissue inhibitor of metalloproteinase

TIAM

T-lymphoma invasion and metastasis-inducing protein

EMP

epithelial membrane protein

CXCL

C-X-C motif chemokine ligand

IL

Interleukin

NF-kB

nuclear factor kappa B

STAT

signal transducer and activator of transcription

ERK

extracellular signal-regulated kinase

HB-EGF

heparin-binding EGF-like growth factor

GATA2

GATA binding protein 2

VEGFR

vascular endothelial growth factor receptor

FGF

fibroblast growth factor

PDGF

platelet-derived growth factor

COL

collagen

PTEN

phosphatase and tensin homolog

nSMase

neutral sphingomyelinase

hnRNP

heterogeneous nuclear ribonucleoproteins

miRISC

miRNA-induced silencing complex

AGO

argonaute

SNARE

soluble N-ethylmaleimide attachment protein receptor

UCMSC

umbilical cord-derived mesenchymal stem cell

TLR

toll-like receptor

MAPK

mitogen-activated protein kinase

TNF

tumor necrosis factor

PI3K

phosphatidylinositol 3-kinase

AKT

protein kinase B

TGF

transforming growth factor

ADMSC

adipose mesenchymal-derived stem cell

ECM

extracellular matrix

CCL

chemokine (C-C motif) ligand

HaCaT

human epidermal keratinocyte

mTOR

mammalian target of rapamycin

pS6

ribosomal protein

PTEN

phosphatase and tensin homolog

EPC

endothelial progenitor cell

VEGF

vascular endothelial growth factor

VEGFR

vascular endothelial growth factor receptor

eNOS

endothelial nitric oxide synthase

HUVEC

human umbilical vein endothelial cell

NF-kB

nuclear factor kappa B

UCB

umbilical cord blood

BMSC

bone mesenchymal stem cell

AFSC

amniotic fluid stem cell

MSC

mesenchymal stem cell

ESC

embryonic stem cell

FB

fibroblast

ADSC

adipose-derived stem cell

EPSC

epidermal stem cell

EPC

endothelial progenitor cell

miR

microRNA

lncRNA

long noncoding RNA

TMTME

too many targets for miRNA effect

Funding Statement

The author(s) reported there is no funding associated with the work featured in this article.

Authors contributions

Mai Thi Le: Investigation, Software, Visualization, Writing – original draft; Huy Hoang Dao: Investigation, Writing – original draft; Xuan-Hung Nguyen: Conceptualization, Writing – original draft, Writing – review & editing; Tinh Van Nguyen: Conceptualization, Visualization, Writing – review & editing; Tu Dac Nguyen: Conceptualization, Visualization, Writing – review & editing; Uyen Thi Trang Than: Conceptualization, Supervision, Writing – original draft, Writing – review & editing.

Disclosure statement

No potential conflict of interest was reported by the author(s).